JOURNAL OF BACTERIOLOGY, Nov., 1965Copyright © 1965 American Society for Microbiology

Vol. 90, No. 5Printed in U.S.A.

Sedimentation Counting and Morphologyof MycoplasmaHAROLD W. CLARK

Arthritis Research Unit, Department of Mledicine, The George Washington University, Washington, D.C.

Received for publication 14 June 1965

ABSTRACT

CLARK, HAROLD W. (The George Washington University School of Medicine, Wash-ington, D.C.). Sedimentation counting and morphology of Mycoplasma. J. Bacteriol.90:1373-1386. 1965.-The sedimentation technique for counting viral particles was ap-

plied to the quantitation and morphological identification of Mycoplasma in broth cul-tures. Mlycoplasma, apparently in their native form, firmly adhered to the surface,wheni sedimented on glass cover slips or onto electron microscope grids. The sedimentedcover slip preparations stained with crystal violet could be readily counted in the lightmicroscope. The cultures sedimented onto electron microscope grids were readilycounted at low magnification and provided excellent preparations for morphologicalexamination at higher magnifications. It was found that air-dried Mycoplasma particleswere enlarged considerably because of excessive flattening. Fixation of sedimentedMycoplasma particles in diluted OS04 prior to air drying yielded a more realistic mor-

phology, with various sizes and shapes in the stages of the growth cycle exhibited. Anew technique of differentially staining Mycoplasma colonies on agar plates was de-veloped to facilitate the quantitation of viable colony-forming units for comparisonwith total counts. The use of plastic or Parafilm gaskets for dry mounting was de-veloped to facilit,ate the handling and examination of the stained cover slippreparations. The results of this investigation indicated that the growth cycle of some

Mycoplasma species includes a stage of hexadic fission with the cleavage of minimalreproductive units (less than 100 mu) containing a limited deoxyribonucleic acid geneticcoding molecule (approximately 4 X 106).

A need exists for the standardization of Myco-plasma cultures to provide comparison of resultsand to improve interpretation of analytical data.It has become apparent in evaluating the chem-ical and immunological differences within a spe-cies and among different species that a moredirect and aceurate method of quantitating My-coplasma is necessary. The methods currentlyused for the quantitation of Mycoplasma growth,i.e., colony counts, turbidimetric analysis, endpoint dilutions, dry weight, total protein, or totalnitrogen, are indirect means of estimating thenumber or amount of Mycoplasma present. Theanalytical results obtained by these methods areinfluenced by a variety of factors, including via-bility, growth conditions and age of culture,species variations, residual and adsorbed mediaimpurities, range of particle size, and aggregation.

Before any viral particle-counting techniquecan be applied to Mycoplasma broth cultures, itis first necessary to establish standard conditionsand a countable unit. The hot-water fixationtechnique (Clark et al., 1961) is a rapid and ac-curate method for identifying Mycoplasma colo-nies. However, no satisfactory method exists forthe identification of the naturally occurring

Mlycoplasma particles because of their pliablenature and resultant heterogenic size and shape.Consequently, a technique previously establishedfor the enumeration and quantitation of virusparticles was investigated. The sedimentationcounting method developed by Sharp (1958) wasmodified and applied to the study of Mycoplasmacultures. When properly controlled, this methodyields preparations suitable for both light andelectron microscopy. The technique, originallydeveloped to readily quantitate Mycoplasma inliquid culture, is also an acceptable method forthe identification of Mycoplasma morphology.

In addition, correlation of Mycoplasma particlecount with viable colony-forming units (CFU)was greatly simplified by the developmet of arapid differential colony-staining technique. Sucha technique facilitated the determination of CFUyield in a series of filtration studies made to com-pare the number, size, and shape of Mycoplasmaparticles with those obtained by sedimentation.

MATERIALS AND METHODS

Mycoplasma strains. M. hominis, type 1, strainJJ and M. hominis, type 2, strain CH used in thisstudy were isolated from the human genital tract.

