Phaco DAPI GFP FM4-64 Overlay Model

P. li

mno

phila

f

E. c

oli

d

c

e

g

h

P. li

mno

phila

0 1 2 30

2000

4000

6000

8000

10000

Size [µm]

Inte

nsity

FM4-64 GFP DAPI

P. limnophilaE. coli

Ctr 30% sucrose0

20

40

60

80

100

Nuc

leoi

d si

ze [%

]

Ctr 30% sucrose0

20

40

60

80

100

Cyt

osol

size

[%]

Ctr 30% sucrose0

20

40

60

80

100

Cyt

osol

size

[%]

P = 0.0001P = 0.0001

P = 0.0001

Ctr 30% sucrose0

20

40

60

80

100

Nuc

leoi

d si

ze [%

] P = 0.0009

i j lk

Fields of view

Cel

ls [%

]

invaginations no invaginationsa bP.

lim

noph

ila

0 1 2 30

2000

4000

6000

8000

10000

Size [µm]

Inte

nsity

FM4-64 GFP DAPI

100

80

60

40

20

01 2 3 4 5 6 7 8 9 10

SupplementaryFigure1:PlasmolyticstressincreasesthecytoplasmicinvaginationsofP.

limnophila

(a+b) Representative section of one field of view with Fm4-64 (red) stained cells. In (b) the

rationbetweenmorphotype1(noinvaginations)andmorphotype2(invaginations)isgivenfor

1838cellsamong10randomlychosenfieldsofview.Scalebar:10µm.

(c-f)Artificialinvaginationofthecytoplasmicmembranebyplasmolysis.GFP(green)expressing

E.coli(c+d)andP.limnophila(e+f)cellswereuntreated(c+e)orosmoticallystressedwith30%

sucrose(d+f).E.colicellsformcharacteristicperiplasmicbays(d,Phaco).Both,nucleoid(blue)

andcytoplasm(green)ofE.colicellsareshrunkwhilethemembranesignal(red)resemblesthe

stainingpatternofmorphotype2P.limnophilacells(eandFigure2c+d).Suchnaturaloccurring

cytoplasmicmembrane invaginationsofP. limnophilamorphotype2 cells (e andFigure2c+d)

increased by osmotic stress and became visible in phase contrast (f, Phaco). The cytoplasm

shrunk further (f, GFP) and a black dot became visible in membrane staining experiments (f,

FM4-64).Scalebars:1µm.

(g+h)OverlayofWF,Phaco,Fm4-64(red),GFP(green)andDAPI(blue)signalsofuntreated(g)

andtreated(h)P.limnophilacells.Theoverviewdemonstratesthatin(e)and(f)selectedcells

were representative. Corresponding fluorescence intensity profiles of untreated- (g) and

osmoticallystressedcells(h)revealedonecytosolicGFPmaximaoftheuntreatedcontrolwhile

aseparationofthecytosolicGFPsignalintotwodistinctintensitymaximawasobservedamong

treatedcells(h).Scalebars:5µm.

(i-l)Quantification revealeda significantdecreaseofboth, sizeof thenucleoidand sizeof the

cytosol in E. coli (i+j) and P. limnophila (k+l), if treated with 30% sucrose compared to the

respectively untreated control. (T-test: P=0.0001, P=0.0001, P=0.0009 and P=0.0001

respectively,errorbars:mintomaxof10measuredcellsamong2individualexperiments).

Gem

mat

a ob

scur

iglo

bus

Rho

dopi

rellu

la b

altic

a

Phaco DAPI FM4-64 Overlay Model Overlay Nile Red DAPI

WF SR-SIM

e

d

g

f

Pla

ncto

piru

s lim

noph

ila

h i

0 1 2 30

2000

4000

6000

8000

Size [µm]

Inte

nsity

a

SR-SIMWF

0 1 2 30

2000

4000

6000

8000

Size [µm]

Inte

nsity

DAPI GFPNile RedGFPFM4-64 DAPI

b c

Pla

ncto

piru

s lim

noph

ila

GAT

TA-S

IM 1

20b

Supplementary Figure 2: The innermost planctomycetal membrane is the cytoplasmic

membrane

(a-c) Control of super resolution structured illumination (SR-SIM) performance under

experimentalconditions. (a+b)Representative fluorescence intensity lineprofilesofFM4-64/

Nile Red (red) and DAPI (blue) stained, GFP (green) expressingP. limnophila cells imaged in

wide field- (a:WF)andSR-SIM(b).Thewidthofouter-and innermembranepeaks(red)was

442+/-68nmforWF-and225+/-44nmforSR-SIMexperiments(20randomlychosencells).

