JOURNAL OF BACTERIOLOGY, Apr. 1974, p. 275-284 Vol. 118, No. 1Copyright i 1974 American Society for Microbiology Printed in U.S.A.

Subunit Cell Wall of Sulfolobus acidocaldariusRICHARD L. WEISS1

Department of Microbiology, Indiana University, Bloomington, Indiana 47401

Received for publication 27 August 1973

The cell wall of Sulfolobus acidocaldarius has been isolated. Cells weremechanically disrupted with a French press, and the cytoplasmic membrane wasremoved by extracting cell-envelope fragments with Triton X-100. The Triton-insoluble cell wall material retained the characteristic subunit structure whenexamined in the electron microscope. Isolated cell wall fragments formed in opensheets that were easily separated from cytoplasmic contamination. Chemicalstudies showed that the Triton-insoluble cell wall fragments consisted oflipoprotein with small amounts of carbohydrate and hexosamine. The amino acidcomposition indicated a highly charged hydrophobic cell surface. The presence ofdiaminopimelic acid with only traces of muramic acid indicates that the cellenvelope does not have a rigid peptidoglycan layer. The results of chemicalanalyses and electron microscopy suggest a wall-membrane interaction stabiliz-ing the cell envelope. The chemical and physical properties of this type of cellenvelope would appear to form the basis for a new major division of bacteria withthe definitive characteristics of a morphologically distinct subunit cell walldevoid of peptidoglycan.

The ability to survive at low pH and hightemperature distinguishes Sulfolobus aci-docaldarius from other bacteria (4). The cellwall and perhaps the cytoplasmic membraneare the only cellular components that must beexposed to both low pH and high temperature.The subunit cell wall lacks a morphologicallydefined peptidoglycan layer which is character-istic of the cell wall of gram-negative bacteria.Subunit cell walls, devoid of peptidoglycan,have also been reported on bacteria from ex-treme saline environments (6, 8). Althoughsome information on these cell walls is available(18), little is known of the structured cell wall ofS. acidocaldarius. In a previous paper, wereported on the biochemical composition ofwhole-cell hydrolysates of S. acidocaldarius,but no chemical analysis of the cell envelopewas presented (4). In this report, the effects ofseveral treatments on the cell envelope areexamined. This work was undertaken for thefollowing reasons: (i) to provide a means forseparating the cell wall from the cytoplasmicmembrane, (ii) to obtain information on thenature of covalent and noncovalent bonds in-volved in maintaining the subunit structure ofthe cell wall and (iii) to provide information onthe chemical composition of the cell wall.

'Present address: The Biological Laboratories, HarvardUniversity, Cambridge, Mass. 02138

MATERIALS AND METHODSS. acidocaldarius (ATCC no. 27360) was used in

most experiments. Cultures were grown at 70 C, pH 2,in basal salts medium with 0.1% yeast extract (4).The following chemical reagents were used: diso-

dium ethylenediaminetetraacetic acid (EDTA)(Sigma Chemical Co., St. Louis, Mo.), sodium do-decyl sulfate (SDS) (Matheson, Coleman and Bell,Cincinatti, Ohio), dithiothreitol, Triton X-100 (Calbi-ochem, Los Angeles, Calif.) and guanidine hydrochlo-ride, and dimethyl sulfoxide (Eastman OrganicChemicals, Rochester, N.Y.).The enzymes used, trypsin (EC 3.4.4.4), 2x-crys-

tallized, and lysozyme (EC 3.2.1.17), 3x-crystallized,were obtained from Nutritional Biochemicals Corp.,Cleveland, Ohio. Pronase (protease, B grade) waspurchased from Calbiochem, La Jolla, Calif. Cellswere washed three times in 10 mM N-2-hydroxyethyl-piperazine-N'-2'-ethanesulfonic acid (HEPES)buffer (pH 7) before treatment with various enzymesand reagents. Treatments with proteolytic enzymeswere carried out in 10 mM HEPES buffer (pH 7) for24 h at 37 C. Treatment with dimethyl sulfoxide wasfor 24 h at 60 C. Lysozyme treatment has recentlybeen described (15). Escherichia coli was used as acontrol to show that the lysozyme system was active.

