Influence of osmolarity of the growth medium on the outer membrane protein pattern of Escherichia coli

(1)

CopyrightC1977 American Society for Microbiology Printed in U.S.A.

Influence

of

Osmolarity of the Growth Medium

on

the

Outer

Membrane Protein

Pattern of

Escherichia coli

WIM VAN ALPHEN* AND BEN LUGTENBERG

DepartmentofMolecular Cell Biology, SectionofMicrobiology,*andInstituteforMolecularBiology,State University, Transitorium3,Padualaan8, Utrecht, TheNetherlands

Received for publication28March 1977

Supplementation ofthe growth medium with high concentrations of NaCl, KCl, orsucrose caused adrastic change inthe ratio of the two

peptidoglycan-associated majoroutermembrane proteins ofEscherichiacoli K-12 in that the

amountsof proteins bandcpresentincell envelope preparationsdecreased and increased, respectively. Kinetic studies showed that, after the osmolarity of the mediumwaschanged, oneproteinwashardlyincorporatedintothemembrane,

whereasthe otherwasincorporatedwithanincreasedrate.Afterabout1.5 to2

generations, the cell envelopes obtained the b/c ratio characteristic for thenew medium, and both proteins were subsequently incorporated with rates that ensured this new ratio. Once proteins b and c were incorporated in the cell envelope, theywerenotconverted into each otherby changes in osmolarity of thegrowth medium

The outermembraneofEscherichiacoli K-12 contains a family of major outer membrane proteinsthat can be resolved into four protein bands: a, b, c, and d (12). Protein a might be identicaltoSchnaitman's protein 3b (22, 23; P. Manning andP. Reeves, personal communica-tion). Protein bands b and c are identical to Henning'sbands Iaand Ib, respectively (9, 21). Proteinsb and ctogether correspond to Schnait-man's protein 1 (12, 22, 23). Protein d is identi-calto Schnaitman'sprotein 3a (12, 22, 23) and to Henning'sprotein II* (12, 13, 21). The func-tions ofthese proteins are largely unknown. The relative amounts of proteins b and c are dependent on strain, growth medium, growth temperature, and growth phase such that a small amount ofprotein b is more orless com-pensated for byanincreasedamountofprotein cand viceversa(13). Thesetwoproteins are the only proteins that are strongly, but not cova-lently, linked to peptidoglycan (13, 20, 21). Moreover, Schmitges and Henning (21) re-ported that b and c represent modifications of the same polypeptide, which differ from each other inonly one cyanogen bromide fragment that does not correspond with the C- or N-terminal ofthe protein molecule.

In this paper we describe the influence of osmolarityofthegrowth mediumonthe compo-sition of the major outer membrane proteins, especiallyontherelative amountsof the pepti-doglycan-associatedoutermembraneproteinsb andc.Thiseffect wasstudiedin somedetailto

contribute to understanding the functions of theseproteins.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Sources, origins, and relevant characteristics of E. coli K-12 strains PC0205, JC7620 (previously desig-nated as PC1349), JF404, PC0668, P400, and its

pro-tein d-deficient derivative strain, P460, were de-scribedpreviously (13). Strain CE1036 (lacking

pro-teinc) is aderivative of strainAB1859(13).Mutants resistant to bacteriophage Mel were obtained as described earlier (27).

Except where noted, cells were grown under

vig-orous aeration at 37°C. The compositions of brain heart medium and of glucose minimal medium were described earlier (13). If required, the leucine

con-centrationinthe latter mediumwas 45 ,ug/ml, ex-ceptwhenradioactive leucinewasusedas a

precur-sor. In these cases, the leucine concentration used will be described in the text. The composition of yeastbrothwas asdescribedpreviously (13), except that NaCl (85 mM) wasomitted. The various

con-centrationsofNaCl, KCl, andsucroseused will be given below.

Isolation andcharacterization of cellenvelopes.

Exponentially growing cellswere disintegrated by

sonic treatment, andasample wastakento deter-minetotal cell protein. Cellenvelopeswereisolated by differential centrifugation as described

previ-ously (12).2-Keto-3-deoxyoctulosonic acidwas deter-mined asdescribed earlier (1).

Peptidoglycan-asso-ciated proteins were isolated by the method of Rosenbusch (20) asmodified by Lugtenberg et al.