Thus,SR-SIMroughlydoubledtheresolutioncomparedtoWF.(c)Overviewofarepresentative

SR-SIMimageshowingmultipleDNAorigamis,eachwithtwoAlexaFluor488dyes in120nm

distance.Signalsofboth fluorophoresoverlap(c:magnification,whitearrowheads), indicating

thattheresolutionis>120nm.Scalebars:5µmand0.1µmrespectively.

(d-g)G. obscuriglobus (d+e) andR. baltica (f+g) cells analysed in WF and structured SR-SIM.

Phase contrast (Phaco) microscopy revealed intracellular structures. In both species, the blue

nucleoid (DAPI) is highly condensed and the red cytoplasmicmembrane (FM4-64 /NileRed)

exhibitmultiple invaginationsas illustratedinthemodel.SR-SIManalysesresolvedmembrane

invaginationsthatappearasreddotsorlinesinWFmicrographs.Scalebars:1µm.

(h+i) A representative freeze fractured, plunge-frozen P. limnophila cell (h) with enlarged

periplasm, caused by multiple invaginations of the cytoplasmic membrane (selected from 50

cells among 3 independent experiments). (i) False colours correspond to colours used in the

modelsabove(d-g).Scalebars:0.5µm.

a b c

d e f

g h i

j k l

SupplementaryFigure3:Cryo-electrontomographyofPlanctopiruslimnophilacells

(a-f)Exampleofacellwithoutinvaginationsofthecytoplasmicmembrane.Representativeslices

of a tomogram of a FIB-thinned cell (thickness of remaining cell volume 260 nm; inter-slice

distances are17, 21, 17, 21 and17nm). (g-l) Cellwith "tubular"- or "disk"-like invaginations

apparently "enclosing" part of the cytoplasm (second image from top left). Slices are from a

tomogramofaFIB-thinnedcell(thicknessofremainingcellvolume360nm;inter-slicedistances

are41,38,41,38and38nm).Scalebars:250nm.

Phaco DAPI FM4-64 OverlayDiOC6(3)R

hodo

pire

llula

bal

tica

Pla

ncto

piru

s lim

noph

ilaG

emm

ata

obsc

urig

lobu

s

b

a

d

c

f

e

P. limnophila R. baltica G. obscuriglobus

Intensity DiOC6(3)

Inte

nsity

FM

4-64

hg i0,909 +/- 0,0490,932 +/- 0,0280,922 +/- 0,0970,984 +/- 0,018

0,911 +/- 0,0340,928 +/- 0,0280,985 +/- 0,0250,970 +/- 0,021

0,811 +/- 0,0300,843 +/- 0,0250,929 +/- 0,0310,998 +/- 0,002

SupplementaryFigure4:Membranepotentialattheinnermostmembrane

(a-f)Theplanctomycetalspecies,P.limnophila(a+b),R.baltica(c+d)andG.obscuriglobus(e+f)

were stained with DAPI (DNA, blue), FM4-64 (membrane, red) and DiOC6(3) (energized

membrane, green). The overlays show a distinct red signal around the outer rim of all three

species. This indicates that the outermost membrane is not energized. The innermost

membranesappearyellowintheoverlays,whichprovidesevidencethattheyareenergizedand

correspondtothecommonbacterialcytoplasmicmembrane.Scalebars:2µm.

(g-i)IntensityblotsofFM4-64versusDiOC6(3)signalsrevealatypicallinearcorrelationwhich

is slightly blurry for G. obscuriglobus cells (i). Calculation of Pearsons’s correlation (white),

Mander’soverlap(green)andco-localisationcoefficientsc1(blue)andc2(red)furthersupport

co-localisation of FM4-64 and DiOC6(3) signals among all analysed species. For (g+h) 50

individualcellswerecompared.For(i)10fieldsofviewwereused,duetoaggregateformation

ofG.obscuriglobus.

