JOURNAL OF BACTERIOLOGYVol. 88, No. 4, p. 1155-1162 October, 1964Copyright © 1964 American Society for Microbiology

Printed in U.S.A.

DENSITY-GRADIENT PATTERNS OF STAPHYLOCOCCUSAUREUS CELLS AND CELL WALLS DURINGGROWTH AND MECHANICAL DISRUPTIONESKIN HUFF, HARRIET OXLEY, AND CAROL S. SILVERMAN

Laboratory of Infectious Diseases, National Institute of Allergy and Infectious Diseases,U.S. Public Health Service, Bethesda, Maryland

Received for publication 1 June 1964

ABSTRACT

HUFF, ESKIN (National Institute of Allergy andInfectious Diseases, Bethesda, Md.), HARRIETOXLEY, AND CAROL S. SILVERMAN. Density-gradi-ent patterns of Staphylococcus aureus cells and cellwalls during growth and mechanical disruption.J. Bacteriol. 88:1155-1162. 1964.-Procedures ca-pable of rapid disruption of Staphylococcus aureuscells with optimal release of intact cell walls wereinvestigated. This search was implemented by ob-servation of the flotation patterns of cells and sub-cellular particulate matter after centrifugation ina cesium chloride density gradient. A quantitativeevaluation of the light-scatter throughout thegradient was achieved by transfer of the entiredensity gradient into an optical cell with wedge-shaped cross section. When this cell was photo-graphed under indirect illumination, each band oflight-scattering material appeared on the negativeas a shaded curve, with an area proportional toamount of that material present. A series of photo-graphs of known amounts of cells and cell wallswas used to estimate the amounts of these mate-rials in mixtures of the two occurring duringmechanical disruption. With the methods em-ployed, time studies established the optimal timefor release of cell walls as 5 min in a Braun shaker.The use of sucrose gradients in the further puri-fication of cell walls, and chemical analysis of theisolated walls, are described.

Salton (1960) pointed out that the preferredmethod for preparation of undegraded cell wallsinvolves disruption of cells by means of shakingwith glass beads, followed by differential centrif-ugation. The devices of Mickle (1948), Cooper(1953), Nossal (1955), Shockman, Kolb, andToennies (1957), and Merkenschlager, Schloss-mann, and Kurz (1957) were used to disrupt thecells of Staphylococcus aureus and other bacteria.However, no careful study has been made of theoptimal conditions for release of cell walls. Thisappears to be due to lack of a satisfactory method

for characterization of cell walls other than byexamination in an electron microscope.

Meselson, Stahl, and Vinograd (1957) ob-served that the deoxyribonucleic acid (DNA)from an organism has a specific density whendistributed in a cesium chloride equilibriumdensity gradient. It seemed conceivable thatcell walls would have a similar specific densityin CsCl, and could thus be distinguished fromwhole cells. An experiment was carried out inwhich cell walls were prepared from staphylococ-cal cells harvested at various stages during a 12-hrgrowth period. Throughout this period, thedensity of the cell walls was constant at 1.415 ±0.005. This result encouraged the developmentof a new procedure in which cell walls and cellscan be characterized and quantitated in a cesiumchloride gradient. This paper describes methodswhich can be utilized to follow the time courseof release of cell walls, destruction of cells, andformation of intermediate particles. The isolationof cell walls and their characterization by electronmicroscopy, chemical composition, and sedimen-tation characteristics in sucrose and flotation inCsCl are described.

MATERIALS AND METHODS

Organism. S. aureuss 8507 (Oeding) was usedin all experiments. Slant cultures of this organismwere maintained on a medium consisting of 1.5%agar in Trypticase Soy Broth (BBL). A 7-hrTrypticase Soy Broth culture grown at 37 Cwas used to inoculate 100 ml of broth, which inturn, after 16 hr was used to inoculate 1 liter ofbroth. Growth was continued aerobically withoutagitation for 5 hr. A batch of 30 liters of brothculture was harvested in a Sorvall continuous-flowcentrifuge over a 2-hr period. The effectivegrowth time here was defined as the sum of thetime from inoculation to harvest plus one-halfthe harvest time. Thus, the effective growth time

1155

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

HUFF, OXLEY, ANI) SILVERAIAN

of the above l)reparation would be 6 hr. The cellswere washed in 1,700 ml of distilled water at 0 C.The packed cells were then resuspended in 90rnl of distilled water and frozen overnight. Afterthawing, the cells were centrifuged foI 10 minat 20,000 rev/min in a Spinco swinging-bucketrotor (SW-25.1). Cells were resuspended in asmall volume of water to a calculated opticaldensitv of 100 at 660 iniu [50 miig (dry weight) ofcells per ml]. The y-ield of diluted cells was 125to 130 ml, with a viable count of 9 X 10° perml. These suspensions were then frozen in 20-im1lbatches until just before use.

