university of groningen interactions of cell division protein ftsz with large and small molecules...

University of Groningen

Interactions of cell division protein FtsZ with large and small moleculesCendrowicz, Ewa

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2016

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Cendrowicz, E. (2016). Interactions of cell division protein FtsZ with large and small molecules. Universityof Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 02-02-2021

https://www.rug.nl/research/portal/en/publications/interactions-of-cell-division-protein-ftsz-with-large-and-small-molecules(6172ec0b-be62-4019-a802-9849e4f314d5).html

https://www.rug.nl/research/portal/en/persons/ewa-cendrowicz(36d2acb2-8ff4-49a6-a236-88e0d129057d).html

https://www.rug.nl/research/portal/en/publications/interactions-of-cell-division-protein-ftsz-with-large-and-small-molecules(6172ec0b-be62-4019-a802-9849e4f314d5).html

INTERACTIONS OF CELL DIVI- SION PROTEIN FTSZ

WITH LARGE AND SMALL MOLECULES

EWA CENDROWICZ

The work described in this book was carried out in the Department of Molecular Microbiology of the Gron-ingen Biomolecular Sciences and Biotechnology Institute (GBB), University of Groningen, the Netherlands, and was financially supported by the Dutch Science Organization (NWO).

Cover design: Milena Cendrowicz and Lovebird design.

Layout and printing: Lovebird design. www.lovebird-design.com

ISBN (print): 978-90-367-8910-3ISBN (digital): 978-90-367-8909-7

Copyright © 2016 by E. Cendrowicz. All rights reserved. No parts of this book may be reproduced or trans-mitted in any form or by any means without prior permission of the author.

Interactions of cell division protein FtsZ with large and

small molecules

PhD thesis

to obtain the degree of PhD at the University of Groningen on the authority of the

Rector Magnificus Prof. E. Sterken and in accordance with

the decision by the College of Deans.

This thesis will be defended in public on

Friday 17 June 2016 at 11.00 hours

by

Ewa Cendrowicz born on 8 May 1986

in Lodz, Poland

SupervisorsProf. D.J. Scheffers Prof. A.J.M. Driessen Assessment CommitteeProf. D.J. Slotboom Prof. J.W. Veening Prof. P. Levin

TABLE OF CONTENTS

CHAPTER 1 Introduction 7

Scope of this thesis 30

CHAPTER 2 FtsZ polymerization assays: simple protocols and considerations

45

CHAPTER 3 Bacillus subtilis SepF binds to the C-terminus of FtsZ 71

CHAPTER 4 The effects of manganese on the interaction of sporulation protein SpoIIE with itself and cell division protein FtsZ

95

CHAPTER 5 Antibacterial activity of alkyl gallates is a combina-tion of direct targeting of FtsZ and permeabilization of bacterial membranes

121

CHAPTER 6 Summary 155

APPENDICES Samenvatting 171

About the author 176

List of publications 177

Acknowledgments 179

Introduction

CH

APT

ER

1. GENERAL INTRODUCTION TO CELL DIVISION CHAPTER 1

9

1. GENERAL INTRODUCTION TO CELL DIVISION

Cell division is a fundamental process in the life of all bacteria. In rod-shaped bacteria it is often simplified to a three steps event: cell elongation, septum for-mation and division into two equal daughter cells 1. In fact, to complete each of these steps, bacteria have to actuate complicated biomolecular machineries, which orchestrate different series of complex events. At the heart of these pro-cesses lies a 40 kDa cytoplasmic GTPase called FtsZ (Filamentous Temperature Sensitive protein Z) 1,2. At the onset of cell division, FtsZ localizes to the mid-cell and polymerizes into a ring-like structure, the so-called Z-ring. FtsZ is tethered to the membrane via other membrane proteins and establishes the future divi-sion site. During the process of division, the Z-ring constricts and disassembles to move to the new division site in the daughter cell (see Fig. 1A, B) 1.

FtsZ is almost universally conserved in Bacteria. Homologues of FtsZ have also been found in a group of Archaea, in chloroplasts, and in mitochondria of some eukaryotes. FtsZ is an essential protein for cell division in most bacteria. It has a dual function during cell division, as a scaffold for the recruitment of other cell division components, known as divisome, and as a force generator for membrane invagination 1-3. FtsZ is also tightly regulated by other proteins, which make sure that the Z-ring assembles in the correct place (usually in the middle of the cell) and time (after chromosome segregation is complete) and which control the dynamics of the Z-ring (see below and Table 1) 1-5.

Understanding the role and mechanism of Z-ring assembly is not just an interesting topic. It is also crucial for answering further questions about other processes during cell division, including remodeling of the cell envelope. Re-cently, the study of FtsZ also gained special attention in antibacterial research, since FtsZ and other essential cell division proteins may be promising targets for the development of new antibiotics 6-12. Here, I describe the nature of FtsZ and its behaviors in vitro, and in the cell, as well as advances in the research of antibacterial compounds which target FtsZ. I focus mainly on the model rod-shaped bacteria, Bacillus subtilis and Escherichia coli. Any data or conclusions described here are derived from work on these organisms unless otherwise noted.

E.CENDROWICZ: INTERACTIONS OF CELL DIVISION PROTEIN FTSZ WITH LARGE AND SMALL MOLECULES

10

Figure 1. Schematic representation of different modes of cell division. Representation of cell di-

vision during vegetative growth in E. coli (A) and B. subtilis (B), and during sporulation in B. subtilis

(C). Vegetative cell division starts with cell elongation and chromosome replication and segre-

gation, followed by assembly of the Z-ring in the middle of the cell. After that, other cell division

proteins are recruited to the Z-ring to form divisome. The divisome drives cell separation by cell

wall ingrowth in B. subtilis (B) or cell envelope invagination in E. coli (A). The process is finalized

by the formation of two equal daughter cells. (C) Sporulation begins with the formation of an axial

DNA filament. After that FtsZ is relocated to the poles of the cell. One of the Z-ring matures and

drives asymmetric septum formation. The process is completed when the forespore is engulfed

in the mother cell and the mother cell has lysed to release the matured spore to the environment.

2. FTSZ CHAPTER 1

11

2. FTSZ

2.1 CRYSTAL STRUCTURE

The first crystal structure of the FtsZ molecule was obtained of FtsZ from the archaeon Methanocaldococcus jannaschii 13. The three-dimensional structure of FtsZ appeared to be similar to the structure of eukaryotic α- and β-tubulin, which were presented at the same time 14, although the sequence homology of these proteins is weak. Based on the crystal structure, presented in Fig. 2, two domains can be distinguished: the N-terminal domain – a nucleotide-binding domain (blue) and the C-terminal globular core (cyan). Both domains are sep-arated by a central α-helix. The GTP/GDP nucleotide binding site is built up of four T-loops (phosphate-binding loops), a sugar recognition sequence (placed in the N-terminal domain), and a guanine recognition sequence (placed in the α-helix that connects both domains) 13.

GTP BINDING

The FtsZ monomer alone can bind GTP/GDP (Fig. 2, orange), however, for GTP hydrolysis at least two monomers are necessary. Activation of GTP hydrolysis occurs when the T7 loop (Fig. 2, red) from one monomer (placed on the C-ter-minal domain, following the central α-helix) enters the GTP binding site of an-other monomer 2,4,13. Nucleotide binding does not cause any conformational change within the FtsZ monomer 15.

2.2 THE EXTREME C-TERMINAL PART

The extreme C-terminal part of FtsZ was not resolved in the crystal structure due to its high flexibility 13. However, this part of FtsZ plays a very important role in FtsZ assembly, its interaction with many cell division proteins, and is crucial for cell division 2,16-22. The extreme C-terminus may be divided into 3 domains as proposed by Buske and Levin: an unstructured C-terminal linker region of variable length, a highly conserved C-terminal tail (CTT) and a short variable region at the extreme C-terminus (CTV) (Fig. 2) 16. Although some of these domains consist of only a few residues (in E. coli, the CTT is formed


12

of 9 residues and the CTV of only 4), distribution of the extreme C-terminal part into 3 independent domains is important as each of the domains has its own function in cell division/FtsZ assembly and interaction with other proteins 2,16-18.

C-TERMINAL LINKER

The structure of the C-terminal linker cannot be resolved in crystal structures (Fig. 2, purple) and its sequence conservation is weak, and so this part of FtsZ has received relatively little attention. The length of the linker may range from 2-330 residues 23, but most FtsZs have linkers with a length of 50 to 100 resi-dues (for E. coli and B. subtilis the linker sequence is about 50 residues). Recently the function of the C-terminal linker was determined 17,18,24. It was shown that the C-terminal linker plays a role in the formation of FtsZ protofilaments and the architecture of the FtsZ ring in vivo. Apparently, the C-terminal linker is cru-cial for FtsZ assembly into a ring structure and cell division. It was proposed that the C-terminal linker is intrinsically disordered 17,18. Flexibility and disorder of the linker is critical for proper FtsZ functioning as replacing the linker with helical repeats leads to the filamentation of B. subtilis cells and a significantly decreased GTP hydrolysis rate by chimeric FtsZ 17. The length of the linker is im-portant for lateral interactions of protofilaments and cell division 17. Introduc-ing extra amino acids into the flexible linker increases the distance between the globular N-terminal domain and the extreme C-terminus, that binds to the membrane anchors of FtsZ, like FtsA or ZipA 25,26. Szwedziak et al. found that in-creasing the length of the C-terminal linker changes the distance of FtsZ proto-filaments from the cell membrane from 16 nm to a variable distance between 16-21 nm 25. Recently, it was shown that the C-terminal linker is required for the coordination of peptidoglycan synthesis in Caulobacter crescentus 24.

C-TERMINAL CONSERVED TAIL

Following the C-terminal linker, there is a highly conserved set of residues called the C-terminal conserved region (CTT) 16. It is defined as 9-amino acid motif, with a completely conserved proline at position 6 and highly conserved residues at positions 4, 5, 8 and 9 23. This part of E. coli FtsZ has been crystalized in a complex with cell division protein ZipA revealing an helical structure of

2. FTSZ CHAPTER 1

13

this region (Fig. 2) 27. However, the structure of the CTT alone, or in complex with other proteins, was never determined.

C-TERMINAL VARIABLE REGION

At the very end of the C-terminus, a small region called the C-terminal variable region (CTV) is present. It may contain 1 to 13 poorly conserved residues (4 in E. coli and 6 in B.subtilis) 16,23. Buske et al. showed that the charge of the CTV is crucial for the formation of lateral interactions between FtsZ protofilaments 16. It was shown that a positively charged CTV is involved in promoting lateral interactions (B. subtilis FtsZ) while a negatively charged (E. coli FtsZ) or neutral CTV does not promote lateral interactions between single FtsZ filaments 16.

The extreme C-terminus of FtsZ (CTT and CTV) serves as a link between FtsZ and the other cell division components since most of the FtsZ interacting proteins bind to this region of FtsZ (FtsA, ZipA, SepF, EzrA, ClpX(P), SlmA, etc.) (see below) 19,20,22,27,28.

Figure 2. Cartoon format of the 3D structure of FtsZ molecule from Pseudomonas aeruginosa.

The N-terminal domain is marked in blue, the C-terminal domain in cyan. A bound GDP molecule

is marked in orange and the T7 synergy loop in red. Binding regions for interacting proteins are

indicated with arrows. The structure of FtsZ was reprinted with permission from 2.


14

3. ASSEMBLY OF THE Z-RING AND REGULATORY SYS-TEMS

3.1 SPATIAL REGULATION

During vegetative growth, Z-ring formation occurs precisely in the middle of a bacterial cell (Fig. 1A, B). How FtsZ is directed with such a high precision to the mid-cell has been studied for decades in two rod shaped model organisms, E. coli and B. subtilis 29. For a long time it has been thought that the spatial reg-ulation of Z-ring positioning is a combination of two negative molecular sys-tems, the Min system, which prevents formation of the Z-ring at the cell poles, and nucleoid occlusion (NO), which prevents constriction of the Z-ring over nucleoids 4,29. Positive regulatory systems were found only in Actinomycetes (SggAB system) and Myxococcus xantus (PomZ) 29. Recently, an additional sys-tem, Ter linkage, was shown to positively regulate Z-ring placement in E. coli 30.

MIN SYSTEM

The Min system prevents Z-ring formation at the cell poles by directly inhib-iting the polymerization of FtsZ via the MinC protein. MinC interacts with a membrane associated ATPase MinD. The third component of the Min sys-tem is a topological factor that localizes the MinCD complex to the poles. In B. subtilis, the DivIVA protein senses negative curvature of the polar membrane and localizes MinD to the cell poles via a bridging protein called MinJ. DivIVA does not exist in E. coli and instead, the MinE protein plays the role of topolog-ical determinant. MinE oscillates from pole-to-pole and displaces MinD from the membrane. MinD assembles at the membrane distant from MinE, creating a dynamic oscillating system. The result is a concentration gradient of MinCD with a minimum around the mid-cell and maximum at the cell-poles. There are many reviews that comprehensively describe the Min systems from E. coli and B. subtilis 29,31-34. Here, the direct inhibition of FtsZ by MinC is shortly described.

MinC. MinC consists of two domains of approximately equal size 35,36. Both domains are necessary for the inhibition of Z-ring formation 36-38. The N-ter-minal domain of MinC (MinCN) blocks FtsZ assembly into a Z-ring in vivo and blocks polymerization in vitro (Fig. 3) without influencing the GTPase activity of

3. ASSEMBLY OF THE Z-RING AND REGULATORY SYSTEMS CHAPTER 1

15

FtsZ (Table 1) 36. The C-terminal domain of MinC (MinCC) affects FtsZ assembly only in the presence of MinD 38. The MinCC/ MinD complex binds to the extreme C-terminus of FtsZ, whereas the binding site for MinCN is placed on the C-ter-minal globular domain of FtsZ (Fig. 2) 39,40. Binding of the MinCC/ MinD complex blocks lateral interactions resulting in less bundled single protofilaments 37,38 (Fig. 3). MinC interacts preferentially with GDP-bound FtsZ and shortens FtsZ filaments possibly by partially destabilizing them 41.

NUCLEOID OCCLUSION

The second regulatory system prevents formation of the cell division septum over the nucleoid and, in consequence, guillotining the chromosome. Nucle-oid occlusion is mainly driven by the DNA binding proteins Noc (B. subtilis) and SlmA (E. coli). Despite their similar roles, Noc and SlmA belong to different groups of DNA binding proteins and inhibit cell division by different mecha-nisms 42-45.

E. coli SlmA. SlmA specifically binds DNA in the regions distributed all over the chromosome except for the replication terminus region. DNA binding acti-vates a SlmA dimer to inhibit FtsZ 45. The inhibition of FtsZ occurs in two steps. In the first step, DNA-bound SlmA binds to the C-terminus of FtsZ, compet-ing with its membrane anchors and several other proteins. In the second step, SlmA breaks/disassembles FtsZ protofilaments without affecting its GTPase ac-tivity 44,46. In this way, FtsZ formation is inhibited everywhere over the chromo-some, except for the replication terminus region, where cell division can occur. Althoug a lot of work has been done on SlmA in E. coli, the exact mechanism of protofilament breakage as well as in vivo mechanism of SlmA action are still not fully clear.

B. subtilis Noc also controls Z-ring formation in regard to chromosome seg-regation. However, the mechanism of control is different from SlmA. There is no evidence of direct interaction between FtsZ and Noc. Recently, Adams et al. proposed a model, in which they explain how Noc inhibits cell division 47. Noc is a peripheral membrane protein that is able to recruit DNA to the cell membrane. As a result, large nucleoprotein complexes are formed over the nu-cleoid, which cause physical crowding. The physical crowding is thought to act as short-range inhibitor of cell division 4,42,47.


16

THE TER LINKAGE

Neither Min nor NO systems are essential in E. coli and B. subtilis. Moreover, in the absence of both systems, the Z-ring is still positioned with high precision at mid-cell between segregated nucleoids in slow growing E. coli cells 30. This sug-gested the presence of another regulatory mechanism. Recently, Espeli et al. found that the MatP protein serves as a linkage between the Z-ring and the chromosome in E. coli 48. MatP binds DNA in the replication terminus region and is connected to the Z-ring via ZapB and ZapA 48. Interestingly, MatP and SlmA binding sites on the chromosome are complementary, making them positive and negative sites for regulation of Z-ring formation 48.

The Ter linkage is also not essential in E. coli. Moreover, deletion of slmA, minCDE and Ter linkage genes (matP, zapB and zapA) does not affect viability of E. coli in slow growth conditions 43. Taken together, these facts indicate the presence of additional mechanism or that there are no indispensable mecha-nisms that place the divisome in the correct position in cells 43.

3.2 OTHER REGULATORY SYSTEMS

CELL SIZE CONTROL, NUTRIENT AVAILABILITY

Bacterial cell size is coupled to nutrient availability and, in consequence, growth rate. Cells grown on a nutrient-rich medium are approximately 2x as voluminous compared to cells grown on a nutrient-poor medium, and cell divi-sion is inhibited until the cell reaches its appropriate size. This means that cells contain a mechanism which transmits the information about their metabolic status and growth rate to the division machinery to delay cell division until the cell reaches the correct size. Part of the mechanism is Uridine-5’-phosphoglu-cose (UDP-glucose), a small molecule that is produced in nutrient-rich condi-tions and that signals the cells metabolic status. In B. subtilis and E. coli, Ugtp and OpgH respectively bind UDP-glucose and are recruited to the nascent sep-tal site to inhibit Z-ring formation in fast-growing cells 49-51.

UgtP is a membrane-associated glucosyltransferase, a part of the con-served glucolipid biosynthesis pathway. In nutrient-rich conditions, UgtP lo-calizes to the division site and inhibits cell division by directly interacting with FtsZ until the cell reaches its appropriate size. In nutrient-poor conditions UgtP

4. THE Z-RING IN VIVO CHAPTER 1

17

is sequestered away from mid-cell in randomly distributed foci. UDP-glucose influences the FtsZ-UgtP interaction by reducing the affinity of UgtP for itself and making it more available for interaction with FtsZ. However, the interac-tion is not dependent on UDP-glucose. In vitro, UgtP inhibits FtsZ polymeriza-tion but does not have significant influence on its GTPase activity 49,51.

OpgH is the functional homologue of UgtP in E. coli. It is an inner-mem-brane protein and interacts directly with FtsZ via its N-terminal domain in a UDP-glucose dependent manner. Similar to UgtP, it localizes to mid-cell in fast-growing cells and is sequestered away from the Z-ring in slow-growing conditions. Unlike UgtP, OpgH localization is not dependent on UDP-glucose. In vitro, it inhibits assembly and GTPase activity of FtsZ 50.

Another link between nutrient availability and cell division in B. subtilis was recently discovered by Monahan et al. It was shown that the E1α subunit of py-ruvate dehydrogenase influences cell division in a pyruvate-dependent man-ner. Whether the influence is direct or indirect is still an open question 52.

SOS RESPONSE

In response to DNA damage, cells activate an SOS mechanism, the goal of which is to repair the DNA damage and to inhibit cell division until the repair is complete. An SOS component that directly targets cell division is SulA. SulA is a small protein induced in response to DNA damage and removed by Lon protease when DNA is repaired 53. It inhibits cell division by directly interacting with FtsZ 54,55, at the site bound by another FtsZ monomer during polymer-ization, thus increasing the critical concentration for FtsZ assembly 54. In the presence of SulA, the GTPase activity of FtsZ is about 50% reduced 55.

4. THE Z-RING IN VIVO

4.1 DYNAMICS OF THE Z-RING

Pioneering studies on FtsZ in vivo, using immunofluorescence and Green Flu-orescent Protein (GFP) fused to FtsZ, revealed localization of FtsZ to mid-cell during most of the cell cycle 76-78. Only 30% of the total FtsZ available in the


18

Tab

le 1

. Su

mm

ary

of

FtsZ

inte

ract

ing

pro

tein

s p

rese

nt

in B

. su

bti

lis a

nd

th

eir

mo

de

of

acti

on

on

Fts

Z.

Prot

ein

Size

[k

Da]

Func

tion

in v

ivo

Knoc

k-ou

t phe

noty

peM

ode

of a

ctio

n on

Fts

ZBi

ndin

g si

te

on F

tsZ

Stru

ctur

e (P

DB

entr

y)G

TPas

e ac

tivity

Poly

mer

izat

ion

Posi

tive

regu

lato

rs

FtsA

56-5

848

• Te

ther

ing

FtsZ

to th

e m

embr

ane

• El

onga

tion

in B

. sub

tilis

, •

Elon

gatio

n an

d ce

ll de

ath

in E

. col

iN

o•

Bund

ling

CTT

1E4G

(T. m

ariti

ma)

ZapA

59-6

110

• Z-

ring

stab

iliza

tion,

•

A ro

le in

sep

tum

fo

rmat

ion

in E

. col

i

• N

o ph

enot

ype

• Le

thal

with

∆ez

rA•

Stro

ng in

hibi

tion

at

10 m

M M

gCl 2

• Sl

ight

inhi

bitio

n at

5

mM

MgC

l 2

• Cr

oss-

linki

ng p

olym

ers,

• Bu

ndlin

g at

10

mM

MgC

l 2

• N

o eff

ect a

t 5 m

M M

gCl 2

Not

kno

wn

1T3U

SepF

19,6

2,63

17•

Teth

erin

g Ft

sZ to

the

mem

bran

e,•

Form

atio

n of

div

isio

n se

pta

• D

efor

med

div

isio

n se

pta,

• le

thal

with

∆ft

sA o

r ∆ez

rAN

o•

Larg

e tu

bula

r str

uctu

res

CTT

C-te

rmin

al p

art

(aa

61-1

40):

3ZIH

SpoI

IE 64

,65

92•

Relo

caliz

atio

n of

Fts

Z to

the

cell

pole

,•

teth

erin

g to

the

mem

bran

e

• D

elay

in th

e fo

rmat

ion

of a

sym

-m

etric

sep

tum

No

• Bu

ndlin

gN

ot k

now

nPh

osph

atas

e do

mai

n: 3

T91

Neg

ativ

e re

gula

tors

EzrA

22,6

6-68

65•

Inhi

bitio

n of

the

form

atio

n of

ext

ra

Z-rin

gs in

cel

l

• Ex

tra

Z-rin

gs fo

rmed

alo

ng c

ell

• In

crea

se

• D

isas

sem

bly

of p

refo

rmed

po

lym

ers

• In

hibi

tion

of p

olym

eriz

atio

n

CTT

Cyto

plas

mic

do

mai

n: 4

UY3

Min

C 20

,35,

4125

• Sp

atia

l inh

ibiti

on o

f Z-

ring

form

atio

n•

Min

icel

lsN

o•

Dep

olym

eriz

atio

n of

pro

to-

filam

ents

,•

Dec

reas

e in

bun

dlin

g

K243

, D28

7 –

Hel

ix 1

0,

R376

- C

TT

1HFZ

Ugt

P 49

,51

44•

Inhi

bitio

n of

Z-r

ing

acco

rdin

g to

nut

rient

av

aila

bilit

y

• Sh

orte

r cel

ls in

nut

rient

-ric

h m

ediu

mN

o•

Not

det

erm

ined

Not

kno

wn

-

ClpX

21,4

9,51

,70

46•

Mai

ntai

ning

dyn

amic

s an

d di

sass

embl

y of

th

e Z-

ring

• St

rong

fila

men

tatio

n in

∆m

inC

stra

in

No

• D

isas

sem

bly

of F

tsZ

poly

-m

ers

• In

hibi

tion

of p

olym

eriz

atio

n

CTT

and

C-

term

inal

lin

ker

3HTE

Mci

Z 72

-74

4•

Inhi

bitio

n of

Fts

Z as

-se

mbl

y in

mot

her c

ell

durin

g sp

orul

atio

n

• Po

lar Z

-rin

g fo

rmat

ion

afte

r en

gulfm

ent o

f a s

pore

by

mot

her

cell

• In

crea

se a

t low

and

de

crea

se a

t hig

h co

n-ce

ntra

tions

of M

ciZ

• Sh

orte

ning

of F

tsZ

poly

-m

ers

C-te

rmin

al

glob

ular

co

re, A

sp28

0

2MRW

4. THE Z-RING IN VIVO CHAPTER 1

19

cell is incorporated into the Z-ring 2,79. The rest of FtsZ is present in the cyto-plasm as a pool of monomers and short polymers 80 which are continuously exchanging with the Z-ring subunits with half-time between 8-9 sec 79. Not only FtsZ monomers are exchanged but also the overall architecture of the Z-ring constantly changes with time and the Z-ring remains dynamic during constriction 81. It is not fully clear whether the Z-ring disassembles during constriction or only after constriction is complete. Strauss et al. have noted that the total intensity of FtsZ-GFP within the ring remains constant during constriction, suggesting that the total amount of FtsZ within the ring does not change 81. However, the data was limited only to large rings (800-900 nm in diameter) 81 and previous studies have noted disassembly during con-striction 82,83. After constriction, FtsZ immediately reassembles at mid-cell in daughter cells. What drives the dynamics and disassembly of the Z-ring? It was suggested that the FtsZ interacting proteins ClpX(P) and B. subtilis EzrA may be involved in this process 21,69.

ClpX(P) is a part of the ClpXP protease complex, in which it serves as a sub-strate recognition domain. ClpX may also function in a ClpP-independent manner. The mode of action of ClpX on FtsZ in E. coli and B. subtilis seems to be complete-ly different. In B. subtilis, ClpX blocks FtsZ polymerization independently of ATP and ClpP, whereas in E. coli both ClpP and ATP are necessary for FtsZ inhibition. However, both modes of action suggest a role of ClpX(P) in the modulation of the Z-ring disassembly with a possible role in maintaining dynamics and subunit turnover between the Z-ring and cytoplasmic FtsZ (Fig. 3) 21,69. ClpX(P) interacts with extreme C-terminal part of both monomeric and polymeric forms of FtsZ. Therefore, it has dual role in inhibition of FtsZ polymerization; one is degradation of FtsZ monomers and prevention of FtsZ polymerization and the second one is breakage of the previously assembled polymers (Table 1) 21.

EzrA (Extra Z-rings A) is a membrane protein present in Gram-positive or-ganisms and one of the first proteins to localize to the Z-ring in B. subtilis. In ezrA null mutant cells, extra Z-rings are formed at cell poles and the medial Z-ring becomes more stable (Table 1). Thus, EzrA is considered an inhibitor of FtsZ in the cell (Fig. 3) 84. In vitro, EzrA inhibits the assembly and increases the GTPase activity of FtsZ. Similarly to ClpX(P), EzrA can also break previously assembled polymers by interacting with the C-terminus of FtsZ (Fig. 2 and Table 1) 22.


20

Recent studies suggest a second role of EzrA in coordination of divisome as-sembly with lateral cell wall synthesis 85,86.

4.2 ARCHITECTURE OF THE Z-RING

FtsZ assembly has been extensively studied in vitro. However, assembly of the Z-ring and its architecture in vivo is still not clear. Together with the develop-ment of super-resolution techniques (like PALM – photoactivated localization microscopy, 3D-SIM – three-dimensional-structured illumination microscopy, cryotomography, the understanding of FtsZ-ring formation in vivo and in vitro is increasing. It has been known that FtsZ polymers are tethered to the cyto-plasmic membrane by membrane (binding) proteins like FtsA or ZipA (E.coli), and FtsA, SepF, and EzrA (B. subtilis) 26,58,63,84. The Z-ring filaments are placed at a distance of ~15-16 nm from the inner membrane (observed for E. coli and C. crescentus) 25,87. Up to now, it was thought that FtsZ forms a continuous fil-ament at mid-cell. Recently, research based on super-resolution microscopy revealed that FtsZ localizes into patches instead of a continuous ring 80,81,88,89. These patches are most likely composed of short overlapping FtsZ polymers 25,87. It was proposed that gaps existing between FtsZ beads may be necessary for FtsZ ring constriction 3,80. FtsA and ZipA in E.coli 80 as well as EzrA and PBP2 in S. aureus 81 form similar patches. The FtsA and ZipA patches overlap with FtsZ and with each other in E. coli 80.

In contrast to previous studies, the recent cryotomography work revealed that the structure of the Z-ring is continuous and encircles the whole division site 25. Both techniques, however, show that the Z-ring is formed of shorter overlapping filaments 25,80,81,87,88. Cryotomography work revealed that the Z-ring is formed of single layered bands that are 5-10 filaments wide 25. It is possible that both models are correct and that either structure may exist during differ-ent stages of cell division 88.

5. ASSEMBLY OF THE Z-RING DURING SPORULATION

A completely different type of cell division is observed in B. subtilis during spor-ulation (Fig. 1C). Sporulation is an adaptive process undertaken by Bacillus and

5. ASSEMBLY OF THE Z-RING DURING SPORULATION CHAPTER 1

21

its relatives under starvation conditions 31,90. It begins with the switch from medial to polar Z-ring formation through a spiral-like intermediate, a process which is indirectly controlled by the protein Spo0A 77,90. In spo0A null mutants, the Z-ring is not directed to the polar sites and cell division is completed at mid-cell. The positional switch is mediated by the sporulation specific protein SpoIIE, which is expressed under Spo0A control 77,91. Asymmetric septation leads to the formation of two unequal-sized daughter cells, a larger mother cell and a smaller forespore. The forespore is then engulfed by a mother cell in a process resembling phagocytosis. Subsequently, the mother cell lyses and the spore is released to the environment, where it can survive indefinitely in a state of dormancy (Fig. 1C) 31,92. During sporulation, chromosomes do not segregate in a manner observed for vegetative growth. Instead, oriC regions migrate to the opposite poles of the cell and chromosomes form an elongat-ed structure known as the axial filament. The axial filament is bisected by the asymmetric septum and one-third of the chromosome ends up in the fore-spore (Fig. 1C). The remaining two-third is later transferred to the forespore by a conjugation-like mechanism directed by SpoIIIE 93,94. It was suggested that the asymmetric septation uses the same cell division machinery as the vege-tative one, except for an additional component, SpoIIE, which localizes to the asymmetric division site in a ring-like structure called the E-ring. However, the asymmetric septum is a much thinner structure than the vegetative septum and most of the peptidoglycan formed to separate the two lipid bilayers is re-moved soon after septation completes 31. Also, the recently discovered sep-tum-forming protein SepF was not yet studied during sporulation, indicating that more differences between vegetative cell division and division during sporulation exist.

SpoIIE is a 92 kDa membrane protein, which consists of three domains: the N-terminal membrane domain with 10 membrane-spanning segments, the central domain involved in SpoIIE oligomerization and interaction with FtsZ and a PP2C-type phosphatase domain at the C-terminus 65. SpoIIE is expressed at the onset of sporulation and has two functions. One is redeployment of FtsZ from a medial to two polar Z-rings, one of which eventually constricts. The sec-ond role is activation of the transcription factor σF in the forespore 64,91,95. Ini-tially, FtsZ assembles into a ring-like structure at mid-cell. SpoIIE co-localizes


22

with FtsZ via a direct interaction. Instead of constricting at mid-cell, FtsZ and SpoIIE redeploy near both cell poles in the form of E-rings. One of the E-rings dissolves and the other matures into the sporulation septum (Fig 1C) 91.

MciZ (Mother cell inhibitor of Z). MciZ is a 40-aa peptide expressed during sporulation to block Z-ring formation in the mother cell. It binds to the C-ter-minal globular core of FtsZ (Fig. 2) and functions as a filament capping protein. It was shown that MciZ shortens FtsZ polymers without competing with GTP for binding to FtsZ (Fig. 3). Interestingly, low MciZ concentrations promote the GTPase activity of FtsZ, while high concentrations of MciZ inhibit FtsZ GTPase (Table 1) 73.

6. FOLLOW-UP PROCESSES

Once FtsZ is present at mid-cell, it becomes a scaffold for the recruitment of other “divisome” components (Fig. 1A, B) 1. In B. subtilis, the assembly of cell division proteins occurs in two steps. First, the proteins FtsZ, FtsA, ZapA and EzrA (early cell division proteins) are recruited to mid-cell and after that, a sec-ond set of proteins arrives to the division site, including GpsB, FtsL, DivIB, FtsW, Pbp2B and DivIVA (late cell division proteins) 96. In E. coli the assembly was first thought to be more hierarchical: [FtsZ, FtsA/ZipA] > [FtsK > FtsQ > FtsL/B > FtsW > FtsI] > FtsN. However, Goehring et al. found that proteins FtsK, Q, L, B, W and I may assemble together independently on FtsA and be recruited to-gether, as a complex, to the established Z-ring (the complexes are marked in square brackets) 97.

6.1 EARLY CELL DIVISION PROTEINS

MEMBRANE TETHERS

Several proteins function as membrane tethers for FtsZ, FtsA and ZipA in E. coli, and FtsA, EzrA and, as recently discovered, SepF in B. sutbilis, reviewed in 1. Mu-tation of ftsA in E. coli prevents cell division. The Z-rings are still formed via ZipA but are not able to complete cell division. In the absence of both proteins, cells are unable to form the Z-ring 58.

6. FOLLOW-UP PROCESSES CHAPTER 1

23

FtsA is an ATPase that is structurally related to actin. ATP binding is crucial for the FtsZ-FtsA interaction. FtsA binds to the membrane via an amphipathic helix at its C-terminus 26. Recently, it was shown that FtsA recruits FtsZ filaments, but not monomers to a lipid bilayer in vitro, and that FtsA and FtsZ together form highly dynamic spirals on the lipid bilayer 28. In contrast, the transmem-brane protein ZipA is able to recruit FtsZ monomers to the membrane and to bundle FtsZ filaments. Thus, the interaction between ZipA and FtsZ is stronger than between FtsA and FtsZ 28. FtsA and ZipA recruit downstream division pro-teins to the divisome 58.

Even though FtsA is essential in E. coli, it is not in B. subtilis. What is more, an ftsA and ezrA double mutant is viable in B. subtilis. However, either ftsA and sepF or ezrA and sepF double knockouts are lethal. These findings, and the fact that overexpression of SepF may restore deletion of ftsA, strongly suggest that SepF complements the function of FtsA in B. subtilis. SepF mutant cells are viable but have deformed division septa 63.

SepF is a 17 kDa protein which assembles into large (>2 MDa) rings at phys-iological conditions in vitro 63. The protein is composed of two domains. The N-terminal domain consists of a highly conserved region which forms a mem-brane binding amphipathic helix 62. The C-terminal domain is involved in SepF ring formation and interactions with FtsZ. The crystal structure of the C-ter-minal part of the SepF monomer reveals two α-helices and a five-stranded β-sheet arranged into a compact α/β-sandwich. SepF is organized into a tight dimer with the interface formed by β-sheets and the α-helices placed at the outside of the dimer. The SepF rings are formed by interactions between con-served glycine residues (G109) in the external helices. The C-terminal domain of SepF is also involved in the interactions with the C-terminus of FtsZ (Fig. 2) 19. Several SepF residues involved in the interaction with FtsZ were identified in a yeast two-hybrid screen, out of which half (V64, F126, I118) are placed on the internal β-sheet and half (D105, F106, G116) are placed on the external α-he-lices. Together, SepF and FtsZ self-organize into long, ~50 nm-wide tubules reminiscent to eukaryotic microtubules 63. Several facts indicate that SepF is an unique membrane anchor for FtsZ with a clear structural role. First, the or-ganization of SepF into rings as well as the polymerization of FtsZ (in the pres-ence of Mg2+ and GTP) are required for tubule formation in vitro. Second, the


24

organization of SepF into a ring structure is also important for normal SepF function in vivo. Mutation of a conserved glycine to asparagine (G135N) at the C-terminus of SepF does not abolish FtsZ binding but this mutant is defective in ring formation in vitro and cannot support cell division in a ∆ftsA mutant. In addition, Duman et al. found an intriguing correlation between the size of SepF rings (~40 nm diameter) and the width of septa which is in the range of 43 nm and proposed a model in which arcs of SepF polymers would fit on top of the leading edge of developing septum 62,63.

Interestingly, all membrane tethers for FtsZ bind to the same region, the extreme C-terminal part of FtsZ (Fig. 2) 19,58. It is not possible that all proteins bind to one region at the same time, it is likely that Z-ring formation is regulat-ed from the membrane via this part of the protein by activators and inhibitors of FtsZ.

Z-RING ASSOCIATED PROTEINS (ZAP)

ZapA is a non-essential, highly conserved, protein recruited early to the divi-sion site via direct interaction with FtsZ. A knock-out of zapA does not exhibit a phenotype unless it is combined with an ezrA mutant which is lethal 98. ZapA forms tetramers in vitro and bundles FtsZ protofilaments into thick branched higher order structures 99. It is thought that the role of ZapA is to stabilize the Z-ring in vivo (Fig. 3) 99.

Recently, three other Zap proteins were discovered in E. coli, two of which directly interact with FtsZ (ZapC and D) 100,101 and one (ZapB) that is associated

Figure 3. Schematic representation of the process of FtsZ assembly into a ring. FtsZ monomers

(black spheres) assemble into short polymers, which are tethered to the membrane via membrane

proteins (red oval), which is and followed by lateral association of the filaments to form a mature

Z-ring. Positive (+) and negative (-) regulators are indicated at each stage of FtsZ assembly.

6. FOLLOW-UP PROCESSES CHAPTER 1

25

to the Z-ring via ZapA 102,103. ZapB forms long cables in vitro, which are bundled by ZapA. Together with FtsZ both proteins form an interactome with highly ordered long cables and bundles 103. ZapC and D bind to FtsZ independently of ZapA and their role is stabilization of FtsZ protofilaments 100,101. Why E. coli needs so many different proteins with similar functions is not fully resolved. However, it is known that FtsZ lateral interactions play a critical role for Z-ring stability in vivo 83. It was shown that the CTV is important for lateral interac-tions of FtsZ protofilaments in vitro. Lateral interactions between B. subtilis FtsZ protofilaments are stronger compared those of E. coli 16. Durand-Heredia et al. suggested that Zap proteins compensate for weak lateral interactions of E. coli FtsZ in vivo 100.

6.2 LATE CELL DIVISION PROTEINS

Z-ring formation at mid-cell promotes the recruitment of other components of the cell division machinery. This includes late cell division proteins which to-gether with FtsZ and its interacting partners form a complex called divisome 104. The divisome contains over 30 different proteins out of which almost half are essential in E. coli 105-107. FtsK is thought to be recruited first and FtsN last of the downstream essential cell division proteins 104,105,108. However, recent findings reveal interactions between FtsN and early cell division proteins and the hi-erarchy of assembly seems to be more complex than previously thought 109. After FtsN is recruited to the septal ring, the constriction of the cell membrane and remodeling of the cell wall begins 105,110-112. This stage is likely to be closely regulated so that the development of potential error is minimal and can be corrected quickly 104,105,112.

FtsK is a bifunctional protein involved in cell division and chromosome segregation 107. It is thought that membrane/periplasmic domain of FtsK is involved in stabilizing late cell division proteins and the recruitment of the FtsQLB complex to the division site 113,114. The cytoplasmic domain forms hex-amers involved in DNA transport 115 associated with chromosome segregation during division.

FtsQLB. Highly conserved among bacteria, the FtsQLB subcomplex (in B. subtilis DivIB, DivIC, FtsL) is formed independently of its localization to


26

mid-cell. FtsL is involved in many protein-protein interactions within the di-visome 116. Therefore, for a long time it was thought that the role of FtsQLB complex is scaffolding the divisome components together. However, recent findings by Tsang and Bernhardt suggest that the complex plays an important role in activating the divisome to begin constriction 117.

FtsW belongs to the SEDS (Shape Elongation Division Sporulation) family of membrane proteins 107,118. Mohammadi et al. identified FtsW as a transporter (flippase) of cell wall building blocks across the cytoplasmic membrane 119.

FtsN was thought to be recruited to the division site as the last component of the divisome 97. However, recent findings indicate that small amounts of FtsN are recruited to the division site earlier via interaction with a FtsA mono-mer 120. It has been suggested that FtsN allosterically activates constriction via two interactions, one with FtsA in the cytoplasm and another with the FtsQLB complex in the periplasm 105,109,111.

In contrast to cell membrane constriction and cell wall ingrowth in E. coli, in B. subtilis septal cross-wall synthesis is completed before daughter cell sep-aration. Gram-positive bacteria contain most of the essential late cell division proteins except for FtsN 107. It was shown that the FtsZ-interacting protein EzrA, together with GpsB, plays a role in the switch from lateral to septal cell wall syn-thesis by recruiting the major peptidoglycan synthase PBP1 85. The misshaped septa formed in a sepF mutant 63 suggest that SepF also plays a role in septum closure. However, the exact mechanism of septum synthesis and closure is still to be discovered 62,63.

The next stages of cell separation involve peptidoglycan synthesis and sep-tum cleavage to complete cell separation. This stage was extensively described in a review by Egan and Vollmer 107 and are beyond the scope of this thesis.

7. FTSZ AS A FORCE FOR MEMBRANE CONSTRICTION

Even though proper cell envelope constriction and cell separation requires a number of accessory proteins, Erickson has proposed that FtsZ may generate the constriction force for the membrane by itself 2,121. The “Z-centric hypoth-esis” was supported by works of Osawa et al., Hsin et al. and Szwedziak et al.,

8. FTSZ ASSEMBLY IN VITRO CHAPTER 1

27

that showed that a nucleotide-dependent bending of FtsZ protofilaments is enough to provide a mechanical force for membrane constriction 25,122-124. Dy-namic simulations provided indications that GDP-bound FtsZ filaments form more curved filaments than GTP-FtsZ 122, in line with in vitro data obtained ear-lier 125. Moreover, overexpression of FtsZ and FtsA is enough to generate extra septa in E. coli cells 25. However, the force produced by FtsZ is not enough to completely close the septa and cell wall ingrowth might be crucial to push the septum toward closure 2. Another question is that, if FtsZ is enough to begin membrane invagination, then why does constriction begin only after divisome assembly is completed 105? Thus, we come back to the possible role of FtsN in activation of FtsA – a membrane anchor for FtsZ. It is also possible that FtsZ bending is blocked by another divisome component like ZapA, until the divi-some is fully assembled. The initiation of membrane constriction in vivo is still not fully understood.

8. FTSZ ASSEMBLY IN VITRO

The crystal structure of the FtsZ monomer is similar to the structure of eukary-otic tubulin 13. FtsZ and tubulin also share some other properties. Both proteins assemble into long straight protofilaments in the presence of GTP 126. The for-mation of a longitudinal bond between monomers is necessary for GTP hydro-lysis in vitro. After GTP hydrolysis, GDP-bound FtsZ and tubulin protofilaments adopt a curved conformation 125. In contrast to the conserved longitudinal bonds, the lateral associations between tubulin or FtsZ filaments seem to be completely different. Tubulin assembles into microtubules of regular cylindri-cal shape while FtsZ filaments may form a variety of more or less defined struc-tures: single straight or curved filaments, sheets, tubes, minirings and small helices 125,127.