1373

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

The avian strain 5969, M. gallisepticum, was ob-tained from M. E. Tourtellotte at the Universityof Connecticut. Other types of Mycoplasma wereexamined but are not included in this report.

Cultures. Mycoplasma strains were cultured in afreshly prepared broth of pancreatin-digested(PD) beef heart enriched with 10% heated bovineserum. The serum and broth were mixed just priorto use and passed through a large 220-mgs Milli-pore filter at 10 psi to minimize media sediment inthe Mycoplasma preparations. The inocula,unless otherwise designated, were from 24-hrbroth cultures and were added in 1% volumes.

Particle sedimentation. A Sorvall refrigeratedcentrifuge, RC-2, with a type SU particle-count-ing rotor was used (Ivan Sorvall, Inc., Norwalk,Conn.). The method of procedure as described bySharp (1958) was used with few modificationsnoted below.

Preparations for the electron microscope wereobtained by sedimentation of diluted Mycoplasmacultures onto nitrocellulose-covered stainless-steel grids at 27,000 X g for 20 min. The gridswere held in place by a magnet mounted in Luciteat the base of the counting cell chamber (Fig. 1).The counting cells were filled with 1.0 ml of suit-ably diluted Mycoplasma cultures or suspensions.Supernatant fluids from the centrifuged cultureswere drained off, and the cell sections containingthe grid were immersed in a 0.4% OS04 solutionof 0.8% NaCl and 5% acetic acid for 2 min. Thecells were then rinsed with distilled water andallowed to air dry before the grids were removed.In the 0804 control preparation, the cell wasrinsed with distilled water only and was fixed byair-drying. Shadow casting of specimens was donewith Pallatium-palladium alloy at approximately30°. When standard latex particles were sedi-

r-

*~ b6

FIG. 1. Center section of the particle-counting cellwith the magnet and electron microscope grid (right)and the glass cover slip showing the mounting tapeof Parafilm gaskets and the final stained andmounted cover slips.

mented along with Mycoplasma for verificationof size and number, a 0.5% gelatin-coated nitro-cellulose film was used as recommended by Sharp(1958).Preparation for light microscopy. Cover slips

(8 X 10 mm) were cut from optical grade no. 1glass and were acid-cleaned. A slip was mountedon the base (cm2) of each counting cell with atrace amount of petrolatum. The counting cellswere filled with 1.25 ml of a suitable dilution ofMycoplasma broth cultures (10-2) and centrifugedat 27,000 X g for 20 min. The cells were drainedand then immersed immediately in a 0.4% 0804solution of NaCl and acetic acid for 2 min. Thecells were drained again, and the cover slips weregently rinsed with water and were allowed todrain dry. The 0504 control preparations withoutfixation were also rinsed with water. After airdrying, the cover slips were transferred to a micro-scope slide and were held in place with the residualpetrolatum. One drop of 1.0% aqueous crystalviolet was placed on each cover slip, and the slidewas placed in a moist chamber for 1 hr. The stainedcover slips were rinsed with distilled water andallowed to air dry.

Plastic mounted cover slips. Tapes of plasticmounting gaskets were prepared by cutting 1-cmwide strips from clear four-gauge plastic (Flex-O-glass, Warp Bros., Chicago, Ill.). These strips weremounted on centimeter graph paper and perfor-ated with 6-mm holes at 1-cm intervals. Astrip of four gaskets cut from the (plastic) mount-ing tape was centered on a clean microscopeslide, and four of the stained cover slips wereplaced face down (inverted) over the center ofeach gasket hole. The microscope slide wasplaced on a low-temperature hot plate (approxi-mately 100 C) for a few seconds until the plasticstarted to melt. The slide was removed, and thecover slips were gently pressed in place so as toexclude trapped air bubbles. After cleaning theresidual petrolatum from the backs of the coverslips with soap and water, the slide was readyfor microscopic examination and particle count-ing. Similar mounting gaskets were also preparedfrom Parafilm. Because of their organic solventsolubility, it is recommended that the plastic-or Parafilm-mounted cover slips be cleaned withsoap and water. Figure 1 shows the center sectionof the counting cell which holds the cover slip orthe magnet and grid. A portion of the Parafilm-mounting tape gasket and a final slide preparationare also demonstrated. Larger cover slip prepara-tions of hot water-fixed Mycoplasma colonieswere also permanently mounted with larger plas-tic gaskets (22 mm square) onto microscope slides.I Colony counts. Broth cultures or similar sus-pensions of washed Mycoplasma were usuallydiluted 10-2 or 10-8 for particle sedimentationcounting. Higher dilutions, 10-4, 106, and 108,were made for colony counts, and 0.05 ml of eachdilution was spread over the surface of 60-mmagar plates. The plates were sealed with Parafilmand incubated at 37 C for 2 to 3 days. The colonies