(13). Protein was determined by the method of Lowry et al. (11). Separation of cell envelope pro-teins by polyacrylamide gel electrophoresis and

staining ofthe protein bands was carried out as

described earlier (12), except that shorterstaining anddestaining procedureswereused.Forfixingand staining, the gelwasincubated for15min at room 623

on January 5, 2017 by WALAEUS LIBRARY/BIN 299

http://jb.asm.org/

(2)

temperature in 50% methanol-10% acetic acid and furtherincubated for1hat60°Cin asolution of 0.1% fast green FCF in 50% methanol-10% acetic acid. Thisprocedure allows interpretation of the results onthe samedayelectrophoresis is carriedout. Sam-ples were alwaysappliedonthegelinatleasttwo

different concentrations. To estimate the relative

amountof proteinineachband, the gelwasscanned with aVitatron TLD 100densitometer at a rateof 1.0cm/min. For autoradiography, the gels were fur-ther soaked in 50% methanol-5% glycerol with gentleshakingat37°C foratleast1.5hand subse-quently dried (12).Autoradiographywascarriedout

for5 to 10daysat4°C with KodakRapid Processing

Royal X-Omat medical X-ray film (RP/R-14). The relative amount of radioactivity in each protein band was determined by scanning the autoradi-ogram, taking into account that the results were

only used when the surfacearea wasproportionalto

the protein concentration.

[14C]leucineincorporation. Anovernightculture ofthe leucineauxotrophic strain JC7620inglucose minimal medium wasdiluted 1:10 inglucose

mini-malmedium (withoutNaCl)containing[14C]leucine (Radiochemical Centre, Amersham, England) (10

jug/ml;

specific activity,6mCi/mmol) and incubated under aeration at 37°C. After three generations, judged from absorbance measurements witha

Uni-cam SP 600 spectrophotometer at a 660-nm

wave-length, a sample of 50 ml wastaken for isolation of cellenvelopes, and the remainder of the culturewas

centrifugedat37°C. The cellsweresuspendedinto 5

volumes of nonradioactive glucose minimal medium (45 ug of leucine per ml; 300 mMNaCl) and

incu-bated at 37°C. During atleasttwogenerations,

sam-ples were taken at various times for the determina-tion oftotal cell protein and for theisolation of cell envelopes.

In an analogous experiment, cells of strain JC7620 werelabeledinthe presence ofahigh NaCl concentration (glucose minimal medium; 10 ,ug of leucine perml; 300mMNaCl)and, after centrifuga-tion, further incubated in nonradioactive medium without NaCl (glucose minimal medium; 45 ,ugof leucine per ml).

Lipopolysaccharide analysis. 32P-labeled lipo-polysaccharide (LPS) was isolated from cells grown

in low-phosphate medium (2) supplemented with tryptophan (20

jig/ml)

and Casamino Acids (0.2%). Samples containing 10,000 to 15,000 cpm were ap-plied to Whatman 3 MM chromatography paper (40 by35cm) andchromatographedinisobutyric acid-1

Mammonium hydroxide (7:3, vol/vol). After drying, thechromatogram was exposed for 2 to 7 days to X-ray film (Kodak X-Omat R film/XR-1).

RESULTS

Effect ofosmolarityof the medium on the relative amounts of

major

outer membrane proteins. During studies on the membrane pro-tein pattern of a temperature-sensitive fabB mutant, which can grow at the restrictive tem-perature without exogenous unsaturated fatty acid provided that the growth medium is

sup-plemented with high concentrations of NaCl, KCI, or sucrose (3, 4), we found that these additions to the medium influenced thepattern of the major outer membrane proteins. This influence was found to be independent of the fabB mutation.Theaddition of 300 mM NaClor

KClor600 mMsucrosetoyeastbroth causeda

decrease in the amount ofprotein b,whichwas

accompaniedby an increase in the amountof proteinc(Fig. 1).That, inadditiontoNaCland KCI, sucrosealsocaused this effect shows that itwasprobablycaused by the osmolarity ofthe medium rather than a specific ion. No other significant changes inthe pattern of cell

enve-lope proteinsweredetected (Fig. 1). The ratios of cellenvelopeproteintototal cell protein and of themajoroutermembrane proteins(aplusb

67K-_

60K-w

,a

J

_-

b

li

1I-

c

d\c.

45K

-36

K-25K-

m

i4K-124

₂

₃

₄

₅

FIG. 1. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of (1) molecular weight

stan-dards; the otherslots contain samples of cell

enve-lopes ofstrainJC7620 grownin(2)yeastbroth, (3) yeast brothwith300mMNaCl,(4) yeast brothwith

300 mM KC1, and (5) yeast broth with 600 mM sucrose.Thestandard protein bands are indicatedat

the left by their molecular weights (e.g., 67 K =

67,000 molecularweight). Thepositionsofproteins a,b, c, and dareindicatedattheright.

http://jb.asm.org/

(3)

MEDIUM OSMOLARITY AND OUTER MEMBRANE

pluscplus d)tototal cell envelopeproteinwere

not influenced by increased osmolarity of the growthmedium,aswascalculated from protein

determinations andscanningof gels.