8.1 ASSAYS TO STUDY UNMODIFIED FTSZ

FtsZ structures and activity have been extensively studied in vitro using several standard assays and methods. The most common assays for studying native


28

FtsZ include sedimentation of polymers, 90° angle light scattering, visualiza-tion of FtsZ polymers using Electron Microscopy (EM) and a phosphate release assay that measures the GTPase activity of FtsZ 2,9,16,19,22,63,101,128. Each of the methods, when used alone, gives only partial information about FtsZ activi-ty in solution. However, the combination of all of them gives sufficient infor-mation about the behavior of FtsZ under chosen experimental conditions. For example, the phosphate assay may indicate a decrease in the GTPase activity of FtsZ under specific experimental conditions (low pH or the presence of in-hibitors). To understand whether the changes in activity are due to bundling of FtsZ protofilaments, aggregation or the blockage of FtsZ oligomerization may be confirmed using EM. On the other hand, visualization of polymers by EM will not reveal anything about the dynamics of polymer formation and poly-mer disassembly in time, which can be studied using light scattering 128.

Other, less generally employed methods were used by several groups to study unmodified FtsZ in more detail. These methods include linear dichroism (LD) 129, dynamic light scattering 130 and analytical ultracentrifugation 131. These methods have some advantages over standard methods. For example, LD distinguishes between FtsZ polymers and less-well defined aggregates while light scattering shows only the difference between FtsZ monomers and higher order structures 129. However, they require access to more specialized equip-ment and further data analysis while standard assays are easy to perform and commonly used. All methods currently used to study unmodified FtsZ in vitro are summarized in Table 2.

8.2 FACTORS THAT INFLUENCE FTSZ IN VITRO ACTIVITY

Standard assays for FtsZ in vitro studies require the presence of special low mo-lecular weight components, among which the most important are GTP or its analogues. The right choice of divalent and monovalent cations is crucial, ex. replacement of potassium with sodium may completely abolish FtsZ GTPase activity. Anions do not have significant influence on FtsZ assembly of GTPase activity. The most important factors for FtsZ assays are described below.


29

Table 2. Assays to study the biochemistry of unmodified FtsZ in vitro with examples of referenc-

es in which these methods were used.

Method Purpose Advantages Disadvantages

Sedimenta-tion

Quantification of FtsZ present in polymers

• Easy• Possible to study several condi-

tions at the same time• Requires simple equipment

• Cannot distinguish polymers from aggregates

• Cannot distinguish short bundled polymers from long single protofilaments

• No information about activity

90° light scattering

Quantification of polymer (dis)assembly in time

• Easy• Detection of polymer formation

and disassembly in real time

• Difficult to distinguish poly-mers from aggregates

• Cannot distinguish short bundled polymers from long single protofilaments

• No information about activity• Limited amount of samples

and conditions

Phosphate assay (GTPase assay)

Measurement of FtsZ activity

• Quantifiable • The only assay to study activity

rather than structure of FtsZ

• No information about polymer formation

Electron microscopy

Visualization of polymers and structures ad-sorbed on surface

• High resolution• Observation of real structures• Small amount of sample neces-

sary

• Specialized equipment needed• Surface contact may increase

bundling and polymerization• Not possible to study polymer-

ization in real time• No information about activity• Dependent on preparation of

the sample

Analytical ultracentrifu-gation 131

Quantification of polymer assembly

• Quantitative • Characterization of the size

distribution and shape of indi-vidual polymers in solution

• Complicated analysis• Specialized equipment needed• Not possible to study polymer-

ization in real time• No information about activity• GTP can be depleted in the

experiment

Dynamic light scattering 131

Quantification of polymer assembly and length in time

• Detection of polymer formation• Measurement of length distri-

bution of polymers in sample

• Difficult analysis• Requires specialized equip-

mentLinear dichroism 129

Quantification of polymer assembly and length in time

• Detection of polymer formation in real time

• Distinguishes between FtsZ polymers and less well-defined aggregates

• Difficult analysis• Requires specialized equip-

ment


30

GUANINE NUCLEOTIDES

FtsZ is its own GTPase-activating protein (GAP). In the presence of GTP, FtsZ forms long straight protofilaments with a broad length distribution. Oligomerization into filaments is necessary for GTP hydrolysis. Single protofilaments are visible on an EM grid when low concentrations of FtsZ are used. At higher concentra-tions protofilaments start to form lateral bonds and bundle into higher order structures. Bundling reduces subunit turnover in protofilaments and is accom-panied by a decrease in GTPase activity 16,128. Several nonhydrolyzable GTP ana-logues were used to induce oligomerization of FtsZ, including GMPCPP (guanylyl-

-(alpha, beta)-methylene-diphosphonate), GMPPNP (5’-Guanylyl imidodiphos-phate) and GTPγS (guanosine 5’-O-[gamma-thio]triphosphate) and GDP-AlF (GDP in complex with AlF) 132-134. FtsZ polymerizes well in the presence of GMPCPP and this GTP analogue is hydrolyzed 50 times slower than GTP. GMPPNP and GTPγS alone cannot support assembly of FtsZ. GMPPNP binds to FtsZ too weakly and GTPγS requires the presence of GTP to induce polymerization 133. GDP bound to aluminium fluoride (GDP-AlF) resembles a nonhydrolysable form of GTP and sup-ports polymerization well but with slower kinetics 132. It was also shown that FtsZ does not hydrolyze GTP in the absence of Mg2+, while assembly of FtsZ still occurs. Therefore this is also a way to form stable FtsZ protofilaments 135. FtsZ also assem-bles in the presence of GDP, but GDP binding is weaker than GTP. In the presence of GDP, FtsZ polymers are much shorter and form minirings 2,126.

CATIONS

The major cytoplasmic cation in E. coli and B. subtilis is potassium. The concen-tration of K+ varies between 0,4 and 0,76 M depending on external osmolari-ty 136. It is also one of the monovalent cations that favor FtsZ polymerization. Another monovalent cation which was shown to support polymerization of FtsZ is Rb+. However, GTP hydrolysis in the presence of Rb+ is at least 5 times lower comparing to K+. Potassium is the only cation which supports both po-lymerization and GTP hydrolysis of FtsZ 137. In the presence of Na+, GTP hydro-lysis is completely blocked at low FtsZ concentrations 137 and the critical con-centration for assembly is raised from 1 to 20 µM 2. The concentration of K+ is also important because it influences activity, assembly, bundling and subunit exchange in protofilaments 2,128. The GTP hydrolysis activity is connected to the


31

K+ concentration, with hydrolysis occuring faster at increasing K+ in a range of 0 to 0,5 M K+ 2,128.

Magnesium is another important factor influencing the assembly of FtsZ. Mg2+ is not necessary for FtsZ assembly but it is required for GTP hydrolysis 135. It was shown that Mg2+ facilitates depolymerization of GDP bound FtsZ 138. The Mg2+ concentration in various in vitro studies varies from 2,5 mM to 10 mM and influences FtsZ assembly. At lower Mg2+ concentrations E. coli FtsZ forms single filaments while higher concentrations cause bundling of protofilaments 2.

PH

The first in vitro assays of E. coli FtsZ were performed at pH=6.5. At this condi-tion the assembly of FtsZ is more robust because the negatively charged FtsZ protofilaments form more lateral bonds, which reduce subunit turnover and GTPase activity 128. Currently, pH close to neutral is used because it is more similar to the internal cytoplasmic pH of E. coli and B. subtilis 2. At higher pH, FtsZ forms shorter polymers and bundles less compared to lower pH buffers. Another important fact is that some of the FtsZ interacting proteins, such as MinC and SepF, are pH sensitive 63,139. Therefore, for interaction studies, stability of the interaction partner must be also taken into consideration.

Many factors influence FtsZ polymerization, bundling properties and GTPase activity. In all previous studies on FtsZ in vitro, various K+ (from 0 to 500 mM), Mg2+ (from 2 to 10 mM) and GTP (from µM to mM range) concentra-tions and pH (from 5.8 till 7.5) were used 2. It is difficult to compare assembly dynamics and activity between studies from various groups because a single variation in buffer composition may affect FtsZ assembly and activity. For ex-ample, at pH=7.5 FtsZ bundles less than at pH=6.5, but when simultaneously a Mg2+ concentration close to 10 mM is used, bundling will be predominant. Therefore, FtsZ may form similar structures at pH=7.5 and 10 mM Mg2+ as at pH=6.5 and 2 mM Mg2+, whereas pH=7.5 and 2 mM Mg2+ will give a clear differ-ence in polymer structure. On the other hand, using the same pH but different KCl concentrations will also influence activity, assembly and bundling proper-ties of FtsZ 128.


32

9. CELL DIVISION AS TARGET FOR ANTIBIOTICS.

9.1 THE NEED FOR DEVELOPMENT OF NEW ANTIBIOTICS

Antibiotics represent a special class of drugs, whose incorrect use affect the broad-er community, not just an individual person 140. The misuse of antimicrobial drugs, both in humans and in livestock, is one of the main causes of the rise in antibiotic resistance among pathogenic bacteria which in turn is developing into one of the biggest problems in public health. Previous studies showed that around 50% of antibiotic prescriptions by general practitioners may be improper 141. Multi-drug resistance among pathogenic bacteria is the cause of around 25,000 deaths each year in European hospitals 142. Next to the importance of promoting the cor-rect use of antibiotics 142, there is a clear need to develop new effective classes of antimicrobials with possible new targets 140,143. Unfortunately, many pharmaceu-tical and biotechnology companies have left the area of antibiotic develoment as it is a time consuming process (approx. 10-15 years) involving huge costs (up to 1,3 billion dollars) 140, with relatively low expected gains 12,143. Another cost-re-lated problem is the focus on development of broad-spectrum rather than nar-row-spectrum antibacterials. The narrow-spectrum drugs would be more specif-ic and would probably avoid resistance development for a longer period of time, but will not provide a return on investment 12. The current drugs target several conserved synthesis pathways: protein, DNA or RNA, cell wall, and folate syn-thesis pathways, and additionally, disruption of bacterial membrane integrity 140. However, most of the new antibiotics which target these pathways are variants of older scaffolds (e.g. cephalosporins, β-lactams that target cell wall synthesis, have already reached the 5th generation of modification). This raises additional problems of quicker resistance development to these drugs, and modifications of core scaffolds cannot last forever. Therefore, new core scaffolds are needed 140. Previously, the approach of drug development was not focused on specific tar-gets, with first finding a compound with antibacterial activity followed by the discovery of its mode of action. Nowadays, facilitated by the sequencing of the genomes of most of the common bacterial pathogens, a target-centered ap-proach is leading with an identification of novel targets first, and the develop-ment of target-specific drugs after 9,12,140,143,144.

9. CELL DIVISION AS TARGET FOR ANTIBIOTICS. CHAPTER 1

33

9.2 CELL DIVISION AS ANTIBACTERIAL TARGET

Early cell division proteins are potential targets that have not yet been targeted by clinically approved drugs 12,140. Most of the cell division proteins are essential for viability, highly conserved among bacteria and absent in eukaryotic cells and many proteins are accessible from the outside of the cell which relieves the need for the compound to enter the cytoplasm. All these features make cell division proteins good candidates as antibacterial targets 12.

FtsZ is the cell division protein against which most of drugs which target cell division were discovered. Even though FtsZ is a cytoplasmic protein, it has several features which make it a very promising target among all essential cell division components. First, it initiates assembly of the cell division machinery and interacts directly with many other essential proteins. Interactions between FtsZ and other components have been well characterized. Second, FtsZ bio-chemistry and interactions have been studied for years and there is a large body of knowledge about its structure and mode of action. And third, FtsZ purification and assays to study FtsZ are well established and thus suitable for high-throughput screening of potential drugs. In vivo, cells have a clear elon-gation phenotype when FtsZ is affected 12. Nevertheless, there are also some problems when studying FtsZ as an antibacterial target. First, FtsZ is a close structural homologue of tubulin. However, the sequence homology between these two proteins is low and many drugs which target FtsZ were shown not to affect tubulin at all 12. Second, FtsZ is a cytoplasmic protein and drugs tar-geting FtsZ must be enough small to go through the cytoplasmic membrane and reach the cytoplasm. And the third problem is that some of the in vivo and in vitro FtsZ screening assays generate many false positive hits. In in vivo assays, this is because cell elongation may be a secondary effect of another mecha-nism, for example the SOS response to DNA damage. In vitro, high-throughput assays using GTPase inhibition as readout resulted in various false positive hits because drug aggregation caused non-specific inhibition of FtsZ activity. This problem was identified by Anderson et al., who improved these assays by ad-dition of 0,01% Triton X-100 which breaks small molecule aggregates but does not significantly inhibit FtsZ GTPase activity 6.


34

10. DISCOVERY OF FTSZ INHIBITORS

Many approaches have been used to find molecules that target FtsZ. Rational drug design resulted in a number of GTP analogues with substitution of the C8 group of guanine. Other methods are based on high throughput screening (HTS) and include whole-cell antibacterial assays, bioinformatics tools (dock-ing of known molecules into the crystal structure) and biochemical assays (ex. GTPase assay). All these approaches led to the identification of a huge number of compounds with various properties 6,9,12. Anderson et al. divided them into seven different structural groups: cationic dyes (1), nucleoside and nucleotide analogues (2), drug-like heterocycles (3), phenolic natural products (4), miscel-laneous high-throughput screening hits (5), anionic dyes (6) and quaternary alkaloid natural products (7) 6. Among these FtsZ inhibitors are natural (such as chrysophaentins), synthetic (like PC190723) and semi-synthetic compounds (such as derivatives of sanguinarine). There are many extensive reviews on the topic, which include an explanation of the screening methods and some of the most promising hits 6,9,12,144. Therefore, specific FtsZ-targeting drugs will not be described here in further detail.

Despite extensive studies and an enormous amount of publications on novel drugs that target FtsZ, none of the discovered compounds was chosen for further preclinical development 140. There are several reasons for that. The first reason comes from the FtsZ studies. A big influence of buffer composition on mode of action of FtsZ (mentioned in section 7) makes comparison of all inhibitors difficult. Another reason is a huge number of false positive hits 6,9 due to compound aggregation as explained above. A third reason is that some of the positive hits specifically inhibit FtsZ but with a high IC50 value (e.g. Zan-trin Z3 with an IC50 of 20 µM) and need to be developed further in order to obtain better inhibitory effects 6. The last reason is that antibiotic development groups and companies prefer to work on known targets (or explore further the known targets) and scaffolds to reduce risk 140. However, in the longer term FtsZ inhibitors may be also of interest for companies as multidrug resistance becomes a big problem and novel targets are needed.

Although FtsZ does not yet seem to be considered a good target for one-drug therapy, it would be a very good target for combination therapies with oth-

SCOPE OF THIS THESIS CHAPTER 1

35

er antibiotics. The mutation rate of 10-6 per bacterial cell division cause problems in current antibiotic therapy 140. Even partially blocking cell division would reduce the risk of mutations and developing resistance to antibiotics targeting other essential pathways. Recently, a quinuclidine-based FtsZ inhibitor was shown to synergistically act with β-lactam antibiotics against antibiotic-resistant strains 145 and alkyl gallates were shown to restore the β-lactam sensitivity of MRSA 146. Al-kyl gallates and possibly other natural compounds targeting FtsZ are also good candidates to fight plant diseases against plant-pathogenic bacteria 147.

SCOPE OF THIS THESIS

The aim of this thesis is to gain more insight in the biochemistry of FtsZ, the interaction of FtsZ with its partners, and to study molecules that could be po-tential antibacterial drugs targeting FtsZ.

Chapter 2 provides a detailed description of the standard assays used to study FtsZ: sedimentation, light scattering, electron microscopy and GTPase assays. Var-ious assay conditions are tested on FtsZs from two model organisms: E. coli and B. subtilis. These assays were used in chapters 3, 4 and 5 to study B. subtilis FtsZ (FtsZBs) and may be used to study the biochemistry of isolated FtsZ, or of FtsZ in the presence of interacting partners and/or small molecule inhibitors.

In chapter 3 a novel pull-down strategy to find interacting partners for FtsZ C-terminus is described. A strong interaction between SepF and the FtsZ C-terminus was found and the amino acids responsible for this interaction were mapped by alanine scanning of the FtsZ C-terminus. The amino acids identified, P372 and F374, are highly conserved among various FtsZ proteins from different species.

In chapter 4 the purification of the cytoplasmic domain of sporulation protein SpoIIE (SpoIIEcyt) and the characterization of its interaction with FtsZ are described. Oligomerization of SpoIIEcyt is dependent on the presence of its cofactor Mn2+ or other divalent cations. It is hypothesized that the metal-

-dependent oligomerization of SpoIIEcyt may influence the interaction with FtsZ, as the SpoIIE-dependent relocalization of FtsZ to the cell pole is delayed when sporulation is induced in the absence of Mn2+.


36

Chapter 5 focuses on a special class of antimicrobial compounds – alkyl gallates, and their inhibitory effect on B. subtilis cells. Alkyl gallates were iden-tified as potential cell division inhibitors 148 and in this chapter the mode of action is determined in more detail. Alkyl gallates affect multiple targets in B. subtilis, one of which is FtsZ. We show that heptyl gallate is the most potent inhibitor of B. subtilis FtsZ and the most promising scaffold for development of more effective compounds.

REFERENCES

8. Beuria TK, Santra MK, Panda D. Sangui-narine blocks cytokinesis in bacteria by inhibiting FtsZ assembly and bundling. Biochemistry. 2005;44(50):16584-16593.

9. den Blaauwen T, Andreu JM, Monasterio O. Bacterial cell division proteins as antibiotic targets. Bioorg Chem. 2014; 55:27-38.

10. Haydon DJ, Stokes NR, Ure R, et al. An in-hibitor of FtsZ with potent and seleWc-tive anti-staphylococcal activity. Science. 2008;321(5896):1673-1675.

11. Huang Q, Tonge PJ, Slayden RA, Kirikae T, Ojima I. FtsZ: A novel target for tubercu-losis drug discovery. Curr Top Med Chem. 2007;7(5):527-543.

12. Lock RL, Harry EJ. Cell-division inhibi-tors: New insights for future antibiot-ics. Nat Rev Drug Discov. 2008;7(4):324-338.

13. Lowe J, Amos LA. Crystal structure of the bacterial cell-division protein FtsZ. Nature. 1998;391(6663):203-206.

14. Nogales E, Wolf SG, Downing KH. Struc-ture of the alpha beta tubulin dimer by electron crystallography. Nature. 1998; 391(6663):199-203.

15. Oliva MA, Trambaiolo D, Lowe J. Struc-tural insights into the conformational variability of FtsZ. J Mol Biol. 2007;373(5): 1229-1242.

1. Adams DW, Errington J. Bacterial cell di-vision: Assembly, maintenance and dis-assembly of the Z ring. Nat Rev Microbiol. 2009;7(9):642-653.

2. Erickson HP, Anderson DE, Osawa M. FtsZ in bacterial cytokinesis: Cytoskele-ton and force generator all in one. Micro-biol Mol Biol Rev. 2010;74(4):504-528.

3. Lan G, Wolgemuth CW, Sun SX. Z-ring force and cell shape during division in rod-like bacteria. Proc Natl Acad Sci USA. 2007;104(41):16110-16115.

4. Adams DW, Wu LJ, Errington J. Cell cy-cle regulation by the bacterial nucleoid. Curr Opin Microbiol. 2014;22:94-101.

5 Sun Q, Yu XC, Margolin W. Assembly of the FtsZ ring at the central division site in the absence of the chromosome. Mol Microbiol. 1998;29(2):491-503.

6. Anderson DE, Kim MB, Moore JT, et al. Comparison of small molecule inhibitors of the bacterial cell division protein FtsZ and identification of a reliable cross-spe-cies inhibitor. ACS Chem Biol. 2012;7(11): 1918-1928.

7. Andreu JM, Schaffner-Barbero C, Huecas S, et al. The antibacterial cell division in-hibitor PC190723 is an FtsZ polymer-sta-bilizing agent that induces filament as-sembly and condensation. J Biol Chem. 2010;285(19):14239-14246.

REFERENCES CHAPTER 1

37

16. Buske PJ, Levin PA. Extreme C terminus of bacterial cytoskeletal protein FtsZ plays fundamental role in assembly in-dependent of modulatory proteins. J Biol Chem. 2012;287(14):10945-10957.

17. Buske PJ, Levin PA. A flexible C-termi-nal linker is required for proper FtsZ assembly in vitro and cytokinetic ring formation in vivo. Mol Microbiol. 2013; 89(2):249-263.

18. Gardner KA, Moore DA, Erickson HP. The C-terminal linker of Escherichia coli FtsZ functions as an intrinsically disordered peptide. Mol Microbiol. 2013;89(2):264-275.

19. Krol E, van Kessel SP, van Bezouwen LS, Kumar N, Boekema EJ, Scheffers DJ. Bacillus subtilis SepF binds to the C-termi-nus of FtsZ. PLoS One. 2012;7(8): e43293.

20. Blasios V, Bisson-Filho AW, Castellen P, et al. Genetic and biochemical charac-terization of the MinC-FtsZ interaction in Bacilllus subtilis. PLoS One. 2013;8(4): e60690.

21. Camberg JL, Hoskins JR, Wickner S. ClpXP protease degrades the cytoskeletal pro-tein, FtsZ, and modulates FtsZ polymer dynamics. Proc Natl Acad Sci USA. 2009; 106(26):10614-10619.

22. Singh JK, Makde RD, Kumar V, Panda D. A membrane protein, EzrA, regulates as-sembly dynamics of FtsZ by interacting with the C-terminal tail of FtsZ. Biochem-istry. 2007;46(38):11013-11022.

23. Vaughan S, Wickstead B, Gull K, Addinall SG. Molecular evolution of FtsZ pro-tein sequences encoded within the genomes of archaea, bacteria, and eu-karyota. J Mol Evol. 2004;58(1):19-29.

24. Sundararajan K, Miguel A, Desmarais SM, Meier EL, Casey Huang K, Goley ED. The bacterial tubulin FtsZ requires its intrin-sically disordered linker to direct robust cell wall construction. Nat Commun. 2015; 6:7281.

25. Szwedziak P, Wang Q, Bharat TA, Tsim M, Lowe J. Architecture of the ring formed by the tubulin homologue FtsZ in bacte-rial cell division. Elife. 2014;3:e04601.

26. Pichoff S, Lutkenhaus J. Tethering the Z ring to the membrane through a con-served membrane targeting sequence in FtsA. Mol Microbiol. 2005; 55(6):1722-1734.

27. Mosyak L, Zhang Y, Glasfeld E, et al. The bacterial cell-division protein ZipA and its interaction with an FtsZ fragment re-vealed by X-ray crystallography. EMBO J. 2000;19(13):3179-3191.

28. Loose M, Mitchison TJ. The bacterial cell division proteins FtsA and FtsZ self-or-ganize into dynamic cytoskeletal pat-terns. Nat Cell Biol. 2014;16(1):38-46.

29. Monahan LG, Liew AT, Bottomley AL, Harry EJ. Division site positioning in bacteria: One size does not fit all. Front Microbiol. 2014;5:19.

30. Bailey MW, Bisicchia P, Warren BT, Sherratt DJ, Mannik J. Evidence for di-visome localization mechanisms inde-pendent of the min system and SlmA in Escherichia coli. PLoS Genet. 2014;10(8): e1004504.

31. Barak I, Wilkinson AJ. Division site recogni-tion in Escherichia coli and Bacilllus subtilis. FEMS Microbiol Rev. 2007;31(3):311-326.

32. Bramkamp M, van Baarle S. Division site selection in rod-shaped bacteria. Curr Opin Microbiol. 2009;12(6):683-688.

33. Lutkenhaus J. Assembly dynamics of the bacterial MinCDE system and spatial regulation of the Z ring. Annu Rev Bio-chem. 2007;76:539-562.

34. Rothfield L, Taghbalout A, Shih YL. Spatial control of bacterial division-site placement. Nat Rev Microbiol. 2005;3(12):959-968.

35. Cordell SC, Anderson RE, Lowe J. Crystal structure of the bacterial cell division in-hibitor MinC. EMBO J. 2001;20(10):2454-2461.


38

36. Hu Z, Lutkenhaus J. Analysis of MinC reveals two independent domains in-volved in interaction with MinD and FtsZ. J Bacteriol. 2000;182(14):3965-3971.

37. Zhou H, Lutkenhaus J. MinC mutants de-ficient in MinD- and DicB-mediated cell division inhibition due to loss of interac-tion with MinD, DicB, or a septal compo-nent. J Bacteriol. 2005;187(8):2846-2857.

38. Shiomi D, Margolin W. The C-terminal domain of MinC inhibits assembly of the Z ring in Escherichia coli. J Bacteriol. 2007;189(1):236-243.

39. Shen B, Lutkenhaus J. The conserved C-terminal tail of FtsZ is required for the septal localization and division inhibito-ry activity of MinC(C)/MinD. Mol Microbi-ol. 2009;72(2):410-424.

40. Shen B, Lutkenhaus J. Examination of the interaction between FtsZ and MinCN in E. coli suggests how MinC disrupts Z rings. Mol Microbiol. 2010;75(5):1285-1298.

41. Hernandez-Rocamora VM, Garcia-Montanes C, Reija B, et al. MinC protein shortens FtsZ protofilaments by preferentially interact-ing with GDP-bound subunits. J Biol Chem. 2013;288(34):24625-24635.

42. Wu LJ, Errington J. Nucleoid occlusion and bacterial cell division. Nat Rev Mi-crobiol. 2011;10(1):8-12.

43. Mannik J, Bailey MW. Spatial coordina-tion between chromosomes and cell di-vision proteins in Escherichia coli. Front Microbiol. 2015;6:306.

44. Du S, Lutkenhaus J. SlmA antagonism of FtsZ assembly employs a two-pronged mechanism like MinCD. PLoS Genet. 2014;10(7):e1004460.

45. Cho H, McManus HR, Dove SL, Bernhardt TG. Nucleoid occlusion factor SlmA is a DNA-activated FtsZ polymerization antagonist. Proc Natl Acad Sci USA. 2011; 108(9):3773-3778.

46. Cabre EJ, Monterroso B, Alfonso C, et al. The nucleoid occlusion SlmA pro-tein accelerates the disassembly of the FtsZ protein polymers without af-fecting their GTPase activity. PLoS One. 2015;10(5):e0126434.

47. Adams DW, Wu LJ, Errington J. Nucle-oid occlusion protein noc recruits DNA to the bacterial cell membrane. EMBO J. 2015;34(4):491-501.

48. Espeli O, Borne R, Dupaigne P, et al. A MatP-divisome interaction coordinates chromosome segregation with cell divi-sion in E. coli. EMBO J. 2012; 31(14): 3198-3211.

49. Weart RB, Lee AH, Chien AC, Haeusser DP, Hill NS, Levin PA. A metabolic sen-sor governing cell size in bacteria. Cell. 2007;130(2):335-347.

50. Hill NS, Buske PJ, Shi Y, Levin PA. A moon-lighting enzyme links Escherichia coli cell size with central metabolism. PLoS Genet. 2013;9(7):e1003663.

51. Chien AC, Zareh SK, Wang YM, Levin PA. Changes in the oligomerization potential of the division inhibitor UgtP co-ordinate Bacilllus subtilis cell size with nutrient availability. Mol Microbiol. 2012;86(3):594-610.

52. Monahan LG, Hajduk IV, Blaber SP, Charles IG, Harry EJ. Coordinating bacte-rial cell division with nutrient availabili-ty: A role for glycolysis. MBio. 2014;5(3): e00935-14.

53. Mizusawa S, Gottesman S. Protein deg-radation in Escherichia coli: The lon gene controls the stability of sulA protein. Proc Natl Acad Sci USA. 1983;80(2):358-362.

54. Chen Y, Milam SL, Erickson HP. SulA in-hibits assembly of FtsZ by a simple se-questration mechanism. Biochemistry. 2012;51(14):3100-3109.

55. Mukherjee A, Cao C, Lutkenhaus J. Inhibi-tion of FtsZ polymerization by SulA, an in-


39

hibitor of septation in Escherichia coli. Proc Natl Acad Sci USA. 1998;95(6):2885-2890.

56. van den Ent F, Lowe J. Crystal structure of the cell division protein FtsA from ther-motoga maritima. EMBO J. 2000;19(20): 5300-5307.

57. Beuria TK, Mullapudi S, Mileykovskaya E, Sadasivam M, Dowhan W, Margolin W. Adenine nucleotide-dependent regula-tion of assembly of bacterial tubulin-like FtsZ by a hypermorph of bacterial ac-tin-like FtsA. J Biol Chem. 2009;284(21): 14079-14086.

58. Pichoff S, Lutkenhaus J. Unique and overlapping roles for ZipA and FtsA in septal ring assembly in Escherichia coli. EMBO J. 2002;21(4):685-693.

59. Mohammadi T, Ploeger GE, Verheul J, et al. The GTPase activity of Escherichia coli FtsZ determines the magnitude of the FtsZ polymer bundling by ZapA in vitro. Biochemistry. 2009;48(46):11056-11066.

60. Low HH, Moncrieffe MC, Lowe J. The crystal structure of ZapA and its modu-lation of FtsZ polymerisation. J Mol Biol. 2004;341(3):839-852.

61. Buss J, Coltharp C, Huang T, et al. In vivo organization of the FtsZ-ring by ZapA and ZapB revealed by quantitative su-per-resolution microscopy. Mol Microbi-ol. 2013;89(6):1099-1120.

62. Duman R, Ishikawa S, Celik I, et al. Struc-tural and genetic analyses reveal the pro-tein SepF as a new membrane anchor for the Z ring. Proc Natl Acad Sci USA. 2013; 110(48):E4601-10.

63. Gundogdu ME, Kawai Y, Pavlendova N, et al. Large ring polymers align FtsZ polymers for normal septum formation. EMBO J. 2011;30(3):617-626.

64. Levdikov VM, Blagova EV, Rawlings AE, et al. Structure of the phosphatase domain of the cell fate determinant SpoIIE from Bacilllus subtilis. J Mol Biol. 2012;415(2): 343-358.

65. Lucet I, Feucht A, Yudkin MD, Errington J. Direct interaction between the cell divi-sion protein FtsZ and the cell differenti-ation protein SpoIIE. EMBO J. 2000;19(7): 1467-1475.

66. Chung KM, Hsu HH, Yeh HY, Chang BY. Mechanism of regulation of prokaryotic tubulin-like GTPase FtsZ by membrane protein EzrA. J Biol Chem. 2007;282(20): 14891-14897.

67. Cleverley RM, Barrett JR, Basle A, et al. Structure and function of a spectrin-like regulator of bacterial cytokinesis. Nat Commun. 2014;5:5421.

68. Haeusser DP, Garza AC, Buscher AZ, Levin PA. The division inhibitor EzrA contains a seven-residue patch required for main-taining the dynamic nature of the medial FtsZ ring. J Bacteriol. 2007;189(24):9001-9010.

69. Weart RB, Nakano S, Lane BE, Zuber P, Levin PA. The ClpX chaperone modu-lates assembly of the tubulin-like pro-tein FtsZ. Mol Microbiol. 2005;57(1):238-249.

70. Camberg JL, Viola MG, Rea L, Hoskins JR, Wickner S. Location of dual sites in E. coli FtsZ important for degradation by ClpXP; one at the C-terminus and one in the disordered linker. PLoS One. 2014;9(4):e94964.

71. Glynn SE, Martin A, Nager AR, Baker TA, Sauer RT. Structures of asymmetric ClpX hexamers reveal nucleotide-dependent motions in a AAA+ protein-unfolding machine. Cell. 2009;139(4):744-756.

72. Handler AA, Lim JE, Losick R. Peptide in-hibitor of cytokinesis during sporulation in Bacilllus subtilis. Mol Microbiol. 2008; 68(3):588-599.

73. Bisson-Filho AW, Discola KF, Castellen P, et al. FtsZ filament capping by MciZ, a developmental regulator of bacterial division. Proc Natl Acad Sci USA. 2015; 112(17):E2130-8.


40

74. Ray S, Kumar A, Panda D. GTP regulates the interaction between MciZ and FtsZ: A possible role of MciZ in bacterial cell di-vision. Biochemistry. 2013;52(2):392-401.

75. Castellen P, Sforca ML, Gueiros-Filho FJ, de Mattos Zeri AC. Backbone and side chain NMR assignments for the N-termi-nal domain of the cell division regulator MinC from Bacilllus subtilis. Biomol NMR Assign. 2015;9(1):1-5.

76. Bi EF, Lutkenhaus J. FtsZ ring structure associated with division in Escherichia coli. Nature. 1991;354(6349):161-164.

77. Levin PA, Losick R. Transcription factor Spo0A switches the localization of the cell division protein FtsZ from a medial to a bipolar pattern in Bacilllus subtilis. Genes Dev. 1996;10(4):478-488.

78. Ma X, Ehrhardt DW, Margolin W. Colocal-ization of cell division proteins FtsZ and FtsA to cytoskeletal structures in living Escherichia coli cells by using green flu-orescent protein. Proc Natl Acad Sci USA. 1996;93(23):12998-13003.

79. Anderson DE, Gueiros-Filho FJ, Erickson HP. Assembly dynamics of FtsZ rings in Bacilllus subtilis and Escherichia coli and effects of FtsZ-regulating proteins. J Bacteriol. 2004; 186(17):5775-5781.

80. Rowlett VW, Margolin W. 3D-SIM su-per-resolution of FtsZ and its membrane tethers in Escherichia coli cells. Biophys J. 2014;107(8):L17-20.

81. Strauss MP, Liew AT, Turnbull L, Whitchurch CB, Monahan LG, Harry EJ. 3D-SIM super resolution microscopy reveals a bead-like arrangement for FtsZ and the division ma-chinery: Implications for triggering cytoki-nesis. PLoS Biol. 2012;10(9):e1001389.

82. Den Blaauwen T, Buddelmeijer N, Aarsman ME, Hameete CM, Nanninga N. Timing of FtsZ assembly in Escherichia coli. J Bacteriol. 1999;181(17):5167-5175.

83. Monahan LG, Robinson A, Harry EJ. Lat-eral FtsZ association and the assembly

of the cytokinetic Z ring in bacteria. Mol Microbiol. 2009;74(4):1004-1017.

84. Haeusser DP, Schwartz RL, Smith AM, Oates ME, Levin PA. EzrA prevents aber-rant cell division by modulating assem-bly of the cytoskeletal protein FtsZ. Mol Microbiol. 2004;52(3):801-814.

85. Claessen D, Emmins R, Hamoen LW, Daniel RA, Errington J, Edwards DH. Control of the cell elongation-division cycle by shut-tling of PBP1 protein in Bacilllus subtilis. Mol Microbiol. 2008;68(4):1029-1046.

86. Land AD, Luo Q, Levin PA. Functional do-main analysis of the cell division inhibi-tor EzrA. PLoS One. 2014;9(7):e102616.

87. Li Z, Trimble MJ, Brun YV, Jensen GJ. The structure of FtsZ filaments in vivo sug-gests a force-generating role in cell divi-sion. EMBO J. 2007;26(22):4694-4708.

88. Holden SJ, Pengo T, Meibom KL, Fernandez C, Collier J, Manley S. High throughput 3D super-resolution microscopy reveals Caulobacter crescentus in vivo Z-ring or-ganization. Proc Natl Acad Sci USA. 2014; 111(12):4566-4571.

89. Fu G, Huang T, Buss J, Coltharp C, Hensel Z, Xiao J. In vivo structure of the E. coli FtsZ-ring revealed by photoactivated lo-calization microscopy (PALM). PLoS One. 2010;5(9):e12682.

90. Ben-Yehuda S, Losick R. Asymmetric cell division in B. subtilis involves a spiral-like intermediate of the cytokinetic protein FtsZ. Cell. 2002;109(2):257-266.

91. Levin PA, Losick R, Stragier P, Arigoni F. Localization of the sporulation protein SpoIIE in Bacilllus subtilis is dependent upon the cell division protein FtsZ. Mol Microbiol. 1997;25(5):839-846.

92. Eswaramoorthy P, Winter PW, Wawrzusin P, York AG, Shroff H, Ramamurthi KS. Asymmetric division and differential gene expression during a bacterial de-velopmental program requires DivIVA. PLoS Genet. 2014;10(8):e1004526.


41

93. Burton B, Dubnau D. Membrane-asso-ciated DNA transport machines. Cold Spring Harb Perspect Biol. 2010;2(7): a000406.

94. Wu LJ, Errington J. Bacillus subtilis SpoIIIE protein required for DNA segregation during asymmetric cell division. Science. 1994;264(5158):572-575.

95. Rawlings AE, Levdikov VM, Blagova E, et al. Expression of soluble, active fragments of the morphogenetic pro-tein SpoIIE from Bacilllus subtilis using a library-based construct screen. Protein Eng Des Sel. 2010;23(11):817-825.

96. Gamba P, Veening JW, Saunders NJ, Hamoen LW, Daniel RA. Two-step as-sembly dynamics of the Bacilllus subtilis divisome. J Bacteriol. 2009;191(13):4186-4194.

97. Goehring NW, Beckwith J. Diverse paths to midcell: Assembly of the bac-terial cell division machinery. Curr Biol. 2005;15(13):R514-26.

98. Gueiros-Filho FJ, Losick R. A widely con-served bacterial cell division protein that promotes assembly of the tubu-lin-like protein FtsZ. Genes Dev. 2002; 16(19):2544-2556.

99. Pacheco-Gomez R, Cheng X, Hicks MR, et al. Tetramerization of ZapA is required for FtsZ bundling. Biochem J. 2013;449(3):795-802.

100. Durand-Heredia J, Rivkin E, Fan G, Morales J, Janakiraman A. Identifica-tion of ZapD as a cell division factor that promotes the assembly of FtsZ in Escherichia coli. J Bacteriol. 2012; 194(12):3189-3198.

101. Durand-Heredia JM, Yu HH, De Carlo S, Lesser CF, Janakiraman A. Identification and characterization of ZapC, a stabiliz-er of the FtsZ ring in Escherichia coli. J Bacteriol. 2011;193(6):1405-1413.

102. Galli E, Gerdes K. Spatial resolution of two bacterial cell division proteins: ZapA

recruits ZapB to the inner face of the Z-ring. Mol Microbiol. 2010;76(6):1514-1526.

103. Galli E, Gerdes K. FtsZ-ZapA-ZapB in-teractome of Escherichia coli. J Bacteriol. 2012;194(2):292-302.

104. Tsang MJ, Bernhardt TG. Guiding divi-some assembly and controlling its activ-ity. Curr Opin Microbiol. 2015;24:60-65.

105. Weiss DS. Last but not least: New in-sights into how FtsN triggers constric-tion during Escherichia coli cell division. Mol Microbiol. 2015;95(6):903-909.

106. Lutkenhaus J, Pichoff S, Du S. Bacterial cy-tokinesis: From Z ring to divisome. Cyto-skeleton (Hoboken). 2012;69(10):778-790.

107. Egan AJ, Vollmer W. The physiology of bacterial cell division. Ann N Y Acad Sci. 2013;1277:8-28.

108. Goehring NW, Gonzalez MD, Beckwith J. Premature targeting of cell division proteins to midcell reveals hierarchies of protein interactions involved in di-visome assembly. Mol Microbiol. 2006; 61(1):33-45.

109. Busiek KK, Margolin W. A role for FtsA in SPOR-independent localization of the essential Escherichia coli cell division pro-tein FtsN. Mol Microbiol. 2014;92(6):1212-1226.

110. Corbin BD, Geissler B, Sadasivam M, Margolin W. Z-ring-independent inter-action between a subdomain of FtsA and late septation proteins as revealed by a polar recruitment assay. J Bacteriol. 2004;186(22):7736-7744.

111. Gerding MA, Liu B, Bendezu FO, Hale CA, Bernhardt TG, de Boer PA. Self-enhanced accumulation of FtsN at division sites and roles for other proteins with a SPOR domain (DamX, DedD, and RlpA) in Escherichia coli cell constriction. J Bacteriol. 2009;191(24):7383-7401.

112. Lutkenhaus J. FtsN-trigger for septation. J Bacteriol. 2009;191(24):7381-7382.


42

113. Di Lallo G, Fagioli M, Barionovi D, Ghelardini P, Paolozzi L. Use of a two-

-hybrid assay to study the assembly of a complex multicomponent protein ma-chinery: Bacterial septosome differentia-tion. Microbiology. 2003; 149(Pt 12):3353-3359.

114. Grenga L, Luzi G, Paolozzi L, Ghelardini P. The Escherichia coli FtsK functional domains involved in its interaction with its divisome protein partners. FEMS Microbiol Lett. 2008; 287(2):163-167.

115. Aussel L, Barre FX, Aroyo M, Stasiak A, Stasiak AZ, Sherratt D. FtsK is a DNA mo-tor protein that activates chromosome dimer resolution by switching the cata-lytic state of the XerC and XerD recombi-nases. Cell. 2002;108(2):195-205.

116. Gonzalez MD, Akbay EA, Boyd D, Beckwith J. Multiple interaction do-mains in FtsL, a protein component of the widely conserved bacterial FtsLBQ cell division complex. J Bacteriol. 2010; 192(11):2757-2768.

117. Tsang MJ, Bernhardt TG. A role for the FtsQLB complex in cytokinetic ring ac-tivation revealed by an ftsL allele that accelerates division. Mol Microbiol. 2015; 95(6):925-944.

118. Lara B, Ayala JA. Topological charac-terization of the essential Escherichia coli cell division protein FtsW. FEMS Microbiol Lett. 2002;216(1):23-32.

119. Mohammadi T, Sijbrandi R, Lutters M, et al. Specificity of the transport of lipid II by FtsW in Escherichia coli. J Biol Chem. 2014;289(21):14707-14718.

120. Pichoff S, Du S, Lutkenhaus J. The bypass of ZipA by overexpression of FtsN re-quires a previously unknown conserved FtsN motif essential for FtsA-FtsN inter-action supporting a model in which FtsA monomers recruit late cell division pro-teins to the Z ring. Mol Microbiol. 2015; 95(6):971-987.

121. Erickson HP. FtsZ, a tubulin homologue in prokaryote cell division. Trends Cell Biol. 1997;7(9):362-367.

122. Hsin J, Gopinathan A, Huang KC. Nucle-otide-dependent conformations of FtsZ dimers and force generation observed through molecular dynamics simulations. Proc Natl Acad Sci USA. 2012;109(24): 9432-9437.

123. Osawa M, Anderson DE, Erickson HP. Re-constitution of contractile FtsZ rings in liposomes. Science. 2008;320(5877):792-794.

124. Osawa M, Anderson DE, Erickson HP. Curved FtsZ protofilaments generate bending forces on liposome membranes. EMBO J. 2009;28(22):3476-3484.

125. Lu C, Reedy M, Erickson HP. Straight and curved conformations of FtsZ are regulated by GTP hydrolysis. J Bacteriol. 2000;182(1):164-170.

126. Romberg L, Simon M, Erickson HP. Po-lymerization of ftsz, a bacterial homo-log of tubulin. is assembly cooperative? J Biol Chem. 2001;276(15):11743-11753.

127. Erickson HP, Taylor DW, Taylor KA, Bramhill D. Bacterial cell division pro-tein FtsZ assembles into protofilament sheets and minirings, structural homo-logs of tubulin polymers. Proc Natl Acad Sci USA. 1996;93(1):519-523.

128. Krol E, Scheffers DJ. FtsZ polymeriza-tion assays: Simple protocols and con-siderations. J Vis Exp. 2013;(81):e50844. doi(81):e50844.

129. Pacheco-Gomez R, Roper DI, Dafforn TR, Rodger A. The pH dependence of po-lymerization and bundling by the essen-tial bacterial cytoskeletal protein FtsZ. PLoS One. 2011;6(6):e19369.