Biochemical assays were performed on intact cellsor subfractions of cells. For protein assays, sampleswere extracted with 1 N NaOH for 60 min at 23 C andassayed by the method of Lowry et al. (12). Crystal-line lysozyme was used as a standard. For hexosamineassays, samples in sealed tubes were hydrolyzed with3 N HCl for 4 h at 95 C. The sample was neutralized

275

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

with 3 N NaOH, and hexosamine was determined (19)by using glucosamine as a standard. Carbohydrateassays were performed by the phenol-sulfuric acidmethod of Hodge and Hofreiter (10), with glucose as astandard.

To prepare cell envelopes, midlogarithmic-phasecultures were harvested by centrifugation for 15 minat 8,000 x g and washed three times in glass-distilledwater. The cells were suspended in 10 mM HEPESbuffer (pH 7) and lysed by raising the pH to 10.5 with1 N NaOH. The suspension was then quickly neutral-ized with 1 N HCl, centrifuged for 20 min at 10,000 xg, washed three times in 10 mM HEPES buffer (pH7), and used for chemical analysis.To isolate cell walls, cultures were grown to the

early logarithmic phase (107 to 108 bacteria per ml),harvested by centrifugation for 15 min at 8,000 x g,and washed three times in 10 mM HEPES buffer (pH7). A concentrated cell suspension in 10 mM HEPESbuffer (pH 7) was passed through a French press at apressure just sufficient to obtain cell breakage (6,000lb/in2). The cell-envelope fragments were then cen-trifuged at 17,000 x g for 10 min and washed threetimes in 10 mM HEPES buffer (pH 7). Next, thesample was suspended in 20 ml of 10 mM HEPESbuffer (pH 7), and an equal volume of 4% Triton X-100in 10 mM HEPES buffer (pH 7) was added to removethe cell membrane. The Triton-insoluble cell wallfraction was washed three times in 10 mM HEPESbuffer (pH 7), concentrated by centrifugation (17,000x g) for 10 min, and used for chemical analysis.The amino acid composition was determined by the

general method of Spackman, Stein, and Moore (16).Muramic acid and diaminopimelic acid were detectedby modification of the column temperatures and thebuffer system (14) by using a Beckman model 121amino acid analyzer with 12-mm cuvettes and rangecard.Tryptophan was determined by the method of

Spies and Chambers (17). Cystine was determined ascysteic acid after performic acid oxidation (20: 1,HCOOH-H202, vol/vol) for 30 min at 25 C.

Samples were prepared for thin-sectioning by adouble fixation technique (7). Samples in 10 mMHEPES buffer (pH 7) were prefixed overnight at 0 to 4C in 1% glutaraldehyde. After fixation, samples werewashed twice in 10 mM HEPES (pH 7) and fixed for 1h at 23 C with 3'S glutaraldehyde in 5 mM HEPES(pH 7.2). The preparation was then washed free ofglutaraldehyde with 5 mM HEPES buffer (pH 7.2)and fixed for 1 h at 23 C with 1%' OS04 in 5 mMHEPES (pH 7.4). Dehydration was carried out withincreasing concentrations of ethanol after which sam-ples were transferred to acetone, embedded, andsectioned as previously reported (4). For negativestaining, 1% aqueous uranyl acetate was used. Themethods used for negative staining have been de-scribed (4). For freeze-fracturing and -etching, log-phase cells were harvested by centrifugation for 15min at 8,000 x g and washed three times in eitherdistilled water or 10 mM HEPES buffer (pH 7). Cellswere concentrated by centrifugation for 15 min at8,000 x g and mounted on a copper specimen holder.Cells were frozen in partially solidified Freon 22 cooled

with liquid nitrogen. Freeze-fracturing and -etchingwere performed with a Balzers unit as previouslydescribed (13). Freeze-fracturing was carried out at110 C, and freeze-etching was done at 104 C. All

samples were observed with a Hitachi HU-11C elec-tron microscope operated at 50 kV or a Philips 300operated at 60 kV.