WF

Immunogold overview

Outer rim 0.4 µm

Inner partOuter rim Inner part

0

20

40

60

80

Inte

nsity

/ A

rea

[%]

Immunogold

Bgr OM IM Unspc.0

20

40

60

Loca

lisat

ion

/ 3.5

µm

2 [%

]

a b

Outer membrane Inner membrane Background Unspecific in cell

****

********

c

d

SupplementaryFigure5:QuantitativeATPaselocalisationanalysisinG.obscuriglobus

(a+b)Thewidefieldimmunofluorescencequantificationwasachievedbydividingcellsintotwo

parts(a).Theouterrimcomprisedadiameterof0.4µmandthuscoverstheentirerangewhere

signals associated with the outer membrane (OM) could be expected. In contrast, only

invaginationsofthecytoplasmicmembrane(CM)cancauseATPasesignalsfromtheinnerpart

ofthecell.(b)Comparisonoftheintensityperareabetweentheinnerandouterpartof50cells

in5fieldsofview.39.21%ofthesignalsrelatetotheouterrimand60.79%tothecellcentre.T-

test:P=0.0001,errorbarsindicate95%confidenceinterval.Scalebar:0.5µm.

(c) Analysis of immunogold ATPase labelling of 10 micrographs from three independent

experiments.Mean-valuesareshownanderrorbarscorrespondtothe95%confidenceinterval.

T-testsrevealedsignificantdifferencesbetweenlocalizationattheinnermembranecomparedto

background(p=0.0001),totheoutermembrane(p=0.0001)andtounspecificsignalsinsidethe

cells(p=0.0001).

(d)Representativemicrographof immunogold labelledATPase localization inG.obscuriglobus

cells.Goldparticles(blackdots)areassignedtofourdifferentlocalizationsasshowninFig.4c.

Diameter of coloured circles indicatesuncertainty of localizationdue to the size of antibodies

andgoldparticle(22.5nm∅).Scalebar:0.5µm.

Type I Type IIType III Type Va Type Vb Type Vc Type IV Type VI

TolC

HlyD

HlyB

YscW

YscFYscOPX

YscJ

YscVURTS

YscQL YscNGspE GspO

GspFLM

GspGHIJK

GspC

GspS

GspDYscC

TatABCESecDEFGY

YajCYidC

SecAFtsY

VirB4 VirD4

VirB11

VirB6VirB8VirB10

VirB1

VirB3VirB5

VirB7VirB9

VirB2Hcp

VgrG

Lip

IcmFDotU

PpkAClpV

Fha1

PppASecB ffh

YscF

YscO YscP YscX

YscC

YscJ YscR YscS

YscT YscU YscV

YscQ YscL

Type III

YscW

YscN

HlyB

Type I

GspS

GspC

GspH GspI GspJ

GspL GspM

GspO

Type II

VacA ShlB

ShlA YadB/C

YadA

Type V

SecE SecG

SecM

SecB

Sec-Secretion pathway

TatB TatE

Twin arginine targeting (Tat)

VirB1

VirB2

VirB7

VirB6

VirB4

VirB3

VirB9

VirB8

VirB11

VirB5

VirB10

VirD4

Type IV

VgrG

Hcp

Lip

IcmF DotU

Type VI

Planctopirus limnophila Gemmata obscuriglobus Rhodopirellula baltica

Planctopirus limnophila Proteome verification

TolC

HlyD

GspD

GspF GspG

GspK

GspE

ClpV

PpkA Fha1 PppA

SecD/F

YajC YidC

SecA

FtsY

ffh

TatA TatC

OM

PP

IM

CP

SecY

a

b

Supplementary Figure 6: Planctomycetes lack most proteins of canonical secretion

systems

(a) Schematic drawing of bacterial secretion systems I to VI, according to the KEGG database

bacterialsecretionsystemsection.

(b) Proteins that correspond to the individual secretion system are indicated in boxes.

HomologousproteinsencodedinthegenomesofP.limnophila(pink),G.obscuriglobus(green)or

R.baltica(grey)areindicatedinrepresentativecolours.IncaseofP.limnophilamostpredicted

proteinswereconfirmedinproteomicexperiments(shadedpinkandSupplementaryTable2).

While all Planctomycetes possess a slightly modified Sec secretion pathway, only few

homologous proteins of the other secretion systems are present. Noteworthy, R. baltica

harboursalmostallgenesforatypeVIsecretionsystem.