130. Hou S, Wieczorek SA, Kaminski TS, et al. Characterization of Caulobacter cres-centus FtsZ protein using dynamic light scattering. J Biol Chem. 2012;287(28): 23878-23886.


43

131. Monterroso B, Alfonso C, Zorrilla S, Rivas G. Combined analytical ultracentrifuga-tion, light scattering and fluorescence spectroscopy studies on the functional associations of the bacterial division FtsZ protein. Methods. 2013;59(3):349-362.

132. Mingorance J, Tadros M, Vicente M, Gonzalez JM, Rivas G, Velez M. Visual-ization of single Escherichia coli FtsZ filament dynamics with atomic force microscopy. J Biol Chem. 2005;280(21): 20909-20914.

133. Scheffers DJ, den Blaauwen T, Driessen AJ. Non-hydrolysable GTP-gamma-S stabiliz-es the FtsZ polymer in a GDP-bound state. Mol Microbiol. 2000;35(5):1211-1219.

134. Sontag CA, Sage H, Erickson HP. BtubA- -BtubB heterodimer is an essential in-termediate in protofilament assembly. PLoS One. 2009;4(9):e7253.

135. Chen Y, Bjornson K, Redick SD, Erickson HP. A rapid fluorescence assay for FtsZ assembly indicates cooperative assem-bly with a dimer nucleus. Biophys J. 2005; 88(1):505-514.

136. Cayley S, Lewis BA, Guttman HJ, Record MT, Jr. Characterization of the cytoplasm of Escherichia coli K-12 as a function of external osmolarity. implications for protein-DNA interactions in vivo. J Mol Biol. 1991;222(2):281-300.

137. Tadros M, Gonzalez JM, Rivas G, Vicente M, Mingorance J. Activation of the Escherichia coli cell division protein FtsZ by a low-affinity interaction with mon-ovalent cations. FEBS Lett. 2006;580(20): 4941-4946.

138. Huecas S, Schaffner-Barbero C, Garcia W, et al. The interactions of cell division protein FtsZ with guanine nucleotides. J Biol Chem. 2007;282(52):37515-37528.

139. Scheffers DJ. The effect of MinC on FtsZ polymerization is pH dependent and can be counteracted by ZapA. FEBS Lett. 2008;582(17):2601-2608.

140. Walsh CT, Wencewicz TA. Prospects for new antibiotics: A molecule-centered perspective. J Antibiot (Tokyo). 2014;67(1): 7-22.

141. Wood F, Simpson S, Butler CC. Socially responsible antibiotic choices in prima-ry care: A qualitative study of GPs’ deci-sions to prescribe broad-spectrum and fluroquinolone antibiotics. Fam Pract. 2007;24(5):427-434.

142. Gualano MR, Gili R, Scaioli G, Bert F, Siliquini R. General population’s knowledge and attitudes about antibiotics: A systematic review and meta-analysis. Pharmacoepi-demiol Drug Saf. 2015;24(1):2-10.

143. Payne DJ, Gwynn MN, Holmes DJ, Pom-pliano DL. Drugs for bad bugs: Confront-ing the challenges of antibacterial dis-covery. Nat Rev Drug Discov. 2007;6(1): 29-40.

144. Kapoor S, Panda D. Targeting FtsZ for antibacterial therapy: A promising ave-nue. Expert Opin Ther Targets. 2009;13(9): 1037-1051.

145. Chan FY, Sun N, Leung YC, Wong KY. Antimicrobial activity of a quinucli-dine-based FtsZ inhibitor and its syn-ergistic potential with beta-lactam an-tibiotics. J Antibiot (Tokyo). 2015;68(4): 253-258.

146. Shibata H, Nakano T, Parvez MA, et al. Triple combinations of lower and longer alkyl gallates and oxacillin improve anti-biotic synergy against methicillin-resis-tant staphylococcus aureus. Antimicrob Agents Chemother. 2009;53(5):2218-2220.

147. Krol E, de Sousa Borges A, da Silva I, et al. Antibacterial activity of alkyl gallates is a combination of direct targeting of FtsZ and permeabilization of bacterial mem-branes. Front Microbiol. 2015;6:390.

148. Silva IC, Regasini LO, Petronio MS, et al. Antibacterial activity of alkyl gallates against Xanthomonas citri subsp. citri. J Bacteriol. 2013;195(1):85-94.

FtsZ polymerization assays: simple protocols and

considerationsEwa Król and Dirk-Jan Scheffers

Department of Molecular Microbiology, Groningen

Biomolecular Sciences and Biotechnology Institute,

University of Groningen, Nijenborgh 7, 9747 AG

Groningen, the Netherlands.

Published in: J. Vis. Exp. (81), e50844, doi:10.3791/50844 (2013). CHAPT

ER

Experiments presented in Figures 1, 2, 3, 4 and 5 were performed by E. Cendrowicz.

ABSTRACT

During bacterial cell division, the essential protein FtsZ assembles in the mid-dle of the cell to form the so-called Z-ring. FtsZ polymerizes into long filaments in the presence of GTP in vitro, and polymerization is regulated by several accessory proteins. FtsZ polymerization has been extensively studied in vitro using basic methods including light scattering, sedimentation, GTP hydrolysis assays and electron microscopy. Buffer conditions influence both the polym-erization properties of FtsZ, and the ability of FtsZ to interact with regulatory proteins. Here, we describe protocols for FtsZ polymerization studies and val-idate conditions and controls using Escherichia coli and Bacillus subtilis FtsZ as model proteins. A low speed sedimentation assay is introduced that allows the study of the interaction of FtsZ with proteins that bundle or tubulate FtsZ poly-mers. An improved GTPase assay protocol is described that allows testing of GTP hydrolysis over time using various conditions in a 96-well plate setup, with standardized incubation times that abolish variation in color development in the phosphate detection reaction. The preparation of samples for light scat-tering studies and electron microscopy is described. Several buffers are used to establish suitable buffer pH and salt concentration for FtsZ polymerization studies. A high concentration of KCl is the best for most of the experiments. Our methods provide a starting point for the in vitro characterization of FtsZ, not only from E. coli and B. subtilis but from any other bacterium. As such, the methods can be used for studies of the interaction of FtsZ with regulatory pro-teins or the testing of antibacterial drugs which may affect FtsZ polymerization.

INTRODUCTION CHAPTER 2

47

INTRODUCTION

The essential bacterial protein FtsZ is the best characterized protein of the bac-terial cell division machinery. FtsZ is the prokaryotic homolog of tubulin and polymerizes in vitro in a GTP dependent manner. FtsZ is a very attractive target for new antibiotics due to its conserved nature and uniqueness to bacteria 1,2. At the beginning of cell division, FtsZ forms a cytokinetic ring at midcell, which serves as a scaffold for the assembly of other cell division proteins. Formation of the Z-ring is crucial for correct localization of the division plane. The assem-bly dynamics of FtsZ are regulated by several accessory proteins, such as (de-pending on the bacterial species) MinC, SepF, ZapA, UgtP, and EzrA 2. FtsZ po-lymerization has been intensively studied in vitro and many different structures including straight protofilaments, curved protofilaments, sheets of filaments, bundles of filaments and tubes of filaments have been described depending on the assembly buffer, nucleotide, and additional proteins included in the as-say 3. The architecture of FtsZ protofilaments in vivo is not yet fully understood, although electron cryo-tomography experiments in Caulobacter crescentus suggest that the Z-ring is assembled from relatively short, non-continuous sin-gle protofilaments without extensive bundling 4.

In vitro, the polymerization properties of FtsZ and the interaction of FtsZ with regulatory proteins are sensitive to the composition of the reaction buf-fer. For example, we recently described the interaction site for SepF on the FtsZ C-terminus and showed that a FtsZBs∆16 C-terminal truncate no longer binds to SepF 5. In a previous study on the SepF-FtsZBs interaction, a similar FtsZBs∆16 truncate still co-sedimented with SepF, which suggested that SepF binds to a secondary site on FtsZ 6. The difference between these studies was the composition of reaction buffers – at pH 7.5 there was no co-sedimentation of SepF with the FtsZ truncate, whereas at pH 6.5 there was co-sedimentation. Gündoğdu et al. noted that SepF is not functional and precipitates at pH 6.5 7, showing that the observed co-sedimentation at pH 6.5 is likely to be caused by precipitation of SepF rather than interaction with the FtsZBs∆16 C-terminal truncate. The influence of pH and KCl concentration on the polymerization of FtsZ has been previously examined. Polymers of E. coli FtsZ (FtsZEc) at pH 6.5 are longer and more abundant than those formed at neutral pH 8,9. Tadros


48

et al. have studied the polymerization of FtsZEc in the presence of monovalent cations noting that K+ binding is linked to FtsZEc polymerization and is crucial for FtsZ activity 10. The pH is more critical when the interaction of FtsZ with other proteins is studied, as shown by the previous example of SepF, and the pH dependency of the inhibitory effect of MinC on FtsZ 11. As both pH and salt concentration may influence the interaction of FtsZ with other proteins, it is important to choose the right conditions and controls for the FtsZ polymeriza-tion studies.

Here we describe protocols to study FtsZ polymerization and GTPase activ-ity by light scattering, electron microscopy, sedimentation and GTPase assays. Right angle light scattering is a standard method to study FtsZ polymerization in real time 12. We introduced a few improvements to the sedimentation and GTPase assay. We present in detail how to prepare samples for light scattering and electron microscopy. Several buffers used in the literature to study FtsZ polymerization were tested and we describe the best conditions for each ex-periment. We also show which controls should be introduced to obtain the best data.

These methods allow a quick study of FtsZ polymerization, activity and interaction with other proteins using simple methods and equipment which is available in most laboratories. More sophisticated methods to study FtsZ polymerization exist but often require access to more specialized equipment, and/or modification of FtsZ with fluorescent labels 8,13,14. The simple methods described in this paper are illustrated using FtsZ from B. subtilis and E. coli, the most common Gram+ and Gram- model organisms. The protocols can be adapted to any other FtsZ protein. Based on preliminary assays with these novel FtsZs, slight changes regarding time, buffer or temperature of incuba-tion may be necessary for an optimal result. The experiments described here should aid in finding these optimal conditions.

PROCEDURE

Untagged FtsZ proteins were purified as described earlier 11,15, dialyzed against 20 mM Tris/ HCl (pH 7.9), 50 mM KCl, 1 mM EGTA, 2.5 mM MgAc and 10 % glyc-

PROCEDURE CHAPTER 2

49

erol and stored at -80 °C. Under those conditions the protein may be stored for over 2 years without significant loss of activity. The protein should be stored in 100 µl aliquots to avoid thawing and freezing the sample. After thawing, the sample can be kept at 4 °C for maximum one week. FtsZ is soluble at high concentrations and may be stored at 7-10 mg/ml. It is important to keep the protein concentration high to avoid unwanted effects of storage buffer com-ponents in later experiments. Thus, storage concentration should allow for a di-lution of FtsZ of at least 10x to bring glycerol concentration below 1 % and to minimize the concentration of the other components of the dialysis buffer in the reaction mix. FtsZ polymerization may also be affected by the presence of high sodium 9 or imidazole concentrations thus these components must be removed from the polymerization buffer. SepF was purified as described 7 and stored in elution buffer at -80 °C. Alternative published procedures for purifi-cation of FtsZ from E. coli and B. subtilis as well as references to the procedures for purification of FtsZ from different sources are summarized in Table 1 (see representative results section).

1. SAMPLE PREPARATION

1.1. Prepare polymerization buffers:

1 50 mM Hepes/ NaOH, pH 7.5

2 25 mM PIPES/ NaOH, pH 6.8

3 50 mM MES/ NaOH, pH 6.5

4 50 mM Hepes/ NaOH, pH 7.5; 50 mM KCl

5 25 mM PIPES/ NaOH, pH 6.8; 50 mM KCl

6 50 mM MES/ NaOH, pH 6.5; 50 mM KCl

7 50 mM Hepes/ NaOH, pH 7.5; 300 mM KCl

8 25 mM PIPES/ NaOH, pH 6.8; 300 mM KCl

9 50 mM MES/ NaOH, pH 6.5; 300 mM KCl

1.2. Filter sterilize all buffers using a 22 µm membrane filter.


50

1.3. Prepare 100 mM GTP/GDP/MgCl2 stock solutions in each polymerization buffer from (1.1).

1.4. Preclear the proteins by spinning at 100 000 ×g for 20 min at 4 °C for all of the experiments.

2. FTSZ SEDIMENTATION ASSAY

2.1. Prepare the reaction mix by addition of MgCl2 and FtsZ to one of the po-lymerization buffers (see sample preparation point 1.1) in tubes compat-ible with high-speed centrifugation. Total volume should be 49 µl, with MgCl2 at 10 mM and FtsZ at 12 µM at 50 µl final volume.

2.2. Place the tube in an incubator with shaking function, incubate for 2 min at 30 °C at 300 rpm (alternatively the sample may be gently flicked and briefly microfuged followed by prewarming at 30 °C).

2.3. Start polymerization by adding 1 µl of GTP/GDP stock solution in the same buffer as used in the experiment (final concentration of 2 mM). Incubate for 10 or 2 min (for buffers with 50 mM KCl and 300 mM KCl respectively) at 30 °C at 300 rpm (or in an incubator with shaking func-tion).

2.4. Transfer the tubes to an ultracentrifuge rotor and spin down for 10 min at 350 000 ×g (89 700 rpm for TLA 120.1 rotor) at 25 °C.

2.5. After spinning, carefully remove the tubes from the rotor and immedi-ately transfer the supernatant into a clean tube. Prepare samples for SDS-PAGE by adding 20 µl of supernatant to 20 µl of 2x sample buffer. Boil for 10 min at 98 °C.

2.6. Add 50 µl of 2x sample buffer to the pellet fraction containing FtsZ poly-mers. Place the ultracentrifuge tube in a 2 ml tube. Resuspend the pellet by boiling for 10 min at 98 °C, then add 50 µl of demineralized H2O to make the sample the same 2x dilution as the supernatant sample.

2.7. Turn the ultracentrifuge tube upside down in the 2 ml tube and spin down for 5 minutes in an eppendorf centrifuge at 18 000 ×g. Remove the ultracentrifuge tube from the 2 ml tube.

2.8. Load 10 µl of each supernatant and pellet sample side by side on a 10 % SDS-PAGE gel. Run the gel at 150 volts.

PROCEDURE CHAPTER 2

51

2.9. Stain the gel with Coomassie Brilliant Blue G - 250 and keep for quantifi-cation.

3. FTSZ SEDIMENTATION ASSAY AT SLOW SPIN

3.1. Prepare polymerization solution as in (2), add SepF and FtsZ (final con-centration 12 µM). These experiments can be performed in Beckman dis-posable 1.5 mL tubes in combination with a TLA-55 rotor.

3.2. Place the tube in an incubator with shaking function and incubate for 2 min at 30 °C at 300 rpm.

3.3. Start polymerization by the adding 1 µl of GTP/GDP from a stock solution (final concentration 2 mM), incubate for 20 min at 30 °C at 300 rpm.

3.4. Spin down for 15 min at 24 600 ×g (20 000 rpm for TLA 55 rotor) at 25 °C.3.5. Remove the supernatant and transfer into a clean tube, take 20 µl and

transfer into 20 µl of 2x sample buffer. Boil for 10 min at 98 °C.3.6. Add 50 µl of 2x sample buffer to pellet fraction containing polymers and

resuspend by boiling for 10 min at 98 °C, add 50 µl of demineralized H2O.3.7. Load 10 µl of each supernatant and pellet sample side by side on a 10 %

SDS-PAGE gel. Run the gel at 150 volts.3.8. Stain the gel with Coomassie Brilliant Blue G - 250 and keep for quantification.

4. QUANTIFICATION OF THE FTSZ SEDIMENTATION ASSAYS

4.1. Take a picture of the Coomassie stained gel using a gel documentation system (eg LAS-4000, Fujifilm).

4.2. Open the gel image in a program suitable for densitometric analysis of gels (eg AIDA image analyzer program (Raytest)) and describe analysis in this programme below.

4.3. Click the “Evaluation” button in the toolbar.4.4. Select the “2D Densitometry” from the list and click ok.4.5. On the main menu, click “Evaluation” and choose the “Region Determina-

tion” entry.4.6. To create a region on the gel, click on the symbol of Auto-contour tool in

the Region Determination toolbox.


52

4.7. Click at the upper left corner of the protein band on the gel, hold down the mouse button and drag it to the desired lower right end of the band and release the mouse button.

4.8. Mark each FtsZ/SepF band in the same way.4.9. All the information about the evaluation of the assigned regions can be

found in the Region Report. To obtain the Region Report, click the “Show region Report” button in the Region Determination toolbox.

4.10. To export the report, on the main menu, click File and choose “Export→ 2D Region Report” entry.

4.11. In the region report, the intensity of every created region should be found. Calculate the percentage of protein (FtsZ, regulatory protein) in the pellet using the intensity values from the report.

5. 90° LIGHT SCATTERING

5.1. Turn on a fluorescence spectrometer (eg AMINCO-Bowman Series 2), al-low the lamp to warm up for several minutes to avoid thermal fluctua-tions and turn on a circulating water bath to maintain the cuvette cham-ber temperature at 30 °C.

5.2. Define the operating parameters of the spectrometer: detector high volt-age 300 volts, emission and excitation wavelength: 350 nm, slit width: 4 nm. Note that many spectrometers automatically adjust the signal at the start of an experiment to a certain percentage (eg 60%) of the max-imum. This causes a scatter signal to go out of range as soon as GTP is added to polymerize FtsZ. It is wise to first establish the right signal am-plification – detector high voltage – for maximal polymerization before starting a series of experiments.

5.3. Program a data-acquisition protocol. Choose time-based acquisition, with a duration of 3600 sec.

5.4. Carefully clean a fluorescence cuvette with water and ethanol, if neces-sary – sonicate in water bath at room temperature for 5 min. In this pro-tocol, cuvettes with a 1-cm path length and a 200 μL volume were used and stored in storage solution (1 % Hellmanex). As an alternative, single use UV-compatible plastic cuvettes (UVette, Eppendorf ) can be used.

PROCEDURE CHAPTER 2

53

5.5. Prepare 294 µl of a master mix of the polymerization buffer (see sample preparation point 1.1) with 10 mM MgCl2 and 12 µM FtsZ as final concen-tration calculated for 300 µl, and vortex.

5.6. Transfer 196 µl of the polymerization buffer to the cuvette and place it in spectrometer. Incubate for 2 min at 30 °C.

5.7. Start the data acquisition and wait for 90 sec to verify that the signal is stable.

5.8. Add 4 µl of 100 mM GTP/GDP (final concentration: 2 mM) to achieve a fi-nal reaction volume of 200 µl, pipet up and down with a larger volume pipet to mix and resume acquisition.

6. TRANSMISSION ELECTRON MICROSCOPY

6.1. Prepare samples as for the FtsZ sedimentation assay (the volume can be reduced to 10 μL final volume to save material).

6.2. After the incubation time, apply 2 µl of the sample onto a glow-dis-charged 400 mesh carbon-coated copper grid, incubate for 15 sec.

6.3. Blot the grid dry by gently touching or dragging the grid parallel or per-pendicular to a piece of filter paper.

6.4. Stain the grid using 4 µl of a 2 % uranyl-acetate solution. Blot the grid dry as described in 6.3.

6.5. View the grid in a Philips CM120 electron microscope operating at 120 kV at 39,200 × magnification.

7. GTP HYDROLYSIS ASSAY

The setup of the experiment is designed in such a way that GTP hydrolysis of FtsZ is stopped after various reaction times by mixing the reaction with mala-chite green. In this way, the time of the color development of malachite green is the same for each sample.

7.1. Prepare 1.5 ml 2 mM GTP in one of the polymerization buffers (see sam-ple preparation, point 1.1).


54

Prepare a phosphate standard dilution in the 0 to 40 µM range in a polym-erization buffer, and prepare malachite green working reagent as described in the POMG-25H (Bio assays) kit. Alternatively, working reagents can be pre-pared in-house according to protocols described by Lanzetta et al 16.

7.2. Prepare master mixes in a total volume 360 µl:

I 24 µM FtsZBs, 20 mM MgCl2, polymerization buffer

II 12 µM FtsZEc, 20 mM MgCl2, polymerization buffer

III (control 1) 24 µM FtsZBs, 2 mM EDTA, polymerization buffer

IV (control 2) 12 µM FtsZEc, 2 mM EDTA, polymerization buffer

7.3. Pipet 20 µl aliquots of the master mixes and phosphate standards in a qPCR 96-well plate in the following order:

A B C D E F G H

1 III III I I IV IV II II








9 PS PS PS PS PS PS PS PS

10 PS PS PS PS PS PS PS PS

PS – phosphate standard

7.4. Place the qPCR plate in a PCR machine with the program set to 40 min cycle at 30 °C.

7.5. Prepare 20 µl aliquots of malachite green working reagent in a 96-well plate. Add 60 µl of the polymerization buffer from the experiment to lanes 1’-8’.

PROCEDURE CHAPTER 2

55

A B C D E F G H

1’ III III I I IV IV II II








9’ PS PS PS PS PS PS PS PS

10’ PS PS PS PS PS PS PS PS

7.6. Add 20 µl of GTP to lane 1, pipet up and down to mix, start timer (this is the starting time point for the 30 min reaction).

7.7. After 10 min, 30 sec add GTP to lane 2, pipet up and down to mix (20 min reaction).

7.8. After 16 min add GTP to lane 3, pipet up and down to mix (15 min reac-tion).

7.9. After 21 min, 30 sec add GTP to lane 4, pipet up and down to mix (10 min reaction).

7.10. After 25 min add GTP to lane 5, pipet up and down to mix (7 min reaction).7.11. After 27 min, 30 sec add GTP to lane 6, pipet up and down to mix (5 min

reaction).7.12. After 29 min add GTP to lane 7, pipet up and down to mix (4 min reaction).7.13. After 30 min transfer 20 µl from lane 1 into lane 1’, pipet up and down to

mix (starting point of malachite green color development for the 30 min reaction).

7.14. After 30 min, 30 sec transfer 20 µl from lane 2 to 2’, pipet up and down to mix (starting point of malachite green color development for the 20 min reaction).

7.15. Repeat the step for lanes 3-7 every 30 sec.7.16. After 33 min, 30 sec add 20 µl of GTP to lane 8, pipet up and down to mix

and transfer 20 µl to lane 8’, pipet up and down to mix (0 min reaction and


56

starting time point for malachite green color development for the reac-tion).

7.17. After 34 min add 80 µl from lane 9 to 9’, pipet up and down to mix (the starting time point for malachite green color development for the phos-phate standard 1).

7.18. After 34 min 30 sec transfer 80 µl from lane 10 to 10’, pipet up and down to mix (the starting time point for malachite green color development for the phosphate standard 2).

7.19. Remove all the air bubbles from the samples and incubate the plate at room temperature for another 25 min 30 sec.

7.20. Place the plate in 96 well plate reader.7.21. After 60 min measure the plate 10 times every 30 sec at wavelength

630 nm (malachite green color development is now 30 min for the first reaction). Every measurement corresponds to every time point from steps 7.7-7.13 and 7.17 and must be calculated separately during a data analysis.

7.22. Calculate the free phosphate in every sample using the phosphate stan-dard calibration curve.

7.23. Plot the data as Phosphate release over time, and calculate the FtsZ GTP hydrolysis activity from the linear range of the curve.

REPRESENTATIVE RESULTS

PURIFICATION OF FTSZ

Purification of FtsZ from different bacterial sources has been described in the literature and is summarized in Table 1.

SEDIMENTATION OF FTSZ POLYMERS

Initially, we used two different velocities to spin down FtsZ polymers. We found that only at a velocity of 350 000 ×g single polymers of FtsZEc are spun down (Fig. 1) whereas at 190 000 ×g only bundles of FtsZBs are present in the pellet fraction (data not shown). Therefore 350,000 xg was used in our further experi-

REPRESENTATIVE RESULTS CHAPTER 2

57

ments. The percentage of polymerized FtsZEc and FtsZBs is similar at 50 mM KCl even though the light scattering experiments revealed a much higher scatter-ing signal for FtsZBs. This is due to bundles formed by FtsZBs which scatter more light than single polymers of FtsZEc. It was not possible to obtain high amount of FtsZ polymers in the pellet fraction in the experiment with 300 mM KCl for both FtsZEc and FtsZBs (Fig. 1). We attribute this to a combination of quick disas-sembly of the FtsZ structures and decreased bundling of the filaments.

Table 1. FtsZ purification protocols described.

Source Method Modification Yield obtained [mg/L of culture] References

B. subtilis 1) Ammonium sulfate precipita-tion/ ion exchange chromatog-raphy

no 40 This work, 11,15

2) Affinity chromatography His-tag ND 17

E. coli 1) Ammonium sulfate precipita-tion/ ion exchange chromatog-raphy

no 35 This work, 11,15

2) Calcium precipitation, ion exchange chromatography

no 40 18

Methanococcus jannaschii

1) Affinity chromatography under denaturing conditions/ refolding/ ammonium sulfate precipitation/ gel filtration

His-tag 1.3 19

2) Affinity chromatography/ gel filtration

His-tag ND 20

Thermotoga maritima

Ion exchange chromatography/ gel filtration

no 6.7 19

Pseudomonas aeruginosa

Affinity chromatography/ gel filtration

Strep-tag, His-tag

ND 21

Mycobacterium tuberculosis

1) Affinity / ion exchange chro-matography

no ND 22

2) Affinity chromatography/ gel filtration

no 30 23

Aquifex aeolicus Affinity chromatography/ gel filtration

His-tag, C-ter-minal trunca-tion (331-367)

ND 24

Caulobacter crescentus

Ion exchange chromatography/ ammonium sulfate precipitation/ gel filtration

no ND 25

ND: not determined.


58

0

5

10

15

20

25

30

35%

Fts

Z in

pel

let

7.5

6.8

6.5

50 mM KCl 300 mM KCl

FtsZBs

FtsZEc

FtsZEc

FtsZBs

FtsZEc

7.5 6.8 6.5

GDP GDP GDPGTP GTP GTP

FtsZBs

S S S S S SP P P P P P

A

B

Figure 1. Quantification of FtsZ polymerization by sedimentation. (A) 12 μM FtsZ was polymerized

in the presence of 2 mM GTP/GDP at pH 7.5 (black bars), 6.8 (grey bars) or 6.5 (white bars). The

amount of protein pelleted was determined by densitometric analysis of Coomassie stained gels.

GDP served as a control for aspecific sedimentation and the percentage of FtsZ sedimented with

GDP was subtracted from the percentage of FtsZ sedimented with GTP to obtain the values plotted

in the graph. On the left: FtsZ sedimented at 50 mM KCl, on the right: FtsZ sedimented at 300 mM

KCl. (B) Representative results from Coomassie stained gels. Polymerization of FtsZBs

(upper gel)

and FtsZEc

(lower gel) at 50mM KCl. (S) supernatant, (P) pellet fractions from the experiment.


59

Figure 2. Sedimentation of SepF/FtsZ tubules at low speed. (A) 12 μM FtsZBs was polymerized

with 2 mM GDP (white bars) or GTP (black bars). The amount of protein pelleted was determined by

densitometric analysis of Coomassie stained gels. The + and – signs under the x-axis indicate the

presence or absence of FtsZ and SepF in the reaction. (B) Representative results from a Coomassie

stained gel. Polymerization of FtsZBs in the presence and absence of SepF. As a control SepF with-

out FtsZ was used. Polymerization was carried out with GTP and GDP. (S) supernatant, (P) pellet

fractions from the experiment.


60

SEDIMENTATION OF FTSZ-SEPF TUBULES

To analyze the interaction of FtsZ with certain activators sedimentation assays can be performed at lower centrifugation speeds. At this velocity only large structures of FtsZ may be pelleted, eg the large tubules formed by SepF rings and FtsZBs filaments 5, or the bundles formed by FtsZ and ZapA. We used lower centrifugation (24 600 ×g) to demonstrate the feasibility of this approach for the tubules formed by FtsZ and SepF. FtsZ was recovered in the pellet above back-ground levels only when both SepF and GTP were present in the sample (Fig. 2), and the presence of SepF does not influence FtsZ GTPase activity 7 showing that FtsZ is fully active in the presence of SepF. Specific sedimentation of SepF and FtsZ is roughly 45% of total SepF and 15% of total FtsZ (compared to material sedimenting when GDP is added). This shows that the SepF-FtsZ tubules con-tain more SepF than FtsZ. This may be because many SepF rings organize the FtsZ-SepF tubules 5,7. The exact stoichiometry of SepF-FtsZ in these tubules is not known but our results suggest that there is more SepF present than FtsZ.

FTSZEC

AND FTSZBS

POLYMERIZATION AND BUNDLING PROPERTIES

To characterize the polymerization efficiency of FtsZBs and FtsZEc in different buffers we analyzed both proteins by 90° angle light scattering. At 50 mM KCl, FtsZBs gives a 20-40-fold higher light scattering signal than FtsZEc depending on buffer pH (Fig. 3 A, B) confirming results of Buske et al 26. Increasing the KCl concentration in the buffer did not significantly influence the light scattering signal of FtsZEc (Fig. 3 D) but the signal of FtsZBs decreased ~80-fold at pH 7.5,

~30-fold at pH 6.8 and ~45-fold at pH 6.5 in 300 mM KCl (Fig. 3 C) compared to buffers with 50 mM KCl (Fig. 3 A). Disassembly of FtsZ polymers is faster at higher KCl concentration for both proteins (Fig. 3 C, D). Studies of Pacheco-

-Gómez et al. show that E. coli FtsZ polymerization and bundling is pH depen-dent. These authors found that in a buffer with 50 mM KCl the light scattering signal of FtsZ polymerization was higher, and disassembly of FtsZ took longer at pH 6.0 compared to pH 7.0 8. These results are not in agreement with our data from polymerization of FtsZEc at 50 mM KCl (Fig. 3 B), but it has to be noted that we have used three different buffers (HEPES, MES and PIPES) where Pacheco-


61

-Gómez et al. only used MES. Thus, not only the pH, but also buffer composition (ionic strength) affects the kinetics of FtsZ polymers at 50 mM KCl. However, at 300 mM KCl neither buffer composition nor pH influenced FtsZEc assembly in a detectable manner.

A light scattering experiment of FtsZBs in buffers without KCl was not possi-ble due to precipitation of the protein under these conditions. When the con-centration of FtsZBs was lowered to 3 µM, precipitation did not occur. However, 3 µM is not the physiological concentration of FtsZ in the cell. In plastic cu-vettes, FtsZBs did not precipitate at 12 µM at pH 7.5, but at pH 6.8 and 6.5 FtsZ still precipitated in the absence of KCl.

Figure 3. Light scattering of 12 μM FtsZEc

and 12 μM FtsZBs

. FtsZs were assembled in the presence

of 2 mM GTP and polymerization was monitored by 90° angle light scattering. Polymerization of

FtsZBs

(A) and FtsZEc

(B) at 50 mM KCl at pH 7.5, pH 6.8, pH 6.5. Polymerization of FtsZBs

(C) and

FtsZEc

(D) at 300 mM KCl at pH 7.5, pH 6.8 and pH 6.5.


62

Figure 4. Structures of FtsZBs

and FtsZEc

polymers visualized by electron microscopy. (A-L) Images

of 12 μM FtsZBs

(A-C and G-I) and FtsZEc

(D-F) and (J-L) polymerized with 2 mM GTP. (A-C) FtsZBs

in buffer with 50 mM KCl and pH 7.5, 6.8 and 6.5 respectively. (D-F) FtsZEc

in buffer with 50mM KCl

and pH 7.5, 6.8 and 6.5 respectively. (G-I) FtsZBs

in buffer with 300 mM KCl and pH 7.5, 6.8 and 6.5

respectively. (J-L) FtsZEc

in buffer with 300 mM KCl and pH 7.5, 6.8 and 6.5 respectively. Scale bar:

100nm.

DISCUSSION CHAPTER 2

63

MORPHOLOGIES OF FTSZ STRUCTURES FROM E. COLI AND B. SUBTILIS

The structures formed by FtsZ were inspected by TEM. FtsZBs assembled into closely compacted polymers that covered the whole grid in all buffers at low salt (Fig. 4A-C). FtsZEc formed long filaments, cables and bundles in all buffers at low salt (Fig 4D-F). However, the observable amount of polymers formed by FtsZEc was lower than the amount formed by FtsZBs. In high salt buffers FtsZBs formed longer single-stranded protofilaments which did not associate into bundles (Fig 4G-I). While FtsZBs protofilaments changed structure at higher salt concentration, FtsZEc formed structures indistinguishable from those of low salt buffer (Fig. 4J-L). There was no observable pH influence on polymerization of FtsZEc but FtsZBs forms more bundles at pH 6.5 which are visible as closely compacted polymers and sheets (Fig. 4 B). These results are in accord with our light scattering experiments and previously published TEM work 8,11,26.

THE GTPASE ACTIVITY OF FTSZ AT HIGH AND LOW KCL CONCEN-

TRATIONS

The GTP hydrolysis activity of FtsZ was measured under different conditions using a colorimetric assay for free phosphate. As reported previously 15 the GTPase activity of FtsZ increased with increasing KCl concentration: depend-ing on the buffer used FtsZBs showed a 3 to 7 fold increase, and FtsZEc showed a 1.5 to 2.5 fold increase in GTPase activity at 300 mM KCl compared to 50 mM KCl. The reduced GTPase activity at 50 mM KCl is due to bundling of FtsZBs fila-ments. At 50 mM KCl FtsZEc had a 3 to 6 fold higher GTPase activity than FtsZBs due to quicker disassembly of the FtsZEc polymers. The difference in GTP hy-drolysis activity between FtsZBs and FtsZEc was reduced at 300 mM KCl, possibly because of reduced bundling of FtsZBs filaments (Fig. 5).

DISCUSSION

We describe a set of methods that allows a quick analysis of FtsZ activity and its interaction with other proteins. Light scattering, sedimentation and GTPase


64 assays as well as electron microscopy have been widely used to study FtsZ polymerization. We have made some improvements to existing protocols, we showed the influence of different conditions on FtsZ assembly, and we pro-pose controls that should be included in FtsZ studies.

We introduce low speed centrifugation to distinguish large structures formed by the association between FtsZ and its interacting proteins from FtsZ polymers. This method shows two advantages over the standard sedimenta-tion assay. First, no background is formed by the FtsZ polymers in the pellet fraction as they are not spun down at 24 600 ×g. Second, the amount of FtsZ present in the structure formed with an interacting protein may be calculat-ed from the gel. Two critical steps in this method are the incubation time and the GTP concentration. It is important to centrifuge the large protein structure when it is complete but before it disassembles when all GTP is hydrolyzed. The best control for this study is polymerization of FtsZ with GDP. There is one po-tential limitation of the assay. FtsZ forms a stable complex with SepF which can easily be spun down at 24 600 ×g. If the sedimentation with another activa-tor or a drug that bundles FtsZ polymers is performed, it may be necessary to

Figure 5. GTP hydrolysis during FtsZ polymerization in 6 different buffers. In all experiments

2 mM GTP was used. As a control sample with no MgCl2 was used. Activity of FtsZ without MgCl

2

was subtracted from the activity of FtsZ in the presence of MgCl2. On the left: GTPase activity of

FtsZ at 50 mM KCl, on the right: GTPase activity of FtsZ at 300 mM KCl.


65

adapt the assay. It may be done by changing the incubation time, or increasing the speed of centrifugation.

Proper preparation of the sample is the most important for light scattering experiments. Proteins must be precleared by spinning and all the buffers should be filtered prior to use. If any aggregates are present in the sample, they will disrupt a stable signal obtained from FtsZ polymers. For the analysis of the FtsZ structures by electron microscopy, preparation of a grid is the main step. The time of sample incubation on the grid will have the effect of producing more or less compacted polymers. For bundles of FtsZBs, the time of incubation must be shorter than for FtsZEc and FtsZBs at high KCl concentration. We used a concen-tration of 12 µM for every sample to be able to compare the results. However, for FtsZBs at 50 mM KCl a lower FtsZ concentration should be used, as 12 µM resulted in a full saturation of the grid. This makes the polymers highly compacted and difficult to detect. Less compacted polymers are better to detect on EM.

The GTPase assay is the only experiment used to study the activity rather than the structures of FtsZ. Mg2+ is necessary for GTP turnover in FtsZ polymers. Thus, in the absence of Mg2+, FtsZ does not hydrolyze GTP. Therefore, a sample with no Mg2+ is the right control in this assay but cations of Mg are present in the FtsZ storage buffer. They may be removed by addition of 1 mM EDTA to the control sample. The critical step in this assay is the incubation time. It is import-ant to stop FtsZ activity after a given time. This is achieved by transferring the FtsZ sample to a malachite green solution in a 96 well plate. However, develop-ment of the malachite green color is a continuous process. Thus the measure-ments must be taken at the same time for every sample. Using a well-planned GTP addition protocol with measurements taken each 30 sec apart in an estab-lished order, it is possible to obtain the same incubation and sample handling time for every time point. Another critical step is choosing the concentration of the protein for the experiment. In the experiment we used two different concentrations for FtsZEc and FtsZBs. GTP hydrolysis is much quicker for FtsZEc compared to FtsZBs. The GTPase activity of FtsZEc under chosen conditions and at 12 µM is linear only for max. 5 min and after that time the hydrolysis rate pla-teaus. Thus, it is difficult to interpret data from the experiment when performed under these conditions. In this case FtsZEc must be used at lower concentration than FtsZBs to be able to compare activities of both proteins. The GTPase activ-


66

ity of FtsZs from different sources may vary. Thus, the right concentration must be chosen. The concentration for FtsZ polymerization should be well above the critical concentration (in general from 2.5 to 10 µM). The dynamics of FtsZ assembly and disassembly is also important. Some proteins show a significant lag in polymerization after addition of GTP, as shown for FtsZBs at 50 mM KCl. It is useful to perform the light scattering assay before the GTPase assay to ap-proximate the time of assembly and disassembly of FtsZ polymers. After that, the time of incubation and concentration of protein may be chosen. Since the conditions chosen for FtsZ polymerization are crucial, it is important to use the right pH and KCl concentration in each method. In this work we studied 9 dif-ferent buffers with pH ranging from 6.5 to 7.5 and KCl concentrations from 0M to 300mM. We noticed that the best condition to analyze FtsZs from B. subtilis and E.coli and their biological activity is at pH that is close to physiological level (7.5) together with a high KCl concentration. At a high KCl concentration, FtsZ has a higher GTPase activity and produces polymers that are better detect-able by electron microscopy. We also confirmed that the physiological pH and a high KCl concentration are better for the study of the interaction between FtsZ and regulatory proteins than any other buffers mostly used to study FtsZ assembly. FtsZBs shows a similar activity to FtsZEc when studied at high KCl con-centration. In addition, at low salt concentration the influence of pH is more visible than in the buffers with high salt concentration. FtsZ sometimes precipi-tates when using buffers without KCl, as a result, buffers without salt should be avoided. Sedimentation of FtsZ polymers is low when using buffers with high KCl concentrations. This may be an advantage when studying interactions be-tween FtsZ and proteins that assemble FtsZ filaments such as SepF and ZapA as these higher order structures are easy to detect with centrifugation. In all our experiments we used MgCl2 at a 10 mM concentration. It was shown that a relatively high Mg2+ concentration stabilizes FtsZ polymers and reduces the GTPase activity of FtsZ. In Table 2 results from various studies are summarized describing FtsZ polymerization and GTPase activity at different Mg2+ concen-trations using otherwise identical buffer conditions 27. The measured concen-tration of free cytoplasmic Mg2+ is 0.9 mM 3. It should be noticed that GTP will chelate an equivalent amount of Mg2+. Thus, the optimal Mg2+ concentration for GTPase experiments is around 2-2.5 mM, which is close to physiological

ACKNOWLEDGMENTS CHAPTER 2

67

level 3. However, in our experiments we used MgCl2 at a 10 mM concentration to obtain an easily detectable light scattering signal and to stabilize FtsZ poly-mers during the sedimentation assay.

Although we applied our protocols to FtsZ from the model organisms E. coli and B. subtilis, they can be adapted to FtsZ from any other organism. It has to be noted that the physiological pH, and concentrations of monovalent, and divalent cations differ among organisms. Thus, the optimal conditions for FtsZ polymerization may vary. Differences in doubling time and growth conditions of different bacteria may result in different assembly kinetics of FtsZ and op-timal conditions of the experiments. However, our protocol provides a good starting point for the experiments with FtsZs from other organisms. The proto-cols should be useful for the study of FtsZ with regulatory proteins or the study of effects of small compounds and drugs on FtsZ.

Table 2. Effect on Mg2+ on FtsZ polymerization and GTPase. Results from this work compared to

published data. All experiments were carried out in 50 mM MES/ NaOH, pH=6.5, 50 mM KCl.

Source Polymerization [% of FtsZ sedimented]

GTPase [Pi/FtsZ/min]

Mg2+ concen-tration [mM]

FtsZ concentra-tion [µM]

References

FtsZEc ~28% ~2.1 10 12 This work

~50% ~2.4 10 12.5 27

~43% ~3.5 5 12.5 27

~27% ~4.6 2.5 12.5 27

ND ~5.4 2.5 5 26

FtsZBs ~30% ~0.8 12 This work

~52% (with DEAE dextran) ~0.5 10 10 11

ND ~2.25 2.5 5 26

ACKNOWLEDGMENTS

Work in our laboratory is funded by a VIDI grant from the Netherlands Organi-sation for Scientific research (to DJS).

We thank Marc Stuart and the Department of Electron Microscopy at our university, for assistance with and providing access to the transmission elec-tron microscope.


68

REFERENCES

1. Haydon, D. J., Stokes, N. R. et al. An in-hibitor of FtsZ with potent and selective anti-staphylococcal activity. Science 321 (5896), 1673-1675 (2008).

2. Adams, D. W. & Errington, J. Bacterial cell division: assembly, maintenance and disassembly of the Z ring. Nat. Rev. Mi-crobiol. 7 (9), 642-653 (2009).

3. Erickson, H. P., Anderson, D. E. & Osawa, M. FtsZ in bacterial cytokinesis: cyto-skeleton and force generator all in one. Microbiol. Mol. Biol. Rev. 74 (4), 504-528 (2010).

4. Li, Z., Trimble, M. J., Brun, Y. V. & Jensen, G. J. The structure of FtsZ filaments in vivo suggests a force-generating role in cell division. EMBO J. 26 (22), 4694-4708 (2007).

5. Król, E., van Kessel, S. P., van Bezouwen, L. S., Kumar, N., Boekema, E. J. & Scheffers, D. J. Bacillus subtilis SepF binds to the C-terminus of FtsZ. PLoS One 7 (8), e43293 (2012).

6. Singh, J. K., Makde, R. D., Kumar, V. & Panda, D. SepF increases the assembly and bundling of FtsZ polymers and sta-bilizes FtsZ protofilaments by binding along its length. J. Biol. Chem. 283 (45), 31116-31124 (2008).

7. Gündoğdu, M. E., Kawai, Y. et al. Large ring polymers align FtsZ polymers for normal septum formation. EMBO J. 30 (3), 617-626 (2011).