RESULTSCharacteristic multilobed cells of S.

acidocaldarius are illustrated in Fig. 1. The cellwall is an unusual feature of this organism. Inmany sections, the subunits forming the cellwall are evident. Figure 1 shows the subunits incross section and in transverse grazing section.The well-defined cytoplasmic membrane is ob-served in profile as two dense outer layersseparated by a light inner layer. The denselystaining cytoplasm is interrupted by clear areaspresumed to be deoxyribonucleic acid.The electron-transparent subunits forming

the cell wall of S. acidocaldarius were clearlyshown in negatively stained preparations. Thegenerally polygonal subunits appear over thesurface and on a cell wall fragment (Fig. 2).Deposits of negative stain in the central core ofthese subunits suggest that the polygonal unitshave a hollow core. The diameter of the sub-units as measured in negatively stained prepa-rations was 15.5 nm with a center-to-centerspacing of about 20 nm. The size of subunits didnot vary when cells were grown autotrophicallyon sulfur or heterotrophically on yeast extract.The subunits forming the cell wall of autotroph-ically grown cells were identical in appearancewith those shown in negatively stained prepara-tions (Fig. 2). Under the experimental condi-tions used, the size of the subunits remainedconstant after prolonged periods of incubationat low pH and high temperature.Negatively stained cells emphasize the lobed

shape of S. acidocaldarius (4) as seen in Fig. 2,but the surface contour of the cellular lobes isshown best in freeze-etched preparations.Freeze-fracturing did not expose the interior ofthe cell wall or the hydrophobic interior of thecytoplasmic membrane. The cells were cross-fractured, revealing a somewhat granular cyto-plasm (Fig. 3). Subsequent etching of the frac-tured cells exposed a regular hexagonal array ofsubunits on the outer convex cell wall surface(Fig. 4). The individual subunits forming thecell wall appear to be identical with thoseobserved in thin-sectioned and negativelystained cells. Measurements made on subunitsin freeze-etched cells show a diameter of 15.9nm which is comparable to the 13- to 15-nmdiameter of subunits seen in thin-sectionedpreparations.

276 WEISS

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SUBUNIT CELL WALL OF S. ACIDOCALDARIUS

#.A-.

0f

k, I*

FIG. 1. Thin section of multilobed S. acidocaldarius showing the subunit cell wall in cross-section and ingrazing-section. Triple-layered cytoplasmic membrane appears in profile. Bar, 1.0 gm.

FIG. 2. Negatively stained S. acidocaldarius showing a face view of a fragment torn away from the cell wall.The electron-transparent subunits contain deposits of negative stain, suggesting a hollow core. Stained with 1%uranyl acetate. Bar, 0.5 im.

277VOL. 118, 1974

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

S

X.

A

-4To e*

e ......--VaF; .......... .'

..C,#._w

a. s. ,, W,, ;s5-

A

FIG. 3. Freeze-fractured S. acidocaldarius. Cells were frozen in 10 mM HEPES buffer (pH 7) and fracturedat -110 C. Shadow direction is indicated by the arrow. Cells were cross-fractured. Bar. 0.5 pm.

FIG. 4. Freeze-etched S. acidocaldarius. Cells were etched 45 s at - 104 C. Etching exposed the regularhexagonal array of subunits on the outer cell wall surface. Cells uere frozen in distilled water. Shadow directionis indicated by the arrow. Bar, 0.5 pm.

The biochemical characteristics of the cell 37 C) had no effect on the cell wall. Lysozyme-wall were investigated by testing the effects of EDTA (15), which converted intact cells of E.various enzymes and reagents on the structure coli into spheroplasts, had no effect on S.of intact cells. Trypsin and Pronase (0.5 mg/ml, acidocaldarius. Treatment of cells with 6 M

278 WEISS

ver-,or1

.1

,

:. -7,

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SUBUNIT CELL WALL OF S. ACIDOCALDARIUS

guanidine hydrochloride (pH 7, 37 C), a hydro-gen bond-breaking reagent, caused a partial orcomplete loss of the organized cell wall struc-ture (Fig. 5).The effect of SDS on the structure and

integrity of the cell envelope was controlled bythe concentration of the reagent as well as thetemperature of incubation. An aqueous 1% SDSsolution (pH 5.5) at 23 C caused a 55% reduc-tion in optical density. The addition of 10 mMEDTA (pH 7) to the 1% SDS (23 C) solutionreduced the optical density by 23% after 30 min,suggesting that metal ions may play a role inthe stability of the cell envelope. The morphol-ogy of SDS (1%)-treated cell walls was essen-tially the same after treatment with EDTA.After extraction with 1% SDS at 90 C, an arrayof subunits with frayed edges and a centralhollow core was observed in negatively stainedpreparations. Treatment of cells with 10% SDSat 23 C fragmented the cell envelope intosmaller straight-sided cell wall sheets, withoutaltering the appearance of negatively stained

subunits. Further treatment with 10% SDS at90 C, which extracts lipoproteins and lipopoly-saccharides (3), solubilized the cell wall.