Theosmoticeffectontheamountsof proteins

b andc wasgeneral for all strains tested that were wild type with respect to these proteins

(Table1).In strains that containahighamount

of proteinb in their cell envelopes after growth inyeastbroth, theadditionofNaCl reduced the amount of this protein (e.g., strain PC0205), whereas in strains containing asmallamount

of proteinbafter growth inyeastbroth, supple-mentation withNaCl ledtotheabsence of

pro-teinb (e.g., strainAB1859). Itseems,therefore, that by supplementation ofyeast broth with NaCltheamountof protein b in cell envelopes canbe reduced by a certain amount and that

the straincanphenotypicallylack protein b in

yeast broth (300 mM NaCl) only when the

amountof protein bisalready low after growth inyeastbroth (no NaCl). Thedecrease in

pro-tein bwasmore orless compensated for by an increase in protein bandc.Theamountof

pro-tein a wasnot significantly influenced by the addition of NaCl. The amount of protein d seemed to increase in strain PC0205 and to

decreaseinstrainP400, whereasitwashardly

influencedinthe other strains (Table 1).

The question arose whether the increase in

proteinbandcwasdue eithertoareal increase ofproteinc ortothesynthesis ofanewprotein withthe sameelectrophoretic mobilityas pro-tein c. Two experiments showed that the in-crease wasduetoareal increase of theamount ofprotein c; namely, (i) after growth in both yeast broth with 0 mM NaCl andyeast broth with 300 mM NaCl, more than 90% of protein bandccouldbe recoveredassociated with pepti-doglycan, and (ii) growth of the protein c-defi-cient strain, CE1036, in yeast broth with 300 mMNaCl resulted in the absence of protein b

but not in the appearance ofaprotein c band

(Table 1). The same result was obtained with another

protein

c-deficientmutant, isolated as a

bacteriophage

Mel-resistant derivative of strainPC0205, which possessesahighrelative

amountofproteinb in itsoutermembrane after

growth

in yeast broth.

Figure2A shows the amounts of four

major

outer membrane proteins (a,

b,

c, and d) of

strain JC7620 relative to total cell

envelope

proteinas afunctionof theNaClconcentration in yeast broth. The amount of protein b de-creased with

increasing

NaCl concentrationup

to about 200 mM and was rather constant at

higher concentrations. No evidence for excre-tion ofprotein b into thegrowthmedium could be obtained, as was measured in the

acid-pre-cipitable

material of the supernatant obtained aftercentrifugation ofthe cells. Theamountof proteincincreasedmore orless inparallelwith thedecrease inprotein

b,

whereas theamount

ofproteinddecreasedgradually.Theamountof proteinawasratherconstantexceptfora

slight

tendencytodecreaseathigherNaCl

concentra-tions. Theinfluence of the NaCl concentration on the generation time isgiven in Fig. 2B.

It has been reported that the b/c ratio is dependent on the composition of the growth medium and on the growth temperature (13).

We tested the effect of NaCl in glucoseminimal medium, yeastbroth, and brainheartmedium

on strain JC7620 at temperatures between 30 and 42°C. Theresults showed that thepresence

ofNaCl under all conditions causedadecrease in protein b accompanied by an increase in

protein c. The highest relative amountof

pro-tein b was found in cells grown at 30°C in glucose minimal medium (b/c ratio, 2.5); the lowest occurred in cellsgrownat420C in brain heart medium supplemented with 300 mM NaCl. Underthe latter growth conditions,

pro-tein b could notbe detectedanymore.

TABLE 1. Effect ofNaCIinthegrowth mediumonthecontentof majoroutermembraneproteins

Relative amt of individual majoroutermembraneproteins(% of totalmajorouter membrane protein)a

Strain Protein

defi-ciency Yeast broth(OmMNaCI) Yeast broth (300 mMNaCI)

a b c d a b c d PC0205 None 6 39 22 33 4 10 32 54 JC7620 None 9 27 20 44 10 5 41 44 AB1859 None 10 21 28 41 7 0 51 42 P400 None 3 16 15 66 5 0 43 52 JF404 None 7 10 40 43 4 0 47 49 PC0668 None 6 4 49 41 6 0 53 41 CE1036 c 2 15 0 83 2 0 0 98

a Cellsweregrown in theindicated growth medium foratleast 10 generations. Cell

envelope proteins

wereseparated by gelelectrophoresis. Thedatawerecalculatedfrom scansof stainedgels.