DAPI Nile Red OverlayGFP

SR-SIM

WF

Ove

rlay

Epi

fluor

esce

nce

cba fe

GFP + CCCP GFP GFP + CCCP GFP

3D re

cons

truct

ion

Epi

fluor

esce

nce

Gem

mat

a ob

scur

iglo

bus

i

h

0 %

50 %

100 %d

0 %

50 %

100 %g

Gemmata obscuriglobus Planctopirus limnophila

SupplementaryFigure7:G.obscuriglobustakesupGFPintotheperiplasm

(a-d)G.obscuriglobuscellswerefedwithGFP(green)andanalysedwithwidefieldmicroscopy

(WF).CellspoisonedwithCCCPshowednosignal(a),whilesomeuntreatedcells tookupGFP

(b).Magnifiedcellsrevealedregionsofintensefluorescence,indicatinglocallyconcentratedGFP

(c). For quantification, among 6 fields of view 1281 cells were counted. 10.2% of the cells

internalizedGFP(d).Errorbarsindicatestandarddeviation.Scalebars:1µm.

(e-f) P. limnophila cells do not show GFP uptake in similar experiments. For quantification

among6 fieldsofview1294cellswerecountedwhilenotasingleone internalizedGFP(f+g).

Scalebars:10µm.

(h+i)G.obscuriglobuscellswerefedwithGFPwhilemembraneswerestainedwithNileRedand

the nucleoid with DAPI (blue). Super-resolution structured illumination microscopy (SR-SIM)

revealed an associationof theGFP signalwith the lipid staining (h).The red signal illustrates

complexinvaginationsofthecytoplasmicmembranewhiletheGFPsignalcorrespondstopartof

theenlargedperiplasmicspace in theoverlayandthethree-dimensionalreconstruction(iand

SupplementaryMovie2).Scalebars:1µm.

SR-Tesseler: Voronoi tessellation and Clusterprediction

ba

c

Alexa 647 labelled Dextran on 0.01% Polylysine

Overlay DIC & dSTORM dSTORM

SR-Tesseler: Voronoi tessellation and Clusterprediction

ed

f

dSTORMOverlay DIC & dSTORM

Gemmata obscuriglobus with Alexa 647 labelled Dextran

SR-Tesseler: Voronoi tessellation and Clusterprediction

hg

ic

dSTORMOverlay DIC & dSTORM

Planctopirus limnophila with Alexa 647 labelled Dextran

SupplementaryFigure8:DetaileddSTORMdextranuptakeanalysis.

(a-c) As negative control Alexa Fluor 647 labelled dextran was imaged with differential

interference contrast (a, DIC) and Super-resolution direct stochastic optical reconstruction

microscopic (b, dSTORM). No clusters >40 nm were detected with SR-Tesseler applying a

minimumof20detections.

(d-f)Detailedanalysisof thedSTORMdextran feedingexperimentswithG. obscuriglobus cells

(see Fig. 5e for details). The overlay of a dSTORM and a DIC image demonstrates that red

dextransignalslocalizewithinthecells(d+e).Clusters>40nmwerecalculatedwithSR-Tesseler

applying a minimum of minimum 20 detections and results are visualized in green (f). Scale

bars:10µm.

(g-i)DetailedanalysisofthedSTORMdextranfeedingexperimentswithP.limnophilacells(see

SupplementaryFig.9eforoverview).TheoverlayofdSTORMandDICimagesdemonstratesthat

thereddextransignalslocalizewithinthecells(g+h).Clusters>40nmwerecalculatedwithSR-

Tesselerapplyingaminimumofminimum20detectionsandresultsarevisualizedingreen(i).

Scalebars:10µm.

Ove

rvie

wCtr Dextran +CCCP Dextran -CCCP DextranDextran Detail

WF dSTORM

DAPI Nile Red Dextran Overlay

SR-SIM

b

Epi

fluor

esce

nce a c

3D re

cons

truct

ion

g

f

Epi

fluor

esce

nce

no uptakedextran uptake 0 %

50 %

100 %d e

SupplementaryFigure9:P.limnophilatakesupdextranintotheperiplasm

(a-d)P. limnophila cells, except thoseof thenegative control (Ctr),were fedwith fluorescein-

labelled 40 kDa dextran (green). (a+b) Poisoned cells showed no signal (+CCCP), while other

cells (-CCCP) took up dextran. Scale bars: 10 µm. (c) Magnified cells revealed foci of intense

fluorescence, indicating locally concentrated dextran. Scale bar: 1 µm. (d) For quantification,

among24fieldsofview15,557cellswerecountedand22.17%ofthecellsinternalizeddextran.

Errorbarsindicatestandarddeviation.

(e)Super-resolutiondirectstochasticopticalreconstructionmicroscopic(dSTORM)analysisof

P.limnophilacellsfedwithAlexa647-labelled10kDadextran.Incontrasttodiffractionlimited

micrographs (c), no distinct foci are visible, but the red fluorescence signal shows an almost

homogeneousdistributionofdextraninaconfinedcompartmentofthecell(seeSupplementary

Fig.8g-Ifordetails).Scalebar:1µm.