Glass beads. Superbrite glass beads (0.003 in.in diameter) were purchased fr-om the M\innesotaMlining and Manufacturing Co., St. I'aul, M\inn.MWagnetic particles were remnov-ed by severalpassages over a strong magnet an(l flotation ofthe beads oIn bromoform. Bromoformi was re-moved by washing with acetone, and the beadsweere then air-dried at room temperature. Beadsrecovered from a Nossal or BIraun shaker run werewashed with water, treated with a sodium di-chromate-sulfuric acid mixtutre, washed withwater, washed with acetone, and finally air-dried.Because of the presence of colloidal stainless steel,beads used with a Blranson 20-ke sonifier werediscarded.

Nossal disintegrator. A large-scale mnechanicalshaker consttructed by the Instrument Shop,National Institutes of Health, according to Nossal(1955) was fitted with tightly capped, cylindricalnyilon inserts (2.17 ciil, inside diameter, by 9.8cm long; total volume = 36.6 ml). For disinte-gration, 20 ml of cell suspension [22 to 44 Imig(dry weight) per ml] were miiixed witlh 24 g ofglass beads. lBecause of the rapid frictional heat-ing of 1 C per see, eaclh nylon insert was shakenfor no longer than 20 sec at a time, an(1 then wasremoved to an ice batlh for 15 to 20 min to coolback to 0 C before the subsequenit shlake.Braun shaker. Reagent bottles (60 ml, Corning

no. 1500; total volumiie 67 nil) weie loaded with36 m-il of cell suspension and 44 g of glass beads.Shaking was carried out in the ap)paratus (le-scribed by AMerkenschlager et al. (1957) andmanufacturedl by The Bratun Co., Aelsungen,Germany. In this method, the temperature ofthe cell suspension is maintained at 0 to 5 Cthroughout disintegration.Branson 20-kc sonifier. A suspension of 20 ml

of cells was treated in a stainless-steel jacketed

chamber ma(le to fit the probe of the l3ranson20-kc sonifier. Water at 5 C was circulatedthrouigh the jacket, and the temperature withinthe chamber was thus maintained at less than 10C.

Suicrose gradients. Linear sucrose gradients wereprel)ared as described by Yoshida et al. (1961),with the use of a gradient mixing device similarto that of Blritten and Roberts (1960). Allsucrose solutions w-ere filtered through 'MilliporeHA filters. A volume of 18.6 mil of 51,r% sucrosein I M Cl was placed in the left unstirred cham-ber of the gradient miixing dev-ice. The rightstirre(1 chamber contained 16.7 ml of 40%sucrose in 1 NI KCl. E'ach gradient was deliveredinto a centrifuge tube (Sorvall no. 204). Aftercentrifugation at 3,100 rev/min in an Interna-tional I'R-1 centrifuge for 30 or 70 min, thegradient tubes w-ere l)unctured in a device similarto that described by -Martin and Ames (1960),and a total of 37 1-mil fractions was collected.The turbidity of cell walls ml-easured at 660 mywas rieduced almost threefold when measuredin 40%' sucrose. Therefore, it was necessary tomultiply the optical density mneasurements at660 mi, of cell walls in sucrose by a correctionfactor which varied from 1 to 3, depending onthe final sucrose coneentration in wlhich theoptical density was imieasured.

Chemtical analyses. Analyses Nere carried outon samples of cell-wall susl)ensions dried in avacuum desiccator ov-er anhvdrous CaSO4.Total phosl)hate was determined by the proce-dure of Cheri, Toribara, and AVarner (1956) asmodified bv Ames and Dubin (1960), as well asby the method3 of Goim1ori (1942). Determinationof ribonucleic acid (RN'A) and DN-A was by theorcinol and diphenylamine methods, res.pectively,as (lescribed by Schneider (1957).