1374 CLARK

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MYCOPLASMA COUNTING AND MORPHOLOGY

were differentially stained by flooding the surfaceof the plate with approximately 2 ml of a filtered0.1% Giemsa stain in chloroform (w/v) and gentlyrotating for 60 sec. The staining time varied some-what depending on the agar hardness. The bestresults were obtained after evaporation of ex-cessive moisture and condensation from the sealedplates. Plastic petri plates can be used becausethe chloroform solvent exposure is brief. Thestained colonies were easily seen and rapidlycounted with a stereomicroscope or hand lenswith the whole plate in view. The stained colonieson a heavily inoculated plate (Fig. 2) were readilyvisible and were counted by placing it on a whitesheet of paper containing centimeter square grid,thus eliminating the need for ruled plates. Thestained M. hominis colonies selected for illustra-tioIn in Fig. 2 required no magnification; theywere exceptionally large because of culture anddilution phenomena. The smaller stained coloniesreadily counted at low magnification were alsoexamined at higher magnification to observecolony morphology and to insure accuracy ofcount.

RESULTSOne of the most striking observations made in

this investigation was the extremely pliablenature and size variation of the Mycoplasmaparticles. Although this morphological propertyhas been described by Borrell et al. (1910) andlater by Turner (1935) and many other investi-gators, it undoubtedly has also been the sourceof repeated misinterpretation. With only a mem-brane to confine their cellular contents, it would

FIG. 2. Mycoplasma hominis, type 2, coloniesdifferentially stained intact with Giemsa for easierexamination and counting.

seem logical to expect multiple sizes and shapesdependent upon their physical-chemical environ-mental exposure.A look at the basic unit of Mycoplasma identi-

fication, the colony, that has been physicallydisrupted just prior to being fixed on a glass slidewith hot water, will reveal several subunits. Thedark-field photomicrograph of the M. hominisstrain shown in Fig. 3 illustrates the variablelarge body composition of the colony as oftendescribed by Dienes (1945). It further illustratesthat the membrane envelope holding the largebodies together is also pliable enough to stretchinto filaments that hold the smallest particulatecomponents in chains. Although the cotton fibertechnique, used by Orskov (1927) and Freundt(1958) to demonstrate filaments in colonies, wasnot used in this preparation, it does seem possiblethat a more filamentous preparation could havebeen obtained under the right conditions. An-other human Mycoplasma colony shown in Fig.3b illustrates the chains of coccoidal particlesbranching out from the colony edge. The prob-lem remains: What particle entity can be countedand what is the self-reproducing unit? The small-est separate particles seen in the colony chainsare less than 300 m,u in diameter and cannot beused to measure Mycoplasma growth, especiallyin broth cultures.An investigation of the effect of various dilu-