(f+g) P. limnophila cells were fed with fluorescein-labelled 40 kDa dextran (green), while

membraneswere stainedwithNileRed and the nucleoidewithDAPI (blue). Super-resolution

structuredilluminationmicroscopy(SR-SIM)revealedthatthegreendextransignalislocalized

ininvaginationsoftheinnermembrane(red)thatenlargetheperiplasmicspaceasshowninthe

overlay image and in the three-dimensional reconstruction (g and Supplementary Movie 4).

Scalebar:1µm.

1100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

NU

P120

S. ce

revi

siae

ZP

_02732338.1

G. obsc

uriglo

bus

D5S

XW

5P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1972

D5S

U03

P. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1153

D5S

VK

0P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1436

D5S

WP

7P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1564

D5S

QU

5P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_2734

D5S

R45

P. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_0666

D5S

U68

P. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_3308

D5S

XI6

P. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1723

D5S

MG

2P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_0029

D5S

VH

1P

. lim

nophila

Ord

ere

d L

ocu

s N

am

es:

Plim

_1407

alp

ha

he

lix

be

ta s

tran

d

tran

sme

mb

ran

e h

elix

α-he

lixβ-

stra

ndtra

nsm

embr

ane

helix

Supplementary Figure 10: Putative membrane coat-like proteins of Planctopirus

limnophila

P.limnophilahomologstotheMC-likeproteinZP_02732338.1ofG.obscuriglobuswereidentified

bysequenceanalysesandsecondarystructureprediction.Theblackhorizontal linesrepresent

the sequence of each protein. The predicted α-helices (magenta) and β-strands (cyan) are

indicatedbybarsaboveeachline.Theheightofthebarsisproportionaltotheconfidenceofthe

prediction (Psipred). Predicted transmembrane helices are shown in green boxes below the

sequences(TMMH).Theestimatedlevelofglobularityisindicatedbyalinebelowthesequence

(pred), the higher the line, the more globular. Plim_1972 is the most similar P. limnophila

homologtoZP_02732338.1intermsofsequencecomparisonandstructureprediction.

P. limnophilaWT

P. limnophilaΔ Plim 1972

Dex

tran

10 k

Da

Ove

rlay

ca

3 kb

1 kb

3.628 kb

Plim_1972

1 kb

KanR

Flank PCR

Kan PCR

Flank PCR Kan PCR

M WT Δ1972 - WT Δ1972 - M M WT Δ1972

Anti-Plim_1972

130 kDa100 kDa

b

d e

0

20

40

60

80

f

P: 0.21

Δ Plim_1972WT

Dex

tran

upta

ke p

er c

ell [

%]

SupplementaryFigure11:UptakeofdextraninPlanctopiruslimnophiladoesnotrequire

theMC-likeproteinPlim_1972

(a)ThePlim_1972geneofP.limnophilawasreplacedbyakanamycinresistancecassette(Kan).

ThecontrolPCRs for themutantvalidationconsistedof twodifferentprimersets. InWTcells

the flankingprimersetshallamplifya3.628kb fragmentwhile theKan-primersetshouldnot

resultinanyproduct.IntheDPlim_1972mutant,theflankingprimersetshouldresultina1.1kb

fragmentwhiletheKan-primersetshouldamplifya1kbregionofthemutant’sgenome.

(b) PCR products for P. limnophila WT cells and the DPlim_1972 mutant correspond to the

predictedfragmentlength(a),indicatingintegrityofthemutants’genotype.

(c)WesternblotanalysisofP.limnophilaWTandtheΔPlim_1972mutant.Theantibodyagainst

Plim_1972 detects a protein band around 120 kDa in WT cells while the ΔPlim_1972 mutant

shows no respective band. Employing the protein molecular weight calculator

(www.sciencegateway.org)amassof120.82kDawasestimatedforPlim_1972whichfitstothe

sizeofthedetectedbandforWTcells.

(d-f)P.limnophilaWT(d)andΔPlim_1972mutants(e)werefedwithfluorescence-labelled10

kDa dextran. Wide field epifluorescence and epifluorescence / phase contrast overlays

demonstratethatbothWTandmutantcellstakeupdextran(red).Nosignificantdifferencesin

the uptake among three independent experiments could be detected while 15,556 WT and

19,502 mutant cells were analysed (f, T-test: P=0.21, error bars indicate standard deviation).