RE SULTS

Establishmnent of flotation patterns of cells andcell walls. After-varying periodls of disruption, asamnple of 0.10 ml or less waas placedl in a testtube in ice. Water was added to bring the volumeto 0.10 ml, and then CsCl (0.90 ml; density,1.390) was added to bring the final (lensitv to1.352; 12 hr to 7 days were allowed for equilibra-tion with CsCl. After longer peliods of time, twobands instead of one were plesent in the whole-cell region of the CsCl gradient. After equilibra-tion, a CsCl step gradient was prepared by

1156 J . BACTERIOL .

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

DENSITY-GRADIENT PATTERNS OF S. AUREUS

placing 0.33 ml of tetrabromoethane in the bot-tom of a cellulose nitrate tube, followed by thesequential addition of 1.00 ml of each of thecesium chloride solutions of densities 1.472,1.409, 1.352 (containing the disrupted cells),and 1.306. 'I'o prevent evaporation of water, thegradient was then overlaid with light paraffinoil. The tubes were centrifuged (Beckman modelL ultracentrifuge) for 3 hr at 33,000 rev/min in aSpinco SW-39 rotor. Control experiments showedthat equilibrium had been reached by this time,since further centrifugation for 22 hr did notchange the distribution patterns.

Transfer of gradient to optical cell. After centrif-ugation, the tubes were removed from the rotor,and the bulk of the paraffin oil was removedby suction. Remaining oil was removed by over-laying the gradient with benzene, followed bysuction removal of the benzene. The bottom ofthe centrifuge tube wNas punctured by a no. 21hy,podermic needle connected via polyethylenetubing to a reservoir of tetrabromnoethane (Fig.1). A specially constructed optical glass cell(Optical Cell Co., Inc., Brentwood, Nd.) ofwedge-shaped cross section, adapted to fit thecentrifuge tube, was placed above the centrifugetube, forced onto it to provide a water-tight seal,and held in place by means of a clamp. Theentire gradient was raised up into the wedge-shaped section of the cell by lifting the reservoirabove this cell and allowing the tetrabromo-ethane to displace the gradient. When the gra-dient was properly positioned for photography,the flow of tetrabromoethane was stopped by aclamp. The cell was then illuminated by indirectlight (Fig. lb). Photographs were taken withKodak Panatomic X roll film, and were devel-oped with a fine-grained developer understandardized conditions of time and temperature.The image size on the negative was 1.5 timesnormal, allowing direct observation of the pat-terns on the negative. Only the wedge-shapedportion of the optical cell, indicated by the thicklines above the dashed line in Fig. I a, was re-produced in the photographs presented here.Figure 2 shows examples of the patterns of cellsand cell walls obtained in CsCl by this procedure.

Determination of cells and cell walls in a gradient.Known amounts of whole cells (1 to 40 ,ug, dryweight) and cell walls (14 to 140 ,ug, dry weight)were centrifuged in a CsCl step gradient, andphotographs of the gradients were l)repared as

FIG. 1. Transfer of CsCl gradient fromrl. centrifugetube to optical cell.

described above. In the range studied, there wasa direct relationship between the shaded area onthe photographic negative and the dry weight(Fig. 3). Unknown amounts of cell walls and cellsin a mixture could be determined with an ac-curacy of ±7%7 by comparison of the photographof the CsCl gradient to these standard photo-graphs.

Time study of the disruption of cells by glassbeads. Three devices were studied for theirability to release cell walls from S. aureus inthe presence of beads: the Braun shaker, the Nos-sal disintegrator, and the Branson 20-ke sonifier.Of these, the Braun shaker breaks the largestnumber of cells (1.5 g, dry weight) in the shortestlength of time (5 min of total shaking time overa 15- to 20-min period) without the temperatureof the suspension rising above 5 C. Consequently,most time studies of disruption were carried outwith this apparatus. The time course of disrup-tion of S. aureus cells in a Nossal disintegratorwas similar to that observed in the Braun shaker.The presence of an air space (equivalent to 20%of the total volume) and of glass beads wasessential to the disruption of cells in both ma-chines. Disruption of cells in the Branson 20-kesonifier resulted in low yields of cell walls. Here,the formation of cell walls was dependent onthe presence of glass beads, and was directlyrelated to the amounts added in the range of0.13 to 1.2 g of beads per ml of cell suspension;whole cell walls were released only transiently,followed by further breakage to particles not

A"OL. 88, 1964 11057

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

HUFF, OXLEY, AND SILV'ERMANB

1.5I 1

1.4

CsCI DENSITFIG. 2. Light-scattering patterns obse

disruption of Staphylococcus aureus cellsfor various periods of time in a Braun sh

ples were removed at intervals duiring disriplaced in CsCl for 2 to 3 days prior to .fCsCl. The volume of original material plgradient and time of removal were as follot,uliter, 0 time; (B) 2.1 ,uliters, 0.61 mix,uliters, 1.5 min; (D) 13.3 ,uliters, 2.67 mi7,uliters, 4 min. The single peak in A andpeak in E demonstrate the positions of i

(of effective growth time of 6 hr) and cel,spectively.