http://jb.asm.org/

(4)

626 cI 0,) z o uJ 0o oa, c: a c-c) Z O 0 z 4) o X 2 0 0-u. 100- 80-Z _ ° C 604

<_

E 40-z 20-(.9 100 200 300 400 NaCl CONCENTRATION (mM) 560

FIG. 2. (A) Relative amounts ofthe majorouter

membrane proteins a, b, c,anddaspercentages of total cell envelope protein of cells of strain JC7620 grownat37°C inyeastbrothsupplementedwith

var-ious NaCl concentrations. Datawerecalculatedfrom scansof stained gels.Symbols: 0,proteina; O, pro-tein b; A,protein c; V, protein d. (B) Generation

times of cells ofstrain JC7620 grown at 37°C in yeastbrothsupplemented with various NaCl

concen-trations.

Kinetics of the change in the protein pat-tern. Tostudythe effect ofachangein

osmolar-ity of the growth medium on the kinetics of

changes in amounts ofmajor outermembrane proteins, a series of experiments was carried

out in which cells were labeled with

I14C]leucine in a medium with a low or high

NaCl concentration, respectively, and

subse-quently shifted to a nonradioactive medium

containing a high or low NaCl concentration,

respectively.The change in NaCl concentration

hardly influencedthe generation time. No

sig-nificant increaseintheamountof acid-precipi-table radioactivity of cells or cell envelopes

couldbe measured after the shifttothe

nonra-dioactive medium.

Inthe firsttypeofexperiment, cellsofstrain

JC7620 were labeled with [14C]leucine in glu-coseminimalmedium (10 ,ug of leucine per

ml;

0 mM NaCl). After about three generations, 85% of theradioactivitywas incorporated into the cells. After centrifugation, the cells were

incubated (zero time) at 37°Cinnonradioactive high-salt medium (glucose minimal medium; 45

g.g

of leucine per ml; 300 mM NaCl). Cell envelopes, isolated fromsamplestaken at

var-ious times during at least two generations, were analyzed by polyacrylamide gel electro-phoresis. Stained gels as well as autoradi-ogramswerescanned. With respecttoboth pro-tein contentandradioactivity, the ratios of cell envelope protein to total cell protein and of total major outer membrane protein (a plus b plusc plus d) tototal cellenvelope protein did notchange during the experiment. In Fig. 3A and B the amounts ofradioactivity and pro-tein,respectively, ofthe individual majorouter

membrane proteins a,b, c, anddare

expressed

as percentages of the amount of total major

outermembraneprotein. Therelativeamounts

ofradioactivityinthe fourproteinsincell

enve-lopes didnotchangesignificantly aftertheshift to nonradioactive medium without NaCl (Fig. 3A). The relative amounts of protein a and d remained constant (Fig. 3B), whereas the amountof protein b decreasedandthat of pro-teincincreased. Figure 3C showsspecific activ-ity in the individual major outer membrane proteins as a function of the number of cell divisions after the shift to medium supple-mented with NaCl. The specific activities of proteins a and d decreased at equal rates, whichhardlydiffer fromtherateexpectedfrom

a protein that neither incorporates radioactiv-ityafter theshift,norissubjecttoturnover(see theoreticallinein

Fig.

3C). Thespecificactivity ofprotein b remained constantduring about1.5

generations and subsequently decreased with about the same slope as the theoretical line. During the first 1.5 generations, the specific activity of proteincstrongly decreasedand sub-sequently diminishedataboutthe samerateas the specific activities ofthe other three

pro-teins.

The resultsdiscussed above are explained as follows. The relative amounts of proteins a and d in cell envelopes were hardly, or not at all, influenced by the presence of NaCl in the growth medium. Immediately after the shift,

newprotein b was incorporated into the mem-brane at a strongly decreased rate (Fig. 3A). Protein c was then incorporated at a strongly increased rate. After 1.5 generations, the

amountsofproteins bandcreached levels

char-acteristic for the new growth medium.

