STRUCTURAL AND BIOCHEMICAL INVESTIGATIONS OF THE MECHANISM OF

HEME CAPTURE BY THE HEMOPHORE HasAp SECRETED BY

PSEUDOMONAS AERUGINOSA

BY

GRACE JEPKORIR LAGAT

B.Ed. (Sci.), Kenyatta University, 1999 Nairobi, Kenya

M.Ed. (Adm.), University of Eastern Africa, Baraton, 2002 Eldoret, Kenya

Submitted to the graduate degree program in Chemistry and the Graduate Faculty of the University of Kansas in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

Committee members:

___________________________ Mario Rivera, Ph. D. (Chair) ___________________________ Robert C. Dunn, Ph. D. ___________________________ Susan Lunte, Ph. D. ___________________________ Mikhail Barybin, Ph. D. ___________________________ Minae Mure, Ph. D.

___________________________ Wonpil Im, Ph. D.

Date Defended: April 19, 2011

ii  

The Dissertation Committee for Grace Jepkorir Lagat certifies that

this is the approved version of the following dissertation:

STRUCTURAL AND BIOCHEMICAL INVESTIGATIONS OF THE MECHANISM OF

HEME CAPTURE BY THE HEMOPHORE HasAp SECRETED BY

PSEUDOMONAS AERUGINOSA

Committee:

___________________________ Mario Rivera, Ph. D. (Chair) ___________________________ Robert C. Dunn, Ph. D. ___________________________ Susan Lunte, Ph. D. ___________________________ Mikhail Barybin, Ph. D. ___________________________ Minae Mure, Ph. D.

___________________________ Wonpil Im, Ph. D.

Date Approved: April 19, 2011

iii  

ABSTRACT

HasAp is a hemophore secreted by Pseudomonas aeruginosa to the extracellular media

under iron limited conditions to sequester heme from the host cell. The heme HasAp complex is

captured by the specific cell surface receptor HasR for subsequent internalization. X-ray and

solution NMR structures of holo and apo hemophores have been solved but so far no studies

have been done to elucidate the mechanism of heme loading. In the apo form, the loop bearing

the distal heme iron ligand His 32 is located nearly 30 Å away from its position in the holo form.

In contrast, the loop bearing the proximal ligand Tyr 75 maintains structural integrity in the apo

and holo forms. This study aimed at investigating the sequential steps that lead to heme binding

by HasAp and the role played by the axial-ligand bearing loops in the process of heme capture. A

chromatographic method was developed to isolate and purify wild type and H32A apo HasAp.

Heme binding was monitored by stopped flow spectroscopy. The results show that heme loading

proceeds in two discrete kinetic phases; an initial rapid phase which entail loading of the heme

onto the Tyr 75 loop followed by a slow phase where the His 32 loop undergoes a large

conformational rearrangement and subsequent coordination of the heme. Molecular dynamic

simulations provided more insights into the molecular motions that relay the presence of the

heme in the Tyr 75 loop to the His 32 loop.

To investigate the role played by the axial ligand bearing loops in heme binding, crystal

structures and solution NMR studies of the apo and holo H32A-HasAp, as well as H32A-

imidazole complex were carried out. The results reveal that the His 32 loop assumes a position

similar to the wild type protein even in the absence of a coordinating residue. This implies that

the presence of the heme in the Tyr 75 loop triggers the relocation of the His 32 loop and that this

iv  

loop is important in protecting the macrocycle from the aqueous media. The crystal structures of

Y75A and H83A show that the proteins assume the original fold although there are some

conformational changes in the His 32 loop and in the secondary structure elements in regions of

the protein implicated to interact with the receptor during heme release. These observations

imply that information is relayed between the Tyr 75 loop and the secondary structure elements

affected likely via the His 32 loop. It was also established that His 83 does not necessarily

coordinate the heme in the absence of Tyr 75.

Heme transfer experiments using methemalbumin as a heme source revealed that the rate

of heme release from methemalbumin correlates with the rate of heme uptake by apo-HasAp.

These rates are comparatively higher than the rate of heme uptake from other host heme proteins;

however, they are comparable to the rate of dissociation of heme from methemalbumin in the

absence of the hemophore. These results mean that methemalbumin is likely the potential target

for HasAp and that heme transfers by passive diffusion and is affinity driven.

v  

ACKNOWLEDGMENTS

First and foremost I thank God for life, wisdom, and knowledge, and for sustaining me,

through it all, during the course of my studies.

Northing worthwhile is ever done alone under the sun. My achievement is a worthwhile

course that has taken the positive minds and caring hearts of all the people who have crossed my

path ever since I took the initial step to climb the academic ladder. I am always and will ever be

grateful to my dear parents for raising me and bringing me this far through their endless love,

self-sacrifice, encouragement, and most of all for believing in me to make the best out of every

circumstance. “Dad and Mom, I was always reassured of your great love, care, and unfailing

support whenever you made a direct transatlantic call just to check on us. I reiterate my

appreciation - THANK YOU SO MUCH. We made it because of you.”

I am also deeply and humbly grateful to every other individual who in one way or another

made it possible for me to complete this journey. I especially thank my advisor, Dr. Mario Rivera

for granting me the opportunity to learn from him and work with him, and for guidance,

encouragement, and continuous support throughout my studies. I am also indebted to my

committee members: Dr. Robert Dunn, Dr. Susan Lunte, Dr. Mikhail Barybin, Dr. Minae Mure

and Dr. Wonpil Im for all their help and advice, and for taking the time to critique my work.

I owe great appreciation to the current and past members of Dr. Rivera Lab: Dr. Juan

Carlos Rodriguez, Dr. Yuhong Zeng, Dr. Christopher O. Damaso, Dr. Andy Wang, Dr. Aileen Y.

Alontaga, Bailey Morgan, Jordan Stobaugh, Maggie Murphy, Casey Gee, Dr. Huili Yao, Dr.

Saroja Weeratunga, Yan Wang, Pavithra N. Vani, Ritesh Kumar and Kate Eshelman. Thank you

vi  

for your support, positive criticisms, and encouragement more so when I had to burn the

midnight lamp working in the lab.

I thank our collaborator Dr. Pierre Moënne- Loccoz for the EPR, RR, and SFS results; Dr.

Scott Lovel for his helpful advice on screening for protein crystal growth and for solving the

crystal structures of the proteins; Kevin. P. Battaile for collecting crystallographic data; and Dr.

Wonpil Im and Huan Rui for the molecular dynamics simulation results.

I also wish to thank Pr. Tony Cash and all the Lawrence (KS) and Philadelphia (IA)

Adventist church members, all my relatives and neighbors, and everyone else who stood by me

throughout this journey. I cannot enlist all your names but please know that your prayers, your

love and your support were a constant source of strength and encouragement.

To my great mentor Prof. Denford Musvosvi, and my true and exemplary friends Dr.

Donita Massengill-Shaw, Dr. Jeraldine Wilson, Dr. Janet Tarus, Dr. Maru Kering, Sarah and

Luka Kapkiai, Louisa Murunga, Dr. Aileen Alontaga, Rebecca Llasaca, Sayool Oh, and Dr.

Jackie Obey, thank you so much for your friendship, selfless actions and kind thoughts. You

were my inspiration to strive and to work harder.

Last but not least I owe a lot to all my wonderful siblings: Luke, Joshua, Janet, Japheth,

Damaris, Abigael, Lucy, Rupo, and Gladys. Words cannot express how much I appreciate you

all for the many ways you supported me and gave me your shoulders to lean on at all times. I

will never be able to give back anything in return but please know that I deeply appreciate each

and every one of you.

To my husband Steve, “you are one of a kind. Thank you for opening the door for me,

bearing my absence from home for the longest possible period, and for enduring all the

challenges to the very end.” My precious daughter Beverly, “your unfailing love, your smile, and

vii  

your patience gave me a reason to rise up every day determined to achieve my goal. This

milestone is our achievement. God has brought us this far and together we have made it!” Thank

You Bev. Thank You All.

viii  

DEDICATION

To

My Dear Parents

(Mr. and Mrs. Eliud Lagat)

ix  

TABLE OF CONTENTS

CHAPTER I 1 INTRODUCTION….……………………………………………………………………….…...1

Iron uptake by bacteria………………………………………………….…………….…...2

Heme uptake systems in Gram negative bacteria................................................................ 4

Hemophore-dependent heme acquisition systems…...…………………..................…......6

Proposed working mechanism of hasA...............................................................................7

Structure of the model hemophore HasAs…………………………..………………….. 12

Mechanism of heme uptake in HasA……………………………..…….…………….….16

Heme binding studies…………………………………………………………………….16

RESEARCH PROBLEM AND RATIONALE…….………………………………………….....17

REFERENCES ……………………………….………………………………….…...……..…. 20

CHAPTER II 27

Structural, NMR Spectroscopic and Computational Investigation

of Heme Loading in the Hemophore HasAp from Pseudomonas

aeruginosa

INTRODUCTION ……….……………….……………………………….…………….………27

EXPERIMENTAL PROCEDURES ….……………..……………………………......................30

Construction of H32A-HasAp ………………….…………………………….…………30

Expression and Purification of apo-proteins………………………………………….…30

Expression and Purification of Uniformly Labeled Proteins …………...………………33

Expression and Purification of Selectively Labeled Proteins……………………………34

Sample preparation……………………………………………………………..………..35

Determination of molecular weight …………………………………………….……….36

Resonance Raman (RR) and EPR Spectroscopy……………………….……………..…38

X-Ray Crystallography…………………………………………………………………..38

NMR Spectroscopy…………………………………..….……………………………….39

x  

Targeted Molecular Dynamics (TMD) Simulations …….…………………..…………..42

RESULTS…………………………………………………………………………………..…....46

Purification of the apo-proteins ………………………………………………………....46

Structural Characterization of apo-HasAp ….…………………………………………..46

Conformational changes observed upon hemin binding onto apo-HasAp……...……….51

Characterization of H32A-HasAp mutant……………………………………………….53

Structural Characterization of H32A-holo-HasAp by X-ray crystallography…………...57

Imidazole complex of monomeric H32A-holo-HasAp…………………………………..62

X-ray crystal structure of the H32A-imidazole complex………………………………...62

Characterization of monomeric H32A-HasAp by NMR Spectroscopy………………….65

Targeted Molecular Dynamics Simulations (TMD)………………………….….………72

DISCUSSION …………………………………………………………………………………...76

Structure of apo-HasAp and conformational changes that occur upon heme loading…..76

Characterization of H32A-HasAp mutant…………………………………………….…77

Kinetic and spectroscopic studies reveal a two-step mechanism of heme-loading

by HasAp…………...………………………………………………………………...….78

Structural characterization of H32A-HasAp mutant by X-ray crystallography

and NMR spectroscopy…………………………………………………………………..80

Relocation of His32 loop proceed via a well-defined path ………………………….…..82

APPENDICES….………………………………………………………………………………..84

REFERENCES ………………………………………………………….……………...…...…..97

CHAPTER III 103

Effects of Proximal Ligand Substitutions on Heme Coordination

and Structural Integrity of HasAp

INTRODUCTION ……………………………………………………………………………..103

EXPERIMENTAL PROCEDURES……….………………………………………...................106

Construction of Y75A and H83A HasAp mutants….……………….………….………106

Expression and purification of Y75A and H83A mutants..…………………………….107

Preparation of holo protein………………………………………………………...…...110

xi  

Determination of extinction coefficient for the Y75A- and H83A-heme complex.……110

Protein crystallization……………………………………………………………...…...111

Data collection, structure determination and refinement..……………..........................112

RESULTS ………………………..…………………………………………………………….117

Purification of Y75A and H83A mutants…………………………………………..…..117

Characterization by UV-vis spectroscopy……………………………...………………118

X-ray Crystallography……………………………………………………………...…..121

Crystallization of Y75A and H83A HasAp mutants ………………………..…122

Crystal structures of Y75A and H83A HasAp mutants ……………………….…...…...123

Heme iron coordination and conformational changes involving His 32 and

Tyr 75 loops…. …...........................................................................................................129

Overall structure of HasAp mutants .………………………………………………….130

DISCUSSION ………………………………………………………………………………….136

APPENDICES …………………………………………………………………………………140

REFERENCES ………………………………………………………………………………...145

CHAPTER IV 148

Methemalbumin is the potential target for the hemophore HasAp

as a heme source; inferences from heme transfer studies

INTRODUCTION ………………………………………………………………......................148

EXPERIMENTAL PROCEDURES …………………………………………………………...150

Expression and Purification of apo-HasAp……...……………………………………...150

Preparation of Human Serum Albumin (HSA)…………………………….…………...151

Quantitation of heme in methemalbumin…………………………………...………….151

Heme transfer assays….………………………. …………………………………….....152

Evaluation of heme transfer by polyacrylamide gel electrophoresis (PAGE)…….……152

Quantitation of heme transfer……………………………………………...………...…153

xii  

RESULTS AND DISCUSSION …………………………………………………....................153

Sample preparation…………………………………………….……………………….153

Spectroscopic changes observed upon incubation of apo-HasAp

with methemalbumin……………………………………………………………...........154

Heme uptake by apo-HasAp correlates with heme loss from methemalbumin………...157

Evaluation of heme transfer by PAGE………………………………………………….160

Heme transfer from methemalbumin to H32A-HasAp mutant……………………........164

REFERENCES…………………………………………………………………………………167

CHAPTER V 170

SUMMARY ……………………………………………………………………………………170

REFERENCES ………………………………………………………………………………...175

xiii  

LIST OF TABLES

CHAPTER 2

Table 1. Crystallographic data of HasAp structures........................

44

CHAPTER 3

Table 1. Data collection and Refinement Statistics of HasAp mutants …………………………………........................

116

Table 2. Distances between key residues or molecules and heme-

iron……………………………………………………….

129

xiv  

LIST OF FIGURES

CHAPTER 1

Figure 1. Map of the P. aeruginosa heme uptake locus containing the phuR gene and the phuSTUVW operon. Three Fur binding elements are shown as white boxes (adapted from ref 34)……....

4

Figure 2. Genetic organization of the has systems in P.a., Pseudomonas aeruginosa and S.m., Serratia marcescens. HasI and hasS encode sigma and antisigma factors, respectively. The white boxes indicate consensus Fur boxes (adapted from ref 41)……...

6

Figure 3. Simplified schematic of the proposed working mechanism of the has system………………………………………………………..

8

Figure 4. Schematic illustration of the protein complex involved in energy transduction from the cytoplasmic membrane to the outer membrane in Gram-negative bacteria. OM: outer membrane; PP: periplasmic space; CM: cytoplasmic membrane (adapted from ref 54)…………………………………………………………….

9

Figure 5. Comparison of the crystal structure of Holo-HasASM (PDB ID. 1DKO) (A), and the crystal structure of HasAp (PDB ID. 3ELL) (B) showing the elements of secondary structure. The -helices are shown in cyan, -sheet in magenta, loops in green, heme in red and ligands in blue. His 83 is colored orange…………….....

13

Figure 6. Zoomed in view of the heme binding site in the crystal structure of HasAs. (A) solvent exposed prosthetic group; heme is shown in red while the water molecules are represented by blue spheres. (B) Hemin coordination ligands His 32 and Tyr 75 (blue) and the adjacent residue His 83 (orange)……………………………..

14

xv  

Figure 7. Amino acid sequence alignment of full-length HasA sequences

from P. aeruginosa (Pa), Serratia marcescens (Sm), P. entomophila (Pe), P. fluorescens (Pf), Yersinia pseudotuberculosis (Yps), Pectobacterium carotovorum (Pc), Erwinia carotovora (Ec), Yersinia pestis (Yp), and Yersinia pestis bv. Antiqua (Ypa) aligned against the sequence of HasA from P. aeruginosa. Truncated HasAp used in these investigations stops at residue 184 (full-length minus 21 amino acids) [13]. Residues highlighted in red are conserved. The alignment was performed using ClustalW [68]…………………..

15

CHAPTER 2

Figure 1. X-ray crystal structure of truncated HasAp monomer B (PDB: 3ELL) showing -helices (cyan), -strands (magenta), loops (wheat), heme (red), and heme axial ligands (green)………...................………………………. 29

Figure 2. DNA and amino acid sequence of HasAp-TR. The Nde I and BamHI restriction endonuclease sites were constructed at the 5' and 3' ends, respectively, for subcloning. The underlined codon represents the mutation of His to Ala at position 32…….……………………………………….......

32

Figure 3. The simulation system of hemin bound apo-HasAp. (A) Hemin bound apo-HasAp. (B) Hemin bound apo-HasAp with 0.15 M KCl. (C) Hemin bound apo-HasAp with 0.15 M KCl aqueous solution. Apo-HasAp (yellow) is shown in cartoon representation with the His32 and Tyr75 loops highlighted in magenta. Hemin is shown in red sticks, hemin iron as an orange sphere, K+ as magenta spheres, and Cl- as green spheres. Waters molecules are represented by lines (marine)……………….……………….

43

Figure 4. X-ray crystal structure of apo-HasAp (PDB: 3MOK). His32 loop is shown in green, Tyr75 loop is shown in red, while His 32 and Tyr 75 are shown in magenta and orange, respectively………………………………………………....………… 47

Figure 5. Overlaid x-ray crystal structures of HasAp showing the position of the loops in apo- (green) and holo-(red) wild type protein. Hemin is colored blue, axial ligands in holo-HasAp are shown in orange …………………………………………….………………...………

48

xvi  

Figure 6. Amino acid sequence alignment of full-length HasA sequences from P. aeruginosa (Pa), Serratia marcescens (Sm), P. entomophila (Pe), P. fluorescens (Pf), Yersinia pseudotuberculosis (Yps), Pectobacterium carotovorum (Pc), Erwinia carotovora (Ec), Yersinia pestis (Yp), and Yersinia pestis bv. Antiqua (Ypa) aligned against the sequence of HasA from P. aeruginosa. Truncated HasAp used in these investigations stops at residue 184 (full-length minus 21 amino acids) [13]. Residues highlighted in red are conserved while those in blue are the His32 loop “hinge” residues in HasAp. The alignment was performed using ClustalW [45]…………………….……………………………………………………. 49

Figure 7. (A) Overlay of apo (magenta) and holo (turquoise, PDB: 3ELL) HasAp showing a zoomed in view of the “hinge” region of the His32 bearing loop.(B) Apo-HasAp showing the aromatic side chains (green) that stabilize the �1-loop near the Gly 28 hinge; the H32 loop is shown in magenta. (C) Apo-HasAp showing the aromatic side chains (green) that stabilize the structure near the Gly 44 hinge. The hemin molecule of holo-HasAp (PDB: 3ELL) is shown in red for reference…..…………................................................................................... 50

Figure 8. Overlay of apo (magenta) and holo (turquoise, PDB: 3ELL) HasAp (A) Zoomed in view of the heme binding pocket showing conformational differences between certain residues upon hemin binding. (B) Displacement of helix �2 and the short loop comprised by G143 and D144 relative to the apoprotein upon hemin binding; movement of the latter enables hinge residue G 28 in the holo structure to adopt its hemin bound conformation…………………….…………………………………………..

52

Figure 9. (A) Elution profile of H32A holo-HasAp from a Sephadex G-75 column; open circles track the Soret band absorption (∼A400 nm) and black circles track A280 nm. (B) Electronic absorption spectra of fractions in the fast eluting peak (black, dimer) and in the slow eluting peak (red, monomer). The ratio of dimer to monomer obtained under these experimental conditions is ∼15%. (B) Elution volumes (Ve) from a calibrated Superdex 75 column and estimated molecular weights for monomer holo H32A HasAp (red triangle), dimer holo H32A HasAp (green diamond), and apo H32A HasAp (blue triangle); the average molecular weight was obtained from two experiments. The Superdex 75 column was calibrated with albumin (67 kDa), ovalbumin (44 kDa), chymotrypsin (25 kDa), and cytochrome c (12.4 kDa) shown in black filled circles….……………….…

54

xvii  

Figure 10. (A) Resonance Raman spectra (obtained at 105 K) of the first and second fractions of H32A-holo-HasAp (λexc = 413 nm, 20 mW; the spectra were normalized on the intensity of the 4 observed at 1375 cm-1). (B) X-band EPR spectra of the first and second fractions of H32A-holo-HasAp obtained at 10 K (0.03 mW microwave power, 9.66 GHz microwave frequency; 10 G modulation amplitude). The two spectra were normalized based on their hemin content determined by the pyridine hemochrome assay.………………………………………………………..………….…… 55

Figure 11. 15 % SDS-PAGE of HasAp pure samples; M- protein ladder, 1- slow eluting band (monomer), 2- fast eluting band (dimer), and 3-apo-HasAp.………………………………………...……………………………. 56

Figure 12. (A) Structure of the H32A holo-HasAp dimer. Subunits A is colored light pink and B is colored slate. Disordered residues in subunit B are located between the asterisks. Heme in subunit A and B is colored pink and blue, respectively. (B) Proximal site of heme in subunit A showing Tyr 75 coordination, and H-bonding interactions involving the propionates and Arg 129 from the same subunit and Tyr 138 from subunit B. Tyr75 loop is colored wheat. (C) 2Fo – Fc electron density maps of the hemin molecules in the dimer contoured at 1σ represented as a green surface ………................................................................................................

58

Figure 13. (A) Superimposed crystal structures showing H32 loops of apo wt (magenta), holo wt (cyan), and holo H32A (red) HasAp. The heme molecules associated with holo wt and holo-H32A are colored blue and gold, respectively (B) Spherical representation of subunit A and associated heme (orange). The H32 loop is colored red, and hydrophobic residues Phe 51, Tyr 75, Leu 77, Phe 78, and His 83 in the heme binding site are colored green………….…………………………………………………………….. 59

Figure 14. UV-Vis spectra of monomeric holo-H32A-HasAp (red) and induced dimer (black). …………………………………………………….....…………….

61

Figure 15. Family of electronic absorption spectra (A) and the binding curve (B) obtained upon mixing imidazole and H32A holo-HasAp. The complex was formed by adding fresh protein (final concentration 5 uM) to known concentrations of imidazole buffer (0-100 mM), pH 7.0. Each data point was recorded after 2 min incubation.……………………………………..… 63

xviii  

Figure 16. X-ray crystal structure of H32A-imidazole complex. (A) Molecule A; heme and imidazole are colored magenta while His 32 and Tyr 75 loops are colored light pink. (B) Molecule B; heme and imidazole are colored cyan while the His32 and Tyr75 loops are colored light cyan. Disordered residues are indicated between the asterisks in both molecules. (C) 2Fo - Fc electron density maps of the heme and heme-bound imidazole in molecule A (top) and molecule B (bottom) of the H32A-imidazole complex contoured at 1σ. (D) Surface representation of the H32 loop in the structure of the H32A-imidazole complex. The imidazole molecule (magenta) is locked in place by Val 38 and Met 141 side chains, and a hydrogen bond formed with the carbonyl oxygen of Asn 42. (E) Superposition of molecule A of the H32A-imidazole complex (magenta) with molecule A of wt holo-HasAp (cyan). Residues in the H32 loop are colored darker…………………………………………………………...….. 64

Figure 17. 1H-15N-HSQC-TROSY spectra of H32A HasAp (A) and H32A-imidazole (B). Spectra were taken at 32 oC using a Varian Inova spectrometer operating at 600 MHz 1H frequency. Acquisition parameters are as follows: 1,710 (1H) x 400 (15N) complex points; 8.9 kHz (1H) and 2.3 kHz (15N) spectral widths; 8 scans per increment; d1=1s. Spectra were processed using NMRPipe software as follows: data in the 15N dimension were linear predicted to twice the number of points and zero-filled in each dimension to yield a data matrix of 4k x 4k points. Spectra were referenced to internal DSS using the methyl peak at 0 ppm for the 1H dimension and indirectly using the ratio of gyromagnetic ratios for the 15N dimension…………………………………………………………………… 66

Figure 18.

Resonance assignments available for H32A-HasAp (A), H32A-imidazole (B), and wt HasAp (C) are mapped in gray on the structure of wt holo-HasAp; residues for which resonances could not be obtained are shown in green, prolines are in blue, and the hemin is in red…...…………………….

67

Figure 19. HSQC spectra obtained with samples of 15N-Gly-H32A-imidazole using standard (A) and fast-repetition (B-E) conditions. Spectra were taken at 32 °C (A and D), 15 °C (B), 25 °C (C), and 38 °C (E). Cross-peaks are labeled according to the corresponding sequential assignment. The cross-peak identified as Gly44 is observed only in spectra collected at 32 and 38 °C using a short relaxation delay (50 ms) and short acquisition time (35 ms). The spectra were recorded with 100 increments in t1 (15N); 630 complex points in t2 (1H), over a spectral width of 8.9 kHz (1H) and 2.3 kHz (15N), with 256 scans per increment for all temperatures except 15 °C, for which 512 scans were used. The data were processed using NMRPipe applying a 90°-shifted squared sine bell apodization function in each dimension,

xix  

followed by zero filling to 4k data points in t2, and linear prediction to twice the number of points and zero filling to 4k data points in t1. Subsequent Fourier transformation and phasing rendered matrices of 4k × 4k points………………………………………………….…...……………

69

Figure 20. HSQC spectra obtained with samples of 15N-Val-H32A-HasAp (A and B) and 15N-Val-H32A-imidazole (C and D). Spectra in (A) and(C) were acquired with standard acquisition parameters, whereas spectra in (B) and (D) were collected with fast-repetition conditions. Dashed lines in spectra (C) and (D) highlight one of the cross-peaks attributed to Val 37 or Val 38, which is clearly visible only in the HSQC spectrum (and corresponding 1D trace) acquired with fast repetition conditions. Cross-peaks are labeled according to their sequential assignment. Cross-peaks originating from backbone amides of Ala residues are present because of isotopic scrambling, and are clearly identified by comparison with the HSQC spectrum obtained with uniformly labeled protein.…………………………

71

Figure 21. A possible path for closing of the H32 loop upon hemin binding; the heavy atom rmsd between initial (apo-) and final (holo-) structures is 5.73 Å, including snapshots at rmsd 5.42 Å (A), 4.87 Å (B), 4.06 Å (C), 2.64 Å (D), 1.61 Å (E), and 0.61 Å (F). The apo- (gray) and intermediate (cyan) structures are shown in cartoon representation. Residues involved in important interactions are shown in either stick or sphere representation, and when appropriate, the same residues in the apo-structure are also shown to indicate their initial positions. The His32 loop and the turn (Ser 60 to Ser 63) are highlighted in magenta and yellow, respectively. The hemin and hemin-iron are shown as red sticks and orange spheres, respectively..…………………………….......................................................

74

Figure 22. (A) Absorption spectra of wild type HasAp (black), H32A variant (red), and heme (blue). Stopped-flow absorption traces obtained with apo-wt (B) and apo-H32A (C) upon mixing with hemin……………………………….. 80

CHAPTER 3

Figure 1. Zoomed in view of the heme pocket in the hemophore HasAp. Heme coordination ligands His 32 and Tyr 75 are colored blue. His 83 is colored orange. ………………………………...……… 103

xx  

Figure 2. DNA and amino acid sequence of HasAp-TR. The Nde I and BamHI restriction endonuclease sites were constructed at the 5' and 3' ends, respectively, for subcloning. The underlined codon represents the mutation of Tyr or His to Ala at position 75 and 83, respectively ………...…………………………………………….

109

Figure 3. X-ray fluorescence scan for zinc using a H83A crystal ………… 114

Figure 4. Phased anomalous difference map calculated using data collected at λ=1.0000 Å (red mesh) contoured at 4σ showing the Zinc ion sites in H83A ……………………………..…….………………… 114

Figure 5. Phased anomalous difference map using data collected at λ=1.0000 Å (red mesh) and 1.28414 Å (low energy remote, green mesh) contoured at 4σ. The anomalous difference density increases at the hemin Fe atom and disappears at the Zinc ion site when using the low energy remote wavelength data……………...

115

Figure 6.

Elution profile of Y75A-HasAp from Sephadex G-75 column; open circles track the soret band abruption (~406 nm) and the black circles track absorbance at 280 nm. A similar profile was observed with the H83A-HasAp mutant….……………………… 117

Figure 7.

Elution volumes (Ve) from a calibrated Superdex 75 column and estimated molecular weight of Y75A HasAp (red triangle) and H83A HasAp (green diamond). The Superdex 75 column was calibrated with albumin (67 kDa), ovalbumin (44 kDa), chymotrypsin (25 kDa), and cytochrome c (12.4 kDa) shown in black filled circles…………………………………………………

118

Figure 8. Electronic absorbance spectra of Y75A (black) and H83A (red) recorded in sodium phosphate buffer (µ = 0.1, pH 7.8).…….…….

120

Figure 9. pH titration spectra of H83A over a pH range of 5.0 to 9.5 in sodium phosphate buffer , µ = 0.1.……………………………….. 120

xxi  

Figure 10. (A) Original crystals of H83A obtained from 10% isopropanol, 100 mM sodium cacodylate pH 6.5, 200 mM zinc acetate; (B) Refined crystals obtained from 10% isopropanol, 100 mM sodium cacodylate pH 6.5, 200 mM Zinc acetate, 3% dioxane. (C) Single crystals of Y75A obtained from Hampton Crystal Screen 2 #32 ((1.6M (NH4)2SO4, 100 mM HEPES pH 7.5, 200 mM NaCl)……………………………….………………………... 123

Figure 11. 2Fo-Fc electron density map contoured at 1; (A) Electron density at the heme showing that it adopts a single orientation; (B) Electron density of the refined formate ion; (C) Zoomed in view of heme binding pocket. (D) Sphere representation of Tyr75 loop and coordinated formate ion.………………………………...

125

Figure 12. 2Fo-Fc electron density map at the heme contoured at 1 (A), and zoomed in view of heme binding pocket in H83A crystal form 2 (B) and H83A crystal form 1(C)………………………………….

126

Figure 13. Heme-iron ligands and residues that make direct contact with the

propionates in WT, Y75A, and H83A HasAp mutants…………..

127

Figure 14. Zoomed in view of His32 and Tyr75 loops of wild type HasAp (magenta), Y75A (green), H83A crystal form 1(cyan), and H83A crystal form 2 (lemon)……………………………………………

128

Figure 15. Conformational changes in the backbone and side chains of amino acid residues that constitute His 32 loop and Tyr 75 loop in Y75A (green) and H83A (cyan) compared to WT (magenta)…....

130

Figure 16. X-ray crystal structures of (A) WT HasAp monomer B (PDB: 3ELL), (B) Y75A (C) H83A crystal form 2, (D) H83A crystal form 1. Secondary structures are colored cyan (a-helices), magenta (b-strands) and green (loops). 310-helices, reflecting a gain in secondary structure elements are colored purple (Fig. B and C). 5* and 7* (Fig C and D) are new hairpin loops

xxii  

resulting from loss of corresponding beta sheet character. Missing electron density is marked with white asterisks (Fig D)…………. 132

Figure 17. Conserved hydrogen bonding interactions involving residues in 2 (Asn 47), 3 (Ser 54, Gln 55) and the loop connecting 6 and 7 (Ala 102 and Tyr 107)…………………………………….……

134

Figure 18. Putty representation of B-factors color traced on the structures

increasing from blue to red and corresponding per residue

plots.................................................................................................. 135

CHAPTER 4

Figure 1. UV-vis spectra of methemalbumin (black), and holo-HasAp red)…………………………………………………….………….

155

Figure 2. A family of electronic absorption spectra of methemalbumin (~10 M) incubated with apo-HasAp (40 M) at 37 o C recorded as a function of time. Arrows show the direction of spectral changes as the reaction progressnot perturbed are in green. Residues which disappear are in marine. The heme is in magenta………..... 155

Figure 3. UV-Vis spectra of the reaction product (red) and holo-HasAp (black) with similar concentrations. The spectrum of methemalbumin (blue) is shown for reference …………….…….. 156

Figure 4. Elution profile of serum albumin and HasAp from a Sephadex G-75 column; blue circles track A280nm while red circles track the soret ………………………………………………………..……. 157

Figure 5. A plot showing changes in absorbance at 407 nm when methemalbumin is incubated with 4-fold excess apo-HasAp with constant stirring at 37 oC. ………………………………………. 158

Figure 6. A plot showing absorbance changes occurring in the visible region when methemalbumin was incubated with four-fold apo-

xxiii  

HasAp. The reaction was monitored as a function of time at 500, 535, 574, and 618 nm.…………………………………………… 158

Figure 7. 13 % native polyacrylamide gel showing the reaction products obtained from a mixture of apo-HasAp and methemalbumin incubated for 2, 5, 10, 20, 30, 60 min at 37 oC.…………………..

162

Figure 8. 13 % Native gel (A) and SDS gel (B), stained with coomassie brilliant blue. Samples were aliquoted at certain time intervals (1, 3, 5, 7, 12, 20 min), immediately flash frozen in liquid nitrogen, and thawed just before loading. Lanes C, C`, S, and S` are controls apo-HSA, methemalbumin, apo-HasAp, and holo-HasAp, respectively ………………………..……………………..

163

Figure 9. Elution profile of serum albumin and HasAp from a Sephadex G-75 column; blue circles track A280nm while red circles track the soret. ……………………………………………………………… 163

Figure 10. UV-Vis spectra of the reaction product (red) and holo-HasAp (black) with similar concentrations. The spectrum of methemalbumin (blue) is shown for reference..………………….. 165

Figure 11. Elution volumes (Ve) from a calibrated Superdex 75 column and

estimated molecular weights for the reaction products obtained

from a reaction mixture of methemalbumin and apo-H32A-

HasAp: holo-H32A-HasAp (21 kDa, green diamond), and serum

albumin ( 63 kDa, red triangle). The Superdex 75 column was

calibrated with ribonuclease A (75 kDa), Carbonic anhydrase (44

kDa), ovalbumin (29 kDa), and conalbumin (13 kDa) shown in

black filled circles...………………...............................................

165

xxiv  

LIST OF SCHEMES

CHAPTER 1

Scheme 1. Oxidative heme degradation catalyzed by heme oxygenase…….. 11

CHAPTER 3

Scheme 1. Proposed coordination state of the heme iron in H83A-HasAs at different pH. ………………………………………………..........

105

xxv  

LIST OF APPENDICES

CHAPTER 2

Appendix 1. Protein crystallization conditions-Wizard II Random Sparse Matrix ……..………………………………………………..….

84

Appendix 2. Backbone resonance assignments for H32A HasAp and H32A-imidazole ……………………………………………...……….

87

CHAPTER 3

Appendix 1. Protein crystallization conditions-Wizard II Random Sparse Matrix ………………………………………………………….

140

Appendix 2. ESI/MS spectrum of Y75A-HasAp ……………...……………. 143

Appendix 3. ESI/MS spectrum of H83A-HasAp ………………………...…. 144

xxvi  

ABBREVIATIONS

DMSO Dimethyl Sulfoxide

EPR Electron Paramagnetic Resonance

Has Heme Acquisition System

HasAp Heme Acquisition System Pseudomonas aeruginosa

HasR Heme Acquisition System Receptor

HasAs Heme Acquisition System Serratia marcescens

Hb Hemoglobin

Hp Haptoglobin

HPX Hemopexin

IPTG Isopropyl--D-thiogalactopyranoside

NMR Nuclear Magnetic Resonance

OD Optical Density

PCR Polymerase Chain Reaction

PDB Protein Data Bank

SDS PAGE Sodium dodecyl Sulfate Polyacrylamide Gel Electrophoresis

TMB Tetramethylbenzidine

TROSY Transverse relaxation optimized spectroscopy

RR Resonance Raman

UV-Vis Ultraviolet-Visible

1  

CHAPTER 1

Introduction

Pseudomonas aeruginosa (P. aeruginosa) is an opportunistic infectious microorganism

that is ubiquitous in nature; it inhabits soil, water, plants and animals including humans. It rarely

causes disease in healthy persons but has the potential to severely infect individuals whose local

or systemic immune defense is compromised [1,2]. As a nosocomial pathogen, it is the third most

commonly isolated organism, leading primarily to urinary track or wound infections in

immunocompromised patients [3,4]. According to the Center of Disease Control and Prevention

(CDC), the overall incidence of P. aeruginosa infections in US hospitals averages about 0.4

percent (4 per 1000 discharges), while an estimated 50 % of patients hospitalized with cancer,

cystic fibrosis (CF), and burns succumb to p. aeruginosa infection [5].

It has been established that growth and subsequent pathogenicity in P.aeruginosa depends

on the ability of this bacterium to sequester and utilize iron from the host cell [6]. The most

convincing findings were reported by Singh and coworker [7,8] who demonstrated that the iron

binding protein lactoferrin can block biofilm development (the common mode of growth of P.

aeruginosa in chronically infected patients [9]) at concentrations below those that kill or prevent

growth. It was proposed that by chelating iron, thereby lowering intracellular levels, lactoferrin

stimulates twitching motility, which in turn causes the bacteria to wander across the surface

instead of forming cell clusters and biofilms. This and other findings strongly suggest that

growth and hence pathogenicity in P. aeruginosa can be minimized or prevented by depriving the

pathogen of the most essential nutrient, iron. It is therefore important to understand the

mechanisms by which this bacterium sequesters iron from the host cell because this information

2  

is expected to give more insights into rational drug design to combat P. aeruginosa and other

related bacterial infections.

Iron uptake by bacteria

Iron is an essential nutrient that plays significant roles in biological processes. It is

generally required for metabolic processes in organisms that consume oxygen. However, in

aerobic conditions, Fe3+ is extremely insoluble and the concentration of free iron is very low (10-

18 M) at neutral pH [10]. The concentration of iron available to mammalian bacterial pathogens is

even significantly lower (10-24 M), because Fe3+ is sequestered by host iron carrier proteins such

as lactoferrin and transferrin, or by heme. However, since this element is essential for growth and

pathogenesis of bacteria, invading microorganisms have developed intricate iron uptake

pathways to overcome iron restriction and assimilate this nutrient from one or more of the

diverse mammalian iron sources.

One way by which invading microbes acquire iron involves production and secretion of

small molecular weight compounds called siderophores that strongly and specifically chelate

ferric iron, solubilize it from mineral or organic complexes, and deliver it to the microbial cells

via specific cell surface receptors. In Gram-negative bacteria, iron-siderophore complexes are

recognized and bound by specific outer membrane receptors at the cell surface and then

transported through the inner membrane by periplasmic protein-dependent ABC permeases. P.

aruginosa produces two siderophores; pyoverdine and pyochelin [11,12] under the regulation of

Fur (Ferric uptake regulator) [13]. Under iron-replete conditions, the biosynthesis of pyochelin

and the uptake system of the ferric-pyochelin complex are repressed by Fur. Pyoverdine

synthesis is indirectly regulated in response to iron through the Fur-controlled sigma factor PvdS

3  

[14]. When Fur is complexed with ferrous iron (under iron-replete conditions), it binds to a

conserved sequence element (Fur-box) in the promoter regions of iron regulated genes and

blocks their transcription [15]. In contrast, when iron is limiting, Fur dissociates from the Fur-

box allowing a basal level of gene expression to occur. Expression of siderophore receptors are

also regulated by Fur. A combination of negative regulation by Fur and transcriptional activation

by the AraC regulator PchR [16] determines the level of pyochelin receptor FptA in the cell. The

regulation of expression of iron uptake, transport and utilization proteins in general is complex

and involves a number of feedback mechanisms for sensing the iron status of the cell and

availability of these proteins in the cell. This tight control is responsible for the virulence factors

of the bacteria.