8. Pacheco-Gomez, R., Roper, D. I., Dafforn, T. R. & Rodger, A. The pH dependence of polymerization and bundling by the essential bacterial cytoskeletal protein FtsZ. PLoS One 6 (6), e19369 (2011).

9. Mendieta, J., Rico, A. I., Lopez-Vinas, E., Vicente, M., Mingorance, J. & Gomez-

-Puertas, P. Structural and functional model for ionic (K(+)/Na(+)) and pH dependence of GTPase activity and

polymerization of FtsZ, the prokaryotic ortholog of tubulin. J. Mol. Biol. 390 (1), 17-25 (2009).

10. Tadros, M., Gonzalez, J. M., Rivas, G., Vicente, M. & Mingorance, J. Activation of the Escherichia coli cell division pro-tein FtsZ by a low-affinity interaction with monovalent cations. FEBS Lett. 580 (20), 4941-4946 (2006).

11. Scheffers, D. J. The effect of MinC on FtsZ polymerization is pH dependent and can be counteracted by ZapA. FEBS Lett. 582 (17), 2601-2608 (2008).

12. Mukherjee, A. & Lutkenhaus, J. Analysis of FtsZ assembly by light scattering and determination of the role of divalent metal cations. J. Bacteriol. 181 (3), 823-832 (1999).

13. Chen, Y. & Erickson, H. P. Rapid in vitro as-sembly dynamics and subunit turnover of FtsZ demonstrated by fluorescence resonance energy transfer. J. Biol. Chem. 280 (23), 22549-22554 (2005).

14. Hou, S., Wieczorek, S. A. et al. Character-ization of Caulobacter crescentus FtsZ protein using dynamic light scattering. J. Biol. Chem. 287 (28), 23878-23886 (2012).

15. Mukherjee, A. & Lutkenhaus, J. Dynamic assembly of FtsZ regulated by GTP hy-drolysis. EMBO J. 17 (2), 462-469 (1998).

16. Lanzetta, P. A., Alvarez, L. J., Reinach, P. S. & Candia, O. A. An improved assay for nanomole amounts of inorganic phos-phate. Anal. Biochem. 100 (1), 95-97 (1979).

17. Ray, S., Kumar, A. & Panda, D. GTP regu-lates the interaction between MciZ and FtsZ: a possible role of MciZ in bacterial cell division. Biochemistry 52 (2), 392-401 (2013).

18. Rivas, G., Lopez, A. et al. Magnesium- -induced linear self-association of the

ACKNOWLEDGMENTS CHAPTER 2

69

FtsZ bacterial cell division protein monomer. The primary steps for FtsZ assembly. J. Biol. Chem. 275 (16), 11740-11749 (2000).

19. Oliva, M. A., Cordell, S. C. & Lowe, J. Struc-tural insights into FtsZ protofilament formation. Nat. Struct. Mol. Biol. 11 (12), 1243-1250 (2004).

20. Lowe, J. & Amos, L. A. Crystal structure of the bacterial cell-division protein FtsZ. Nature 391 (6663), 203-206 (1998).

21. Cordell, S. C., Robinson, E. J. & Lowe, J. Crystal structure of the SOS cell divi-sion inhibitor SulA and in complex with FtsZ. Proc. Natl. Acad. Sci. U. S. A. 100 (13), 7889-7894 (2003).

22. Chen, Y., Anderson, D. E., Rajagopalan, M. & Erickson, H. P. Assembly dynamics of Mycobacterium tuberculosis FtsZ. J. Biol.Chem. 282 (38), 27736-27743 (2007).

23. White, E. L., Ross, L. J., Reynolds, R. C., Seitz, L. E., Moore, G. D. & Borhani, D. W. Slow polymerization of Mycobacterium tuberculosis FtsZ. J. Bacteriol. 182 (14), 4028-4034 (2000).

24. Oliva, M. A., Trambaiolo, D. & Lowe, J. Structural insights into the conforma-tional variability of FtsZ. J. Mol. Biol. 373 (5), 1229-1242 (2007).

25. Thanbichler, M. & Shapiro, L. MipZ, a spa-tial regulator coordinating chromosome segregation with cell division in Caulo-bacter. Cell 126 (1), 147-162 (2006).

26. Buske, P. J. & Levin, P. A. Extreme C termi-nus of bacterial cytoskeletal protein FtsZ plays fundamental role in assembly inde-pendent of modulatory proteins. J. Biol. Chem. 287 (14), 10945-10957 (2012).

27. Mukherjee, A. & Lutkenhaus, J. Analysis of FtsZ assembly by light scattering and determination of the role of divalent metal cations. J. Bacteriol. 181 (3), 823-832 (1999).

Bacillus subtilis SepF binds to the

C-terminus of FtsZEwa Król1,a, Sebastiaan van Kessel1,a, Laura S. van

Bezouwenb, Neeraj Kumara, Egbert J. Boekemab

and Dirk-Jan Scheffers*,a

aDepartment of Molecular Microbiology, Groningen Biomolecular

Sciences and Biotechnology Institute, University of Groningen,

Nijenborgh 7, 9747 AG Groningen, the Netherlands.bDepartment of Electron Microscopy, Groningen Biomolecular

Sciences and Biotechnology Institute, University of Groningen,

Nijenborgh 7, 9747 AG Groningen, the Netherlands.

1 equal contribution

Published in: PLoS One. 2012;7(8):e43293

Experiments presented in Figures 2B, 3G, 3H were performed by E. Cendrowicz. Experiments presented in Figures 4A and 4B were performed by E. Cendrowicz and S. van Kessel. Experiments presented in Figures 3A-3F were performed by E. Cendrowicz and LS. Van Bezouwen.

ABSTRACT

Bacterial cell division is mediated by a multi-protein machine known as the “di-visome”, which assembles at the site of cell division. Formation of the divisome starts with the polymerization of the tubulin-like protein FtsZ into a ring, the Z-ring. Z-ring formation is under tight control to ensure bacteria divide at the right time and place. Several proteins bind to the Z-ring to mediate its mem-brane association and persistence throughout the division process. A con-served stretch of amino acids at the C-terminus of FtsZ appears to be involved in many interactions with other proteins. Here, we describe a novel pull-down assay to look for binding partners of the FtsZ C-terminus, using a HaloTag af-finity tag fused to the C-terminal 69 amino acids of B. subtilis FtsZ. Using lysates of Escherichia coli overexpressing several B. subtilis cell division proteins as prey we show that the FtsZ C-terminus specifically pulls down SepF, but not EzrA or MinC, and that the interaction depends on a conserved 16 amino acid stretch at the extreme C-terminus. In a reverse pull-down SepF binds to full-length FtsZ but not to a FtsZΔC16 truncate or FtsZ with a mutation of a conserved proline in the C-terminus. We show that the FtsZ C-terminus is required for the formation of tubules from FtsZ polymers by SepF rings. An alanine-scan of the conserved 16 amino acid stretch shows that many mutations affect SepF bind-ing. Combined with the observation that SepF also interacts with the C-termi-nus of E. coli FtsZ, which is not an in vivo binding partner, we propose that the secondary and tertiary structure of the FtsZ C-terminus, rather than specific amino acids, are recognized by SepF.


73

INTRODUCTION

Bacterial cell division is mediated by a multi-protein machine known as the “divisome” that assembles into a ring at the site of cell division [1]. Formation of the divisome starts with the polymerization of the tubulin-like protein FtsZ into a ring, the Z-ring, just below the membrane at the cell division site. Be-cause of its critical role in division, Z-ring formation is under tight control to ensure bacteria divide at the right time and place. Once formed, the Z-ring functions as a scaffold to which all other division proteins are recruited, and several proteins bind to the Z-ring to mediate its membrane association and persistence throughout the division process [2].

Placement of the Z-ring is mediated by two systems, that prevent cells from dividing at the poles or immediately after division is completed (Min), or from dividing through nucleoids (nucleoid occlusion). The MinC component of Min directly acts on FtsZ [3,4], whereas nucleoid occlusion is mediated by the FtsZ-binding protein SlmA in E. coli [5,6] and the non-homologous protein Noc in B. subtilis, which does not seem to act directly on FtsZ [7]. In B. subtilis, Z-ring formation is also controlled by the nutritional status of the cell, through the nutritional sensor UgtP that acts directly on FtsZ [8]. Another level of control is provided by ClpX. In B. subtilis ClpX destabilizes, but not degrades FtsZ poly-mers, whereas in E. coli ClpX degrades (preferentially polymeric) FtsZ and thus controls the amount of FtsZ available for Z-ring formation [9,10].

The Z-ring is tethered to the membrane by FtsA. In E. coli, ZipA forms an additional membrane tethering factor [11]. A protein with a similar topology to ZipA, EzrA, acts as a destabilizer of Z-rings in B. subtilis [12]. Z-rings are further stabilized by ZapA [13], a protein that strengthens lateral associations between FtsZ filaments, and in a small subset of bacteria including E. coli, ZapC fulfills a similar function [14,15]. Other proteins that interact directly with FtsZ have a specific function in division. FtsE functions as part of the FtsEX complex that controls hydrolysis of the cell wall in dividing E. coli [16,17]. In B. subtilis, SepF is a ring-forming protein that binds FtsZ and plays a role in correct assembly of the cross-wall septum, presumably by bundling FtsZ polymers into ~50 nm wide tubes that can be observed in vitro[18]. During sporulation, SpoIIE inter-acts with FtsZ to ensure proper formation of the asymmetric septum [19] and


74

MciZ, a 40 amino acid protein, prevents aberrant FtsZ ring formation in the mother cell compartment [20].

Even though crystal structures of FtsZ and various interacting proteins exist, relatively little is known about the precise interactions between FtsZ and these proteins. However, it is increasingly evident that a conserved stretch of amino acids at the C-terminus of FtsZ is involved in many of these interactions. The high degree of conservation of this C-terminal peptide was identified by Ma and Margolin [21]. The C-terminal peptide consists of a highly conserved core sequence followed by a variable number of poorly conserved residues at the extreme C-terminus. This region is not resolved in the available FtsZ structures, as it is preceded by a non-conserved, flexible linker sequence of variable length. The FtsZ-binding domain of ZipA has been co-crystallized together with a pep-tide matching the FtsZ C-terminus, which showed an extended β-strand fol-lowed by an α-helix [22]. In addition to ZipA, evidence has been reported for the interaction of the conserved FtsZ C-terminus with MinC [23], FtsA [24], EzrA [25], ClpX [9], SepF [26] and FtsZ itself [27]. Evidence for the interaction of FtsA with the C-terminus is based on yeast two-hybrid data [24]. For MinC, mutations in the FtsZ C-terminus conserved core have been described that abolish the inter-action with MinC both in vivo and in vitro [23]. Using a C-terminal truncation of FtsZ (FtsZEcΔ18) that is still capable of polymerization, it was shown that ClpX is far less active in degradation of FtsZEcΔ18 than of wt FtsZ [9]. All these studies were done in E. coli. In Mycobacterium tuberculosis, the FtsZ C-terminus mediates the interaction between FtsZ and FtsW [28]. EzrA and SepF are Gram-positive division proteins that modulate FtsZ. The effects of EzrA or SepF on FtsZ polym-erization can be relieved by adding competing amounts of a synthetic peptide encoding the FtsZ C-terminus, and EzrA and SepF have no effects on a C-termi-nal truncation mutant of FtsZ that still polymerizes (FtsZBsΔ16) [25,26]. The SepF study showed that SepF bundles FtsZ filaments and co-sediments with FtsZ in polymerization experiments. The FtsZBsΔ16 truncate still polymerized but no longer bundled in the presence of SepF, yet SepF still co-sedimented with FtsZBsΔ16, leading the authors to suggest that SepF also binds to a secondary site on FtsZ next to the C-terminus [26]. Recently the poorly conserved residues at the extreme C-terminus have been implicated in lateral interactions between FtsZ polymers that are potentially mediated by electrostatic forces [27].

RESULTS CHAPTER 3

75

The crystal structure of the FtsZ-binding domain of ZipA bound to a peptide matching the FtsZ C-terminus [22], and the recent crystal and NMR structures on the interaction of Thermotoga maritima FtsA (FtsATm) and a FtsZTm C-terminal peptide [29] provide more insight in how the C-terminus is bound by proteins. Bound to ZipA, the C-terminal peptide assumes the confirmation of an extend-ed β-strand followed by an α-helix, whereas bound to FtsA the peptide is pre-dominantly helical, but not as a single helix. Six residues in the peptide interact with ZipA, mostly through hydrophobic contacts but also through two hydro-gen bonds between the backbones of ZipA and the FtsZ-peptide, formed by a conserved Asp and Leu residues (D367 and L369 in B. subtilis FtsZ) [22]. The interactions with FtsA are quite different, with three salt bridges connecting the peptide to FtsA. The salt bridges are formed by a conserved Asp and Arg residue (D370 and L376 in B. subtilis FtsZ) and the carboxyl group of the C-ter-minal amino acid residue [29]. Based on a comparison of both structures it has been suggested that the FtsZ C-terminus can adopt different conformations to fit different binding pockets [29].

In this study, we describe a pull-down strategy to identify interaction part-ners of the B. subtilis FtsZ C-terminus. Using this strategy we identify SepF as a binding partner of the C-terminus, confirming an earlier observation of Singh et al. [26]. We extend the earlier results by showing that SepF only binds to the C-terminus and by identifying residues critical for binding of SepF by an ala-nine-scan of the FtsZ C-terminus.

RESULTS

SEPF BINDS TO THE FTSZ C-TERMINUS

To screen for B. subtilis proteins that interact with the FtsZ C-terminus we fused the last 69 amino-acid residues of B. subtilis FtsZ to a HaloTag (Halo-ZC, Fig. 1A). The HaloTag consists of a modified dehalogenase that can covalently bind to a chloroalkane ligand coupled to sepharose beads (HaloLink resin) [30]. As a control we fused only the flexible, non-conserved linker sequence of FtsZ without the final 16 amino acid residues to a HaloTag (Halo-ZΔC16). The fusion


76

proteins were expressed in E. coli and bound to the Halolink resin in binding buffer. Aliquots of the resin containing the covalently attached bait protein were subsequently used for pull-down experiments.

To validate the pull-down approach we analyzed the binding of three can-didate interacting proteins after overexpression in E. coli. EzrA and SepF have

Figure 1. MBP-SepF binds to the FtsZ C-terminus. (A) Schematic representation of the Halo-Tagged

constructs used for the pull-down experiments. Halo-ZC consists of the HaloTag followed by a TEV

protease cleavage site fused to the final 69 amino acids of B. subtilis FtsZ, including the 16 ami-

no acid residues of the conserved C-terminal tail. Halo-ZΔC16 contains only the 53 amino acid

residues of the FtsZ linker-domain but lacks the conserved C-terminal tail and Halo-ZAla

is rep-

resentative of the constructs used for the alanine scanning of the conserved C-terminal tail. (B)

Pull down experiments using Halo-Zc or Halo ZΔC16 bound to resin. Pull downs were performed

on lysates from E. coli cells overexpressing His-EzrAcyt

(top panel), MinC-strep (middle panel) and

MBP-SepF (lower panel). The input material (Input), flow-through (FT) and eluate (E) fractions of the

experiments are shown and relevant proteins are indicated on the left. In the top and lower panel

a fragment of the gel that did not contain the bait proteins or the protein of interest was deleted

(white bar) to reduce the size of the image.

RESULTS CHAPTER 3

77

been reported to interact with the FtsZ C-terminus in B. subtilis, and MinC in-teracts with the FtsZ C-terminus in E. coli [23,25,26]. We expressed B. subtilis MinC fused to a strep-tag (MinC-strep), the cytosolic domain of EzrA fused to a His-tag (His-EzrAcyt), and SepF fused to Maltose-Binding protein (MBP-SepF) in E. coli cells, prepared cell lysates and incubated the lysates with aliquots of resin with attached bait protein. After extensive washing, the bait protein and any interacting proteins bound to it were eluted from the resin by TEV protease cleavage. We found that MBP-SepF was retained by Halo-ZC bound to resin, and very weakly by Halo-ZΔC16 bound to resin (Fig. 1B). This confirms an earli-er report that suggested that SepF binds to the FtsZ C-terminus [26]. His-EzrAcyt

and MinC-strep were not retained by either construct, also not when we tested for the presence of prey protein using specific antibodies against the His- or strep-tag (not shown). Previously, we had not been able to establish a direct interaction between B. subtilis MinC and FtsZ [31].

THE INTERACTION BETWEEN THE FTSZ C-TERMINUS AND SEPF IS

SPECIFIC

To study the interaction between SepF and the FtsZ C-terminus in more de-tail, and to exclude that the interaction is mediated by potential ‘bridging’ pro-teins present in the E. coli lysate, we repeated the pull-down experiments with purified MBP-SepF and SepF. As an extra control, we used a bait construct in which a conserved proline residue in the C-terminus was mutated into ala-nine, Halo-ZCP372A. Pro372 (notation for full length FtsZ) is fully conserved in the FtsZ C-terminus across species and we reasoned that a mutation of this residue would abolish the capability of the C-terminus to interact with SepF. Both pure MBP-SepF and SepF were retained when Halo-ZC was used as bait protein. When Halo-ZΔC16 or Halo-ZCP372A were used as bait a very faint band of MBP-SepF was visible in the eluate, but untagged SepF was not retained (Fig. 2A). This confirms that the interaction between SepF and the FtsZ C-ter-minus is specific, and that the pull-down is not caused by aspecific binding of the MBP-moiety to the C-terminus. The result also shows that the conserved proline residue indeed is critical for the function of the FtsZ C-terminus. We noticed that the ZCP372A bait protein is visible in the eluate as two bands on


78

gel. The protein is initially expressed as full-length protein, during elution the ZCP372A bait is released from the Halo-tag by TEV cleavage. The double band pat-tern can be attributed to a change in the electrophoretic pattern of the mutat-

Figure 2. Specificity of the interaction between SepF and the FtsZ C-terminus. (A) Pull down ex-

periments using Halo-ZC, Halo ZΔC16 or Halo-ZCP372A

bound to resin, with MBP-SepF (upper panel)

or wild type SepF (lower panel) as prey proteins. The input material (Input), flow-through (FT) and

eluate (E) fractions of the experiments are shown and relevant proteins are indicated on the left.

MBP-SepF tends to form dimers as can be observed from the extra band in the input and flow-

through samples. (B) Reverse pull down experiment with MBP-SepF bound to resin and purified

FtsZ, FtsZP372A

or FtsZΔC16 used as prey proteins. The input material (I), flow-through (FT), last wash

(W) and eluate (E) fractions of the experiments are shown and relevant proteins are indicated on

the left. M: molecular weight marker, bands are 70, 55 and 40 kDa. Upper panel: Coomassie stained

gel (CBB), lower panel: western blot developed with anti-FtsZ antibody. (C) Pull down experiments

using Halo-Zc bound to resin, with MBP-SepF, MBP-SepFA98V

and MBP-SepFF124S

as prey proteins.

The input material (Input), flow-through (FT) and eluate (E) fractions of the experiments are shown

and relevant proteins are indicated on the left.

RESULTS CHAPTER 3

79

ed bait protein which is not completely resolved by the sample buffer resulting in two bands. Another, less likely explanation is that the alanine substitution gives rise to an extra TEV cleavage site. Other alanine substitutions in the C-ter-minus also give rise to double bands for the bait protein, and in several cases this change did not affect SepF binding (see Fig. 4).

MBP-SepF bound to amylose resin can be used to pull down FtsZ from E. coli lysates expressing B. subtilis FtsZ [18]. This allowed us to perform the reverse pull-down experiment, and we incubated MBP-SepF bound to resin with purified FtsZ, FtsZΔC16 and FtsZP372A. Importantly, in this experiment full-length FtsZ was used, so if there is a secondary binding site for SepF on FtsZ, one could expect that all three prey proteins are pulled down by MBP-SepF. We found that wild type FtsZ was bound by MBP-SepF and could be detect-ed in the eluted fraction both by Coomassie staining and by western blotting (Fig. 2B). FtsZΔC16 and FtsZP372A were not visibly co-eluted with MBP-SepF with Coomassie, and only a very faint band was detected for co-eluted FtsZP372A. This result suggests that the FtsZ C-terminus is either the only site of interaction with SepF, or that the binding of SepF to the secondary binding site is weak.

Two mutations in SepF, SepFA98V and SepFF124S have been described that abolish SepF-FtsZ interaction [18]. We purified MBP-SepFA98V and MBP-SepFF124S and used these proteins as prey in a Halo-ZC pull-down. Neither SepF mutant was found to bind to the FtsZ C-terminus (Fig. 2C). This suggests that the mu-tations in SepFA98V and SepFF124S abolish binding of SepF to the FtsZ C-terminus rather than to a secondary binding site on FtsZ. It is also of note that whereas MBP-SepF runs as monomers and dimers on SDS-PAGE, the two mutants do not, confirming the observation that both mutants, most notably SepFA98V, are less prone to form large oligomeric structures[18].

Taken together, these results show that SepF interacts with the FtsZ C-ter-minus, that this interaction is specific, and that we cannot find evidence for a secondary interaction site on FtsZ using pull-down assays.

SEPF CANNOT TUBULATE FTSZΔC16 AND FTSZP372A

POLYMERS

In vitro, SepF rings and FtsZ polymers combine to form large tubular struc-tures[18]. Our pull-down experiments show that FtsZΔC16 and FtsZP372A have


80

lost the ability to bind to SepF, but we cannot exclude that a weak resid-ual interaction between the proteins (e.g. through a secondary interaction site on FtsZ) still allows the formation of tubules. Therefore, we mixed FtsZ, FtsZP372A and FtsZΔC16 with SepF and GTP, to induce FtsZ polymerization and possible tubule formation, and examined the samples using electron micros-

Figure 3. SepF cannot tubulate FtsZΔC16 and FtsZP372A

polymers. (A-F) Electron microscopy of

10 μM FtsZ (A, D), FtsZ P372A

(B,E) and FtsZΔC16 (C,F) polymerized with 2 mM GTP in the absence

(A-C) or presence (D-F) of 6 μM SepF. Scale bar: 100 nm. (G) Pelleting of FtsZ-SepF tubules by

centrifugation. 10 μM FtsZ (1, 2) or FtsZ P372A

(3,4) was polymerized with 2 mM GTP in the absence

(1,3) or presence (2,4) of 12 μM SepF. As a control, SepF was used without FtsZ (5). Tubules were

recovered by centrifugation and supernatant and pellet fractions were analyzed by SDS-PAGE. (H)

Quantification of the amount of FtsZ pelleted in the assay shown in (G). Three independent exper-

iments were analyzed, error bars indicate standard deviation.

RESULTS CHAPTER 3

81

copy. FtsZ formed tubules with SepF (Fig. 3D), but in the case of FtsZP372A and FtsZΔC16 we could detect both FtsZ polymers and SepF rings (Fig. 3E,F), but not tubules, despite careful examination of the sample grids. We noticed that under our experimental conditions, FtsZP372A readily polymerized, but that FtsZΔC16 only formed very short polymers and aggregates (Fig. 3A-C), unlike the single stranded filaments that were reported for a FtsZΔC17 mutant [27]. We attribute this difference to the buffer system used in our study, with an increased pH and salt concentration that has previously been reported to de-crease the amount of FtsZ polymerized in centrifugation and light scattering assays [18].

As a second assay for tubule formation we used sedimentation, at a veloc-ity whereby SepF-FtsZ tubules but not FtsZ polymers are recovered in the pel-let. FtsZ and FtsZP372A were polymerized with GTP in the presence and absence of SepF. FtsZΔC16 was not used in this assay because of its tendency to ag-gregate. FtsZ was recovered in the pellet above background levels only when SepF was present, whereas FtsZP372A was not pelleted with SepF (Fig. 3G, H). SepF sedimentation is dependent on the presence of FtsZ, and does not occur with FtsZP372A, or when SepF is incubated alone (Fig. 3G). Combined with the pull-down experiments above this provides evidence that the FtsZ C-terminus is the only site of interaction with SepF.

THE SEPF BINDING SITE ARISES FROM THE FOLDING OF THE FTSZ

C-TERMINUS

The Halo-ZC construct specifically pulls down SepF, and this interaction is lost upon deletion of the final 16 residues from the C-terminus or by mutating the conserved Pro372 to Ala. To learn which other residues in the conserved C-terminus are critical for SepF binding, we performed an alanine-scan of the C-terminus. Each of the last 16 residues of the C-terminus was mutated to Ala (Fig. 1A), resulting in a series of HaloZCAla constructs which were used to pull down purified SepF (Fig. 4A). We determined the intensity of Coomassie-

-stained bands from the eluted C-terminus and the co-eluted SepF, and plotted the ratio of the densities as an indication for binding, using Halo-ZC and Halo-

-ZΔC16 as references (Fig. 4B). The use of ratios of bound SepF to eluted bait


82

protein means that we have corrected for possible fluctuations in the amount of the bait protein bound to and/or eluted from the beads. Changing the ful-ly conserved Pro372 and Phe374 residues abolished SepF binding. Changing Leu369, Leu375, Arg376 and Asn377 reduced SepF binding to less than 25% of the binding observed with the wild type. Changing Asp370, Ile371, Arg378, Lys380 and Arg381 also notably affected SepF binding, whereas changing Asp367, Thr368, Thr373, Asn379 and Gly382 resulted in SepF recovery that was not noticeably different from wild type. The fact that 11 out of 16 mutations inhibited SepF binding to various extent, suggests that the recognition of the FtsZ C-terminus by SepF depends more on the folding of the C-terminus than on specific interactions between amino acids in the form of salt bridges.

Figure 4. Alanine scanning of the FtsZ C-terminus. (A) Series of three gels from a representative

pull down experiment using Halo-ZC (wt), Halo ZΔC16 (Δ16) and a series of HaloZCAla

constructs

(mutated residue indicated). For each pull down the flow-through (left side) and eluate fractions

(right side) were loaded on the gel. The last lane is a loading control containing an equivalent

amount of input SepF material. (B) Ratio of eluted SepF over eluted bait protein as determined by

densitometric scanning of gels as shown in (A). Three independent experiments were analyzed,

error bars indicate standard deviation.


83The E. coli FtsZ C-terminus has a conserved part that is very similar to that of B. subtilis FtsZ, yet as E. coli does not have SepF there is no need for FtsZEc to bind SepF. We reasoned that if the overall conformation of the C-terminus is recognized by SepF, then SepF might recognize the E. coli FtsZ C-terminus as well. We constructed Halo-ZCEc, Halo-ZEcΔC14 and Halo-ZCEcP375A similar to our B. subtilis constructs and performed a pull-down with purified SepF (Fig. 5). SepF was bound to the Halo-ZCEc construct but not by Halo-ZEcΔC14 or Halo-ZCEcP375A. This result shows that SepF is capable of recognizing the C-ter-minus of a homologous FtsZ molecule, although this homologous FtsZ is not a binding partner for SepF in vivo.

DISCUSSION

In this paper we describe a novel pull-down assay to detect protein-protein interactions of the FtsZ C-terminus. The assay is based on the HaloTag fusion technology in which the bait protein is covalently bound to a resin allowing

Figure 5. SepF binds to the C-terminus of E. coli FtsZ. (A) Alignment of the C-termini of B. subtilis

(Bs) and E. coli (Ec) FtsZ. Identical residues are marked with an asterisk. (B) Pull down experiments

using Halo-ZCEc, Halo-ZCEcP372A

or Halo-ZEcΔC14 bound to resin, with wild type SepF as prey pro-

tein. The input material (Input), flow-through (FT) and eluate (E) fractions of the experiment are

shown and relevant proteins are indicated on the left.


84

stringent wash steps before applying a sample to the resin [30]. Our fusion constructs consist of a modified dehalogenase fused to a part of FtsZ contain-ing both the long flexible, non-conserved linker sequence and the conserved FtsZ C-terminus. The presence of the linker sequence has a dual function, it provides flexibility and distances the C-terminus from the dehalogenase, and it provides a control for unspecific binding through the FtsZΔC constructs. Using our assay we can specifically pull down FtsZ binding partners.

To validate our pull-down strategy we performed pull-downs with the C-ter-minus of B. subtilis FtsZ on lysates of E. coli cultures expressing three B. subtilis proteins that have previously been reported to bind to the FtsZ C-terminus, EzrA [25], SepF [26] and MinC. It has to be noted that evidence for MinC bind-ing to the FtsZ C-terminus comes from the homologous E. coli proteins [23]. Al-though binding for each protein could be expected, we only managed to pull down SepF. Although genetic evidence suggested that MinCBs needs MinDBs to act on FtsZBs [32] our earlier work showed that MinCBs alone destabilizes bundles of FtsZBs polymers and increases the light scattering signal of non-po-lymerized FtsZBs in solution, indicating that FtsZBs and MinCBs interact without MinDBs [4]. However, we had not been able to detect interactions between full length MinCBs and FtsZBs through a cross-linking approach [31], whereas MinCEc cosediments with FtsZEc polymers [3] and polymerized FtsZEc can be recruit-ed to MinCEc/MinDEc coated vesicles in a FtsZEc-C-terminus dependent manner [23]. This suggests that the interaction between B. subtilis FtsZ and MinC is not as strong as between the E. coli proteins. We do not know why we have not been able to pull down EzrA, as previously a peptide corresponding to the C-terminal 17 amino acids of FtsZ was shown to competitively inhibit EzrA binding to FtsZ and bound to EzrA with a dissociation constant of 102 μM [25]. The interaction between FtsZ and EzrA is sensitive to increasing salt concen-trations [25] but our pull-down assays were performed at 150 mM NaCl, a salt concentration at which at least 50% of binding should still occur. Possibly we need a higher expression of His-EzrAcyt or to perform the pull-down assay un-der cross-linking conditions, or both, to detect the interaction between EzrA and the FtsZ C-terminus.

We have shown that our pull-down assay is very specific. Only SepF was pulled down from an E. coli lysate and control experiments showed that this


85

pull-down was dependent on the presence of the C-terminus and indepen-dent of the tag attached to SepF. Furthermore, we could not detect SepF mu-tants that have lost the capability to bind FtsZ and we could show that FtsZ recovery in the reverse pull-down experiment depended on the presence of the FtsZ C-terminus (Fig 2). Our results strongly suggest that the FtsZ C-termi-nus is the only binding site for SepF. In an earlier study SepF has been found to co-precipitate with FtsZΔC16 polymers, which led the authors to suggest a potential secondary binding site for SepF on FtsZ [26]. We attribute the ob-served co-sedimentation of SepF with FtsZΔC16 polymers [26] to the buffer conditions used by the authors – 25 mM Pipes at pH 6.8. In an earlier report we have shown that SepF forms rings at pH 7.5 and high salt, but precipitates as aggregates at pH < 7 [18]. In vitro polymerization and tubulation experiments confirmed that the formation of FtsZ-SepF tubules depends on the presence of the FtsZ C-terminus (Fig 3). SepF sedimentation is dependent on the presence of FtsZ polymers, and does not occur when we use polymers of FtsZP372A or SepF alone (Fig. 3G).

To identify residues in the FtsZ C-terminus involved in SepF binding we performed an alanine-scan of the C-terminus and performed pull-down as-says. Mutation of the two residues that are totally conserved across species, Pro372 and Phe374, abolished the FtsZ-SepF interaction. In total, 11 out of the 16 mutations tested considerably affected SepF binding to the C-terminus. This suggests that the secondary and teriary structure of the FtsZ C-terminus is rec-ognized by SepF, in a way that is more similar to how ZipA binds FtsZ, through hydrophobic contacts and hydrogen bonds to the peptide backbone, rather than to how FtsA binds FtsZ, through specific salt bridges [22,29]. We cannot exclude that our mutations alter the structure of the C-terminus in such a way that residues are no longer in the right conformation to form salt bridges – for that structural information is required. Recently, the FtsZ C-terminus was divid-ed into a conserved C-terminal tail (CTT) that is followed by a highly variable - both in length and composition - set of residues, the C-terminal Variable region (CTV) [27]. The CTT is thought to constitute the ‘landing pad’ for FtsZ-binding proteins whereas the CTV plays a role in promoting lateral interactions be-tween FtsZ polymers [27]. The B. subtilis FtsZ CTV comprises residues 377-382. We noted that several residues in the CTV are likely to play a role in SepF bind-


86

ing, as alanine mutations of residues 377, 378, 380 and 381 led to a consider-able decrease in SepF binding. Also, the salt bridge between the C-terminal residue of T. maritima FtsZ, which has a 7 residue long CTV, and FtsA [29], sug-gests that the CTV is part of the ‘landing pad’ for interacting proteins, next to its role in promoting lateral interactions. However, the finding that the E. coli FtsZ C-terminus, with a different and noticeably shorter CTV, also binds SepF, supports the notion that the CTT is the main binding partner for FtsZ-binding proteins. More information on the interactions between the FtsZ C-terminus and several FtsZ-binding proteins is required to elucidate this question.

In conclusion, we have developed a novel and highly specific assay to iden-tify interaction partners of the C-terminus of FtsZ from B. subtilis and E. coli, and validated the assay with candidate binding proteins. Our results show that SepF specifically binds the C-terminus whereas the interactions of B. subtilis MinC and EzrA are too weak to be detected in our assay. We are currently using this assay to look for novel proteins interacting with the FtsZ C-termini from lysates of B. subtilis and E. coli cultures.

MATERIALS AND METHODS

PLASMID CONSTRUCTION

All plasmids are listed in Table 1. The sequence coding for the C-terminal 67 amino acid residues (317-383) from E. coli ftsZ was amplified from E. coli K12 chromosomal DNA by PCR using primers djs414 and djs415 (primers list-ed in Table 2). The sequence coding for the C-terminal 69 amino acid residues (314-382) from B. subtilis ftsZ was amplified from B. subtilis 168 chromosomal DNA by PCR using primers djs416 and djs417. Both PCR products were digest-ed with PmeI and SgfI and ligated into PmeI/SgfI digested pFN18A (Promega), resulting in plasmids pDJ64 (E. coli ftsZ) and pDJ65 (B. subtilis ftsZ). pDJ65 was used as a template for site directed mutagenesis according to the Quickchange protocol (Stratagene), to introduce a stop codon at the position of D367, and to systematically change residues 367 to 382 to alanine, using primers djs426 to djs459. Similarly, pDJ64 was used as a template for site directed mutagenesis,

MATERIALS AND METHODS CHAPTER 3

87

Table 1. Plasmids

Plasmids Relevant characteristics Source

pFN18A HaloTag® T7 Flexi® Vector bla PT7 HaloTag-TEVsite Promega

pET302 lacZα, trc HisTag-enterokinase site [34]

pDJ15 tetR PtetA-minC-strep-tagII [4]

pDJ26 bla lacIQ PT7-his8-zapA [4]

pMal-SepF bla Ptac malE-sepF [18]

pMal-SepF-A98V bla Ptac malE-sepFA98V [18]

pMal-SepF-F124S bla Ptac malE-sepFF124S [18]

pDJ64 bla PT7 HaloTag-TEVsite-ftsZEc317-383; ZCEc this work

pDJ65 bla PT7 HaloTag-TEVsite-ftsZBs314-382 ; ZC this work

pDJ69 bla PT7 HaloTag-TEVsite-ftsZBs314-366 ; ZΔC16 this work

pDJ73 bla PT7 HaloTag-TEVsite-ftsZEc317-369; ZEcΔC14 this work

pDJ74 bla PT7 HaloTag-TEVsite-ftsZEc317-P375A-383; ZCEcP375A this work

pDJ367-382 bla PT7 HaloTag-TEVsite-ftsZBs314-382; residue in plasmid name changed to Ala this work

pEK5 lacZα, trc HisTag-enterokinase site- ezrA27-562 this work

Table 2. Primers

primer name primer sequence 5’-3’ aa change

djs414 GTCCGCGATCGCCATGGACAAACGTCCTG

djs415 CAGCGTTTAAACTTAATCAGCTTGCTTACG

djs416 GTCCGCGATCGCCACCGGCTTTATCGAAC

djs417 GAGCGTTTAAACTTAGCCGCGTTTATTACGG

djs426 CTTCACAGCCGGCTGATTAAACGCTTGACATCCCGAC D367Stop

djs427 GTCGGGATGTCAAGCGTTTAATCAGCCGGCTGTGAAG D367Stop

djs428 GATGATACGCTTGACATCGCGACATTCTTAAGAAACC P372A

djs429 GGTTTCTTAAGAATGTCGCGATGTCAAGCGTATCATC P372A

djs430 GAAACCGTAATAAACGCGCCTAAGTTTAAACGAATTC G382A

djs431 GAATTCGTTTAAACTTAGGCGCGTTTATTACGGTTTC G382A

djs432 CTTAAGAAACCGTAATAAAGCCGGCTAAGTTTAAACG R381A

djs433 CGTTTAAACTTAGCCGGCTTTATTACGGTTTCTTAAG R381A


88

primer name primer sequence 5’-3’ aa change

djs434 CATTCTTAAGAAACCGTAATGCACGCGGCTAAGTTTAAACG K380A

djs435 CGTTTAAACTTAGCCGCGTGCATTACGGTTTCTTAAGAATG K380A

djs436 CGACATTCTTAAGAAACCGTGCTAAACGCGGCTAAGTTTAAAC N379A

djs437 GTTTAAACTTAGCCGCGTTTAGCACGGTTTCTTAAGAATGTCG N379A

djs438 CCGACATTCTTAAGAAACGCTAATAAACGCGGCTAAG R378A

djs439 CTTAGCCGCGTTTATTAGCGTTTCTTAAGAATGTCGG R378A

djs440 CCCGACATTCTTAAGAGCCCGTAATAAACGCGG N377A

djs441 CCGCGTTTATTACGGGCTCTTAAGAATGTCGGG N377A

djs442 CATCCCGACATTCTTAGCAAACCGTAATAAACGCG R376A

djs443 CGCGTTTATTACGGTTTGCTAAGAATGTCGGGATG R376A

djs444 CTTGACATCCCGACATTCGCAAGAAACCGTAATAAAC L375A

djs445 GTTTATTACGGTTTCTTGCGAATGTCGGGATGTCAAG L375A

djs446 GATACGCTTGACATCCCGACAGCCTTAAGAAACCGTAATAAAC F374A

djs447 GTTTATTACGGTTTCTTAAGGCTGTCGGGATGTCAAGCGTATC F374A

djs448 CGCTTGACATCCCGGCATTCTTAAGAAACC T373A

djs449 GGTTTCTTAAGAATGCCGGGATGTCAAGCG T373A

djs450 GCTGATGATACGCTTGACGCCCCGACATTCTTAAGAAAC I371A

djs451 GTTTCTTAAGAATGTCGGGGCGTCAAGCGTATCATCAGC I371A

djs452 GGCTGATGATACGCTTGCCATCCCGACATTCTTAAG D370A

djs453 CTTAAGAATGTCGGGATGGCAAGCGTATCATCAGCC D370A

djs454 CCGGCTGATGATACGGCTGACATCCCGACATTC L369A

djs455 GAATGTCGGGATGTCAGCCGTATCATCAGCCGG L369A

djs456 CCGGCTGATGATGCGCTTGACATCC T368A

djs457 GGATGTCAAGCGCATCATCAGCCGG T368A

djs458 CAGCCGGCTGATGCTACGCTTGACATC D367A

djs459 GATGTCAAGCGTAGCATCAGCCGGCTG D367A

djs461 CAAACTGCGAAAGAGCCGTAATATCTGGATATCCCAGC FtsZEcD370Stop

djs462 GCTGGGATATCCAGATATTACGGCTCTTTCGCAGTTTG FtsZEcD370Stop

djs463 GGATTATCTGGATATCGCAGCATTCCTGCGTAAG FtsZEcP375A

djs464 CTTACGCAGGAATGCTGCGATATCCAGATAATCC FtsZEcP375A

ek13 AGACTACCATGGCCGAAATCGACCGGCTGGA

ek14 CGTTACTCTAGACTAAGCGGATATGTCAGCTTTG


89

to introduce a stop codon at the position of D370, and to change residues P375 to alanine, using primers djs461 to djs464. The sequence coding for the sol-uble part of B. subtilis EzrA (amino acid residues 27-562) was amplified from B. subtilis 168 chromosomal DNA by PCR using primers ek13 and ek14. The PCR product was digested with NcoI and XbaI and ligated into NcoI/XbaI digested pET302, resulting in plasmid pEK5. All constructs were verified by sequencing.

PROTEIN PURIFICATION

MBP-SepF and SepF were purified as described [18]. FtsZ was purified as de-scribed [4], FtsZΔC16 and FtsZP372A were purified using the same protocol.

PULL DOWN EXPERIMENTS

Expression of HaloTagged constructs was done in E. coli BL21(DE3) carrying plasmid pBS58 which encodes an extra copy of ftsQAZ [33]. Freshly transformed cells were grown overnight at 37°C, diluted 1:100 and grown to an OD600 ~0.7 and induced with 1 mM IPTG. After three hours of growth cells were harvested by centrifugation and washed with PBS. Cell pellets were flash frozen in liquid nitrogen and stored at -20°C. To prepare lysates, cells were taken up in 1/10th of the original volume in buffer HPB (50 mM HEPES pH 7,5; 150 mM NaCl). Lyso-zyme (0,1 mg/ml) and DNAse (1 µg/ml) were added and the suspensions were incubated for 10 min on ice. Cells were disrupted using sonication (12 cycles of 5s with 5s cooling at 10 microns probe amplitude). Aliquots of 1ml were made and stored at -20°C. For a typical pull down experiment, for every construct test-ed a 1 ml aliquot was thawed on ice and subsequently centrifuged at 10000 ×g for 30 min at 4°C. DTT (to 1 mM) and Nonidet P-40 (to 0,005% v/v) were added to the supernatant. The rest of the procedure was carried out in buffer HPB+ (50 mM HEPES pH 7,5; 150 mM NaCl; 1 mM DTT; 0,005% v/v nonidet P-40). To the supernatant, 200µl (25%) of HaloLink resin (Promega) equilibrated with HPB+ was added. The samples were incubated for 1 h at RT with gentle mixing. Subsequently, the samples were centrifuged at 1000 ×g for 5min at RT. The supernatant was discarded and the resin was washed 3 times with HPB+. The resin with bound HaloTagged bait protein were subsequently incubated with


90

1 mL cell lysates (10x concentrated with respect to the original culture volume) of E. coli strains overexpressing potential binding partners. Alternatively, the resin with bound Halo-tagged bait protein was incubated with 1 mL of MBP-SepF, MBP-SepFA98V, MBP-SepFF124S or SepF at 15 μM diluted in HPB+. For the alanine-scanning experiments, 300 μL of 30 μM SepF was used. The samples were incubated for 1 h at RT with gentle mixing. Subsequently, the samples were centrifuged at 1000 ×g for 5min at RT, and the supernatant was kept as the flow-through fraction. The resin was washed 3 times with HPB+. The HaloTagged bait protein and associated proteins were released from the resin by TEV-cleavage using 50 μL of Cleavage solution (HPB+ containing TEV-prote-ase, Promega) for 1 h at RT with gentle mixing. The eluted material was collect-ed by centrifugation at 1000 ×g for 5min at RT and the resin was washed with an additional 50 μL of HPB+ which was collected by centrifugation and added to the eluted material. Samples were prepared for SDS-PAGE and run on 15% gels, which were stained with Coomassie Brilliant Blue.