Subse-B

http://jb.asm.org/

(5)

627

A o100 u 080 I->60 L _ O b 020 Z20 O a u I 2 1 2: B 100 m80 0 0I Z60 9 -- d L 40 0 z

020

₂₀ <_ b a 1 2

NUMBER OF CELL DIVISIONS

FIG. 3. Kinetics ofchanges in relativeamounts ofproteins a, b, c, and daftergrowth ofstrain JC7620 cells inglucose minimal medium (10 pgof['4C]leucineperml, 0 mMNaCl) at 37°C and a shift at zero time to glucose minimal medium (45 pgofunlabeled leucine per ml, 300 mMNaCl). (A) Relative amounts of radioactivity in the individual proteins aspercentages of the total radioactivity in a plus b plus c plus d. As explained in the text, the ratio (a + b + c +d)/totalcell protein remained constant during the experiment. (B) Relative amounts ofprotein in the individualproteins as percentages of the total amount ofprotein in a plus b pluscplus d. (C) Specific activity (arbitrary) units ofthe individual proteins. Note the logarithmic scale of the y axisin(C). The broken line indicates the curve expected for a protein that is neither synthesized nor subject to turnover in the courseof the experiment. Symbols: 0, protein a; O, protein b; A, protein c; V, protein d.

quently, these proteins were incorporated at newrates specific for this medium.

In the second type of experiment, strain JC7620cells were labeled with [14C]leucine in glucoseminimal medium (10 ugofleucine per ml, 300 mM NaCl). After centrifugation, the cellswereincubated at

370C

innonradioactive medium without NaCl (glucose minimal

me-dium,

45 ug of leucine/ml, 0 mM NaCl). The

latterpartof theexperiment was carriedout in amannersimilartothat for the former. Again it Wasfoundthat, with respect to bothprotein contentandradioactivity, the ratios of total cell envelope protein to total cell protein and of total major outer membrane protein (a plus b

plus

cplus d) tototal cellenvelope protein did notchange during theexperiment.The relative amountsofradioactivityinthe proteins(a,b, c, and

d)

incell

envelopes

didnot

change

after the shift(Fig. 4A). The relativeamountsofproteins a and din cell envelopes did not significantly change after the shift, whereas the amountof protein b strongly increased during about the first 1.5 generations after the shift (Fig. 4B). The relative amount ofprotein c strongly de-creasedduring this period. The specific activi-ties ofproteins aandddecreasedatequalrates, which correspond rather well with a

simple

dilution of theirradioactivities

(Fig.

40). The specific activityofproteinb

strongly

decreased

during the first 1.5 generations and subse-quently followed the kinetics of dilution. The results of the experimentplotted inFig. 4 can be explained in essentially the same way as those of Fig. 3, except that band c should be reversed.

Interferencecontrast microscopy on exponen-tially growing cells of strain JC7620 did not indicatea majorinfluence of the osmolarity of thegrowthmedium on theshapeand size of the cells, except that some

wrinkling

of the cell surface wasobserved for cells grown in a

me-diumwith ahigh osmolarity.Anincreaseofthe osmolarity of the

growth

medium did not

change

the cell size but resulted in

slightly

irregularly shaped plasmolyzed cells with a slight

tendency

to form clusters.

Influence of the osmolarity on LPS. The existence ofa

relationship

betweenthe

struc-ture of LPS and the relative amounts of the major outer membrane proteins has been de-scribed

previously

(1, 7, 10, 13, 26). To test

whether a changed b/c ratio, caused by the presence of NaCl in the growth medium, was

accompanied

bya

change

inthe LPSstructure, 32P-labeled LPS was isolatedfrom cells grown in the absenceand presence of300 mMNaCl. After paper chromatography,

equal Rf

values were obtained for the two preparations. It is unlikely, therefore, that the NaCl

concentra-a

http://jb.asm.org/

(6)

628 VAN ALPHEN AND LUGTENBERG '100-0 * 80-~ 0 a < 40-Z 20 D

I

0 2 1 2 B .9100-Y .0 80 o-z cr60- 9 0. Z40 2 b 20 C a 2 3

NUMBER OFCELL DIVISIONS

FIG. 4. Kinetics ofchangesintherelativeamountsofproteinsa,b,cand dafter growthinglucoseminimal

medium (10 pgof [14C]leucineperml,300 mMNaCl)at37°Candashiftatzerotime toglucoseminimal

medium (45 pgofunlabeled leucineperml, 300 mMNaCI). (A)Relativeamounts ofradioactivity inthe

individualproteinsaspercentagesofthetotalradioactivityinaplusbpluscplusd.(B)Relativeamountsof

protein inthe individual proteinsaspercentagesofthe totalamountofproteininaplusbpluscplusd.(C) Specific activity ofthe individualproteins.Forfurtherdetails,seethelegend ofFig.3.Symbols:0,proteina;

D,proteinb; L.,proteinc; V, proteind.

tion in the growth medium affected the LPS structure.