0 50 100 150

DRY WEIGHT (/grams)FIG. 3. Quantitative relationship between dry

weight of cells or cell walls and the intensity of light-scattering of these materials distributed in a cesiumchloride gradient. Relative area under the curve re-fers to the shaded area measured on photographs,taken under standard conditions of illutmination,when the indicated amount of cell walls (or cells) wasdistributed in a cesium chloride gradient in a wedge-shaped optical cell. A relative area value of 100equals 1 in.2 on a final enlargement 5.1 times normalsize.

I detectable in an indirectly illuminated CsCl1.3 gradient.

To study the time course of breakage, astandard suspension of S. aureus cells (37 mg,

1 dry weight, per ml) was shaken with glass beads.At intervals, the beads were allowed to settle,

rved after and samples of the supernatant fluid were takenby shaking for determination of the optical density at 660zker. Sam- mA, viable count, and CsCl gradient pattern. Aitption and representative series of CsCl gradient patternslotation in observed during disruption of staphylococcalaced in the cells in a Braun shaker is shown in Fig. 2. Thews: (A) 0.9 amounts of cell walls and unruptured cells weren; (c) 4.0 determined from these patterns by measure-the single ment of the area under the light-scatter curve

whole cells and comparison with the standard curves of1 walls, re- Fig. 3. These data are presented in Table 1. In

four of five experiments, it was observed that

E

I I 11 I0

1158 J. BACTERI OL.

I

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

DENSITY-GRAI)IENT PATTEIRNS OF S. AUREUS

after 4 to 6 min of shaking in the Braun shakerthere was no band other than the main density1.426 cell-wall band detectable, all traces of thedensity 1.380 band of unruptured cells havingdisappeared. The positions of the indicated CsCldensities throughout the cell in Fig. 2 were calcu-lated from data given bv Ifft, Voet, and Vinograd(1961). All densities were determined from suchpatterns by measurement of the distance fromthe peak to the uppermost portion of the lowermeniscus, at which point the density is approxi-nmately 1.50.The disappearance of the density 1.380 band

of unruptured cells and the decrease in theviable counts are first order. In one experiment,the cells were not completely disrupted, even

after 10 min of shaking. However, these cellsdid differ from the others used in having an

effective growth time of 5.1 hr as comparedwith 6.0 to 6.8 hr for all other experiments.

During the early stages of disruption in mostexperiments, a band of density 1.402 appeared(faintly visible in patterns B, C, and D of Fig. 2),reaching a maximum by 1 min and disappearingby 4 min of shaking. The nature of this band,which may be an intermediate stage in therelease of wall material, is unknown.A logarithmic plot of the rate of reduction

in viable counts during disruption in a Braunshaker revealed that a resistant portion of thepopulation (1 in 106 organisms) was killed athi2 the rate of the others. This observation is

similar to that of King and Alexander (1948),who observed that in a mechanical shaker 1 in106 S. aureus cells was killed at a rate J46 thatof the bulk of the organisms. In our experimentswith 50 times as many cells per ml, decrease inviability in a Braun shaker was at a rate three-fold greater than that reported by King andAlexander (1948).A comparison of the viable counts and un-

ruptured cells in Table 1 shows that after 1 minof shaking the unruptured cells are predominantlynonviable. This result is in contrast to the resultsof Cooper (1953) who, with a mechanical shaker,found a close correlation between viable countand rate of disruption of S. aureuis. In the case of

the Branson 20-ke sonifier, it was observed thatdisappearance of the unruptured cells parallelsthe loss of viable count.

Wl hole-cell and cell-wall densities during growth.