Extraction of cells with 67% DMSO at 60 C,which removes lipopolysaccharides (1), had noeffect on cell wall stability or the appearance ofcells. A 1% SDS (23 C) solution at pH 10.5caused cell lysis (Fig. 6), but did not alter thecell wall structure. The use of dithiothreitol, adisulfide bond-breaking reagent, at pH 10.5with 1% SDS (23 C) brought about the completedisintegration of the cell wall (Fig. 6). Nega-tively stained preparations of this materialappeared amorphous.The procedure adopted for the isolation of cell

walls makes use of the removal of the cellmembrane by Triton X-100 (2). The morphol-ogy of Triton-treated cells is illustrated in Fig.7, in which a Triton-treated cell is matched withan untreated cell. The untreated cell (Fig. 7a)contains densely staining cytoplasmic materialthat consists of seemingly typical ribonuclearprotein particles. Surrounding the cytoplasm is

0

FIG. 5. S. acidocaldarius treated with 6 M guanidine hydrochloride (pH 7). Smooth cell wall surface andextracted material are evident. Intact log-phase cells were harvested by centrifugation at 8,000 x g for 15 min,washed three times in 10 mM HEPES buffer (pH 7), and suspended in 6M guanidine hydrochloride (pH 7) for45 min at 37 C. Cells were negatively stained with 1% uranyl acetate (compare with Fig. 2) Bar, 0.5 Aim.

-

279VOL. 118, 1974

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

0.2-

Toi.

a~~~~~~~~

MI11NUTES

FIG. 6. Lysis of S. acidocaldarius after treatmentwith dithiothreitol. Cells were harvested in log phalse,washed three times in 5 mM HEPES buffer (pH 7),adjusted to an optical density of 0.25, and incubatedat 23 C with 1% SDS (pH 10.5) (a) and l1o SDS (pH10.5)-50 mM dithiothreitol (A). Untreated cells in 5mM HEPES buffer (pH 7) showed no decrease froman optical density of 0.25 during the 8-min incubationperiod.

the cell wall and a "unit" membrane with anoverall width of about 7.5 nm. The cytoplasmicmembrane is not observed in the Triton-treatedcells (Fig. 7b). The Triton-insoluble cell wallsurrounds lightly stained cytoplasmic material.In the procedure used to isolate cell walls, intactcells were initially disrupted by passing washedbacteria through a French press twice. Theresulting fraction consisted mainly of open cell-envelope fragments that retained the character-istic cellular shape. These fragments were thenwashed in buffer and treated with Triton X-100.After Triton treatment, the cell wall fragmentslost the characteristic cellular shape and formedopen sheets with the characteristic subunitstructure (Fig. 8). The results of ultravioletspectra at 255 to 260 nm showed that these cellwall sheets were easily freed of nucleic acids bywashing.When cells were lysed at pH 10.5, it was

found that on a per cell basis about 65% of theprotein was released, whereas only 34% of thetotal carbohydrate was released. On the otherhand, 47% of the total hexosamine was released.These data indicate that a significant amount ofcarbohydrate and hexosamine is located in thecell envelope. Chemical analysis of the cellenvelope shows a protein-hexosamine-carbohy-drate ratio of 1:0.029:0.241 (wt/wt/wt). Whenthe cell membrane is removed with TritonX-100, a protein-hexosamine-carbohydrateratio of 1:0.026:0.140 (wt/wt/wt) is obtained.The reduction in the carbohydrate fraction from0.241 to 0.140 after Triton treatment impliesthat carbohydrate is associated with both thecell wall and the cytoplasmic membrane. Triton

FIG. 7. Effect of Trition X-100 on S. acidocaldarius. (a) Section through an intact cell; (b) section through acell treated with 2% Triton X-100 at 23 C for 20 min. Cell wall on the intact cell is matched with thecorresponding cell wall on the Triton-treated cell. Bar, 0.5 Am.