In addition to siderophore secretion, some bacteria directly bind host iron-binding

proteins [17,18], reduce the insoluble form of iron (Fe3+) to the usable soluble form (Fe2+) [19],

secrete extracellular proteinases that degrade host iron-binding proteins [20,21], or produce

lethal compounds (exotoxins) that may eliminate competitors for usable iron resources [22,23].

Bacteria can also acquire iron from heme. However, like iron, free heme is toxic and

causes oxidative damage to cells by catalyzing formation of reactive oxygen intermediates

generated by Fenton-type reactions. To avoid this deleterious effect free heme appearing in

plasma is rapidly sequestered by the host heme binding proteins hemopexin (Hpx), and human

serum albumin (HSA). Under pathologic conditions HPX, HSA, hemoglobin (Hb) and

hemoglobin binding haptoglobin (Hp) are potential heme-iron sources to invading bacteria. Since

these proteins bind heme with high affinity, under iron-limiting conditions, bacteria employ

sophisticated mechanisms to sequester the heme from these sources and transport it into the

cytoplasm where iron is extracted by heme oxygenanses that break the tetrapyrrole ring [24], or

4  

iron is released and the tetrapyrole ring is left intact (de-ferration) [25]. Different heme uptake

mechanisms exist in gram positive and gram negative bacteria. Examples of heme uptake

systems in gram negative bacteria will be discussed.

Heme uptake systems in Gram negative bacteria

Heme-uptake systems in Gram-negative bacteria have been extensively studied and can

be classified into three groups. The first group involves direct binding of heme or host heme-

containing proteins (HSA, Hpx, Hb, or Hp-Hb complex) to specific outer membrane receptors

located on the bacterial surface whereupon the heme is translocated across the membrane to the

cytoplasm. Examples of systems in this category are the the hmuRSTUV system in Yersinia pestis

[26], the hemR-hemSTUV system in Yersinia enterocolitica [27], the shuASTUV system of

Shigella dysenteriae [28,29] and the phuRSTUVW system in Pseudomonas aeruginosa

[30,31,32]. The phu (Pseudomonas heme uptake) system in P. aeruginosa involves a single

outer membrane heme receptor gene (PhuR), a periplasmic heme transport protein (PhuT), inner

membrane proteins (PhuUVW) typical for an ABC transport system, and a cytoplasmic protein

(PhuS) [33] (Figure 1, adapted from ref. 34).

Figure 1. Map of the P. aeruginosa heme uptake locus containing the phuR gene and the

phuSTUVW operon. Three Fur binding elements are shown as white boxes adapted from

5  

The second heme-uptake system closely related to the preceding category is the bipartite

Hb / Hp-hb receptor HpuAB in Neisseria meningitidis [35,36]. HpuAB consists of two outer

membrane proteins; a TonB dependent outer membrane receptor HpuB similar to other single

component TonB-dependent receptors and an accessory outer membrane lipoprotein HpuA. Both

proteins are required for Hb, Hp-hb, and apo-haptoglobin binding to whole cells. HpuAB is

analogous to the bipartite transferrin receptors Tbp1 and Tbp2 [37]; Tbp1 is an integral TonB-

dependent outer membrane protein while Tbp2 is a lipoprotein. Whereas both receptors are

necessary for efficient transferrin binding, the latter discriminates between the iron-bound and

iron-free transferrin and binds the former with 100 times higher affinity.

Secretion of small heme-binding proteins called hemophores to the extracellular medium

constitutes the third category of heme-uptake systems. Secreted hemophores bind hemin (free or

hemoprotein-bound) and deliver it to a TonB dependent specific outer membrane receptor on the

cell surface. This category include the hemophore systems hasRADEBF in S. marcescens and P.

aeruginosa [38], and the hxuCBA system found only in H. influenza [39,40]. Figure 2 shows

genetic organization of the has (heme acquisition system) systems [41]. The has operon in

Serratia, which is similar to that of Pseudomonas, is negatively regulated by the Fur repressor in

the presence of iron and the extracytoplasmic function (ECF) sigmas. The latter is dependent on

the receptor HasR and the two genes located upstream of HasR; HasI (extrecytoplasmic sigma

factor) and its modulator HasS. Bivelle and coworkers [42] have shown that HasI and HasS

genes cooperate with Fur-dependent mechanism to tune the has system under iron limiting

condition. Binding of HasA (the hemophore) to HasR (the receptor) activates the sigma gene

HasI but inactivates HasS and in effect the transcription of has signal cascade is enabled.

6  

By contrast, at optimal iron concentrations, HasI induces transcription of HasS turning off its

anti-sigma. As a consequence, inactivated HasS accumulates and only turns on when the heme

concentration is low or/ and when hemin loaded HasA is not bound to the receptor. This

feedback mechanism is analogous to that involving iron uptake and transport proteins where the

bacterium senses the concentration of iron in the cell and only turns on the genes necessary for

transcription of relevant enzymes [38,43,44].

Hemophore-dependent heme acquisition systems

In H. influenza, the hemophore HxuA is secreted by a signal-peptide dependent pathway

involving one outer membrane helper protein [45]. Once in the extracellular medium, HxuA

acquires heme from HPX and shuttles it to a cell surface receptor which is thought to be HxuC.

The only other known hemophore in gram negative bacteria, HasA, is found in P. aeruginosa, P.

Figure 2. Genetic organization of the has systems in Pseudomonas aeruginosa (P.a.) and

Serratia marcescens (S.m.). HasI and hasS encode sigma and antisigma factors, respectively.

The white boxes indicate consensus Fur boxes (adapted from ref 41).

7  

fluorescens, S. marcescens, Y. pestis, and Y. enterocolitica [46,47,48,49]. HasA belongs to a

novel family of proteins without homology to any other known proteins. They do not have a

signal peptide and are secreted by ABC transporters directly into the extracellular medium where

they bind free heme or extract heme from hemoproteins and deliver it to the specific outer

membrane receptor HasR. Whereas the receptor is important in heme acquisition, the hemophore

is not but the synergy increases the efficiency of free heme or hemoglobin acquisition 100-fold

[50]. This observation portrays the importance of the hemophore in enhancing the capacity of the

system and broadening the spectrum of substrate sources when the bacterium is faced with

limited amounts of iron in the cell. This study focuses on the structural and biochemical

characterization of the mechanism of heme capture by the hemophore HasAp from P.

aeruginosa. Therefore a general overview of the working mechanism of the has system,

structural information of characterized HasA, and heme binding studies are discussed in this

introduction.

Proposed working mechanism of hasA

A simplified schematic of the proposed working mechanism of HasA is shown in Figure

3. Under iron limiting conditions, the hemophore HasA is secreted to the extracellular medium

where it sequesters heme from host heme binding proteins and shuttles it to the specific outer

membrane receptor HasR [51]. Once the heme is delivered to the receptor, apo HasA is released

[52]. The heme is then translocated across the membrane into the cytosol where it is thought to

be uptaken by a cytoplasmic protein (PhuS). PhuS is thought to deliver the heme to heme

oxygenase [30] which is thought to degrade the macrocycle to release iron [53]. The protein

8  

complex involved in energy transduction from the cytoplasmic membrane to the outer

membrane, and the complex involved in translocation of the heme from the outer membrane to

the cytosol are discussed in detail in the following sub-sections. A schematic illustration is shown

in Figure 4.

Periplasm Inner membrane

Ext

race

llula

r m

ed

ium

Outer membrane

Hemoproteins

HO

HasR

HasA

ABC

Fe, CO, Biliverdin

Cyt

oso

l

- Hemin

PhuS

Hemophore secretion

The proteins making the secretion apparatus in hasA system are an ABC protein HasD,

and two accessory proteins; HasE (a member of the membrane fusion protein family), and HasF

(a TolC-like outer membrane protein) (Figure 4) [54]. HasA hemophores are secreted by ATP-

Binding Cassette (ABC) secretion pathway into the extracellular medium in a secB-dependent

manner [55]. Like most proteins secreted by this pathway, hemophores have a -helical C-

Figure 3. Simplified schematic of the proposed working mechanism of the has system

9  

terminal secretion signal characterized by the presence of a negatively charged amino acid

residue followed by several hydrophobic residues [56]. This secretion signal interacts with the

ABC transport protein, regulates its ATPase activity and induces a multiprotein complex

formation comprising the secreted protein (HasA) and the three secretion proteins (HasD, HasE

and HasF) [52]. The location of the ABC secretion signal (C` terminus) indicates that the protein

is fully synthesized before the signal is recognized. The absence of the secretion apparatus, secB,

presents two main problems to the bacterium; first, HasA accumulates in the cytoplasm, folds,

and binds heme from other cytoplasmic proteins [38]; second, folding renders the protein

incompetent for secretion yet it can still interact with the transporter; as a result secretion of

newly synthesized proteins is inhibited [57]. SecB chaperon is therefore important in keeping the

newly synthesized proteins in a conformation that is competent for secretion [58].

Figure 4. Schematic illustration of the protein complex involved in energy transduction from the

cytoplasmic membrane to the outer membrane in Gram-negative bacteria. OM: outer membrane;

PP: periplasmic space; CM: cytoplasmic membrane. Adapted from ref. 54.

10  

Heme binding to HasA and interaction of HasA with the receptor HasR

HasA has a high affinity for heme (Kd = 10-11 M) [59]. When it is secreted to the

extracellular medium, it may extract heme from a large range of substrates including

hemoglobin, hemoglobin-haptoglobin complex, myoglobin or serum albumin. Although heme

sequestration from these hemeproteins has been documented, previous to our study, the

mechanism by which heme binds to HasA was not known. The mechanism of hemin delivery to

the receptor is also not clearly understood. Of interest is the fact that the hemophore has a

markedly higher affinity for heme than the cognate receptor (Kd ~ 10−6 M) [52] and that both

hemin-free and hemin-loaded HasA bind the receptor HasR on two overlapping sites (-strands,

S51-T60 and G95-S107) with the same apparent affinity (Kd 5 nm) [60]. In addition, it has been

suggested that HasA interaction with HasR induce some conformational changes which may

move heme away from the distal ligand and as a result the central iron is reduced or coordinates

one of HasR residue facilitating transfer.

Translocation of hemin across the membrane

Hemin bound to the receptor is translocated from the outer membrane to the periplasm in

a TonB-dependent process and involve formation of a HasB-ExbB-ExbD protein complex. Once

in the periplasm, the heme is bound by a heme-specific periplasmic transport protein (HTP) (e.g

PhuT in P. aeruginosa), which mediates heme transfer to the cytoplasm through ATP-binding

cassette (ABC) transporter. Association of the ABC transporter and HTP forms a complete heme-

uptake permease [61,62].

11  

Hemin degradation and iron releasing in the cytosol

Free hemin is toxic to cells and so once it is delivered to the cytoplasm, it is believed to

be sequestered by a PhuS analogue which passes it on to heme oxygenase (HO) via protein-

protein interactions [63]. HO then catalyzes the oxidation degradation of heme to release iron

which can then be utilized by the bacterium. The process of hemin catabolism is summarized in

Scheme 1 below [64,65]. Three oxygen molecules and seven electrons (supplied by NADPH in

the case of P. aeruginosa) are involved. The catabolic process begins with reduction of the

central metal iron (Fe3+) to form a metastable oxyferrous complex (Fe2+-O2). This complex then

accepts a proton (from the distal pocket water) and a second electron to form an intermediate

species (Fe3+-OOH). The terminal oxygen of Fe3+-OOH then attacks the α-meso-carbon of the

porphyrin ring to form ferric α-meso-hydroxyheme which oxidizes to ferric verdoheme in the

presence of oxygen releasing CO. A third oxygen molecule inserts into the macrocyle thereby

breaking the tetrapyrole ring to release iron.

 

N

NN

NF eII

H3C H3C

CH3

CH3HO2C

HO2C

OO

O2

e-

H+e-

O2

CO

O2

Fe2+ H

N

NN

NF eII I

H3C H3C

C H3

C H3HO2C

O2C

O H

-meso-hydroxyheme

N

O+NN

NF eII

H3C H3C

C H3

C H3HO2C

HO2C

verdoheme

N

NN H

HN

H3C H3C

C H3

C H3HO2C

HO2C

OO

β

δ

γ α

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

heme

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

OOH

β

δ

γ α

β

δ

γα

β

δ

γα

β

δ

γ

α-biliverdin

α

N

NN

NF eII

H3C H3C

CH3

CH3HO2C

HO2C

OO

N

NN

NF eII

H3C H3C

CH3

CH3HO2C

HO2C

OO

O2

e-

H+e-

O2

CO

O2

Fe2+ H

N

NN

NF eII I

H3C H3C

C H3

C H3HO2C

O2C

O H

H

N

NN

NF eII I

H3C H3C

C H3

C H3HO2C

O2C

O H

-meso-hydroxyheme

N

O+NN

NF eII

H3C H3C

C H3

C H3HO2C

HO2C

N

O+NN

NF eII

H3C H3C

C H3

C H3HO2C

HO2C

verdoheme

N

NN H

HN

H3C H3C

C H3

C H3HO2C

HO2C

OON

NN H

HN

H3C H3C

C H3

C H3HO2C

HO2C

OO

β

δ

γ α

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

heme

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

heme

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

OOH

N

NN

NFeII I

H3C H3C

CH3

CH3HO2C

HO2C

OOH

β

δ

γ α

β

δ

γα

β

δ

γα

β

δ

γ

α-biliverdin

α

Scheme 1. Oxidative heme degradation catalyzed by heme oxygenase

12  

Structure of the model hemophore HasAs

HasAs from S. marcescens was the first hemophore to be identified, and it is the most

characterized of all the hemophores studied so far. It is a 19 kDa monomeric protein that binds b-

type heme in a 1:1 stoichiometry and with a high affinity (Kd ~ 10−11 M) [67]. The crystal

structure determined at 1.9 Å resolution [58] (Figure 5-A) shows an original α/β fold in which

seven β-strands form an antiparallel β-sheet wall on one face of the protein, while four α-helices

form an -helix wall on the opposite face. Hemin is coordinated by a unique pair of ligands His

32 and Tyr 75 located in two flexible loops (His 32 and Tyr 75 loop, respectively) at the interface

between the α-helix and β-sheet walls. Adjacent to Tyr 75 is another His residue (His 83), which

is within hydrogen bonding distance (2.8 Å) with Oη of Tyr 75 (Figure 6-B). This hydrogen

bonding interaction is thought to modulate the Tyr-Fe bond. Sequence alignment of identified

hemophores reveal that the proximal pair, Tyr 75 and His 83 is conserved but the His 32 is not

(Figure 7) [66]. The crystal structure also reveals that a large portion of the prosthetic group is

exposed to the solvent (186 Å) (Figure 6-A). Nonetheless, the macrocycle is tightly anchored by

coordination of the axial ligands, and favorable stacking and hydrophobic interactions with

residues within the binding site.

The high affinity for hemin, and the flexible secondary structure elements constituting

hemin binding site likely makes the hemophore efficient for hemin sequestration from host heme

proteins, and for release of hemin to the receptor HasR. Similar structural elements are observed

in the crystal structure of HasAp (Figure 5-B) except that HasAp has an extra -sheet (8) at the

C` terminal.

13  

Figure 5. Comparison of the crystal structure of Holo-HasASM (PDB ID. 1DKO) (A), and the

crystal structure of HasAp (PDB ID. 3ELL) (B) showing the elements of secondary structure.

The -helices are shown in cyan, -sheet in magenta, loops in green, heme in red and ligands in

blue. His83 is colored orange.

A

B

14  

Solution NMR structure of apo-HasAs has been solved. Overall, the protein has a similar

polypeptide fold with the holo counterpart except for the loop harboring the distal ligand His 32

which is relocated nearly 30 Å away (“open” conformation) relative to its position in the holo

protein (“closed” conformation) [67]. In this work, we solved the crystal structure of apo-HasAp

and investigated the overall structural changes accompanying hemin loading. Capitalizing from

the availability of the apo and holo HasAp crystal structures, we determined a plausible path

taken by the His 32 loop as it transitions from the “open” to the “closed” conformation using

molecular dynamics simulations. The results are discussed in detail in chapter II.

A B

Tyr 75

His 32

His 83

Figure 6. Zoomed in view of the heme binding site in the crystal structure of HasAs.( A)

solvent exposed prosthetic group; heme is shown in red while the water molecules are

represented by blue spheres. (B) Hemin coordination ligands His 32 and Tyr 75 (blue) and

the adjacent His 83 (orange).

15  

10 20 30 40 50 60 | | | | | | Pa MSISISYSTTYSGWTVADYLADWSAYFGDVNHRPGQVV-DGSNTGGFNPGP--FDGSQYALKS-TASDAAFIA Sm MAFSVNYDSSFGGYSIHDYLGQWASTFGDVNHTNGN-V-TDANSGGFYGGS--LSGSQYAISSTANQVTAFVA Pe MTISVNYDAAFASSTIDDYLAFWSAGFVTAGHG-------YSNTGGFSNGT--YDGDQYATHGRNSSDYAFIA Pf MSISISYSATYGGNTVAQYLTDWSAYFGDVNHRPGEVV-DGTNTGGFNPGP--FDGTQYAIKS-TASDAAFVA Yps MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHGNPEGMIA Pc MSFAITYDAYYANYSIASYLTEWSAAFGDVNHTAGNTQVGGNNTGGFYGGDTFIDGTQYAITSTQNDFSALIA Ec MSFAITYDAYYANYSIANYLTEWSAAFGDVNHTTGNTQVGGNNTGGFYGGDTFIDGSQYAITSTQNDFSALIA Yp MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHSNPEGMIA Ypa MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHSNPEGMIA 70 80 90 100 110 120 | | | | | | Pa GGD----LHYTLFS------NPSHTLWGKLDSIALGDTLT---GGASSGG-YALDSQEVSFSNLGLDSPIAQ- Sm GGN----LTYTLFN------EPAHTLYGQLDSLSFGDGLS---GGDTSP--YSIQVPDVSFGGLNLSSLQAQ- Pe DSNNANGLHYVFDPSKAPGDNLNHYLWGSLDNVSLGEVLG---GGSGSD--FTLGNYVVNFNGLDLDAAHGA- Pf DGN----LHYTLFS------NPSHTLWGSVDTISLGDTLA---GGSGSN--YNLVSQEVSFTNLGLNSLKEE- Yps EGN----LKYSFM--------PQHTFYGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG Pc GGD----LTYSLFT------PPAHTLYGDLDSLSFGNVLQ---GGTTAGTTYSLAEPEVTFSGLDLSTDVAN- Ec GGD----LTYSLFS------PPAHTLYGDLDSLSFGNVLQ---GGTTAGTTYSLVEPEVTFSGLDLSTDVAN- Yp EGD----LKYSFM--------PQHTFHGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG Ypa EGD----LKYSFM--------PQHTFHGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG 130 140 150 160 170 180 190 | | | | | | | Pa ---GR-DG-TVHKVVYGLMSGDSSALQGQIDALLKAVDPSLSINSTFDQLAAAGVAHATP-AAAAAEVGVVGV Sm ---GH-DG-VVHQVVYGLMSGDTGALETALNGILDDYG--LSVNSTFDQVAAA----T-----------AVGV Pe ---GR-AGNEVQTVIYGLMQGNTAGLEGVLDTLLASYG--VSTNDTFDVVAAA----LGG-ASGAASAQAVGV Pf ---GR-AG-EVHKVVYGLMSGDSSALAGEIDALLKAIDPSLSVNSTFDDLAAAGVAHVNPAAAAAADVGLVGV Yps KSMTENHQGDMHKAILGLRKGNA----DPMLEVMKAKG--FDVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV Pc --LTVSDRGVVHDVIYGLMSGQV----QPLLDALTNAG--IDINASLDSLSFATATSDAALSADT-VVDVVGV Ec --LTVSDRGVVHDVIYGLMGGQV----QPLLDALTNAG--IDINASLDSLSFATATSDAALSADT-VVDVVGV Yp KSMTENHQGDMHKSVRGLMKGNP----DPMLEVMKAKG--INVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV Ypa KSMTENHQGDMHKSVRGLMKGNP----DPMLEVMKAKG--INVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV 200 205 | | Pa QELPHDLALAA Sm QHADSPELLAA Pe QALPEDLALAA Pf QDVAQDWALAA Yps MDSN-DMLLAA Pc AETA-DLLAA- Ec AETA-DLLAA- Yp VDCH-DMLLAA Ypa VDCH-DMLLAA

Figure 7. Amino acid sequence alignment of full-length HasA sequences from P. aeruginosa

(Pa), Serratia marcescens (Sm), P. entomophila (Pe), P. fluorescens (Pf), Yersinia

pseudotuberculosis (Yps), Pectobacterium carotovorum (Pc), Erwinia carotovora (Ec), Yersinia

pestis (Yp), and Yersinia pestis bv. Antiqua (Ypa) aligned against the sequence of HasA from P.

aeruginosa. Truncated HasAp used in these investigations stops at residue 184 (full-length minus

21 amino acids) [13]. Residues highlighted in red are conserved while those in blue are the His32

loop “hinge” residues in HasAp. The alignment was performed using ClustalW [68].

16  

Mechanism of heme uptake in HasA

Prior to our study, the mechanism of hemin loading onto HasA was unknown. However,

Wolff and coworkers [67] proposed a two-step mechanism where hemin is initially loaded onto

the loop harboring the proximal hemin-iron ligand Tyr 75 followed by closure of the His 32 loop.

Two recent independent studies strengthened this proposal. First, Lukat-Rodgers et al [69]

observed that when wild type holo-HasAs is reduced with dithionite in the presence of CO, a

ferrous hexacoordinate complex in which His 32 is displaced by CO at the distal site is formed.

Second, the crystal structure of HasAs in complex with its receptor (HasR), recently reported by

Krieg et al [70], shows that hemin in the hemophore is coordinated only by Tyr 75 at the

proximal site while the His 32 loop is turned away from its heme-binding conformation.

Together, these studies suggest that His 32 is more labile and that the loop harboring it is more

flexible relative to the Tyr 75 loop. In this study extensive investigations were carried out to

elucidate the mechanism of hemin loading onto HasAp using wild type apo HasAp and H32A-

HasAp mutant. The results of this investigation are presented in chapter II.

Heme binding studies

Hemin binding studies in HasA have been extensively carried out with the hemophore

HasAs using free heme in solution as the heme source [71]. Biochemical characterization of

heme loaded proteins revealed that one ligand is efficient and sufficient to coordinate heme in the

heme-HasAs complex. It was thus proposed that His 83 (adjacent to the proximal heme-iron

ligand Tyr 75 in wild type HasA) can act as an alternate ligand in the absence of Tyr 75.

However, there is no reliable structural information to support this proposal. In this study,

17  

corresponding HasAp mutants in which the proximal residues were replaced by a

noncoordinative residue Ala (Y75A and H83A) were prepared and characterized by X-ray

crystallography. The overall architecture of the mutants and the coordination state or hemin-iron

are discussed in Chapter III.

Studies in hemophore acquisition of heme from hemoproteins have focused primarily on

hemoglobin as the heme source. Kinetic analysis of heme transfer from hemoglobin to HasAs or

HasAp revealed that the rate of heme transfer is relatively slow [72]. It was concluded that heme

transfer is by passive diffusion and affinity driven. We hypothesize that methemalbumin (heme

loaded serum albumin) is the potential target for the secreted hemophore HasAp from P.

aeruginosa. This argument is based on two lines of thought; i, the association between serum

albumin and heme is relatively weak (Kd = 10 -8 M) [73], and ii, under hemolytic conditions as

is the case during Pseudomonas infection, the heme carrier protein HPX is saturated leaving

enormous amounts of heme bound to serum albumin . We carried out heme transfer studies using

apo HasAp and methemalbumin to test this possibility. The results are described in detail in

chapter IV.

RESEARCH PROBLEM AND RATIONALE

Iron is an essential element for virtually all living organisms. Its ligand binding

capability, reactivity and redox properties make it attractive for diverse biological processes.

However, the same properties that make iron attractive in biological applications pose enormous

problems. First, the easy one electron inter-conversion of Fe2+and Fe 3+ makes iron very toxic by

catalyzing the generation of damaging oxygen-derived radicals through Fenton-type reactions.

18  

Second, under aerobic conditions Fe3+ is thermodynamically favored, leading to problems of

solubilization. For these reasons, living organisms have evolved molecular systems to limit iron

availability; by sensing adequate iron levels, limiting uptake, and storing excess iron in storage

proteins. Iron, in the form of heme is similarly very toxic to cells, thus once heme appears in

plasma it is rapidly sequestered by heme binding proteins (hemopexin and serum albumin) and

subsequently cleared from circulation via receptor-mediated endocytosis.

Invading pathogenic bacteria including P. aeruginosa require iron for growth and for the

successful colonization of the host cells. However, as alluded to earlier, the human host presents

a hostile, iron restrictive environment to bacterial invaders. To overcome this iron limitation, P.

aeruginosa has developed sophisticated iron and heme-iron acquisition mechanisms to

circumvent host sequestration of iron sources. It follows therefore that molecules or/and

compounds that are capable of inhibiting or interrupting iron acquisition paths may constitute a

new class of antibiotics to fight Pseudomonas infections. Prospects for rational drug design of

these antibiotics, structure-based or otherwise, require molecular level understanding of the

intricate pathways employed by this bacterium to sequester iron from the host cell.

Under iron limiting conditions, P.aeruginosa secretes the hemophore HasAp to the

extracellular medium where it sequesters heme and shuttles it to the cognate outer membrane

receptor HasR. However, the mechanism of heme binding to the hemophore HasAp and the

target host heme binding protein are not known. This investigation focused on elucidating the

mechanism of heme loading as well as determining the potential host heme iron source targeted

by HasAp.

19  

To achieve this objective, a chromatographic method was developed to isolate apo-

HasAp from the holo fraction and to purify it to homogeneity. Pure apo-HasAp was then

characterized by X-ray crystallography and solution NMR spectroscopy. To be able to elucidate

the mechanism of heme binding to the hemophore, HasAp variant in which the distal ligand His

32 was replaced with the non-coordinative residue Ala (H32A) was purified in the apo form.

Kinetic studies of heme binding were carried using a stopped flow spectrometer. A parallel

experiment was done using the wild type protein for comparison. Heme loaded H32A-HasAp

was further characterized by X-ray crystallography, and by EPR and Resonance Raman

Spectroscopy to elucidate the coordination state and spin state of heme-iron. H32A-HasAp

mutant in complex with imidazole (H32A-Imidazole) was characterized by X-ray

crystallography and solution NMR to establish the topological position of His 32 loop in the

heme loaded protein. The results and discussion are presented in chapter II. Capitalizing on the

availability of the structures of holo and apo HasAp, molecular dynamic simulations were carried

out to obtain information pertaining to His 32 loop closing as it transitions from the “open” to the

“closed” state. The results of these studies are also presented in detail in chapter II.

The crystal structures of HasAp mutants involving the proximal heme residues Tyr 75

and His 83 (Y75A and H83A) are presented in chapter III. The overall protein fold, hemin

coordination states and spin states are compared to wild type HasAp.

Chapter IV describes preliminary studies aimed at understanding heme acquisition from

the host heme-binding protein serum albumin. Experiments were carried out using UV-vis

spectroscopy, gel filtration, and polyacrylamide gel electrophoresis.

20  

References

1. Woods, D. E., and Vasil, M. L., Pathogenesis of Pseudomonas aeruginosa infections. In

Pseudomonas aeruginosa Infections and Treatment, 1994. Marcel Dekerr Inc. New York. 2. Ferguson, D., Cahill, O. J., and Quilty, B., Phenotypic, molecular and antibiotic

resistance profiling of nosocomial Pseudomonas aeruginosa strains isolated from two Irish Hospitals. J Med, 2007. 1: np. http://www.scientificjournals.org/journals2007/articles/1055

3. Goldsworthy, M. J. H., Gene expression of Pseudomonas aeruginosa and MRSA within a

catheter associated urinary track biofilm model. Biosci Horz, 2008. 1: p. 28-37. 4. Goldberg, J. B., Hancock, R. E., Parales, R. E., Loper, J. and Cornalis, P., Pseudomonas

2007. J Bacteriol. , 2008. 190: p. 2649-2662. 5. http://www.polyphor.com/infections-caused-by-pseudomonas-aeruginosa.html 6. Weinberg, E. D., Iron and infection. Microbiol. Rev., 1978. 42: p. 45-66. 7. Singh, P. K., Iron sequestration by human lactoferrin stimulates P. aeruginosa surface

motility and blocks biofilm formation. Biometals, 2004. 17: p. 267-70. 8. Singh. P. K., Parsek, M. R., Greenberg, E. P., and Welsh, M. J., A component of innate

immunity prevents bacterial biofilm development. Nature, 2002. 417: p. 552-555. 9. Bjarnsholt, T., and Givsko, M., Costerton, J, W., The role of quorum sensing in the

pathogenisity of the cunning aggressor Pseudomonas aeruginosa. Anal Bioanal Chem, 2007. 387: p. 409-414.

10. Braun, V., and Killman, H., Bacterial solutions to the iron-supply problem. Trends

Biochem Sci, 1999. 24: p. 104-109. 11. Bodilis, J., and Cornelis, P., A survey of TonB-dependent receptors in fluorescent

pseudomonads. Environ Microbiol Rep, 2009. 1: p. 256-262. 12. Cornelis, P., Iron uptake and metabolism in pseudomonads. Appl Microbiol Biotechnol,

2010. 86: p. 1637-1645. 13. Ochsner, U. A., Vasil, A. I., and Vasil, M. L., Role of the Ferric Uptake Regulator of

Pseudomonas aeruginosa in th Regulation of Siderophores and Exotoxin A Expression: Purification and Activity on Iron-Regulated Promoters. J. Bactriol., 1995. 177: p. 7194-7201.

21  

14. Vasil., M. L., and Ochsner, U. A., The response of Pseudomonas aeruginosa to iron:

genetics, biochemistry and virulence. Molecular Microbiology, 1999. 34: p. 399-413. 15. Escolar, L., Perez-Martin, J., and de Lorenzo, V., Opening the iron box: transcriptional

metalloregulation by the Fur protein. J Bactriol, 1999. 181: p. 6223-6229. 16. Michel, L., Gonzalez, N., Jagdeep, S., Nguyen-Ngoc, T. , and Reimmann, C., PchR-box

recognition by the AraC-type regulator PchR of pseudomonas aeruginosa reguires the siderophore pyochelin as an effector . Mol Microbiol, 2005. 58: p. 495-509.

17. Cornelissen, C., and Sparling, P., Iron piracy: acquisition of transferrin-bound iron by

bacterial pathogens. Mol Microbiol, 1994. 14: p. 843-850. 18. Mietzner, T. A., and Morse, S. A., The role of iron-binding proteins in the survival of

pathogenic bacteria. Annu. Rev. Nutr., 1994. 14: p. 471-493. 19. Coulanges, V., Andre, P., Ziegler, O., Bucheit, L., and Vidon, D., Utilization of iron-

catecholamine complexes involving ferric reductase activity in Listeria monocytogenes. Infect Immun, 1997. 65: p. 2778-2785.

20. Wolz, C., Hohloch, K., Ocaktan, A., Poole, K., Rochel, N., Albrecht-Gary, A., et al. Iron

release from transferrin by pyoverding and elastase from Pseudomonas aeruginosa. Infect Immun, 1994a. 62: p. 4021-4027.

21. Wilderman, P. J., et al, Charactrization of an endoprotease (PrpL) encoded by PvdS-

regulated gene in Pseudomonas aeruginosa. Infect Immun, 2001. 69: p. 5385-5394. 22. Bjorn, M., Iglewski, B.H., Sadoff, J., and Vasil, M.L., Effect of iron on yields of exotoxin

A in cultures of Pseudomonas aeruginosa PA103. Infect Immun, 1978. 19: p. 785-791. 23. Calderwood, S.B., and Mekalanos, J.J., Iron regulation of Shiga-like toxin expression in

Escherichia coli is mediated by the fur locus. J Bacteriol, 1987. 169: p. 4759-4764. 24. Yoshida, T., Noguchi, M., Kikuchi, G., and Sano, S., Degradation of mesoheme and

hydroxymesoheme catalyzed by the heme oxygenase system: involvement of hydroxyheme in the sequence of heme catabolism. J Biochem., 1981. 90: p.125 -131.

25. Cornelis, P., Iron uptake and metabolism in pseudomonads. Appl Microbio Biotechnol,

2020. 86: p 1637-1645. 26. Hornung, J. M., Jones, H. A., Perry, R. D., The hmu locus of Yersinia pestis is essential

for utilization of free haemin and haem--protein complexes as iron sources. Mol Microbiol., 1996. 20: p. 725-39.

22  

27. Stojiljkovic I, and Hantke K., Hemin uptake system of Yersinia enterocolitica:

similarities with other TonB-dependent systems in gram-negative bacteria. EMBO J., 1992. 11: p. 4359 - 4367.

28. Eakanunkul, S., Lukat-Rodgers, G. S., Sumithran, S., Ghosh, A., Rodgers, K. R., Dawson

J. H., and Wilks, A., Characterization of the periplasmic heme-binding protein shut from the heme uptake system of Shigella dysenteriae. Biochemistry, 2005. 44: p.13179-13191.

29. Burkhard, K. A., Wilks, A., Characterization of the outer membrane receptor ShuA from

the heme uptake system of Shigella dysenteriae. Substrate specificity and identification of the heme protein ligands. J Biol Chem., 2007. 282: p. 15126-36.

30. Lansky, I. B., Lukat-Rodgers, G. S., Block, D., Rodgers, K. R., Ratliff, M., and Wilks,

A., The cytoplasmic heme-binding protein (PhuS) from the heme uptake system of Pseudomonas aeruginosa is an intracellular heme-trafficking protein to the delta-regioselective heme oxygenase. J Biol Chem., 2006. 281: p. 13652-13662.

31. Ho, W. W., Li, H., Eakanunkul, S., Tong, Y., Wilks, A., Guo, M.,and Poulos, T.L., Holo-

and apo-bound structures of bacterial periplasmic heme-binding proteins. J Biol Chem., 2007. 282: p. 35796-802.

32. Ochsner, U. A., Johnson, Z., and Vasil, M.L., Genetics and regulation of two distinct

haem-uptake systems, phu and has, in Pseudomonas aeruginosa. Microbiology, 2000. 146 : p. 185-198.

33. Tong, Y., Guo, M., Bacterial heme-transport proteins and their heme-coordination

modes. J Biol Inorg Chem, 2007.12: p. 735-750. 34. Tong, Y., and Guo, M., Cloning and characterization of a novel periplasmic heme-

transport protein from the human pathogen Pseudomonas aeruginosa. J Biol Inorg Chem., 2007. 12: p. 735-750.

35. Lewis, L. A., Gray, E., Wang, Y. P., Ro, B. A., and Dyer, D. W., Molecular

characterization of hpuAB, the haemoglobin-haptoglobin-utilization operon of Neisseria meningitidis. Mol Microbiol., 1997. 23: p. 737-749.

36. Lewis, L. A., and Dyer, D. W., Identification of an iron-regulated outer membrane

protein of Neisseria meningitidis involved in the utilization of hemoglobin complexed to haptoglobin. J Bacteriol., 1995. 177: p.1299-1306.

37. Cornelissen, C., and Sparling, P., Iron piracy: acquisition of transferrin-bound iron by

bacterial pathogens. Mol Microbiol, 1994. 14: p. 843-850. 38. Cescau, S., Cwerman, H., Létoffé, S., Delepelaire, P., Wandersman, C., and Biville, F.,

Heme acquisition by hemophores . Biometals, 2007. 20: p. 603-613.

23  

39. Hanson, M. S., Pelzel, S. E., Latimer, J., Muller-Eberhard, U., and Hansen, E. J.,

Identification of a genetic locus of Haemophilus influenzae type b necessary for the binding and utilization of heme bound to human hemopexin . Proc Natl Acad Sci USA, 1992. 89: p. 1973-1977.

40. Cope, L. D., Thomas, S. E., Hrkal, Z., and Hansen, E. J., Binding of heme-hemopexin

complexes by soluble HxuA protein allows utilization of this complexed heme by Haemophilus influenzae. Infect Immun., 1998. 66: p. 4511-4516.

41. Poole, K., and McKay, G. A., Iron Acquisition and Its Control in Pseudomonas

aeruginosa: Many Roads Lead To Rome. Front. Biosci., 2003. 8: p. 661-686. 42. Bivelle, F., Cwerman, H., Létoffé, S., Rossi, M.-S., Drouet, V., Ghigo, J. M., and

Wandersman, C., Haemophore-Mediated Signalling in Serratia marcescens: A New Mode of Regulation for an Extra Cytoplasmic Function (ECF) Sigma Factor Involved in Haem Acquisition. Mol. Microbiol., 2004. 53: p. 1267-1277.

43. Rossi, M.-S., Paquelin, A., Ghigo, J. -M., and Wandersman, C., Haemophore-Mediated

Signal Transduction Across the Bacterial Cell Envelope in Serratia marcescens: The Inducer and the Transported Substrate are Different Molecules. Mol. Microbiol., 2003. 48: p. 1467-1480.

44. Cwerman, H., Wandersman, C., and Biville, F., Heme and a Five-Amino-Acid

Hemophore Region Form the Bipartite Stimulus Triggering the has Signalling Cascade. J. Bacteriol., 2006. 188: p. 3357-3364.

45. Hanson, M.S., Pelzel, S. E., Latimer, J., Muller-Eberhard, U., and Hansen, E. J.,

Identification of a genetic locus of Haemophilus influenzae type b necessary for the binding and utilization of heme bound to human hemopexin. Proc. Natl. Acad. Sci. USA., 1992. 89: p. 1973-1977.

46. Létoffé S, Omori K, Wandersman C. Functional characterization of the HasAPF

hemophore and its truncated and chimeric variants: determination of a region involved in binding to the hemophore receptor J Bacteriol , 2000. 182: p. 4401-4405.

47. Létoffé S, Ghigo JM, Wandersman C., Secretion of the Serratia marcescens HasA

protein by an ABC transporter. J Bacteriol, 1994.176: p. 5372-5377. 48. Rossi, M. S., Fetherston, J. D., Letoffe, S., Carniel, E., Perry, R. D., and Ghigo, J. M.,

Identification and Characterization of the Hemophore-Dependent Heme Acquisition System of Yersinia pestis. Infect Immun, 2001. 69: p. 6707-6717.