TABLE 1. Release of cell walls front Staphylococcuisaureus by disruption of the cells in a Braun

shaker with glass beadsa

Shaking Vibecut Cells Walls Opticaltime Viable count remainingb releasedc densityd

nin 1 o,C ,c No

O 100 100 0 1000.6 1.8 32 53 501.5 0.014 15 83 272.6 3 X 10-4 4.1 93 134.0 1 X 10-4 1.1 96 7.75.0 7 X 10-5 0.2 100 6.3

a Cell and cell-wall gradient patterns wereprepared as in Fig. 2, and dry weights were cal-culated by comparison with the standard patternsused to prepare Fig. 3.

1 The percentage of cells remaining was calcu-lated from the dry weights; the initial dry weightof the cells was assumed to be 100%.

c The dry weight of cell walls determined afterall traces of cells had disappeared was consideredto be 100%.

d The per cent optical density (OD) at 660 mp(100) (final OD at 66) m,u)starting 01) at 660 m,u

The densitv of S. aureus cell walls has been foundto be 1.412, with little or no change during thecourse of growth. Cell walls reisolated from aCsCl gradient showed no change in the ratio oftotal phosphate to hexosamine released onhydrol-sis. This would indicate little change inthe chemiiical nature of cell walls in a CsCl gra-dient. In contrast to the cell walls, the density ofthe whole cell changes considerably duringgrowth. Stationary and lag-phase cells have anaverage density which varies from 1.375 to1.384. Early logarithmic phase cells have anaverage density of 1.450. By the time the cellcount is 90% of the final value, this averagedensity has dropped to 1.425, and another speciesof density 1.400 appears. This species becomesdominant, and is replaced by the 1.375 to 1.384species after the viable count becomes stationary.It is of interest that no gradual change of onedensity type to another was observed. Instead,two bands of material were present during theperiod in which cells were changing from onedensity type to another. This would suggestthat individual cells rapidly- change from onedensity type to another, so that few cells of

V-oi,. 88, 1964 11.;9

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

HUFF, OXLEY, AND SILVERMAN

TABLE 2. Chemttical analysis of cell-wall soibfrac-tions obtained fromtt a sucrose gradient*

Per cent dryweight as

Fraction Totalwegta

OD6ssRegion of gradient no. dry wt dry wt

RNA Phos-phorus

mg

Precipitate... 37t 8. 6 0.51 2.72 0.42A, lower ...... 12-16 9.3 0.16 2.70 0.47B, middle ..... 17-19 17.3 0.17 2.63 0.36C, upper...... 20-26 20.9 0.22 2.72 0.29

* Material from the indicated regions of fouridentical gradients was pooled, washed twice withwater, and resuspended in 1.00 ml of water.Samples were dried and weighed, and IRNA andphosphorus were determined.

t The precipitate came off partially with thelast milliliter of sucrose, and the remainder of theprecipitate was combined with this fraction.

intermediate densitv are observed at any onetime.

Isolation and characterization of cell walls. Cellwalls were prepared by shaking in a Braunshaker with glass beads, followed by washingwith KCI and fractionation in a sucrose gradient.These cell walls were examined for size hetero-geneity in sucrose gradients, with an electronmicroscope, and for phosphorus and RNAcontent. In a typical experiment, 1 g (dry weight)of cells in 36 ml of water was shaken in a Braunshaker with glass beads for 5 min. The disruptedcell suspension was decanted from the beadsand combined with a 1 M KCl wash of the beadsto give a final volume of 30 ml. This suspensionwas centrifuged 15 min at 11,000 X g, the super-natant fluid was decanted, and the residue waswashed three additional times with 30 ml of 1M KCl, followed by centrifugation for 10 min at46,000 X g. The final pellet was resuspended in 8ml of 1 M KCl buffered with 0.05 M tris(hydroxy-methyl)aminomethane (tris) at pH 6.4 to 6.8and frozen at -15 C. At lower pH and at alower molarity of KCl, clumping of cell wallsoccurred to a greater extent during a freeze-thaw cycle. The frozen cell-wall suspension wasthawed, and was treated for 2 min in a Ray-theon 10-ke oscillator at 5 C to break up clumpswhich formed during the freezing process. Re-maining clumps were removed by centrifugationfor 10 min in a conical centrifuge tube at 700 Xg. The sul)ernatant susl)ension Nas removed, and

a 2-ml sample was placed on a sucrose gradientand centrifuged for 30 min at 3,100 rev/min asdescribed under M\Iaterials and MIethods.