280 WEISS

.4. .- ---w i

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SUBUNIT CELL WALL OF S. ACIDOCALDARIUS

.,a,

.f*

@t; , ': .~~~

,~~~~~~~~~d.

P

FIG. 8. Cell wall preparation obtained by passing S. acidocaldarius through a French pressure cell afterwhich the cell fragments were treated with 2%o Triton X-100 at 23 C and washed in 10 mM HEPES buffer (pH7.4). Cell wall subunits are evident on the open sheets. Bar, 0.5 um.

treatment caused the release of protein withoutchanging the ratio of protein to hexosamine,suggesting that hexosamine is present in thecytoplasmic membrane. When the protein-hex-osamine-carbohydrate ratio of alkali-treatedcell wall fractions (1:0.026:0.140, wt/wt/wt) iscompared with that obtained after Triton treat-ment of French press extracts (1:0.006:0.065,wt/wt/wt), it is clear that most of the hexosa-mine and carbohydrate are associated with thecytoplasmic membrane; these two componentsare removed more readily from French pressextracts, presumably because the cell wall frag-ments form as open sheets free of contamina-tion.The data from the chemical analysis of the

Triton-insoluble fraction show that the cellwalls contain a large proportion of protein. Thecomposition of this protein is presented in Table1. It appears that the cell wall fraction isenriched with charged amino acids (aspartate,glutamate, and lysine) as well as withbranched-chain hydrophobic amino acids (va-line, leucine, and isoleucine). Special methods(14) were used to detect components of the rigid

layer, diaminopimelic acid and muramic acid.Cell walls hydrolyzed for 22 h at 110 C or 8 h at100 C contained small amounts of diaminopi-melic acid with only traces of muramic acid. Inaddition, two unidentified peaks were observed(Fig. 9). The first peak appeared after lysineand before histidine with about 30% of theabsorbance displayed by histidine. The secondpeak, which was seen after leucine and in frontof tyrosine, had about 10% of the absorbance oftyrosine. One peak representing hexosaminewas also observed.

DISCUSSION

The unusual cell wall of S. acidocaldarius isof interest, because it comes in contact with lowpH and high temperature. The procedure de-scribed in this study for isolating the cell wallhas two advantages over other procedures whichhave been used with morphologically similarcell walls (18). First, extreme ionic conditionsare not required during preparation; isolationcan be carried out at neutral pH with thepreservation of the subunit structure of the cell

VOL. 118, 1974

,5

281

'i

I

- '.,I ae

"

SNM- I.

li ".".1. ..

I

vl-.I...

qa.. -1. X

94,

.k..

WA ..

Al,

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

TABLE 1. Amino acid composition of the cell wall ofS. acidocaldarius

Amino acid M per 100 mg Mo) (%-

Aspartic acid ........... 29.47 9.14Glutamic acid .......... 33.27 10.32Lysine ............... 22.30 6.92Arginine ............... 13.58 4.21Histidine.3.60 1.12Glycine ............ 32.59 10.11Alanine ................ 23.18 7.19Valine ........... 25.31 7.85Leucine ........... 28. 46 8.83Isoleucine .............. 21.26 6.60Serine ... ......... 21.40 6.64Threonine .............. 19.79 6.14Cysteic acida ........... 3.35 1.04Methionine .......... 1.31 0.41Proline .......... 16.02 4.97Hydroxyprolineb ........Phenylalanine .......... 10.73 3.33Tyrosine ............... 13.46 4.18Tryptophanc ........... 3.14 0.97Diaminopimelic acidd 1.12 0.35

Total 322.34 100.32a Detected after performic acid oxidation.b Not detected.c Detected by the method of Spies and Chambers

(17).d Detected by the method of Peterson and Bernlohr

(14).