24  

49. Stojiljkovic I, Hantke K., Transport of haemin across the cytoplasmic membrane

through a haemin-specific periplasmic binding-protein-dependent transport system in Yersinia enterocolitica. Mol Microbiol. 1994.13: p. 719-732.

50.  Ghigo, J., Le´Toffe´, S., and Wandersman, C., A New Type of Hemophore-Dependent

Heme Acquisition System of Serratia marcescens Reconstituted in Escherichia coli. J of Bacteriol, 1997. 179: p. 3572-3579.

 

51. Létoffé, S., Nato, F., Goldberg, M. E., and Wandersman, C., Interactions of HasA, a Bacterial Haemophore with Haemoglobin and with Its Outer Membrane Receptor HasR. Mol. Microbiol., 1999. 33: p. 546-555.

52.  Le´toffe´, S., Delepelaire, P., and Wandersman, C., Free and Hemophore-Bound Heme Acquisitions through the Outer Membrane Receptor HasR Have Different Requirements for the TonB-ExbB-ExbD Complex. J of Bacteriol, 2004. 186: p. 4067- 4074.

 

53. Ratliff, M., Zhu, W., Deshmukh, R., Wilks, A., and Stojiljkovic, I., Homologues of Neisserial Heme Oxygenase in Gram-Negative Bacteria: Degradation of Heme by the Product of the pigA Gene or Pseudomonas aeruginosa. J. Bacteriol., 2001. 183: p. 6394-6403.

 

54. Poole, K., and McKay, G. A., Iron Acquisition and Its Control in Pseudomonas aeruginosa: Many Roads Lead To Rome. Front. Biosci., 2003. 8: p. 661-686.

55. Sapriel, G., Wandersman, C., and Delepelaire, P., The N Terminus of the HasA Protein

and the SecB Chaperone Cooperate in the Efficient Targeting and Secretion of HasA via the ATP-binding Cassette Transporter. J of Biol Chem., 2002. 277: p. 6726-6732.

 

56. Ghigo, J. M, and Wandersman, C., A carboxyl-terminal four-amino acid motif is required for secretion of the metalloprotease PrtG through the Erwinia chrysanthemi protease secretion pathway. J Biol Chem, 1994. 269: p. 8979 - 8985.

 

57. Debarbieux, L., and Wandersman, C., Folded HasA inhibits its own secretion through its ABC exporter. EMBO J., 2001. 20: p. 4657-4663.

 

58.  Wolff, N., Sapriel, G., Bodenreider, C., Chaffotte, A., and Delepelaire, P., Antifolding Activity of the SecB Chaperone is Essential for Secretion of HasA, a Quickly Folding ABC Pathway Substrate. J. Biol. Chem., 2003. 278: p. 38247-38253.  

 

59. Izadi-Pruneyre. N., Huché, F., Lukat-Rodgers, G. S., Lecroisey, A., Gilli, R., Rodgers, K. R., Wandersman, C., and Delepelaire, P., The heme transfer from the soluble HasA hemophore to its membrane-bound receptor HasR is driven by protein-protein interaction from a high to a lower affinity binding site. J Biol Chem., 2006 281: p. 25541-25550.

25  

60. Létoffé, S., Debarbieux, L., Izadi, N., Delepelaire, P., and Wandersman, C., Ligand

Delivery by Haem Carrier Proteins: The Binding of Serratia marcescens Haemophore to Its Outer Membrane Receptor is Mediated by Two Distinct Peptide Regions. Mol. Microbiol., 2003. 50: p. 77-88.

61.  Rossi, M.-S., Paquelin, A., Ghigo, J. -M., and Wandersman, C., Haemophore-Mediated

Signal Transduction Across the Bacterial Cell Envelope in Serratia marcescens: The Inducer and the Transported Substrate are Different Molecules. Mol. Microbiol., 2003. 48: p. 1467-1480.

 

62. Ghigo, J.-M., Létoffé, S., and Wandersman, C., A New Type of Hemophore-Dependent Heme Acquisition System of Serratia marcescens Reconstituted in Escherichia coli. J. Bacteriol., 1997. 179: p. 3572-3579.

63. Block, D. R., Lukat-Rodgers, G. S., Rodgers, K. R., Wilks, A., Bhakta, M. N., Lansky, I.

B., Identification of Two Heme-binding Sites in the Cytoplasmic Heme-trafficking Protein PhuS from Pseudomonas aeruginosa and Their Relevance to Function Biochemistry, 2007. 46: p. 14391-14402.

64. Wilks, A., and Burkhard, K. A., Natural product reports, 2007. 24: p. 511-522. 65. Rivera, M. and Zeng, M., Heme oxygenase, steering dioxygen activation toward heme

hydroxylation. J Inorg Biochem. 2005. 99: p. 337-354.  

66. Jepkorir, G., Rodríguez, J.C., Rui, H., Im, W., Lovell, S., Battaile, K.P., Alontaga, A.Y., Yukl, E.T., Moënne-Loccoz P., and Rivera, M., Structural, NMR Spectroscopic, and Computational Investigation of Hemin Loading in the Hemophore HasAp from Pseudomonas aeruginosa. J., Am Chem Soc., 2010. 132: p. 9857-9872.

67. Wolff, N., Izadi-Pruneyre, N., Couprie, J., Habeck, M., Linge, J., Riepeng, W.,

Wandersman, C., Nilges, M., Delepierre, M., and Lecroisey, A., Comparative Analysis of Structural and Dynamic Properties of the Loaded and Unloaded Hemophore HasA: Functional Implications. J. Mol. Biol., 2008. 376: p. 517-525.

68. Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. R.,

McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R., Thompson, J. D., Gibson, T. J., and Higgins, D. G., Clustal W and Clustal X version 20. Bioinformatics, 2007. 23: p. 2947-2948.

69. Lukat-Rodgers, G. S., Rodgers, K. R., Caillet-Saguy, C., Izadi-Pruneyre, N., Lecroisey, A., Novel heme ligand displacement by CO in the soluble hemophore HasA and its proximal ligand mutants: implications for heme uptake and release. Biochemistry, 2008. 47: p. 2087-2098.

26  

70. Krieg, S., Huche´, F., Diederichs, K., Izadi-Pruneyre, N., Lecroisey, A., Wandersman, C.,

Delepelaire, P., and Welte, W., Heme uptake across the outer membrane as revealed by crystal structures of the receptor-hemophore complex. Proc. Natl. Acad. Sci. U.S.A., 2009. 106: p. 1045-1050.

71. Deniau, C., Gilli. R., Izadi, N., Létoffé, S., Delepierre, M., Wandersman, C., Briand, C., Lecroisey, A., Thermodynamics of heme binding to the HasA(SM) hemophore: effect of mutations at three key residues for heme uptake. Biochemistry, 2003. 42: p. 10627-10633.

72. Alontaga, A. Y., Rodrı´guez, J. C., Scho¨nbrunn, E., Becker, A., Funke, T., Yukl, E. T.,

Hayashi, T., Stobaugh, J., Moe¨nne-Loccoz, P., and Rivera, M., Structural characterization of the hemophore HasAp from Pseudomonas aeruginosa: NMR spectroscopy reveals protein-protein interactions between Holo-HasAp and hemoglobin. Biochemistry, 2009. 48: p 96-109.

73. Adams, P. A., and Berman, M. C., Kinetics and mechanism of the interaction between human serum albumin and monomeric haemin. Biochem. J., 1980. 191: p 95-102.

                              

27  

CHAPTER II

Structural, NMR Spectroscopic and Computational Investigation of Heme

Loading in the Hemophore HasAp from Pseudomonas aeruginosa

Introduction

The opportunistic pathogen Pseudomonas aeruginosa (P. aeruginosa) prefers aerobic

metabolism which necessitates respiratory enzymes that need significant amounts of iron for

their function [1]. Typically, bacterial organisms (such as E.coli) require approximately 0.3-1.8

M of iron for optimal growth; however in a mammalian host iron is tightly sequestered by

dedicated iron binding and iron storage proteins such that the concentration of free iron is very

low (10-9 mM) [2]. Furthermore, under aerobic conditions at neutral pH, Fe3+ is extremely

insoluble (10-18 M) [3]. To thrive in mammalian host cells therefore, pathogenic bacteria have

evolved remarkable sophisticated mechanisms to circumvent host-mediated sequestration. One

of the systems developed by some Gram-negative pathogens (such as P.aeruginosa) involve

secretion of hemin binding proteins (hemophores) called HasA. The has (hemin acquisition

system) operon encodes for hasRADEBF genes where HasR is the hemophore-specific outer

membrane receptor, HasA is the hemophore, HasD an inner membrane ABC protein, HasE a

membrane fusion protein, HasB a TonB homologue and HasF an outer membrane protein [4].

HasA is secreted to the extracellular media where it binds heme (free or hemoprotein-bound) and

delivers it to the outer membrane hemophore-specific receptor HasR [5]. The gene coding for the

hemophore of P. aeruginosa, HasAp, is upregulated under iron limiting conditions and is

28  

essential for growth of P. aeruginosa with hemoglobin as the only source of iron [6]. These

observations point to an important role of HasAp in the capture and subsequent utilization of

heme iron. The hemophore from Serratia marcescens (HasAs) [7] which shares function and

sequence similarities with HasAp binds heme with a high affinity (~ 5.3 × 1010 M-1) [8] and

shuttles it the outer membrane receptor HasR [9]. The high affinity of HasAs for heme has been

attributed to the presence of the proximal ligand Tyr 75 and its hydrogen bond partner His 83,

acting in concert to the hemin-HasAs complex [10].

In the holo form both HasAs [11,12] and HasAp [13] have similar structures composed of

an “-helix wall” and a “-sheet wall” connected by two extended loops that harbor the heme

coordinating ligands (His 32 and Tyr 75) (Figure 2). Henceforth, the loops which harbor His 32

and Tyr 75 will be referred to as His 32 loop and Tyr 75 loop respectively. Solution NMR

structure of apo-HasAs showed that His 32 loop is spatially re-oriented and is approximately 30

Å away relative to its position in the holo-protein, while the Tyr 75 loop and the rest of the

structure remain virtually unaffected [14]. It is thought that binding of hemin onto the Tyr 75

loop causes the His 32 loop to move toward the closed conformation where heme binding is

mediated by His 32 coordination and hydrogen bonding interactions between the propionate

groups and the protein to form a stable complex. When His 32 is replaced with a noncoordinative

residue Ala (H32A) in HasAs, the mutant still binds hemin albeit with a lower affinity (Ka ~ 1.0

X 1010 M-1); a similar replacement in Tyr 75 (Y75A) causes an appreciable decrease in affinity

by nearly 2 orders of magnitude (Ka ~ 1.3 X 108) [8]. However, to date, there is no structural

information to complement these observations and hence neither the coordination state of the

heme iron in the mutants nor the topological position of the His 32 loop in the H32A mutant is

known. In addition, the mechanism by which heme loads onto the protein has not been

29  

elucidated. Nevertheless it has been proposed that heme binding is a two-step where initially, the

heme loads onto the Tyr 75 loop followed by closure of the His32 loop [14]. Recent independent

studies in support of this proposition have been reported; (i) reduction of holo-HasAs with

dithionite in the presence of CO results in formation of a ferrous hexacoordinate complex in

which His 32 is displaced by CO at the distal site [15], (ii) the crystal structure of HasAs in

complex with its receptor (HasR) show that the heme in the hemophore is coordinated only by

Tyr 75 at the proximal site while the His32 loop is turned away from its heme-binding

conformation and either assumes multiple conformations or is disordered [16]. In both cases, His

32 is deemed more labile than Tyr 75 hence the hypothesis that the heme first “latch” onto the

Tyr 75 loop followed by relocation of the His 32 loop. If this mechanism applies, it would be

distinct from that elucidated for the Shp hemophore secreted by the Gram-positive Streptococcus

pyogenes [17], which binds heme with two axial Met ligands [18] without detectable

intermediates [19].

Figure 1: X-ray crystal structure of truncated HasAp monomer B (PDB: 3ELL) showing -helices

(cyan), -strands (magenta), loops (wheat), heme (red), and heme axial ligands (green).

8

5

30  

In this study, we present experimental evidence which demonstrates that heme loading

proceeds in two distinct sequential steps: a fast kinetic phase where the heme initially loads onto

the Tyr 75 loop, followed by a slow phase which involve conformational rearrangement of the

His 32 loop and subsequent coordination of the heme. We also present tangible evidence that the

presence of the heme in the Tyr 75 loop triggers relocation of the His 32 loop regardless of the

presence or absence of the heme-iron distal ligand. We used targeted molecular dynamics

simulations utilizing the structures of heme-free and heme-loaded HasAp to gain more insights

in the molecular motions that accompany the transition of the His 32 loop from the “open” to

“closed” conformation and suggest a plausible path for its relocation.

Experimental Procedures

Construction of H32A-HasAp

A gene coding for a truncated form of HasAp missing the last 21 C′-terminal amino acids

relative to the full-length protein was constructed as described previously [13]. H32A-HasAp

mutant was constructed on the background of the truncated form of HasAp by site directed

mutagenesis using the QuickChange mutagenesis kit (Stratagene) (Figure 3) and the mutation

was confirmed by DNA sequencing (SeqWright) prior to transforming the plasmid into E. coli

BL 21 GOLD (DE3) competent cells.

31  

Expression and Purification of apo-proteins

An overnight starter culture of 50 mL Luria-Bertani (LB) medium containing 100μg/mL

ampicillin was grown from a single colony of freshly transformed plasmid. 30 mL of the culture

was used to inoculate 1 L of M9 minimal medium supplemented with 200 μg/mL ampicillin.

Cells were grown at 37 oC with continuous shaking at 235 rpm. Once OD600 was ~ 0.8-0.9 the

temperature was lowered to 30 °C and incubated further until OD600 reached ∼1.0. At this point

protein expression was induced by addition of isopropyl-D-thiogalactopyranoside (IPTG) to a

final concentration of 1 mM. The cells were grown for 5 h at 30 °C, harvested by centrifugation

(4800 rpm for 12 min at 4 °C) and stored at -20 °C. The cell paste was resuspended in 20 mM

Tris-HCl, pH 7.6, (3 mL/g cell paste) and disrupted by sonication (amplitude-70%) in the

presence of DNase (Sigma-Aldrich, St. Louis, MO). Large particles and cell debris were

separated by initial centrifugation at 2800 rpm and 4 °C for 10 min using a table centrifuge,

followed by complete clarification by ultracentrifugation at 19500 rpm and 4 °C for 40 min.

HasAp expressed in E.coli produces a mixture of apo- and holo-protein and is recovered

in the soluble fraction. The apo form was gently isolated from the holo fraction using two

chromatographic technique; ion exchange chromatography (IEC) at 4 °C and hydrophobic

interaction chromatography (HIC) at ambient temperature. Specifically, the clarified supernatant

was loaded onto a Q-Sepharose Fast Flow column (2.6 cm i.d. × 15 cm; GE Healthcare) pre-

equilibrated with 20 mM Tris-HCl (pH 7.6). The column was then washed with at least 3 bed

volumes of the same buffer before the protein was eluted with the same buffer and a linear

gradient (0-600 mM) of NaCl. Fractions containing HasAp were pooled and to it ammonium

sulfate was added to a final concentration of 0.7 M. The sample was then loaded onto a Butyl

Sepharose Fast Flow (GE Healthcare) column (2.6 cm i.d. × 12 cm) pre-equilibrated with 50 mM

32  

sodium phosphate/0.7 M ammonium sulfate, pH 7.0. Elution of the protein was achieved in two

stages: (i) 50 mM sodium phosphate/0.5 M ammonium sulfate, pH 7.0 was used to elute weakly

S N T G G F N P

CCG TTT GAT GGC AGC CAG TAT GCG CTG AAC AGC ACC GCG AGC GAT GCGP F D G S Q F A L K S T A S D A

TTT ATT GCG GGC GGC GAT CTG CAT TAT ACC CTG TTT AGC AAC CCG AGC CAT F I A G G D L H Y T L F S N P S

ACC CTG TGG GGC AAA CTG GAT AGC ATT GCG CTG GGC GAT ACC CTG ACC GGCT L W G K L D S G

GGC GCG AGC AGC GGC GGCTAT GCG CTG GAT AGC CAG GAA GTG AGC TTT AGC G A S S G G Y A L D S Q E V S F S

AAC CTG GGC CTG GAT AGC CCG ATT GCG CAG GGC CGT GAT GGC ACC GTGN L G L D S P I A Q G R D G T V

AAA GTG GTG TAT GGC CTG ATG AGC GGC GAT AGC AGC GCG CTG CAG GGC CAGK V V Y G L M S G D S S A L Q G

ATT GAT GCG CTG CTG AAA GCG GTT GAT CCG AGC CTG AGC ATT AAC AGC ACC I D A L L K A V D P S L S I N S

TTT GAT CAG CTG GCG GCG GCG GGC GTG GCG CAT GCG ACC CCG GCG GCG GCGF D Q L A A A G V A H A T P A A A

GCG GCG GAA GTG GGC GTG GTG GGC GTG CAG GAA CTG CCG CAT GAT CTG GCGA A E V G V V G V Q E L P H D L

Bam H ICTG GCG GCG TA

A GGATCC GGG

L A A

-

5 ' Nde IGGGCATATG AGC ATT AGC ATT AGC TAT AGC ACC ACCTAT AGC GGC TGG ACC

M S I S I S Y S T T Y S G W T

GTG GCG GAT TAT CTG GCG GAT TGG AGC GCG TAT TTT GGC GAT GTG AAC GCGV A D Y L A D W S A Y F G D V N A

CGT CCG GGC CAG GTG GTG GAT GGC AGC AAC ACC GGC GGC TTT AAC CCG GGCR P G Q V V D G G

GCGP F D G S Q F A A

TTT ATT GCG GGC F I A G G D L H H

T L W G K L D S I A L G D T L T

GGC GCG AGC

CATN L G L D S P I H

K V V Y G L M S Q

ATT GAT GCG CTG T

TTT GAT CAG CTG GCG

GCG GCG GAA GTG GGC GTGA

Bam H I

L A A

-

3 '

Figure 2. DNA and amino acid sequence of HasAp-TR. The Nde I and BamHI

restriction endonuclease sites were constructed at the 5' and 3' ends, respectively, for

subcloning. The underlined codon represents the mutation of His to Ala at position 32.

33  

bound proteins and holo-HasAp, and (ii) a linear gradient of sodium phosphate buffer (50-20

mM)/ammonium sulfate (0.50-0 M) was used to elute apo-HasAp. It is noteworthy to mention

here that apo-HasAp tenaciously binds to the column and elutes toward the ends of the linear

gradient (~ 20 mM Napi). Thus following elution from the butyl column, fractions containing the

apo-protein were pooled and again ammonium sulfate was added to a final concentration of 0.7

M before being loaded onto a Phenyl Sepharose 6 Fast Flow high substitution (GE Healthcare)

column (2.6 cm i.d. × 12 cm) pre-equilibrated with the same buffer as in the butyl column. The

protein was similarly eluted using steps (i) and (ii) described above. The purity of the protein in

the resultant fractions was assessed by 15 % SDS-PAGE (sodium dodecyl sulfate polyacrylamide

gel electrophoresis), and if necessary, fractions containing apo-HasAp were purified again on a

second Butyl Sepharose column.

Expression and Purification of Uniformly Labeled Proteins

Uniformly labeled U-15N-H32A-HasAp and U-13C/U-15N-H32A-HasAp were prepared

using a minor modification of the above-described protocol: Cells grown to OD600 ~ 0.8-0.9 were

harvested by centrifugation (4800 rpm, at 4 oC, for 12 min) and resuspended in fresh 1 L M9

medium supplemented with 200 g/mL ampicillin containing 1.0 g of 15NH4Cl as the sole

nitrogen source for preparing U15N-H32A-HasAp or 1.0 g of 15NH4Cl and 2.0 g of 13C6-D-

glucose as the sole nitrogen and carbon sources for preparing U-13C/U-15N-H32A- HasAp. The

cells were grown until OD600 reached ~1.0 after which biosynthesis of polypeptide was induced

by adding isopropyl--D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM. The

cells were incubated for 5 h at 30 °C, harvested, lysed and purified as described above.

34  

Expression and Purification of Selectively Labeled Proteins

Expression of selectively labeled samples was done following previously described

protocols [20, 21, 22] with minor modifications. Prior to autoclaving, synthetic amino acids (g/L)

except L-Trp (0.05) were added to the M9 media: L-Ala (0.5), L-Arg (0.4), L-Asn (0.4), L-Asp

(0.4), L-Cys (0.05), L-Gln (0.4), L-Glu (0.65), Gly (0.55), L-His (0.1), L-Ile (0.23), L-Leu

(0.23), L-Lys hydrochloride (0.42), L-Met (0.25), L-Phe (0.13), L-Pro (0.1), L-Ser (2.10), L-Thr

(0.23), L-Tyr (0.17), and L-Val (0.23).

Expression of [15N-Gly]

1.0 L of M9 medium supplemented with all the amino acids except Gly was inoculated

with 30 mL of overnight E. coli BL21 GOLD (DE3) starter culture. When the OD600 approached

1.0, the temperature was lowered and biosynthesis of polypeptide was induced by adding IPTG

to a final concentration of 1 mM, followed by the addition of 0.55 g [15N-Gly]. The cells were

incubated for 5 h at 30 °C, harvested, lysed and purified as described above.

Expression of [15N-Ala]

1.0 L of M9 medium supplemented with all the amino acids except L-Ala was inoculated

with 30 mL of overnight E. coli BL21 GOLD (DE3) starter culture. When the OD600 approached

1.0, the temperature was lowered and biosynthesis of polypeptide was induced by adding IPTG

to a final concentration of 1 mM, followed by the addition of 0.50 g [15N-L-Ala]. The cells were

incubated for 5 h at 30 °C, harvested, lysed and purified as described above.

35  

Expression of [15N-Val]

1.0 L of M9 medium supplemented with all the amino acids except L-Val was inoculated

with 30 mL of overnight E. coli BL21 GOLD (DE3) starter culture. When the OD600 approached

1.0, the temperature was lowered and biosynthesis of polypeptide was induced by adding IPTG

to a final concentration of 1 mM, followed by the addition of 0.23 g [15N- Val]. The cells were

incubated for 5 h at 30 °C, harvested, lysed and purified as described above.

Sample preparation

Preparation of H32A-holo-HasAp

1.8-2.0 μmol of apo-H32A (8-10 μM) in sodium phosphate buffer μ = 0.1, pH 7.8 were

reconstituted with hemin by titrating with a freshly prepared solution of hemin in DMSO (1 mM)

under constant stirring until the absorbance ratio (A280/A402) remained constant. The resultant

holo-protein was incubated overnight at 4 °C, concentrated to 2 mL using 10 kDa MWCO

Amicon ultracentrifuge filter devices, and passed through a G-75 Sephadex column (2.6 cm i.d.

× 90 cm; GE Healthcare) equilibrated with phosphate buffer (μ = 0.1, pH 7.8).

Preparation of H32A-Imidazole complex

H32A-imidazole complex was prepared by titrating a solution of H32A-holo-HasAp with

0.1 M imidazole (pH 7.0) until soret maximum remained constant. The reaction was monitored

with UV-vis spectrophotometer. The affinity of imidazole for H32A-HasAp was determined as

follows; fresh protein (final concentration of 5 μM) was added to 2 mL each of increasing

36  

concentration of imidazole buffer (1, 3, 5, 10 …100 mM) pH 7.0, contained in a cuvette. In

each case, the mixture was incubated with constant stirring for 2 minutes after which the Soret

maximum was recorded. The imidazole-protein complex formed was followed

spectrophotometrically by monitoring changes in the Soret band (red shifting) until there was no

further change. The binding constant (KD) was determined by measuring the amount of

imidazole-H32A-HasAp complex formed at different concentrations of imidazole and fitting the

hyperbolic curve to equation 1 [23];

AB = AB max {[B]/ ([B] +KD)} ------------------------------- Eqn 1

Where AB = amount of complex formed; AB max = maximum amount of complex that can form

at saturating [B], [B] = concentration of imidazole, and KD = binding constant.

For crystal growth screens the protein was exchanged to 150 mM imidazole buffer by

subjecting the sample to three cycles of buffer exchange using 10 kDa MWCO Amicon Ultra-15

centrifugal filter units, replenishing the solution with 150 mM Imidazole pH 7.0 after every

concentration step.

Determination of molecular weight

(a) Gel filtration chromatography

Molecular weight of apo- and holo-H32A-HasAp was estimated with the aid of a high-

resolution gel filtration column (Superdex 75 prep grade, GE Healthcare) equilibrated with 50

37  

mM Tris-HCl (pH 7.5), 100 mM KCl. The column was calibrated with a set of molecular weight

standards (GE Healthcare) including cytochrome c (12.4 kDa), chymotrypsin (25 kDa),

ovalbumin (44 kDa), and albumin (67 kDa) wherefrom a calibration curve was constructed. Pure

protein samples were passed through the same column and the molecular weight was determined

using their respective elution volumes.

(b) Electrospray Ionization Mass Spectrometry (ESI/MS)

Exact molecular mass of H32A-HasAp samples was determined by mass spectrometry

using a Waters capLCtm system equipped with an autoinjector coupled to ESI-MS-Q-TOF. Prior

to injection, ~ 5 M sample in 100 mM sodium phosphate buffer, pH 7.4, was diluted with H2O

in a 1:1 ratio. Separation was achieved by running acetonitrile/isopropyl alcohol (IPA)/water

linear gradient of 4 % acetonitrile/min at 10 L/min from 20 -60 % acetonitrile (buffer A- 99 %

water/ 1 % acetonitrile/ 0.08 % formic acid, buffer B- 80 % acetonitrile/ 10 % IPA/ 10 % water/

0.06 % formic acid) using a 0.5 mm ID C4 reverse phased column (MC-5-C4, 300Å pore size,

3.5 M particles; Micro-Tech).

ESI spectra were acquired on a Q-TOF-2 (Micromass Ltd, Manchester UK) hybrid mass

spectrometer operated in MS mode and acquiring data with the time of flight analyzer. The

instrument was operated for maximum resolution with all lenses optimized on the M2H2+ ion

from the cyclic peptide Gramicidin S. The cone voltage was 30 eV and Ar was admitted to the

collision cell at a pressure that attenuates the beam to about 20 %. The cell was operated at 5eV

(maximum transmission). Spectra were acquired at 11,346 Hz pusher frequency covering the

38  

mass range of 100 to 3000 amu and accumulating data for 5 seconds per cycle. Time to mass

calibration was made with cesium iodide (CsI) cluster ions acquired under the same conditions.

Resonance Raman (RR) Spectroscopy

RR spectra were obtained using a custom McPherson 2061/207 spectrograph (0.67 m

with variable gratings) equipped with a Princeton Instruments liquid N2-cooled CCD detector

(LN-1100PB). Excitation at 413 nm was provided by a krypton laser (Innova 302, Coherent), and

a Kaiser Optical supernotch filter was used to attenuate Rayleigh scattering. Spectra at room

temperature were collected in a 90° scattering geometry on samples mounted on a reciprocating

translation stage. Frequencies were calibrated relative to indene and CCl4 and are accurate to ±1

cm-1. CCl4 was also used to check the polarization conditions. The integrity of the RR samples,

before and after laser illumination, was confirmed by direct monitoring of their UV-vis spectra in

the Raman capillaries. Low temperature spectra were obtained in a backscattering geometry on

samples maintained at ∼105 K in a liquid nitrogen coldfinger. Frequencies were calibrated

relative to aspirin and are accurate to ±1 cm-1.

EPR Spectroscopy

EPR spectra were obtained on a Bruker E500 X-band EPR spectrometer equipped with a

superX microwave bridge and a dual mode cavity with a helium-flow cryostat (ESR900, Oxford

Instruments, Inc.). The experimental conditions such as temperature, microwave power, and

39  

modulation amplitude, were varied to optimize the detection of all potential EPR active species

and to ensure nonsaturating conditions.

X-Ray Crystallography

Protein Crystallization

Crystallization studies were carried out in the Protein Structure Laboratory at the Delbert

M. Shankel Structural Biology Center (University of Kansas, Lawrence, KS). Crystal growth

screening conditions were carried out using the sitting drop vapor diffusion method in Compact

Jr. Crystallization plates (Emerald BioSystems) at 20 °C. 0.5 L of pure proteinwere mixed with

an equal volume of reservoir solution and equilibrated against 100 μL of the latter.

Needle-like crystals of apo-HasAp (approximately 300 μm long) were obtained from 50

mg/mL protein in 100 mM sodium phosphate, pH 7.8, within 2 days from Wizard II (Emerald

BioSystems) condition #19 (1.6 M NaH2PO4/0.4 M K2HPO4, 100 mM phosphate-citrate pH 4.2).

Before freezing in liquid nitrogen for data collection, the samples were immersed in a

cryoprotant containing 80% crystallization solution and 20% glycerol for approximately 30 s.

Prismatic crystals of dimeric H32A holo-HasAp (approximately 700 μm long) were obtained

from 15 mg/mL protein in 100 mM sodium phosphate, pH 7.0, within 24 h from Wizard II

condition #15 (1.26 M (NH4)2SO4, 100 mM HEPES pH 7.5). Crystals were transferred and

equilibrated for 60 s in a solution containing 2 M (NH4)2SO4, 100 mM HEPES pH 7.5, 10% d-

sorbitol and then in the same solution containing 20% d-sorbitol before freezing in liquid

nitrogen. Prismatic crystals of imidazole bound H32A-holo-HasAp obtained after approximately

4 weeks from 9.3 mg/mL protein in 150 mM imidazole pH 7.0 grew from Wizard II condition

40  

#34 (10% (w/v) PEG 8000, 100 mM imidazole pH 8.0). Samples were transferred to a

cryoprotectant containing 80% crystallization solution and 20% (v/v) PEG 400 before freezing in

liquid nitrogen for data collection. Wizard II random sparse matrix crystallization screen is

shown in Appendix 1.

X-ray Data Collection

Data for apo- and dimeric holo-HasAp were collected at 100 K at the Advanced Photon

Source (APS) IMCA-CAT, beamline 17BM using an ADSC Quantum 210r CCD detector.

Diffraction data for imidazole bound H32A-holo-HasAp were collected at 93 K in-house using a

Rigaku RU-H3R rotating anode generator equipped with Osmic Blue focusing mirrors and an R-

axis IV+2 image plate detector. Apo-HasAp and imidazole bound H32A-holo-HasAp diffraction

data were integrated and scaled with the HKL2000 package [24] while dimeric H32A-holo-

HasAp data were processed with the XDS [25] package. The Laue class for each data set was

checked using Pointless [26].

Structure Solution and Refinement

Structure solution of apo-HasAp was obtained by molecular replacement with Molrep [27]

using the holo-HasAp structure (PDB: 3ELL) [13] as the search model. Automated model

building was conducted with ARP/wARP [28]. The final model was used for subsequent

molecular replacement searches with the dimeric H32A-holo-HasAp and imidazole bound

H32A-holo-HasAp data sets. Structure refinement was performed with Refmac [29] for apo and

imidazole bound H32A-holo-HasAp and with Phenix [30] for dimeric H32A-holo-HasAp,

respectively. Manual model building was carried out with Coot [31]. The crystals of dimeric

41  

H32A-holo-HasAp were indexed in a monoclinic C lattice with a = 110.34 Å, b = 52.70 Å, c=

71.74 Å, = 129.05°, and space group C2. However, the unit cell and space group were

transformed to the nonstandard setting I2 using the reindexing matrix (0 0 -1, 0 1 0, 1 0 1) to give

a = 71.74 Å, b = 52.70 Å, c = 85.71 Å, = 91.48°. For dimeric H32A-holo-HasAp, anisotropic

displacement parameters were modeled by TSL refinement [32] using 10 and 11 TLS groups for

subunits A and B respectively. Molecule B appeared to display increased thermal motion as

indicated by the larger overall B-factor (∼2-fold) relative to molecule A. For the imidazole bound

H32A-holo-HasAp, 7 TLS groups were refined for both molecules of the noncrystallographic

dimer in the asymmetric unit. Structure validation was performed using MOLPROBITY [33],

and figures were prepared using CCP4 mg [34]. A summary of crystallographic data is provided

in Table 1. For apo-HasAp, phosphate and sodium ions were identified from the difference

electron density and assigned based on the coordination and expected nonbonded distances [35].

NMR Spectroscopy

All experiments were conducted using a Varian Unity Inova 600 NMR spectrometer

equipped with a triple resonance probe. Sequential backbone resonance assignments were

obtained with the aid of 1H-15N-HSQC, HNCA, HN(CO)CA, HNCACB, CBCA(CO)NH,

HNCO, and HN(CA)CO experiments conducted with H32A-HasAp sample uniformly labeled

with 13C and 15N. The samples were contained in phosphate buffer (μ = 0.1, pH 7.0) with 5-10%

D2O and the concentrations ranged between 3.9 and 6.7 mM. After collecting the data that was

required to carry out resonance assignments of H32A- HasAp, the samples were exchanged to

150 mM imidazole buffer, pH 7.0 using 10 KDa MWCO Amicon Ultra-15 centrifugal filter

42  

units. The protein samples in imidazole buffer (H32A-imidazole) were finally concentrated to the

same volume as the original H32A-HasAp sample to maintain the protein concentration constant.

Samples of 15N-Gly (4.6 mM), 15N-Ala (4.4 mM), and 15N-Val (5.2 mM) selectively labeled

proteins were used to collect two-dimensional 1H-15N-HSQC spectra using standard and rapid

repetition conditions. For rapid repetition HSQC experiments, the acquisition time and relaxation

delay were shortened to 35 and 50 ms, respectively from the standard 95 ms and 1 s, and the

number of scans was increased to at least 256 from the typical 16 scans per increment used in

standard experiments. All other acquisition parameters were identical. Unless otherwise stated,

the experiments were conducted at 32 °C. Other relevant acquisition and processing parameters

used to obtain specific spectra are included in the corresponding figure caption. The data were

processed using NMRPipe [36] and analyzed with the program Sparky [37]. 1H chemical shifts

were referenced directly to internal DSS using the methyl 1H resonance at 0.00 ppm, whereas

15N and 13C chemical shifts were referenced indirectly using the ratio of gyromagnetic ratios.

Targeted Molecular Dynamics (TMD) Simulations

All calculations were performed using the biomolecular simulation program CHARMM

[38] with the all-atom parameter set PARAM22 for proteins [39], including the dihedral cross-

term corrections (CMAP) [40] and a modified TIP3 water model [41]. The simulation system,

Figure 3, was built by using the Quick MD Simulator module in CHARMM-GUI [42] after the

hemin-iron was coordinated to the phenol group of Tyr 75 in the structure of apo-HasAp. Holo-

HasAp was solvated in a 80 Å × 80 Å × 80 Å water box with 0.15 M KCl, resulting in a total

of 48 292 atoms; the number of water molecules, K+ ions, and Cl- ions are 15 191, 50, and 37,

43  

respectively; different numbers of K+ and Cl- ions were added to make the simulation system

neutral. After a 350 ps equilibration time, 25 independent 1.2 ns TMD simulations were

performed starting from the “open” (apo) and progressing to the “closed” (holo) structure. A time

step of 2 fs was used for the NVT (constant volume and temperature) dynamics with the SHAKE

algorithm [43]. The same options as those in the input scripts provided by CHARMM-GUI were

used for nonbonded interactions [42]. In total, 50 structures (“images”) were generated at equal

intervals of the root mean squared deviation (rmsd) between the open and closed structures for

each run, followed by clustering the images obtained at each of the rmsd intervals. Analysis of

the trajectories was also performed to identify important interactions during the transition from

“open” to “closed” structures starting from the “open” (apo) and progressing to the “closed”

(holo) structure. A time step of 2 fs was used for the NVT (constant volume and temperature)

dynamics with the SHAKE algorithm [44]. The same options as those in the input scripts

provided by CHARMM-GUI were used for nonbonded interactions [42]. In total, 50 structures

(“images”) were generated at equal intervals of the root mean squared deviation (rmsd) between

the open and closed structures for each run, followed by clustering the images obtained at each

of the rmsd intervals. Analysis of the trajectories was also performed to identify important

interactions during the transition from “open” to “closed” structures.

44  

  A B CA B C

Table 1. Crystallographic data of HasAp structures.

Apo-HasAp HasAp H32A Dimer HasAp H32A Imidazole

Data Collection

Unit-cell parameters (Å, o) a = 32.98, b = 65.25,

c = 38.34, = 110.5

a = 71.74 Å, b = 52.70 Å,

c = 85.71 Å, = 91.48

a = 41.37 Å, b = 46.38 Å,

c = 81.09 Å, = 97.01

Space group P21 I2 P2

Resolution (Å)i 50.0-1.55 (1.61 - 1.55)

50.0-1.20 (1.30-1.20) 30.0-2.25 (2.33-2.25)

Wavelength (Å) 1.0000 1.0000 1.5418

Observed reflections 78,274 443,848 47,076

Unique reflections 21,268 95,175 16,389

<I/(I)>1 22.9 (2.1) 17.0 (3.4) 15.6 (2.3)

Completeness (%)i 98.5 (72.4) 95.0 (91.9) 96.6 (91.8)

Figure 3. The simulation system of hemin bound apo-HasAp. (A) Hemin bound apo-HasAp. (B) Hemin

bound apo-HasAp with 0.15 M KCl. (C) Hemin bound apo-HasAp with 0.15 M KCl aqueous solution.

Apo-HasAp (yellow) is shown in cartoon representation with the His32 and Tyr75 loops highlighted in

magenta. Hemin is shown in red sticks, hemin iron as an orange sphere, K+ as magenta spheres, and Cl-

as green spheres. Waters molecules are represented by lines (marine).

45  

Redundancyi 3.7 (2.1) 4.7 (3.5) 2.9 (2.6)

Rsymi,ii 7.6 (32.3) 4.1 (35.0) 411.6 (53.5)

Refinement

Resolution (Å) 35.92-1.55 19.25-1.20 24.65-2.25

Reflections (working/test) 20,125 / 1,095 87,374 / 4,379 15,450 / 821

Rfactor / Rfree (%)iii 15.3 / 17.8 17.7 / 19.7 21.4 / 26.7

No. of atoms protein (chain A, B)/heme/water)

1344 / - / 158 1353, 1273 / 86 / 305 1259, 1269 / 86 / 64

Model Quality

R.m.s deviations

Bond lengths (Å) 0.014 0.010 0.018

Bond angles (o) 1.434 1.421 1.617

Average B factor (Å2)

All Atoms 15.7 21.5 42.3

Protein 15.4 14.9 / 27.7 40.4 / 44.9

Water 25.8 25.7 36.8

Heme - 20.7 35.6

Phosphate or imidazole 26.0 - 41.9

Coordinate error based on maximum likelihood (Å)

0.05 0.16 0.21

Ramachandran Plot

Most favored (%) 97.3 97.2 95.9

Additionally allowed (%) 2.2 2.2 4.1

Disallowed (%) 0.5 0.4

i) Values in parenthesis are for the highest resolution shell.

ii) Rsym = hkli |Ii(hkl) - <I(hkl)>| / hkli Ii(hkl), where Ii(hkl) is the intensity

measured for the ith reflection and <I(hkl)> is the average intensity of all reflections with indices hkl.

iii) Rfactor = hkl ||Fobs (hkl) | - |Fcalc (hkl) || / hkl |Fobs (hkl)|; Rfree is calculated in an identical manner using 5% of randomly selected reflections that were not included in the refinement.