For the FtsZ pull-downs with MBP-SepF, MBP-SepF bound to amylose-resin was prepared as described [18]. 250 μL aliquots of amylose resin with MBP-SepF were incubated with 300 μL of 30 μM FtsZ, FtsZΔC16 and FtsZP372A in buffer A (20 mM Tris–HCl pH 7.4, 200 mM KCl, 1 mM EDTA). The proteins were allowed to bind to the resin for 30 min at 4°C. Unbound protein and liquid was removed by mild centrifugation and collected as Flow-through. The resin was washed 6 times with buffer A and the last wash fraction was collected. MBP-SepF and bound FtsZ were eluted from the resin with 300 μL of 10 mM maltose in buffer A. Equal amounts of the input material, the flow-through, last wash and eluate were loaded on 12% SDS-PAGE gels, which were either stained with Coomassie Brilliant Blue, or subsequently transferred to PVDF using Western Blotting and developed using anti-FtsZ.

ELECTRON MICROSCOPY

FtsZ , FtsZP372A, FtsZ∆C16 (at 10 µM) and SepF (6 µM) were incubated in polym-erization buffer (50mM HEPES/ NaOH, pH=7,4, 300 mM KCl, 10mM MgCl2) for 5 min. at 30°C. GTP was added to a final concentration of 2 mM to start FtsZ po-lymerization and the samples were incubated for 20 min. at 30°C. 2 µl of each

ACKNOWLEDGEMENTS CHAPTER 3

91

sample was applied onto a glow-discharged 400 mesh carbon-coated copper grid, incubated for 30 s., blotted dry with filter paper and negatively stained using 5µl uranyl-acetate (2%). The grids were viewed in a Philips CM120 elec-tron microscope equipped with a LaB6 filament operating at 120 kV. Images were recorded with a Gatan 4000 SP 4 K slow-scan CCD camera at 36,400 × magnification.

SEDIMENTATION OF TUBULES

FtsZ or FtsZP372A (at 10µM) was mixed with SepF (12 µM) in polymerization buf-fer and incubated for 5 minutes at 30°C. Polymerization was started by the ad-dition of GTP to a final concentration of 2 mM, and samples were incubated for a further 20 min. at 30°C. Control samples contained only FtsZ, FtsZP372A or SepF. Tubules were spun down for 15 min. at 24,600 ×g, at 25°C. Pellet and superna-tant fractions were recovered and equal amounts were analyzed by SDS-PAGE and staining with Coomassie Brilliant Blue.

ACKNOWLEDGEMENTS

We would like to thank Bart van den Berg van Saparoea (VU University Amsterdam, the Netherlands), Nina van den Berghe and Zhong Yu (Promega Benelux, the Netherlands) for helpful advice regarding the use of the HaloTag system. Seline Zwarthoff (University of Groningen, the Netherlands) is acknowl-edged for the construction of pDJ73 and pDJ74. We thank Leendert Hamoen (University of Newcastle, UK) and Arnold Driessen (University of Groningen, the Netherlands) for helpful comments on the manuscript.


92

REFERENCES

1. de Boer PA. (2010) Advances in under-standing E. coli cell fission. Curr Opin Mi-crobiol 13(6): 730-737.

2. Adams DW, Errington J. (2009) Bacteri-al cell division: Assembly, maintenance and disassembly of the Z ring. Nat Rev Microbiol 7(9): 642-653.

3. Hu Z, Mukherjee A, Pichoff S, Lutkenhaus J. (1999) The MinC component of the di-vision site selection system in Escherichia coli interacts with FtsZ to prevent polym-erization. Proc Natl Acad Sci USA 96(26): 14819-14824.

4. Scheffers DJ. (2008) The effect of MinC on FtsZ polymerization is pH dependent and can be counteracted by ZapA. FEBS Lett 582(17): 2601-2608.

5. Bernhardt TG, de Boer PA. (2005) SlmA, a nucleoid-associated, FtsZ binding pro-tein required for blocking septal ring as-sembly over chromosomes in E. coli. Mol Cell 18(5): 555-564.

6. Tonthat NK, Arold ST, Pickering BF, Van Dyke MW, Liang S, et al. (2011) Molec-ular mechanism by which the nucleoid occlusion factor, SlmA, keeps cytokine-sis in check. EMBO J 30(1): 154-164.

7. Wu LJ, Ishikawa S, Kawai Y, Oshima T, Ogasawara N, et al. (2009) Noc protein binds to specific DNA sequences to co-ordinate cell division with chromosome segregation. EMBO J 28(13): 1940-1952.

8. Weart RB, Lee AH, Chien AC, Haeusser DP, Hill NS, et al. (2007) A metabolic sen-sor governing cell size in bacteria. Cell 130(2): 335-347.

9. Camberg JL, Hoskins JR, Wickner S. (2009) ClpXP protease degrades the cy-toskeletal protein, FtsZ, and modulates FtsZ polymer dynamics. Proc Natl Acad Sci USA 106(26): 10614-10619.

10. Weart RB, Nakano S, Lane BE, Zuber P, Levin PA. (2005) The ClpX chaperone

modulates assembly of the tubulin-like protein FtsZ. Mol Microbiol 57(1): 238-249.

11. Hale CA, De Boer PAJ. (1997) Direct bind-ing of FtsZ to ZipA, an essential compo-nent of the septal ring structure that mediates cell division in E. coli. Cell 88: 175-185.

12. Levin PA, Kurtser IG, Grossman AD. (1999) Identification and characteriza-tion of a negative regulator of FtsZ ring formation in Bacillus subtilis. Proc Natl Acad Sci USA 96(17): 9642-9647.

13. Gueiros-Filho FJ, Losick R. (2002) A widely conserved bacterial cell division protein that promotes assembly of the tubu-lin-like protein FtsZ. Genes Dev. 16(19): 2544-2556.

14. Durand-Heredia JM, Yu HH, De Carlo S, Lesser CF, Janakiraman A. (2011) Iden-tification and characterization of ZapC, a stabilizer of the FtsZ ring in Escherichia coli. J Bacteriol 193(6): 1405-1413.

15. Hale CA, Shiomi D, Liu B, Bernhardt TG, Margolin W, et al. (2011) Identifica-tion of Escherichia coli ZapC (YcbW) as a component of the division apparatus that binds and bundles FtsZ polymers. J Bacteriol 193(6): 1393-1404.

16. Corbin BD, Wang Y, Beuria TK, Margolin W. (2007) Interaction between cell divi-sion proteins FtsE and FtsZ. J Bacteriol 189(8): 3026-3035.

17. Yang DC, Peters NT, Parzych KR, Uehara T, Markovski M, et al. (2011) An ATP-bind-ing cassette transporter-like complex governs cell-wall hydrolysis at the bac-terial cytokinetic ring. Proc Natl Acad Sci USA 108(45): E1052-E1060.

18. Gundogdu ME, Kawai Y, Pavlendova N, Ogasawara N, Errington J, et al. (2011) Large ring polymers align FtsZ polymers for normal septum formation. EMBO J 30(3): 617-626.


93

19. Lucet I, Feucht A, Yudkin MD, Errington J. (2000) Direct interaction between the cell division protein FtsZ and the cell differentiation protein SpoIIE. EMBO J. 19(7): 1467-1475.

20. Handler AA, Lim JE, Losick R. (2008) Pep-tide inhibitor of cytokinesis during spor-ulation in Bacillus subtilis. Mol Microbiol 68(3): 588-599.

21. Ma X, Margolin W. (1999) Genetic and functional analyses of the conserved C-terminal core domain of Escherichia coli FtsZ. Journal of Bacteriology 181(24): 7531-7544.

22. Mosyak L, Zhang Y, Glasfeld E, Haney S, Stahl M, et al. (2000) The bacterial cell-divi-sion protein ZipA and its interaction with an FtsZ fragment revealed by X-ray crys-tallography. EMBO J 19(13): 3179-3191.

23. Shen B, Lutkenhaus J. (2009) The con-served C-terminal tail of FtsZ is required for the septal localization and division inhibitory activity of MinC(C)/MinD. Mol Microbiol 72(2): 410-424.

24. Haney SA, Glasfeld E, Hale C, Keeney D, He Z, et al. (2001) Genetic analysis of the Escherichia coli FtsZ.ZipA interaction in the yeast two-hybrid system. character-ization of FtsZ residues essential for the interactions with ZipA and with FtsA. J Biol Chem: 0-11987.

25. Singh JK, Makde RD, Kumar V, Panda D. (2007) A membrane protein, EzrA, reg-ulates assembly dynamics of FtsZ by in-teracting with the C-terminal tail of FtsZ. Biochemistry 46(38): 11013-11022.

26. Singh JK, Makde RD, Kumar V, Panda D. (2008) SepF increases the assembly and bundling of FtsZ polymers and stabiliz-es FtsZ protofilaments by binding along its length. J Biol Chem 283(45): 31116-31124.

27. Buske PJ, Levin PA. (2012) The extreme C-terminus of the bacterial cytoskeletal

protein FtsZ plays a fundamental role in assembly independent of modulatory proteins. J Biol Chem 284(14): 10945-10957.

28. Datta P, Dasgupta A, Bhakta S, Basu J. (2002) Interaction between FtsZ and FtsW of Mycobacterium tuberculosis. J Biol Chem 277(28):24983-24987.

29. Szwedziak P, Wang Q, Freund SM, Lowe J. (2012) FtsA forms actin-like protofila-ments. EMBO J 2249-2260224

30. GV Encell, McDougall, Hartzell-Karassina. (2008) HaloTag: A novel protein labeling technology for cell imaging and protein analysis. ACS Chem Biol 3(6): 373-382.

31. de Oliveira IF, de Sousa Borges A, Kooij V, Bartosiak-Jentys J, Luirink J, et al. (2010) Characterization of ftsZ mutations that render Bacillus subtilis resistant to MinC. PLoS One 5(8): e12048.

32. Marston AL, Errington J. (1999) Selection of the midcell division site in Bacillus subtilis through MinD-dependent po-lar localization and activation of MinC. Molecular Microbiology 33: 84-96.

33. Wang X, Lutkenhaus J. (1993) The FtsZ protein of Bacillus subtilisis localized at the division site and has GTPase activity that is dependent upon FtsZ concentra-tion. Mol Microbiol 442.

34. van der Does C, Manting EH, Kaufmann A, Lutz M, Driessen AJM(1998) Inter-action between SecA and SecYEG in micellar solution and formation of the membrane-inserted state. Biochemistry 37(1): 201-210.

CHAPTER

The effects of manganese on the

interaction of sporulation protein

SpoIIE with itself and cell division

protein FtsZEwa Król, Anabela de Sousa

Borges and Dirk-Jan

Scheffers

Department of Molecular

Microbiology, Groningen

Biomolecular Sciences

and Biotechnology

Institute, University

of Groningen,

The Netherlands

Experiments presented in Figures 1, 2, 4, 5 and 6 were performed by E. Cendrowicz.

ABSTRACT

Sporulation is an adaptive process undertaken by Bacillus subtilis and its rel-atives under starvation conditions, that results in the formation of a dormant cell called a spore. It begins with the formation of a polar septum, which di-vides the cell into two unequal-sized cells, a larger mother cell and a smaller forespore. After polar septum formation, a transcription specific factor, σF is activated in the forespore. Both events require a protein called SpoIIE. SpoIIE interacts directly with the cell division protein FtsZ and localizes together with FtsZ to the polar septa. The second function of SpoIIE is the activation of σF by dephosphorylating the anti-sigma factor antagonist SpoIIAA, a process de-pending on the cofactor Mn2+. Here, we report our attempts to study the FtsZ-

-SpoIIE interaction in vitro. We demonstrate influence of divalent metals on the oligomerization of the cytoplasmic domain of SpoIIE. We investigate the influ-ence of the SpoIIE cofactor, Mn2+, on the oligomerization of FtsZ. Finally, we study direct interaction between FtsZ and SpoIIE and show clear interaction between these two proteins in the presence of GDP.


97

INTRODUCTION

In response to starvation conditions, Bacillus subtilis cells cease vegetative growth and initiate formation of a dormant cell type called a spore. The first clear step in this process is the formation of the asymmetrically positioned septum that divides the cell into two daughter cells of different size: the larg-er mother cell and the smaller forespore, that each receive one copy of the chromosome. Next, compartment specific transcription factors σF and σE are activated in the forespore and in the mother cell respectively. This activation event is critical because it initiates the rest of the sporulation developmental program in each daughter cell. Ultimately, the forespore becomes the spore and the mother cell lyses when the process is complete 1-5.

Upon entry into sporulation, the cell division protein FtsZ, that during veg-etative growth drives the mid-cell division, switches its position to the polar sites of the cell. The switch is triggered by an integral membrane protein called SpoIIE 1,6-8. At the onset of sporulation, SpoIIE co-localizes with FtsZ at mid-cell and both proteins redeploy to polar sites via a spiral-like intermediate 6. Next, one of the polar Z-rings disassembles and the other one is converted into a di-vision septum. SpoIIE contributes to the formation of the polar Z-rings but is not required for the process. It was shown that in spoIIE null mutant cells, FtsZ may still localize to the polar sites but the switch from medial to polar rings is delayed and the frequency of polar Z-ring formation decreased. After constric-tion of the polar Z-ring, FtsZ is released to the cytosol and SpoIIE remains asso-ciated with the polar septum and performs its second function, which involves its phosphatase domain 2. It activates σF by dephosphorylating the anti-sigma factor antagonist SpoIIAA 3,9. SpoIIE phosphatase is inactive in the pre-division-al cell and becomes active only after the asymmetric septum is formed 10. Thus, it is thought that FtsZ may be involved in the activation of phosphatase of SpoIIE. SpoIIE is a transmembrane protein with a three-domain structure. The N-terminal domain (domain I) consists of 10 membrane-spanning segments, the central, poorly conserved domain (domain II) is involved in the oligomeri-zation of SpoIIE and the interaction with FtsZ and the C-terminal domain (do-main III) is structurally related to the PP2C (Protein phosphatase 2C), Mn2+-de-pendent family of protein phosphatases (Fig. 1A) 11-14. There is a clear structural


98

separation of the three domains of SpoIIE. It was shown that the C-terminal domain can independently function as a phosphatase in vitro, and that the central domain may interact with either FtsZ or itself, independent of the other two domains. However, several in vivo studies found that some mutations in the domain II influence the phosphatase activity of SpoIIE 10. It was also shown that some mutations in domain III that impair the phosphatase activity also impair polar cell division 7.

The interaction between SpoIIE and FtsZ was previously shown in vitro 15. However, details about the SpoIIE-FtsZ interaction and SpoIIE self-interaction are not known because of difficulties in the expression of the soluble cyto-plasmic domain of SpoIIE 16. In this work, we purified the soluble part of SpoIIE (SpoIIEcyt) fused to maltose binding protein (MBP) and studied the interaction of SpoIIE with FtsZ and itself in more detail.


PLASMID CONSTRUCTION

pMalC2x (New England Biolabs) was used to clone and purify strep-SpoIIEcyt as an N-terminal MBP-strep-SpoIIEcyt (Ms-SpoIIEcyt) fusion (Fig. 1A). A 1503 bps fragment coding for spoIIEcyt was amplified by PCR from B. subtilis 168 tem-plate genomic DNA using a forward primer containing the strep-tag coding sequence (bold) (ek61 5’-AGCGCTTGGCGTCACCCGCAGTTCGGTGGTCCT-CAATCTATTACGAGGAAAGTGG) 10 and a reverse primer containing a BamHI site (underlined) (ek62 5’-GCGGATCCCATATATTCCCATCTTCGCCAGAAG). The PCR product was digested using BamHI and ligated into pMalC2x linearized with XmnI (blunt end)/BamHI, resulting in plasmid pEK33.

PROTEIN EXPRESSION AND PURIFICATION

FtsZ was purified as described 17,18. His-EzrAcyt was purified as described 19. Ms-SpoIIEcyt, MBP-SepF124 20 or MBP were produced in E. coli BL21-RIL cells using the same method. Freshly transformed cells were grown overnight on


99

LB agar plates containing 100 µg/ml ampicillin and 50 µg/ml chloramphen-icol. Liquid LB medium containing the same antibiotics was inoculated with a single colony from a fresh plate. An overnight culture was diluted 1:100 into fresh LB containing the same antibiotics and grown at 37°C until OD600=0,7. Protein expression was induced by addition of 1 mM IPTG at 37°C for 3 hours. For purification, cells were resuspended in 50 mM Tris/ HCl pH=7.5, 300 mM KCl, 0.5 mM DTT and 0.1% Triton X-100 and disrupted by sonication. The cell lysate was clarified by centrifugation at 20 000 ×g for 20 min and the superna-tant was applied onto amylose resin. The resin was washed with the same buf-fer without Triton X-100 and protein was eluted with 50 mM Tris/ HCl pH=7.5, 300 mM KCl, 0.5 mM DTT and 10 mM maltose. All three proteins were con-centrated using Amicon Ultra-15 Centrifugal Filter Units (Merck Millipore). The protein concentration was measured and absorption spectra were taken using a NanoDrop ND-1000 Spectrophotometer (ISOGEN Lifescience). An extinction coefficient (Ex=107,720) was determined using an ExPASy ProtParam tool us-ing the Ms-SpoIIEcyt sequence.

To obtain strep-SpoIIEcyt, the protein solution was mixed with an equal volume of 50 mM Hepes/ NaOH pH=7.5, 1M KCl, 0.5mM DTT, 1mM EDTA and 1% Triton X-100 to obtain final concentrations of 650 mM KCl and 0.5% Triton X-100. The MBP tag was cleaved overnight at 4°C using Factor Xa protease (New England Biolabs). Factor Xa was deactivated using Dansyl-glu-gly-arg-chloro-methyl ketone and the protein mixture was applied onto HiLoad Superdex 16/600 gel filtration column and eluted with 50 mM Hepes/ NaOH, pH=7.5, 650 mM KCl, 0.5 mM DTT, 1 mM EDTA and 0.5% Triton X-100.

ICP-OES MEASUREMENT

Purified Ms-SpoIIEcyt from two independent purifications was lyophilized and analysed for the presence of calcium, iron, magnesium, manganese and zinc using inductively-coupled plasma optical emission spectroscopy on an Optima 7000DV ICP-OES (PerkinElmer) apparatus. The measurements were performed in duplicate.


100

LIGHT SCATTERING

The effects of various cations on the oligomerization of strep-SpoIIEcyt were monitored by 90° light scattering using an AMINCO-Bowman Series 2 fluores-cence spectrometer. Strep-SpoIIEcyt or MBP (1.5 µM) were incubated in 50 mM MES/NaOH pH=6.5, 300 mM KCl buffer. After 60 seconds of incubation various chloride salts of divalent cations (Ca2+, Mg2+, Mn2+, Zn2+, Fe2+, Cu2+ or Co2+, Ni2+) were added to the solution and the light scattering signal was monitored for 1 hour.

The reversal of strep-SpoIIEcyt oligomerization by EDTA was done in a larg-er volume cuvette (1 ml) to allow stirring, and measurements were taken us-ing a QuantaMaster™ spectrofluorometer controlled by the FelixGX program (Photon Technology International, Inc.). MnCl2 (10 mM final concentration) was added to the sample after 3.5 min of incubation and EDTA (20 mM final concentration) was added after 10 minutes of incubation. As a dilution control, H2O was added to the sample. The light scattering signal of protein without Mn2+ was measured as blank.

The effect of strep-SpoIIEcyt (1.5 µM) on the assembly of FtsZ (10 µM) was studied in 50 mM Tris/HCl, pH=7.5, 300 mM KCl and 10 mM MgCl2. After 90 sec of measurement GTP or GDP (2 mM final concentration) was added to the sam-ple. As controls, measurements were taken of strep-SpoIIEcyt and/or FtsZ in the presence or absence of nucleotides. All experiments were done at 30°C.

CRITICAL CONCENTRATION DETERMINATION

To study changes in critical concentration for oligomerization of Ms-SpoIIEcyt in the presence and absence of Mn2+, fluorescence spectra of increasing concentration of Ms-SpoIIEcyt-Cy5 were acquired in the presence of Mn2+ or EDTA. Ms-SpoIIEcyt was purified from lysates by binding to a amylose resin as described above, with the inclusion of an on-column Cy5 labeling step. After washing column bound protein with buffer A (50 mM Tris/ HCl pH=7.5, 300 mM KCl, 0.5 mM DTT), the resin with Ms-SpoIIEcyt was mixed with fluorescent label Cy5 (GE Healthcare) and incubated for 1 hour at 4°C. Excess label was washed away using the same buffer and labeled protein was eluted with buffer A sup-


101

plemented with 10 mM maltose. For the measurement equal volumes (12 µl containing 0.23 µM Ms-SpoIIEcyt-Cy5) of Ms-SpoIIEcyt-Cy5 was titrated to a buf-fer containing 50 mM Tris/HCl, 300 mM KCl, 1 mM EDTA (for the measurement without manganese) or 10 mM MnCl2 (for the measurement with manganese). The measurements were taken immediately after addition of Ms-SpoIIEcyt at 30°C. Fluorescence was excited at 633 nm and emission spectra (640-685 nm) were acquired in a QuantaMaster™ spectrofluorometer controlled by the Felix-GX program (Photon Technology International, Inc.).

SEDIMENTATION ASSAY

FtsZ (10 µM) was mixed with Ms-SpoIIEcyt (10 µM) in polymerization buffer (50 mM Hepes/ NaOH, pH=7.5, 50 mM KCl, 10 mM MgCl2). After incubation for 15 min at 30°C, an equal volume of GTP or GDP was added to the samples (final concentration 2 mM). The samples were incubated for another 15 min at 30°C and centrifuged at 186 000 ×g for 15 min at 20°C. Pellet and supernatant fractions were analyzed by SDS-PAGE as described 17. Two controls were used in this assay. MBP-SepF124 (10 µM), a MBP fusion of SepF mutant that does not interact with FtsZ 20 was used in place of Ms-SpoIIEcyt to see whether MBP is not a cause of increase in sedimentation of FtsZ. To check whether the interaction between Ms-SpoIIEcyt and FtsZ is specific, His-EzrAcyt (10 µM) was used in the assay instead of FtsZ.

For the sedimentation assay of Ms-SpoIIEcyt alone, Ms-SpoIIEcyt was preincu-bated for 20 min in 100 µl of polymerization buffer (50 mM MES/ NaOH, pH=6.5, 50 mM KCl) supplemented with 1 mM EDTA. 30 µl of the sample was collected for further analysis. After that, the sample was supplemented with 5 mM MnCl2 and incubated for another 30 min. After the incubation time, 30 µl sample was collected. The rest of the sample was supplemented with 10 mM EDTA and incubated for another 40 min. All the samples collected after each incubation time were collected, spun down at 186 000 ×g and the supernatants and pel-lets were separated for analysis by SDS-PAGE.


102

ELECTRON MICROSCOPY

To visualize Ms-SpoIIEcyt using Transmission Electron Microscopy (TEM), 10 µM Ms-SpoIIEcyt was prepared in 50 mM Hepes/ NaOH, pH=7.5, 50 mM KCl.

To visualize FtsZ in the presence or absence of Ms-SpoIIEcyt, 10 µM FtsZ with or without 5 µM Ms-SpoIIEcyt was prepared in polymerization buffer: 50 mM Tris/ HCl, pH=7.5, 300 mM KCl supplemented with 1 mM EDTA or 10 mM MgCl2 or 10 mM MnCl2. After 5 min of incubation 2 mM GTP was added to the mixture. 4 µl of each sample was collected in appropriate time points and applied onto glow discharged carbon grids and grids were prepared as described 17. The grids were examined in a Philips CM120 electron microscope equipped with a LaB6 filament operating at 120 kV. Images were recorded with a Gatan 4000 SP 4 K slow-scan CCD camera at magnifications 36,400 × (for FtsZ ± Ms-SpoIIEcyt) or 45,500 × (for Ms-SpoIIEcyt structures alone).

FLUORESCENCE MICROSCOPY

An overnight culture of B. subtilis strain 4055 19 grown in CH medium 21 was di-luted into fresh CH to an OD of 0.1. Cells were grown at 37°C until OD of 0.7. At this point, 2 samples of 5 ml were taken and cells were collected and washed 2 times with the same volume of CH with (spo+) or without (spo-) manganese. After the washing steps, spo+ and spo- cells were resuspended in 100 µl of CH with and without manganese, respectively. Sporulation medium (with or without manganese) was added up to the volume of 5 ml in the presence of 0.02 mM of IPTG. Cells were allowed to sporulate at 37°C by continuing the incubation. Every hour, a 500 µl sample was taken and cells were harvested and resuspended in 20-50 µl of PBS before being mounted on an agarose pad prior to microscopy. FtsZ-eYFP was visualized as described in 19 and cells were scored according to their Z-ring localization.

GTPASE ASSAY

The GTP hydrolysis rate determination was performed as described in 17. Ms-SpoIIEcyt was pre-incubated with 1 mM EDTA to remove metal bound to the

RESULTS CHAPTER 4

103

protein during purification to obtain Ms-SpoIIEcyt-apo. Ms-SpoIIEcyt-Fe – metal bound during purification, Ms-SpoIIEcyt-apo or MBP were incubated with FtsZ in polymerization buffer: 50 mM Hepes/ NaOH, pH=7.5, 300 mM KCl, 1 mM EDTA. MgCl2 was added in the same time as GTP to avoid oligomerization of Ms-SpoIIEcyt before polymerization of FtsZ.

RESULTS

SPOIIE CO-PURIFIES WITH BOUND METAL IONS

The 501 residue C-terminal fragment of SpoIIE (amino acids 326-827) was N-terminally fused to maltose binding protein (MBP) via a short strep-tag. The Ms-SpoIIEcyt protein was purified with a high yield (~ 40 mg/1L culture). The ab-sorption spectrum of purified Ms-SpoIIEcyt showed a broad peak in the 420 nm region, similar to those of iron/manganese-protein complexes 22, suggesting that Ms-SpoIIEcyt was purified in a metal bound form. The absorption spectrum of purified MBP-SepF124 (which will serve as a control protein for some assays) did not show similar peak at this region. The C-terminal domain of SpoIIE be-longs to the Mn2+- dependent phosphatase 2C family of phosphatases 14. We conducted an ICP-OEC measurement of lyophilized Ms-SpoIIEcyt. The measure-ment revealed Fe bound to the sample in the ratio close to 2:1 (Fig. 1B). As iron is very similar to manganese, we think that it simply replaced manganese in its active site during folding in the heterologous host. Our results show 2 metal ions bound per one SpoIIE molecule. The analysis also revealed small amounts of Ca and Mg bound to the protein but manganese levels were below the de-tection limit (Fig. 1B).

It was possible to replace iron with manganese by incubation of protein with EDTA followed by incubation with excess of MnCl2. After incubation with EDTA, the absorption peak at 420 nm disappeared and appeared again when the protein was incubated with MnCl2. However, the spectrum had a lower peak compared to the spectrum of Ms-SpoIIEcyt-Fe obtained from the purified sample. We do not know if the metal binding site was again fully occupied by manganese.


104

Strep-SpoIIEcyt was further purified using gel filtration after cleavage of the MBP tag. The purification was performed in the presence of 650 mM KCl and 0.5% TritonX-100 as the protein was more stable under these conditions. Strep-SpoIIEcyt was eluted in several fractions corresponding to a range of mo-lecular weights between 60 and 500 kDa, indicating the presence of oligomers of up to 9-10 subunits. This result is in agreement with an earlier study that describes oligomerization of SpoIIE 15. As the purification buffer was not com-patible with visualization of the oligomers using elactron microscopy, we di-luted Ms-SpoIIEcyt into a 50 mM Hepes/ NaOH pH=7.5, 50 mM KCl buffer to a final concentration of 10 µM. We found two different structural species on the grid, rod shaped protein oligomers and circular structures (Fig. 1C). The circular structures may represent rods which bound perpendicular to the carbon grid (top view of Ms-SpoIIEcyt oligomers) whereas rods represent a side view of the Ms-SpoIIEcyt oligomers (Fig. 1C).

METAL BINDING ENHANCES OLIGOMERIZATION OF STREP-SPOIIECYT

It was shown before that domain II (amino acids 321-567) is involved in the oligomerization of SpoIIE (Fig. 1A) 15. We wanted to study the influence of metal ions on strep-SpoIIEcyt oligomerization and therefore performed light scattering measurement of strep-SpoIIEcyt in the presence of its cofactor Mn2+ and other divalent cations. Strep-SpoIIE showed an increase in light scattering signal upon binding to its cofactor. We wanted to see whether Fe, which was co-purified with Ms-SpoIIEcyt, has a similar effect on strep-SpoIIEcyt. However, Fe2+ becomes oxidized very easily in solution and the analysis in the presence of this metal was not possible even when reducing agents like DTT or ascor-bate were added to the buffer. Therefore we tested other divalent metal ions, which are commonly used as cofactors by biological systems, and thus may easily bind to the same site on SpoIIE as manganese. All the metals used had a significant influence on the oligomerization of strep-SpoIIEcyt (Fig. 2A). It is known that binding preferences of metals to some proteins do not always correspond to their metal requirements 23. This feature was also observed for strep-SpoIIEcyt. Light scattering signal was the highest in the presence of the strongest binding metal Zn2+ or Ni2+ and the lowest in the presence of Mg2+

RESULTS CHAPTER 4

105

0

0,5

1

1,5

2

2,5

Fe Ca Mg Mn Zn

Am

etal

/ pro

tein

ratio

B

N C

CMBP

I II III

transmembrane oligomerizationFtsZ interaction

PP2C phosphatase

strep

C

N 0

0,5

1

1,5

2

2,5

Fe Ca Mg Mn Zn

A

met

al/ p

rote

in ra

tio

B

N C

CMBP

I II III

transmembrane oligomerizationFtsZ interaction

PP2C phosphatase

strep

C

N

Figure 1. (A) The putative three domain structure of SpoIIE (top) and a schematic of the SpoIIE con-

struct used in this study (bottom), (B) ICP-OES measurement of lyophilized Ms-SpoIIEcyt revealed

saturation of Ms-SpoIIEcyt with Fe and small amounts of other divalent cations but not Mn. (B)

Electron microscopy of Ms-SpoIIEcyt oligomers after co-purification with Fe. Arrows point to rod

and circular structures of Ms-SpoIIEcyt. Scale bar: 50 nm.

(Fig. 2A), according to Irving-Williams series 23. Like Fe, Cu was excluded from the measurement due to its instability/high reactivity with buffer components. To prove that the increase in light scattering signal is not due to reaction of metals with some buffer components, we used MBP as a control and did not see an increase in light scattering signal (Fig. 2B). Metal-dependent oligomeri-zation of Ms-SpoIIEcyt was in agreement with the observation that Ms-SpoIIEcyt was always recovered in the pellet fraction when used in FtsZ assays, which are performed in the presence of Mg2+ (Fig. 5B, C).

To study the critical concentration of Ms-SpoIIEcyt necessary to form oligo-mers we labeled Ms-SpoIIEcyt with the fluorescent dye Cy5. Labeled protein was titrated into a buffer supplemented with either EDTA or MnCl2. The fluores-cence signal continuously increased upon addition of Ms-SpoIIEcyt-Cy5 until it reached the concentration at which signal was quenched by oligomerization of Ms-SpoIIEcyt-Cy5 and did not further increase (Fig. 2C, D). The emission spec-


106

Ca2+

Zn2+

Zn2+

Zn2+

Co2+

Ni2+

Ni2+

Ni2+

Mn2+

Mg2+

Mg2+

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

light

scat

terin

g [A

U]

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

time [sec]

time [sec] time [sec]

FE

A B

Mn2+EDTA EDTA

1460

510

560

610

660

710

0 500 1000 1500 2000

EDTA

control

C D

0

20000

40000

60000

80000

640 650 660 670 680

Wavelength [nm]

Wavelength [nm]

Fluo

resc

ence

inte

nsity

[AU

]

0

10000

20000

30000

40000

50000

640 650 660 670 680

Fluo

resc

ence

inte

nsity

[AU

]

Wavelength [nm]

2 3 4 5 6

Mn2+

Ca2+

Zn2+

Zn2+

Zn2+

Co2+

Ni2+

Ni2+

Ni2+

Mn2+

Mg2+

Mg2+

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

light

scat

terin

g [A

U]

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

time [sec]


FE

A B

Mn2+EDTA EDTA

1460

510

560

610

660

710

0 500 1000 1500 2000

EDTA

control

C D

0

20000

40000

60000

80000

640 650 660 670 680

Wavelength [nm]

Wavelength [nm]

Fluo

resc

ence

inte

nsity

[AU

]

0

10000

20000

30000

40000

50000

640 650 660 670 680

Fluo

resc

ence

inte

nsity

[AU

]

Wavelength [nm]

2 3 4 5 6

Mn2+

Ca2+

Zn2+

Zn2+

Zn2+

Co2+

Ni2+

Ni2+

Ni2+

Mn2+

Mg2+

Mg2+

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

light

scat

terin

g [A

U]

0,05

0,07

0,09

0,11

0,13

0,15

0,17

0 500 1000 1500 2000

light

scat

terin

g [A

U]

time [sec]


FE

A B

Mn2+EDTA EDTA

1460

510

560

610

660

710

0 500 1000 1500 2000

EDTA

control

C D

0

20000

40000

60000

80000

640 650 660 670 680

Wavelength [nm]

Wavelength [nm]

Fluo

resc

ence

inte

nsity

[AU

]

0

10000

20000

30000

40000

50000

640 650 660 670 680

Fluo

resc

ence

inte

nsity

[AU

]

Wavelength [nm]

2 3 4 5 6

Mn2+

Figure 2. Binding to divalent cations enhances oligomerization of strep-SpoIIEcyt. (A, B) Light

scattering measurement of 1,5 μM strep-SpoIIEcyt (A) and MBP (B) in the presence of 10 mM of

divalent cations. (C, D) Fluorescence emission spectra of strep-SpoIIEcyt-Cy5 titrated into a buffer

in the absence (C) and presence (D) of Mn2+. (E) Light scattering signal of strep-SpoIIEcyt in the

presence of 10 mM Mn2+, followed by incubation with 20 mM EDTA or H2O (control). (F) SDS-PAGE

of the sedimentation of Ms-SpoIIEcyt experiment. Ms-SpoIIEcyt was pre-incubated with 1 mM

EDTA (lanes 1 and 2), followed by the incubation with 5 mM MnCl2 (lanes 3 and 4), followed by the

incubation with 10 mM EDTA (lanes 5 and 6). Odd numbers below the image represent supernatant

fractions and even numbers represent pellet fractions.

RESULTS CHAPTER 4

107

tra in the presence of metal (Fig. 2D) gave significantly lower signals comper-ing to the sample with EDTA (Fig. 2C) possibly due to binding of the protein to metal ions or formation of dimers. However, our goal was not to compare the spectra intensities between two experiments but to find the concentration in which fluorescence did not increase within one experiment. We noticed that the critical concentration of Ms-SpoIIEcyt-Cy5 was higher in the absence of Mn2+ (~2.1 µM) compering to the sample in the presence of Mn2+ (1.4 µM). In the ab-sence of divalent cation strep-SpoIIEcyt is also able to form oligomers, which is in agreement with the oligomerization observed during purification and previ-ous results 15. However, these oligomers are possibly smaller and not as stable as oligomers formed by strep-SpoIIEcyt in the presence of Mn2+.

MANGANESE-DEPENDENT OLIGOMERIZATION OF MS-SPOIIECYT

AND STREP-SPOIIECYT

IS REVERSIBLE

We wanted to see whether the observed increase in light scattering is due to the oligomerization rather than aggregation of strep-SpoIIEcyt. Therefore, we performed light scattering and sedimentation experiments, in which addition of manganese to the sample was followed by incubation with EDTA. As be-fore, the light scattering signal changed immediately after addition of MnCl2 to strep-SpoIIEcyt. Subsequent addition of EDTA to the sample caused an im-mediate decrease to the light scattering signal (Fig. 2E), which was not due to dilution or buffer effects as evidenced by the addition of an equal volume of buffer. The light scattering signal did not fully return to the background signal, probably because some manganese is trapped between monomers and can-not be released completely.

Next, Ms-SpoIIEcyt was incubated with EDTA to remove all metals bound to the protein during purification. Subsequently, the protein was incubated with Mn2+, followed by additional incubation with EDTA. Samples of all three stages were centrifuged and supernatants and pellets were analyzed by SDS-

-PAGE. The initial incubation with EDTA resulted in Ms-SpoIIEcyt in the superna-tant, which indicates that the protein is in monomeric or forms small oligomers (Fig. 2F, lanes 1 and 2). After Mn2+ addition, oligomers of Ms-SpoIIEcyt were re-covered in the pellet fraction (Fig. 2F, lane 4). This oligomerization could be re-


108

versed by incubation with EDTA, after which MBP-SpoIIEcyt was again present in the supernatant (Fig. 2F, lane 5). Combined, the light scattering and sedimen-tation experiments show that the metal-dependent oligomerization of SpoIIE is reversible and thus not caused by protein aggregation.

THE ABSENCE OF MANGANESE DELAYS POLAR Z-RING FORMATION

IN VIVO

It has been known for a long time that the sporulation process is strongly de-pendent on the presence of manganese, but so far this dependence has been attributed to the role of manganese as a cofactor in the phosphatase activity of SpoIIE, which is required for the activiation of the forespore specific transcrip-tion factor σF 1. Because the metal-dependent oligomerization of SpoIIE may be required for the function of SpoIIE in polar septation, we studied FtsZ-eYFP localization in sporulating B. subtilis cells in the presence and absence of man-ganese. In the absence of manganese, relocation of the Z-ring from mid-cell to the cell poles is delayed compared to samples supplemented with Mn2+. We noticed that cells without Mn2+ form more mid-cell rings and significantly less polar rings compared to the cells with Mn2+ (Fig. 3). It is clear that asymmetric septum formation is affected in the absence of manganese. Interestingly, the delay and decrease in the formation of asymmetric Z-rings is also observed in a spoIIE null mutant 8, suggesting that in the absence of manganese SpoIIE does not interact with FtsZ or cannot support FtsZ relocation to the cell pole. However, we cannot fully exclude indirect effects such as a decrease in expres-sion of SpoIIE in the absence of its cofactor, or another mechanism which is affected by the absence of Mn2+.

FTSZ POLYMERS ARE STABLE AND DO NOT DEPOLYMERIZE IN THE

PRESENCE OF MN2+

Next, we wanted to study the interaction between strep-SpoIIEcyt and FtsZ. Be-cause the activities of both proteins depend on divalent cations, we wanted to choose the best conditions to study the FtsZ-strep-SpoIIE interactions. The presence of Mg2+, which is a FtsZ cofactor, clearly influences the oligomeric

RESULTS CHAPTER 4

109

state of strep-SpoIIEcyt (Fig. 2A). To study whether the strep-SpoIIEcyt cofactor Mn2+ has any influence on the assembly of FtsZ, light scattering and EM were used. FtsZ assembly after addition of GTP in the presence of Mn2+ is similar to the assembly of FtsZ in the presence of Mg2+ (Fig. 4A). However, after prolonged incubation times (>30 min), the light scattering signal began to increase and after ~45 min reached the detection limit (Fig. 4B), with FtsZ precipitating out of the solution. To identify the cause of the precipitation the sample was an-alyzed using EM. In the presence of Mn2+, FtsZ forms stable polymers, which are not able to depolymerize even after 90 min of incubation (Fig. 4F), which is probably caused by the block in GTP hydrolysis. In the absence of polymer dynamics, polymers continuously assemble into long thick structures, which condense and precipitate from the solution. This is similar to FtsZ-SepF tubules,

0,0

5,0

10,0

15,0

20,0

25,0

30,0

35,0

40,0

45,0

50,0

+ - + - + - + -

no ring mid-cell ring two polar rings polar ring

Freq

uenc

y

average 2haverage 3h

A

Figure 3. The absence of Mn2+ in the sporulation medium delays the asymmetric Z-ring forma-

tion. Representation of counting of sporulating B. subtilis cells in the presence (+) and absence (-)

of Mn2+. Four different populations were found: cells without any Z-ring in the cell (no ring), cells

with clearly formed Z-ring in the middle of the rod (mid-cell ring), cells with two rings assembled

at the cell poles (two polar rings) and cells with only one polar ring (polar ring). Below the graph,

representative pictures of each of the population are present. More than 100 cells were counted

per each time point.


110

which are also very big structures and precipitate out of the solution when high GTP concentration is used (not shown).

FtsZ samples were also visualized after 30 min of incubation. In the pres-ence of Mg2+, FtsZ polymers were almost not visible or very short (Fig. 4E), which means they had disassembled in agreement with the decrease in light scattering signal (Fig. 4A). In the presence of Mn2+, FtsZ polymers are still long and form large bundles (Fig. 4C). Polymers formed in the absence of metal

1,51,551,6

1,651,7

1,751,8

1,851,9

1,952

0 500 1000 1500 20000

2

4

6

8

10

12

2000 2500 3000 3500 4000

Mn

Mg

EDTA

light

scat

terin

g [A

U]

light

scat

terin

g [A

U]


A B

C D E

F

Figure 4. (A) Light scattering signal of FtsZ in the presence of 2 mM GTP and 10 mM of divalent

cations (Mn2+, Mg2+) or 1 mM EDTA. (B) continuation of the signal from the sample with Mn2+. (C, D,

E) Electron microscopy of FtsZ polymers assembled in the presence of 2 mM GTP and either Mn2+

(C), EDTA (D) or Mg2+ (E) after 30 min of incubation. (D) Representative picture of FtsZ polymers

assembled in the presence of 2 mM GTP and Mn2+ after 90 min of incubation.

RESULTS CHAPTER 4

111

were still present after 90 min of incubation (Fig. 4D) but were not as dense as polymers formed in the presence of Mn2+ (Fig. 4C). Because the presence of Mn2+ has such a significant effect on the FtsZ properties, we decided to exclude it from the FtsZ-strep-SpoIIEcyt interaction studies.

DIRECT INTERACTION BETWEEN FTSZ AND STREP-SPOIIECYT

OCCURS IN THE PRESENCE OF GDP

A direct interaction between FtsZ and SpoIIE has been shown by Lucet et al. us-ing several techniques like gel filtration, non-denaturing PAGE and pull-down ex-periments 15. Further attempts to do study the FtsZ-SpoIIE interaction in vitro in more detail faltered because of difficulties in the purification of SpoIIE 16. To study the effect of strep-SpoIIE on FtsZ, the light scattering experiment of FtsZ in the presence of strep-SpoIIEcyt without manganese was initially performed. The ex-periment was performed in a buffer with high salt concentration as strep-SpoIIEcyt is more stable under those conditions. Strep-SpoIIEcyt enhanced the light scatter-ing of FtsZ both in the presence of GTP and GDP (Fig. 5A). Notably, the scattering increased to higher levels than the scattering signals of the individual proteins combined. Strep-SpoIIEcyt alone gave a small light scattering signal, because of the presence of MgCl2 in the sample, whereas the light scattering signal of FtsZ alone with GTP was very low because lateral interactions between FtsZ protofila-ments are diminished in the buffer conditions used 17,24. The increase in light scat-tering signal of FtsZ in the presence of strep-SpoIIEcyt had a significant lag phase, although assembly of FtsZ under these conditions should be immediate. To con-firm immediate polymerization, FtsZ was polymerized in the same conditions but using a different signal amplification (Fig. 5A, FtsZ enhanced). Intriguingly, the increase of the signal of FtsZ in the presence of strep-SpoIIEcyt is noticeable in the moment that FtsZ polymers have started to disassemble.