I

£0CEE

c._b._

wall. This permits the morphological identifica-tion of the cell wall for subsequent biochemicalanalysis. Second, the isolated cell walls remainas open fragments rather than closed vesicles,which yield pure preparations of isolated cellwalls. It is likely that the formation of openfragments is related to the difference betweenthe isolation temperature and that required forreannealing. This difference may be importantin future studies on cell walls and cytoplasmicmembrane vesicles.The results presented here allow some conclu-

sions to be drawn about the types of covalentand noncovalent bonds that contribute to thesubunit structure of the intact cell wall. Non-covalently-linked protein in the cell wall isdisaggregated only under rather severe condi-tions, namely by extraction of lipoprotein andlipopolysaccharide with hot SDS (3). Becauseextraction of lipopolysaccharide alone with di-methyl sulfoxide had essentially no effect oncells (1), it appears that the cell wall is aprotein-lipid complex. Although the arrange-ment of the components of the cell wall isunknown, it seems clear that hydrogen bondsparticipate in maintaining the integrity of thesubunit structure, because high concentrationsof guanidine hydrochloride or urea effectivelydisorganized the subunit structure of the cellwall. It is likely that the resistance to treat-

Short Column

120 150 0 30 90 120 1o go*

_. _, \Long Column

Elution Time (minutes)

FIG. 9. Elution profile of the Triton-insoluble cell wall fragments of S. acidocaldarius. Three curves showabsorbance at 570 nm using two cell depths (dashed-line and large dots) and at 440 nm (small dots) fordetermination of proline. Known and unknown ninhydrin positive peaks are labeled. Short column (basicamino acids) was eluted with 0.35 M sodium citrate. pH 5.28, while the long column (acidic amino acids) waseluted with 0.2 M sodium citrate at pH 3.25 and after methionine at pH 4.25.

282 WEISS

:2

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

SUBUNIT CELL WALL OF S. ACIDOCALDARIUS

ments that affect noncovalent bonds is due inpart to the stabilization of lipoprotein by diva-lent cations, which would explain the effect ofEDTA during treatment of cells with low con-centrations of SDS. In addition to observationsof noncovalently-linked protein, it was notedthat the cell wall was sensitive to treatmentwith dithiothreitol, a disulfide bond-breakingreagent. This strongly suggests that the lipo-protein component is responsible for cell wallintegrity, presumably by hydrophobic protein-protein interaction.The loss of the characteristic cellular shape

after the removal of the cytoplasmic membrane,in combination with the results obtained on thecarbohydrate and hexosamine content of thecell wall, indicates a strucural association be-tween the cell wall and the cytoplasmic mem-brane. The wall-membrane interaction couldoccur through insertion of carbohydrate andhexosamine of the cytoplasmic membrane intothe cell wall. This interaction would tend tostabilize both the cell wall and cytoplasmicmembrane, making the cell envelope morerigid. The cell wall contains a high proportion ofcharged amino acids. Without correcting foramide, which presumably is rapidly degraded atpH 2 above 70 C, it was found that the cell wallcontains an excess of acidic over basic aminoacids at a percentage of about 9 mol. Thisvalue is similar to that of mesophilic bacterialiving in neutral habitats (11), but well belowthat of other bacteria from extreme ionic envi-ronments (9, 18).

It seems clear from the results described herethat the cell wall of S. acidocaldarius is unique.The inability to detect significant amounts ofmuramic acid and the observation that lyso-zyme-EDTA has no effect on the integrity of thecell wall leave little doubt that the wall of S.acidocaldarius is fundamentally different fromthat of gram-positive or gram-negative bacteria.Very few organisms with similar cell walls havebeen studied chemically or by electron micros-copy. In addition to S. acidocaldarius, onlyHalobacterium (6, 8, 18), thermophilic sulfurbacteria (5), and rod-shaped acido-thermoph-ilic bacteria (Weiss, unpublished data) seemto have subunit cell walls that lack a rigidpeptidoglycan layer. Organisms of this typewould appear to form a new major division ofbacteria distinguished from gram-positive andgram-negative bacteria by the physicochemicalfeatures of the subunit cell wall.The results presented here support the idea

that survival of S. acidocaldarius may dependon the existence of an unusual cell wall charac-

terized as follows: (i) morphologically distinctsubunit structure devoid of peptidoglycan, (ii)lipoprotein subunits stabilized by divalent cat-ions, (iii) wall-membrane interaction stabilizingthe cell envelope, (iv) highly charged hydro-phobic cell surface. The most unusual charac-teristic of the cell wall is the lack of peptidogly-can. Indeed, one can speculate that, in low pH,high temperature habitats, peptidoglycan isunstable and the nonpeptidoglycan wall of S.acidocaldarius is a factor permitting adaptationto the extreme conditions.