46  

Results

Purification of the apo-proteins

In E.coli, HasAp is expressed as a mixture of apo (~ 80%) and holo (~ 20 %) protein.

Previously, isolation of the apo protein was achieved by reconstituting the entire sample with

hemin followed by extraction of the heme [13] using the cold acid-acetone method as reported

for the preparation of apo-HasAs [9]. Briefly, pure holo-protein is concentrated and added

dropwise with constant stirring into cold acid-acetone solution (0.2 % V/V of 12 M HCl). Under

these conditions, heme is extracted and remains in solution while the protein precipitates.

Following gentle centrifugation, the precipitate is dissolved in appropriate buffer containing 7.0

M urea followed by extensive dialysis to rid urea and refold the protein. When this method is

applied to HasAp, extraction is not complete. The final apo-protein obtained contains

approximately 3-10 % residual holo-protein [13]. Undoubtedly this method would not give

authentic results in the experiments designed for this study. Thus, a chromatographic method

which capitalizes on tenacious binding of the apo-protein to functional groups in hydrophobic

affinity interaction columns relative to the holo-protein was developed. This approach allows

complete separation of the holo-form and gentle isolation of pure apo-protein for subsequent

crystallographic and spectroscopic studies.

Structural Characterization of apo-HasAp

The crystal structure of apo-HasAp shows that except for the His 32 loop, the overall fold

of HasAp is very similar in the apo and holo forms (Figure 4). Structural alignment of both

proteins shows that the position of the “β-sheet wall” comprised of eight antiparallel β-strands

47  

and the “α-helix wall” comprised of 4 -helices are nearly identical (Figure 5). These secondary

structure elements are stabilized by tight networks of hydrogen bonding and hydrophobic

interactions. For the same reasons, the position of the loop that harbors the hemin-iron proximal

ligand, Tyr 75, is conserved. In contrast, the loop that harbor the distal ligand, His 32, is

relocated nearly 30 Å away relative to its position in the holo form. Solution NMR structure of

apo-HasAs [14] shows that the equivalent loop is also relocated by the same distance relative to

its position as seen in the crystal structure of holo-HasAs [11,12]. It is therefore very likely that

the large reorganization of the His 32 loop upon heme binding is a common feature of heme

capture by equivalent hemophores from different bacteria.

 

Figure 4. X-ray crystal structure of apo-HasAp (PDB: 3MOK). His 32 loop is shown in green, Tyr75

loop is shown in red, while His 32 and Tyr 75 are shown in magenta and orange, respectively.

48  

  ~ 90o

26.1 Å

31.2 Å

Comparison of the amino acid sequence to other hemophores (Figure 6) in the context of

the structural alignment of apo- and holo-HasAp shows that most of the conserved residues are

hydrophobic and majority of them are found around the beginning and the end of His32 loop

(Gly 28 to Gly 45), and Tyr 75 loop (His 74 to Leu 85). Small aliphatic residue Gly 28 (C`-end

of 1) and Gly 45 (beginning of 2) constitute the “hinges” about which His32 loop pivots

(Figure 7-A). In close proximity are conserved aromatic residues; Phe 27, near hinge residue

Gly 28, and Phe 46, near hinge residue Gly 45 (Figure 7-B & C). The side chain of Phe 27 is

located at the end of 1 and packs against the side chain of conserved Trp 23, which in turn is

stabilized by additional hydrophobic interactions with aromatic side chains upstream in 1

Figure 5: Overlaid x-ray crystal structures of HasAp showing the position of the loops in apo- (green)

and holo-(red) wild type protein. The heme is colored blue, axial ligands in holo-HasAp are shown in

orange.

49  

10 20 30 40 50 60 | | | | | | Pa MSISISYSTTYSGWTVADYLADWSAYFGDVNHRPGQVV-DGSNTGGFNPGP--FDGSQYALKS-TASDAAFIA Sm MAFSVNYDSSFGGYSIHDYLGQWASTFGDVNHTNGN-V-TDANSGGFYGGS--LSGSQYAISSTANQVTAFVA Pe MTISVNYDAAFASSTIDDYLAFWSAGFVTAGHG-------YSNTGGFSNGT--YDGDQYATHGRNSSDYAFIA Pf MSISISYSATYGGNTVAQYLTDWSAYFGDVNHRPGEVV-DGTNTGGFNPGP--FDGTQYAIKS-TASDAAFVA Yps MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHGNPEGMIA Pc MSFAITYDAYYANYSIASYLTEWSAAFGDVNHTAGNTQVGGNNTGGFYGGDTFIDGTQYAITSTQNDFSALIA Ec MSFAITYDAYYANYSIANYLTEWSAAFGDVNHTTGNTQVGGNNTGGFYGGDTFIDGSQYAITSTQNDFSALIA Yp MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHSNPEGMIA Ypa MSTTIQYNSNYADYSISSYLREWANNFGDIDQAPAETK----DRGSFSGSSTLFSGTQYAIGSSHSNPEGMIA 70 80 90 100 110 120 | | | | | | Pa GGD----LHYTLFS------NPSHTLWGKLDSIALGDTLT---GGASSGG-YALDSQEVSFSNLGLDSPIAQ- Sm GGN----LTYTLFN------EPAHTLYGQLDSLSFGDGLS---GGDTSP--YSIQVPDVSFGGLNLSSLQAQ- Pe DSNNANGLHYVFDPSKAPGDNLNHYLWGSLDNVSLGEVLG---GGSGSD--FTLGNYVVNFNGLDLDAAHGA- Pf DGN----LHYTLFS------NPSHTLWGSVDTISLGDTLA---GGSGSN--YNLVSQEVSFTNLGLNSLKEE- Yps EGN----LKYSFM--------PQHTFYGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG Pc GGD----LTYSLFT------PPAHTLYGDLDSLSFGNVLQ---GGTTAGTTYSLAEPEVTFSGLDLSTDVAN- Ec GGD----LTYSLFS------PPAHTLYGDLDSLSFGNVLQ---GGTTAGTTYSLVEPEVTFSGLDLSTDVAN- Yp EGD----LKYSFM--------PQHTFHGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG Ypa EGD----LKYSFM--------PQHTFHGQIDTLQFGKDLATNAGGPSAG--KHLEKIDITFNELDLSGEFDSG 130 140 150 160 170 180 190 | | | | | | | Pa ---GR-DG-TVHKVVYGLMSGDSSALQGQIDALLKAVDPSLSINSTFDQLAAAGVAHATP-AAAAAEVGVVGV Sm ---GH-DG-VVHQVVYGLMSGDTGALETALNGILDDYG--LSVNSTFDQVAAA----T-----------AVGV Pe ---GR-AGNEVQTVIYGLMQGNTAGLEGVLDTLLASYG--VSTNDTFDVVAAA----LGG-ASGAASAQAVGV Pf ---GR-AG-EVHKVVYGLMSGDSSALAGEIDALLKAIDPSLSVNSTFDDLAAAGVAHVNPAAAAAADVGLVGV Yps KSMTENHQGDMHKAILGLRKGNA----DPMLEVMKAKG--FDVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV Pc --LTVSDRGVVHDVIYGLMSGQV----QPLLDALTNAG--IDINASLDSLSFATATSDAALSADT-VVDVVGV Ec --LTVSDRGVVHDVIYGLMGGQV----QPLLDALTNAG--IDINASLDSLSFATATSDAALSADT-VVDVVGV Yp KSMTENHQGDMHKSVRGLMKGNP----DPMLEVMKAKG--INVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV Ypa KSMTENHQGDMHKSVRGLMKGNP----DPMLEVMKAKG--INVDTAFKDLSIASQYPDSGYMSDAPMVDTVGV 200 205 | | Pa QELPHDLALAA Sm QHADSPELLAA Pe QALPEDLALAA Pf QDVAQDWALAA Yps MDSN-DMLLAA Pc AETA-DLLAA- Ec AETA-DLLAA- Yp VDCH-DMLLAA Ypa VDCH-DMLLAA

Figure 6. Amino acid sequence alignment of full-length HasA sequences from P. aeruginosa

(Pa), Serratia marcescens (Sm), P. entomophila (Pe), P. fluorescens (Pf), Yersinia

pseudotuberculosis (Yps), Pectobacterium carotovorum (Pc), Erwinia carotovora (Ec), Yersinia

pestis (Yp), and Yersinia pestis bv. Antiqua (Ypa) aligned against the sequence of HasA from P.

aeruginosa. Truncated HasAp used in these investigations stops at residue 184 (full-length minus

21 amino acids) [13]. Residues highlighted in red are conserved while those in blue are the His32

loop “hinge” residues in HasAp. The alignment was performed using ClustalW [45].

50  

Figure 7. (A) Overlay of apo (magenta) and holo (turquoise, PDB: 3ELL) HasAp showing a zoomed in view of the “hinge” region of the His32 bearing loop.(B) Apo-HasAp showing the

aromatic side chains (green) that stabilize the 1-loop near the Gly 28 hinge; the H32 loop is shown in magenta. (C) Apo-HasAp showing the aromatic side chains (green) that stabilize the structure near the Gly 44 hinge. The hemin molecule of holo-HasAp (PDB: 3ELL) is shown in red for reference

A

B

C

51  

(Figure 7-B). Seemingly, Phe 27 is placed immediately before the “hinge” to act as a “hinge

anchor” to preserve the structural integrity of 1 while the loop pivots between conformations.

The side chain of Phe 46 located at the beginning of 2 near the hinge residue Gly 44 packs

against the side chain of conserved Tyr 56. As shown in Figure 7-C, the side chains of both

aromatic residues interact with the vinly group of the macrocycle suggesting that they could be

playing an important role in stabilizing the heme complex.

Conformational changes observed upon heme binding onto apo-HasAp

As shown in Figure 8, heme binding induces a few subtle changes in the side chains of

residues near the heme binding site. In the Tyr 75 loop, the orientation of the side chain of Tyr

75 is nearly identical in both apo and holo forms possibly because it is the ligand that coordinates

to the heme. On the other hand, the side chain of the heme binding partner, His 83, undergoes

minimal movement upon heme binding (Figure 7-A), which reduces the hydrogen bond length

between His 83 δN and ηO of Tyr 75 by approximately 0.7 Å. Reduction of this bond length

may be a necessity to enhance the tyrosinate character of the heme ligand Tyr 75, in turn

strengthening the Tyr- Fe bond. Previous studies have shown that this hydrogen bonding

interaction is conserved in heme-bound hemophores [11,13,46 ]. Another significant residue that

is affected by heme loading is Arg 129. In the apo-protein the side chain of Arg 129 forms a

hydrogen bond (3.6Å) with ηO of Tyr 138 located in 2 (Figure 8-A). Upon heme binding the

side chain of Arg 129 reorients such that the NH2 group initially hydrogen bonded to Tyr 138

moves away by ~4.7 Å to form a salt bridge with one of the heme propionates (3.4 Å). On the

other hand, the side chain of Tyr 138 moves away to accommodate the heme and in turn forms a

52  

hydrogen bond (2.7 Å) with the carbonyl group of His 32. This new position likely provides

hydrophobic stabilization of the macrocycle and also may help stabilize the His 32-Fe

coordination. The side chain of Met 141, also located in 2, relocates upon heme binding. The

terminal methyl group moves toward the heme and engage in hydrophobic interaction, likely

stabilizing the complex. The only secondary structure element that is significantly affected when

the heme loads onto the hemophore is His 32 loop. Its movement, which also facilitates His 32

coordination to the heme, is accompanied by conformational change in the carbonyl oxygen of

the “hinge” residue Gly 28. The new conformation of the carbonyl oxygen of Gly 28 forces the

short loop comprised of Gly 143 and Asp 144 ( flanking 2 and 3) to move, causing

displacement in the C`-end of helix 2 (Figure 8-B). The opposite end of the loop (near Gly 45

hinge) is less obstructed, and thus heme loading does not bring about significant observable

changes on or around it.

A B

Characterization of H32A-HasAp mutant

Figure 8. Overlay of apo (magenta) and holo (turquoise, PDB: 3ELL) HasAp (A) Zoomed in

view of the heme binding pocket showing conformational differences between certain residues

upon hemin binding. (B) Displacement of helix 2 and the short loop comprised by G143 and

D144 relative to the apoprotein upon hemin binding; movement of the latter enables hinge

residue G28 in the holo structure to adopt its hemin bound conformation.

53  

The molecular weight of pure H32A-apo-HasAp estimated by gel filtration revealed that

in the apo form, the mutant is monomeric (Figure 9-C). However, when the apo-protein is

reconstituted with hemin and passed through a gel filtration column (Sephadex G-75), two

chromatographically distinguishable peaks are observed (Figure 9-A). The electronic absorption

spectrum obtained from fractions corresponding to the fast-eluting peak exhibits a broad Soret

band with a maximum at 399 nm, Q bands at 504 and 536 nm, and a CT band at 630 nm (black

trace in Figure 9-B). On the other hand, fractions corresponding to the slow-eluting peak exhibit

spectra with a relatively narrower Soret band at 401 nm, Q bands at 502 and 534 nm, and a CT

band at 623 nm (red trace in Figure 9-B). The position of the soret and the presence of the CT

bands above 600 nm indicate that fractions in both peaks contain high spin (hs) species. The

presence of / bands near 530 and 500 nm are indicative of the presence of low spin (ls)

species. Thus, judging from UV-vis spectra, protein fractions in both peaks exist as a mixture of

hs and ls species. These observations are corroborated by resonance Raman and EPR

spectroscopies. The fast eluting peak was identified by low-temperature resonance Raman as

having predominantly five-coordinate high-spin (5c-hs) ferric heme species while a greater

contribution of 6-coordinate species exists in the second fraction (Figure 10-A). In the EPR

spectra (Figure 10-B), both fractions exhibit intense g ≈ 6 signatures consistent with axial S=5/2

high spin iron (III) configurations, with evidence of signal broadening in the first fraction.

54  

Fraction number

0 5 10 15 20 25 30

Abs

orba

nce

0.0

0.2

0.4

0.6

0.8

1.0

A

dimer

monomer

Fraction number

0 5 10 15 20 25 30

Abs

orba

nce

0.0

0.2

0.4

0.6

0.8

1.0

A

Fraction number

0 5 10 15 20 25 30

Abs

orba

nce

0.0

0.2

0.4

0.6

0.8

1.0

A

dimerdimer

monomermonomer

Wavelength (nm)

300 400 500 600 700

Abs

orban

ce

0.00

0.05

0.10

0.15

0.20

0.25

0.30

401

5X

630623

534

504536

502

399

B monomerdimer

Wavelength (nm)

300 400 500 600 700

Abs

orban

ce

0.00

0.05

0.10

0.15

0.20

0.25

0.30

401

5X

630623

534

504536

502

399

B

Wavelength (nm)

300 400 500 600 700

Abs

orban

ce

0.00

0.05

0.10

0.15

0.20

0.25

0.30

401401

5X

630623

534

504536

502

5X5X

630630623623

534534

504504536536

502502

399399

B monomerdimermonomermonomerdimerdimer

Ve (mL)

12 14 16 18

log

MW

1.0

1.2

1.4

1.6

1.8

2.0

Albumin

Ovalbumin

Chymotrypsin

Cytochrome C

apo H32A

holo H32A monomer

holo H32A dimer

Ve (mL)

12 14 16 18

log

MW

1.0

1.2

1.4

1.6

1.8

2.0

Albumin

Ovalbumin

Chymotrypsin

Cytochrome C

apo H32Aapo H32A

holo H32A monomerholo H32A monomer

holo H32A dimerholo H32A dimer

C

Ve (mL)

12 14 16 18

log

MW

1.0

1.2

1.4

1.6

1.8

2.0

Albumin

Ovalbumin

Chymotrypsin

Cytochrome C

apo H32A

holo H32A monomer

holo H32A dimer

Ve (mL)

12 14 16 18

log

MW

1.0

1.2

1.4

1.6

1.8

2.0

Albumin

Ovalbumin

Chymotrypsin

Cytochrome C

apo H32Aapo H32A

holo H32A monomerholo H32A monomer

holo H32A dimerholo H32A dimer

C

Figure 9. (A) Elution profile of H32A

holo-HasAp from a Sephadex G-75

column; open circles track the Soret band

absorption (∼A400 nm) and black circles

track A280 nm.

(B) Electronic absorption spectra of

fractions in the fast eluting peak (black,

dimer) and in the slow eluting peak (red,

monomer). The ratio of dimer to

monomer obtained under these

(C) Elution volumes (Ve) from a

calibrated Superdex 75 column and

estimated molecular weights for monomer

holo H32A HasAp (red triangle), dimer

holo H32A HasAp (green diamond), and

apo H32A HasAp (blue triangle); the

average molecular weight was obtained

from two experiments. The Superdex 75

column was calibrated with albumin (67

kDa), ovalbumin (44 kDa), chymotrypsin

(25 kDa), and cytochrome c (12.4 kDa)

shown in black filled circles.

55  

1000 2000 3000

First Fraction Second Fraction

Magnetic Field (G)

1400 1500 1600

F irs tF ra c tio n

S e co n dF ra c tio n

R am an S h ift (cm -1)

1 4 8 1 1 4 8 9

3

A B

Further analysis was done to determine whether formation of the species in either one of the two

peaks is equilibrium driven. Thus, immediately following the initial chromatographic separation,

fractions from each peak were pooled and loaded independently onto the same Superdex G-75

column. In each case there was no evidence of interconversion since fractions from each

independent run eluted in a single band. Clearly, the two species are stable and once they are

formed they do not interconvert. This observation is handy in a sense that the species in solution

can easily be identified by its electronic absorption signature. Denaturing polyacrylamide gel

electrophoresis indicated that the protein from both peaks migrates at the same rate as the apo-

protein (Figure 11) , implying that the observed difference in elution volumes from the Sephadex

Figure 10. (A) Resonance Raman spectra (obtained at 105 K) of the first and second fractions

of H32A-holo-HasAp (λexc = 413 nm, 20 mW; the spectra were normalized on the intensity of

the 4 observed at 1375 cm-1). (B) X-band EPR spectra of the first and second fractions of

H32A-holo-HasAp obtained at 10 K (0.03 mW microwave power, 9.66 GHz microwave

frequency; 10 G modulation amplitude). The two spectra were normalized based on their

hemin content determined by the pyridine hemochrome assay.

56  

G-75 column are not due to degradation or covalent modification but likely an indication of

different stable oligomeric states. In pursuit of this notion, pooled fractions from each peak were

loaded independently onto a calibrated, high resolution Superdex 75 column. Protein obtained

from the fast-eluting peak exhibited an elution volume (Ve) of 13.2 ± 0.2 mL (red triangle in

Figure 9-C), which corresponds to an estimated molecular mass of ∼38.3 ± 2 kDa (dimer),

whereas protein from the slow-eluting peak exhibited a Ve of 14.6 ± 0.1 mL (green diamond),

which corresponds to ∼24.3 ± 0.8 kDa. The elution volume of the latter is close to that of H32A-

apo-HasAp (Ve = 14.3 ± 0.1 mL, blue triangle), corresponding to an estimated molecular mass of

∼26.7 ± 0.4 kDa (monomer). Thus based on the elution volume and the estimated molecular

mass, the fast eluting peak is a dimer and it is only formed when the apo-sample is loaded with

heme. On the contrary the slow eluting peak is a monomer. Analysis of heme content revealed a

1:1 heme to protein ratio in both the monomeric and dimeric species. The nature of this dimer

and the coordination of the heme were revealed by X-ray crystallography.

 

1 32

Figure 11. 15 % SDS-PAGE of HasAp

pure samples; M- protein ladder, 1- slow

eluting band (monomer), 2- fast eluting

band (dimer), and 3-apo-HasAp.

57  

Structural Characterization of H32A-holo-HasAp by X-ray crystallography

Crystals of H32A-holo-HasAp were grown from the monomeric fraction of heme

reconstituted sample. Unexpectedly, the crystal structure obtained reveal a noncrystallographic

dimer linked by cofacial interaction of two heme molecules (Figure 12-A). The two heme planes,

defined by their corresponding pyrrole-nitrogen planes, intersect at an angle of 3.6° and are

separated by an average distance of ∼3.4 Å. Moreover, there is roughly 60% overlap in the

tetrapyrrole planes and the two central iron atoms are only 5.51 Å apart. Close inspection of the

heme molecules show that there is no electron density between the two heme planes that could

be attributed to a bridging ligand (Figure 12-C), as would be the case if the dimer was linked by

a μ-oxo ligand. Furthermore, the approximately 60% macrocycle overlap is in contrast to the

nearly complete overlap in a μ-oxo dimer analogue [47]. Another key feature in the two heme

molecules is that the central metal iron in subunits A and B are displaced by 0.21 and 0.17 Å,

respectively with respect to the mean plane defined by the coordinating pyrrole-nitrogen atoms,

toward the coordinating Tyr. This observation is consistent with a 5-coordinate iron, as observed

in catalase [48] and a H93Y mutant of myoglobin [49].

58  

**

**

A

C

*

*

R129

Y138

B

Figure 12. (A) Structure of the H32A holo-HasAp dimer. Subunits A is colored light pink and

B is colored slate. Disordered residues in subunit B are located between the asterisks. Heme in

subunit A and B is colored pink and blue, respectively. (B) Proximal site of the heme in subunit

A showing Tyr 75 coordination, and H-bonding interactions involving the propionates and Arg

129 from the same subunit and Tyr 138 from subunit B. Tyr75 loop is colored wheat. (C) 2Fo –

Fc electron density maps of the hemin molecules in the dimer contoured at 1σ represented as a

green surface.

59  

Heme coordination at the proximal site is very similar to that of wild type protein where

Tyr 75 coordinates the heme while His 83 side chain is poised to accept a hydrogen bond from

the coordinating ηO of Tyr 75. Similarly, the propionates engage in hydrogen bonding

interactions with Arg 129 and Tyr 138, the only difference being Tyr 138 is from the other

subunit in the dimer (Figure 12-B). Conservation of the propionate-Arg 129 hydrogen bond in

each of the subunits of the dimer may be attributed to in-plane orientation of the heme which

places one of the propionates in close proximity to Arg 129 located in the loop connecting 7 and

2 as seen in the wild type holo-HasAp.

Superimposed crystal structures of H32A-holo-HasAp (subunit A), wild type apo- and

holo-HasAp are shown in Figure 13-A. A striking observation is the fact that His 32 loop in the

crystal structure of heme loaded H32A assumes an open conformation nearly identical to that of

the apo-protein, exposing the distal face of the heme to the aqueous media. This directly suggests

A B

Figure 13. (A) Superimposed crystal structures showing H32 loops of apo wt (magenta), holo wt

(cyan), and holo H32A (red) HasAp. The heme molecules associated with holo wt and holo-H32A are

colored blue and gold, respectively (B) Spherical representation of subunit A and associated heme

(orange). The H32 loop is colored red, and hydrophobic residues Phe 51, Tyr 75, Leu 77, Phe 78, and

His 83 in the heme binding site are colored green.

60  

that formation of dimers in holo H32A mutant (Figure 12-A) may be a consequence of the need

to satisfy the hydrophobic demand of exposed hydrophobic surface of the heme, which, in the

aqueous media can only be satisfied by favorable interactions between two heme molecules.

Intriguingly, this phenomenon is not observed in wild type protein in which both of the heme

axial ligands are present. This issue will be revisited in a moment.

Figure 13-B is a spherical representation of subunit A of H32A dimer; as shown,

underneath the heme (orange) is a cluster of hydrophobic residues Phe 51, Tyr 75, Leu 77, Phe

78, and His 83 (green) that interact hydrophobically with the macrocycle. Thus, besides the

stability imparted by propionate interactions with the protein, the heme is stabilized and

anchored in place by this hydrophobic table.

To establish why dimeric H32A-holo-HasAp crystals were obtained from solutions

containing monomeric protein, all the crystal growth screening conditions that yielded crystals

were checked to establish whether there was a common denominator. Certain enough,

ammonium sulfate was present in all solutions. Nonetheless, it was not clear whether ammonium

sulfate induced dimerization prior to, or during protein crystallization. As mentioned earlier, it is

possible to distinguish monomeric and dimeric species of H32A-holo-HasAp in solution with the

aid of UV-vis spectrophotometer. Thus to determine if ammonium sulfate promoted dimer

formation prior to protein crystallization, pure monomeric holo protein was added with constant

stirring to a buffer containing ammonium sulfate at a concentration similar to that in the

crystallization solution (1.26 M) and the changes in absorption spectrum were followed

spectrophotometrically. In a few seconds, the electronic absorption signature of the monomeric

species (red) changed to that corresponding to a dimer (black) (Figure 14). These observations

61  

clearly demonstrated that ammonium sulfate in the crystallization solution indeed promoted

dimerization of the protein, prior to its crystallization.

High concentrations of antichaotrophic salts are known to favor hydrophobic interactions [50].

Cutler (1995) observed that Green Fluorescent Protein (GFP) monomers interact at hydrophobic

surfaces when they form dimers in aqueous solutions particularly in the presence of high

concentration of ammonium sulfate (1M) [50]. Thus in a similar manner, high concentration of

ammonium sulfate in the crystallization solution of H32A-holo-HasAp favors dimer formation

by promoting hydrophobic interactions between protein-bound heme molecules in the

monomers. Consistent with the relevance of this salt for H32A dimerization and subsequent

crystallization, additional experiments aimed at finding conditions to crystallize the monomer

using other salts were unsuccessful. However, since to this end it is not possible to

unequivocally tell the position of the His 32 loop from the crystal structure of the dimer, we

opted to introduce a ligand in solution (prior to crystallization) that would bind the heme at the

sixth position eliminatie cofacial interaction of the heme molecules by steric exclusion. Among

Figure 14. UV-Vis spectra of monomeric holo-H32A-HasAp (red) and induced dimer (black).

62  

the ligands tested, imidazole and cyanide were potential candidates. The cyano-complex formed

was not stable under our experimental conditions and thus the imidazole-complex of H32A-holo-

HasAp (H32A-imidazole) was pursued.

Imidazole complex of monomeric H32A-holo-HasAp

Titrating a solution containing monomeric H32A-holo-HasAp with imidazole while

monitoring the reaction with the aid of electronic absorption spectroscopy induces a shift of the

Soret band from 401 to 407 nm (Figure 15-A). Absorbance at 407 nm is in agreement with that

observed in wild type holo-HasAp, strongly suggesting that in the complex formed, the heme

axially coordinated by Tyr 75 and imidazole. The magnitude of the dissociation constant, Kd =

4.7 mM, was obtained by fitting the binding curve (Figure 15-B) to equation 1. Based on the

dissociation constant, the complex is relatively weak and so crystal screens and solution NMR

studies were done in the presence of relatively high concentrations of the ligand (0.15 M).

X-ray crystal structure of the H32A-imidazole complex.

A noncrystallographic dimer was observed in the asymmetric unit of the H32A-imidazole

complex. The overall fold of each of the two molecules (Figure 16-A and B) is nearly identical to

that of wild type holo-HasAp. In the distal site, exogenous imidazole coordinates the heme (2.3

Å) and makes hydrogen bonding contacts with the carbonyl group of Asn 42 (2.9 Å) likely

stabilizing the complex. Further stabilization is imparted by packing interactions with the side

63  

[Imidazole] mM

0 20 40 60 80

Fra

ctio

n bo

und

0.0

0.2

0.4

0.6

0.8

1.0

1.2B

Wavelength (nm)

360 380 400 420 440

Abs

orb

ance

0.0

0.1

0.2

0.3

0.4

0.5

0.6

401

407A

[Imidazole] mM

0 20 40 60 80

Fra

ctio

n bo

und

0.0

0.2

0.4

0.6

0.8

1.0

1.2B

[Imidazole] mM

0 20 40 60 80

Fra

ctio

n bo

und

0.0

0.2

0.4

0.6

0.8

1.0

1.2B

Wavelength (nm)

360 380 400 420 440

Abs

orb

ance

0.0

0.1

0.2

0.3

0.4

0.5

0.6

401

407A

Wavelength (nm)

360 380 400 420 440

Abs

orb

ance

0.0

0.1

0.2

0.3

0.4

0.5

0.6

401

407

401

407A

chains of Val 38 (located in the His32 loop) and Met 141 (located at the C`-end of 2) both of

which keep the imidazole plane in one specific orientation (Figure 16-D). That the coordinated

imidazole is certainly not assuming more than one orientation is corroborated by the well-

defined electron density as shown in the heme and heme-bound imidazole electron density maps

in Figure 16-C. At the proximal site heme coordination is also conserved; Tyr 75 coordinates the

heme-iron (2.2Å) and the distance between N of His 83 and the phenolate oxygen of Tyr 75

(2.6Å) is maintained. Zoomed in view of superimposed structures of the wild type (cyan) and

H32A-imidazole complex (magenta) (Figure 16-E) clearly reveal that overall, the His 32 loop

assumes a “closed” conformation similar to that of wild-type protein. In particular, the

conformation of the hinge residues (Gly 28, and Gly 45), and the topological position of the

residues exhibiting discernible electron densities are nearly identical. In addition, the plane of the

Figure 15. Family of electronic absorption spectra (A) and the binding curve (B) obtained upon

mixing imidazole and H32A holo-HasAp. The complex was formed by adding fresh protein (final

concentration 5 uM) to known concentrations of imidazole buffer (0-100 mM), pH 7.0. Each data point

was recorded after 2 min incubation.

64  

A B

*

*

*

*

C D

E

bound imidazole and that of His 32 in wild type are nearly superimposable. Complementary

information in solution was obtained with the aid of NMR spectroscopy.

Figure 16. X-ray crystal structure of H32A-imidazole complex. (A) Molecule A; heme and imidazole are

colored magenta while His32 and Tyr75 loops are colored light pink. (B) Molecule B; heme and imidazole are

colored cyan while the His32 and Tyr75 loops are colored light cyan. Disordered residues are indicated

between the asterisks in both molecules. (C) 2Fo - Fc electron density maps of the heme and heme-bound

imidazole in molecule A (top) and molecule B (bottom) of the H32A-imidazole complex contoured at 1σ. (D)

Surface representation of the His 32 loop in the structure of the H32A-imidazole complex. The imidazole

molecule (magenta) is locked in place by Val 38 and Met 141 side chains, and a hydrogen bond formed with

the carbonyl oxygen of Asn 42. (E) Superposition of molecule A of the H32A-imidazole complex

(magenta) with molecule A of wt holo-HasAp (cyan). Residues in the H32 loop are colored darker.

65  

Characterization of monomeric H32A-HasAp by NMR Spectroscopy

Three-dimensional structure and the details regarding the position of His 32 loop that

could not be explicitly obtained from the high-resolution X-ray crystal structures of H32A and

H32A-imidazole complex were studied in solution by NMR. Sequential backbone amide

resonance assignments were done on uniformly isotopic labeled samples using conventional

heteronuclear 2D and 3D NMR experiments and amino acid selective labeled proteins as

employed in backbone resonance assignments in wild type holo-HasAp [13]. This allowed

assignment of 125 (~70 %) and 154 (~80%) nonproline amide cross-peaks for H32A-HasAp and

H32A-imidazole, respectively. Representative HSQC spectra obtained with samples of H32A-

HasAp and H32A-imidazole are shown in Figure 17; respective assignments are summarized in

Table 2 and Table 3 in the Appendix. Since the assignments of wild type holo-HasAp had already

been obtained, assignments of most of the residues in the mutant were greatly simplified. A

comparison of resonances assigned in H32A- HasAp, H32A-imidazole and wild type holo-

HasAp mapped on the structures is shown in Figure 18. Clearly, most of the residues in the Tyr

75 loop that are located near hemin in all the three proteins could not be assigned. As expected,

they are influenced by hemin-iron paramagnetism which enhances relaxation and causes severe

line broadening and signal degradation [51]. Assignments in other sections of the protein in the

two proteins differ appreciably. In H32A-HasAp (Figure 18-A) resonances of a few residues

constituting the His 32 loop, all residues in 2 (Val 136-Met 141) and the loop connecting 2 to

3, the turn connecting 3 and 4, and the C′-end of 1 could not be obtained.

66  

A BAA B

In H32A-imidazole most of the aforementioned residues are assigned, however a stretch of

residues (Arg 33- Asn 42) in the His 32 loop are not assigned (Figure 18-B). This differential

accessibility to resonance assignments are addressed below. Illustrations are made using HSQC

spectra collected from amino acid selective labeled samples (15N-Gly and 15N-Val).

Figure 17. 1H-15N-HSQC-TROSY spectra of H32A HasAp (A) and H32A-imidazole (B).

Spectra were taken at 32 oC using a Varian Inova spectrometer operating at 600 MHz 1H

frequency. Acquisition parameters are as follows: 1,710 (1H) x 400 (15N) complex points; 8.9

kHz (1H) and 2.3 kHz (15N) spectral widths; 8 scans per increment; d1=1s. Spectra were

processed using NMRPipe software as follows: data in the 15N dimension were linear predicted

to twice the number of points and zero-filled in each dimension to yield a data matrix of 4k x

4k points. Spectra were referenced to internal DSS using the methyl peak at 0 ppm for the 1H

dimension and indirectly using the ratio of gyromagnetic ratios for the 15N dimension.

67  

A B C

(i) Conformational exchange affects the hinge regions of H32A-HasAp and H32A-Imidazole

differently.

As mentioned earlier, the large conformational change of the His32 loop that occurs

when wild type apo-HasAp binds heme is centered about hinge residues Gly 28 (∼157° about its

φ angle) and Gly 45 (∼36° about its ψ angle). Remarkably, dynamic processes affect these hinge

regions in H32A-HasAp and in H32A-imidazole differently: Near the Gly 28 hinge, cross-peaks

were not detected in the HSQC spectrum of H32A-HasAp for Tyr 26 and Phe 27. In contrast, the

corresponding cross-peaks in the spectrum of H32A-imidazole were readily identified because

they exhibit cross-peaks with intensities similar to those for corresponding cross-peaks in the

spectrum of wild type HasAp. At the other end of the loop (near Gly 45 hinge), neither Asn 42

nor Thr 43 had detectable cross-peaks in the HSQC spectrum of H32A-HasAp or in the spectrum

of H32A-imidazole. However, Gly 44 gave rise to a cross-peak in the spectrum of uniformly

labeled H32A-HasAp but was undetectable in the spectrum of uniformly labeled H32A-

Figure 18. Resonance assignments available for H32A-HasAp (A), H32A-imidazole (B), and wt

HasAp (C) are mapped in gray on the structure of wt holo-HasAp; residues for which resonances

could not be obtained are shown in green, prolines are in blue, and the hemin is in red.

68  

imidazole acquired with standard conditions (Figure 19-A). It was possible to observe the Gly 44

amide cross-peak in H32A-imidazole only if HSQC spectra were acquired with fast repetition at

32 °C from a sample selectively labeled with 15N-Gly (Figure 19-D). Observation of this cross-

peak enabled its subsequent sequential assignment using data from the set of 3-D spectra

acquired with a sample uniformly labeled with 13C and 15N. Variable temperature studies

revealed that below 32 °C the cross-peak is undetectable (Figure 19-B and C) but appears as a

collection of low intensity cross-peaks in the spectrum obtained at 32 °C (Figure 19-D) and

becomes more intense and likely dominated by a more abundant population at 38 °C (Figure 19-

E). This temperature dependent behavior indicates conformational exchange affecting Gly-44 in

H32A-imidazole, which at the lower temperatures explored falls in the intermediate exchange

regime, where exchange broadening renders the cross-peaks undetectable. At 32 °C the rate of

exchange is accelerated, and at least three cross-peaks can be observed; increasing the

temperature to 38 °C causes the cross-peaks to migrate toward similar chemical shifts,

resembling a single broad peak, which suggests that the distinct conformations seen at 32 °C are

in rapid exchange with a thermally accessible state that becomes predominant at the higher

temperatures. A similar, albeit less severe attenuation of intensity at low temperature is observed

for the cross-peak corresponding to the hinge residue Gly 45 (Figure 19). However, attenuation

of the peak intensity at lower temperatures is less pronounced, suggesting that a lower energetic

barrier separates the alternate conformations of Gly 45 relative to Gly 44. Since Gly 45 is a

hinge residue, the degree of freedom to explore large conformational rearrangements is

expectedly restricted relative to conformational changes in Gly 44.

69  

A B

C D

E

Figure 19. HSQC spectra obtained with samples of 15N-Gly-H32A-imidazole using standard (A) and fast-

repetition (B-E) conditions. Spectra were taken at 32 °C (A and D), 15 °C (B), 25 °C (C), and 38 °C (E).

Cross-peaks are labeled according to the corresponding sequential assignment. The cross-peak identified as

Gly44 is observed only in spectra collected at 32 and 38 °C using a short relaxation delay (50 ms) and short

acquisition time (35 ms). The spectra were recorded with 100 increments in t1 (15N); 630 complex points in t2

(1H), over a spectral width of 8.9 kHz (1H) and 2.3 kHz (15N), with 256 scans per increment for all

temperatures except 15 °C, for which 512 scans were used. The data were processed using NMRPipe applying

a 90°-shifted squared sine bell apodization function in each dimension, followed by zero filling to 4k data

points in t2, and linear prediction to twice the number of points and zero filling to 4k data points in t1.