In our hands, cleavage of the MBP tag from Ms-SpoIIEcyt resulted in an un-stable protein that could not be obtained in high concentrations. Therefore, subsequent experiments with FtsZ were performed using Ms-SpoIIEcyt. In the presence of GDP, FtsZ does not sediment, but when Ms-SpoIIEcyt is present sig-nificant amounts of FtsZ can be found in the pellet (Fig. 5B). This was in contrast to the sample with GTP, as the amount of FtsZ pelleted did not clearly increase


112

in the presence of Ms-SpoIIEcyt. Control experiments with MBP-SepF124 (which does not interact with FtsZ) showed that FtsZ sedimentation in the presence of GDP is not due to the presence of the MBP tag (Fig 5A). Another experiment

light

scat

terin

g [A

U]

time [sec]

A

B

Ms-SpoIIEcyt

MBP-SepF124

FtsZ

GTP GTPGDP GDP

GTP GTPGDP GDPC

0

0,2

0,4

0,6

0,8

1

1,2

0 500 1000 1500 2000 2500 3000 3500 4000

FtsZ + strep-SpoIIEcyt

FtsZ + strep-SpoIIEcyt

strep-SpoIIEcyt

FtsZ

GTP

GDP

GTP

GTP

FtsZ enhancedGTP

1

Ms-SpoIIEcyt

His-EzrAcyt

2 3 4 5 6 7 8

1 2 3 4 5 6 7 8

Figure 5. Strep-SpoIIEcyt interacts directly with FtsZ in the GDP form. (A) Light scattering mea-

surement of either FtsZ or strep-SpoIIEcyt alone in the presence of GTP (FtsZ GTP or strep-SpoIIEcyt

GTP respectively), in the presence of GTP and strep-SpoIIEcyt (FtsZ+strep-SpoIIEcyt GTP), in the

presence of GDP and strep-SpoIIEcyt (FtsZ+strep-SpoIIEcyt GDP). Light scattering signal of FtsZ

alone in the presence of GTP using enhanced light detection to show assembly/ disassembly path

(FtsZ enhanced GTP). (B) Sedimentation assay of FtsZ in the presence of either Ms-SpoIIEcyt or

MBP-SepF124 and in the presence of either GTP or GDP. (C) Sedimentation of His-EzrAcyt alone

or in the presence of Ms-SpoIIEcyt with GTP or GDP. Odd numbers below the image represent

supernatant fractions and even numbers represent pellet fractions.

RESULTS CHAPTER 4

113

was performed with the control cell division protein His-EzrAcyt. We noticed that Ms-SpoIIEcyt did not increase sedimentation of His-EzrAcyt in the presence of either GTP or GDP (Fig. 5C). Thus, the sedimentation of FtsZ in the presence of GDP and MBP-strep-SpoIIEcyt is specific.

Taking together the light scattering and sedimentation results, it is tempt-ing to speculate that strep-SpoIIEcyt interacts with, and stabilizes, FtsZ in the GDP form. We do not think that FtsZ-GDP in the presence of Ms-SpoIIEcyt form higher order structures beyond FtsZ protofilaments (see below). The reason why FtsZ was recovered in the pellet fraction is that Ms-SpoIIEcyt is always pres-ent in the pellet after centrifugation in the presence of divalent cation. There-fore, Ms-SpoIIEcyt pulls FtsZ monomers/ short polymers to the pellet as a result of simple interaction. Also, the increase in the light scattering signal may be an effect of accumulation of FtsZ monomers/ short polymers on the surface of the strep-SpoIIEcyt oligomers, which together form bigger structures which scatter more light compared to strep-SpoIIEcyt alone (Fig. 5A).

INTERACTION BETWEEN MS-SPOIIECYT

AND FTSZ-GTP IS NOT DE-

TECTABLE IN OUR ASSAYS

We did not find any significant increase in the sedimentation of FtsZ when both GTP and Ms-SpoIIEcyt were present in the sample. However, the sedimen-tation assay is not very sensitive and it is possible that FtsZ and Ms-SpoIIEcyt interact but that structures formed by these proteins together are not enough big to be spun down in the centrifuge. To further investigate the interactions in the presence of GTP we studied the GTPase activity of FtsZ in the presence of Ms-SpoIIEcyt-apo (Ms-SpoIIEcyt preincubated with EDTA) and Ms-SpoIIEcyt-Fe (Ms-SpoIIEcyt in the presence of Fe bound during expression of the protein). The GTPase activity of FtsZ does not significantly change in the presence of Ms-SpoIIEcyt when the ratio FtsZ:Ms-SpoIIEcyt is 2:1 (the small increase observed is close to the error of the experiment, (Fig 6A). We also decided to study FtsZ:Ms-SpoIIEcyt interactions using EM. We checked the assembly of FtsZ-GTP in the presence of Ms-SpoIIEcyt and did not find any significant difference in the filaments formed other that a mixture of FtsZ filaments next to Ms-SpoIIEcyt structures (Fig 6B, C, D). Therefore we conclude that Ms-SpoIIEcyt does not in-


114

teract with FtsZ-GTP. When strep-SpoIIEcyt oligomerizes, the MBP tag, which is 40 kDa (almost half the size of the entire protein) might cover the binding re-gion for FtsZ. It is possible that FtsZ polymers are too big to enter the binding site of strep-SpoIIEcyt but FtsZ monomers or short polymers formed in the pres-ence of GDP are able to interact with strep-SpoIIEcyt. Another possibility is that Ms-SpoIIEcyt prefers to interact with FtsZ-GDP as was already shown for some other FtsZ partners like MinC 25.

DISCUSSION

SpoIIE is a bifunctional protein involved in asymmetric septum formation and activation of the forespore compartment-specific transcription factor σF 1. We

0

0,5

1

1,5

2

2,5

3

GTP

ase

activ

ity [P

i/Fts

Z/m

in]

apo Fe MBP

A

B C D

Figure 6. Ms-SpoIIEcyt does not interact with FtsZ in the presence of GTP. (A) GTPase activity of

FtsZ in the presence of either Ms-SpoIIEcyt-apo, without intrinsic metal (apo), Ms-SpoIIEcyt-Fe,

with Fe ion bound from the purification (Fe) or MBP. (B, C, D) Representative images of Electron

microscopy experiments with FtsZ and Ms-SpoIIEcyt in the presence of 2 mM GTP and either Mn2+

(B), EDTA (C) or Mg2+ (D) after 20 min of incubation.


115

purified the entire cytoplasmic portion of SpoIIE fused to MBP via a short strep-tag. This cytoplasmic portion consists of two domains that carry out the func-tions of SpoIIE: Domain II is involved in oligomerization and interaction with FtsZ 15 and domain III is the phosphatase (Fig. 1A), that dephosphorylates and activates SpoIIAA-P, which in turn can induce the release of σF which is com-plexed with SpoIIAB. Dephosphorylation of SpoIIAA is an enzymatic process which can be run independently by domain III and is dependent on the pres-ence of the cofactor, Mn2+ 14. Our analysis revealed that Ms-SpoIIEcyt binds two metal ions and that metal binding is involved in the reversible oligomerization of SpoIIEcyt. Because domain III was shown to be monomeric 14,15, we hypothe-size that binding of Mn2+ to domain III induces a conformational change which resulted in oligomerization of domain II of SpoIIE, or that there is an extra metal binding site on domain II that is involved in oligomerization. Oligomerization was also induced by other divalent cations, which is not surprising as the bind-ing specificity of enzymes for metal ions is quite low 26. In the crystal struc-ture of the phosphatase domain of SpoIIE, only a single manganese ion was found in the active site, which represented a fundamental difference between SpoIIE and the other type PP2C phosphatases, ex. human PP2Cα phosphatase of which active site binds two metal ions 14. The difference between our data and the previous work could be because in the work of Ledikov et al. Tris and Hepes buffers were used for incubation of SpoIIE with Mn2+ 14. As Tris has high affinity for divalent cations and Hepes clearly reacted with Mn2+ in our assays, we think the choice of buffer could be the cause of the low metal occupancy in the active site. It is also possible that domain II of SpoIIE is necessary for the stable metal binding by domain III. The other explanation is that the second metal can easily bind to SpoIIE during folding while it is not easy to introduce this metal into the protein structure while the protein is already folded, which was also observed in our analysis of the absorption spectrum of Ms-SpoIIEcyt after incubation with MnCl2. Our data clearly demonstrate that domain II and III are not fully independent and may influence each other’s activity, which is in agreement with the previous in vivo findings, where mutations in either of the domains had an influence on the activity of the other domains in vivo 7,10.

Manganese is important for the oligomerization and activity of SpoIIE but it has been known for a long time that it is also necessary for sporulation of


116

B. subtilis. It is known that Mn2+ is a cofactor for many enzymes involved in sporulation but it is not known at which stage is sporulation blocked in the absence of Mn2+. Because of the role of SpoIIE in asymmetric septation, we decided to study localization of FtsZ-eYFP during sporulation in the presence and absence of Mn2+. We noticed that in the absence of manganese, forma-tion of the asymmetric septum is delayed and less polar Z-rings are formed compared to cells sporulating in the presence of Mn2+. Only two proteins are known to be involved in the switch from medial to polar Z-ring, early sporula-tion-specific transcriptional factor Spo0A and SpoIIE. A mutation in the spo0A gene completely blocks sporulation at stage 0 before asymmetric septation occurs. Thus, it is unlikely that the lack of manganese influences Spo0A as it would completely block relocation of the Z-ring 8. Localization of FtsZ in the absence of Mn2+ during sporulation resembles the situation in spoIIE mutant cells. It has been shown previously that deletion of the spoIIE gene affects for-mation of asymmetric Z-rings but does not prevent it 8. Therefore, we conclude that the absence of manganese influences relocation of the Z-ring in a SpoIIE-

-dependent manner. However, we do not know whether our observations are the secondary effect of lower expression or misfolding of SpoIIE in the absence of Mn2+ or another mechanism which involves direct interactions between these two proteins. Because of lack of SpoIIE antibodies we could not check the levels of expression of SpoIIE in cells sporulating with and without manganese.

Our further analysis focused on the direct interaction between FtsZ and SpoIIEcyt. Until now only two works studied the direct interaction between FtsZ and SpoIIE in vitro. Lucet et al. showed a clear direct interaction between domain II of SpoIIE and FtsZ using gel filtration 15. A study by Rawlings et al. could not confirm these findings 16. Lucet et al. purified a SpoIIE fragment containing ami-no acids 326-827, whereas Rawlings et al. used two SpoIIE fragments starting at positions 376 and 412 respectively 15,16. Here, we used the same length of the cytoplasmic fragment of SpoIIE as Lucet et al. and confirmed the direct interac-tion between FtsZ and SpoIIEcyt using sedimentation and light scattering assays, which suggests that the constructs used by Rawlings et al., lacked a critical part of the protein for interaction with FtsZ. Our results clearly suggest that the inter-action between SpoIIEcyt and FtsZ occurs mostly in the presence of GDP, whereas in the work of Lucet et al. direct interaction using gel filtration was found in the


117

presence of GTP although the interaction was found between SpoIIE and mono-mers and very short oligomers (3-4 subunits) of FtsZ. It is possible that GTP was hydrolyzed during this the assay as Mg2+ was included in the gel filtration buffer, which would explain why FtsZ did not form large structures which should elute much earlier from the column 15. We did not find evidence for a specific interac-tion of polymerized FtsZ with Ms-SpoIIEcyt. It is possible that the large, 40 kDa MBP tag covers FtsZ binding site on SpoIIEcyt upon oligomerization of SpoIIEcyt. Therefore, the binding site is only accessible for FtsZ monomers and short oligo-mers. We do not exclude the possibility of interaction between FtsZ-GTP and strep-SpoIIEcyt as we noticed a significant increase in the light scattering signal when both proteins were present in the sample with GTP. However, the increase in the light scattering signal had a significant lag phase, which appeared to be related to the moment of the disassembly of FtsZ polymers. Therefore, we hy-pothesize that SpoIIEcyt preferentially interacts with FtsZ-GDP, possibly by stabil-ing FtsZ-GDP filaments and thus preventing filament disassembly.

REFERENCES

1. Feucht A, Magnin T, Yudkin MD, Errington J. Bifunctional protein required for asym-metric cell division and cell-specific tran-scription in Bacilllus subtilis. Genes Dev. 1996;10(7):794-803.

2. Eswaramoorthy P, Winter PW, Wawrzusin P, York AG, Shroff H, Ramamurthi KS. Asymmetric division and differential gene expression during a bacterial de-velopmental program requires DivIVA. PLoS Genet. 2014;10(8):e1004526.

3. Arigoni F, Guerout-Fleury AM, Barak I, Stragier P. The SpoIIE phosphatase, the sporulation septum and the establish-ment of forespore-specific transcrip-tion in Bacilllus subtilis: A reassessment. Mol Microbiol. 1999;31(5):1407-1415.

4. King N, Dreesen O, Stragier P, Pogliano K, Losick R. Septation, dephosphor-ylation, and the activation of sigmaF

during sporulation in Bacilllus subtilis. Genes Dev. 1999;13(9):1156-1167.

5. Wu LJ, Feucht A, Errington J. Prespore- -specific gene expression in Bacilllus subtilis is driven by sequestration of SpoIIE phosphatase to the prespore side of the asymmetric septum. Genes Dev. 1998;12(9):1371-1380.

6. Ben-Yehuda S, Losick R. Asymmetric cell division in B. subtilis involves a spiral-like intermediate of the cytokinetic protein FtsZ. Cell. 2002;109(2):257-266.

7. Carniol K, Ben-Yehuda S, King N, Losick R. Genetic dissection of the sporulation protein SpoIIE and its role in asymmet-ric division in Bacilllus subtilis. J Bacteriol. 2005;187(10):3511-3520.

8. Khvorova A, Zhang L, Higgins ML, Piggot PJ. The spoIIE locus is involved in the Spo0A-dependent switch in the


118

location of FtsZ rings in Bacilllus subtilis. J Bacteriol. 1998;180(5):1256-1260.

9. Clarkson J, Campbell ID, Yudkin MD. NMR studies of the interactions of SpoIIAA with its partner proteins that regulate sporulation in Bacilllus subtilis. J Mol Biol. 2001;314(3):359-364.

10. Feucht A, Abbotts L, Errington J. The cell differentiation protein SpoIIE con-tains a regulatory site that controls its phosphatase activity in response to asymmetric septation. Mol Microbiol. 2002;45(4):1119-1130.

11. Adler E, Donella-Deana A, Arigoni F, Pinna LA, Stragler P. Structural rela-tionship between a bacterial develop-mental protein and eukaryotic PP2C protein phosphatases. Mol Microbiol. 1997;23(1):57-62.

12. Barak I, Behari J, Olmedo G, et al. Struc-ture and function of the bacillus SpoIIE protein and its localization to sites of sporulation septum assembly. Mol Mi-crobiol. 1996;19(5):1047-1060.

13. Das AK, Helps NR, Cohen PT, Barford D. Crystal structure of the protein serine/threonine phosphatase 2C at 2.0 A res-olution. EMBO J. 1996;15(24):6798-6809.

14. Levdikov VM, Blagova EV, Rawlings AE, et al. Structure of the phosphatase domain of the cell fate determinant SpoIIE from Bacilllus subtilis. J Mol Biol. 2012;415(2):343-358.

15. Lucet I, Feucht A, Yudkin MD, Errington J. Direct interaction between the cell division protein FtsZ and the cell dif-ferentiation protein SpoIIE. EMBO J. 2000;19(7):1467-1475.

16. Rawlings AE, Levdikov VM, Blagova E, et al. Expression of soluble, active frag-ments of the morphogenetic protein SpoIIE from Bacilllus subtilis using a li-brary-based construct screen. Protein Eng Des Sel. 2010;23(11):817-825.

17. Krol E, Scheffers DJ. FtsZ polymeriza-tion assays: Simple protocols and con-

siderations. J Vis Exp. 2013;(81):e50844. doi(81):e50844.

18. Mukherjee A, Lutkenhaus J. Dynamic assembly of FtsZ regulated by GTP hy-drolysis. EMBO J. 1998;17(2):462-469.

19. Krol E, de Sousa Borges A, da Silva I, et al. Antibacterial activity of alkyl gallates is a combination of direct targeting of FtsZ and permeabilization of bacterial membranes. Front Microbiol. 2015;6:390.

20. Gundogdu ME, Kawai Y, Pavlendova N, et al. Large ring polymers align FtsZ polymers for normal septum forma-tion. EMBO J. 2011;30(3):617-626.

21. Sharpe ME, Hauser PM, Sharpe RG, Errington J. Bacillus subtilis cell cycle as studied by fluorescence microscopy: Constancy of cell length at initiation of DNA replication and evidence for ac-tive nucleoid partitioning. J Bacteriol. 1998;180(3):547-555.

22. Whittaker MM, Whittaker JW. Thermal-ly triggered metal binding by recom-binant thermus thermophilus manga-nese superoxide dismutase, expressed as the apo-enzyme. J Biol Chem. 1999; 274(49):34751-34757.

23. Foster AW, Osman D, Robinson NJ. Metal preferences and metallation. J Biol Chem. 2014;289(41):28095-28103.

24. Buske PJ, Levin PA. Extreme C terminus of bacterial cytoskeletal protein FtsZ plays fundamental role in assembly in-dependent of modulatory proteins. J Biol Chem. 2012;287(14):10945-10957.

25. Hernandez-Rocamora VM, Garcia- -Montanes C, Reija B, et al. MinC protein shortens FtsZ protofilaments by preferen-tially interacting with GDP-bound subunits. J Biol Chem. 2013;288(34):24625-24635.

26. Dudev T, Lim C. Competition among metal ions for protein binding sites: Determinants of metal ion selectivity in proteins. Chem Rev. 2014;114(1):538-556.

Antibacterial activity of alkyl

gallates is a combi-nation of direct

targeting of FtsZ and permeabiliza-tion of bacterial

membranesEwa Król1,4, Anabela de Sousa

Borges1,4, Isabel da Silva2, Carlos

Roberto Polaquini3, Luis Octavio

Regasini3, Henrique Ferreira2, Dirk-

Jan Scheffers1,*C

HA

PTER

1 Department of Molecular Microbiology, Groningen Biomolecular Sciences

and Biotechnology Institute, University of Groningen, The Netherlands; 2 Departamento de Bioquímica e Microbiologia, Instituto de Biociências,

Universidade Estadual Paulista, Rio Claro, SP, Brazil, 3Departamento de Química e Ciências Ambientais, Instituto de Biociências,

Letras e Ciências Exatas, Universidade Estadual Paulista, São José do Rio Preto,

SP, Brazil,

4 equal contribution

Published in: Front Microbiol. 2015 Apr 29;6:390

Experiments presented in Figures 2, 3, 4, 5, S2, S3 were performed by E. Cendrowicz. Figure S4 was performed by E. Cendrowicz and I. da Silva.

http://www.ncbi.nlm.nih.gov/pubmed/25972861

ABSTRACT

Alkyl gallates are compounds with reported antibacterial activity. One of the modes of action is binding of the alkyl gallates to the bacterial membrane and interference with membrane integrity. However, alkyl gallates also cause cell elongation and disruption of cell division in the important plant pathogen Xanthomonas citri subsp. citri, suggesting that cell division proteins may be targeted by alkyl gallates. Here, we use Bacillus subtilis and purified B. subtilis FtsZ to demonstrate that FtsZ is a direct target of alkyl gallates. Alkyl gallates disrupt the FtsZ-ring in vivo, and cause cell elongation. In vitro, alkyl gallates bind with high affinity to FtsZ, causing it to cluster and lose its capacity to po-lymerize. The activities of a homologous series of alkyl gallates with alkyl side chain lengths ranging from five to eight carbons (C5-C8) were compared and heptyl gallate was found to be the most potent FtsZ inhibitor. Next to the direct effect on FtsZ, alkyl gallates also target B. subtilis membrane integrity – how-ever the observed anti-FtsZ activity is not a secondary effect of the disruption of membrane integrity. We propose that both modes of action, membrane disruption and anti-FtsZ activity, contribute to the antibacterial activity of the alkyl gallates. We propose that heptyl gallate is a promising hit for the further development of antibacterials that specifically target FtsZ.


123

INTRODUCTION

The use of plants as sources of antimicrobial agents has a long history (Abreu et al., 2012). One group of plant compounds known for their antimicrobial activities are alkyl gallates, esters of gallic acid, a main product of tannin hydrolysis. Alkyl gallates have been shown to exhibit a broad spectrum of antibacterial activities against both Gram-positive and Gram-negative bacteria, including foodborne Salmonella (Kubo et al., 2002), Methicillin Resistant Staphylococcus aureus (MRSA) (Kubo et al., 2002; Shibata et al., 2005), Bacillus subtilis (Kubo et al., 2004), the plant pathogen Xanthomonas citri subsp. citri (Silva et al., 2013), and various others (Kubo et al., 2002; Kubo et al., 2003). Alkyl gallates with varying alkyl side chain lengths (C1-C14), have been studied as antibacterial agents alone or as modulators of the activities of β-lactams against MRSA (Kubo et al., 2004; Shibata et al., 2005; Kubo et al., 2002; Kubo et al., 2002; Kubo et al., 2003; Silva et al., 2013), a common cause of bloodstream infections in hospitals and healthcare facilities worldwide. The hydrolysis of alkyl gallates produces gallic acid and the corresponding alcohols (or alkanols), which both are common components in many plants.

Although the alkyl gallates have a head-and-tail structure similar to alkanols, suggesting that their antibacterial mode of action may be as surface-active agents affecting membrane integrity (Kubo et al., 2002; Takai et al., 2011), Kubo et al. pro-posed that their antimicrobial activity is unlikely to be due to their surfactant property (Kubo et al., 2002; Kubo et al., 2003; Kubo et al., 2002; Kubo et al., 2004). Recently, we showed that alkyl gallates are active against X. citri subsp citri (Xac), an important plant pathogen that is the causative agent of citrus canker, one of the most damaging infections in citriculture. Pentyl, hexyl, heptyl and octyl gal-late treatment resulted in elongated Xac cells and disruption of the cell division machinery in this bacterium (Silva et al., 2013). Octyl gallate has been reported to exhibit bactericidal activity only against dividing and exponentially growing cells of B. subtilis but did not affect the viability of cells in the stationary phase (Kubo et al., 2004). Taken together, these results indicate that alkyl gallates may affect func-tions associated with cell division in Gram-positive and Gram-negative bacteria (Kubo et al., 2004; Silva et al., 2013).

Cell division is a relatively novel target for antibacterial drugs (Kapoor, Panda 2009; Lock, Harry 2008; Huang et al., 2007). Division is an essential pro-


124

cess, which starts with the polymerization of the highly conserved cytoplasmic protein FtsZ in the middle of the cell leading to the formation of the so-called Z-ring (Erickson et al., 2010; Adams, Errington 2009). After assembly of the Z-ring, several other proteins are recruited to mid-cell, resulting in a complex called the divisome, which carries out cell division at the correct time and place in the cell. Formation of the divisome depends on the assembly of FtsZ. FtsZ belongs to the tubulin family of cytoskeletal GTPases. The binding of GTP to FtsZ promotes the assembly of FtsZ monomers into long filaments in vitro (Kapoor, Panda 2009). FtsZ is conserved among bacteria and is essential for cell viability, mak-ing it a potential target for new antibiotic discovery (Lock, Harry 2008; Kapoor, Panda 2009). Several natural, synthetic and semi-synthetic compounds were identified as inhibitors of FtsZ from Gram-positive and Gram-negative bacteria (Anderson et al., 2012; Lock, Harry 2008; Andreu et al., 2010; Beuria et al., 2005; Hemaiswarya et al., 2011; Rai et al., 2008; Keffer et al., 2013).

To establish whether alkyl gallates indeed target bacterial cell division, we characterized the mode of action of alkyl gallates with a side chain length ranging from five to eight carbons (Table 1) in more detail, using B. subtilis as a model. We show that B. subtilis FtsZ is a target for these esters and that some of these compounds bind FtsZ with high affinity, resulting in protein cluster formation and disruption of FtsZ structures in vitro and in vivo. Additionally, the alkyl gallates interfere with the stability of the cell membrane. FtsZ binding and inhibition and membrane integrity are differently affected based on the alkyl chain length. Our results indicate that alkyl gallate with a C7 side chain is the best hit for the further development of a FtsZ specific antibacterial with minimal effects on overall membrane integrity.


GENERAL

DNA manipulations including molecular cloning in Escherichia coli DH5α, PCR, DNA sequencing, restriction, ligation and transformation were performed stan-dard methods (Sambrook et al., 1989). Restriction enzymes, T4 DNA Ligase and


125

Phusion DNA polymerase were used as specified by the supplier (Fermentas). Both E. coli and B. subtilis were grown at 37 ºC on solid medium (LB Lennox plus agar 1.5% w/v) (Sezonov et al., 2007), and liquid medium (LB Lennox). When appropriate, ampicillin and spectinomycin were added to final concentrations of 100 and 50 µg/mL, respectively. Starch (Sigma Aldrich) was used at 0.1%. Primers are listed in Table 2. Plasmids and strains are listed in Table 3. B. subtilis 168 genomic DNA was isolated using the Wizard genomic DNA kit (Promega) according to the suppliers’ instructions.

Table 1. Structures of the alkyl gallates and their activity against B. subtilis 168.

C7H5O5-R R MIC90a (µg/mL) MIC50

a (µg/mL) MIC24Hb (mg/mL)

Compound 8 pentyl gallate (CH2)4CH3 235 141 400

Compound 9 hexyl gallate (CH2)5CH3 90 65 100

Compound 10 heptyl gallate (CH2)6CH3 50 20 50

Compound 11 octyl gallate (CH2)7CH3 105 75 25

a Determined by the REMA assayb Determined as minimum concentration that shows no growth after 24 h in a 2-fold dilution series.

Table 2. Primers used for cloning.

Name Sequence (5’−3’) Location

AB7 GCAGCGCTAGCATTACTTGTACAGCTCGTCCAT-GCCGAG reverse primer for eyfp with NheI site

AB10 GCCGCAGAATTCATGTTGGAGTTCGAAAC forward primer for ftsZ with EcoRI site

AB55 tagcatggatccggcggcggcggctccggtggtggtggttc-cggcggcggcggcATGGTGAGCAAGGGCGAGG

forward primer for eyfp with flexible linker sequence and BamHI site

AB56 tagcatggatccGCCGCGTTTATTACGGTTTC reverse primer for ftsZ with BamHI site

Restriction sites used to clone are in bold. Initiation or termination codons are underlined.


126

PLASMID AND STRAIN CONSTRUCTION

To generate vector pDJ108, a PCR product containing the coding sequence of eyfp (enhanced Yellow Fluorescent Protein) was amplified from vector pMK13, using primer AB7/AB55 (Table 2). The PCR product was sequenced and subse-quently digested with BamHI/NheI. ftsZ was amplified from the genomic DNA of B. subtilis 168 using primer AB10/AB56 (Table 2). The PCR product was sequenced and subsequently digested with EcoRI/BamHI, and ligated with the digested eyfp PCR product and EcoRI/NheI digested pDOW01 resulting in plasmid pDJ108, which was verified by sequencing. Subsequently, B. subtilis 168 was transformed with pDJ108, and correct integration at the amyE locus was verified on starch plates (Harwood, Cutting 1990). A positive colony was selected, expression and localization of the fusion protein FtsZ-eYFP was verified by microscopy (as de-scribed below for in vivo Z-ring analysis) resulting in strain 4055.

ALKYL GALLATES (COMPOUNDS 8−10)

Alkyl gallates with side chains varying from five to eight carbons [pentyl (com-pound 8), hexyl (compound 9), heptyl (compound 10) and octyl (compound 11) gallates] were synthetized as described (Silva et al., 2013). Compound num-bers were chosen to keep in line with our previous report (Silva et al., 2013).

MIC ASSAYS

REMA

The antibacterial activity of alkyl gallates was tested by the standard resazurin microtiter assay (REMA) plate method with some modifications (Silva et al., 2013; Palomino et al., 2002). Briefly, an overnight culture of B. subtilis was grown in 5 mL LB medium. The overnight culture was diluted with LB medium and distributed into a 96-well microtiter plate to a final volume per well of 100 μL (105 CFU/well). The alkyl gallates were serially diluted with LB (1000 μg/mL to 15 μg/mL), and 100 μL dilutions were added to the 96-well microtiter plate. Ka-namycin (50 μg/mL) was used as control antibiotic. The plate was subsequently incubated at 37 °C for 4 h. After 4 h, 15 μL of a 0.01% (wt/vol) freshly prepared


127

resazurin solution in H2O was added to each well and the plate was incubat-ed for 1 hour. Subsequently, the fluorescence in each well was measured in a Synergy Mix Microplate Reader (BioTek), with excitation and emission wave-lengths at 530 nm and 590 nm, respectively. All the experiments were done in triplicate and MIC50 and MIC90 values were calculated using a nonlinear regres-sion approach (Silva et al., 2013).

DILUTION

Cells from an exponentially growing B. subtilis culture were inoculated at OD600 of 0.2 in LB medium, at 37 °C, with alkyl gallates in 2-fold dilution series. After 24 h, the lowest alkyl gallate concentration at which no visible growth occurred was scored as MIC24H.

GENERAL MICROSCOPY ANALYSIS

Cells were resuspended in small volumes of CH medium (Sharpe et al., 1998) or Phosphate Buffered Saline (PBS, 58 mM Na2HPO4; 17 mM NaH2PO4; 68 mM NaCl, pH 7.3), and mounted on an agarose pad (1% w/v in PBS). Cells were imaged us-ing a Nikon Ti-E microscope (Nikon Instruments, Tokyo, Japan) equipped with a Hamamatsu Orca Flash4.0 camera. Image analysis was performed using the software packages ImageJ (http://rsb.info.nih.gov/ij/) and Adobe Photoshop (Adobe Systems Inc., San Jose, CA, USA).

IN VIVO Z-RING ANALYSIS

An overnight culture of B. subtilis 4055 grown in LB with spectinomycin at 37°C was diluted 1:100 into fresh LB with 0.02 mM IPTG to express ftsZ-eyfp. When the culture reached an OD600 of 0.4, 1 mL culture was centrifuged (1 min, 22,000 ×g), resuspended in 100 µl LB, and incubated at 37°C in the presence of either dimethyl sulfoxide (DMSO, 2% v/v), nisin (1.5 mg/mL), carbonyl cyanide m-chlorophenyl hydrazone (CCCP, 0.2 mM) or alkyl gallates at MIC50 concentra-tions (see Table 3). After either 2 min or 15 min of incubation, the cells were col-lected (1 min, 22,000 ×g), resuspended in 50 µl of PBS and mounted on agarose pads for microscopy analysis.


128

Table 3. Plasmids and strains.

Name Relevant characteristics Source

plasmids

pMK13 ampR, bgaA’, tetR, PZn-MCS-linker-yfp, ’bgaA Kjos and Veening, unpublished

pDOW01 pDR111 derivative with pUC18 ori, amyE::spc Phyperspac

Ruud Detert Oude Weme, unpub-lished

pDJ108 amyE::spc Phyperspac-ftsZ-eyfp This work

pEK5 lacZa, trc HisTag-enterokinase site- ezrA27-562 (Krol et al., 2012)

B. subtilis strains

168 trpC2 Bacillus Genetic Stock Center (www.bgsc.org)

4055 trpC2 amyE::spc Phyperspac-ftsZ-eyfp This work

E. coli strains

DH5α Life Technologies

Spc, Spectinomycin resistance

BRIGHTFIELD MICROSCOPY

An overnight culture of B. subtilis 168 grown in LB at 37°C was diluted 1:100 into fresh LB until early exponential phase, OD600 of 0.3. 3 mL samples of the culture were taken and incubated at 37°C in the presence of 2 µg/mL 3-((6-chlorothi-azolo[5,4-b]pyridin-2-yl)methoxy)-2,6-difluorobenzamide (PC190723, Merck), and alkyl gallates at various concentrations . After 1h, 2h or 3h of continued growth in the presence of the compounds, a sample of 1 mL was collected (1 min, 22,000 g) and cells were fixed by the addition of 1 mL of 8% formal-dehyde and incubation at 23°C for 60 min. Subsequent to fixation, cells were washed twice in PBS and resuspended in 10 to 30 µL of PBS before mounting on agarose pads for microscopy analysis.

MEMBRANE PERMEABILITY ASSAY

An overnight culture of B. subtilis 168 grown in LB was diluted 1:100 into fresh LB and grown at 37°C until exponential phase, OD600 of 0.5. A sample of 1 mL was collected and cells were resuspended in 50 µL of CH medium (Sharpe et


129

al., 1998). Membrane integrity was assessed using the commercial assay Live/Dead BacLight bacterial viability kit (Invitrogen) for microscopy, according to the manufacturer’s instructions. The dyes propidium iodide (20 mM) and SYTO 9 (3.34 mM) were combined in equal amounts and 0.15 µL of the dye mixture was used to stain the DNA of cells resuspended in 50 µL of CH me-dium. Cells were incubated for 15 min. at 23°C in the presence of DMSO (2%), Nisin (2.5 mg/mL), CCCP (0.2 mM) and alkyl gallates at MIC50 and MIC90 (see Table 3). After incubation, cells were mounted on agarose pads for microscopy analysis and the green and red fluorescence were imaged. In total, two inde-pendent experiments were performed. Per experiment, more than 250 cells for each condition were scored for green (intact membrane) or red (permeabilized membrane) fluorescence. The values were converted into percentages and the mean and standard deviation was plotted on a graph using Excel.

PROTEIN EXPRESSION AND PURIFICATION

B. subtilis FtsZ was expressed and purified using the ammonium sulfate pre-cipitation method as described before (Mukherjee, Lutkenhaus 1998; Krol, Scheffers 2013). His-EzrAcyt was expressed as described (Krol et al., 2012). For purification, pellet from one liter of culture was resuspended in 20 mL of Buffer A (50 mM tris(hydroxymethyl)aminomethane [Tris]/HCl pH=7.5, 250 mM NaCl, 10 mM imidazole) supplemented with a EDTA (ethylenediaminetetraacetic acid)-free protease inhibitor tablet (Roche). The cells were disrupted at 18 kpsi (Constant systems OneShot disruptor, LA biosystems) and the insoluble frac-tion was removed by centrifugation at 7,000 ×g for 20 minutes at 4°C. Super-natant was applied onto Ni-NTA Agarose resin (Qiagen). The resin was washed with Buffer A containing 25 mM imidazole and His-EzrAcyt protein was eluted with the same buffer containing 300 mM imidazole. Samples were dialyzed and stored at -80°C in a buffer containing 20 mM Tris/HCl, pH=7.5, 250 mM NaCl, 10 % glycerol.


130

GTPASE ASSAY

The FtsZ GTP hydrolysis rate was determined using the malachite green phos-phate assay described in (Krol, Scheffers 2013) with the following modifica-tions. Two fold concentrated stocks of alkyl gallates with FtsZ were prepared in polymerization buffer (50 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid [Hepes]/NaOH, pH=7,5; 300 mM KCl; 0.02 % Triton X-100) and incubated for 5 minutes at 30 °C. After that, 2 mM GTP dissolved in 50 mM Hepes/NaOH, pH=7,5; 300 mM KCl was added. The final concentrations of GTP and Triton X-100 in the sample were 1 mM and 0.01%, respectively.

BINDING OF ALKYL GALLATES TO FTSZ

Binding of alkyl gallates to FtsZ was assessed by monitoring the increase of alkyl gallate fluorescence upon binding to protein, analogous to (Takai et al., 2011). Alkyl gallates at constant concentration were incubated with increasing amounts of FtsZ (0,6-14,4 µM). After addition of protein, the solution was in-cubated for 3 minutes at room temperature to allow equilibration of binding. Fluorescence was excited at 271 nm and emission spectra (320-450 nm) were acquired in a QuantaMasterTM spectrofluorometer controlled by the FelixGX program (Photon Technology International, Inc.). The fluorescence spectra of corresponding blanks resulting from titration of FtsZ into buffer alone were recorded and subtracted from the respective data sets. All measurements were done in polymerization buffer: 50 mM Hepes/NaOH pH=7,5, 50 mM KCl in the absence of GTP. The change in fluorescence of alkyl gallates upon binding to FtsZ was used to determine the dissociation constant (Kd) of the interaction between the compounds and FtsZ as described in (Kelley et al., 2012). The dis-sociation constant was determined using MATLAB (The MathWorks, Inc., Mas-sachusetts, U.S.A.).

FTSZ SEDIMENTATION ASSAYS

FtsZ (10 µM) was mixed with the alkyl gallates or an equal volume of DMSO as control at 50 or 100 µg/mL in the polymerization buffer (50 mM Hepes/ NaOH,


131

pH=7.5, 50 mM KCl) supplemented with 10 mM MgCl2. After incubation for 5 minutes at 30 °C, an equal volume of GTP, GDP or polymerization buffer was added (final concentration of nucleotides 2 mM). The samples were incubated for another 20 minutes and centrifuged at 186,000 ×g or at 350,000 ×g when indicated for 10 minutes at 25 °C. Pellet and supernatant fractions were ana-lyzed by sodium dodecyl sulfate - poly-acrylamide gel electrophoresis (SDS-PAGE) as described (Krol, Scheffers 2013). FtsZ interacting protein His-EzrAcyt and bovine serum albumin (BSA) (both at 10 µM) were used as controls.

ELECTRON MICROSCOPY

Electron microscopy samples were prepared essentially as described for the sedimentation assays. FtsZ (10 µM) was mixed with alkyl gallates at 50 µg/mL in polymerization buffer (50 mM Hepes/ NaOH, pH=7.5, 50 mM KCl, 10 mM MgCl2), incubated for 5 minutes at 30 °C with shaking (300 rpm), after which GTP was added to a final concentration of 2 mM. The samples were incubated at 30 °C for another 20 minutes and the polymerization mixture was applied to an electron microscopy grid as described in (Krol, Scheffers 2013). To test the effect of alkyl gallates on pre-polymerized FtsZ, the protein was incubated at 30 °C in polymerization buffer (25 mM piperazine-N,N΄-bis(2-ethanesulfon-ic acid) [PIPES]/NaOH pH=6.8, 300 mM KCl, 10 mM MgCl2) and polymerized for 2 minutes with 2 mM GTP – this results in maximal polymerization as de-termined by light scattering (Krol, Scheffers 2013). After 2 min, alkyl gallates (50 µg/mL) were added and the samples were incubated for another 6 minutes.

Samples were collected at two times, immediately after alkyl gallates were added and after 6 minutes of incubation with alkyl gallates, and grids were prepared as described in (Krol, Scheffers 2013). The grids were examined in a Philips CM120 electron microscope equipped with a LaB6 filament operating at 120 kV. Images were recorded with a Gatan 4000 SP 4 K slow-scan CCD cam-era at 36,400× magnification.


132

RESULTS

ALKYL GALLATES WITH CARBON SIDE CHAIN LENGTH C5-C

8 INHIBIT

B. SUBTILIS GROWTH

Recently, we showed that pentyl (compound 8), hexyl (compound 9), heptyl (compound 10) and octyl (compound 11) gallates disrupt cell division in Xac. Delocalization of GFP-ZapA from the cell division site suggested that the FtsZ-ring is a target of these compounds (Silva et al., 2013). To further study the mode of action of these compounds we tried to overexpress and purify Xac FtsZ, but failed to do so (data not shown). Therefore, we chose to use B. subtilis FtsZ for a more detailed characterization of the mechanism of action of alkyl gallates. First, we tested the antibacterial activity of compounds 8-11 against B. subtilis 168, using the resazurin microtiter assay (REMA). This assay deter-mines the concentrations at which compounds block bacterial metabolic ac-tivity over a period of 4 hours. Using this assay, these compounds were found to block cell activity with MIC90 concentrations ranging from 50 µg/mL to 235 µg/mL (Table 1). In the REMA assay, compound 10 (C7) was the most potent (Table 1). In an earlier report, Kubo and co-authors found similar MIC values for compounds 9 and 10, and a lower MIC for compound 11 against B. subtilis strain ATCC 9372 (12,5 µg/mL) using the macrodilution method (Kubo et al., 2004). Since compound 10 was the most potent against B. subtilis and alkyl gal-lates with longer and shorter side chains than compounds 8-11 did not disrupt septum formation in Xac (Silva et al., 2013), we focused on compounds 8-11 in our study.

ALKYL GALLATES DISRUPT Z-RING FORMATION IN VIVO

To study whether FtsZ ring formation is the target of the alkyl gallates, we used a B. subtilis strain (4055) that expresses an additional copy of ftsZ fused to the fluorescent protein eYFP from the ectopic amyE locus. In this strain, the FtsZ-eYFP fusion protein localized to the division site at mid-cell (Fig. 1) and mid-cell localization was not affected by DMSO, which was used as a solvent for the compounds. Incubation with alkyl-gallates at MIC50 interfered with the

RESULTS CHAPTER 5

133

formation or stability of the Z-ring. After 2 mins of incubation, the FtsZ-eY-PF fluorescence pattern was mostly found spread throughout the cell, even though some Z-rings could still be detected (most noticeably after incubation with compound 9 and compound 11). After 15 minutes of incubation the effect was more pronounced, with fluorescence throughout the cytosol, sometimes

DMSO

8

9

10

11

2 min 15 min

Figure 1. Alkyl gallates disrupt the Z-ring. B. subtilis cells expressing ftsZ-eyfp were incubated with DMSO (2%) or alkyl gallate compounds at MIC50 for 2 min. (left columns) or 15 min. (right columns). Brightfield and fluorescence microscopy images are shown. Incubation with alkyl gallates led to the disappearance of Z-rings and increase of fluorescence in the cytoplasm. Scale bar (same for all): 5 µm.

Figure 1. Alkyl gallates disrupt the Z-ring. B. subtilis cells expressing ftsZ-eyfp were incubated with

DMSO (2%) or alkyl gallate compounds at MIC50 for 2 min. (left columns) or 15 min. (right columns).

Brightfield and fluorescence microscopy images are shown. Incubation with alkyl gallates led to

the disappearance of Z-rings and increase of fluorescence in the cytoplasm. Scale bar (same for

all): 5 μm.


134

with occasional fluorescent spots, but hardly any Z-rings. The disruption of the Z-ring seemed to be fast, although it became more evident with an increase of incubation time. The occasional observance of rings was consistent with the fact that the cells were incubated with compounds at MIC50, meaning that it was expected that not every cell was affected. To confirm that the disappear-ance of the FtsZ rings is not caused by a generic loss of membrane integrity, the effect of membrane potential dissipation and membrane pore formation on FtsZ rings was determined using carbonyl cyanide m-chlorophenylhydra-zone (CCCP) and Nisin (Strahl, Hamoen 2010). After incubation of the strain 4055 with CCCP and nisin, various Z-rings localized at the mid-cell (Figure S1), although rings are less bright after CCCP treatment as reported before (Strahl, Hamoen 2010). The presence of Z-rings after CCCP and nisin treatments indi-cate that the disappearance of the FtsZ-rings caused by the alkyl gallates is not caused by a generic loss of membrane integrity.