The ratios of 2-keto-3-deoxyoctulosonic acid tototal cellproteinand of

2-keto-3-deoxyoctulo-sonic acid to total cell envelope protein

de-creasedbyabout 15 to 25% when the cellswere

grown in thepresenceof 300mMNaCl.

DISCUSSION

The results presented in this paper have

shown thatsupplementationof thegrowth

me-dium of E. coli K-12 strainswith high

concen-trations ofNaCl, KCl, or sucroseresults in a strong decrease of outer membrane protein band b, accompanied by a roughly equal

in-creaseinouter membraneproteinbandc (Fig. 1, Table 1).Sinceproteinc wasnot foundinthe

cell envelope of protein c-deficient mutants

after growth under these conditions, the in-crease inprotein band cmust be due to areal

increase in the amount ofproteinc and not to productionofa newproteinwiththesame elec-trophoretic mobility asproteinc.

The b/c ratio differs strongly forvariousE.

coli K-12 strains (13). In addition to the osmo-larity of the growthmedium, theb/cratioofa particularstraincanalso beinfluencedbythe

nutrient composition of the growth medium andbythegrowthtemperature (13). Thus, the

b/cratiocanbechangeddramatically,whereas thetotal amount of thesetwoproteinsremains

about constant. A change in the amount(s) of

protein(s) b and/or c can also influence the

amountofproteind (23; Fig. 2). The resultsof thispapershow, aswasreported earlier (8, 13, 23, 26), that the composition of theouter mem-brane, with respect tothe amountsofvarious major outer membrane proteins, is extremely flexible.

Inourlaboratorywehave applied this knowl-edge of the influence of the growthconditions onmutantsdeficient inone ortwo outer

mem-braneproteins (b, c, andd) for thepurification of theseproteins. Strain CE1034 (13), grownin

yeast broth supplementedwith 300mM NaCl, lacks proteins b and d and contains large

amounts of protein c. Strain CE1036, which lacks proteinc (13) and contains smallamounts

of protein b after growth in brain heart me-dium, was found to be an excellent source of protein d (25a). Strain CE1041 (13), grown in

yeast broth, lacked proteins c and d and

con-tained large amounts ofprotein b. The flexi-bilityofthe composition of theoutermembrane can be shown extremely well by growing the latter strain in brain heart medium

supple-mented with 300 mM NaCl, resulting in the

absence ofall three outer membrane proteins

(b, c, and d).

Bothproteins b andcarenoncovalently

asso-ciated withthepeptidoglycan layer (6,9,13, 20, 21). Schmitges and Henning (21) reported that thesetwo proteinsarealmost chemically

iden-tical. The only difference is a cyanogen

bro-mide fragment that corresponds neither with

A 2 c aa OLi - -b J. BACTERIOL. L t I c

http://jb.asm.org/

(7)

629 the N-terminal part nor with the C-terminal

part of the protein molecule. Based on this striking chemical similarity, it was suggested that both proteins b and c might be products of one structural gene and that the difference is introduced by post-translational modification (21). Our kinetic data (Fig. 3 and 4)show that, once these proteins are incorporated into the outermembrane, a change ofosmolarity of the growthmedium does not result in conversion of the bulk of one protein into the other.So if post-translational modification occurs, itmost likely takes place before the proteins become inserted into the outer membrane.

Anothercommon property ofproteins b (13, 17) and c (17, 21) is their interactionwith LPS. However, this is not an exclusive property of thesetwo proteins, since it wasrecently shown thatproteind alsohas this attribute (25a). The affinity for LPSmightevenbe thereasonwhy these proteins are located in the outer mem-brane andnot inthecytoplasmicmembrane.

Nakae has shown that the incorporation of certain outer membrane proteins ofSalmonella typhimurium (14, 15) and of the peptidoglycan-associatedmatrix protein b of E. coliB (16, 20) into phospholipid-LPS vesicles loadedwith dex-tran and sucrose results in leakage of sucrose but not of dextran. This system mimics the aquaous pores of the outer membrane that are supposed to be responsible for the extremely high permeability of the outer membrane for hydrophilic low-molecular-weight substances (18). Lugtenberg et al. showed that the outer membrane proteins of Salmonella typhimu-rium, which produce aquaous pores in phospho-lipid-LPS vesicles (15), are also peptidoglycan associated and proposed that the formation of aquaous pores might be a property of peptido-glycan-associated proteins in general (lla). The affinity of both peptidoglycan-associated proteins of E. coli K-12 for LPS andthe pres-enceof LPSinthe vesiclesdescribedby Nakae suggest thatLPSisalsorequiredfor the forma-tion of aquaous pores (13a) and also that pro-tein cisinvolvedinporeformation. Onemight evenspeculate that (partof) the particles that canbe seen-byfreeze-fracture electron micros-copy on the outer fracture face of the outer membrane (5, 24, 25, 28, 29) are identical to aquaous pores, since strong evidence was pre-sented for the idea that these particles are