Subsequent Fourier transformation and phasing rendered matrices of 4k × 4k points

70  

(ii) Conformational exchange affects the internal regions of the His32 loop differently in H32A-

imidazole and in H32A-HasAp

The difference in effects of dynamic processes in the internal regions of His 32 loop in

H32A-HasAp and in H32A-imidazole is depicted in Figure 18; as shown, most residues in the

His32 loop of H32A-imidazole could not be assigned, whereas the similar stretch of sequence in

H32A-HasAp is nearly fully assigned. This implies that paramagnetic effect of hemin-iron

affects resonances from residues in the His32 loop in H32A-imidazole complex more than

equivalent resonances in H32A- HasAp. Key observations were made from HSQC spectra

acquired with standard and fast repetition rates from samples of H32A-HasAp and H32A-

imidazole labeled selectively with 15N-Val. In total there are 9 Val residues in the molecule three

of which are located in the His 32 loop (Val 30, Val 37, and Val 38). Val 37 and Val 38 are

located in the middle of the loop and in wild type holo HasAp they are ~8.1 and ~9.3 Å away

from hemin-iron, respectively. In H32A-HasAp all the valines were sequentially assigned from

conventional 2D and 3D NMR data and cross-peaks corresponding to Val 37 and Val 38, are

readily observed in HSQC spectra acquired with standard or with fast repetition conditions,

Figure 20-A and 20-B, respectively. In comparison, only one of two cross-peaks corresponding

to Val 37 and Val 38 in H32A-imidazole could be observed in the HSQC spectrum obtained with

standard conditions from a sample of 15N-Val H32A-imidazole (Figure 20-C). However, when

the HSQC spectrum is acquired with fast repetition conditions, two peaks are clearly observed in

spectra obtained from a sample of 15N-Val H32A-imidazole (Figure 20-D). These two cross-

peaks could not be sequentially assigned because correlations to neighboring carbon nuclei could

not be found in 3D spectra. Nonetheless, since they appear in the HSQC spectrum of the 15N-Val

71  

sample, and all the other remaining valines had already been assigned, these peaks could be

attributed to Val 37 or Val 38.

Figure 20. HSQC spectra obtained with samples of 15N-Val-H32A-HasAp (A and B) and 15N-Val-H32A-

imidazole (C and D). Spectra in (A) and(C) were acquired with standard acquisition parameters, whereas spectra

in (B) and (D) were collected with fast-repetition conditions. Dashed lines in spectra (C) and (D) highlight one

of the cross-peaks attributed to Val 37 or Val 38, which is clearly visible only in the HSQC spectrum (and

corresponding 1D trace) acquired with fast repetition conditions. Cross-peaks are labeled according to their

sequential assignment. Cross-peaks originating from backbone amides of Ala residues are present because of

isotopic scrambling, and are clearly identified by comparison with the HSQC spectrum obtained with uniformly

labeled protein.

72  

The fact that conformational exchange differentially affects identical portions of the His 32 loop

in the two proteins is also demonstrated by differences in the intensity exhibited by the cross-

peaks corresponding to Val 37 and Val 38. Although similar protein concentration was used to

collect the HSQC spectra of both proteins, Val 37 and Val 38 peaks in H32A-imidazole spectrum

are of low intensity relative to their counterparts in the spectrum of H32A HasAp. This strongly

suggests that these residues are in close proximity and likely interact with the distal face of

hemin, in good agreement with the packing interaction between Val 38 and the imidazole ring in

the crystal structure of H32A-imidazole complex (Figure 16-C). Hence, despite the missing

electron densities of some of the residues flanking the His32 loop in the crystal structure of

H32A-imidazole, the loop is likely in the closed conformation as inferred from analysis of

solution NMR studies.

Targeted Molecular Dynamics Simulations (TMD)

Molecular dynamics is a computational method that calculates the time dependent

behavior of a molecular system. MD simulations have been used to provide detailed information

on the fluctuations and conformational changes of proteins [52]. In this study, targeted

molecular dynamics simulations were used to visualize and trace a plausible path for closing of

the His32 loop in HasAp utilizing the available structures of apo and holo wild type protein.

Simulation runs are typically short: typically t ~ 103-106 MD steps, corresponding to a few

nanoseconds of real time [51]. In this study, 25 independent TMD simulations were run starting

with the open loop (apo-structure) and ending with the closed loop (holo-structure)

conformation. Each run lasted 1.2 ns. For each of the 25 independent simulations, 50 structures

73  

(“images”) were generated at equal intervals of rmsd between the apo- and holo-structures (i.e. a

total of 1250 structures). The images obtained at each rmsd for all the 25 independent

simulations were then clustered and analyzed. Very similar structural features were observed

among the clustered images, suggesting that the path that His32 loop takes, while relocating from

the apo-conformation to the holo-conformation, may be well-defined. A summary of this

plausible path showing the most prominent features of loop closing is presented in six-frames

(Figure 21 A-F). In each frame, an overlay of the starting (gray) and intermediate (cyan)

structures are shown in cartoon representation. Residues that are involved in important

interactions are highlighted in stick or CPK representation, and when appropriate the same

residues in the starting structure are also shown in stick representation for comparison. The

His32 loop is highlighted in magenta. At the start of the simulation, the peptide backbone root

mean square deviation (rmsd) between apo- and holo-HasAp is 5.73 Å. Frame 21-A (rmsd 5.42

Å); when heme loads onto the Tyr 75 loop, helix 2 tilts relocating its C`-end from the original

position. As a result, the succeeding short loop (flanking 2 and 3) moves outward breaking the

hydrogen bonds between Gly 143 (short loop) and Ser 24 (1). On the side of the -sheet wall,

2, near the hinge residue Gly 45, moves away from hemin likely to avoid steric clashes between

the side chain of Phe 46 and macrocyle. Frame 21-B (rmsd 4.87 Å); further movements of 2

and the short loop connecting it to 3 in the directions described in frame A above creates space

which allows the side chain of Phe 27 (1) to move downward and interact adeptly with the side

chains of Leu 58 (3) and Met 141 (2). These attractive hydrophobic interactions and the

outward movement of the loop connecting 2 and 3 likely provide the force that causes helix

1 to bend. This effect is transmitted to the turn connecting 3 and 4 (yellow) via hydrogen

bonding interactions between the carbonyl groups of Tyr 26 and Phe 27 (1) and the

74  

Figure 21. A possible path for closing of the H32 loop upon hemin binding; the heavy atom rmsd

between initial (apo-) and final (holo-) structures is 5.73 Å, including snapshots at rmsd 5.42 Å (A),

4.87 Å (B), 4.06 Å (C), 2.64 Å (D), 1.61 Å (E), and 0.61 Å (F). The apo- (gray) and intermediate

(cyan) structures are shown in cartoon representation. Residues involved in important interactions are

shown in either stick or sphere representation, and when appropriate, the same residues in the apo-

structure are also shown to indicate their initial positions. The His32 loop and the turn (Ser 60 to Ser

63) are highlighted in magenta and yellow, respectively. The hemin and hemin-iron are shown as red

sticks and orange spheres, respectively.

75  

hydroxyl groups of Ser 60 and Ser 63 in the turn. Frame 21-C (rmsd 4.06 Å); stooping of 1

pulls down the portion of the His32 loop near the Gly 28 hinge but leaves the stretch toward the

Gly 45 hinge intact. In turn Val 30, Val 37, and Val 38, which in the apo structure interact

hydrophobically with one another, relocate near the yellow turn and instead interact with the

backbone of the turn (Ser 60 to Ser 63). These rearrangements bring Asn 31 and Gly 35 within

hydrogen bonding distance. Presumably, these interactions maintain the loop fairly organized as

it slides over the yellow turn. Frame 20-D (rmsd 2.64) shows that as the His32 loop slides over

the yellow turn, Val 30, Val 37, and Val 38 re-establish hydrophobic interactions and

concurrently Gln 36 and Arg 33 take their turn to interact with the yellow turn. As shown also, at

this point 1 and 2 (and the loop connecting it to 3) have relaxed back nearly to their initial

positions. This partial relaxation is necessary to provide space for Gly 28 carbonyl which to this

end has adopted its closed loop conformation. Frame 21-E (rmsd 1.61 Å) shows that as Val 37

and Val 38 approach the heme, the hydrogen bond between Arg 33 and Ala 62 breaks. It is also

clear that the side chain of the axial ligand His 32 (green) is still pointing away from the heme.

Analysis of motions between frame 21-E and frame 21-F (rmsd 0.61 Å) suggests that Arg 33

explores different conformations until its guanidinium group forms a salt bridge with the

carboxyl group of Asp 29 presumably forcing the side chain of His 32 to flip toward the heme

and subsequently coordinate hemin-iron as seen in frame 21-F.

76  

Discussion

Structure of apo-HasAp and conformational changes that occur upon hemin loading

X-ray crystal structure of apo-HasAp is very similar to that of the holo form except for

the His 32 loop which relocates ~ 30 Å away relative to its position in the holo-protein (Figure

4). The original / fold and the position of the Tyr75 loop are conserved implying that the

conformation of these secondary structures is independent of the presence or absence of the

heme. Possibly, networks of hydrophobic and polar interactions between the Tyr 75 loop and

nearby residues are sufficient to constrain their conformations both in the heme-free and heme-

loaded protein. In the only other characterized hemophore HasA, similar observations were made

regarding the overall structure of apo and holo conformers. Dynamics studies revealed that in

apo-HasAs His 32 loop does not oscillate between the open and the closed conformation but

rather adopts a single and well defined conformation [14]. Solution NMR and X-ray crystal

structure of apo-HasAp does not show any evidence of the His 32 loop exploring different

conformations. Thus it is likely that repositioning and stability of this loop in the absence of the

ligand has a functional role in all the hemophores. Characterized proteins that are involved in

binding, transport and release of iron or heme, such as transferrins, hemopexin, FbpA, and HemS

consist of two domains connected by a flexible linker loop that acts as a hinge between the two

domains [53,54]. It was proposed and recently demonstrated for the heme transport

proteobacteria-protein HemS [55] that in these proteins, ligand binding induces the transition

from a flexible “open” form of the two domains to a “closed” form where the two domains

tightly associate. In comparison, HasA is a mono-domain protein and shares no structural

77  

homology with other known iron or heme-binding proteins. In addition, in place of domain

movements as seen in other heme-binding proteins, it is the loop harboring one of the heme-iron

axial ligands (His 32) that undergoes a large conformational rearrangement. It is proposed that

heme binding to HasA involve a two-step mechanism where initially, the heme interacts with Tyr

75 loop and its flanking regions, followed by a large shift of the His 32 loop to close the binding

pocket [14]. Additional observations that have been interpreted as providing support for this

proposition arise from; (i) studies in HasAs-HasR complex which show that the His 32 ligand is

not coordinated and that the His 32 loop is disordered [15] and, (ii) studies in wild type HasAs

which showed that when the holo protein is reduced with dithionite in the presence of CO, His

32 rather than Tyr 75 heme ligand is displaced by CO [16]. Although these findings suggest

lability of His 32 ligand relative to Tyr 75, conclusive evidence demonstrating that the heme is

loaded first onto Tyr 75 followed by relocation of His 32 loop is lacking. In this work, kinetic,

structural, and spectroscopic studies were conducted to provide evidence of possible

intermediates formed in the process of hemin loading in HasAp. Thus HasAp mutant in which

the nonconserved His 32 heme iron ligand is replaced by a noncoordinative residue Ala (H32A)

was constructed, purified in the apo form and used in the heme loading experiments.

Characterization of H32A-HasAp mutant by UV-vis, EPR, and RR Spectroscopy.

The apo form of the H32A mutant was determined to be monomeric in solution but a

mixture of monomers and dimers after reconstitution with heme. Gel filtration chromatography

revealed that both the monomeric and the dimeric species are not in dynamic equilibrium; that it,

once formed there is no evidence of monomers associating or dimers dissociating. Furthermore,

78  

although the heme-iron in both species is in the high-spin state, their electronic absorption

spectra are quite distinct. These observations are helpful in a sense that at any given time, the

species in solution can easily be identified from its electronic absorption spectrum. Resonance

Raman characterization also support a high-spin ferric heme configuration but the species in the

monomer and the dimer are predominantly pentacoordinate (5c) and hexacoordinate (6c),

respectively. The very intense g ≈ 6 signature exhibited by both species in EPR spectra is

consistent with S=5/2 high spin ferric heme configuration with coordination states in agreement

with that observed in RR.

Kinetic and spectroscopic studies reveal a two-step mechanism of hemin-loading by HasAp

Wild type and H32A-holo-HasAp have distinct spectroscopic signatures (Figure 22-A).

However, heme uptake is very rapid that any intermediate species could not be detected by UV-

vis spectrophotometer and requiried monitoring by rapid spectroscopic techniques. Stopped-

Flow Spectroscopy (SFS) experiments were therefore conducted in a collaborating laboratory,

supplied with purified apo samples of HasAp. SFS results reveal that heme binding to wild type

apo-HasAp proceeds in two distinct phases via formation of an intermediate species that exhibit

electronic absorption signature typical of a 5c-hs species; by contrast H32A-apo-HasAp binds

hemin in a single phase [56]. The initial phase, identical for both wild type and H32A-HasAp

(accompanied by growth of a peak at 618-619 nm) occurs within 20 ms (Figure 22-B and C).

Over time, the spectral signature of H32A-HasAp remains unchanged while that of the

intermediate species observed in wild type continues to evolve (a new peak appears at 570 nm)

until the spectrum characteristic of holo-HasAp is attained in approximately 1 s (Figure 22-B) .

79  

Since the spectral signatures of the two proteins are very similar in the fast phase, it was

suggested that hemin-iron coordinates Tyr 75 as the initial step. The second (slow) phase

observed only in wild type HasAp is unambiguously attributed to closing of the His32 loop and

coordination of the heme-iron since this phase is only observed in wild type protein in which His

32 distal ligand is present. The spectral features of the final hemin loaded species are in

agreement with those observed by conventional UV-vis clearly spectroscopy demonstrating that,

as hypothesized, heme loads onto the protein in two phases, the initial step being loading of the

heme onto the Tyr 75 loop followed by the large conformational change of the His 32 loop to

close the binding site. It is therefore likely that the His 32 loop adopts an open conformation in

the apo protein by necessity to provide an accessible binding site for the heme to “latch” onto;

once the heme is bound it relocates to shield the hydrophobic surface from the aqueous media.

Similar conformational changes involving functional loops have also been observed in diverse

enzymes including β1, 4-galactosyltransferase-I (4Gal-T1), enolase, lipase and HIV protease

[57]. In the case of HIV protease, an intrinsically flexible flap region acts as a gate for ligand

entry and exit [58]; in fungal lipase, the movement of a helical loop (~ 8 Å) opens the solvent

[59] while in yeast enolase, one of three functional loops approach the active site located nearly

8 Å away bringing functional residues into position for catalysis [60,61]. The largest

conformational change reported is that of a 21 amino acid residue long loop in 4Gal-T1 which

undergoes approximately 20 Å rearrangement upon binding to UDP-galactose (UDP-gal) and a-

lactalbumin [62,63]. The loop relocation creates binding sites for the oligosaccharide and -

lactalbumin. Thus, it can be concluded that regardless of the protein function, sequence or

overall tertiary structure, loop movements serve to allow binding or release of substrates or

ligands. This however does not exclude other possibilities.

80  

A CB

enzyme`s hydrophobic active site which is otherwise effectively shielded from the aqueous

Structural characterization of H32A-HasAp by X-ray crystallography and NMR spectroscopy

Despite direct evidence from absorption spectroscopic studies which show that heme

loads onto HasAp in two discrete sequential steps, complementary studies using X-ray

crystallography and NMR spectroscopy were deemed necessary to determine the coordination

state of hemin-iron in H32A-HasAp, and establish the position of the His 32 loop in solution,

respectively. Crystals of H32A-HasAp grown from a solution of isolated monomer unexpectedly

revealed a dimer associated via two stacked heme molecules (Figure 11); furthermore, the

His 32 loop is in the open conformation similar to that in apo-HasAp (Figure 13-B). As shown in

the kinetic studies, heme rapidly loads onto the Tyr 75 loop followed by a relatively slow

relocation of the His 32 loop and His 32 coordination. Presumably, the closed conformation

adopted by the His 32 loop in wild type not only serves to allow coordination of hemin by the

Figure 22. (A) Absorption spectra of wild type HasAp (black), H32A variant (red), and

hemin (blue). Stopped-flow absorption traces obtained with apo-wt (B) and apo-H32A (C) upon

mixing with hemin.

81  

distal ligand but also shields the active site from the aqueous solvent. In this respect, it is likely

that in the absence of the distal site coordinating residue, either the His 32 loop movement is

slower compared to wild type or actually assumes an open conformation as revealed in the

crystal structure of dimeric H32A. In either case, the propensity of the protein to dimarize via

dihemin stacking is increased. The same argument may apply for the formation of dimeric

species observed during heme reconstitution in the mutant. Similar - stacking of hemin

molecules has been reported in the nitrate scavenger cytochrome c NapB, from Haemophilus

influenza [64], the heme transport protein chaN, in Campylobacter jejuni [65] and more recently

in the periplasmic heme-binding protein HmuT, from Yersinia pestis [66]. Since His 32 is not

conserved among HasA hemophores, it is likely that dimerization may be physiologically

relevant in other bacterial species.

To this end, the proximal ligand to heme-iron in the mutant is overtly discernible;

however, the conformation of the His 32 loop could be unequivocally ascertained based on the

nature of the H32A dimer. As pointed out earlier, crystals were obtained only from crystallization

conditions containing high concentration of ammonium sulfate. Extensive attempts to crystalize

the mutant in the absence of ammonium salt were unsuccessful. In order to determine the

topological position of the His 32 loop therefore, the imidazole complex was crystallized. Its

crystal structure reveals that the His 32 loop is in close proximity to the distal face of the heme

since discernible parts of the loop assume similar position as that of wild type holo-HasAp. It can

be deduced from this observation that (i) besides preventing diheme pi-pi stacking by steric

exclusion, bound imidazole imparts stability to the His32 loop albeit to a lesser extent through its

interactions with residues located in the same loop (Figure 16-C); (ii) in solution, the

conformation of the distal ligand bearing loop, once the heme loads onto the protein is

82  

independent of the presence or absence of heme-iron distal ligand. Solution NMR studies

provide further convincing evidence that the His32 loop in the imidazole complex is in close

proximity to hemin. First, the extent of loop resonance assignments in the presence and absence

of imidazole differ significantly. Whereas most of the loop resonances in the imidazole complex

spectrum are broadened beyond detection, corresponding resonances can be detected and

sequentially assigned in the spectrum of H32A as was the case in the spectrum of wild type

protein [13]. Secondly, the protein bound imidazole alters the dynamic properties of residues in

the His 32 loop, notably those of Val 37 and Val 38. These observations indicate that in solution

the His 32 loop is close to the distal face of the heme and likely interacts with nearby residues

altering the time scale of the motions affecting the loop. Such a close proximity to the heme and

the implied interactions between hemin and the loop residues is also observed in the crystal

structure of the imidazole complex where the side chain of Val 38 packs against the imidazole

ring (Figure 16-C).

Relocation of His32 loop proceed via a well-defined path

The proximal ligand partners Tyr 75 and His 83 and a network of hydrophobic residues in

the Tyr 75 provide packing interactions that stabilize the macrocylcle irrespective of the presence

or absence of the distal heme iron ligand. It has been proposed [14] and implied by the

preceding discussion that heme loading onto the Tyr 75 loop triggers conformational

rearrangement of the loop harboring the distal heme-iron ligand. As observed in sequential

assignment coverage in H32A, imidazole complex and wild type (Figure 16-A), resonances of

certain portions of the protein (C′- terminal of 1 prior to the hinge residue Gly 28, the entirety

83  

of 2, the short loop connecting 2 to 3, 2 at the end of Gly 45 hinge, and the turn connecting

3 and 4) were difficult to assign due to peak broadening caused by dynamic processes

affecting these sections of the structure. The same regions were independently implicated by

targeted molecular dynamics simulations as undergoing motions important to the closing of the

His 32 bearing loop in the wild type protein (Figure 21). Taken together, these observations

suggest that His 32 loop transitions from the “open” to the “closed” conformation via a well-

defined path. Assenting information provided by NMR and TMD simulations regarding

dynamic properties in common sections of the structure can be utilized in future investigations

aimed at elucidating the mechanism of His 32 loop closing.

84  

Appendices

Appendix I: Protein crystallization conditions-Wizard II Random Sparse Matrix

Kit# crystallant buffer (0.1 M) salt (0.2 M)

1 10% (w/v) PEG-3000 acetate pH 4.5 Zn(OAc)2

2 35% (v/v) 2-methyl-2,4-pentanediol MES pH 6.0 Li2SO4

3 20% (w/v) PEG-8000 Tris pH 8.5 MgCl2

4 2.0 M (NH4)2SO4 cacodylate pH 6.5 NaCl

5 20% (v/v) 1,4-butanediol HEPES pH 7.5 NaCl

6 10% (v/v) 2-propanol phosphate-citrate pH 4.2 Li2SO4

7 30% (w/v) PEG-3000 Tris pH 7.0 NaCl

8 10% (w/v) PEG-8000 Na/K phosphate pH 6.2 NaCl

9 2.0 M (NH4)2SO4 phosphate-citrate pH 4.2 none

10 1.0 M (NH4)2HPO4 Tris pH 8.5 none

11 10% (v/v) 2-propanol cacodylate pH 6.5 Zn(OAc)2

12 30% (v/v) PEG-400 cacodylate pH 6.5 Li2SO4

13 15% (v/v) ethanol citrate pH 5.5 Li2SO4

14 20% (w/v) PEG-1000 Na/K phosphate pH 6.2 NaCl

15 1.26 M (NH4)2SO4 HEPES pH 7.5 none

16 1.0 M sodium citrate CHES pH 9.5 none

17 2.5 M NaCl Tris pH 7.0 MgCl2

85  

18 20% (w/v) PEG-3000 Tris pH 7.0 Ca(OAc)2

19 1.6 M NaH2PO4/0.4 M K2HPO4 phosphate-citrate pH 4.2 none

20 15% (v/v) ethanol MES pH 6.0 Zn(OAc)2

21 35% (v/v) 2-methyl-2,4-pentanediol acetate pH 4.5 none

22 10% (v/v) 2-propanol imidazole pH 8.0 none

23 15% (v/v) ethanol HEPES pH 7.5 MgCl2

24 30% (w/v) PEG-8000 imidazole pH 8.0 NaCl

25 35% (v/v) 2-methyl-2,4-pentanediol HEPES pH 7.5 NaCl

26 30% (v/v) PEG-400 CHES pH 9.5 none

27 10% (w/v) PEG-3000 cacodylate pH 6.5 MgCl2

28 20% (w/v) PEG-8000 MES pH 6.0 Ca(OAc)2

29 1.26 M (NH4)2SO4 CHES pH 9.5 NaCl

30 20% (v/v) 1,4-butanediol imidazole pH 8.0 Zn(OAc)2

31 1.0 M sodium citrate Tris pH 7.0 NaCl

32 20% (w/v) PEG-1000 Tris pH 8.5 none

33 1.0 M (NH4)2HPO4 citrate pH 5.5 NaCl

34 10% (w/v) PEG-8000 imidazole pH 8.0 none

35 0.8 M NaH2PO4/1.2 M K2HPO4 acetate pH 4.5 none

36 10% (w/v) PEG-3000 phosphate-citrate pH 4.2 NaCl

37 1.0 M K/Na tartrate Tris pH 7.0 Li2SO4

38 2.5 M NaCl acetate pH 4.5 Li2SO4

86  

39 20% (w/v) PEG-8000 CAPS pH 10.5 NaCl

40 20% (w/v) PEG-3000 imidazole pH 8.0 Zn(OAc)2

41 2.0 M (NH4)2SO4 Tris pH 7.0 Li2SO4

42 30% (v/v) PEG-400 HEPES pH 7.5 NaCl

43 10% (w/v) PEG-8000 Tris pH 7.0 MgCl2

44 20% (w/v) PEG-1000 cacodylate pH 6.5 MgCl2

45 1.26 M (NH4)2SO4 MES pH 6.0 none

46 1.0 M (NH4)2HPO4 imidazole pH 8.0 NaCl

47 2.5 M NaCl imidazole pH 8.0 Zn(OAc)2

48 1.0 M K/Na tartrate MES pH 6.0 none

All formulations are made with ultrapure ASTM Type I water and sterile-filtered stock solutions.

Store at 4-25 ºC.

87  

Appendix II

Table 2. Backbone resonance assignments for H32A HasAp.

Residue 1HN 15NH 13C 13C 13CO

M1 S2 I3 S4 I5 S6 Y7 S8 T9 T10 Y11 S12 G13 W14 T15 V16 A17 D18 Y19 L20 A21 D22 W23 S24 A25 Y26 F27 G28 D29 V30 N31 A32 R33 P34 G35 Q36 V37 V38 D39 G40 S41

8.77 8.92 7.98

8.74 8.80 7.33 8.55 8.44 8.83 8.82 8.18 7.81 8.80 8.90 8.42

7.23 7.40 7.34 7.93 8.06 7.30

8.56 7.92 8.34 8.31 10.58 9.30 8.18

127.30 124.40 120.40

122.98 109.13 120.12 119.85 119.81 122.61 119.40 119.96 122.57 120.34 114.55 131.27

102.32 119.34 127.85 115.53 121.62 115.87

112.23 118.77 117.07 124.95 131.55 112.58 116.97

56.44 60.06 56.38 54.73

57.10 60.10 66.53 55.48 57.45 63.36 58.40 55.29 58.29 63.72 63.63

45.68 54.36 63.71 54.66 54.34 53.34 63.74 45.38 55.04 60.09 61.80

45.01 58.27

65.04 42.62 65.10 41.21

29.21 70.98 30.92 18.63 39.60 38.76 40.89 18.28 44.79 28.44

38.16 18.83

31.83

29.77 32.67 30.59

64.09

174.29 175.08 172.76

174.97 176.23 175.69 178.62 180.24 180.24 176.43 179.23 180.26 177.61 178.10

177.00 172.94 174.82 174.57 175.54

178.17

176.86 174.91 174.88

174.15

88  

N42 T43 G44 G45 F46 N47 P48 G49 P50 F51 D52 G53 S54 Q55 Y56 A57 L58 K59 S60 T61 A62 S63 D64 A65 A66 F67 I68 A69 G70 G71 D72 L73 H74 Y75 T76 L77 F78 S79 N80 P81 S82 H83 T84 L85 W86 G87

8.00 8.26

8.12

6.67 8.69 8.96

7.22

8.10 7.22

7.28 9.15

9.27 10.51 9.72 8.88 9.58 8.53

7.70 7.80

8.32

109.91 106.58

109.57

113.61 106.65 114.12

129.16

125.98 111.90

123.52 121.21

115.65 135.70 111.45 111.94 129.28 123.12

120.08 117.56

105.77

60.85 44.68 45.05

44.51

52.00 47.68 58.27

52.97 54.15

61.59 51.87 56.09 53.98 53.35 51.54 55.58 61.78 50.79 43.86 46.56 54.59 54.56

65.67 58.29

55.23 43.20

45.75

38.88 19.93

19.99 64.86 40.20 18.69 25.10 45.40 42.06 21.20

41.63 46.93

69.60 38.15

172.13 178.86

170.87 174.92 171.48 176.89

173.56 182.28

174.73 175.37 171.85 175.05 176.01

173.18 176.39 175.55 173.76 171.65 173.69 176.12

174.58 175.33

172.23 172.21

89  

K88 L89 D90 S91 I92 A93 L94 G95 D96 T97 L98 T99

G100 G101 A102 S103 S104 G105 G106 Y107 A108 L109 D110 S111 Q112 E113 V114 S115 F116 S117 N118 L119 G120 L121 D122 S123 P124 I125 A126 Q127 G128 R129 D130 G131 T132 V133

8.46 8.93 9.53 8.79 9.24 9.22 9.01 8.63 8.92

9.24 8.44 6.94 8.83

7.57 8.07 7.82 9.60 8.84 8.36 9.25 7.84 9.02 8.64 7.36 8.73 9.45 9.57 8.40 8.98 8.80 7.39 8.57 9.29

7.44 6.82 6.79 7.75

7.62 7.84 8.72 8.88

120.84 126.44 125.02 111.47 120.78 131.06 121.46 108.42 127.27

116.93 107.24 106.75 123.75

114.92 109.18 107.67 121.17 124.03 120.19 125.09 110.52 127.87 129.88 115.82 117.80 124.19 118.52 118.59 115.13 102.79 118.59 125.14 122.93

124.91 118.23 113.70 106.92

114.84 107.77 120.28 122.40

54.87 54.00 55.18 56.99 59.06 50.82 53.76 47.67 54.27

54.47 60.00 45.14 43.92 55.42 59.77 57.85 45.58 44.23 59.29 51.53 52.99 57.06 55.86 56.55 58.67 60.45 56.38 56.18 57.73 53.92 57.37 57.39 54.77 53.99 55.96 61.69 64.39 53.66 56.64 46.15 57.13 54.72 45.83 67.26 69.40

34.84 46.34 44.57 66.66 41.99 21.51 44.20

71.29

19.00

63.79

41.11 22.33 43.15 42.13 63.18 29.60 28.73 37.73 65.84 41.69 65.58 37.30 41.90

42.87 43.35 66.68 32.53 36.21 18.63 29.29

29.72 41.99

69.19 31.54

175.13 173.05 176.88 173.80 173.12 175.24 177.08 170.57

176.36 173.60 173.48 173.93

175.10 174.64 174.36 174.64 175.78 175.04 177.27 177.11 173.85 175.07 174.93 173.81 172.73 174.16 173.00 175.87 178.96 175.10 176.27 175.87 172.35 177.32 176.08 178.90 177.46

174.93 177.24 173.34 175.48 176.86

90  

H134 K135 V136 V137 Y138 G139 L140 M141 S142 G143 D144 S145 S146 A147 L148 Q149 G150 Q151 I152 D153 A154 L155 L156 K157 A158 V159 D160 P161 S162 L163 S164 I165 N166 S167 T168 F169 D170 Q171 L172 A173 A174 A175 G176 V177 A178 H179

7.59

8.64 8.90 9.05 8.32 9.21 8.00 7.67 8.58 8.33 7.26 8.17 8.59 7.96 7.34 7.56 7.77 8.95

8.23 8.29 9.01 8.49 7.96 8.42

8.12 8.40 9.01 8.10 7.43 7.88 7.68 6.96 8.71

112.89

118.67 121.56 127.37 120.05 118.56 103.72 119.50 118.63 118.50 117.31 119.58 120.65 118.63 118.33 120.17 118.82 115.04

124.49 109.75 113.52 116.55 119.47 111.36

121.77 120.76 126.49 124.03 117.28 105.44 113.98 119.68 116.32

52.97 57.86 63.31 55.67 58.51 60.80 47.50 57.59 63.98 59.10 55.34 57.98 58.10 58.52 53.70 65.58 52.43 64.53 59.90 53.97 58.29 62.41 53.65 60.24 59.99

59.19 57.50 54.92 54.91 52.35 45.70 62.16 49.07 55.26

40.96 63.97 62.85 17.69 40.87 29.10

27.72 36.33 42.63 18.39 42.76 41.39 32.78 18.51 31.79 44.00 32.13 63.99 43.30 67.54 38.66 39.07 64.24 71.34

28.07 41.39 18.61 17.69 20.19

33.08 24.47 32.29

176.00 175.13 177.45 181.35 177.48 177.65 177.11 178.05 177.43 177.28 181.24 180.24 178.52 178.43 179.48 176.70 174.02 176.92 174.57 174.85 175.27 174.36 174.90 173.37 176.96

180.18 180.27 179.44 182.20 179.02 177.67 174.55 173.70 175.35 172.54

91  

A180 T181 P182 A183 A184

8.45 8.56

8.26 7.89

127.62 117.87

126.00 129.03

50.84 59.14 63.16 52.53 53.86

19.59 69.26 32.29 19.64 20.12

178.66 172.50 176.73 176.43 182.42

92  

Table 3. Backbone resonance assignments for H32A-Imidazole.

Residue 1HN 15NH 13C 13C 13CO

M1 S2 I3 S4 I5 S6 Y7 S8 T9 T10 Y11 S12 G13 W14 T15 V16 A17 D18 Y19 L20 A21 D22 W23 S24 A25 Y26 F27 G28 D29 V30 N31 A32 R33 P34 G35 Q36 V37 V38 D39 G40 S41 N42 T43

8.59 8.04 8.55 8.80 7.92 8.17 8.44 7.68 7.82 7.60

8.57 8.81 7.33 8.50 8.22 8.89 8.81 8.38 8.01 9.10 8.51 7.45 7.31 8.25 7.94 7.48 7.85 7.89 6.91

122.15 118.68 127.14 124.41 119.97 115.59 118.63 120.10 117.44 115.66

122.90 109.47 120.79 119.99 118.35 121.57 119.24 122.04 120.69 120.63 114.21 124.39 120.21 123.60 107.18 116.56 120.20 119.79 128.51

57.46 61.53 56.51 59.98 56.39 54.88 56.88 65.47 65.74 58.71 63.08 44.96 57.35 60.07 66.59 55.62 57.41 63.51 58.23 55.35 58.08 63.31 63.28 54.62 58.16 61.20 44.75 52.37 62.35 52.81 53.97

61.48

64.26 38.90 65.19 43.00 65.37 41.49 64.09 68.87 69.54 38.28 64.12

29.23 70.88 31.02 19.04 39.93 39.17 41.65 18.58 44.36 28.84 63.46 18.24 38.71 38.90

42.03 31.41 39.24 19.97

170.61 171.98 174.24 175.01 172.69 172.50 177.65 176.29 174.69 175.02

175.06 176.26 175.62 178.51 179.61 180.24 176.63 179.70 180.26 178.54 177.95 176.52 178.84 177.87 176.83 171.46 177.68 175.66 173.57 182.08

93  

G44 G45 F46 N47 P48 G49 P50 F51 D52 G53 S54 Q55 Y56 A57 L58 K59 S60 T61 A62 S63 D64 A65 A66 F67 I68 A69 G70 G71 D72 L73 H74 Y75 T76 L77 F78 S79 N80 P81 S82 H83 T84 L85 W86 G87 K88 L89

7.78 8.36 9.60 9.03

8.01

8.93 6.66 8.66 8.90 8.16 9.48 8.23 7.03 7.70 8.08 7.74 8.18 7.43 7.83 7.39 9.13 8.19 8.72 10.45 9.49 8.68 9.56 8.48 8.10

5.95 6.54 7.48

7.95

9.46 8.52 8.46 8.81

110.38 107.09 117.15 116.28

105.30

117.20 113.50 106.79 114.04 117.27 122.75 129.48 120.96 124.23 117.31 115.43 125.05 114.75 114.69 123.54 120.55 113.25 115.07 135.52 110.80 111.30 129.46 123.20 118.48

119.75 118.00 117.73

124.32

125.70 106.17 120.76 126.50

45.75 46.26 57.45 49.26 65.55 46.42 64.70 56.56 52.01 47.65 58.16 54.69 57.61 49.72 54.20 53.50 57.40 62.44 52.37 56.53 54.68 52.86 52.03 55.32 60.71 50.51 43.60 46.55 54.51 54.53 55.15

56.40 57.53 51.33 61.30 57.96

54.26 55.30 43.23 54.78 53.98

44.63 43.19 34.61

34.70 40.70 46.06

64.12 37.44 42.72 24.30 45.53 35.23 63.03 68.82 21.18 66.10 41.43 20.15 25.35 44.36 42.55 21.23

41.76 47.30 31.58

38.52 62.81 39.25 32.60 63.19

46.78 30.53

34.95 46.46

173.94 170.96 176.53 175.31 178.32 171.55 177.75 171.06 174.93 171.35 176.93 175.45 172.29 174.26 172.97 174.71 174.63 174.79 176.23 172.23 175.55 175.47 174.51 172.81 176.30 175.25 173.92 171.67 173.80 176.16 173.96

170.30 173.64 176.28 172.97

174.46 172.26 172.24 175.13 173.01

94  

D90 S91 I92 A93 L94 G95 D96 T97 L98 T99

G100 G101 A102 S103 S104 G105 G106 Y107 A108 L109 D110 S111 Q112 E113 V114 S115 F116 S117 N118 L119 G120 L121 D122 S123 P124 I125 A126 Q127 G128 R129 D130 G131 T132 V133 H134 K135

9.51 8.66 9.00 8.98 8.77 8.20 8.89 8.32 8.30 9.26 8.30 6.75 8.70

7.38 7.89 7.64 9.31 8.75 8.26 9.26 7.72 8.77 8.83 7.36 8.58 9.20 9.41 8.25 8.92 8.70 7.39 8.50 9.29

7.58 7.00 6.75 7.98 8.64 7.58 7.75 8.60 8.76 8.57 8.73

125.09 111.75 120.16 130.97 120.37 108.85 123.84 111.14 128.30 120.53 106.35 106.31 123.68

114.71 108.91 107.31 121.48 123.78 119.71 124.70 111.69 126.84 128.25 116.39 117.56 122.80 117.84 117.94 115.34 102.56 118.51 124.61 122.45

124.65 118.71 113.31 107.42 122.75 114.79 107.81 120.00 122.39 120.03 116.30

55.08 56.90 59.52 50.87 53.74 47.33 53.73 64.77 55.19 60.11 45.00 43.59 55.25 59.83 57.77 45.46 44.26 58.69 51.71 52.81 56.69 55.74 56.67 58.42 60.21 56.34 56.16 57.69 53.82 57.24 47.42 54.67 54.00 56.07 61.76 64.61 53.91 56.62 46.30 57.08 54.61 45.68 67.25 69.27 62.65 61.29

45.06 66.82 42.53 21.72 43.61

45.61 65.35 43.42 72.11

19.42 63.31 63.94

41.33 22.49 43.41 41.82 63.19 30.11 30.89 37.92 66.08 42.11 65.70 37.35 42.15

43.22 43.94 66.94 32.79 36.42 19.17 29.60

29.83 42.09

69.16 31.83 31.63 33.29

176.78 173.63 172.97 175.00 176.88 171.07 174.29 172.64 176.43 173.65 173.33 173.88 180.04 174.96 174.53 174.24 174.43 175.81 174.75 177.05 177.48 173.37 175.23 175.26 173.71 172.58 174.21 172.97 175.74 178.80 175.04 176.37 175.91 172.41 177.54 176.41 179.03 177.48 175.20 174.89 177.12 173.30 175.39 176.62 178.49 181.54

95  

V136 V137 Y138 G139 L140 M141 S142 G143 D144 S145 S146 A147 L148 Q149 G150 Q151 I152 D153 A154 L155 L156 K157 A158 V159 D160 P161 S162 L163 S164 I165 N166 S167 T168 F169 D170 Q171 L172 A173 A174 A175 G176 V177 A178 H179 A180 T181

9.04 8.30 10.33 8.12 7.17 7.84 7.11 7.61 8.37 8.86 8.72 9.03 8.26 8.97 8.06 7.56 8.42 8.40 7.17 7.93 8.55 7.94 7.25 7.47 7.76

8.88 8.11 8.15 9.02 8.39 7.87 8.38

8.05 8.28 8.85 8.01 7.32 7.81 7.57 6.88 8.61 8.43 8.52

118.54 119.46 122.80 106.09 122.86 115.15 110.53 110.10 119.18 119.43 121.21 126.97 120.17 117.62 103.95 119.31 118.55 118.78 117.19 119.46 120.56 118.55 118.33 119.88 118.74

114.90 124.23 109.71 114.51 116.68 119.26 111.24

121.58 120.43 126.55 123.96 117.06 105.31 113.89 119.54 116.29 127.81 117.64

66.76 67.69 62.49 48.38 58.30 59.26 57.17 46.58 52.58 58.63 63.46 55.61 58.35 60.86 47.49 57.51 64.38 59.16 55.28 57.92 58.02 58.49 53.70 65.45 52.35 64.47 59.87 53.96 58.13 62.20 53.65 60.19 59.98

57.56 59.18 57.39 54.85 54.86 52.28 45.62 62.16 48.98 55.12 50.82 59.06

31.93 32.46 40.75

41.41 33.83 63.70

42.22 64.15 62.45 17.68 41.17 29.29

27.81 36.80 42.95 18.42 43.01 41.56 33.08 18.54 32.04 44.19 32.32 64.11 43.70 67.70 39.25 39.27 64.40 71.52

41.21 28.30 41.51 18.84 17.81 20.59

33.38 24.58 32.01 19.69 69.24

178.41 177.11 179.86 176.45 179.02 175.73 173.82 173.39 175.61 175.11 177.16 181.10 177.83 177.78 177.06 178.12 177.35 177.26 181.18 180.23 178.53 178.38 179.43 176.67 174.04 176.85 174.60 174.84 175.14 174.26 174.86 173.35 176.91

180.15 180.26 179.48 182.17 178.98 177.67 174.50 173.69 175.40 172.35 178.61 172.52

96  

P182 A183 A184

8.20 7.84

125.94 129.00

63.14 52.49 53.83

32.46 19.75 20.50

176.75 176.40 182.34

97  

References

1. Ochsner, U. A., Johnson, Z., and Vasil, M. L., Genetics and regulation of two distinct

haem- uptake systems, phu and has, in Pseudomonas aeruginosa. Microbiology, 2000. 146: p. 185-198.