ents, rinses, and fixatives on Mycoplasma sus-pensions was made to determine optimal condi-tions for obtaining maximal yield, uniform-sizedparticles, and nonaggregated preparations. Theeffect of decreased osmolarity on Mycoplasmahas been variously interpreted by different inves-tigators to cause complete lysis and bactericidalor no apparent effect. Marked differences wereobserved in the centrifuged pellets of M. hoministype 2 (CH) washed with distilled water or hy-potonic NaCl (less than 0.9%) and the pelletswashed with hypertonic NaCl (2 to 8%). Thehypotonic solution produces a diffuse and trans-lucent pellet, whereas the hypertonic solutionproduces a smaller and white opaque pellet.However, no significant difference could be de-tected in the amount of protein released from an18-hr culture washed three times with distilledwater or with NaCl solutions of increasing con-centrations (0.5 to 8%) or a fourth wash of10% ethyl alcohol. The similar analysis of thewater and NaCl washings was confirmed by thealmost identical amounts of protein found in thewashed Mycoplasma pellets obtained by centrifu-gation at 27,000 X g for 20 min. To avoid mis-interpretation, it must be emphasized that theseresults obtained with an 18-hr culture do not

VOL. 90, 1965 1375

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

rt-- . 4.,

lt ,

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f- .w .:'

r,,S~~~~ ~~~ ~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~k;,:..

, e ..... u -.' ..... ?: '." . ......~~~~~~~~~~~~~~~~~~~~~~~~~~~~u~~~~~~~~~~~~~f

*3?j '.:.:ifzfsS3k?rSfgos-; _'~~~~~~~~~~~2

1A ARSK S .. *g... .'' ,S '';":..

FIG. 3. Dark-field photomicrographs of hot water-fixed Mycoplasma hominis, type 1, colonies, X 450.a) Variable large body and particulate chainlike filaments disrupted from colony. b) Chains of particlesbranching out from edge of colony.

1376 CLARK

r* .ir ow

*.1-.1;.. W.

0-

I

4i,: I.". r,

k,::!. ':

-,f

Jh

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MYCOPLASMA COUNTING AND MORtPHOLOGY

necessarily apply to older cultures and otherstrains. Earlier experiments on chemical compo-sition of MI1ycoplasma revealed marked instabilityof older cultures and aged preparations with therelease of soluble components. Unless properlycontrolled, this variable would introduce consid-erable influence in both qualitative and quantita-tive analysis.

These results on osmolarity were confirmed bymicroscopic analysis of particles sedimented fromMIycoplasma cultures diluted in various NaClconcentrations. There was a definite tendency forthe MIycoplasma particles from the high salt con-centrations to be smaller and more aggregatedand usually to give slightly higher counts. How-ever, the particles sedimented from water wereless aggregated and slightly larger, which madethese preparations (stained) easier to count.Some of these results on osmolarity effect areillustrated in Fig. 4a to 4d, with photomicro-graphs of the sedimented Mycoplasma particlesdiluted in various solutions and washed withwater or 10% ethyl alcohol.To simplify the counting procedures, photo-

magnification of particles sedimented on coverslips was kept constant at X 1,700. It was thuspossible to use a 1.7-cm square (equivalent to10 ,u) overlay of the photos (Fig. 4d) to represent10-6 of the centimeter square cell base uponwhich the 1.25 ml of Mycoplasma suspension wassedimented. The number of particles (n) countedper square, times 106, times the culture dilution(102), and divided by 1.25-ml volume gives afairly accurate estimate of Mlycoplasma particlesper milliliter of culture. An area of averageparticle density was selected by use of lowermagnification, and the average number of par-ticles counted in five or more squares was used.Counts were also made direct from the projectionof the microscope field on a ground glass platewhen immediate analysis was required.The appropriate dilution and condition of

Mlycoplasma suspensions for electron microscopeanalysis was pretested on cover slips. The elec-tron microscope grids with sedimented Myco-plasma were pre-examined in the light microscopeand photographed to assure a suitable specimenand to confirm the final results. Other than thevariable size, shape, and particle distribution, themost striking and characteristic feature noted inthe unfixed air-dried specimens was the markedflattening and stretching of particles. Similarresults have been shown and variously inter-preted by other investigators in earlier electronmicroscope photographs.