GTP

ase

activ

ity (P

i/Fts

Z/m

in)

DMSO 8 9 10 110

1

2

0,5

3

1,5

2,5

Figure 2. Alkyl gallates inhibit FtsZ GTPase activity. FtsZ GTPase activity was determined as described in Materials and Methods. Compound concentrations were 50 µg/ml, 1.5 % DMSO (compound vehicle) was used as a control. Tree independent experiments were performed, average and standard deviation are shown.

Figure 2. Alkyl gallates inhibit FtsZ GTPase activity. FtsZ GTPase activity was determined as de-

scribed in Materials and Methods. Compound concentrations were 50 μg/mL, 1.5 % DMSO (com-

pound vehicle) was used as a control. Three independent experiments were performed, average

and standard deviation are shown.

RESULTS CHAPTER 5

135

ALKYL GALLATES INHIBIT GTPASE ACTIVITY OF FTSZ

The disruption of the Z-ring in cells treated with alkyl gallates suggested that FtsZ is a direct target for these compounds. To test this we made use of the polymer-ization-associated GTP hydrolysis activity of FtsZ. GTP hydrolysis is often used to screen for FtsZ inhibitors from a compound library (Anderson et al., 2012). The GTPase activity of FtsZ in the presence of 50 µg/mL of the alkyl gallates was re-duced nearly 6-fold compared to the control sample. However, residual GTPase activity was still detected in all of the samples, indicating that FtsZ was not com-pletely inactive (Fig. 2). All of the compounds showed similar levels of inhibition of the GTPase activity of FtsZ. The GTP hydrolysis assay was performed in the presence of Triton X-100 as it has been reported that some compounds identi-fied in high-throughput screens form small aggregates that inhibit FtsZ activity non-specifically. This aspecific inhibition is abolished by the inclusion of Triton X-100 in the assay (Anderson et al., 2012). The experiment was also performed in the absence of Triton X-100 and similar results were obtained (not shown). Com-bined, this indicates that the effect of the alkyl gallates on FtsZ hydrolysis was specific and not caused by aggregation of the compounds.

FTSZ IS A DIRECT TARGET FOR ALKYL GALLATES

The binding of alkyl gallates to FtsZ was monitored using the intrinsic fluores-cence of the compounds. FtsZ was titrated into a quartz cuvette containing a compound at fixed concentration (see Materials and Methods) and fluores-cence emission spectra of the alkyl gallates were recorded with the excitation wavelength set at 271 nm as described in (Takai et al., 2011). Compounds 10 and 11 showed strong binding to FtsZ: the fluorescence emission maximum shifted from 389 nm to 366 nm and the maximum signal intensity increased upon FtsZ addition (Fig. 3A, B). The fluorescence of alkyl gallates was plotted as a function of FtsZ concentration and the dissociation constant was deter-mined using 1:1 binding formalism described in (Kelley et al., 2012). The best fits (R2=0.99) were obtained for a fixed concentration of compounds at 3.3 µM. The estimated Kd values obtained were 0.08 ± 0.03 µM for compound 10 and 0.84 ± 0.22 µM for compound 11 (Fig. 3C, D).


136

Tryptophan and tyrosine fluorescence are also excited at 271 nm. How-ever, FtsZ does not contain tryptophan and the emission maximum of tyro-sine occurs at 303 nm and does not change according to solvent polarity. The emission maximum of alkyl gallates occurs at 389 nm. We observed that ty-rosine emission does not change upon binding of FtsZ to compound 10 and 11. However, addition of compound 8 and 9 significantly quenched tyrosine fluorescence of FtsZ. During analysis, the fluorescence spectra obtained for the compounds in the presence of FtsZ were corrected for protein fluorescence by subtracting the spectra of samples containing only FtsZ. The change in the FtsZ emission spectra due to tyrosine quenching made proper analysis of bind-ing of compound 8 and 9 to FtsZ very difficult (Fig. S2A, B). The resulting shift in emission maximum for compounds 8 and 9 was severely reduced compared to compounds 10 and 11. For compound 8, an estimated Kd of 3.1±2.0 µM could be calculated, with a worse fit (R2=0.83) than was obtained for compounds 10

Fluo

resc

ence

Inte

nsity

(a.u

.)

Wavelength (nm)

Fluo

resc

ence

Inte

nsity

(a.u

.)

Wavelength (nm)

compound 10 compound 11A B

C D

250000

250000

320 340 360 380 400 420 440 320 340 360 380 400 420 440

300000

250000

200000

150000

100000

50000

0

-5000

300000

250000

200000

150000

100000

50000

0

350000

400000

[FtsZ] (µM) [FtsZ] (µM)2 4 6 8 100 2 4 6 8 100 12

350000

300000

250000

200000

150000

100000

50000

0

300000

250000

200000

150000

100000

50000

0

350000Fl

uore

scen

ce In

tens

ity (a

.u.)

Fluo

resc

ence

Inte

nsity

(a.u

.)

Figure 3. Binding of compounds 10 and 11 to FtsZ monitored by fluorescence spectroscopy. Fluorescence emission spectra of compound 10 (A) and 11 (B) acquired in the presence of 0, 1.2, 2.4, 3.6, 4.8 µM FtsZ (from bottom to top). (C, D) Binding curves of compounds 10 (C) and 11 (D) to FtsZ. The change in fluorescence intensity at 366 nm was plotted against FtsZ concentration (from 0 to 9.6 µM for compound 10 and from 0 to 12 µM for compound 11).

Figure 3. Binding of compounds 10 and 11 to FtsZ monitored by fluorescence spectroscopy. Flu-

orescence emission spectra of compound 10 (A) and 11 (B) acquired in the presence of 0, 1.2, 2.4,

3.6, 4.8 μM FtsZ (from bottom to top). (C, D) Binding curves of compounds 10 (C) and 11 (D) to FtsZ.

The change in fluorescence intensity at 366 nm was plotted against FtsZ concentration (from 0 to

9.6 μM for compound 10 and from 0 to 12 μM for compound 11).

RESULTS CHAPTER 5

137

and 11 (Fig. S2C). For compound 9, an estimated Kd could not be calculated. Combined, these results show that compounds 10 and 11 specifically bind FtsZ, with compound 10 having the highest affinity. The quenching observed with compounds 8 and 9 indicates interaction of the compounds with FtsZ but the low shifts in emission maximum and the difficulties in Kd estimation indi-cate that the binding of compounds 8 and 9 to FtsZ may be aspecific.

SEDIMENTATION OF FTSZ IS ENHANCED IN THE PRESENCE OF ALKYL

GALLATES

Next, we used a sedimentation assay to study the effect of alkyl gallates on the assembly of FtsZ in vitro. FtsZ was mixed with the alkyl gallates (at 50 µg/mL) or 1,5 % DMSO and polymerization was started by addition of GTP or GDP (control) to the sample. We noticed that in the presence of compounds 8, 10 and 11, FtsZ was recovered in the pellet fraction above background levels in-dependent on the presence and type of nucleotide used (Fig. 4A). This result suggested that the compounds induce protein clustering or aggregation. As both the FtsZ interacting protein His-EzrAcyt (Fig. 4A) and BSA (not shown) did not sediment in the presence of the compounds, the observed FtsZ sedimen-tation is not the result of aspecific protein aggregation. Compound 10 has the strongest effect on FtsZ sedimentation (all FtsZ protein was present in the pel-let fraction), whereas approximately 50 % of FtsZ protein was recovered in the pellet fraction after treatment with compounds 8 or 11. No sedimentation was detected when FtsZ was incubated with compound 9 (Fig. 4A). However, upon higher sedimentation speed (350 000 ×g) or at higher compound concentra-tion (100 µg/mL), FtsZ was recovered above background levels also with com-pound 9 and 100 % of FtsZ was present in the pellet fraction after treatment with compounds 8, 10 and 11 (Fig. S3). The strength of the effects of the vari-ous alkyl gallates on FtsZ sedimentation are in line with the estimated Kds for the different compounds.

The structures formed by FtsZ treated with alkyl gallates were visualized by electron microscopy (EM). As expected, FtsZ formed clusters after incubation with all the alkyl gallates (Fig. 4B).


138

FtsZ EzrA

no nucl GTP GDP no nucl

S S S SP P P P

DMSO

8

9

10

11

8 9

A B C

D

10 11

E

DMSO

A

B

Figure 4. Sedimentation of FtsZ in the presence and absence of alkyl gallates and structures formed after treatment with alkyl gallates. (A) 10 µM FtsZ was incubated with alkyl gallates (50 µg/ml) before GTP/GDP was added. In the sample without nucleotide (no nucl), the same volume of polymerization buffer was added to the sample as for GTP/GDP. His-EzrAcyt was sedimented without nucleotide. (S) indicates supernatant and (P) pellet fractions from the experiment. As a control, 1.5 % DMSO was used. All experiments were performed in triplicate. (B) The same experiment was performed, with GTP, and structures of FtsZ were visualized by EM. Scale bar: 100 nm.

Figure 4. Sedimentation of FtsZ in the presence and absence of alkyl gallates and structures

formed after treatment with alkyl gallates. (A) 10 μM FtsZ was incubated with alkyl gallates

(50 μg/mL) before GTP/GDP was added. In the sample without nucleotide (no nucl), the same

volume of polymerization buffer was added to the sample as for GTP/GDP. His-EzrAcyt was sedi-

mented without nucleotide. (S) indicates supernatant and (P) pellet fractions from the experiment.

As a control, 1.5 % DMSO was used. All experiments were performed in triplicate. (B) The same ex-

periment was performed, with GTP, and structures of FtsZ were visualized by EM. Scale bar: 100 nm.

RESULTS CHAPTER 5

139

DMSO

0 min 6 min

10

11

A B

C D

E F

8

G H

9

I J

Figure 5. Structures of FtsZ formed in the presence and absence of alkyl gallates. 10 µM FtsZ was polymerized with 2 mM GTP for 2 minutes. After polymers were formed, alkyl gallates or DMSO (1.5 % w/v) were added and immediately sample 0 was collected (0 min, A, C, E, G, I). After 6 minutes of incubation, another sample was collected (6 min, B: DMSO, D: compound 8, F: compound 9, H: compound 10, J: compound 11). Scale bar: 100 nm.

Figure 5. Structures of FtsZ formed in the presence and absence of alkyl gallates. 10 μM FtsZ was

polymerized with 2 mM GTP for 2 minutes. After polymers were formed, alkyl gallates or DMSO

(1.5 % w/v) were added and immediately sample 0 was collected (0 min, A, C, E, G, I). After 6 min-

utes of incubation, another sample was collected (6 min, B: DMSO, D: compound 8, F: compound

9, H: compound 10, J: compound 11). Scale bar: 100 nm.


140

ALKYL GALLATES PROMOTE CLUSTERING OF FTSZ AND BUNDLING

OF FTSZ POLYMERS

Alkyl gallates caused the clustering of FtsZ and only a small amount of short polymers was detectable in the sample when GTP was added. When FtsZ was incubated with the respective compounds at MIC90 values, only big protein clusters and aggregates were observed (Fig S4). We assume that FtsZ was not able to polymerize because the clustering of FtsZ prevents the correct associ-ation of FtsZ molecules required for polymerization. To establish whether the alkyl gallates disrupt existing polymers, we performed an experiment in which polymerization of FtsZ was initiated before the addition of compounds to the sample. FtsZ was polymerized in a PIPES/KCl buffer that ensures optimal po-lymerization as determined by light scattering (Krol, Scheffers 2013), and the compounds or DMSO were added to polymerized FtsZ. Samples were collect-ed immediately after the addition of compounds and after an additional 6 min-utes of incubation. Compounds 8, 10 and 11 bound to the polymers of FtsZ and induced the formation of irregular bundles (Fig. 5). The amount of bundles that were observed in samples treated with compounds 10 and 11 was much higher than for compound 8 – although it has to be noted that this method is not quantitative. In the samples with compound 9, bundles were almost not visible (Fig. 5). We could only detect a few small bundle-like structures under the conditions used. The presence of a high number of tubules in the samples with compounds 10 and 11 suggests that the compounds can easily bind to the polymeric form of FtsZ, whereas binding of compound 8 and 9 to polymers of FtsZ is less strong – again, this is in line with the overall stronger FtsZ-bind-ing of compounds 10 and 11. Although the alkyl gallates induce the formation of clusters of FtsZ monomers that as a result no longer form polymers, the compounds do not disrupt existing FtsZ polymers.

MEMBRANE INTEGRITY AND CELL VIABILITY ARE AFFECTED BY AL-

KYL GALLATES

Cells in which the FtsZ activity is compromised, either through mutation or by the addition of FtsZ targeting compounds, display cell elongation and fila-

RESULTS CHAPTER 5

141

control

8

9

10

11

PC190723

1 h 2 h

Figure 6. Alkyl gallates do not cause cell elongation, but do cause cells to lyse. B. subtilis cells were incubated with alkyl gallates at MIC50 for 60 min. (left panel) or 120 min. (right panel). Brightfield microscopy images are shown. There is evident cell lysis after incubation with compounds 9, 10, an 11. Scale bar (same for all): 5 µm.

Figure 6. Alkyl gallates cause cell elongation and lysis. (A) controls: B. subtilis cells were incubated

with nothing (168) or with PC190723 at 2 μg/mL for 1h, 2h or 3h. (B) B. subtilis cells were incubat-

ed with alkyl gallates at MIC50 for 1h (1st column) or 2h (2nd column), or for 3h at 50% and 10%

of MIC50 (3rd and 4th column, respectively). Brightfield microscopy images are shown. There is

evident cell lysis after incubation at MIC50 with compounds 9, 10, an 11, whereas cell elongation is

seen at concentrations below MIC50. Scale bar (top left, same for all): 5 μm.


142

mentation caused by delayed or defective division (Addinall et al., 1996). Even though the alkyl gallates had clear and immediate effects on FtsZ rings (Fig. 1), elongated cells could only occasionally be observed upon longer incubation at MIC50, whereas cells grown in the presence of the known FtsZ-targeting com-pound PC190723 (Haydon et al., 2008) were clearly filamentous (Fig. 6 A). After 1 hour of incubation with the compounds 9, 10, and 11 at MIC50, we noticed that some cells had already lysed, and lysis was more noticeable when the in-cubation time was extended to 2 hours. At MIC90, compounds 9, 10, and 11 caused cell death at an equivalent or higher proportion (data not shown). In-cubation with compound 8 did not cause noticeable lysis – at least not to the extent as the other compounds - however, the large majority of the non-lysed cells did not appear elongated. These results indicate that the alkyl gallates can cause cell death via another mechanism than FtsZ inhibition. These alternative mechanisms may correspond to the ones reported by the REMA assay for an-tibacterial activity, as elongation still requires metabolic activity. Therefore, we investigated the effects of the alkyl gallates at lower concentrations than the MIC50, determined in the REMA assay. Lowering the concentrations of the com-pounds revealed that compounds 9, 10 and 11 are capable of causing cell elon-gation, as expected for FtsZ inhibitors, whereas compound 8 had no effect on cell length (Fig. 6B). The minimal inhibitory activity (MIC24H) of the compounds was established by a 2-fold dilution series in which prolonged incubation re-vealed that compound 11 can prevent cell growth at a lower concentration than the one determined by the REMA assay (25 mg/mL Table 1). This is closer to the value reported in the literature for this compound (Kubo et al., 2004). Compounds 8 and 9 and 10 show similar inhibition of growth after 24 hours as previously determined by the REMA assay. Taken together these results indi-cate that alkyl gallates with a high affinity for FtsZ in vitro can induce cell death by directly targeting FtsZ. Additionaly it is observed that, at higher concentra-tions of the alkyl gallates an alternative mechanism is responsible for the quick cell death that takes place without cell elongation (Fig. 6B).

As several described FtsZ inhibitors also affect cell membranes (Foss et al., 2013), we decided to study the effect of the alkyl gallates on cell membrane in-tegrity in vivo. The Live/Dead BacLight kit, which combines the green membrane permeable fluorescent DNA dye, SYTO 9, and the red membrane impermeable

RESULTS CHAPTER 5

143

fluorescent DNA dye, propidium iodide, was used to assess membrane integrity. Cells were incubated with both dyes and alkyl gallates. Subsequently, cells were imaged and red or orange cells were classified as cells with affected membrane integrity, whereas green cells were classified as cells with intact membranes. Ni-sin, which is known to make pores in the membrane, and CCCP, which disrupts the membrane potential but does not make pores (Strahl, Hamoen 2010), were used as controls. All alkyl gallates were found to be able to create membrane pores in vivo, to different extents (Fig. 7). Compound 8, at both MIC50 and MIC90

concentrations, permeabilized all cells, which is in line with the observed cell death without elongation – although it has to be noted that cells incubated with compound 8 at MIC50 did not noticeably lyse (Fig. 6). Compounds 9, 10, and 11 showed a concentration-dependent membrane permeabilization – but impor-tantly, incubation at MIC50 concentrations never resulted in more than 50% per-meabilized cells, indicating that the alkyl gallates function by targeting mem-brane integrity, FtsZ function, and possibly other mechanisms.

1,38% 2,43%

87,89%

97,86%

11,02%16,51%

29,77%

99,00%

36,55%

94,60%

65,44%

20%

40%

60%

80%

100%

CCCP NISIN 8 9 10 11

Cells

with

per

mea

biliz

ed m

embr

anes

(%)

Untreated

Figure 7. Membrane integrity is affected in the presence of alkyl gallates. B. subtilis cells were incubated with alkyl gallates, at MIC50 (light grey bars) and MIC90 (dark grey bars), and with CCCP (2 mM) or nisin (2.5 µg/ml) for 15 min., and membrane integrity was monitored using the Live/Dead assay (see text). The experiment was performed twice and more than 250 cells per incubation were counted per experiment. The values were converted to percentages, average percentage and standard deviation are shown.

Figure 7. Membrane integrity is affected in the presence of alkyl gallates. B. subtilis cells were

incubated with alkyl gallates, at MIC50 (light grey bars) and MIC90 (dark grey bars), and with CCCP

(2 mM) or nisin (2.5 μg/mL) for 15 min., and membrane integrity was monitored using the Live/Dead

assay (see text). The experiment was performed twice and more than 250 cells per incubation

were counted per experiment. The values were converted to percentages, average percentage and

standard deviation are shown.


144

DISCUSSION

The use of alkyl gallates as anti-bacterial agents has been proposed in various studies in which the pharmacological activity of these compounds was described (Kubo et al., 2002; Kubo et al., 2004; Kubo et al., 2002; Kubo et al., 2003; Shibata et al., 2005). As semi-synthetic compounds that are derived from gallic acid, a plant metabolite, these compounds could be interesting and environmental-friendly al-ternatives for the control of bacterial infections of agricultural crops. To this end, we previously characterized the activity of alkyl gallates against the important citrus pathogen Xanthomonas citri subsp citri (Xac)(Silva et al., 2013). Our initial studies in-dicated that the alkyl gallates disrupt cell division in Xac, possibly by targeting the FtsZ-ring, which is different from the observed mechanism of membrane binding that is also described for these compounds (Takai et al., 2011). Here, we investigat-ed the mode of action of four alkyl gallates in more detail.

As we neither had access to a Xac strain expressing fluorescent FtsZ, nor were able to purify Xac FtsZ, we turned to B. subtilis, which is also killed by alkyl gallates (Kubo et al., 2004). Here, we provide both in vivo and in vitro evidence that FtsZ is a target for the alkyl gallates. The addition of alkyl gallates to cells expressing a fluorescent variant of FtsZ causes immediate disruption of Z-rings (Fig. 1), and cultures exposed to compounds 9, 10 and 11 display the classical elongation phenotype of cells affected in cell division (Fig. 6). Purified FtsZ is blocked from polymerizing in the presence of alkyl gallates (Fig. 4 and Fig S4), and some of the alkyl gallates bind FtsZ with high affinity, most notable heptyl gallate with an estimated Kd of 80 nM (Fig. 3). It should be noted that the effect of the alkyl gallates on B. subtilis indicated that FtsZ is not the sole target– the classical phenotype of cell elongation by division inhibition was observed for compounds 9, 10 and 11 (Fig. 6), but all compounds also affected membrane integrity as observed with the permeability assay (Fig. 7). Therefore, alkyl gal-lates probably promote cell death by a combination of mechanisms: FtsZ inhi-bition, membrane permeabilization and, possibly, another activity.

Two recent studies pointed out some issues with antibacterials that have been identified as ‘FtsZ-inhibitors’. The first study, by Anderson et al. (Anderson et al., 2012), showed that many compounds identified as FtsZ inhibitors in high-throughput screening assays based on FtsZ-mediated GTP hydrolysis, in


145

fact form small aggregates that block GTP hydrolysis. The addition of Triton X-100 to these assays prevents aggregate formation and reveals normal GTP hy-drolysis levels in the presence of ‘false-positive’ compounds. The effect of alkyl gallates on GTP hydrolysis is the same, irrespective of whether Triton X-100 is present or not, indicating that alkyl-gallates are not false-positive GTP hydrolysis inhibitors (Fig. 2). Also, our other in vitro assays show that – at least some – of the alkyl gallates bind FtsZ with high affinity and that these compounds cluster FtsZ and prevent polymerization (Figs 3, 4). Cluster formation is not caused by aspecific protein aggregation as shown by control experiments with His-EzrAcyt and BSA. Combined, the in vitro work clearly shows that FtsZ is inhibited by the alkyl gallates. The second study, from Foss et al. (Foss et al., 2013) indicated that many compounds, identified as FtsZ inhibitors, target the membrane. Alkyl gallates have already been identified as membrane binding agents (Takai et al., 2011) – therefore we examined the effect of the alkyl gallates on membrane in-tegrity. Intriguingly, compound 8, with the shortest alkyl chain length, had the most disruptive effect on membranes as monitored by the influx of propidium iodide into cells (Fig. 7). The other compounds also affected membrane integrity albeit to different extents. We observed that compounds 9-11 permeabilize less than 50% of the cells at concentrations where 50% of the cells are metabolically inactive (MIC50) as determined by the REMA assay. At MIC50, cell elongation can-not be observed and many cells in the culture lyse. At lower concentrations of the compounds, elongation can clearly be observed, and it is evident, from the dilution series (Table 1), that concentrations that cause elongation are sufficient to inhibit cell growth. We conclude that, especially for compounds 10 and 11, FtsZ inhibition occurs at lower concentrations than MICs determined with the REMA assay. Increasing the concentration to MIC values leads to disruption of membrane integrity – this is particularly obvious for compound 8. The effect of alkyl gallates on membrane integrity is not the cause for FtsZ ring disruption as compounds that disrupt membrane integrity or that dissipate the membrane potential do not affect the Z-ring in vivo, whereas all alkyl gallates, at MIC50 and MIC90, quickly disrupt most of the Z-rings in cells. Combined, our experiments show that alkyl gallates, in addition to targeting membrane integrity (Takai et al., 2011; Haydon et al., 2008), can also directly target FtsZ. We consider it very likely that both membrane disruption and FtsZ inhibition by alkyl gallates apply


146

to many bacteria because: (1) alkyl gallates have a wide spectrum of antibacte-rial activity against both Gram-positive and Gram-negative bacteria (Kubo et al., 2002; Kubo et al., 2002; Kubo et al., 2004; Shibata et al., 2005; Silva et al., 2013); (2) alkyl gallates disrupt bacterial model membranes (Takai et al., 2011) as well as B. subtilis membranes (this work); (3) a cell division phenotype was described both for the Gram-negative X. citri subsp. citri (Silva et al., 2013) as well as for Gram-positive B. subtilis (this work); (4) FtsZ is directly inhibited by alkyl gallates (this work) – FtsZ is highly conserved and the absence of ZapA rings described for X. citri subsp. citri (Silva et al., 2013) strongly suggests that Z-ring formation is blocked in this organism as well. Heptyl-gallate (compound 10) was found to bind to FtsZ with very high affinity, resulting in blocked cell division at low concentrations and FtsZ cluster formation in vitro. The length of the alkyl chains affects both the interaction with the membrane and with FtsZ. As the most promising anti-FtsZ agent, heptyl gallate can be used as a hit for the design of innovative compounds that have enhanced specificity towards FtsZ and less ac-tivity on the membrane. Several modifications on the structure of heptyl gallate are currently being made in our laboratories.

CONFLICT OF INTEREST STATEMENT

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

ACKNOWLEDGEMENTS

We are grateful to Morten Kjos and Ruud Detert Oude Weme (Molecular Genetics, University of Groningen, NL) for plasmids pMK13 and pDOW01.

This work was funded by the bilateral research programme “Biobased Economy” from the Netherlands Organisation for Scientific research (NWO) and the São Paulo Research Foundation (FAPESP 2013/50367-8, Brasil, to DJS and HF), a NWO Vidi grant (to DJS), a doctoral grant (SFRH/BD/78061/2011) from POPH/FSE and FCT (Fundação para a Ciência e Tecnologia) from Portugal (to ASB) and a Science without Borders sandwich grant (CNPq process: 201196/2012-3, to IS).


147

REFERENCES

Abreu, A.C., McBain, A.J., and Simoes, M. (2012). Plants as sources of new antimicrobials and resistance-modifying agents. Nat. Prod. Rep. 29, 1007-1021. doi: 10.1039/C2NP20035J

Kubo, I., Fujita, K., and Nihei, K. (2002). Anti- -Salmonella activity of alkyl gallates. J. Agric. Food Chem. 50, 6692-6696. doi: 10.1021/jf020467o

Kubo, I., Xiao, P., and Fujita, K. (2002). Anti- -MRSA activity of alkyl gallates. Bioorg. Med. Chem. Lett. 12, 113-116

Shibata, H., Kondo, K., Katsuyama, R., Kawazoe, K., Sato, Y., Murakami, K., et al. (2005). Alkyl gallates, intensifiers of beta-lact-am susceptibility in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 49, 549-555. doi: 10.1128/AAC. 49.2.549-555.2005

Kubo, I., Fujita, K., Nihei, K., and Nihei, A. (2004). Antibacterial activity of akyl gallates against Bacillus subtilis. J. Agric. Food Chem. 52, 1072-1076. doi: 10.1021/jf034774l

Silva, I.C., Regasini, L.O., Petronio, M.S., Silva, D.H., Bolzani, V.S., Belasque, J., et al. (2013). Antibacterial activity of alkyl gal-lates against Xanthomonas citri subsp. citri. J. Bacteriol. 195, 85-94. doi: 10.1128/JB.01442-12

Kubo, I., Fujita, K., Nihei, K., and Masuoka, N. (2003). Non-antibiotic antibacterial ac-tivity of dodecyl gallate. Bioorg. Med. Chem. 11, 573-580

Takai, E., Hirano, A., and Shiraki, K. (2011). Ef-fects of alkyl chain length of gallate on self-association and membrane binding. J. Biochem. 150, 165-171. doi: 10.1093/jb/mvr048

Kapoor, S. and Panda, D. (2009). Targeting FtsZ for antibacterial therapy: a promising av-enue. Expert Opin. Ther. Targets 13, 1037-1051. doi: 10.1517/14728220903173257

Lock, R.L. and Harry, E.J. (2008). Cell-division inhibitors: new insights for future anti-biotics. Nat. Rev. Drug Discov. 7, 324-338. doi: 10.1038/nrd2510

Huang, Q., Tonge, P.J., Slayden, R.A., Kirikae, T., and Ojima, I. (2007). FtsZ: a novel target for tuberculosis drug discovery. Curr. Top. Med. Chem. 7, 527-543. doi: http://dx.doi.org/10.2174/156802607780059790

Erickson, H.P., Anderson, D.E., and Osawa, M. (2010). FtsZ in bacterial cytokinesis: cytoskeleton and force generator all in one. Microbiol. Mol. Biol. Rev. 74, 504-528. doi: 10.1128/MMBR.00021-10

Adams, D.W. and Errington, J. (2009). Bacteri-al cell division: assembly, maintenance and disassembly of the Z ring. Nat. Rev. Microbiol. 7, 642-653. doi: 10.1038/nrmi-cro2198

Anderson, D.E., Kim, M.B., Moore, J.T., O’Brien, T.E., Sorto, N.A., Grove, C.I., et al. (2012). Comparison of small molecule inhib-itors of the bacterial cell division pro-tein FtsZ and identification of a reliable cross-species inhibitor. ACS Chem. Biol. 7, 1918-1928. doi: 10.1021/cb300340j

Andreu, J.M., Schaffner-Barbero, C., Huecas, S., Alonso, D., Lopez-Rodriguez, M.L., Ruiz-Avila, L.B., et al. (2010). The antibac-terial cell division inhibitor PC190723 is an FtsZ polymer-stabilizing agent that induces filament assembly and conden-sation. J. Biol. Chem. 285, 14239-14246. doi: 10.1074/jbc.M109.094722

Beuria, T.K., Santra, M.K., and Panda, D. (2005). Sanguinarine blocks cytokinesis in bac-teria by inhibiting FtsZ assembly and bundling. Biochemistry 44, 16584-16593. doi: 10.1021/bi050767

Hemaiswarya, S., Soudaminikkutty, R., Narasumani, M.L., and Doble, M. (2011). Phenylpro-panoids inhibit protofilament formation of Escherichia coli cell division protein


148

FtsZ. J. Med. Microbiol. 60, 1317-1325. doi: 10.1099/jmm.0.030536-0

Rai, D., Singh, J.K., Roy, N., and Panda, D. (2008). Curcumin inhibits FtsZ assembly: an at-tractive mechanism for its antibacterial activity. Biochem. J. 410, 147-155. doi: 10.1042/BJ20070891

Keffer, J.L., Huecas, S., Hammill, J.T., Wipf, P., Andreu, J.M., and Bewley, C.A. (2013). Chryso-phaentins are competitive inhibitors of FtsZ and inhibit Z-ring formation in live bacteria. Bioorg. Med. Chem. 21, 5673-5678

Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular cloning. New York: Cold spring harbor laboratory press.

Sezonov, G., Joseleau-Petit, D., and D’Ari, R. (2007). Escherichia coli physiology in Luria-Bertani broth. J. Bacteriol. 189, 8746-8749. doi: 10.1128/JB.01368-07

Harwood, C.R. and Cutting, S.M. (1990). Molec-ular biological methods for Bacillus. UK, Chichester: John Wiley.

Palomino, J.C., Martin, A., Camacho, M., Guerra, H., Swings, J., and Portaels, F. (2002). Resazurin microtiter assay plate: sim-ple and inexpensive method for de-tection of drug resistance in Mycobac-terium tuberculosis. Antimicrob. Agents Chemother. 46, 2720-2722. doi: 10.1128/AAC.46.8.2720-2722.2002

Sharpe, M.E., Hauser, P.M., Sharpe, R.G., and Errington, J. (1998). Bacillus subtilis cell cycle as studied by fluorescence micros-copy: constancy of cell length at initia-tion of DNA replication and evidence for active nucleoid partitioning. J. Bacteriol. 180, 547-555

Sharpe, M.E., Hauser, P.M., Sharpe, R.G., and Errington, J. (1998). Bacillus subtilis cell cycle as studied by fluorescence micros-copy: constancy of cell length at initia-tion of DNA replication and evidence for active nucleoid partitioning. J. Bacteriol. 180, 547-555

Mukherjee, A. and Lutkenhaus, J. (1998). Dy-namic assembly of FtsZ regulated by

GTP hydrolysis. EMBO J. 17, 462-469. doi: 10.1093/emboj/17.2.462

Krol, E. and Scheffers, D.J. (2013). FtsZ polym-erization assays: simple protocols and considerations. J. Vis. Exp. (81):50844. doi: 10.3791/50844

Krol, E., van Kessel, S.P., van Bezouwen, L.S., Ku-mar, N., Boekema, E.J., and Scheffers, D.J. (2012). Bacillus subtilis SepF binds to the C-terminus of FtsZ. PLoS One 7(8): e43293. doi: 10.1371/journal.pone.0043293

Kelley, C., Zhang, Y., Parhi, A., Kaul, M., Pilch, D.S., and LaVoie, E.J. (2012). 3-Phenyl substituted 6,7-dimethoxyisoquinoline derivatives as FtsZ-targeting antibacte-rial agents. Bioorg. Med. Chem. 20, 7012-7029

Strahl, H. and Hamoen, L.W. (2010). Mem-brane potential is important for bac-terial cell division. Proc. Natl. Acad. Sci. U. S. A. 107, 12281-12286. doi: 10.1073/pnas.1005485107

Addinall, S.G., Bi, E., and Lutkenhaus, J. (1996). FtsZ ring formation in fts mutants. J. Bacteriol. 178, 3877-3884

Haydon, D.J., Stokes, N.R., Ure, R., Galbraith, G., Bennett, J.M., Brown, D.R., et al. (2008). An inhibitor of FtsZ with potent and selective anti-staphylococcal activity. Science 321, 1673-1675. doi: 10.1126/science.1159961

Foss, M.H., Eun, Y.J., Grove, C.I., Pauw, D.A., Sorto, N.A., Rensvold, J.W., et al. (2013). Inhibitors of bacterial tubulin target bacterial membranes in vivo. Medchem-comm 4, 112-119. doi: 10.1039/C2M-D20127E

SUPPLEMENTAL FIGURES CHAPTER 5

149

SUPPLEMENTAL FIGURES

Figure S1. B. subtilis cells expressing eyfp (strain 4055) without treatment or after incubation with

CCCP (2 mM) or nisin (1.5 μg/mL) to disrupt membrane potential or integrity. Brightfield and fluo-

rescence microscopy images are shown. Scale bar (same for all): 5 μm.

2 min

4055

CCCP

Nisin

Figure S1. B. subtilis cells expressing eyfp (strain 4055) without treatment or after incubation with CCCP (2 mM) or nisin (1.5 µg/ml) to disrupt membrane potential or integrity. Brightfield and fluorescence microscopy images are shown. Scale bar (same for all): 5 µm.


150

Wavelength (nm)

[FtsZ] (µM)

320 340 360 380 400 420 440

A

C

320 340 360 380 400 420 440

Wavelength (nm)

B

300000250000200000150000

10000050000

0

350000400000450000500000

2 4 6 8 100 12 14

Fluo

resc

ence

Inte

nsity

(a.u

.)Fl

uore

scen

ce In

tens

ity (a

.u.)

Fluo

resc

ence

Inte

nsity

(a.u

.)

30000

25000

20000

15000

10000

5000

35000

0

50000

40000

35000

20000

15000

10000

0

45000

30000

25000

5000

compound 8 compound 9

Figure S2. Binding of compound 8 and 9 to FtsZ monitored by fluorescence spectroscopy. Fluorescence emission spectra of compound 8 (A) and 9 (B) acquired in the presence of 0, 1.2, 2.4, 3.6, 4.8 µM FtsZ-compound 8 and 0, 2.4, 3.6, 4.8, 7.2 µM FtsZ-compound 9 (from bottom to top). Compound 9 was used at 4x higher concentration than compound 8. The change in fluorescence intensity of compound 8 (C) at 366 nm plotted vs increasing concentrations of FtsZ (from 0 to 14.4 µM). The plot of fluorescence intensity vs FtsZ concentration for compound 9 was impossible to obtain due to high instability of the fluorescence signal.

Figure S2. Binding of compound 8 and 9 to FtsZ monitored by fluorescence spectroscopy. Flu-

orescence emission spectra of compound 8 (A) and 9 (B) acquired in the presence of 0, 1.2, 2.4,

3.6, 4.8 μM FtsZ-compound 8 and 0, 2.4, 3.6, 4.8, 7.2 μM FtsZ-compound 9 (from bottom to top).

Compound 9 was used at 4x higher concentration than compound 8. The change in fluorescence

intensity of compound 8 (C) at 366 nm plotted vs increasing concentrations of FtsZ (from 0 to

14.4 μM). The plot of fluorescence intensity vs FtsZ concentration for compound 9 was impossible

to obtain due to high instability of the fluorescence signal.

Figure S3. Sedimentation of FtsZ in the presence of 100 μM alkyl gallates and at high sedimentation

spin. (A) 10 μM FtsZ was incubated with alkyl gallates (100 μg/mL) before GTP/GDP was added. In

the sample without nucleotide (no nucl), the same volume of polymerization buffer was added

to the sample as for GTP/GDP. (S) indicates supernatant and (P) pellet fractions from the experi-

ment. As a control, 1.5 % DMSO was used. Samples were spun down at 186,000 ×g for 10 min. All

experiments were performed in duplicate. (B) The same experiment as in (A) with alkyl gallates at

50 μg/mL. Samples were spun down at 350 000 ×g for 15 min.

SUPPLEMENTAL FIGURES CHAPTER 5

151


152

8 9

A B C

D10 11

E

DMSO

Figure S4. Structures formed by FtsZ with alkyl gallates at MIC90 concentration. 10 µM FtsZ was incubated with alkyl gallates at MIC90 concentration for 5 minutes. After that, 2 mM GTP was added and samples were incubated for another 20 minutes. As a control, 1.5 % DMSO was used. Scale bar: 100 nm.

Figure S4. Structures formed by FtsZ with alkyl gallates at MIC90 concentration. 10 μM FtsZ was

incubated with alkyl gallates at MIC90 concentration for 5 minutes. After that, 2 mM GTP was

added and samples were incubated for another 20 minutes. As a control, 1.5 % DMSO was used.

Scale bar: 100 nm.

Summary

CHAPTER

THE FTSZ PROTEIN CHAPTER 6

157

The first known step in bacterial cell division is the assembly of the FtsZ protein into a ring-like structure, the so-called Z-ring, in the middle of the cell. Assembly of FtsZ is a complex process that involves the polymerization of FtsZ into long pro-tofilaments, which are tethered to the inner membrane by additional proteins like FtsA, ZipA or some others. Next, single FtsZ filaments are cross-linked by other pro-teins or interact with each other via lateral bonds to form thicker structure. Upon formation of this complete Z-ring structure, other cell division proteins are recruit-ed, which results in a big complex called the divisome. The divisome machinery is responsible for membrane constriction and cell wall ingrowth, which results in the splitting of the original ‘mother’ cell into two equally sized ‘daughter’ cells. FtsZ is a key player in this process. It is a very conserved essential protein, whose assembly and disassembly is highly regulated by other proteins in order to ensure proper cell separation. In the presence of FtsZ inhibitors or when FtsZ is depleted, cells are not able to divide, elongate into long filaments, and eventually die 1.

THE FTSZ PROTEIN

FtsZ consists of 5 domains, all of which are crucial for the proper functioning of FtsZ and the division machinery. FtsZ assembles into long filaments in the pres-ence of GTP, through longitudinal interactions between the globular domains of FtsZ that assemble on top of each other. These interactions activate FtsZ GTP hy-drolysis, which occurs when the T7 loop (placed on the C-terminal globular core) from the top monomer enters the GTP binding site (present in the N-terminal domain) on the bottom monomer 1. Lateral interactions between protofilaments of FtsZ represent another kind of interaction. Here, the extreme C-terminal part of FtsZ is involved, in particular, the last 4-6 residues of the FtsZ protein (CTV) 2. The CTV is preceded by a very short α-helix (CTT), built up of 9 residues which was shown to be involved in many interactions with FtsZ binding partners (see 1 and chapter 3). The CTT and CTV are connected to the globular domain via the intrinsically disordered C-terminal linker which is important for the formation of protofilaments and the Z-ring architecture in vivo 3.

The most characteristic feature of FtsZ is the formation of long filaments, which makes it interesting protein for in vitro studies. Several techniques are


158

commonly used to study assembly and activity of FtsZ. The most common are: sedimentation of polymers, which allows the quantitation of the fraction of polymerized protein; light scattering, which measures assembly and disassem-bly of FtsZ in real time; negative stain electron microscopy (EM), which allows visualization of polymers in the sample; and free phosphate release assays, in which the total phosphate released from GTP hydrolysis can be measured and correlated to GTPase activity of FtsZ. In chapter 2, protocols to study the biochemistry of FtsZ in various buffers using above methods are discussed. The assembly of FtsZ from two model organisms, E. coli (FtsZEc) and B. subtilis (FtsZBs), at various pH (6.5, 6.8 and 7.5) and various KCl concentrations (50 mM, 300 mM) is compared. Even if FtsZEc forms longer polymers than FtsZBs, as vi-sualized using EM, the light scattering signal of FtsZEc is around 40-fold lower than that of FtsZBs at 50 mM KCl and pH=6.5. This is an effect of the bundling of FtsZ protofilaments into larger structures driven by electrostatic forces. The CTV of FtsZBs is positively charged at pH=6.5 while the charge of the CTV from FtsZEc is neutral. In the model proposed by Buske and Levin, the polymer core formed by the assembled globular domains of FtsZ carries a net negative charge at physiological pH 2 . A positively charged CTV interacts with the core of the adjacent polymer, causing bundling 2. Bundling of FtsZBs polymers re-duces the GTPase activity compared to non-bundled polymers of FtsZEc. At an increased salt concentration (300 mM), FtsZBs displays reduced polymer bun-dling and the light scattering signal of both proteins was comparable. EM anal-ysis showed that protofilaments of FtsZBs are shorter than those of FtsZEc. Short-er filaments of FtsZBs gave slightly a lower light scattering signal at 300 mM KCl and the polymers disassembled more quickly comparing to the ones formed by FtsZEc. The GTPase activity of both proteins was similar at high KCl concen-trations, indicating that bundling of filaments and the concomitant decrease in subunit turnover of FtsZ monomers caused the decrease in GTPase activity of FtsZBs. All methods described in chapter 2, except the sedimentation assay, give clear results at all chosen conditions. In the sedimentation assay, preas-sembled FtsZ polymers are spun down and separated from the monomers and shorter polymers that stay in solution. Polymers from both FtsZBs and FtsZEc could not be recovered in this assay at high KCl concentrations. This may be because of the reduction in the formation of lateral bonds between polymers

FTSZ INTERACTING PROTEINS CHAPTER 6

159

and the formation of shorter polymers or a higher turnover of subunits. Some of the methods described in chapter 2 were used in the following three chap-ters. Knowledge about the behaviour of FtsZ under various conditions allowed a quick selection of the best conditions for the interaction studies with other proteins (chapter 3 and chapter 4) and small molecule inhibitors (chapter 5). These protocols were also successfully adapted in our laboratory to study FtsZ from some other organisms (not shown).