Sucrose gradients of completely (lisruptedorganisms contained a single band of turbidity.Material froml this band was N-ashed free fromsucrose by centrifuging at 46,000 X g andsuspending in distilled water three separate times.Electr on micrographs rievealed the presence ofcell walls and cell-wall fragments, vith no evi-dence of cytop)lasmic contents. When placed ina CsCl gradient, this material gave a singlesyimmetrical band of density- 1.426.

Incompletely disrupted organisms gave riseto two bands of turbidity in sucrose gradients.Material from the up)l)er band appeared to beidentical to that observed in completely disruptedorganisms, and consisted of cell walls. 1\Iaterialfrom the lower band w-as washed free fromsucrose. This material was nonviable and in-distinguishable from whole cells when placed ina CsCl gradient. The amounts of remaining cells,as determined by the turbidity in a sucrosegradient, checked well with the values obtainedfrom a CsCl gradient.Although the cell-wall material was distributed

sy-mmetrically in a sucrose gradient, it was ofinterest to see whether there were physical andchemical differences between cell walls isolatedfrom lower and upper portions of the cell-wallband. A small batch of cell walls from cells withan effective growth time of 6.8 hr was shaken for7.8 min in a rliaun shaker and washed four timeswith 1 At KCl. This material was placed on fouridentical sucrose gradients. Centrifugation timewas extended to 70 min to bring the cell-wallpeak into the center of the gradient. In all gra-dients of cell walls, 10 to 20%,, of the startingmaterial forms a precipitate on the bottom ofthe sucrose gradient tube, owing to the presence ofclumps, which tend to form as cell walls stand inconcentrated suspensions. AIaterial from theprecil)itate and the A, 13, and C regions (Table 2)of each gradient was l)ooled, washed twice withwater, and suspended in 1.0 ml of water. Sampleswere dried and weighed, and RNA and totalphos)horus were determined (Table 2). It canbe seen that the phosphorus content was con-stant at about 2.72% l)hosphorus, as has beenobserved for similar preparations. The RNAcontent was rather constant at less than 0.2%,except in the case of the precip)itate.

Material from the A, ,13 and C regions of

1160 J. BACTERI OL.

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

8,)ENSITY-GIRAD)IENT PATTEIRNS OF S. AURIEUS

Table 2, when rerun in separate sucrose gra-dients, came off respectively in the A, 1B, andC regions of the second sucrose gradient. Thiswould suggest that material from the A region isof different size than that from the C region.This was confirmed by electron micrographswhich show whole cell walls in the A. region andcell-wall fragments in the C region. As yet, itis not known whether cell-wall fragments formdirectly by breakage of cells or are preceded bywhole cell walls. When whole cell walls from theA region of the sucrose gradient were treated for5 min in a Braun shaker with glass beads, nodecrease in the optical density at 660 m,u wasobserved, and there was little or no change in thepattern of sedimentation in sucrose. This wouldsuggest that the smaller fragments may ariseonly during rupture of the whole cells or shortlythereafter. Further work is being carried out, todetermine whether cell-wall fragments differchemically from whole cell walls, and to de-termine whether these fragments are related toany newly synthesized portion of cell-wall ima-terial or are merely formed in a random fashionfrom all parts of the cell wall during disruption.

DISCUSSIONB3ecause of their large size (500 m,u in diameter)

it has been difficult to characterize cell walls bystandard physical and chemical methods utsedfor lproteins and nucleic acids. The only mnethodwhich has been employed extensively withsuccess for characterization of cell walls has beenobservation of electron micrographs, as discussedby Salton (1960). Virus particles of 300 m,ulengths were separated by Steere (1963) on 1 ,agar-gel columns, but sel)aration of cell wallsand cell-wall fragments has not been reported.Electrophoresis has not been used to characterizecell walls, although Roberson and Schwab (1960)carried our electrophoresis of sonically treatedstreptococcal cell walls. I)iffusion, light-scatter,and size distribution mneasurement have notbeen used.

In considering the possible methods for charac-terizing cell walls by l)hysicochemical means,there appeared to be an absence of any study ofthe flotation patterns of cell walls in a CsClgradient. It is clear from the l)resent study thatcell walls from a single strain of S. aureus have aspecific density, regardless of the growth l)hase ofthe parent organism. Although the densities ofcell walls of different bacterial species have not

been dletermined, it seems l)robable that thesealso may be lspecific and constant, and may allowa characterization of cell-wall type.