Actual comparison of pre- and postcentrifuga-tion specimens indicated there was no apparent

effect of the 27,000 X g sedimentation on theflatness because of size similarities. The flatteningof Mlycoplasma particles is primarily caused bysurface drying, because fixation of Mlycoplasmawith OS04 prior to drying yielded more sphericaland thicker (longer shadow) particles. The effectof OS04 fixation on MIycoplasma prior to dryingis shown in the cover slip preparations (Fig. 5aand 5b) and the electron microscope photomicro-graphs (Fig. 6a and 6b). An additional advantageof OS04 fixation, other than producing a firmmembrane, seems to be a differential cytologicalreaction resulting in the demonstration of elec-tron-dense inclusion particles in the unshad-owed preparations. The Os04-fixed preparationsseemed to have maintained their consistent mor-phological growth-stage characteristics and werenot reduced to flattened globs by air-dryingnor altered to various pleomorphic forms.

Analysis of the number of colony-formingunits (CFU) in a 24-hr broth culture of MI. homi-nis, type 2, indicated a 66% viability in the totalparticle count. The per cent viability was de-pendent upon the size and viability of the inocu-lum. The accuracy of both CFU and total par-ticle counts improved with minimal aggregation.Estimation of aggregation and CFU size bypassage through a series of Millipore filters (450,300, 220, and 100 m,ut) indicated considerablevariation. The CFU recovered from passage of a48-hr 31. hominis culture (CH) through a gradedseries of 25-mm diameter Millipore filters underpressures of 2 to 15 psi are shown in Table 1.These results showed that 99% of the viableCFU particles in the 48-hr culture were greaterthan 300 m,u and that one in 10 million wassmaller than 100 m,u. Particle counts and exam-ination of the filtrates below 300 mu were difficultbecause of the low yields. No evidence of fila-ments or extruded forms was observed.

Standardization of Mycoplasma cultures interms of CFU counts, particle counts, and washedsedimented protein or dry weight provide morespecific basis for describing and relating meta-bolic activity, chemical composition, and anti-genic activity. Table 2 illustrates the quantitativeanalysis of Mlycoplasma growth in a 24-hr brothculture by means of colony counts (light andelectron microscopy) and total protein. Theaverage protein mass of a 24-hr 31. hominisviable lparticle is less than 11 X 10-9 ,ig, and,correcting for nonviable particles, the averageprotein mass is 7 X 10-9 ,ug. This range of valuesis in approximate agreement with the dry weightvalue of 6.4 X 10-9 ,ug per 31. gallisepticum(5969) particle calculated by Morowitz et al.(1962). One could approximate that the few

N'OL. 90, 1965 1377

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

1378 CLARK J. BACTERIOL.

.*~*'* Wh'. *. .:.,i

Ti.s. ~~~~~~~~~~~~~~~~~~~~~~~

ii*

* i:.a:B:X u i

o.:d.., <^^.,.,'; -

FIG. 4. Diluted Mycoplasma broth cultures sedimented on cover slips, air-dried, and stained with crystalviolet. (a) M. hominis, type 2, diluted with 0.9% NaCI and rinsed with 10% ethyl alcohol. (b) M. hominis,type 2, diluted with 4% dextrose and rinsed with water. (c) M. gallisepticum (5969) diluted in water andrinsed in 10% ethyl alcohol. (d) M. gallisepticum (5969) diluted in water and rinsed in water with the equiva-lent 10 ju square overlay to facilitate counting. Original magnification, X 1,700.

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

VOL. 90, 1965 MYCOPLASMA COUNTING AND MORPHOLOGY 1379

#'~ ~ ~~, ,A.v **?* ? ^~~~~* 4 *

* 4**

se~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Mzt**.s

:4 .: :1 :u :* :.:

Am'

4:i.

.~~Ny*: s * .,

Eli"g., + .i:

t~~~W:*

.I...G.4-C,o

FIG. 4-Con,tinued

CFU passing the 100-m,u filter are about r/4 or sity. The average deoxyribonucleic acid (DNA)volume/64 with a mass of 1.2 X 10-10 jig, as- content of these preparations was about 5% bysuming the small Mycoplasma and large (400 the diphenylamine test of hot trichloroaceticm,u) were both spherical and had the same den- acid extracts. According to this, the amount of

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

CLARK J. BACTERIOL.