ACKNOWLEDGMENTSThis work was supported by U.S. Public Health Service

grant no. 5T1-GM-503 from the National Institute of GeneralMedical Sciences.

Facilities provided by the Electron Microscope Laborato-ries at the University of California, Berkeley, Calif., andHarvard University, Cambridge, Mass., and Public HealthService grant GM-06637-15 from the National Institute ofGeneral Medical Sciences are acknowledged.

I thank C. W. H. Hirs for helpful discussion and the use oflaboratory facilities. The skilled assistance of G. Mross andJ. J. Schmidt is gratefully acknowledged.

LITERATURE CITED1. Adams, G. A. 1967. Extraction of lipopolysaccharides

from gram-negative bacteria with dimethyl sulfoxide.Can. J. Biochem. 45:422-426.

2. Birdsell, D. C., and E. H. Cota-Robles. 1968. Lysis ofspheroplasts of Escherichia coli by a non-ionic deter-gent. Biochem. Biophys. Res. Commun. 31:438-446.

3. Braun, V., and K. Rehn. 1969. Chemical characteriza-tion, spatial distribution and function of a lipoprotein(murein-lipoprotein) of the E. coli cell wall. Eur. J.Biochem. 10:426-438.

4. Brock, T. D., K. M. Brock, R. T. Belly, and R. L. Weiss.1972. Sulfolobus: A new genus of sulfur-oxidizingbacteria living at low pH and high temperature. Arch.Mikrobiol. 84:54-68.

5. Brock, T. D., M. L. Brock, T. L. Bott, and M. R.Edwards. 1971. Microbial life at 90 C: the sulfurbacteria of Boulder Spring. J. Bacteriol. 107:303-314.

6. Brown, A. D. 1964. Aspects of bacterial response to theionic environment. Bacteriol. Rev. 28:296-329.

7. Carrick, L., Jr., and R. S. Berk. 1971. Membranousinclusions of Pseudomonas aeruginosa. J. Bacteriol.106:250-256.

8. Cho, K. Y., C. H. Doy, and E. H. Mercer. 1967.Ultrastructure of the obligate halophilic bacteriumHalobacterium halobium. J. Bacteriol. 94:196-201.

9. Crum, E. H., and D. J. Siehr. 1967. Thiobacillusthiooxidans cell wall amino acids and monosaccha-rides. J. Bacteriol. 94:2069-2070.

10. Hodge, J. E., and B. T. Hofreiter. 1962. Determination ofreducing sugars and carbohydrates, p. 380-394. In R. L.Whistler, M. L. Wolfrom, J. N. BeMiller, and F.Shafizadeh (ed.), Methods in carbohydrate chemistry,vol. 1. Academic Press Inc., New York.

11. Howe, J. M., W. R. Featherston, W. J. Stadelman, andG. J. Banwart. 1965. Amino acid composition of certainbacterial cell-wall proteins. Appl. Microbiol.13:650-652.

12. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

283VOL. 118, 1974

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

284 WEISS

13. Moor, H., and K. Miihlethaler. 1963. Fine structure infrozen-etched yeast cells. J. Cell. Biol. 17:609-628.

14. Peterson, D. E., and R. W. Bernlohr. 1970. Determinationof muramic acid, ornithine, and diaminopimelic acidduring automatic amino acid analysis. Anal. Biochem.33:238-243.

15. Repaske, R. 1956. Lysis of gram-negative bacteria bylysozyme. Biochim. Biophys. Acta 22:189-191.

16. Spackman, D. H., W. H. Stein, and S. Moore. 1958.Automatic recording apparatus for use in the chroma-tography of amino acids. Anal. Chem. 30:1190-1206.

J. BACTERIOL.

17. Spies, J. R., and D. C. Chambers. 1949. Chemicaldetermination of tryptophan in proteins. Anal. Chem.21:1249-1266.

18. Steensland, H., and H. Larsen. 1969. A study of the cellenvelope of the Halobacteria. J. Gen. Microbiol.55:325-336.

19. Tipper, D. J. 1968. Alkali-catalyzed elimination ofD-lactic acid from muramic acid and its derivativesand the determination of muramic acid. Biochemistry7:1441-1449.

on May 8, 2020 by guest

http://jb.asm.org/

Dow

nloaded from