composedof LPS aggregatesstabilizedby diva-lent cations and possibly also containing pro-tein and/or

phospholipid

(29). Since the pres-ence of NaCl in the growth medium has no

influence on thedensityofparticlesattheouter

fracture face of

wild-type

cells (29), one can

expectthatinthepresenceof NaClthe

density

of pores or particles that contain protein b would decrease, whereas the density ofthose containing protein cwould increase.

We are planning now to test whether both proteins b and c of E. coli K-12 are indeed active in pore formation and, if so, to see whether the two proteins differ in specificity towards various components. Theresultsmight answer the question of whether the observed influence of the osmolarity on the b/c ratio is either"accidental" or meant to protect thecell againstsuch a high osmolarity. An accidental change in the b/c ratio can be the result of influence of osmolarity on theconformation of a structuralor regulatory protein (19).This could result in a strong decrease in the incorporation of protein b or c into the outermembrane. The observed compensation effect then can be ex-plainedby assuming that the total amountofb pluscis regulated either by the spaceavailable in the outer membrane or on thepeptidoglycan or by the amount of LPS available. If the b/c ratiochanges to protect the cell, this protection could be accomplished either to decrease the influx of harmful components or wasteproducts or tofacilitatethe influx of certain nutrients.

ACKNOWLEDGMENTS

The excellent technical assistance of Greet van Loon is gratefully acknowledged.

This work was supported by the Netherlands Foundation forChemical Research (SON) with financial aid from the Netherlands Organization for the Advancement of Pure Research (ZWO).

LITERATURE CITED

1. Ames, G. F., E. N. Spudich, and H. Nikaido. 1974. Proteincompositionof the outer membrane of Salmo-nellatyphimurium:effectoflipopolysaccharide muta-tions. J. Bacteriol. 117:406-416.

2. Boman, H.G., and D. A. Monner.1975. Characteriza-tionoflipopolysaccharidesfrom Escherichia coli K-12 mutants. J.Bacteriol. 121:455-464.

3. Broekman, J. H. F.F., and J.F.Steenbakkers. 1973. Growthin highosmotic medium of an unsaturated fatty acid auxotrophof Escherichia coli K-12. J. Bac-teriol. 116:285-289.

4. Broekman, J.H.F.F., and J.F.Steenbakkers. 1974. Effect of the osmotic pressureof thegrowthmedium onfabB mutants of Escherichia coli. J. Bacteriol. 117:971-977.

5. Gilleland,H.E., Jr.,J. D.Stinnett,andR.G.Eagon.

1974.Ultrastructuraland chemical alteration of the cellenvelope of Pseudomonasaeruginosa, associated with resistance to ethylenediaminetetraacetate

re-sultingfromgrowthinaMg2+-deficientmedium. J. Bacteriol. 117:302-311.

6. Hasegawa, Y., H. Yamada, andS. Mizushima. 1976. Interactions of outer membraneproteins0-8 and0-9 withpeptidoglycansacculus ofEscherichia coli K-12. J. Biochem.80:1401-1409.

7. Havekes,L. M., B. J. J. Lugtenberg, and W. P. M. Hoekstra. 1976.ConjugationdeficientE. coli K12 F-mutants with heptose-less lipopolysaccharide. Mol. Gen.Genet. 146:43-50.

8. Henning, U., and I. Haller. 1975. Mutants of Esche-richia coli K12 lacking all 'major' proteins of the

http://jb.asm.org/

(8)

outercell envelope membrane. FEBS Lett. 55:161-164.

9. Hindennach, I.,and U. Henning. 1975.The major pro-teins ofthe Escherichia coli outer cell envelope mem-brane. Preparative isolation of allmajor membrane proteins. Eur. J. Biochem. 59:207-213.

10. Koplow,J., and H. Goldfine. 1974. Alterations in the outermembrane of the cell envelope of heptose-defi-cient mutants of Escherichia coli. J. Bacteriol. 117:527-543.

11. Lowry,0. H., N.J. Rosebrough, A. L.Farr, and R.J. Randall. 1951. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265-275. lla.Lugtenberg, B., H. Bronstein, N.van Selm, and R.

Peters. 1977. Peptidoglycan-associated outer mem-braneproteinsingram-negative bacteria. Biochim. Biophys. Acta 465:571-578.