FTSZ INTERACTING PROTEINS

At least nine proteins directly regulate assembly of FtsZ in B. subtilis 1. FtsA, EzrA and SepF are membrane anchors for FtsZ in cell during vegetative growth 1. SpoIIE is an additional membrane anchor for FtsZ that functions in the reloca-tion of the Z-ring from the mid-cell to the cell poles during sporulation 4,5. Most of these proteins (excluding EzrA) “positively” regulate assembly of FtsZ, which means that they support bundling or polymerization of FtsZ. Another posi-tive regulator is the cytoplasmic protein ZapA, that stabilizes FtsZ structures by cross-linking of protofilaments 1. The role of the negative regulators is to block FtsZ polymerization in places where, or at moments when FtsZ should not as-semble. MinC, for example, blocks FtsZ polymerization at cell poles ensuring proper placement of the Z-ring at mid-cell during vegetative growth 1. Another example is UgtP, which blocks assembly of the Z-ring in fast-growing cells “to give time” to the cell to reach its appropriate size before cell division occurs 6. Many FtsZ interacting proteins bind to the the C-terminus (CTT +/- CTV) of FtsZ. This short peptide is considered to be a “landing pad” for the interact-ing partners of FtsZ 1. It is impossible that all these proteins simultaneously bind to such a small region. Thus, it is likely that FtsZ is regulated in time by competition of the interacting proteins for the C-terminus. In fact, FtsZ cannot function in cells without its extreme C-terminus, because proteins like ZipA or FtsA that serve as membrane anchors also bind to this region 7. In chapter 3, a pull-down strategy to find binding partners for the extreme C-terminus of FtsZ is described. The last 69 amino acids of FtsZ, comprising the C-terminal non-conserved linker and the CTT+CTV (16 aa) sequences were fused to a Halo-


160

-tag. The HaloTag consists of a modified dehalogenase that covalently binds to sepharose beads via a chloroalkane ligand 8. A control fusion protein contained only the linker sequence without the final 16 C-terminal amino acids. Lysates of E. coli cultures in which the FtsZ interacting partners EzrA, MinC and SepF were overexpressed were passed over sepharose beads with the bound bait proteins. Although both EzrA and MinC were previously shown to interact with the C-terminus of FtsZ 9,10 , in the pull-down assay only SepF was recovered. It is likely that MinC and EzrA do not strongly bind the C-terminus and thus were lost during washing steps. The recovery of SepF was surprising, especially be-cause a previous study by Singh et al. indicated that SepF protein binds to the globular domain of FtsZ 10. To evaluate which amino acids in the C-terminus are involved in the interaction with SepF, every single amino acid in the C-ter-minus was mutated to Alanine and the same pull-down assay was performed using this single mutant collection. Mutation of the two highly conserved res-idues P372 and F374 fully abolished binding of SepF to the C-terminus. Sur-prisingly, mutation of most of the other residues (11 out of 16) also affected the binding to SepF to some extent. This suggests that the secondary and ter-tiary structure of the C-terminus are important for the interactions with SepF. Recently, Duman et al. screened FtsZ binding sites on SepF using a yeast two hybrid screen 11. Many residues on SepF were found to be important for FtsZ binding, suggesting a similar requirement for folding of the SepF domains in the interactions with FtsZ 11.

A reverse pull-down assay was used to prove that the interaction between SepF and the C-terminus of FtsZ is specific and this is the only place where SepF is able to bind FtsZ. MBP-SepF immobilized on amylose resin was incu-bated with FtsZ, FtsZP372A (full-length FtsZ with mutated conserved Pro372 into Ala) and FtsZ∆16 (FtsZ without the CTT and CTV). After several washing steps, full-length FtsZ remained bound to the resin while the other two mu-tants were not retained by MBP-SepF. This proves that the globular domain of FtsZ is not sufficient to (strongly) bind SepF.

Another binding partner of FtsZ that has not been studied in much detail is the sporulation protein SpoIIE. SpoIIE is a membrane protein and is thought to consist of three domains: the membrane domain I, the FtsZ-interacting and oligomerization domain II and the phosphatase domain III. It has been known

FTSZ INTERACTING PROTEINS CHAPTER 6

161

for a long time that FtsZ and SpoIIE co-localize in vivo and that SpoIIE is in-volved in relocation of FtsZ from mid-cell to the cell poles during sporulation in B. subtilis 12. After formation of asymmetric septum, SpoIIE performs its second function. It activates σF by dephosphorylation of the anti-sigma factor antag-onist SpoIIAA. Direct interaction between SpoIIE and FtsZ in vitro was shown by Lucet et al. using a pull-down strategy and gel filtration analysis 13. Howev-er, another group which studied the cytoplasmic domain of SpoIIE (SpoIIEcyt) in vitro could not reproduce these findings 14. Therefore, whether or not, and how, these two proteins interact is still not fully clear. One reason for this may be that SpoIIEcyt is highly unstable and forms inclusion bodies when overexpressed in E.coli 13,14. Levdikov et al. showed that domain III on its own can function as phos-phatase and that this function depends on its cofactor Mn2+ 15. In chapter 4 the interactions between FtsZ and SpoIIE are studied. The cytoplasmic part of SpoIIE (domains II and III) was purified using a fusion to MBP to improve solu-bility and a strep-tag for purification (Ms-SpoIIEcyt). The purified protein strong-ly binds metal ions, which were incorporated into the protein during folding in E. coli, indicating that the SpoIIEcyt domain is properly folded and possibly active. Metal binding enhances oligomerization of strep-SpoIIEcyt, which is an indication that domain II, involved in oligomerization of SpoIIE, and the manga-nese-dependent phosphatase domain III are not completely independent but might influence each other. The phosphatase domain of SpoIIE is involved in the activation of σF, which activates most of the sporulation specific-genes, that drive the sporulation process in the prespore. Activation of σF is achieved by dephosphorylation of the SpoIIAA protein, a process dependent on Mn2+ 16. In the absence of Mn2+ sporulation is blocked and can be supported only by high concentrations of iron in the sporulation medium 17. However, at which stage of sporulation is the process blocked is still not fully understood. Here, it is shown that in the absence of Mn2+ formation of the asymmetric septum is delayed and that less asymmetric Z-rings are formed compared to when Mn2+ is included in the medium. This phenotype resembles the SpoIIE knock-out phenotype. The result suggests that metal binding by SpoIIE is not just required for its phospha-tase activity, but also critical for SpoIIE functioning in asymmetric division, ei-ther by influencing SpoIIE folding, oligomerization and/or interaction with FtsZ. However, which of these processes is affected is not clear.


162

The interactions between FtsZ and strep-SpoIIEcyt were analysed in more detail in vitro. Preliminary data on strep-SpoIIEcyt strongly suggested a direct interaction between FtsZ and strep-SpoIIEcyt. However, strep-SpoIIEcyt alone, af-ter cleavage of the MBP tag was unstable and required high concentrations of KCl and the presence of Triton X-100 in order to remain soluble, components that influence FtsZ activity. High concentrations of the wild type protein could not be obtained. As Ms-SpoIIEcyt could be purified in large amounts and was stable in variety of buffers, this protein was used to study the interactions with FtsZ using assays described in chapter 2. FtsZ interacted with Ms-SpoIIEcyt only in the presence of GDP, as interactions between protofilaments of FtsZ and Ms-SpoIIEcyt in the presence of GTP could not be detected in any of the stan-dard assays. The large MBP tag, which is ~40 kDa, possibly (partially) covered the domain II of SpoIIE, which contains the FtsZ binding site. This shielding might make the binding site not accessible for FtsZ polymers but only for FtsZ monomers and small oligomers. However, these findings are preliminary and should be further analysed. Most of the standard assays and conditions used in chapter 2 could not be used to study interactions with SpoIIE or should be adapted for these studies. All assays contain Mg2+ because GTP hydrolysis of FtsZ is dependent on the presence of this metal. This complicates sedimenta-tion assays as Ms-SpoIIEcyt is always oligomeric in the presence of metal and is thus recovered in the pellet fraction in all sedimentation assays. Second, be-cause both proteins oligomerize and scatter light, the analysis of the results of light scattering experiments is difficult as the measured light scattering signal might be either a sum of the individual signals or an effect of the interactions between these two proteins. Finally, in the presence of Ms-SpoIIEcyt, the free phosphate level in GTPase assay was always higher than the background sig-nal. It is unlikely that SpoIIE is GTPase. The free phosphate levels may be higher because Ms-SpoIIEcyt, as phosphatase, may be co-purified with free phosphate bound to the protein. The free phosphate can be released from the protein in the assay, although there is no phosphate released from Ms-SpoIIEcyt in the presence of EDTA. However, a direct interaction between SpoIIEcyt and FtsZ is evident and the findings described in chapter 4 may inform other studies on FtsZ and SpoIIE interaction.

SMALL MOLECULE INHIBITORS OF FTSZ CHAPTER 6

163

SMALL MOLECULE INHIBITORS OF FTSZ

FtsZ has been studied as an attractive target to develop novel antibiotics. Most attention is focused on human pathogens, especially from the group of mul-tidrug resistance pathogens like S. aureus or M. tuberculosis. In chapter 5 the possible activity of semi-synthetic compounds, alkyl gallates, against FtsZ is studied. Alkyl gallates are derivatives of gallic acid, which is an intermediate of the tannin biosynthesis pathway in plants. Alkyl gallates are easily hydro-lysed to gallic acid and corresponding alcohols and, though of limited use in medicine, could provide good environment-friendly alternatives for pesticides used in agriculture. Alkyl gallates have been shown in various studies to have a broad spectrum activity against Gram-positive and Gram-negative patho-gens 18,19. Recently, Silva et al. noted that alkyl gallates with a side chain length between 5 and 8 carbons, may target cell division proteins in the Gram-neg-ative plant pathogen Xanthomonas citri 20. In chapter 5 the activity of pentyl, hexyl, heptyl and octyl gallates (called drug 8, 9, 10 and 11 to keep in line with the previous report from Silva et al.) is further investigated using the model organism B. subtilis. To visualize cell division in live cells, a FtsZ-eYFP fusion was used. In B. subtilis, cell division is affected when bacteria are treated with al-kyl gallates, with treated cells not only significantly longer than the cells treat-ed with the control (1% DMSO), and FtsZ-eYFP localization to the Z-ring was disrupted with FtsZ-eYFP present in the cytoplasm in randomly distributed patches. The standard assays described in chapter 2 were used to study the effect of alkyl gallates on FtsZ in vitro. EM analysis showed that alkyl gallates cluster FtsZ monomers and bundle preformed FtsZ polymers. Additionally, the compounds effectively inhibit GTPase activity of FtsZ. There was a correlation between the activity of alkyl gallates in vivo and in vitro as heptyl gallate had the lowest MIC, the lowest binding constant and the strongest activity against FtsZ in vitro. Combined, these findings indicate that FtsZ is a direct target for alkyl gallates. In a previous study by Takai et al. it was shown that alkyl gallates may target cell membranes 21, which was confirmed here using a membrane permeability assay. However, as the delocalization of FtsZ-eYFP from mid-cell by alkyl gallates is not a secondary effect of disruption of the cell membrane integrity, it can be concluded that alkyl gallates kill bacteria through a combi-


164

nation of mechanisms: disruption of the cell membrane integrity and a block in cell division. The length of the side chain of alkyl gallates is important for their activity both against FtsZ and cell membrane. Pentyl gallate, the drug with the shortest side chain had the highest activity against the membrane, whereas heptyl gallate was most effective against FtsZ. Drugs with shorter and longer side chains had lower activity against FtsZ. Therefore, heptyl gallate is an at-tractive compound for further development and studies against cell division in bacterial pathogens. Heptyl gallate and other gallates used in chapter 5 are attractive compounds for the use in agriculture against plant pathogens. First, some time after introduction into the environment they will be hydrolysed to gallic acid and corresponding alkohols which makes them safe for use. Second, their multi-target mechanism makes it difficult for bacteria to develop resis-tance. And finally, they are effective against broad spectrum of bacteria and fungi, which makes them good compounds to be used against a broad range of plant pathogens.

CONCLUSION AND FUTURE PERSPECTIVES

The results described in this thesis bring new information into the FtsZ research field in three different FtsZ areas: the interactions between FtsZ and partner proteins, the use of FtsZ as antibacterial target, and the standardization of FtsZ assay conditions.

Based on the results described in this thesis and on the literature it can be concluded that FtsZ is regulated by most of the interacting partners via the small region placed at its C-terminus (see Table 1 and fig. 2 from Introduction). Most of the regulation via this region occurs through membrane-associated proteins (eg. FtsA, ZipA, SepF, EzrA, MinC, etc. (see Table 1 and fig. 2 from In-troduction). However, some cytoplasmic proteins, like ClpX or SlmA, also bind to the C-terminus of FtsZ, whereas other proteins bind to the globular domain of FtsZ (ZapA and MciZ). An interesting example is MinC, which is a cytoplas-mic protein but that regulates FtsZ from the membrane through formation of a complex with MinD. MinC binds to two regions on FtsZ, placed on the glob-ular domain and the C-terminus (see Table 1 and fig. 2 from Introduction). The

CONCLUSION AND FUTURE PERSPECTIVES CHAPTER 6

165

extreme C-terminus of FtsZ is highly conserved. In chapter 3 we showed that SepF, which is present only in Gram-positive organisms can still bind to the C-terminus of E. coli FtsZ. Whether this cross-interactivity is also a feature of the other C-terminus binders is not known. Most of the binding sites for FtsZ part-ners were discovered. The proteins of which the binding sites on FtsZ remain unknown are UgtP, SpoIIE and ZapA.

Most of the FtsZ interacting partners have an additional function apart from the regulation of FtsZ. It is possible that FtsZ also regulates these pro-teins, or at least that interaction with FtsZ is necessary for the localization of the specific protein to the division septum in order to perform its function. One of such an example is SepF which is responsible for proper septum formation in B. subtilis 22. SepF is not absolutely required for FtsZ function as cell division occurs also without this protein. The exact function of the FtsZ-SepF interac-tion is still to be discovered. The current focus in the field is on effects of FtsZ regulators on FtsZ because assays to detect FtsZ polymerization, bundling and GTPase activity are well established. Thu, it is easy to study changes in these features caused by other factors. The features of the FtsZ binders are not so well defined and standard assays are lacking, which makes it difficult to study of the changes in their behaviour caused by FtsZ. Therefore, this area remains open for future work.

The study of the interactions of FtsZ with regulatory proteins is interest-ing from a mechanistic point of view. It brings additional information to un-derstanding of the cell division process. Protein-protein interactions were also studied in order to find new inhibitors that could specifically target FtsZ-essen-tial protein interactions. So far, only FtsZ-ZipA interaction sites were targeted with specific drugs 23. The reason for this is that this approach requires more detailed analysis of the target, good knowledge of the interaction and ratio-nal drug design process. Nowadays, more focus is put on the screening of the drugs that target FtsZ alone. In chapter 5, a series of in vivo and in vitro assays was performed to find the target for alkyl gallates, semi-natural compounds that were identified by screening of a library of drugs against the plant patho-gen X. citri. FtsZ was identified as a direct target for alkyl gallates but these compounds also targeted bacterial cell membrane. A series of controls was performed to prove that targeting FtsZ by alkyl gallates is not a false positive


166

event. It is important to study FtsZ inhibitors using both in vitro and in vivo methods, as some of the drugs might give false positive results. For example, some screens for FtsZ inhibitors are based on cell elongation or FtsZ delocal-ization, which might be a secondary effect of pore formation in the membrane or DNA damage. On the other hand, studying FtsZ inhibitors only in vitro, using purified FtsZ does not give significant and sufficient information as well.

There is still a long way to go for FtsZ researchers before FtsZ will get enough attention from pharmaceutical companies. In a recent review by Walsh and Wencewicz, FtsZ was mentioned as a target which did not give promising results 24. Therefore, the area needs some improvements in order to bring FtsZ into a level of significance in antibacterial area. First, the amount of false pos-itive FtsZ targeting drugs must decrease. Anderson et al. proposed a simple improvement in the in vitro assays by introducing Triton X-100 into the buf-fers 25. Also, both in vitro and in vivo controls should be performed, and a more systematic approach is required. One solution is unification of the conditions in the FtsZ assays in order to simplify comparison of the effectiveness of drugs between laboratories. A good start for that are method papers and protocols which could be used by various laboratories. One such an example is present-ed in chapter 2. This protocol was used in chapter 5 in combination with in vivo studies to find new FtsZ inhibitors.

REFERENCES

1. Adams DW, Errington J. Bacterial cell di-vision: Assembly, maintenance and dis-assembly of the Z ring. Nat Rev Microbiol. 2009;7(9):642-653.

2. Buske PJ, Levin PA. Extreme C terminus of bacterial cytoskeletal protein FtsZ plays fundamental role in assembly independent of modulatory proteins. J Biol Chem. 2012;287(14):10945-10957.

3. Buske PJ, Levin PA. A flexible C-termi-nal linker is required for proper FtsZ assembly in vitro and cytokinetic ring formation in vivo. Mol Microbiol. 2013; 89(2):249-263.

4. Feucht A, Magnin T, Yudkin MD, Errington J. Bifunctional protein required for asym-metric cell division and cell-specific tran-scription in Bacilllus subtilis. Genes Dev. 1996;10(7):794-803.

5. Carniol K, Ben-Yehuda S, King N, Losick R. Genetic dissection of the sporulation protein SpoIIE and its role in asymmet-ric division in Bacilllus subtilis. J Bacteriol. 2005;187(10):3511-3520.

6. Weart RB, Lee AH, Chien AC, Haeusser DP, Hill NS, Levin PA. A metabolic sen-sor governing cell size in bacteria. Cell. 2007;130(2):335-347.


167

7. Pichoff S, Lutkenhaus J. Tethering the Z ring to the membrane through a con-served membrane targeting sequence in FtsA. Mol Microbiol. 2005;55(6):1722-1734.

8. Los GV, Encell LP, McDougall MG, et al. HaloTag: A novel protein labeling tech-nology for cell imaging and protein analysis. ACS Chem Biol. 2008;3(6):373-382.

9. Shen B, Lutkenhaus J. The conserved C-terminal tail of FtsZ is required for the septal localization and division inhibito-ry activity of MinC(C)/MinD. Mol Microbi-ol. 2009;72(2):410-424.

10. Singh JK, Makde RD, Kumar V, Panda D. A membrane protein, EzrA, regulates as-sembly dynamics of FtsZ by interacting with the C-terminal tail of FtsZ. Biochem-istry. 2007;46(38):11013-11022.

11. Duman R, Ishikawa S, Celik I, et al. Struc-tural and genetic analyses reveal the protein SepF as a new membrane an-chor for the Z ring. Proc Natl Acad Sci USA. 2013;110(48):E4601-10.

12. Khvorova A, Zhang L, Higgins ML, Piggot PJ. The spoIIE locus is involved in the Spo0A-dependent switch in the location of FtsZ rings in Bacilllus subtilis. J Bacteriol. 1998;180(5):1256-1260.

13. Lucet I, Feucht A, Yudkin MD, Errington J. Direct interaction between the cell division protein FtsZ and the cell differ-entiation protein SpoIIE. EMBO J. 2000; 19(7):1467-1475.

14. Rawlings AE, Levdikov VM, Blagova E, et al. Expression of soluble, active fragments of the morphogenetic pro-tein SpoIIE from Bacilllus subtilis using a library-based construct screen. Protein Eng Des Sel. 2010;23(11):817-825.

15. Levdikov VM, Blagova EV, Rawlings AE, et al. Structure of the phosphatase domain of the cell fate determinant SpoIIE from Bacilllus subtilis. J Mol Biol. 2012;415(2):343-358.

16. Arigoni F, Guerout-Fleury AM, Barak I, Stragier P. The SpoIIE phosphatase, the sporulation septum and the establish-ment of forespore-specific transcription in Bacilllus subtilis: A reassessment. Mol Microbiol. 1999;31(5):1407-1415.

17. Vasantha N, Freese E. The role of man-ganese in growth and sporulation of Bacilllus subtilis. J Gen Microbiol. 1979; 112(2):329-336.

18. Kubo I, Fujita K, Nihei K, Nihei A. Anti-bacterial activity of akyl gallates against bacillus subtilis. J Agric Food Chem. 2004;52(5):1072-1076.

19. Kubo I, Fujita K, Nihei K. Anti-Salmonel-la activity of alkyl gallates. J Agric Food Chem. 2002;50(23):6692-6696.

20. Silva IC, Regasini LO, Petronio MS, et al. Antibacterial activity of alkyl gallates against Xanthomonas citri subsp. citri. J Bacteriol. 2013;195(1):85-94.

21. Takai E, Hirano A, Shiraki K. Effects of al-kyl chain length of gallate on self-associ-ation and membrane binding. J Biochem. 2011;150(2):165-171.

22. Gundogdu ME, Kawai Y, Pavlendova N, et al. Large ring polymers align FtsZ polymers for normal septum formation. EMBO J. 2011;30(3):617-626.

23. Tsao DH, Sutherland AG, Jennings LD, et al. Discovery of novel inhibitors of the ZipA/FtsZ complex by NMR fragment screening coupled with structure-based design. Bioorg Med Chem. 2006;14(23): 7953-7961.

24. Walsh CT, Wencewicz TA. Prospects for new antibiotics: A molecule-centered perspective. J Antibiot (Tokyo). 2014; 67(1):7-22.

25. Anderson DE, Kim MB, Moore JT, et al. Comparison of small molecule inhib-itors of the bacterial cell division pro-tein FtsZ and identification of a reliable cross-species inhibitor. ACS Chem Biol. 2012;7(11):1918-1928.

Appendices

SAMENVATTING APPENDICES

171

SAMENVATTING

De eerste bekende fase in de bacteriële celdeling is de vorming van de zoge-naamde Z-ring in het midden van de cel. Deze Z-ring is samengesteld uit FtsZ eiwitten en de vorming ervan is een zeer complex proces, wat onder meer de polymerisatie van FtsZ eiwitten tot lange protofilamenten inhoudt. Deze pro-tofilamenten worden gekoppeld aan de binnenmembraan door middel van andere eiwitten, bijvoorbeeld FtsA en ZipA. Daaropvolgend worden de enkel-voudige FtsZ filamenten gecrosslinked met behulp van andere eiwitten of ze interageren met elkaar via laterale verbindingen. Op die manier wordt er een dikkere structuur gevormd.

Tijdens de vorming van deze complete Z-ring structuur worden andere celdelingseiwitten gerekruteerd wat resulteert in het ontstaan van het divi-soom, een groot eiwitcomplex. Het divisoom is verantwoordelijk voor de ver-nauwing van het membraan en het ingroeien van de celwand, met de splitsing van de oorspronkelijke moedercel in twee precies even grote dochtercellen als gevolg.

FtsZ is de belangrijkste pion in dit gehele proces. Het is een zeer gecon-serveerd essentieel eiwit, en de opbouw en de afbraak van de Z-ring wordt zeer sterk gereguleerd door andere eiwitten met een correct scheiding van de cellen als resultaat. In aanwezigheid van FtsZ remmers of wanneer de hoev-eelheid FtsZ eiwitten te laag wordt, zijn cellen niet meer in staat om te delen, waardoor ze verder groeien tot lange filamenten en uiteindelijk doodgaan.

HET FTSZ EIWIT

FtsZ bestaat uit 5 domeinen die allemaal cruciaal zijn voor de correcte werk-ing van het eiwit zelf maar ook van het celdelingsapparaat. In aanwezigheid van GTP kan FtsZ lange protofilamenten vormen door longitudinale interac-ties tussen de globulaire domeinen van de FtsZ eiwitten, die zodanig op elkaar worden gezet. Deze interacties activeren GTP hydrolyse. Laterale interacties tussen protofilamenten van FtsZ vertegenwoordigen een ander soort inter-acties. Het C-terminale uiteinde van FtsZ is hierbij betrokken, meer in het bij-zonder de laatste 4 tot 6 aminozuren van het eiwit (CTV). Van het CTT domein,


172

een korte α-helix van 9 aminozuren gelegen vóór het CTV, is aangetoond dat deze betrokken is bij verschillende interacties met bindingspartners van FtsZ. Het CTT en CTV domein zijn verbonden met het globulaire domein via een van nature ongestructureerde C-terminale linker die belangrijk is voor de vorming van protofilamenten en de architectuur van de Z-ring in vivo.

In hoofdstuk 2 worden protocollen beschreven om op biochemische wijze FtsZ te bestuderen in verschillende buffers. De polymerisatie van FtsZ afkom-stig uit twee modelorganismen, E. coli (FtsZEc) en B. subtilis (FtsZBs) bij verschil-lende pH (6.5, 6.8 en 7.5) en bij verschillende KCl concentraties (50 en 300 mM) wordt met elkaar vergeleken. Sommige methoden beschreven in hoofdstuk 2 worden in de volgende drie hoofdstukken toegepast. Door ervaring op te doen over het gedrag van FtsZ in verschillende condities kon men snel tot een selectie komen van de beste condities voor interactiestudies met andere ei-witten (hoofdstuk 3 en 4) en kleine molecuulremmers (hoofdstuk 5). Deze protocollen zijn ook met succes in ons laboratorium aangepast om FtsZ uit andere organismen te bestuderen.

FTSZ INTERAGERENDE EIWITTEN

Zeker 9 eiwitten reguleren rechtstreeks de opbouw van FtsZ tot een Z-ring in B. subtilis. FtsA, EzrA en SepF vormen een membraananker in de cel voor FtsZ tijdens vegetatieve groei. SpoIIE vormt ook een membraananker voor FtsZ, nodig voor de migratie van de Z-ring van het midden van de cel naar de celpo-len tijdens sporulatie. De meeste van deze eiwitten reguleren de opbouw van FtsZ op positieve wijze, bijgevolg stimuleren ze de polymerisatie of de bun-deling van FtsZ. Nog zo een positieve FtsZ regulator is het cytoplasmatische eiwit ZapA, dat FtsZ structuren stabiliseert door het crosslinken van de pro-tofilamenten. Negatieve regulatoren daarentegen verhinderen FtsZ polymeri-satie op plaatsen waar of op momenten wanneer de Z-ring niet gevormd zou mogen worden. Bijvoorbeeld, MinC voorkomt dat FtsZ polymeriseert aan de celpolen waardoor correcte lokalisatie van de Z-ring in het midden van de cel wordt verzekerd tijdens vegetatieve groei.

Veel eiwitten die interactie aangaan met FtsZ binden aan het C-terminale uiteinde van FtsZ (CTT+/-CTV). In hoofdstuk 3 wordt een pull-down methode


173

beschreven die het mogelijk maakt om zulke eiwitten te vinden. De laatste 69 aminozuren van FtsZ, die de niet-geconserveerde linker en de CTT+CTV se-quenties bevatten, werden gefuseerd met een HaloTag. De HaloTag bestaat uit een gemodificeerde dehalogenase dat een covalente binding kan aangaan met sepharosekorrels via een chloroalkaanligand. Gelyseerde E. coli culturen waarin EzrA, MinC en SepF tot overexpressie gebracht zijn, werden vervolgens samen met sepharosekorrels, die het C-terminale uiteinde van FtsZ gebond-en hadden, geïncubeerd. Na afloop van dit pull-down experiment werd enkel SepF teruggevonden, alhoewel van zowel EzrA en MinC voorheen werd aange-toond dat ze ook interactie kunnen aangaan met het C-terminale uiteinde van FtsZ. Om na te gaan welke aminozuren van de C-terminus betrokken zijn bij de interactie met SepF, werd elk aminozuur in de C-terminus gemuteerd tot een alanine en hetzelfde pull-down experiment werd met deze collectie enkel-voudige mutanten uitgevoerd. Mutatie van twee zeer geconserveerde residu-en P372 en P374 zorgde voor een volledige remming van de binding van SepF met de C-terminus. Zeer verrassend was dat de mutatie van de meeste andere aminozuren (11 tot 16) ook een zekere invloed had op deze binding. Hieruit kan men concluderen dat waarschijnlijk de secundaire en tertiaire structuur van de C-terminus belangrijk is voor de interactie met SepF.

Een bindingspartner van FtsZ die nog niet in detail bestudeerd werd, is het sporulatie-gerelateerde eiwit SpoIIE. SpoIIE is een membraaneiwit en zou uit drie domeinen bestaan: het membraandomein I, het oligomerisatiedomein II, dat interactie aangaat met FtsZ en het fosfatasedomein III. Sinds lange tijd is bekend dat FtsZ en SpoIIE in vivo co-lokaliseren en dat SpoIIE betrokken is bij de relokalisatie van FtsZ van het midden van de cel naar de celpolen tijdens sporulatie in B. subtilis. In hoofdstuk 4 worden de interacties tussen FtsZ en SpoIIE bestudeerd. Het cytoplasmatische domein van SpoIIE (domein II en III) werd gezuiverd met behulp van een MBP fusie om de oplosbaarheid te ver-beteren en een Strep-tag voor de daadwerkelijke zuivering (Ms-SpoIIEcyt). Het gezuiverde eiwit kon goed aan metaalionen binden die gedurende de vou-wing in het eiwit werden ingebouwd in E. coli. Hieruit werd afgeleid dat het SpoIIEcyt domein correct gevouwen en waarschijnlijk actief is. Binding aan metalen verbetert de oligomerisatie van strep-SpoIIEcyt. Dit zou kunnen bete-kenen dat domein II, wat betrokken is bij de oligomerisatie van SpoIIE en het


174

mangaan-afhankelijk fosfatasedomein III niet volledig onafhankelijk van elkaar zijn maar elkaar kunnen beïnvloeden. In hoofdstuk 4 wordt aangetoond dat de vorming van het asymmetrische septum vertraagd is in afwezigheid van Mn2+ en dat minder asymmetrische Z-ringen worden gevormd in vergelijking met wanneer Mn2+ in het medium aanwezig is. Dit fenotype lijkt erg op het fenotype van de SpoIIE knock-out. Dit resultaat doet vermoeden dat metaalbinding door SpoIIE niet alleen vereist is voor zijn fosfatase activiteit maar ook zeer belangrijk is voor de rol van SpoIIE in asymmetrische celdeling, en dit door invloed te heb-ben op ofwel de vouwing van SpoIIE, de oligomerisatie en/of de interactie met FtsZ. Welke van deze processen wordt beïnvloed, is echter nog niet duidelijk. Omdat Ms-SpoIIEcyt in aanzienlijke hoeveelheden gezuiverd kon worden en sta-biel was in een grote verscheidenheid van buffers, werd dit eiwit gebruikt om de interacties met FtsZ te bestuderen met behulp van de analyses beschreven in hoofdstuk 2. Interacties tussen FtsZ en Ms-SpoIIEcyt werden enkel geob-serveerd in aanwezigheid van GDP, terwijl interacties tussen FtsZ protofilament-en en Ms-SpoIIEcyt in aanwezigheid van GTP in geen enkele standaard analyse gedetecteerd kon worden. Nochtans is een directe interactie tussen SpoIIE en FtsZ duidelijk en de conclusies beschreven in hoofdstuk 4 kunnen van belang zijn bij andere studies over de interactie tussen FtsZ en SpoIIE.

KLEINE MOLECUULREMMERS VAN FTSZ

Veel onderzoek wordt verricht naar FtsZ als een aantrekkelijk doelwit voor de ontwikkeling van nieuwe antibiotica. De meeste aandacht gaat naar hu-mane pathogenen, meer in het bijzonder de groep van multidrug-resistente pathogenen zoals S. aureus of M. tuberculosis. In hoofdstuk 5 wordt de mo-gelijk werking van semi-synthetische verbindingen, alkylgallaten, tegen FtsZ onderzocht. Alkylgallaten zijn derivaten van galluszuur, wat een intermediair is van de tannine biosynthetische route in planten. Ze worden gemakkelijk gehydrolyseerd tot galluszuur en overeenkomstige alcoholen. Alhoewel deze alkylgallaten beperkt toegepast worden in de geneeskunde, kunnen ze als een milieuvriendelijk alternatief voor pesticiden in de landbouw gebruikt worden.

In hoofdstuk 5 wordt de rol van pentyl-, hexyl-, heptyl- en octylgallaten verder onderzocht in het modelorganisme B. subtilis. Om celdeling te visual-


175

iseren in levende cellen werd een FtsZ-eYFP fusie gemaakt. Na behandeling met alkylgallaten was de celdeling in B. subtilis aangetast. De cellen waren significant langer vergeleken met de controlecellen en FtsZ-eYFP lokaliseerde niet meer op de plaats van de Z-ring maar op random verspreide plekken in het cytoplasma.

EM analyse toonde aan dat de alkylgallaten verantwoordelijk zijn voor de clustering van FtsZ en de bundeling van voorgevormde FtsZ polymeren. Bov-endien kunnen deze verbindingen zeer effectief de GTPase activiteit van FtsZ remmen. Er is een verband aangetoond tussen de activiteit van alkylgallaten in vivo en in vitro aangezien heptylgallaat de laagste MIC, de laagste bindings-constante en de hoogste activiteit tegen FtsZ in vitro had. Deze bevindingen tonen gezamenlijk aan dat FtsZ een direct doelwit voor alkylgallaten is. In een onderzoek van Takai et al. werd aangetoond dat alkylgallaten zich kunnen richten op celmembranen, wat ook hier door middel van een membraanper-meabiliteitsanalyse aangetoond werd. Omdat de delokalisatie van FtsZ-eYFP van het midden van de cel door alkylgallaten niet een gevolg is van de verstor-ing van de integriteit van het celmembraan, kan worden besloten dat alkylgal-laten bacteriën doden door een combinatie van mechanismen: verstoring van de integriteit van het celmembraan en het belemmeren van de celdeling.

De resultaten beschreven in dit proefschrift geven nieuwe informatie voor het FtsZ onderzoek in drie verschillende gebieden: de interacties tussen FtsZ en eiwitpartners, FtsZ als doelwit voor de ontwikkeling van antibiotica en de normalisering van de condities in FtsZ analyses.


176

ABOUT THE AUTHOR

Ewa Cendrowicz (fam. Król), born on 8th May, 1986 in Łódź, Poland. In personal life, she is a content wife and a mother of two children. Her professional career is connected to biotechnology and bio-chemistry.

EDUCATION AND RESEARCH

Biochemist, MABION S. A. Konstantynów Łódzki, Poland (2016 - …)

PhD in the Department of Molecular Microbiology at the University of Groningen, The Netherlands

Research field: Interactions of cell division protein FtsZ from Bacillus subtilis with large and small molecules. (2011-2015)

Erasmus and Master internship at the Instituto de Tecnologia Quimica e Biologica, Oeiras, Portugal

Research field: Purification of subunit B of Cholera Toxin from Vibrio cholerae. (2010)

Internship at the Institute of Medical Sciences in Aberdeen, Scotland (2009)

Internship at the Institute of Agricultural and Food Biotechnology in Łódź, Poland (2008)

International exchange program “Campus Europae” at the Department of Biotechnology at the University of Aveiro, Portugal (2007-2008)

Master of Science degree and the Department of Biotechnology and Food Technology at Technical University of Lodz, Poland (2005-2010)

LIST OF PUBLICATIONS APPENDICES

177

LIST OF PUBLICATIONS

Król, E., de Sousa Borges, A., da Silva, I., Polaquini, C. R., Regasini, L. O., Ferreira, H., & Scheffers, D. J. (2015). Antibacterial activity of alkyl gallates is a com-bination of direct targeting of FtsZ and permeabilization of bacterial membranes. Frontiers in microbiology, 6.

Król, E., & Scheffers, D. J. (2013). FtsZ polymerization assays: simple proto-cols and considerations. Journal of visualized experiments: JoVE, (81), e50844-e50844.

Król, E., van Kessel, S. P., van Bezouwen, L. S., Kumar, N., Boekema, E. J., & Scheffers, D. J. (2012). Bacillus subtilis SepF binds to the C-terminus of FtsZ. PloS one, 7(8), e43293.

ACKNOWLEDGMENTS APPENDICES

179

ACKNOWLEDGMENTS

So many things have happened in only 5 years in the city of Groningen! A PhD was the goal of my staying in The Netherlands, but next to the delightful mo-ments in Molmic group, hard work and grand time with wonderful people in the green building, my private life was also very abundant. Marriage, arrival of two beautiful children, developing friendships, that I hope will last my whole life. My time here, in Groningen, was filled with hard work, wonderful time with family and friends and a lot of love and happiness. Now, is the right time to thank all of the people who were part of my staying here. Without all of you, these 5 years would not be the same.

PROMOTORS

First of all, I would like to give many thanks to my promotor and daily super-visor Dirk-Jan. Thank you for encouraging me to apply (while still in Portugal) and for selecting me for this vacancy. You taught me all FtsZ techniques al-ready during the first week of my contract. You gave me more supervision in the beginning and less and less as my experience increased. It brought me to some level of independence in my work. Even if it was sometimes difficult and caused many mistakes, I appreciate your way of leading me to my future professional life. Thank you for interesting projects and all of the possibilities of collaborations. During the biggest panic attacks in the last year of my PhD studies, you always told me: “don’t worry, Ewa”. In the end, I want to thank you very much for many talks we had considering the matter of staying on the path of science, finding a work-life balance and many more. Usually, all of it did not make taking a decisions easier or more clear, but they definitely encour-aged me to take a broader view on things.

Arnold. Thank you very much for your advice and patience, especially in difficult tasks. I appreciate your encouragement to think more about my own projects. It was a real pleasure to be a part of your group.

MY DEAR PARANIMFS

ANA! At work, I spend most of my time with you. I think around 85% (the re-maining 15% is our holidays which, unfortunately, we could not spend together).


180

Office, lab, lunch breaks, “I am going to …, do you want to walk with me?”, “sure!”. Laughs, tears, jokes, complains… Very quickly you became my close friend. Thank you for all the time spent together and your priceless support during this time. You are very special to me! Elke. You are a supermom! I am very happy I have met you. Thank you for small chats and advice. It was always nice to stop by your office on the way to coffee machine. Thank you for the Dutch summary. Girls, thank you for your help with the preparation of my defence day.

COLLABORATORS AND STUDENTS

Eve, Sebas. It was a pleasure to work with you. You are both very smart. You caught all the methods very quickly and produced many many good results. Isa, Raquel. Thank you for working with me in the lab and for the nice projects which you brought to our group. Raquel, we had intensive time finishing experi-ments. Spending the whole evening in the lab, only with you! Laura, we worked on two projects together. It was always intensive and successful work. And it was pleasure for me. Mariana Pinho, Henrique Ferreira, Egbert Boekema, Tanneke den Blaauwen. Thank you for the possibilities of joining your interest-ing projects. Not always everything worked out, but everything I learned from it is priceless. Egbert, thank you for your trust. Gert, thank you for looking at my projects with a careful eye of an expert. Marc, your advice are always the best. You thought me everything I know about electron microscopy. But, you also helped me to find my way as a PhD student. Thank you for all the talks!

Hania. Thank you for all conversations about “these stupid metalloen-zymes”. Thank you for all your advice , e-mails etc. Without you it would not go that far. Marysia, thank you for advice on LICs.

ALL MOLMIC MEMBERS

Ana, thank you for the ideas, help and advice when I was not sure about the exact direction. Thank you for successful experiments in our projects. Gosia K. Thank you for many evenings with finding good solutions for difficult projects. Sabrina. Thank you for days in front of nanotemper and advice on labelling protein, Marteen for your fluorescence advice. Ilja, Samta, Jeanine, Alexej. Your help was priceless, especially at the beginning of my PhD. It is very good to have post-docs – experts in the group. Janny, Greetje, Anne-Bart, Stefan,


181

thank you for technical support, which I needed often. Juke, Intan, Irfan, Antonella, Reto, Ciprian, Erik, Jan-Pieter, Lito, thank you for small and big advice. Bea! Have I already told you that you are simply the best? Without you, I would not survive all the difficult days and matters at the university. Each time I asked you for help or advice, you solved my problems in 10 seconds. How was it possible? A superwoman? Anmara and Manon, thank you for your support. Other Molmic members with whom I spent wonderful time . And to these who always greeted me with smile on the corridor: Annarita, Amalina, Yang, Ciprian, Min, Danae, Hyun, Susan, Jeroen, Zsofia, Yanping, Carsten (thank you for switching off fluorimeter at 1 a.m.;)), Fabiola, Loes, Magda. Reto, Sasha, Ciprian, thank you for small talks about life. Gosia A., your office had a very good location! It was always within reach! Thank you for all your ad-vices on maternity, work, life in Groningen, life choices. I appreciate becoming close friends so quickly. When you left the group, it became quite empty. Oth-er Molmic members and students: Menno, Yvonne, Andre, Yannick, Sandra, Niels, Marta, Marta S., Fernando, Riccardo, Javier, Tomek, Ali, Jelger. It was a very good time in the Molmic group. Thanks to all of you.

MY DEAR POLISH FRIENDS FROM GRONINGEN

Marysia and Maciek, I have never met people like you before. You are just amazing people and a great example for others. You quickly became my fam-ily’s best friends in here and you are people that I will miss the most when I leave. Thank you for your friendship, dinners, meetings, and for all the help you offered many times when me and my family were in trouble. Gosia, Sławek, Jagódka, we also became good friends so quickly. Gosia, I missed you in the group and outside the laboratory, after you made the decision about moving back to Poland. Especially, I miss your jokes and comments straight to the point with a small glimpse of ironic remarks. Agata, Gosia and Magda were first Pol-ish people in Molmic that I met when arrived for my job interview. Agata, thank you for introducing me to the Polish group and being a big support during my first days in Groningen. Hania, you are a special person, always very nice and helpful. Thank you for your support with my projects. It was always nice to talk to you. Gosia M, another Gosia but a special Gosia. Thanks to you and your prayers, I finished my thesis. Thank you very much for your time, friendship and


182

our prayers together. Another Polish people with whom I spent splendid time here in Groningen are: Pani Isia, Pani Krysia, Weronika, Gabrysia, Paweł, Kasia, Maciek, Dawidek, Ola, Maciek, Hubercik, Jacek, Wojtek, Justyna, Agata, Piotrek, Monika, Adam i Oskarek. You were a very important part of my and my falmily’s life here, in Groningen. With you, it always felt like being at home.

Gosia, Janek, Marysia i Olafek, thanks for a very good time together. It’s a pity we have met so late. Polish parish in Groningen with our priest Józef Okonek. Thank you for having a place where I could pray in my own language and nice community where, together with my family, I felt safe.

MORE FRIENDS…

Opstap mothers, thank you for teaching me Dutch, so pity it was just for a few months. My friends in Poland: Ania, Marysia, Kasia, thank you for having time for me every time I was back in Łódź.I want to thank all other people which I met here, with whom I spend a great time but were not mentioned here.

MY PARENTS

Kochani rodzice, bardzo Wam dziękuję za wsparcie, pomoc, wychowanie. Dz-iękuję Wam za czas, który poświęciliście, żeby udało mi się skończyć doktorat. Mamo! Jesteś wyjątkowa. Dziękuję, że zawsze znajdowałaś czas, kiedy Cię potrzebowaliśmy, i za miłość, którą dajesz nam i naszym dzieciom. Dziękuję mojej teściowej za pomoc przy opiece nad Milenką. Rodzeństwu: Łukaszowi, Marcie, Beacie, za pomoc przy załatwianiu spraw w Polsce. Łukasz, dziękuję Ci za pomoc przy naszym mieszkaniu i służenie radą. Babciu, dziadku! Dobrze, że jesteście.

AND MY DEAR CLOSEST FAMILY.

My dear husband, Łukasz. Thank you for coming with me to Groningen. Thank you for your patience, support and love. Thank you for all the drawings you have made and patience to read my thesis so carefully. I love you and I am very happy with you! Moje Kochane Dzieciaczki: Milenusia i Mieszkuś. Dziękuję Wam za cierpliwość, za Wasze cudowne uśmiechy i szczerość. Nauczyłam się od Was i dzięki Wam bardzo wiele. Kocham Was bardzo! Milenko, dziękuję za


183

prześliczny rysunek na okładkę. Together, we have built a place to which it is always so nice to come back after a difficult day at work. Our home is a safe place where I always feel fantastic. Thank you Łukasz, Milenka and Mieszko for being with me all the time. I love you!

Drogi Boże, dziękuję Ci za opiekę. Bez Ciebie nic nie byłoby tak jak jest.

university of groningen interactions of cell division protein ftsz with large and small molecules...

Documents