2. Vasil, M. L., and Ochsner, U. A., The response of pseudomonas Aeruginosa to iron:

Genetics, biochemistry and virulence. Molecular Microbiology, 1999. 34: p. 399-413. 3. Braun, V., and Killmann, H., Bacterial solutions to the iron-supply problem. Trends

Biochem Sci, 1999. 24: p. 104-109. 4. Bivelle, F., Cwerman, H., Le´toffe´, S., Rossi, M.-S., Drouet, V., Ghigo, J. M., and

Wandersman, C., Hemophore-Mediated Signalling in Serratia marcescens: A new mode of regulation for an Extra Cytoplasmic Function (ECF) Sigma Factor Involved in Haem Acquisition. Mol. Microbiol., 2004. 53: p. 1267-1277.

5. Letoffe,S., Ghigo,J.M., and Wandersman,C., Secretion of the Serratia marcescens HasA

protein by an ABC Transporter. J. Bacteriol., 1994b. 176: p. 5372-5377. 6. Le´toffe´, S., Redeker, V., and Wandersman, C., Isolation and characterization of an

extracellular haem-binding protein from Pseudomonas aeruginosa that shares function and sequence similarities with the Serratia marcescens HasA haemophore. Mol. Microbiol., 1998. 28: p. 1223-1224.

7. Létoffé, S., Ghigo, J. M., and Wandersman, C., Iron acquisition from heme and

hemoglobin by a Serratia marcescens extracellular protein. Proc. Natl. Acad. Sci. U S A., 1994. 91: p. 9876-9880.

8. Deniau, C., Gilli, R., Izadi-Pruneyre, N., Lettofe, S., Delepierre, M., Wandersman, C.,

Briand, C., and Lecroisey, A., Thermodynamics of hemin Binding to the HasAsm Hemophore: Effects of Mutations at Three Key Residues for Heme Uptake. Biochemistry, 2003. 42: p. 10627-10633.

9. Ghigo, J. M., Létoffé, S., Wandersman, C., A new type of hemophore-dependent heme acquisition system of Serratia marcescens reconstituted in Escherichia coli. J Bacteriol., 1997. 179: p. 3572-3779.

11. Arnoux, .P, Haser, R., Izadi, N., Lecroisey, A., Delepierre, M., Wandersman, C., and

Czjzek M., The crystal structure of HasAs, a hemophore cesreted by Serratia marcescens. Nat Struct Biol., 1999. 6: p. 516-520.

98  

12. Arnoux, P.; Haser, R.; Izadi-Pruneyre, N.; Lecroisey, A.; Czjek, M. Functional aspects of

the heme bound hemophore HasA by structural analysis of various crystal forms. Proteins:Struct., Funct., Genet. 2000. 41: p. 202-210.

13. Alontaga, A. Y., Rodrı´guez, J. C., Scho¨nbrunn, E., Becker, A., Funke, T., Yukl, E. T., Hayashi, T., Stobaugh, J., Moe¨nne-Loccoz, P., and Rivera, M., Structural characterization of the hemophore HasAp from Pseudomonas aeruginosa: NMR spectroscopy reveals protein-protein interactions between Holo-HasAp and hemoglobin. Biochemistry, 2009. 48: p. 96-109.

14. Wolff, N., Izadi-Pruneyre, N., Couprie, J., Habeck, M., Linge, J., Rieping, W., Wandersman, C.; Nilges, M., Delepelairre, P., and Lecroisey, A., Comparative Analysis of Structural and Dynamic Properties of the Loaded and Unloaded Hemophore HasA: Functional Implications. J. Mol. Biol., 2008. 376: p. 517-525.

15. Lukat-Rodgers, G. S., Rodgers, K. R., Caillet-Saguy, C., Izadi-Pruneyre, N., Lecroisey, A., Novel heme ligand displacement by CO in the soluble hemophore HasA and its proximal ligand mutants: implications for heme uptake and release. Biochemistry, 2008. 47: p. 2087-2098.

16. Krieg, S., Huche´, F., Diederichs, K., Izadi-Pruneyre, N., Lecroisey, A., Wandersman, C., Delepelaire, P., and Welte, W., Heme uptake across the outer membrane as revealed by crystal structures of the receptor-hemophore complex. Proc. Natl. Acad. Sci. U.S.A. 2009, 106 : p. 1045-1050.

17. Lei, B., Smoot, L. M., Menning, H. M., Voyich, J. M., Kala, S. V., DeLeo, F. R., Reid, S. D., and Musser, J. M. Identification and Characterization of a Novel Heme-Associated Cell Surface Protein Made by Streptococcus pyogenes. Infect. Immun. 2002, 2002: p. 4494-4500.

18. Aranda, R., Worley, C. E., Liu, M., Bitto, E., Cates, M. S., Olson, J. S., Lei, B., and Phillips, G. N., Bis-methionyl coordination in the crystal structure of the heme-binding domain of the streptococcal cell surface protein Shp.. J. Mol. Biol., 2007. 374: p. 374-383.

19. Nygaard, T. K., Blouin, G. C., Liu, M., Fukumura, M., Olson, J. S., Fabian, M.; Dooley, D. M., Lei, B., The mechanism of direct heme transfer from the streptococcal cell surface protein Shp to HtsA of the HtsABC transporter. J. Biol. Chem., 2006. 28: p. 20761-20771.

20. Cheng, H., Westler, W. M., Xia, B., Oh, B. -H., and Markley, J. L., Protein Expression, Selective Isotope Labeling, and Analysis of Hyperfine-Shifted NMR Signals of Anabana 7120 Vegetative [2Fe-2S] Ferredoxin. Arch..Biochem. Biophys., 1995. 316: p. 619-634.

99  

21. Rodriguez, J.C., Wilks, A., and Rivera, M., Backbone NMR Assignments and H/D

Exchange Studies on the Ferric Azide- and Cyanide-Inhibited Forms of Pseudomonas aeruginosa Heme Oxygenase. Biochemistry, 2006. 45: p. 4578-4592.

22. Cheng, H., and Markley, J. L., NMR Spectroscopic Studies of Paramagnetic Proteins:

Iron Sulfur Proteins. Annu. Rev. Biophys. Biomol. Struct., 1995. 24: p. 209-237. 23. Goodrich, J. A., and Kugel, J. F., Binding and Kinetics for Molecular Biologists; 2007:

Cold Spring Harbor, New York.

24. Otwinowski, Z., and Minor, W. In Methods in Enzymology, Macromolecular Crystallography, Part A; Carter, C. W., Sweet, R. M., Eds., 2007: Academic Press, New York.

25. Kabsch, W., Automatic indexing of rotation diffraction patterns J. Appl. Crystallogr.. 1988, 21: p. 67-72.

26. Evans, P. Scaling and assessment of data quality, Acta Crystallogr. 2006. D62: p. 72-82. 27. Vagin, A., Teplyakov, A., MOLREP: an Automated Program for Molecular

Replacement. J. Appl. Crystallogr., 1997. 30: p. 1022-1025. 28. Langer, G.; Cohen, S. X.; Lamzin, V. S.; Perrakis, A., Automated macromolecular model

building for X-ray crystallography using ARP/wARP version 7. Nat. Protoc., 2008, 3: p. 1171-1179.

29. Murshudov, G. N., Vagin, A. A., and Dodson, E. J., Refinement of Macromolecular Structures by the Maximum-Likelihood Method. Acta Crystallogr., 1997. D53: p. 240-255.

30. Adams, P. D., et al. PHENIX: a comprehensive Python-based system for macromolecular structure solution Acta Crystallogr., 2010. D66: p. 213-221.

31. Emsley, P.; Cowtan, K., Coot: model-building tools for molecular graphics. Acta

Crystallogr., 2004. D60: p. 2126-2132. 32. Painter, J., and Merritt, E. A., Optimal description of a protein structure in terms of

multiple groups undergoing TLS motion. Acta Crystallogr., 2006. D62: p. 439-450. 33. Lovell, S. C., Davis, I. W., Arendall, W. B., III, de Bakker, P. I., Word, J. M., Prisant, M.

G., Richardson, J. S., and Richardson, D. C., Structure validation by Calpha geometry: phi,psi and Cbeta deviation. Proteins, 2003. 50: p. 437-450.

100  

34. Potterton, L.,McNicholas, S., Krissinel, E., Gruber, J., Cowtan, K., Emsley, P.,

Murshudov, G. N., Cohen, S., Perrakis, A., and Noble, M., Developments in the CCP4 molecular-graphics project. Acta Crystallogr., 2004. D60: p. 2288-2294.

35. Harding, M. M., Metal-ligand geometry relevant to proteins and in proteins: sodium and potassium. Acta Crystallogr., 2002. D58: p. 872-874.

36. Delaglio, F., Grzesiek, S., Vuister, G. W., Zhu, W., Pfeifer, J., and Bax, A., NMRPipe: a

multidimensional spectral processing system based on UNIX pipes. J. Biomol. NMR, 1995. 6: p. 277-293.

37. Goddard, T. D., and Kneller, D. G., SPARKY3; University of California, San Francisco: 2004.

38. Brooks, B. R., CHARMM: The biomolecular simulation program. J. Comput. Chem., 2009. 30: p. 1545-1614.

39. MacKerell, A. D., Jr., All-Atom Empirical Potential for Molecular Modeling and

Dynamics Studies of Proteins. J. Phys. Chem. B., 1998. 102: p. 3586-3616. 40. Mackerell, A. D., Feig, M., and Brooks, C. L., Extending the treatment of backbone

energetics in protein force fields: Limitations of gas-phase quantum mechanics in reproducing protein conformational distributions in molecular dynamics simulations. J. Comput. Chem., 2004. 25: p. 1400-1415.

41. Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W., and Klein, M. L., Comparison of simple potential functions for simulating liquid water . J. Chem. Phys. 1983. 79: p. 926-935.

42. Jo, S., Kim, T., Iyer, V. G., and Im, W., CHARMM-GUI: a web-based graphical user interface for CHARMM. J. Comput. Chem., 2008: 29: p. 1859-1865.

43. Ryckaert, J. P., Ciccotti, G., and Berendsen, H. J. C., Numerical integration of the Cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys., 1977. 23: p. 327-341.

44. Ryckaert, J. P., Ciccotti, G., and Berendsen, H. J. C., Numerical integration of the Cartesian equations of motion of a system with constraints: molecular dynamics of n-alkanes. J. Comput. Phys.,1977. 23: p. 327-341.

45. Larkin, M. A., Blackshields, G., Brown, N. P., Chenna, R., McGettigan, P. R., McWilliam, H., Valentin, F., Wallace, I. M., Wilm, A., Lopez, R., Thompson, J. D., Gibson, T. J., and Higgins, D. G., Clustal W and Clustal X version 20. Bioinformatics, 2007. 23. p. 2947-2948.

101  

46. Caillet-Saguy, C., Turano, P., Piccioli, M., Lukat-Rodgers, G. S., Czjzek, M., Guigliarelli, B., Izadi-Pruneyre, N., Rodgers, K. R., Delepierre, M., and Lecroisey, A., Deciphering the Structural Role of Histidine 83 for Heme Binding in Hemophore HasA. J. Biol. Chem., 2008. 283: p. 5960-5970.

47. Cheng, L., Lee, J., Powell, D. R., and Richter-Addo, G. B., mu.-Oxo-bis[(protoporphyrin IX dimethyl ester)iron(III)]. Acta Crystallogr., 2004. E60: m1340-m1342.

48. Putnam, C. D., Arvai, A. S., Bourne, Y., and Tainer, J. A., Active and inhibited human catalase structures: Ligand and NADPH binding and catalytic mechanism. J. Mol. Biol., 2000. 296: p. 295-309.

49. Hildebrand, D. P., Burk, D. L., Maurus, R., Ferrer, J. C., Brayer, G. D., and Mauk, A. G., The proximal ligand variant His93Tyr of horse heart myoglobin. Biochemistry, 1995. 34: p. 1997-2005.

50. Chalfie, M., and kain, S. R., GFP Properties, Applications and Protocols 2nd edition, 2006. Wiley & Sons Inc, Hoboken New Jersey.

51. Scott, A. R., Applications of physical methods to inorganic and bioinorganic chemistry 52. Duan, Y., and Kollman, P. A., Pathways to a protein folding intermediate observed in a

1-microsecond simulation in aqueous solution. Science, 1998. 282: p. 740-744. 53. Wand, A. J., Dynamic activation of protein function: a view emerging from NMR

spectroscopy. Nature Struct. Biol., 2001. 8: p. 926-931.

54. Fontana, A., Polverino de Laureto, P., De Filippis, V., Scaramella, E. and Zambonin, M., Limited proteolysis in the study of protein conformation. In Proteolytic Enzymes: Tools and Targets. (Sterchi, E. E. & Stocker, W., eds), Springer, Heidelberg, 1999. P. 253-280.

55. Hao, M. H. and Scheraga, H. A., Analyzing the normal mode dynamics of macromolecules by the component synthesis method: residue clustering and multi component approach. Biopolymers, 1994. 34: p. 321-335.

56. Yukl, E. T., Jepkorir, G., Alontaga, A., Pautsch, L., Rodriguez, J. C., Rivera, M., and Moe¨nne-Loccoz, P., Kietic and spectroscopic Studies of Hemin Acquisition in the Hemophore HasAp from Pseudomonas Aeruginosa. Biochemistry, 2010. 49: p. 6646-6654.

102  

57. Gunasekaran, K., Ma, B., and Nussinov, R., Triggering Loops and Enzyme Function:

Identification of Loops that Trigger and Modulate Movements. J. Mol. Biol., 2003. 332: p. 143-159.

58. Nicholson, L. K., Yamazaki, T., Torchia, D. A., Grzesiek, S., Bax, A., Stahl, S. J. et al.

Flexibility and function in HIV-1 protease. Nature Struct. Biol., 1995. 2: p. 274-280.

59. Derewenda, Z. S., Structure and function of lipases. Advan. Protein Chem., 1994. 45: p. 1-52.

60. Lebioda, L., Stec, B. and Brewer, J. M., The structure of yeast enolase at 2.25-A° resolution. An 8-fold beta þ alpha-barrel with a novel beta beta alpha alpha (beta alpha) 6 topology. J. Biol. Chem., 1989. 264: p. 3685-3693.

61. Lebioda, L., and Stec, B., Mechanism of enolase: the crystal structure of enolase-Mg2(þ)-2-phosphoglycerate/ phosphoenolpyruvate complex at 2.2-A° resolution. Biochemistry, 1991. 30: p. 2817-2822.

62. Ramakrishnan, B., and Qasba, P. K., Crystal structure of lactose synthase reveals a large conformational change in its catalytic component, the b1,4- galactosyltransferase-I. J. Mol. Biol., 2001. 310: p. 205-218.

63. Ramakrishnan, B., Balaji, P. V., and Qasba, P. K., Crystal structure of b1,4-galactosyltransferase complex with UDP-gal reveals an oligosaccharide acceptor binding site. J. Mol. Biol., 2002. 318: p. 491-502.

64. Sham, Y. Y., Ma, B., Tsai, C. J. & Nussinov, R., Thermal unfolding molecular dynamics

simulation of Escherichia coli dihydrofolate reductase: thermal stability of protein domains and unfolding pathway. Proteins: Struct. Funct. Genet., 2002. 46: p. 308-320.

65. Chan, A. C. K.; Lelj-Garolla, B.; Rosell, F. I.; Pedersen, K. A.; Mauk, A. G.; Murphy, M. E. P., Cofacial Heme Binding is Linked to Dimerization by a Bacterial Heme Transport Protein. J. Mol. Biol. 2006. 362: p. 1108-1119.

66. Mattle, D., Zeltina, A., Woo, J., Goetz, B., and Locher, K.P., Two Stacked Heme Molecules in the Binding Pocketof the Periplasmic Heme-Binding Protein HmuT from Yersinia pestis. J. Mol. Biol., 2010. 404: p. 220-231.

103  

CHAPTER III

Effects of Proximal Ligand Substitutions on Heme Coordination and Structural Integrity of HasAp

Introduction

The heme in the crystal structure of the hemophore HasAp (heme acquisition system)

from Pseudomonas aeruginosa (P. aeruginosa) is coordinated by a rare pair of amino acid

residues; a histidine (His 32) and a tyrosine (Tyr 75) (Figure 1) [1,2]. Tyr 75 is also hydrogen

bonded to a nearby residue (His 83, 2.6 Å) via its Oη and this interaction is thought to modulate

the Fe3+-Tyr bond [3,4]. Residues Tyr 75 and His 83, which are conserved in all identified

hemophores [5], (see Figure 6, Chapter II) have been suggested to play a crucial role in heme

binding and /or release [6]. By contrast His 32 is not conserved.

His 32

His 83Tyr 75

Figure 1. Zoomed in view of the heme pocket in the hemophore HasAp . Heme coordination

ligands His 32 and Tyr 75 are colored blue. His 83 is colored orange.

104  

Deniau et al. replaced His 32, Tyr 75 and His 83 for Ala in the hemophore HasAs from S.

marcescens and showed that one axial iron ligand is sufficient to ligate the heme [7]. It was

proposed that His 83 may act as an alternative iron ligand in the absence of Tyr 75 or both His 32

and Tyr 75 [8]. It is noteworthy, however, that there is no conclusive evidence to support the

proposed coordination state of the heme iron in His 83 to Ala mutant. Similarly, information

regarding the overall architecture of the mutant proteins is very limited and inconclusive.

The X-ray crystal structure of H83A-HasAs has been solved but the low resolution (2.6

Å) prevented definitive determination of the heme-iron ligands [9]. A combination of

spectroscopic techniques (RR, NMR and EPR) was used to investigate the coordination state of

the heme-iron in this mutant over a wide range of pH [9]. A summary is shown in Scheme 1

below. It was concluded that at acidic pH (pH 5.0), the major form of holo-H83A is a

pentacoordinate high spin species (5c Hs) where the heme is axially coordinated by His 32. At

neutral and basic pH values, a mixture of high spin and low spin six coordinate species (6c Hs

and 6c Ls) predominates. It was proposed that the 6c Hs species is a His32/H2O but as the pH is

increased the nearby residue (Tyr 75) is deprotonated and its phenolic oxygen atom forms a

hydrogen bond with the bound water stabilizing the low spin hydroxide form, His 32/OH- (6c

Ls). However, recent investigations contradict this model: (i), Lukat-Rodgers et al [10] reported

that when wild type holo-HasAs is reduced with dithionite in the presence of CO, a ferrous

hexacoordinate complex in which His 32 is replaced by CO at the distal site is formed. (ii), the

crystal structure of HasAs in complex with its receptor (HasR), recently reported by Krieg et al

[11], shows that the heme in the hemophore is coordinated only by Tyr 75 at the proximal site

while the His 32 loop is turned away from its heme-binding conformation. Together, these

observations suggest that His 32 is considerably more labile than Tyr 75. It is therefore likely that

105  

the heme ligand that ionizes at high pH (scheme 1) is the distal His not the Tyr. Ionization of a

His residue close to the heme pocket has been reported in other ferric heme proteins [12].

Fe

His 32

(5c HS)

Fe

His 32

(6c HS)

Fe

His 32

(6c LS)

Tyr 75-O

H

H

O

H

Tyr 75-O

H

O

H

Acidic pH Basic pH

In this study, we carried out an investigation to elucidate the nature of the heme iron

coordination and the overall architecture of the Y75A and H83A HasAp mutants. Electronic

absorption spectroscopy revealed that both mutants bind heme and exist as a mixture of high spin

and low spin species in solution. A pH titration analysis shows that Y75A-HasAp possesses a

unique coordination as a function of pH whereas H83A-HasAp is pH sensitive. The crystal

structure of Y75A obtained at pH 7.5 reveals that the heme is coordinated by His 32 at the distal

site and a chance ion at the proximal position. The side chain of His 83 assumes a nearly

identical position with that of wild type, clearly indicating that in the absence of the Tyr 75 His

83 does not necessarily coordinate the heme. This result contradicts the long held speculation

that His 83 may serve as the heme iron ligand in the absence of Tyr 75.

Sheme 1. Proposed coordination state of the heme iron in H83A-HasAs at different pH.

106  

Two crystal forms of H83A-HasAp were obtained. In the first form, obtained at pH 6.5,

the heme is axially coordinated by His 32 and Tyr 75, and the Tyr forms a hydrogen bond with a

chance ion (ethylene glycol) via its phenolic oxygen. In the second crystal form, also obtained at

pH 6.5, the heme is similarly coordinated by His 32 and Tyr 75 but ethylene glycol is replaced by

a water molecule. In either case neither His 32 nor Tyr is detached from the heme. We therefore

propose that the acid-base transition observed in H83A-HasAp is likely due to ionization of His

32. A crystal structure of H83A at low pH would help ascertain this argument.

Overall, the polypeptide fold of Y75A and H83A HasAp mutants is similar to that of the

wild type protein. The crystal structures reveal a few conformational changes in the loops that

harbor the heme iron ligands, and gain or loss in the secondary structure of the loop implicated to

interact with the receptor HasR during heme release. Loss of beta sheet character or shortening

of beta strands is also observed in the H83A mutant. Unless otherwise stated, the discussions

herein are based on H83A crystal form 1.

Experimental Procedures

Construction of Y75A and H83A HasAp mutants

Y75A and H83A HasAp mutants were constructed in the background of the gene coding

for the truncated form of HasAp missing the last 21 C′-terminal amino acids relative to the full-

length protein. Construction of the truncated form is described in reference 12. Mutagenic

primers were custom synthesized by Integrated DNA Technologies, Inc. and were used (without

further purification) in conjunction with polymerase chain reaction (PCR) and QuickChange TM

107  

site-directed mutagenesis kit (Stratagene; La Jola, CA) to introduce the desired mutations.

Sequences corresponding to the Y75A complementary oligonucleotide primers are

5`-GGCGGCGATCTGCATGCGACCCTGTTT-3`and 5`-GCT AAACAGGGTCGCA TGCAG

ATCGCCGCC-3` while those corresponding to H83A oligonucleotide primers are 5`-CTGTTT

AGCAACCCGAGCGCGACCCTGTGGGGC-3` and 5`-GCCCCACAGGGT CGCGCTCGGG

TTGCTAAACAG-3`. The underlined codons represent target substitutions. Each gene construct

was transformed into XL1-Blue Competent cells (Stratagene) for amplification. DNA sequences

(Figure 2) were verified by SeqWrigh (Huston, Texas). Recombinant DNA plasmid with the

correct sequence was transformed into Escherichia coli BL21 (DE3) GOLD Competent cells for

protein expression.

Expression and purification of Y75A and H83A mutants

Recombinant proteins were expressed in the BL21 DE3 strain of E.coli following a

similar procedure that was used to express H32A and wild type protein. Briefly, a single colony

from the transformant E. coli BL21 (DE3) Gold Competent cells was cultured overnight at 37 °C

in 50 mL of LB medium containing 100 μg/mL ampicillin. 30 mL of the overnight innoculum

was sub-cultured in fresh 1 L M9 medium supplemented with 200 µg/mL ampicillin and the cells

were grown to an OD600 of ~ 1 at 37 o C, shaking at 235 rpm. The temperature was then lowered

to 30 °C before inducing polypeptide synthesis by addition of 1 mM isopropyl-D-

thiogalactopyranoside (IPTG). The cells were allowed to grow for an additional 5 h at 30 o C,

harvested by centrifugation at 4800 rpm and 4 0C for 12 min, and store overnight in -20 °C. Cells

were resuspended in 20 mM Tris-HCl pH 7.6 (3 mL/ wet gram cell) and disrupted by sonication

108  

in ice in the presence of a pinch of DNase (Sigma-Aldrich, St. Louis, MO). The lysate was

pelleted by ultracentrifugation at 19500 rpm and 4 °C for 40 min. The resultant supernatant was

loaded onto an ion exchange column pre-equilibrated with 20 mM Tris-HCl pH 7.6. Following

column washing with at least three bed volumes of the same buffer, the protein was eluted with a

linear gradient of (0-0.6 M) NaCl. The apo fraction (~ 80 %) was gently separated from its holo

counterpart using hydrophobic affinity chromatography and following the same protocol

developed for isolation of apo-HasAp as detailed in the experimental section of chapter II. The

concentration of the pure samples was determined using Bovine Serum Albumin (BSA) protein

kit (Pierce).

109  

S N T G G F N P

CCG TTT GAT GGC AGC CAG TAT GCG CTG AAC AGC ACC GCG AGC GAT GCG P F D G S Q F A L K S T A S D A

TTT ATT GCG GGC GGC GAT CTG CAT GCG ACC CTG TTT AGC AAC CCG AGC GCGF I A G G D L H A T L F S N P S

ACC CTG TGG GGC AAA CTG GAT AGC ATT GCG CTG GGC GAT ACC CTG ACC GGCT L W G K L D S G

GGC GCG AGC AGCGGC GGCTAT GCG CTG GAT AGC CAG GAA GTG AGC TTT AGCG A S S G G Y A L D S Q E V S F S

AAC CTG GGC CTG GAT AGC CCG ATT GCG CAG GGC CGT GAT GGC ACC GTG N L G L D S P I A Q G R D G T V

AAA GTG GTGTAT GGC CTG ATG AGC GGC GAT AGC AGC GCG CTG CAG GGC CAGK V V Y G L M S G D S S A L Q G

ATT GAT GCG CTG CTG AAA GCG GTT GAT CCG AGC CTG AGC ATT AAC AGC ACC I D A L L K A V D P S L S I N S

TTT GAT CAG CTG GCG GCG GCGGGC GTG GCG CAT GCG ACC CCG GCG GCGGCGF D Q L A A A G V A H A T P A A A

GCG GCGGAA GTG GGC GTG GTG GGC GTG CAG GAA CTG CCG CAT GAT CTG GCGA A E V G V V G V Q E L P H D L

Bam H ICTG GCG GCGTAA GGATCC GGG

L A A -

5' NdeIGGGCATATGAGC ATT AGC ATT AGC TAT AGC ACC ACC TAT AGC GGC TGG ACC

M S I S I S Y S T T Y S G W T

GTG GCG GAT TAT CTG GCG GAT TGG AGC GCG TAT TTT GGC GAT GTGAAC CATV A D Y L A D W S A Y F G D V N H

CGT CCG GGC CAG GTG GTG GAT GGC AGC AAC ACC GGC GGC TTT AAC CCG GGCR P G Q V V D G G

GCGP F D G S Q F A A

A

I A L G D T L T

CATH

Q

ATT GAT GCG CTG T

TTT GAT CAG CTG GCG

GCG GCGGAA GTG GGC GTG A

I

L A A -

3'

Figure 2. DNA and amino acid sequence of HasAp-TR. The Nde I and BamHI

restriction endonuclease sites were constructed at the 5' and 3' ends, respectively, for

subcloning. The underlined codons represent mutation of Tyr or His to Ala at positions

75 and 83, respectively.

110  

Preparation of holo protein

Identical reconstitution methodologies were employed on Y75A and H83A. Fresh

solution of 1 mM hemin was prepared by dissolving 6.5 mg of hemin chloride in 10 mL DMSO.

250 µL aliquots of the resultant solution were used to titrate 2 µmoles (10-12 µM) of protein,

while monitoring hemin uptake spectrophotometrically. After every addition, 3 min mixing time

was allowed before absorbance measurement was taken. This was continued until the Rz value

(absorbance of soret/ absorbance at 280 nm) remained constant. The heme-complex mixture was

allowed to incubate overnight at 4 0C, before concentrating to 2 mL using 10 kDa MWCO

Amicon ultracentrifuge filter devices. During concentration, the retentate was constantly mixed

every 15 minutes to minimize or avoid the possible formation of dimers. The concentrated

sample was passed over a gel filtration column (Sephadex G-75; 2.6 cm i.d X 90 cm: GE

Healthcare) to remove excess heme. MW of pure holo samples was checked with electrospray

mass spectrometry following a similar procedure for determining the MW of H32A described in

chapter II. The oligomeric state was determined with the aid of a high resolution gel filtration

column (Superdex 75 prep grade, GE Healthcare) in 50 mM Tris-HCl (pH 7.5), 100 mM KCl

buffer. The column was calibrated with; cytochrome c (12.4 kDa), chymotrypsin (25 kDa),

ovalbumin (44 kDa), and albumin (67 kDa) standards.

Determination of the extinction coefficient for the Y75A- and H83A-heme complex.

The millimolar extinction coefficient at 406 nm (406 ) or at 405 nm (405 ) for Y75A and

H83A heme complexes, respectively, was determined by the pyridine-hemochrome method [13].

Briefly, the absorbance of the purified holo Y75A and H83A samples was measured at their

111  

respective soret maximum. The solution was diluted with alkaline pyridine and a few grains of

dithionite were added. The spectrum of the reduced ferrous pyridine hemochrome was then

recorded. The mean concentration was calculated from the absorbance values at 418, 525, and

557 nm using reported epsilon values of 170 mm-1cm-1, 17.5 mm-1cm-1 and 34.5 mm-1cm-1 ,

respectively, for reduced hemochromes.

Protein crystallization

HasAp mutants in 100 mM sodium phosphate pH 7.8 (Y75A) or 100 mM Tris pH 7.6

(H83A) were concentrated to 15 mg/mL and screened for crystal growth using Compact Jr.

(Emerald biosystems) sitting drop vapor diffusion plates. The sitting drop was prepared by

mixing 0.5 µL of protein and 0.5 µL of crystallization solution then equilibrated against 100 µL

of the latter. Initial crystals of Y75A were obtained within 24 hours at 20 oC from the Hampton

Crystal Screen 2 (Hampton Research) condition #32 (1.6M (NH4)2SO4, 100 mM HEPES pH 7.5,

200 mM NaCl). Crystals were transferred to a solution containing 80% crystallization solution

and 20% glycerol before freezing in liquid nitrogen for data collection. In the case of H83A,

initial crystals were similarly obtained in 24 hours at 20 oC from Wizard II (Emerald

biosystems) condition #11 (10% isopropanol, 100 mM Na cacodylate pH 6.5, 200 mM zinc

acetate) ( Appendix 1). However, the original crystals of H83A grew as clusters of small needles

and did not diffract. Crystallization conditions refined using pHatTM Buffer Screen (Emerald

Biosystems) followed by Hampton Additive Screen resulted in plate shaped clusters that

diffracted to a high resolution from 10% isopropanol, 100 mM Sodium Cacodylate pH 6.5, 200

mM zinc acetate, 3% dioxane. Samples were transferred to a solution containing 75%

112  

crystallization solution and 25% ethylene glycol before freezing in liquid nitrogen for data

collection.

Data collection, structure determination and refinement

Initial x-ray diffraction data were collect at the University of Kansas Protein Structure

Laboratory using a Rigaku R-Axis IV++, image plate detector equipped with an RU-H3R rotating

anode generator (Cu-Kα, 1.5418Å) and Osmic Blue focusing mirrors. High resolution

diffraction data were collected at the Advanced Photon Source IMCA-CAT beamline 17ID using

a Dectris Pilatus 6M pixel array detector.

Diffraction intensities were integrated and scaled using the XDS [14] and Scala [15]

packages, respectively. The Laue class was checked using Pointless [14] which indicated that

mmm and 2/m was the correct class for Y75A and H83A respectively. Coordinates from a

previously determined structure of HasAp (PDB: 3ELL) were used for molecular replacement

search with Molrep[3]. To determine the structure of Y75A, a single monomer of HasAp was

used and residue Y75 was omitted from the search model. Prominent difference (Fo-Fc) electron

density was observed near the heme on the side that would normally be occupied by Y75. Three

peaks were present in this region that were at an angle of approximately 120o and separated by

approximately 1.3Å. Refinement of an ethanol molecule at this site resulted in positive Fo-Fc

electron density at C2 atom of the ethanol indicating an underestimation of electrons whereas

placement of a formate ion did not yield any residual electron density. This electron density was

observed from four diffraction data sets, ranging from 0.95 Å to 1.3 Å resolution obtained from

four unique crystallization conditions. Since the electron density was not consistent with any of

113  

the crystallization components, protein storage buffer or purification buffers, it is likely that this

ion was present in the heme used in the analysis.

To solve the structure of H83A, a single monomer of the previously determined structure

of HasAp (PDB: 3ELL) was used for molecular replacement search with Molrep [16] and

residue H83 was omitted from the search model. Molecular replacement searches in the space

groups P21 and P2 yielded correlation coefficients of 0.62 and 0.30 respectively. Therefore,

subsequent structure refinement was conducted in the space group P21. Initial structure solution

and refinement were conducted using in-house diffraction data collected with Cu-Kα radiation.

Large peaks greater than 10 were observed in the difference Fourier maps (Fo-Fc) that were

likely zinc ions from the crystallization solution. No anomalous signal was observed at these

sites using Cu-Kα data. An X-ray fluorescence scan indicated that zinc was indeed present in the

crystal (Figure 3). Data for structure refinement were collected at a wavelength of 1.0000 Å

which still yields an appreciable anomalous signal for Zn ~2.6e-, (Figure 4). In addition,

diffraction data were collected at a low energy remote wavelength (λ =1.28414 Å, 9,655 eV)

which resulted in the disappearance of the anomalous difference density at the Zn sites but an

increase at the hemin Fe site (Figure 5). This confirmed that the assignment of Zn at these sites

was correct.

Structure refinement of positional and anisotropic atomic displacement parameters and

manual model building for both mutants were performed with Phenix [17] and Coot [18]

respectively. Structure validation was carried out using Molprobity [19] and figures were

prepared using CCP4mg [20] and pyMol [21]. Data collection and refinement statistics are

summarized in Table 1.

114  

 

Figure 3. X-ray fluorescence scan for zinc using a H83A

Figure 4. Phased anomalous difference map

calculated using data collected at λ=1.0000 Å (red

mesh) contoured at 4σ showing the Zinc ion sites in

115  

Figure 5. Phased anomalous difference map using data

collected at λ=1.0000 Å (red mesh) and 1.28414 Å (low

energy remote, green mesh) contoured at 4σ. The

anomalous difference density increases at the hemin Fe

atom and disappears at the Zinc ion site when using the

low energy remote wavelength data.

116  

Table1. Data collection and Refinement Statistics of HasAp mutants

Y75A H83A

Data Collection Unit-cell parameters (Å, o) a = 34.440, b = 46.73,

c = 101.04 a = 34.90, b = 66.23 c = 41.00, = 97.13

Space group P212121 P21 Resolution (Å)1 40.00-0.95 (1.00-0.95) 66.23-1.32 (1.39-1.32) Wavelength (Å) 1.0000 1.0000 Temperature (K) 100 100 Observed reflections 901,848 151,394 Unique reflections 97,317 43,290 Mean (<I>/�<I>)1 18.7 (2.6) 7.7 (2.2) Completeness (%)1 94.1 (80.0) 99.6 (99.8) Multiplicity1 9.3 (5.6) 3.5 (3.5) Rmerge (%)1, 2 5.5 (60.1) 9.5 (68.1)

Rmeas4 5.8 (66.3) 11.2 (80.5)

Rpim4 1.7 (27.3) 5.9 (42.3)

Refinement Resolution (Å) 34.30-0.95 34.63-1.32 Reflections (working/test) 92,362 / 4,861 41,067 / 2,180 Rfactor / Rfree (%)3 13.5 / 14.6 14.3 / 16.8 No. of atoms( protein / heme / zinc / ethlyene glycol / formate/water)

1,357 / 43 / 3 / 16/322 1,356 / 43 / 9 / 16 / 215

Model Quality R.m.s deviations Bond lengths (Å) 0.012 0.010 Bond angles (o) 1.456 1.356 Average B factor (Å2) All Atoms 12.4 13.5 Protein 10.3 11.8 Heme 8.0 9.2 Zinc - 16.3 Ethylene glycol Formate

- 7.4

23.3 -

Water 22.2 24.5 Coordinate error based on Maximum Likelihood (Å)

0.07 0.16

Ramachandran Plot (chain A/B) Favored (%) 98.4 98.9 Allowed (%) 1.6 1.1

1) Values in parenthesis are for the highest resolution shell.

2) Rmerge = ∑hkl ∑ i |Ii(hkl) - <I(hkl)>| / ∑hkl∑i Ii(hkl), where Ii(hkl) is the intensity

measured for the ith reflection and <I(hkl)> is the average intensity of all reflections with indices hkl.

3) Rfactor = ∑hkl ||Fobs (hkl) | - |Fcalc (hkl) || / ∑hkl |Fobs (hkl)|; Rfree is calculated in an identical manner using 5% of randomly selected

reflections that were not included in the refinement

4) Rmeas = redundancy-independent (multiplicity-weighted) Rmerge (P.R.Evans, "Scaling and assessment of data quality", Acta Cryst.,

2006. D62: p. 72-82.

117  

Results

Purification of Y75A and H83A mutants

Both Y75A and H83A accumulate in the E. coli host mostly as apo (~80 %) protein,

judged by hemin absorption features in the UV-vis spectra. Following ion exchange

chromatography, the apo fraction was separated from its holo counterpart using hydrophobic

affinity chromatography and then reconstituted with hemin. Both mutants eluted from a size

exclusion chromatography column in a single band (Figure 6) and run in a 15 % SDS gel with an

apparent mass of ~ 19 kDa. The molecular mass was estimated by size exclusion

chromatography as 16.9 kDa (Y75A) and 17.6 kDa (H83A) (Figure 7) which is very close to the

calculated molecular weight of a monomer.

wavelength (nm)

35 40 45 50 55 60

Abs

orba

nce

0.0

0.5

1.0

1.5

2.0

2.5

Figure 6. Elution profile of Y75A-HasAp from Sephadex G-75 column; open circles track the

soret band absorption (~406 nm) and the black circles track absorbance at 280 nm. A similar

profile was observed with the H83A-HasAp mutant.