12. Lugtenberg, B.,J. Meijers, R. Peters, P. van derHoek, and L. van Alphen.1975.Electrophoreticresolution ofthe 'major outer membraneprotein'ofEscherichia coli K12 intofourbands. FEBS Lett.58:254-258. 13. Lugtenberg, B., R. Peters, H. Bernheimer, and W.

Berendsen.1976.Influence ofcultural conditionsand mutations onthe composition of the outermembrane proteins of Escherichia coli. Mol. Gen. Genet. 147:251-262.

14. Nakae, T. 1975. OutermembraneofSalmonella typhi-murium: reconstitution ofsucrose-permeable mem-brane vesicles. Biochem. Biophys. Res. Commun. 64:1224-1230.

15. Nakae, T. 1976. Outer membraneofSalmonella. Isola-tionof proteincomplexthatproducestransmembrane channels. J.Biol. Chem.251:2176-2178.

16. Nakae, T. 1976. Identificationof the outermembrane protein ofE. coli thatproduces transmembrane chan-nels in reconstituted vesiclemembranes. Biochem. Biophys. Res. Commun. 71:877-884.

17. Nakamura,K., and S. Mizushima. 1975. In vitro reas-semblyof themembranousvesicle from Escherichia coli outer membrane components.Roleofindividual components and magnesium ions in reassembly.

Biochim. Biophys. Acta 413:371-393.

18. Nikaido, H. 1976.Outer membraneofSalmonella typhi-murium. Transmembrane diffusion of some hydro-phobic substances. Biochim. Biophys. Acta 433:118-132.

19. Ricard, M., and Y. Hirota. 1973. Effect des sels et autrescompos6ssurle phenotypedemutants thermo-sensibles de Escherichia coli. Ann. Microbiol. (Paris) 124:29-43.

20. Rosenbusch, J. P. 1974. Characterization of the major envelope protein from Escherichia coli. Regular ar-rangement onthepeptidoglycan and unusual dodecyl sulphate binding. J. Biol. Chem. 249:8019-8029. 21. Schmitges, C. J., and U. Henning. 1976. The major

proteins of the Escherichia coli outer cell envelope membrane. Heterogeneity of protein I. Eur. J. Bio-chem. 63:47-52.

22. Schnaitman, C. A. 1974. Outer membrane proteins of Escherichia coli. III.Evidence that the major protein of Escherichia coli0111 outer membrane consists of four distinct polypeptide species. J. Bacteriol. 118:442453.

23. Schnaitman, C. A. 1974. Outer membrane proteins of Escherichia coli. IV. Differences in outer membrane proteins due to strain and cultural differences. J. Bacteriol. 118:454-464.

24. Schweizer, M., H. Schwarz, I. Sonntag, and U. Hen-ning. 1976. Mutational change of membrane architec-ture. Mutantsof Escherichia coli K12 missing major proteins of the outer cell envelope membrane. Biochim. Biophys. Acta 448:474-491.

25. Smit,J., Y. Kamio, and H. Nikaido. 1975. Outer mem-braneof Salmonella typhimurium: chemical analysis and freeze-fracture studies with lipopolysaccharide mutants.J. Bacteriol. 124:942-958.

25a.van Alphen, L., L. Havekes, and B. Lugtenberg. 1977.Major outer membrane proteind ofEscherichia coli K12. Purification and in vitro activityon bac-teriophage K3 and F-pilus mediated conjugation. FEBS Lett. 75:285-290.

26. vanAlphen, W., B. Lugtenberg, and W. Berendsen. 1976. Heptose-deficient mutantsof Escherichia coli K12deficient in up to three major outer membrane proteins. Mol. Gen. Genet. 147:263-269.

27. Verhoef, C., P. J. de Graaf, and E. J. J. Lugtenberg. 1977.Mappingof a gene for a major outer membrane protein ofEscherichia coli K12 withthe aid of a newly isolated bacteriophage. Mol. Gen. Genet. 150:103-105.

28. VerkleiJ, A.J., E. J. J. Lugtenberg, and P. H.J. T. Ververgaert. 1976. Freezeetch morphology of outer membrane mutants ofEscherichia coli K12. Biochim. Biophys.Acta 426:581-586.

29. Verklei,A., L. van Alphen, J.Bivelt,and B. Lugten-berg. 1977. Architecture of the outer membrane of Escherichia coli K12. H. Freeze fracture morphology of wild type andmutantstrains. Biochim. Biophys. Acta 466:269-282.

http://jb.asm.org/