Purification of stal)hylococcal cell walls wasaccoml)lished by Yoshida (1961), with sucrosedensity gradients. Roberson and Schwab (1960)separated streptococcal cell walls from chemicallysimilar lparticles with the use of sucrose gradients.In the l)resent study, it was possible to separatecell-wall imaterial into subfractions of differentsizes with the use of sucrose density gradients.The kinetics of formation of fragments and sourceof origin will be reported in a future publication.

In the l)resent study, it was l)ossible to deter-mine (onditions necessary for the comi)letebreakage of cells to (ell walls by observation ofCsCl flotation patterns. A further study of thesepatterns revealed a l)ossible intermediate in thedisruption of cell walls. P'attern changes d(luingautolyNsis were also observed, andl vill be reportedelsewhere.

LITERATURE CIrED

AMES, B. N., ANI) D1. T. D)UBIN. 1960. The role ofpolyamines in the neutralization of bacterio-phage deoxyribonucleic acid. J. Biol. Chenm.235:769-775.

BRtITTEN, It. J., AND R. B. ROBERTS. 1960. High-resolution density gradient sedimentationanalysis. Science 131:32-33.

COOPER, P'. 1). 1953. The study of cell rupture inStaph ylococcoxs a urec.s. J. Gen. Microbiol.9:199-206.

CHEN, P. S., T. Y. ToRIBARA, ANI) H. WA.RNER.1956. Microdeteriniilcation of phosphorus.Anal. Chem. 28:1756-1758.

(GoMooli, G1. 1942. A modificationt ()f the col(orimet-ric phosphorus determinationi for use with thephotoelectric colorimeter. J. Lab. Clin. MIed.27:955-960.

IFFT, J. B., 1). H. VOET, ANI) J. VINOCRAD. 19(61.The determination of deensity distributionsanid density gradients in binary solutiotns atequilibrium in the ultracentrifuge. J. Phys.Chem. 65:1138-1145.

KING, H. K., AND H. AlEXANDER. 1948. Themechanical destruction of bacteria. J. (Gen.Microbiol. 2:315-324.

MA.RTIN, IT. G., AND B. N. AMES. 1960. A mnethodfor determining the sedimentation behavior ofenzymes: application to protein mixtures. J.Biol. Chem. 236:1372-1379.

MIERKENSCHLAGER, M., J. SCHLOSSMI\NN, AND WV.Kuiz. 1957. Ein mechanischer Zellhomo-genisator und seine Anweridbarkeit auf biolo-gische Probleme. Biochem. Z. 329:332-340.

1161N OL . 88, 196;4

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

1162 HUFF, OXLEY, AND SILVERMAN

MESELSON, M., F. W. STAHL, AND J. VINOGRAD.1957. Equilibrium sedimentation of macro-molecules in density gradients. Proc. Nati.Acad. Sci. U.S. 43:581-588.

MICKLE, H. 1948. Tissue disintegrator. J. Roy.Microscop. Soc. 68:10-12.

NOSSAL, P. M. 1955. A mechanical cell disinte-grator. Australian J. Biol. Med. Sci. 31:583-589.

ROBERSON, B. S., AND J. H. SCHWAB. 1960. Studieson preparation of bacterial cell walls and cri-teria of homogeneity. Biochim. Biophys. Acta44:436-444.

SALTON, M. R. J. 1960. Microbial cell walls, p. 3-5.John Wiley & Sons, Inc., New York.

J. BACTERIOL.

SCHNEIDER, W. C. 1957. Determination of nucleicacids in tissues by pentose analysis, p. 680-682.In S. P. Colowick and N. 0. Kaplan [ed.],Methods in enzymology, vol. 3. AcademicPress, Inc., New York.

SHOCKMAN, G. D., J. J. KOLB, AND G. TOENNIES.1957. A high speed shaker for the disruption ofcells at low temperatures. Biochim. Biophys.Acta 24:203-204.

STEERE, R. L. 1963. Tobacco mosaic virus: purify-ing and sorting associated particles accordingto length. Science 140:1089-1090.

YOSHIDA, A., C. G. HEDEN, B. CEDERGREN, ANDL. EDEBO. 1961. A method for the preparationof undigested bacterial cell walls. J. Biochem.Microbiol. Technol. Eng. 3:151-159.

on March 19, 2021 by guest

http://jb.asm.org/

Dow

nloaded from