118  

Ve (mL)

12 14 16 18

log

MW

1.0

1.2

1.4

1.6

1.8

2.0

Ovalbumin

Albumin

Chymotrypsin

Cytochrome c

Y75AH83A

This is corroborated by mass values obtained with the aid of ESI-MS; 18,667 Da (Y75A) and

18,693 Da (H83A), in good agreement with the masses calculated from amino acid sequences

18,668 Da and 18,694 Da for Y75A and H83A (excluding initiator methionine), respectively

(Appendix 2 and Appendix 3).

Characterization by UV-vis spectroscopy

Figure 8 shows the electronic absorbance spectra of Y75A (black) and H83A (red)

recorded in sodium phosphate buffer (µ = 0.1, pH 7.8). Y75A exhibit absorbance peaks at 406

Figure 7. Elution volumes (Ve) from a calibrated Superdex 75 column and estimated molecular

weight of Y75A HasAp (red triangle) and H83A HasAp (green diamond). The Superdex 75

column was calibrated with albumin (67 kDa), ovalbumin (44 kDa), chymotrypsin (25 kDa), and

cytochrome c (12.4 kDa) shown in black filled circles.

119  

nm (soret), 535, 574, and 624 nm. The Q bands at 535 and 574 nm indicate the presence of low

spin species in solution, while the soret maximum and the charge transfer band at 624 nm are

indicative of the presence of high spin species [22, 23]. Thus, like wild type HasAp, Y75A

mutant exists in solution as a mixture of high spin and low spin species. It is worth stressing that

the spectral properties did not change over a pH range of 5.0- 9.5.

The electronic absorption spectrum of H83A at pH 7.0 has a broad soret band at 405 nm,

weak Q bands at 495 and 555 nm, as well as very intense characteristic high-spin marker band at

610 nm (red trace in Figure 8). pH titration analysis show that at acidic pH, the soret band is at

404 nm and the species in solution is predominantly high spin as evidenced by soret maximum

and intense high spin marker bands at 500 and 618 nm (Figure 9). Upon increasing the pH, the

soret band decays and concomitantly red shifts by about a nm, while the 500 nm and 618 nm

high spin bands decay giving rise to a new set of species absorbing at 495, 555, 580, and 610 nm.

The 555 nm and 580 nm bands are characteristic of low spin species while the 495 and 610 nm

bands are characteristic of high spin species albeit different from the high spin species observed

at low pH. Thus, unlike wild type and Y75A, H83A exhibit three pH dependent forms; a high

spin species at acidic pH and a different high spin species and a low spin species at neutral and

alkaline pH.

120  

Wavelength (nm)

300 400 500 600 700

Abs

orba

nce

0.0

0.2

0.4

0.6

0.8

1.0

1.2

405 406

495

624

535610

555

Y75AH83A

Wavelength (nm)

450 500 550 600 650 700

0.00

0.02

0.04

0.06

pH 5.0 pH 5.5 pH 6.0 pH 6.5 pH 7.3 pH 8.1 pH 9.5

300 400

Abs

orb

an

ce

0.0

0.2

0.4

0.6

0.8

1.0

pH 5.0

pH 9.5

Figure 8. Electronic absorbance spectra of Y75A (black) and H83A (red) recorded in sodium

phosphate buffer (µ = 0.1, pH 7.8).

Figure 9. pH titration spectra of H83A over a pH range of 5.0 to 9.5 in sodium

phosphate buffer , µ = 0.1.

121  

X-ray Crystallography

Crystallography is a useful tool for determining three-dimensional protein structures at

molecular level. However, producing high-quality crystals that yield good diffraction data

remains a major bottleneck in structure determination [22]. Crystallization is thought to occur in

two steps: (i) nucleation and (ii) growth of nuclei to macroscopic crystals [23]. Nucleation,

which is a prerequisite, requires different conditions from growth but once a critical nucleus has

formed, growth follows spontaneously [24].

A common initial approach to protein crystallization is to set up screening trials using

random sparse-matrix screens [25]. In these screens, the protein sample is exposed to a variety of

formulations with the aim of finding ‘hits’ or ‘leads’ that may be conducive for crystal growth.

Once the leads are identified, the conditions are optimized by varying the concentration of the

protein, buffer composition, temperature, pH, type and concentration of precipitant, or by adding

additives to optimized lead conditions. Alternatively, the crystallization environment in the

sparse-matrix screen can be actively influenced and controlled as crystallization proceeds so as

to lead crystal growth in the direction that will give the best results [22].

Several methods are used to grow crystals of biological macromolecules but vapor

diffusion is the most commonly employed technique [25]. In this method, the protein solution

(containing pure protein and crystallization solution in a 1:1 ratio) is either a hanging drop or a

sitting drop that continuously equilibrates against a reservoir containing similar crystallizing

agents but at a higher concentration than in the drop [25]. The difference between the two

diffusion methods is the vertical orientation of the droplet within the system. In hanging drop,

the droplet is suspended on the cover slip over the reservoir like a hanging drop whereas in

sitting drop the wells are fitted with a support for the droplet to sit on. The sitting drop method

122  

was used in this work. Initially, the droplet of protein solution contains an insufficient

concentration of precipitant to effect crystallization, but as water evaporates from the drop the

precipitant concentration increases to a level optimal for crystallization. The process of

equilibration continues and optimum conditions are maintained until the crystallization is

complete [25].

Crystallization of Y75A and H83A HasAp mutants

Crystals of Y75A mutant (Figure 10-C) were obtained from four unique crystallization

conditions. Since these crystals diffracted to a high resolution, further refinement was deemed

not necessary. In the case of H83A, initial crystals from a random screen (Wizard II #11) grew as

clusters of small needles that did not diffract (Figure 10-A). Crystallization conditions were fine-

tuned using (i) pHat™ buffer Screen to simultaneously explore the effects of pH and buffer

composition on crystal growth, and (ii) Hampton Additive Screen to improve the quality and/or

size of the crystals. From the former refinement screen, small low-quality crystals were observed

only in conditions buffered with sodium cacodylate. This showed that cacodylate buffer is

important in crystal growth and hence it was used in the latter optimization screens. Plate shaped

clusters (Figure 10-B) that diffracted to a high resolution were obtained from 10% isopropanol,

100 mM sodium cacodylate pH 6.5, 200 mM zinc acetate, 3% dioxane. Before the samples were

frozen for data collection, they were placed in a cryoprotectant solution containing 25 %

ethylene glycol

123  

BA

C

A B

Crystal structures of Y75A and H83A HasAp mutants

The crystal structure of the Y75A HasAp mutant in the orthorhombic space group P212121

was solved at 0.95 Å resolution with Rfree of 14.6 %, using holo-HasAp (PDB: 3ELL) as the

search model. The H83A mutant crystallized in the monoclinic space group P21 and its structure

was refined to 1.3 Å resolution with Rfree of 16.8 % (crystal form 1). H83A crystal form 2 was

refined to 0.89 Å. Unlike the wild type protein which has two subunits in an asymmetric unit,

both mutants contain one molecule in the asymmetric unit. Since both molecules (A and B) in

wild type HasAp are similar, from here on, all comparisons are made using molecule B.

Figure 10. (A) Original crystals of

H83A obtained from 10% isopropanol,

100 mM sodium cacodylate pH 6.5,

200 mM zinc acetate; (B) Refined

crystals obtained from 10%

isopropanol, 100 mM sodium

cacodylate pH 6.5, 200 mM Zinc

acetate, 3% dioxane. (C) Single

crystals of Y75A obtained from

124  

Heme iron coordination and conformational changes involving His 32 and Tyr 75 loops.

Heme interaction with the protein can be viewed in reference to the two important

structural elements, namely His 32 loop and Tyr 75 loop at the interface of the -helical and -

sheet wall. The His 32 loop harbors the distal ligand whereas the Tyr 75 loop provides the

proximal ligand to the heme-iron.

In Y75A HasAp mutant, a 2Fo-Fc electron density map contoured at 1 clearly show that

the heme adopts a single orientation (Figure 11-A). At the proximal site is an elongated well-

defined density assignable to a formate ion at a distance of 2.0 Å from the central metal iron

(Figure 11-B). This formate ion is directed toward the -meso carbon of hemin and is identically

superimposable with the CO group of Tyr 75 in the wild type protein. One of its O atoms forms a

hydrogen-bond with Nδ of His 83 (2.6 Å) whereas the other interacts with the backbone carbonyl

group of Thr 76 (3.8 Å) (Figure 11-C). It is noteworthy to mention here that the side chain of His

83 is positioned in a nearly identical position to that of the wild type protein and as such, it does

not coordinate the heme-iron. In the H83A HasAp mutant, 2Fo-Fc electron density map

contoured at 1 (Figure 12-A) show that the heme adopts a single orientation. In crystal form 2

of H83A though, the heme is coordinated by Tyr 75 which in turn forms a hydrogen bond with a

water molecule (Figure 12-B). The water molecule is replaced by a chance ion ethylene glycol

(EG) in H83A crystal form 1 (Figure 12-C). Together, these observations contradict the proposal

that in the absence of His 83, the Tyr 75 to iron coordination bond is severed. Secondly, the high

resolution crystal structure of Y75A mutant shows that in the absence of Tyr 75, His 83 does not

necessarily coordinate the heme.

125  

A B

C D

Tyr75 loop

Hemin

His 32

Formate

Figure 11. 2Fo-Fc electron density map contoured at 1; (A) Electron density at the heme showing that it adopts a single orientation; (B) Electron density of the refined formate ion. (C) Zoomed in view of hemin binding pocket. (D) Sphere representation of Tyr 75 loop and coordinated formate ion.

126  

His 32

Ala 83

Tyr 75EG

(C) H83A_1(B) H83A_2(A)

H2O

As shown in Figure 13, the heme propionates adopt identical hydrogen bonding

interactions with the protein as seen in the crystal structure of wild type HasAp; the polar end of

Arg 129 side chain (3.5 Å) and N atom of Gly 35 (3.0 Å) form hydrogen bonds with the first and

the second propionates, respectively. The H83A crystal form 1 deviates slightly from the wild

type protein in the sense that one of the propionates is coordinated to another chance ion, Zn2+.

Since the propionates in general have a high propensity to interact electrostatically with side

chains of positively charged residues or cations, by virtue of its proximity, Zn2+ ion (2.0 Å, 2.9

Å) was close enough to provide favorable interactions compared to Gly 35 N atom located 4.5Å

away. Additional interactions with a network of four to seven water molecules cap the front ends

of the solvent exposed propionates in both mutants. Lastly, the position of the coordinating

ligands His 32 and Tyr 75 are stabilized by hydrogen bonding interactions with Oη of Tyr 138

(2.7 Å) and carboxyl group of Phe 51 (2.9 Å), respectively. All these interactions seem to

preserve the architecture of the heme pocket and anchor the heme macrocyle tightly within the

binding site despite the disruption of the crucial Tyr 75-His 83 hydrogen bond. It is however

Figue 12. 2Fo-Fc electron density map at the heme contoured at 1 (A), and zoomed in view of heme

binding pocket in H83A crystal form 2 (B) and H83A crystal form 1(C).

127  

notable that the subtle effects contributed by proximal heme coordinating ligands and the chance

ions are large enough to induce conformational changes in the His 32 and Tyr 75 loops.

Y75A

H83A_2

WT

H83A_1

ARG 129GLY 35

TYR 138

PHE 51

HIS 32TYR 75

ALA 83H2O

ARG 129GLY 35

TYR 138

PHE 51

HIS 32

ALA 75

HIS 83

FORMATE

ARG 129

TYR 75

EG

HIS 32

Zn2+

GLY 35

ALA 83

PHE 51

ARG 129

GLY 35

TYR 138

PHE 51

HIS 32TYR 75

HIS 83

Figure 13. Heme-iron ligands and residues that make direct contact with the propionates in WT,

Y75A, and H83A HasAp mutants.

128  

Conformational changes in the His 32 and Tyr 75 loops.

An important feature that becomes apparent in the mutants is the movement of the His 32

and Tyr 75 loops away from the heme relative to their position in wild type HasAp (Figure 14).

In H83A crystal form 2, the movement further extents to the hairpin loop connecting 2 and 3

(Lys 59 - Asp 64). Conformational changes in the backbone of His 32 loop is accompanied by

spatial reorientation of side chains involving nearly all amino acid residues that constitute the

loop. Significant changes are observed in Arg 33, Gln 36, Val 37, Val 38, and Asn 39 (Figure 15)

near the hinge residue Gly 45. Although disruption of native heme-iron coordination may justify

these changes, it is interesting to note that the His 32 loop is more affected yet the mutations are

on the Tyr 75 loop. In addition, the bond lengths between the heme-iron and the axial ligands are

nearly maintained (Table 2). Nonetheless, the slight tilting motion of the heme plane away from

the Tyr 75 loop that accompany His 32 movement, and substitution of a small amino acid

Y75A H83A_1

H32 loop

Y75 loop

H83A_2

K59-D64

Y75 loop

H32 loop

Figure 14. Zoomed in view of His 32 and Tyr 75 loops of wild type HasAp (magenta), Y75A

(green), H83A crystal form 1(cyan), and H83A crystal form 2 (lemon).

129  

residue (Ala) for the bulky proximal heme residues Tyr 75 and His 83 relief the heme pocket of

steric bulk of the amino acid side chains between the heme-iron and the locale of the Tyr 75 loop

and this may have precipitated sliding in and coordination of the formate ion in Y75A (Figure

11-D), and coordination of EG in H83A (Figure 12-C).

Table 2. Distances between key residues and hemin-iron

WT H83A_1 H83A_2 Y75A

Ligand Distance (Å)

His 32 Nε 2.0 2.0 2.0 2.0

Tyr 75 Oη 2.0 2.1 2.2 -

His 83 Nδ 3.9 - - 3.9

Ethylene glycol - 4.2 - -

Water - - 4.1 -

Formate - - - 2.0

130  

ASP 29VAL 30GLN 36

VAL 37VAL 38-A

ASP 39-AVAL 38-B

ASP 39-BHIS 32

ARG 33-B

ARG 33-A

HEME

Y75A H83A

ASP 39-A

ASP 39-B

VAL 38-B

VAL 38-AVAL 37

HIS 32

ARG 33-A

ARG 33-B

ASN 31-A

ASN 31-BASP 29

GLN 36

HEME

ASN 80

PRO 81

SER 82

HIS 83

TRY 75

ALA 75

SER 79

PHE 78

LEU 77

THR 76

THR 84

LEU 85

TRP 86

LEU 85

THR 84

PHE 78SER 82

SER 79

ASN 80

PRO 81

TRP 86

LEU 77

TYR 75

HIS 83 ALA 83

THR 76

H32 Loops

Y75 Loops

Overall structure of HasAp mutants

The overall structure of both mutants is nearly identical to that of wild type with a few

exceptions. First, there is a gain of secondary structure elements in Y75A and in H83A crystal

form 2, in a section of the hairpin loop adjoining 6 and 7 (310-helix, colored purple in Figure

16-B and C). In contrast, electron density of residues constituting the same section of the hairpin

loop in H83A crystal form 1 is missing (marked with white asterisks in Figure 16-D). It is

Figure 15. Conformational changes in the backbone and side chains of amino acid residues that

constitute His 32 loop and Tyr 75 loop in Y75A (green) and H83A (cyan), compared to WT

(magenta).

131  

interesting that a single mutation in the vicinity of hemin can induce structural gain or loss in

areas remote from the point of mutation. This suggests communication between the loop affected

(hairpin loop connecting 6 and 7) and the Tyr 75 loop, in agreement with reports involving

interaction between HasAs and its outer membrane receptor HasR [11].

The second difference which is only exhibited in the crystal structures of H83A is the loss

of beta sheet character in and/orlabeledorandcolored slate in Figure 16-C and

D). Intriguingly, the residues that constitute these regions of the protein are positioned in space to

mimic beta strands 7 and/or 5 as seen in the structure of the wild type protein (Figure 16-A). In

addition, the length of and 1 in H83A crystal form 2 is significantly reduced (Figure

16-C). These observations point to the crucial role of the hydrogen bonding interaction between

Oη of Tyr 75 and Nδ of His 83 in maintaining the overall structural integrity of the protein.

132  

8

5

**

1

8

5

1

5

8

5

(A) WT (B) Y75A

(D) H83A_1(C) H83A_2

Studies with HasAs showed that the loop flanked by 6 and 7 is one of the two distinct

regions in HasAs that mediate its interactions with its cognate receptor HasR during heme

Figure 16. X-ray crystal structures of (A) WT HasAp monomer B (PDB: 3ELL), (B) Y75A, (C)

H83A crystal form 2, (D) H83A crystal form 1. Secondary structures are colored cyan (-helices),

magenta (-strands) and green (loops). 310-helices, reflecting a gain in secondary structure elements

are colored purple (Fig. B and C). 5* and 7* (Fig C and D) are new hairpin loops resulting from

loss of corresponding beta sheet character. Missing electron density is marked with white asterisks

133  

delivery [8]. The other receptor binding region is a 10 amino acid stretch in 3, located behind

and directly opposite this loop. It was proposed that upon binding of the hemophore to the

receptor, the two regions move apart in a TobB-dependent manner and as a result, the heme

binding pocket might open up leading to destabilization around heme iron and iron ligands (e.g

hydrogen bond scission) and subsequent transfer of the heme to the receptor. Close inspection of

crystal structures of wild type HasAp and the Y75A and H83A mutants reveals conserved

hydrogen bonding interactions between Asn 47 (in 2), Ser 54 and Gln 55 (in 3), and Tyr 107

(in the loop adjoining 5 and 6) (Figure 17). These interactions may be important in

maintaining the loop in place and the overall structure of the protein and likely, the bonds are

broken when the hemophore is bound to the receptor to facilitate heme transfer. The gain in the

secondary structure elements in Y75A, and H83A crystal form 2 is reflected by additional

hydrogen bonding interactions between Tyr 107 (C`-end of the loop) / Asn 47 (2) pair, and Ala

102 (N`-end the loop).

Analysis of thermal factors in the structure of Y75A and H83A HasAp (summarized in

Figure 18) show that the B-factors of residues constituting the loop mentioned above (Gly 101-

Gly 106) significantly differ in the three proteins and correlate with observations made in the

crystal structures.

134  

TRY 107

ALA 102GLN 55

SER 54 ASN 47

TYR 107

GLN 55

SER 54

ASN 47

TRY 107

ALA 102

GLN 55

SER 54

ASN 47

TYR 107

GLN 55

SER 54

ASN 47

WT Y75A

H83A_2 H83A_1

Figure 17. Conserved hydrogen bonding interactions involving residues in 2 (Asn 47) 3

(Ser 54, Gln 55) and the loop connecting 6 and 7 (Ala 102 and Tyr 107).

135  

Residue

0 20 40 60 80 100 120 140 160 180

B-v

alu

e (A

2 )

0

10

20

30

40

Residue

0 20 40 60 80 100 120 140 160 180

B-v

alue

(A

2)

0

10

20

30

40

H83A_1

Residue

0 20 40 60 80 100 120 140 160 180

B-v

alue

(A

2 )

0

5

10

15

20

25

Y75A

G106 G101

59-64

38-40

102-105

38-40

59-64

38-4038-40

H83A_2

Figure 18. Putty representation of B-factors color traced on the structures increasing

from blue to red and corresponding per residue plots.

136  

The fact that Y75A mutant has B-factors in the lower range in the above mentioned loop is

indeed surprising. Ideally, mutation of a heme coordination residue would be expected to have a

larger effect than that caused by the adjacent indirectly coordinating His 83. Two other regions of

the protein that exhibit high B factor values are Val 38-Gly 40 located toward the C-end of the

His 32 loop close to the hinge residue Gly 45 , and Lys 59-Asp 64 adjacent to 3. The regions

of the protein were shown by molecular simulations (see Figure 21 in Chapter II) to undergo

motions as the His 32 loop transitions from the open to the closed conformation. We suppose that

the thermal instability of the affected parts in the mutants originate from the heme pocket due to

disruption of the natural Tyr-His hydrogen bond.

137  

Discussion

Y75A and H83A HasAp mutants were expressed and purified to homogeneity in the apo

form before reconstituting with heme to obtain the holo protein. The molecular weight estimated

by size exclusion chromatography and determined by ESI/MS revealed that both mutants are

monomeric. Assessment by UV-visible absorption spectroscopy showed that like wild type, both

mutants exist in solution as a mixture of ls and hs species. However, unlike wild type and Y75A,

H83A is pH sensitive and exhibit three pH dependent forms; a high spin species at acidic pH (pH

5.0) and a mixture of high spin and low spin species at neutral and alkaline pH. Coordination

number and spin states studies for the different forms of holo H83A in HasAs revealed that

pentacoordinate high spin species dominates at acidic pH, and as the pH is increased, the

concentration of this species decreases giving rise to hexacoordinate high-spin and low-spin

species [9]. Electronic absorption spectral features observed for H83A-HasAp are in agreement

with these spin states.

Single X-ray crystal structures of Y75A and H83A HasAp reveal a polypeptide fold

similar to that of the wild type protein. The structural integrity of the protein is maintained by

interactions within and between the secondary structure elements (-strands, -helices, turns),

and a network of water molecules surrounding the protein. Slight, yet significant, conformational

changes are observed in or close to regions of the beta sheet wall in which equivalent residues in

HasAs have been implicated to interact with the receptor HasR [8]. In particular, partial gain or

loss in the secondary structure of the hairpin loop connecting 6 and 7 (Gly 100-Tyr 107)

induced by mutation of residues proximal to the heme suggest that information is relayed

between this loop and the Tyr 75 loop.

138  

Comparison of the heme binding motif between wild type and the mutant proteins shows

that substitution of Ala for the heme proximal residues induces drifting apart of the His 32 and

Tyr 72 loops and conformational rearrangement of residues constituting the His 32 loop. This

observation indicates that the conserved heme iron ligands (Tyr 75 and His 83) in wild type

protein are crucial in maintaining the architecture of the heme binding pocket. The fact that

residues in the His 32 loop are most affected relative to those in the Tyr 75 loop strongly suggests

that communication between the heme pocket and the rest of the polypeptide proceed via this

loop. Further characterization of HasAp (and its mutants) in complex with the cognate receptor

HasR is necessary to shed more light on this line of thought.

The high resolution x-ray crystal structures afforded unequivocal determination of

coordination state of the heme iron in both mutants. Under our experimental conditions, the

heme iron is hexa-coordinate. In Y75A, the sixth ligand is a formate ion while in H83A it is Tyr

75. These findings bring up two important points; i, substitution of a small amino acid residue for

any one of the heme ligands relief the binding cavity from steric cluster and in turn potential

heme ligands may bind the heme. This contradicts the argument attributing restricted

accessibility of heme to anionic ligands in wild type holo-HasAs and holo-H83A-HasAs to the

hydrophobic nature of the hemin pocket [7]. Sequence alignment of Tyr 75 loop residues of

HasAp and HasAs reveal that amino acid residues surrounding the heme pocket are pretty much

similar in both proteins. Besides, we earlier observed that holo H32A HasAp mutant also bind an

anionic ligand (cyanide) albeit weakly. ii, In the absence of Tyr 75, His 83 does not necessarily

serve as an alternate ligand but rather is coordinated to the heme iron via the intermediacy of a

ligand in the proximal site (in our case, a formate ion). This observation contradicts the proposal

that in the absence of Tyr 75, His 83 may coordinate the heme. Studies are underway to obtain

139  

crystals of the Y75A mutant without the chance ion to determine if the coordination state of the

heme iron would be any different from the observations made in this study. In both crystal forms

of His 83, His 32 and Tyr 75 are not detached from the protein. Nonetheless, the nature of

molecule bound to the Oη of Tyr 75 appears to have a profound effect on the overall structure of

the protein, particularly on the beta sheet wall. There is a loss of beta sheet character in both

crystal forms; however, in the structure in which Tyr 75 is coordinated by a water molecule, the

lengths of three beta strands and a helix are significantly reduced. It is possible that the polar

interactions between Tyr and the fortuitous ions (EG) in crystal form 1 stabilizes the Tyr 75 Oη-

Fe3+ bond albeit to a lesser extent compared to the stability imparted by His 83 in the wild type

protein. Crystal structures obtained at low pH would be important to help identify the axial

ligand that ionizes during acid-base transition as depicted in the pH titration results.

140  

Appendices

Appendix 1: Protein crystallization conditions-Wizard II Random Sparse Matrix

Kit# crystallant buffer (0.1 M) salt (0.2 M)

1 10% (w/v) PEG-3000 acetate pH 4.5 Zn(OAc)2

2 35% (v/v) 2-methyl-2,4-pentanediol MES pH 6.0 Li2SO4

3 20% (w/v) PEG-8000 Tris pH 8.5 MgCl2

4 2.0 M (NH4)2SO4 cacodylate pH 6.5 NaCl

5 20% (v/v) 1,4-butanediol HEPES pH 7.5 NaCl

6 10% (v/v) 2-propanol phosphate-citrate pH 4.2 Li2SO4

7 30% (w/v) PEG-3000 Tris pH 7.0 NaCl

8 10% (w/v) PEG-8000 Na/K phosphate pH 6.2 NaCl

9 2.0 M (NH4)2SO4 phosphate-citrate pH 4.2 none

10 1.0 M (NH4)2HPO4 Tris pH 8.5 none

11 10% (v/v) 2-propanol cacodylate pH 6.5 Zn(OAc)2

12 30% (v/v) PEG-400 cacodylate pH 6.5 Li2SO4

13 15% (v/v) ethanol citrate pH 5.5 Li2SO4

14 20% (w/v) PEG-1000 Na/K phosphate pH 6.2 NaCl

15 1.26 M (NH4)2SO4 HEPES pH 7.5 none

16 1.0 M sodium citrate CHES pH 9.5 none

17 2.5 M NaCl Tris pH 7.0 MgCl2

18 20% (w/v) PEG-3000 Tris pH 7.0 Ca(OAc)2

19 1.6 M NaH2PO4/0.4 M K2HPO4 phosphate-citrate pH 4.2 none

20 15% (v/v) ethanol MES pH 6.0 Zn(OAc)2

141  

21 35% (v/v) 2-methyl-2,4-pentanediol acetate pH 4.5 none

22 10% (v/v) 2-propanol imidazole pH 8.0 none

23 15% (v/v) ethanol HEPES pH 7.5 MgCl2

24 30% (w/v) PEG-8000 imidazole pH 8.0 NaCl

25 35% (v/v) 2-methyl-2,4-pentanediol HEPES pH 7.5 NaCl

26 30% (v/v) PEG-400 CHES pH 9.5 none

27 10% (w/v) PEG-3000 cacodylate pH 6.5 MgCl2

28 20% (w/v) PEG-8000 MES pH 6.0 Ca(OAc)2

29 1.26 M (NH4)2SO4 CHES pH 9.5 NaCl

30 20% (v/v) 1,4-butanediol imidazole pH 8.0 Zn(OAc)2

31 1.0 M sodium citrate Tris pH 7.0 NaCl

32 20% (w/v) PEG-1000 Tris pH 8.5 none

33 1.0 M (NH4)2HPO4 citrate pH 5.5 NaCl

34 10% (w/v) PEG-8000 imidazole pH 8.0 none

35 0.8 M NaH2PO4/1.2 M K2HPO4 acetate pH 4.5 none

36 10% (w/v) PEG-3000 phosphate-citrate pH 4.2 NaCl

37 1.0 M K/Na tartrate Tris pH 7.0 Li2SO4

38 2.5 M NaCl acetate pH 4.5 Li2SO4

39 20% (w/v) PEG-8000 CAPS pH 10.5 NaCl

40 20% (w/v) PEG-3000 imidazole pH 8.0 Zn(OAc)2

41 2.0 M (NH4)2SO4 Tris pH 7.0 Li2SO4

42 30% (v/v) PEG-400 HEPES pH 7.5 NaCl

43 10% (w/v) PEG-8000 Tris pH 7.0 MgCl2

142  

44 20% (w/v) PEG-1000 cacodylate pH 6.5 MgCl2

45 1.26 M (NH4)2SO4 MES pH 6.0 none

46 1.0 M (NH4)2HPO4 imidazole pH 8.0 NaCl

47 2.5 M NaCl imidazole pH 8.0 Zn(OAc)2

48 1.0 M K/Na tartrate MES pH 6.0 none

All formulations are made with ultrapure ASTM Type I water and sterile-filtered stock solutions.

Store at 4-25 ºC.

143  

Appendix 2: ESI/MS spectrum of Y75A-HasAp

144  

Appendix 4: ESI/MS spectrum of H83A-HasAp

145  

References

1. Alontaga, A., Rodriguez, J., Schönbrunn, E., Becker, A., Funke, T., Yukl, E., Hayashi,

T., Stobaugh, J., Moënne-Loccoz, P., and Rivera, M., Structural Characterization of the Hemophore HasAp from Pseudomonas aeruginosa: NMR Spectroscopy Reveals Protein-Protein Interactions between Holo-HasAp and Hemoglobin. Biochemistry, 2009. 48: p. 96-109.

2. Arnoux, P., Haser, R., Izadi, N., Lecroisey, A., Delepierre, M., Wandersman, C., and

Czjzek, M. The Crystal structure of HasA, a hemophore secreted by Serratia marcescens. Nat. Struct. Biol., 1999. 6: p. 516-520.

3. Wolff, N., Deniau, C., Letoffe, S., Simenel, C., Kumar, V., Stojiljkovic, I., Wandersman,

C., Delepierre, M., and Lecroisey, A. Histidine pKa shifts changes of tautomeric states induced the binding of gallium-protoporphyrin IX in the hemophore HasAsm. Protein Sci., 2002. 11: p. 757-765.

4. Caillet-Saguy, C., Delepierre, M., Lecroisey, A., Bertini, I., Piccioli, M., and Turano, P.

Direct-Detected 13C NMR to Investigate the Iron(III) Hemophore HasA. J. Am. Chem. Soc., 2006. 128: p. 150-158.

5. Jepkorir, G., Rodríguez, J.C., Rui, H., Im, W., Lovell, S., Battaile, K.P., Alontaga, A.Y.,

Yukl, E.T., Moënne-Loccoz P., and Rivera, M., Structural, NMR Spectroscopic, and Computational Investigation of Hemin Loading in the Hemophore HasAp from Pseudomonas aeruginosa. J., Am Chem Soc., 2010. 132: p. 9857-9872.

6. Wolff, N., Izadi, N., Couprie, J., Habeck, M., Linge, J., Rieping, W., Wandersman, C.,

Nilges, M., Delepierre, M., and Lecroisey, A., Comparative analysis of structural and dynamic properties of the loaded and unloaded hemophore HasA: functional implications. J. Mol. Biol., 2008. 376: p. 517-525.

7. Deniau, C., Gilli. R., Izadi, N., Létoffé, S., Delepierre, M., Wandersman, C., Briand, C.,

Lecroisey, A., Thermodynamics of heme binding to the HasA(SM) hemophore: effect of mutations at three key residues for heme uptake. Biochemistry, 2003. 42: p. 10627-10633.

8. Letoffe, S., Deniau, C., Wolff, N., Dassa, E., Delepelaire, P., Lecroisey, A., and

Wandersman, C. Haemophore-mediated bacterial haem transport: evidence for a common or overlapping site for haem-free and haem-loaded haemophore on its specific outer membrane receptor. Mol. Microbiol., 2001. 41: p. 439-450.

9. Caillet-Saguy, C., Turano, P., Piccioli, M., Lukat-Rodgers, G. S., Czjzek, M.,

Guigliarelli, B., Izadi-Pruneyre, N., Rodgers, K. R., Delepierre, M., and Lecroisey, A.,

146  

Deciphering the structural role of histidine 83 for heme binding in hemophore HasA. J Biol Chem., 2008. 283: p. 5960-5970.

 

10. Lukat-Rodgers, G. S., Rodgers, K. R., Caillet-Saguy, C., Izadi-Pruneyre, N., Lecroisey, A., Novel heme ligand displacement by CO in the soluble hemophore HasA and its proximal ligand mutants: implications for heme uptake and release. Biochemistry, 2008. 47: p. 2087-2098.

11. Krieg, S., Huche´, F., Diederichs, K., Izadi-Pruneyre, N., Lecroisey, A., Wandersman, C., Delepelaire, P., and Welte, W., Heme uptake across the outer membrane as revealed by crystal structures of the receptor-hemophore complex. Proc. Natl. Acad. Sci. U.S.A., 2009. 106: p. 1045-1050.

12. Ishikawa, K., Takeuchi, N., Takahashi, S., Matera, K. M., Sato, M., Shibahara, S., Rousseau, D. L., Ikeda-Saito, M., and Yoshida, T., Heme Oxygenase-2 properties of the heme complex of the purified tryptic fragment of recombinant human heme oxygenase-2. J. Biol. Chem., 1995. 270: p. 6345-6350.

 

13. Falk, J. E., 1963 Part A: pyrrole pigments: chemistry and biochemistry of porphyrins and metalloporphyrins, p. 3–33. In M. Florkin, and E. H. Stotz (ed.), Comprehensive biochemistry. 9. Elsevier Publishing Company, Amsterdam.

14. Kabsch, W., Automatic indexing of rotation diffraction patterns. Journal of Applied

Crystallography, 1988. 21: p. 67-72. 15. Evans, P., Scaling and assessment of data quality. Acta Crystallogr D Biol Crystallogr,

2006. 62: p. 72-82. 16. Vagin, A. A., and Teplyakov, A., MOLREP: an Automated Program for Molecular

Replacement J. Appl. Cryst., 1997. 30: p. 1022-1025. 17. Adams, P.D., et al., PHENIX: a comprehensive Python-based system for macromolecular

structure solution. Acta Crystallogr D Biol Crystallogr, 2010. 66: p. 213-221. 18. Emsley, P. and K. Cowtan, Coot: model-building tools for molecular graphics. Acta

Crystallogr D Biol Crystallogr, 2004. 60: p. 2126-2132. 19. Lovell, S.C., et al., Structure validation by Calpha geometry: phi,psi and Cbeta

deviation. Proteins, 2003. 50: p. 437-450. 20. Potterton, L., et al., Developments in the CCP4 molecular-graphics project. Acta

Crystallogr D Biol Crystallogr, 2004. 60: p. 2288-2294. 21. The PyMOL Molecular Graphics System, Version 1.3, Schrödinger, LLC. 

147  

22. Chayen, N. E., and Saridakis, E., Protein crystallization: from purified protein to

diffraction-quality crystal. Nature methods, 2008. 5: p. 147-153. 23. McPherson, A., Introduction to protein crystallization. Methods, 2004. 34: p. 254-265. 24. Chayen, N. E., Methods for separating nucleation and growth in protein crystallization.

2005. Protein Crystallisation, 2005. 88: p. 329-337. 25. Thaku, A. S., et al, Improved Success of Sparse Matrix Protein Crystallization Screening

with Heterogeneous Nucleating Agents. PLoS ONE, 2007. 2: e1091

148  

CHAPTER IV

Methemalbumin is the potential target for the hemophore HasAp as a heme

source; inferences from heme transfer studies

Introduction

Requirement for iron by invading microbial organisms is a necessity for growth and for

successful colonization of mammalian host cells [1]. However, the human host presents a hostile,

iron restrictive environment to bacterial invaders [2]. Free iron is toxic to the cells and hence it is

often sequestered by iron-binding proteins, or globins, or is sequestered within the cytoplasm of

cells in the form of iron-sulfur clusters, cytochromes or as iron mineral in iron storage enzymes

[3,4]. The binding affinity of iron to these proteins is so high that the amount of free iron

available (< 10-18 M) is significantly below the minimum concentration required for bacterial

growth [5]. Accordingly, bacteria have evolved sophisticated mechanisms to sequester this

important element from these sources and thus fulfill their nutritional requirement.

One of the mechanisms employed by the opportunistic pathogen Pseudomonas

aeruginosa to acquire iron is secretion of a hemophore (HasAp) to the extracellular milieu which

binds hemin (free or globin-bound) and delivers it to the cognate outer membrane receptor

(HasR) for subsequent internalization [6]. Once in the cytosol, hemin is thought to be degraded

by heme oxygenase (pa-HO) to release iron, making the nutrient available to bacteria [7,8].

However, like free iron, heme and heme carrier proteins particularly hemoglobin (outside the

constraints of the red blood cells) are toxic substances and mammalian organisms have evolved

committed molecular pathways to ensure rapid removal once they appear in plasma [9, 10]. Two

149  

major host defense systems that work hand in hand are (i) hemoglobin-binding haptoglobin (Hp)

and its cognate receptor CD163 [11], and (ii) hemin-binding hemopexin (HPX) and the low

density lipoprotein receptor-related protein/LRP1 [12, 13, 14]. Under normal circumstances, the

heterotetrameric glycoprotein Hp binds hemoglobin dimers in a 1:1 stoichiometry to form stable

nontoxic complexes [15] of very high affinity (Kd = 10-23 M) which are then cleared from

circulation via receptor mediated endocytosis [16,17,18]. In situations with pronounced

hemolysis, though (as is the case during bacterial infection), the amount of free hemoglobin in

plasma overwhelms the binding capacity of Hp, and thus the excess hemoglobin rapidly

autoxidizes and spontaneously liberates the heme. Kinetic analysis of time-dependent hemin

distribution in plasma proteins show that within the first seconds after hemin appears in plasma,

more than 80 % of total hemin binds initially to the oxidatively intolerant lipoproteins (low

density lipoprotein, LDL and high density lipoprotein, HDL), and only 20 % binds to the

antioxidants human serum albumin (HSA) and hemopexin (HPX) [19]. Over time, hemin

molecules bound to lipoproteins partially transfer to HSA and HPX. HSA is the most abundant

serum protein (35-55 mg/ mL) and only acts as a transient depot before hemin is transferred to

the carrier protein HPX (0.5-1.2 mg/mL) for subsequent removal from circulation [20,21]. Given

its abundance in plasma and the comparatively weak association of HSA with heme (HSA; Kd =

10-8 M, HPX; Kd = 10-12M) [22], it is probable that increase of hemin plasma levels under

pathological conditions could saturate the carrier hemoprotein HPX leaving substantial amount

of hemin in the reservoir HSA accessible to invading microbes. These, taken together, made us

postulate that methemalbumin (heme loaded serum albumin) is a probable target for HasAp as a

heme source when p. aeruginosa is faced with limiting iron conditions.