Thus,theproteinPlim_1972isnotrequiredfordextranuptake.

SupplementaryTable1:ImmunogoldlocalisationofATPases

Goldparticleswerecountedin10fieldsofview(fov)bythreedifferentresearchers(R#1-R#3).

Localisation was differentiated between background, outer membrane, inner membrane and

unspecific within a cell. Thereby a localization uncertainty of 22.5 nm ∅ caused by antibody

labellingwasacknowledged.

fov

Background OuterMembrane InnerMembrane UnspecificincellR#1 R#2 R#3 R#1 R#2 R#3 R#1 R#2 R#3 R#1 R#2 R#3

1 5 15 12 32 20 17 29 44 57 35 24 132 10 10 18 37 30 24 21 49 53 32 21 223 4 18 17 36 32 25 33 55 59 23 22 194 6 24 18 34 25 30 23 54 67 26 19 185 5 15 21 44 44 15 23 52 67 28 17 206 6 12 10 42 20 13 26 45 52 23 21 247 7 9 16 34 20 22 31 41 49 37 11 148 5 7 13 30 21 18 26 40 50 36 23 239 6 16 21 34 34 24 22 58 63 33 8 16

10 4 15 22 26 26 27 13 30 47 41 9 16

SupplementaryTable2:P.limnophilaproteomeanalysis

List of typical Gram-negative proteins identified by our proteomic approach compared to oursecretion pathway prediction (SECPATH, Supplementary Fig. 6) and to the computationalanalysisbySpethetal.20121.LocalizationpredictionwasachievedwithPSORTB3.0.

ACCESSION NUMBER ANNOTATION

LOCALISATION

WITH PSORTB 3.0

SECPATH SPETH ET AL., 20121

ADG68672.1 Putative TolC Outer membrane + +

ADG69500.1 GspD Outer membrane + +

ADG66639.1 Type II/III ss homologue Outer membrane +

ADG67415.1 Beta barrel transporter Outer membrane +

ADG68191.1 BamA Outer membrane +

ADG68627.1 Beta barrel transporter Outer membrane +

ADG69297.1 TolC homologue Outer membrane +

ADG69651.1 CpaC Outer membrane +

ADG65956.1 LptD Outer membrane

ADG66250.1 Putative fimbrial subunit Outer membrane

ADG66422.1 Hypothetical protein Outer membrane

ADG66532.1 Hypothetical protein Outer membrane

ADG67744.1 OMPT Outer membrane

ADG68522.1 Carboxylase Outer membrane

ADG68596.1 Cytochrome C Outer membrane

ADG67045.1 Hypothetical protein Outer membrane

ADG67290.1 Hypothetical protein Outer membrane

ADG67378.1 Glycanotransferase Outer membrane

ADG69934.1 Hypothetical protein Outer membrane

ADG67060.1 Sulfatase Outer membrane

ADG67760.1 BamB Outer membrane

ADG69252.1 OMPA Outer membrane

ADG69913.1 Porin Outer membrane

ADG69212.1 Esterase Outer membrane

ADG66736.1 Porin Cytoplasm +

ADG69997.1 GspE/PilB Cytoplasm +

ADG67874.1 ClpV/VasG/ClpB Cytoplasm +

ADG68194.1 SecA Cytoplasm +

ADG69980.1 FtsY Cytoplasm +

ADG66008.1 GspF/PilC/PulF Cytoplasmic membrane +

ADG69480.1 PpkA Cytoplasmic membrane +

ADG67183.1 SecD/F Cytoplasmic membrane +

ADG66345.1 SecY Cytoplasmic membrane +

ADG67540.1 YidC Cytoplasmic membrane +

ADG68547.1 TatA Cytoplasmic membrane +

ADG68503.1 TatC Cytoplasmic membrane +

ADG66306.1 Cell adhesion Extracellular +

ADG67429.1 HlyD Periplasm +

ADG69451.1 GspK/ PulK Periplasm +

ADG66488.1 Putative Secretin Unknown

ADG70027.1 Fha1 Unknown +

ADG67044.1 TolC homologue Unknown

ADG67856.1 OMP Unknown +

SupplementaryReferences

1. Speth DR, van Teeseling MC, Jetten MS. Genomic analysis indicates the presence of an asymmetric bilayer outer membrane in planctomycetes and verrucomicrobia. Frontiers in microbiology 3, 304 (2012).