150  

To date, hemin acquisition by hemophores has mainly focused on hemoglobin as a heme

source. The extensively characterized hemophore from S.marcescens HasAs is able to use

hemoglobin-heme as an iron source [23]. Similarly, HasAp from P.aeruginosa and HasAyp from

Y. pestis produced in E.coli are able to acquire heme from hemoglobin [24,25]. The hemophore

system described in H.influenzae Type b, HxuA, which does not share sequence similarity with

HasAs, is able to assimilate heme from the serum glycoprotein HPX and shuttle it to a specific

outer membrane receptor [26]. In this study, we demonstrate for the first time that the hemophore

HasAp from p.aeruginosa is able to acquire hemin from serum albumin. The mechanism of

hemin binding and/or transfer remains to be investigated.

Experimental Procedures

Expression and Purification of apo-HasAp

Expression and purification of apo-HasAp was carried out as described in the

experimental section of chapter 2. Briefly, a single colony of freshly transformed E.coli BL21

(DE3) cells were cultured overnight in 50 mL LB-medium containing 100 g/mL ampicillin. 30

mL of this starter culture was diluted in 1 L M9 medium supplemented with 100 g/mL

ampicillin and grown until OD600 reached approximately 1. Polypeptide synthesis was induced

by addition of isopropyl--D-thiogalactopyranoside (IPTG) to a final concentration of 1 mM

before the cells were grow for an additional 5 h, pelleted by centrifugation (4800 rpm, 12 min,

and 4 0C) and stored in -20 0C. Cell lysis and protein purification were carried out as detailed in

chapter II.

151  

Preparation of Human Serum Albumin (HSA)

A lyophilized sample of “globin and fatty-acid” free human serum albumin (Sigma

Aldrich) was dissolved in 50 mM sodium phosphate buffer, pH 7.4. The concentration of the apo

protein was estimated spectrophotometrically using Ɛ280 = 36.5 mM-1cm-1 [27]. Immediately, the

sample was titrated with 1 mM stock solution of freshly prepared heme in DMSO to a final ratio

of 1:1.2 [28]; the total amount of heme in DMSO in the resultant solution was < 0.9 % (V/V).

After an overnight incubation at 4 0C, the HSA-hemin complex (methemalbumin) solution was

concentrated using 50 kDa MWCO Amicon ultracentrifuge filter units (Millipore) and passed

through a gel filtration column (Sephadex G-75 superfine, GEHealthcare) pre-equilibrated with

50 mM sodium phosphate buffer pH 7.0. The purity of the sample was assessed by sodium

dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and pure fractions were

concentrated and kept frozen in -20 0C until needed.

Quantitation of heme in methemalbumin

The amount of hemin incorporated into the protein was determined as pyridine

hemochromes using = 20.7 mM-1cm-1 [29]. Concisely, 200 L of neat pyridine was added to 6

M of methemalbumin dissolved in 800 L sodium phosphate buffer, pH 7.4, then the resultant

solution was divided into two half parts. A few grains of sodium dithionite was added to one of

the solutions and dissolved by gentle agitation of the cuvette to avoid aerating the solution.

Immediately, readings were taken at the expected -maximum (557 nm), then at the maximum a

little more dithionite was added. If the absorbance at the maximum increased, the process was

rapidly repeated until there was no further increase. The absorbance intensity of the reduced

152  

hemochrome was then recorded at 557 nm. To the other half, 15 L of 1 mM K3Fe(CN)6 was

added, mixed gently, then the absorbance intensity at the minimum (541 nm) was recorded.

Concentration of hemin was determined using and the magnitude of the difference in

absorbance intensity between reduced and oxidized hemochrome.

Hemen transfer assays

To determine whether apo-HasAp can capture hemin from methemalbumin, hemin

transfer assay was done as follows: Methemalbumin (10 µM) was incubated with constant

stirring at 37 0C in 50 mM sodium phosphate buffer pH 7.0. After 5-10 minutes, equimolar

amount (or 4-8 fold excess) of apo-HasAp was added and the reaction was monitored as a

function of time. Changes in absorbance measurements were performed on UV-vis

spectrophotometer at various time intervals after mixing or alternatively the evolution or decay

of absorbance at fixed wavelengths was monitored. Rates of hemin transfer were measured by

fitting a double exponent curve to observed decay in absorbance at 500 nm and 535 nm or to the

observed growth of a peak at 618 nm (corresponding to evolution of a CT band of holo-HasAp).

Evaluation of heme transfer by polyacrylamide gel electrophoresis (PAGE)

The reaction mixture was analyzed by nondenaturing-PAGE to allow complete separation

and identification of the reaction products. In this case, aliquots were taken at certain time

intervals as the reaction progressed, immediately flash frozen in liquid nitrogen, and thawed just

before being loaded onto the gel. Two matching gels were run concurrently and assessed for the

153  

presence of protein (or heme) bands by coomassie brilliant-blue R-250 (Coomassie-G) staining

or by staining with 3, 3`, 5, 5`-tetramethylbenzidine (TMB).

Quantitation of heme transfer

The extent of hemin transfer from methemalbumin to apo-HasAp was estimated by

determining the amount of holo-HasAp formed and comparing it to the starting concentration of

the apo-protein. To achieve this, the reaction mixture was incubated for 20 minutes under the

same experimental conditions and immediately loaded onto a size exclusion column (Sephadex

G-75 superfine, GEHealthcare to separate serum albumin from HasAp. Fractions containing

HasAp were pooled and the concentration of holo-HasAp was determined using 407nm = 136

mM-1cm-1.

Results and Discussion

Sample preparation

Purification of apo-HasAp was carried out following a similar chromatographic method

described in detail in Chapter II. This afforded a high quality hemin free sample suitable for

hemin transfer studies described herein. Hemin concentration in methemalbumin was determined

using ΔƐ = (Ɛ557-Ɛ541) = 20.7 mM-cm-1 from the dithionite-reduced minus ferricyanide-oxidized

pyridine hemochrome spectra. Hemin content of a 6 µM sample was found to be 5.6 µM giving

a 1: 0.93 stoichiometry of serum albumin and heme; in good agreement with a 1:1 heme protein

ratio.

154  

Spectroscopic changes observed upon incubation of apo-HasAp with methemalbumin

Electronic absorption spectra of methemalbumin and holo-HasAp recorded at pH 7.0 are

shown in Figure 1. The spectra are very distinct in the visible region; methemalbumin absorbs at

404, 500, 535, and 623 nm, while holo-HasAp absorbs at 407, 504, 537, 574 and 618 nm. The

differences in the electronic absorption signature of the two proteins are useful in tracking hemin

as it exchanges between the two proteins in solution.

When methemalbumin is incubated with four-fold excess apo-HasAp (Figure 2), absorbance at

500 and 535 nm dramatically decreases. The decay in these peaks corresponds to decrease in the

absorbance of methemalbumin-heme as the heme is transferred to apo-HasAp. Formation of

holo-HasAp is evidenced by concomitant blue shift of the charge transfer band at 623 nm leading

to accumulation of a species absorbing at 618 nm. The latter has been identified as an

intermediate holo-HasAp species whereby the heme is loaded onto the Tyr 75 loop before the

distal ligand coordinates (appearance of a peak at 574 nm) at the sixth position. In the soret

region, the absorbance band red shifts from 404 to 407 nm, and becomes narrower and more

intense as the reaction progresses. This is indicative of accumulation of hexacoordinate holo-

HasAp in solution. A clear isosbestic point at 393 nm reveal that only two species are present in

solution namely; unreacted methemalbumin and holo-HasAp. The fact that we do not see any

baseline drift as the reaction progresses indicates that heme released from methemalbumin is not

accumulating in solution to detectable levels but rather rapidly binds apo-HasAp.

155  

 

Wavelength (nm)

500 550 600 650 700

Wavelength (nm)

250 300 350 400

Abs

orb

anc

e

0.0

0.2

0.4

0.6

0.8

1.0

1.2

500

404 > 407

535

618 < 623574

5X

Superposition of the reaction product (red trace) and HasAp reconstituted with free heme

(black trace) is shown in Figure 3. Clearly, the spectra are nearly identical except for the

difference in intensity of the soret band. Three different sets of experiments where temperature,

apo-HasAp concentration or incubation time were varied yielded similar results. This suggests

Figure 1. UV-vis spectra of methemalbumin (black), and holo-HasAp (red).

Figure 2. A family of electronic absorption spectra of methemalbumin (~10 M) incubated with apo-

HasAp (40 M) at 37 o

C recorded as a function of time. Arrows show the direction of spectral changes as

the reaction progress.

Wavelength

250 300 350 400

Ab

sorb

ance

0.0

0.2

0.4

0.6

0.8

1.0

Wavelength

500 550 600 650 700

156  

that the significant attenuation of the Soret band is not dependent on these experimental

parameters but likely a reflection of changes in properties intrinsic to the protein. To pursue this

notion, the reaction mixture was passed through a gel filtration column to separate the reaction

products (Figure 4) then the molar absorptivity of holo-HasAp was determined by Beers law

using absorbance intensity of the isolated holo-HasAp fraction and the concentration of

corresponding reduced pyridine-hemochromes. Interestingly, absorption coefficient of holo-

HasAp separated from the reaction mixture is comparatively lower (407 = 136 mM-1cm-1) than

that of HasAp reconstituted with free heme in solution (407 = 169 mM-1cm-1); it follows then that

at similar concentrations, the two samples would exhibit different intensities at the soret region

as seen in Figure 3. Based on these observations, we conclude that the local environment of

hemin in HasAp is different when the protein acquires hemin from hemoproteins and when it is

reconstituted with free heme.

Wavelength

250 300 350 400

Abs

orba

nce

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Wavelength

500 550 600 650 700

Holo-HasAp

Reaction product

Methemalbumin

Figure 3. UV-Vis spectra of the reaction product (red) and holo-HasAp (black) with similar

concentrations. The spectrum of methemalbumin (blue) is shown for reference.

157  

Heme uptake by apo-HasAp correlates with heme loss from methemalbumin

The affinity of heme for HasA (Ka = 5.3 X 1010 M-1) [30], which shares 50 % similarity

with HasAp is relatively higher than that of serum albumin (Ka = 1.1 X 107-8 M-1) [31,32]. When

equimolar mixture of apo-HasAp and serum albumin is titrated with free heme in solution, hemin

preferentially binds the former, in good agreement with the reported affinities. Figures 5 and

Figure 6 shows representative plots of changes in absorbance at the soret and at the visible region

recorded as a function of time when methemalbumin and apo-HasAp were incubated in a 1: 4

molar ratio, respectively. As shown, transfer occurs in a biphasic mode. The fast phase, depicted

by loss of hemin from methemalbumin (decay in absorbance at 500 and 535 nm peaks), occurs

concurrently with heme uptake by apo-HasAp and accumulation of species absorbing at 618 nm,

 

Fraction #

10 20 30 40 50

Abs

orb

anc

e

0.0

0.5

1.0

1.5

2.0

2.5

Figure 4. Elution profile of serum albumin and HasAp from a Sephadex G-75 column;

blue circles track A280nm while red circles track the soret.

158  

(Figure 6) and rapid increase in absorbance at 407 nm, (Figure 5).The species absorbing at 618

Time (mins)

0 5 10 15 20 25 30 35

Abs

@ 5

00 n

m

0.185

0.190

0.195

0.200

0.205

0.210

0.215

0.220

0.225

Abs

@ 5

35 n

m

0.15

0.16

0.17

0.18

0.19

0.20

0.21

Abs

@ 5

74 n

m

0.120

0.122

0.124

0.126

0.128

0.130

0.132

0.134

Abs

@ 6

18 n

m

0.114

0.116

0.118

0.120

0.122

0.124

0.126

0.128

0.130

0.132

0.134

Time(min) vs abs @ 500 nm Time(min) vs abs @ 535 nm Time(min) vs abs @ 574 nmTime(min) vs abs @ 618 nm

 

Time (min)

0 5 10 15 20 25 30

ab

sorb

ance

at

407

nm

0.00

0.05

0.10

0.15

0.20

0.25

Figure 5. A plot showing changes in absorbance at 407 nm when methemalbumin is incubated

with 4-fold excess apo-HasAp with constant stirring at 37 oC.

Figure 6. A plot showing absorbance changes occurring in the visible region when

methemalbumin was incubated with four-fold apo-HasAp. The reaction was monitored as a

function of time at 500, 535, 574, and 618 nm.

159  

which has been identified as an intermediate heme-HasAp complex whereby heme is loaded onto

the Tyr 75 loop [33].The slow phase which occurs around 10 min after the onset of the

experiment corresponds to coordination of hemin-iron at the sixth position by His 32. This is

corroborated by the near linear absorbance changes occurring at 574 nm (blue circles) which

gradually increases for about 20 min before it levels off when the HasAp species is

hexacoordinated. The fact that hemin loss from methemalbumin and up-take by apo-HasAp

occur concomitantly suggests two possibilities; (i) that the transfer involves protein-protein

interaction (PPI) whereby apo-HasAp favorably interacts with methemalbumin and “steals”

heme, in an activated complex, or (ii) PPI induces conformational changes in the secondary

structure of methemalbumin releasing the macrocycle, which in turn, is rapidly consumed by

apo-HasAp. A double exponential function fitted to experimental data (absorbance at 618 nm in

Figure 7) reveals that the rate of hemin uptake by apo-HasAp (k-1 = 0.1939 min-1) is comparable

to that of hemin dissociation from methemalbumin in aqueous media (K-1 = 0.198 min-1) [34].

Although low protein concentrations used in this experiment (at most 40 M) do not mimic that

in a host cell, these results do not favor hemin transfer in an activated complex; otherwise the

observed rate of hemin release from methemalbumin would be higher. It therefore appears that

hemin transfer is passive and the rapid up-take by apo-HasAp is affinity driven. More studies

need to be done to elucidate the mechanism of transfer under conditions that closely mimic the

host cell.

From the UV-Vis data, it can be concluded that HasAp can capture hemin from

methemalbumin and that the latter is the potential target for the homophore when bacteria is

faced with limiting iron conditions in the host cell. Complementary data were obtained by gel

electrophoresis.

160  

Evaluation of heme transfer by PAGE.

Electrophoresis was used as an alternative method to separate the reaction products. The

advantage of this method lies in the possibility to tweak the composition of the gel in order to

improve resolution of components in the reaction mixture, as well as do multiple sample analysis

using much smaller sample quantities. In a nondenaturing polyacrylamide gel, samples are

separated based on both size and charge. Besides, hemophores migrate without dissociating from

heme and thus the apo and holo form of the protein can be resolved.

Figure 7 shows results obtained from a reaction mixture of apo-HasAp and

methemalbumin incubated for 2, 5, 10, 20, 30, 60 min and run in 13 % native polyacrylamide

gel. Figure 8-A is a picture of the gel taken immediately following the run before staining. Only

two sets of bands corresponding to serum albumin and HasAp are seen. Whereas the intensity of

HSA band decreases across the gel, that of HasAp increases. This implies that these bands

correspond to the hemin-bound proteins whereby the concentration of methemalbumin decreases

over time as hemin is being transferred to HasAp. The approximate time around which the

reaction is complete was best estimated from the stained gels.

Heme staining using tetramethylbenzidine (TMB) (Figure 7-B), reveals high molecular

weight bands corresponding to hemin-bound HSA. The intensity of these bands decreases

drastically after 10 minutes and completely disappears between 10 and 20 minutes; in agreement

with the UV-vis data where heme loss from the hemoprotein and subsequent binding to the Tyr

75 loop in HasAp occurs within the first 10 minutes. TMB is specific for peroxidases and so

even if substantial amount of nascent holo-HasAp accumulates during the reaction, heme in

161  

HasAp could not be stained to any detectable extent at these concentrations. Saturating amounts

of holo-HasAp reconstituted with free heme (lane C` in Figure 7-B) was used as a control.

The most interesting result is depicted in Figure 7-C where three intense and well

resolved sets of proteins bands are observed in Coomassie brilliant blue R-250 (Coomassie G)

stained gel. Both heme-loaded and heme-free serum albumin could not be resolved and thus

travel the same distance in the gel. In the case of HasAp, the presence of hemin caused a

significant increase in the electrophoretic mobility of heme-bound fraction allowing complete

separation of the apo and holo forms. The intensity of the holo-HasAp band reaches the

maximum in about 10 minutes and does not change thereafter; in contrast, that of the apo-form

seem to be consistent throughout the reaction time. This is because an excess of apo-HasAp was

used with the intent of pushing the equilibrium toward formation of the products. To present

these results more clearly, the same experiment was repeated under similar conditions but using

equimolar amounts of proteins. Apo-HasAp protein bands (Figure 8-A) diminish drastically after

7 min and almost disappears completely in 12 min. A matching SDS gel (Figure 8-B), shows that

intact serum albumin and HasAp subunits are present in solution and there is no indication of

protein degradation or aggregation. The more intense HSA bands, albeit of the same

concentration as HasAp was due to the difference in the proportion of basic amino acids (lysines,

arginines and histidines) in the two proteins that bind Coomassie G dye molecules [35]; HSA has

five times the number of these amino acid residues relative to HasAp and therefore stains better.

Figure 9 shows an elution profile of the reaction mixture passed over a gel filtration column

following 10 min incubation. Taking into account any protein loss, nearly 80 % heme transfers to

HasAp.

162  

 

holo-HSA

4 fold holo-HasAp

A

B

HSA

HasAp

Dye-front

C 2 5 10 20 30 60 C` Incubation time (min)

C

HSA

apo-HasAp

holo-HasAp

Figure 7. 13 % native polyacrylamide gel showing the reaction products obtained from a

mixture of apo-HasAp and methemalbumin incubated for 2, 5, 10, 20, 30, 60 min at 37 oC.

163  

A

B HasAp

C C` 1 3 5 7 12 20 S S`

HSA

apo-HasApholo-HasAp

HSA

Incubation time (min)

Figure 8. 13 % Native gel (A) and SDS gel (B), stained with coomassie brilliant blue. Samples were

aliquoted at certain time intervals (1, 3, 5, 7, 12, 20 min), immediately flash frozen in liquid nitrogen, and

thawed just before loading. Lanes C, C`, S, and S` are controls apo-HSA, methemalbumin, apo-HasAp,

and holo-HasAp, respectively.

Figure 9. Elution profile of serum albumin and HasAp from a Sephadex G-75 column;

blue circles track A280nm while red circles track the soret.

Fraction #

0 5 10 15 20 25 30 35

Ab

sorb

ance

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

164  

Heme transfer from methemalbumin to H32A-HasAp mutant.

A similar experiment but using H32A variant in place of wild type protein was conducted

to determine whether the mutant can sequester heme from methemalbumin. UV-vis data (Figure

10) reveal conceivable changes in absorbance spectra when the two proteins are incubated with

constant stirring. However, it is not possible to explicitly track hemin exchange between the two

proteins because the holo-forms absorb around the same wavelengths. What is clear is that the

soret band blue shifts from 404 nm (lamda max of methemalbumin) to 402 nm (lamda max of

H32A-HasAp monomer), while the charge transfer band shifts to a higher wavelength (630 nm).

As described in chapter II, monomeric H32A-HasAp, soret and charge transfer bands are at 402

and 621 nm, while those of the dimer absorb at 400 and 623 nm, respectively. In the nascent

holo-H32A-HasAp species observed in the reaction mixture, the soret band matches that of the

monomer while the charge transfer band does not match any of the two species. Thus, unlike the

protein reconstituted with free heme, it is not possible to simply identify the species in solution

by UV-vis spectroscopy when heme is acquired from methemalbumin. To further characterize

this specie, the reaction mixture was passed over a high resolution gel filtration column

(Superdex G-75 Superfine) calibrated with conalbumin, ovalbumin, carbonic anhydrase, and

ribonuclease A protein standards. Using elution volume of isolated holo-H32A-HasAp (Ve = 98

mL) and the calibration curve, the molecular weight of holo-H32A-HasAp was determined to be

21.7 kDa, thus it is a monomeric species (Figure 11). UV-vis spectra of pooled holo-H32A-

HasAp fractions and holo-H32A-HasAp monomeric sample reconstituted with free heme in

solution match, in good agreement with the gel filtration data.

165  

Wavelength (nm)

250 300 350 400

Ab

sorb

ance

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Wavelength (nm)

0 500 550 600 650 700

623 > 630

402 < 404

Ve (mL)

60 80 100 120

log

MW

0.8

1.0

1.2

1.4

1.6

1.8

2.0

2.2

Ribonuclease A

HSA Carbonic anhydrase

Ovalbumin

H32A-HasAp

Conalbumin

Figure 10. UV-Vis spectra of the reaction product (red) and holo-HasAp (black) with similar

concentrations. The initial spectrum spectrum of methemalbumin (blue) is shown for reference.

Figure 11. Elution volumes (Ve) from a calibrated Superdex 75 column and estimated molecular

weights for the reaction products obtained from a reaction mixture of methemalbumin and apo-

H32A-HasAp: holo-H32A-HasAp (21 kDa, green diamond), and serum albumin ( 63 kDa, red

triangle). The Superdex 75 column was calibrated with ribonuclease A (75 kDa), Carbonic

anhydrase (44 kDa), ovalbumin (29 kDa), and conalbumin (13 kDa) shown in black filled circles.

166  

It therefore appears that before the reaction products are isolated, coordination of heme in the

mutant is different from that of the established monomer or dimer when this mutant binds free

heme in solution. It is possible that since the distal heme iron ligand (His 32) is missing in the

mutant, heme coordination ligands in nascent holo-H32A in the reaction mixture are contributed

by both proteins. Time dependent crystallographic studies would be useful to establish possible

interaction sites between HasAp and methemalbumin and to elucidate the coordination state of

the heme iron.

167  

References

1.   Bullen, J. J., The significance of iron in infection. Rev Infect Dis, 1981. 3: p. 1127-1138. 2. Weinberg, E. D., Iron withholding: a defence against infection and neoplasia. Physiol

Rev, 1984. 64: p. 65-102. 3. Wagener, F. A., Volk, H. D., Willis, D., Abraham, N. G., Soares, M. P., Adema, G.J., and

Figdor, C.G., Different faces of the hemeheme oxygenase system in inflammation. Pharmacol Rev, 2003. 55: p. 551-571.

4. Arosio, P., and Levi. S., Ferritin, iron homeostasis, and oxidative damage. Free Radic

Biol Med, 2002. 33: p. 457-463. 5. Braun, V., and Killmann, H., Bacterial solutions to the iron-supply problem. Trends

Biochem Sci, 1999. 24: p. 104-109. 6. Ghigo, J. M., Letoffe, S. and Wandersman, C. A new type of hemophore-dependent heme

acquisition system of Serratia marcescens reconstituted in Eschericia coli. J. Bacteriol. 1997. 179: p. 3572-3579.

7. Tenhunen, R., Marver, H., Pimstone, N. R., Trager, W. F., Cooper, D. Y., and Schmid,

R., Enzymatic Degradation of Heme. Oxygenative Cleavage Requiring Cytochrome P-450. Biochemistry, 1972. 11: p. 1716-1720.

8. Ortiz de Montellano, P.R., and Wilks, A., Heme Oxygenase Structure and Mechanism.

Adv. Inorg. Chem., 2000. 51: p. 359-402.  

9. Grinberg, L. N., O'Brien, P. J., and Hrkal, Z., The effects of heme-binding proteins on the peroxidative and catalytic activities of heroin. Free Radic Biol Med, 1999. 27: p. 214-219.

10. Gutteridge, J. M., and Smith, A., Antioxidant protection by haemopexin of haem-

sumulated lipid peroxidation. Biochem J, 1988. 256: p. 861-865.  

11. Kristiansen, M., Graversen, J. H., Jacobsen, C., Sonase, O., Hoffman, H. J., Law, S. K., and Moestrup, S. K., Identification of the haemoglobin scavenger receptor. Nature, 2001. 409: p. 198-201.

 

12. Boucher, P., Gotthard, M., Li, W. P., Anderson, R. G., and Herz, J., LRP: role in vascular wall integrity and protection from atherosclerosis. Science, 2003. 300: p. 329-032.

168  

13. Hvidberg, V., Maniecki, M. B., Jacobsen, C., Hojrup, P., Moller, H.J., and Moestrup,

S.K., Identification of the receptor scavenging hemopexin-heme complexes. Blood, 2005. 106: p. 2572-2579.

14. Lillis, A. P., Mikhailenko, I., and Strickland, D. K., Beyond endocytosis: LRP function in

cell migration, proliferation and vascular permeability. J Thromb Haemost, 2005. 3: p. 1884-1893.

15. Ascenzi, .P, Bocedi, A., Visca, P., Altruda, F., Tolosano, E., Beringhelli, T., and Fasano,

M., Hemoglobin and heme scavenging. IUBMB Life, 2005. 57: p. 749-759. 16. Higa, Y., Oshiro, S., Kino, K., Tsunoo, H., and Nakajima, H., Catabolism of globin-

haptoglobin in liver cells after intravenous administration of hemoglobin-haptoglobin to rats. J Biol Chem, 1981. 256: p. 12322-12328.

17. Kino, K., Mizumoto, K., Watanabe, J., and Tsunoo, H., Immunohistochemical studies on

hemoglobin-haptoglobin and hemoglobin catabolism sites. J Histochem Cytochem, 1987. 35: p. 381-386.

18. Weinstein, M. B., and Segal, H. L., Uptake of free hemoglobin by rat liver parenchymal

cells. Biochem Biophys Res Commun, 1984. 123: p. 489-496. 19. Miller, Y. I., and Shaklai, N., Kinetics of hemin distribution in plasma reveals its role in

lipoprotein oxidation. Biochimica et Biophysica Acta, 1999. 1454: p. 153-164. 20. Morgan, W. T., The binding and transport of heme by hemopexin. Ann Clin Res., 1976.

8: p. 223-232. 21. Morgan, W. T., Liem, H. H., Sutor. R. P., and Muller-Ebergard, U., Transfer of heme

from heme-albumin to hemopexin. Biochim Biophys Acta, 1976. 444: p 4351145, 1976.  

22. Adams, P. A., and Berman, M. C., Kinetics and mechanism of the interaction between human serum albumin and monomeric haemin. Biochem. J. 1980. 191: p. 95-102.

23. Letoffe, S., Ghigo, J. M., and Wandersman, C. Iron acquisition from heam and

haemoglobin by Serratia marcescens extracellular protein. Proc. Natl Acad Sci USA, 1994a. 91: p. 9876-9880.

24. Letoffe, S., Redeker, V. and Wandersman, C., Isolation and characterization of an

extracellular haem-binding protein from Pseudomonas aeruginosa that shares function and sequence similarities with the Serratia marcescens HasA haemophore. Molecular Microbiology, 1998. 28: p. 1223-1234.

25. Rossi, M., et al, Identification and Characterization of the Hemophore-Dependent Heme

Acquisition System of Yersinia pestis. Infection and Immunity, 2001. 11: p. 6707-6717.

169  

26. Cope, L., Thomas, S., Hrkal, Z., and Hansen, E., Binding of heme-hemopexin complexes

by soluble HxuA protein allows utilization of this complexed heme by Haemophilus influenzae. Infect. Immun., 1998. 66: p. 4511-4516.

27. Painter, L., Harding, M. M., and Beeby, P. j., Synthesis and interaction with human

serum albumin of the first 3,18-disubstituted derivative of bilirubin. J. Chem. Soc., Perkin Trans. 1, 1998. 18: p. 3041-3044.

28. Harkal, Z., Kalousek, I., and Vodrazka, Z., Heme binding to albumin and equilibria in

the albumin-ferrihaemoglobin and albumin-haemopexin systems. Int. J. Biochem. 1980. 12: p. 619-624.

29. Fuhrhop, J. M., and Smith, K. M., (1995) In: Smith, K. M., (ed) Porphyrins and

Metalloporphyrins. Elsevier, Amsterdam, p. 757-869. 30. Deniau, C., et al. Thermodynamics of hemin binding to the HasA (SM) hemophore:

effects of mutations at three key residues for heme uptake. Biochemistry, 2003. 42: p. 10627-10633.

31. Adams, P. A., and Berman, M. C., Kinetics and mechanisms of the interaction between

human serum albumin and monomeric hemin . Biochem. J., 1980. 95: p. 95-102. 32. Beaven, G. H., Gen, S., D`Albis, A. and Gratzer W. B., Physical studies on the

ribosomal protein S2 from the Escherichia coli 30S subunit European J. Biochem., 1974. 41: p. 539-546.

33. Yukl, E. T, Jepkorir, G., Alontaga, A. Y., Pautsch, L., Rodriguez, J. C., Rivera, M., and

Moënne-Loccoz, P., Kinetic and Spectroscopic Studies of Hemin Acquisition in the Hemophore HasAp from Pseudomonas aeruginosa. Biochemistry, 2010. 9: p. 6646-6654.

34. Kuzelova, K., Mrhalova, M., and Hrkal, Z., Kinetics of heme interaction with heme-

binding proteins: The effects of heme aggregation states. Biochimica et Biophysica Acta, 1997. 1336: p. 497-501.

35. Tal, M., Silberstein, A., and Nusser, E., Why Does Coomassie Brilliant Blue R Interact

Differently with Different Proteins? J. Biol. Chem., 1980. 260: p. 9976-9980.

170  

CHAPTER V

Summary

The opportunistic pathogen Pseudomonas aeruginosa requires iron for growth and

successful colonization of mammalian hosts cells [1]. However, the host provides a hostile

environment to microbial invaders by tightly sequestering free iron in iron binding and storage

proteins thereby limiting the amount of free iron available. Accordingly, bacteria have developed

committed sophisticated mechanisms to overcome iron limitation and circumvent the host

sequestration mechanisms.

One of the ways through which Pseudomonas aeruginosa acquires iron is by deploying

the has (heme acquisition system) system which constitutes a heme binding protein HasAp and a

cell surface receptor HasR [1]. Under iron limiting conditions, Pseudomonas aeruginosa

secretes HasAp to the extracellular media where it binds heme (free or bound) and shuttles it to

the specific outer membrane HasR for subsequent internalization [2]. Once in the cytosol, the

heme is thought to be degraded by heme oxygenase to release iron which is then available to the

bacteria as an essential nutrient [3]. Since it is well established that Pseudomonas aeruginosa

requires iron to be pathogenic [4], infections caused by this bacterium can be minimized or

stopped at an early stage by depriving or limiting the amount of iron available for growth and

subsequent infections. Molecular level understanding of the mechanisms of iron or heme-iron

uptake and the protein-structure functional relationship is therefore essential. This will pave way

for development of molecules that would inhibit heme acquisition by HasAp from the host heme

binding proteins thereby curbing pseudomonas infections.

171  

To date, the mechanism of heme uptake by HasAp and heme release to its cognate

receptor HasR is not well understood. Moreover, the target host heme protein is not known. In

this study we used a combination of biochemical and spectroscopic techniques to investigate the

heme binding mechanism of HasAp. We also carried out in vitro heme transfer using human

serum albumin as the heme source. The results and conclusions drawn from experimental data

are summarized below.

1. Heme binding to apo-HasAp proceeds in two discrete sequential steps.

The X-ray crystal structure of apo-HasAp reveal that the overall polypeptide fold is

similar to that of holo HasAp except for the His 32 loop that is relocated nearly 30 Å away

relative to its position in the holo protein. Binding of free heme to form the holo conformer

proceed in two discrete sequential steps; a fast kinetic phase corresponding to loading of the

heme to the Tyr 75 loop followed by a slow phase where the His 32 loop undergoes a large

conformational rearrangement and coordinates the heme. The slow movement of the His 32 loop

is corroborated by the single x-ray crystal structure of H32A-HasAp mutant which shows a non-

crystallographic dimer (Figure 12, Chapter 2) associated via cofacial stacking of two heme

molecules.

Upon heme loading, subtle yet significant changes occur at and near the hinge residues

(Gly 28 and Gly 45) of the His 32 loop. Assenting NMR and TMD simulations results suggest

that these changes are necessary to allow a smooth transition of the His 32 loop as it relocates

from the “open” to the “closed” conformation.

172  

2. The presence of the heme in the Tyr 75 loop triggers conformational rearrangement of the

His 32 loop

Loading of the heme onto the Tyr 75 loop triggers conformational rearrangement of the

His 32 loop regardless of the presence or absence of the heme iron distal ligand. This is clearly

demonstrated in the crystal structure of H32A-imidazole where the residues whose electron

densities could be traced (i.e in this His 32 loop) are located in similar positions as those seen in

the crystal structure of wild type holo HasAp. Complementary information was obtained from

solution NMR studies of H32A and H32A-imidazole complex which showed that resonances of

some of the residues located in the His 32 loop are broadened beyond detection due to the

paramagnetic effect of the heme, implying that these residues are in close proximitiy to the face

of the heme. We conclude that the role of the His 32 loop is to protect the heme macrocycle from

exposure to the aqueous media.

3. Hydrogen bonding interaction between the heme proximal residues Tyr 75 and His 83 is

important in maintaining the structural integrity of the protein.

Breaking and formation of the hydrogen bond between Oη of Tyr 75 and Nδ of His 83 is

proposed to modulate the Tyr-Fe3+ bond [5]. X-ray crystal structures of Y75A and H83A show

that scission of this bond brought about by the alanine mutations induces subtle yet conceivable

conformational changes in the protein (Figures 12, 17, 18 and 19 in chapter 3). These changes

involve conformational changes in the backbone and spatial rearrangement of the amino acid

side chains of residues constituting the His 32 loop, and loss or gain in secondary structure

elements on or near the regions of the protein implicated to associate with the receptor HasR

173  

during heme release. Our results are in good agreement with the proposal that conformational

changes that occur upon interaction of HasA with the receptor are remotely communicated to the

Tyr 75 loop though in our case communication is in the reverse direction. We propose that

information is relayed via the His 32 loop since this loop experiences more conformational

changes compared to the Tyr 75 loop. We also propose that besides modulating the Tyr-Fe3+ bond

and hence heme binding or release, the Oη Tyr 75- N His 83 hydrogen bond is important in

maintaining the overall architecture of the protein. In addition, the nature of the molecule

associating with the proximal heme-iron ligand (Tyr 75) in the absence of the auxiliary partner

(His 83) contributes to the stability of the secondary structure elements of the protein. This is

clearly demonstrated by the two crystal forms of H83A (Figure 11-C and D, in chapter 3).

The high resolution x-ray crystal structures of the mutants unequivocally determined the

coordination state of the heme iron. As shown in Figure 14 in chapter 3, the heme in the Y75A

HasAp mutant is hexacoodinate and the axial ligands are His 32 and a chance ion (formate). The

side chain of His 83 is in a position nearly identical to that seen in the crystal structure of wild

type HasAp and it is not coordinating the heme. This observation contradicts Deniau and

coworker`s proposal that in the absence of Tyr 75, His 83 may serve as the proximal ligand to the

heme [6]. Caillet-Saguy and coworkers proposed that in the H83A mutant of HasAs, Tyr 75 is

detached from the heme at acidic pH forming a pentacoordinate high spin species and, upon

increasing the pH, Tyr 75 ionizes and its phenolic oxygen forms a hydrogen bond with the bound

water ligand thereby stabilizing the hydroxide complex [7]. Our results do not concur with this

proposal. In the structure of H83A refined to the highest resolution possible (0.89 Å), Tyr 75 is

not detached from the heme.

174  

3. Methemalbumin is the potential target for the hemophore HasAp as a heme source.

Studies involving heme transfer from host heme proteins to HasA have focused primarily

on hemoglobin as the heme source. However, we propose that methehemalbumin is the potential

target for the secreted hemophores based on two reasons; first, the association between heme and

serum albumin (Kd= 10-8 M) [8] is relatively weak; the weakest among the host heme proteins,

and secondly, the amount of methhemalbumin is blood plasma is very high compared with the

other heme or hemoglobin scavenging proteins hemopexin and haptoglobin, respectively.

Furthermore it is probable that the heme transport protein hemopexin is saturated during

hemolytic conditions leaving substantial amount of heme in serum albumin accessible to

invading microbes. On this basis, investigations were carried out to determine if HasAp can

acquire heme from methemalbumin. Results from UV-vis spectroscopic analysis revealed that

heme transfer from methemalbumin to apo HasAp is very rapid. In addition, the rate of heme

release from methemlbumin correlates with the rate of heme uptake by HasAp. These rates

however are comparable to the rate of dissociation of heme from methemalbumin in the absence

of the hemophore. Complementary information was obtained by polyacrylamide gel

electrophoresis. The high resolution afforded by polyacrylamide gels allowed complete

separation of the reaction products and confirmed that heme is indeed completely transferred

from methemalbumin to HasAp within a short time ( <10 min). We conclude that methemalumin

is the potential target for HasAp and that heme transfer is driven by passive diffusion and

affinity.

175  

References

1. Vasil, M. L., and Ochsner, U. A., The response of pseudomonas Aeruginosa to iron:

Genetics, biochemistry and virulence. Molecular Microbiology, 1999. 34: p. 399-413. 2. Letoffe,S., Ghigo,J.M., and Wandersman,C., Secretion of the Serratia marcescens HasA

protein by an ABC Transporter. J. Bacteriol., 1994b. 176: p. 5372-5377. 3. Yoshida, T., Noguchi, M., Kikuchi, G., and Sano, S., Degradation of mesoheme and

hydroxymesoheme catalyzed by the heme oxygenase system: involvement of hydroxyheme in the sequence of heme catabolism. J Biochem., 1981. 90: p.125-131.

4. Weinberg, E. D., Iron and infection. Microbiol. Rev., 1978. 42: p. 45-66. 5. Wolff, N., Izadi, N., Couprie, J., Habeck, M., Linge, J., Rieping, W., Wandersman, C.,

Nilges, M., Delepierre, M., and Lecroisey, A., Comparative analysis of structural and dynamic properties of the loaded and unloaded hemophore HasA: functional implications. J. Mol. Biol., 2008. 376: p. 517-525.

6. Deniau, C., Gilli. R., Izadi, N., Létoffé, S., Delepierre, M., Wandersman, C., Briand, C.,

Lecroisey, A., Thermodynamics of heme binding to the HasA(SM) hemophore: effect of mutations at three key residues for heme uptake. Biochemistry, 2003. 42: p. 10627-10633.

7. Caillet-Saguy, C., Delepierre, M., Lecroisey, A., Bertini, I., Piccioli, M., and Turano, P.

Direct-Detected 13C NMR to Investigate the Iron(III) Hemophore HasA. J. Am. Chem. Soc., 2006. 128: p. 150-158.

8. Adams, P. A., and Berman, M. C., Kinetics and mechanism of the interaction between

human serum albumin and monomeric haemin. Biochem. J. 1980. 191: p 95-102.