FIG. 5. Mycoplasma hominis, type 2, broth cultures sedimented on cover slips. (a) Air-dried prior tostaining shown with equivalent 10 ,U square overlay. (b) Fixed with OS04 prior to air-drying to yield a greaternumber of irregular shaped and sized particles (same suspension used in Fig. 8b at higher magnification).

1380

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MYCOPLASMA COUNTING AND MORPHOLOGY

FIG. 6. Mycoplasma hominis, type 2, diluted broth cultures sedimented on electron microscope grids andshadowed, X 2900 magnification. (a) Air-dried preparation producing large flattened particles (compareFig. Sa cover slip). (b) Preparation fixed with OsO4 yields spherical particles of more natural size. Un-washed medium fixed in the background does not affect particle counting but does affect chemical and immuno-logical analysis.

VOL. '30, 1965 1381

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

TABLE 1. Recovery of Mycoplasma hominis, type 2,CFU through various-sized filters

Filtrate tested Dilution CFU/ Recoveryplate* fraction

Whole culture 10-6 156 1300 m,u 10-4 104 10-2220 m,u 10-1 10 10-6100 mIA F St 8 10-i

* Average colony count from 0.05 ml of filtratedilution.

t Full strength.

TABLE 2. Quantitative analysis oJ Mycoplasmain broth culture

Analysis Total count Protein*/particle/ml

jig

CFU.26 X 108 11 X 10-'Electron microscope[-grid.23 X 108 12 X 10-9tCover slip. 40 X 108 7 X 10-9Estimated minimalCFU mass(volume/64) 1 X 10-"0

* Protein analysis of sedimented and washedMycoplasma culture by use of a modified Folin-Ciocalteau method (28 /ig/ml).

t Not applicable-low yield in this test condi-tion.

genetic-coding DNA in the smallest reproductiveunits would be approximately 4 X 10i molecularweight per CFU, and suggests a limited syntheticcapability.

Morphology. The accuracy of the average massper Mycoplasma particle calculation is greatlylimited, because of the wide range in size andshape during stages of development and growth.The uniform size and easily countable particlesin a 24-hr culture of Ml. hominis, type 2 (CH), isshown in Fig. 6b along with the fixed but easilydifferentiated media fragments. Examination ofother cultures or strains having different growthrates often yielded preparations exhibiting thevarious stages in development as shown by the24-hr 3ll. gallisepticum (5969) culture in Fig. 7aand 7b. It appears as though a single coccoidalparticle develops from 100 m,u into a large body,500 m,u, whereupon it concentrates at the periph-ery and segments into six minimal reproducingunits that usually outgrow the restricting mem-brane. The number of segments seems to be de-pendent on the strain and culture conditions.

Consequently, in a fluid state, it would be possi-ble to obtain distortions of the growth stages re-sulting in an endless number of shapes. Genera-tion time studies are further complicated by thecleavage failure after segmentation (Fig. 7a).These aggregates of segmented particles undoubt-edly form the basis of colony formation on a sta-tionary agar surface.The results of this study are similar to some of

the original morphological studies on 71I. mycoidesby Borrel et al. (1910) and to the dark-field prepa-rations of contagious bovine pleuropneumonia ob-tained by Turner (1935), and agree in generalwith some of the ideas on Mycoplasma growth andmorphology presented by Klieneberger-Nobel(1962). Most of the pleomorphic sizes and shapesof MIycoplasma particles (rods, filaments, elip-soids, etc.) observed are apparently the result ofthe artificial chemical and physical manipulationsemployed during preparation of the specimen(Liebermeister, 1960; Weibul and Lundin, 1962).The flat, irregular disc-shaped MIycoplasma fre-quently observed in solid media growth were alsoobserved in liquid media and were found to bethe result of surface drying (Fig. 6a).The size of colony growth on solid media is

primarily dependent on such conditions as agarfluidity or hardness, chemical composition, andthe inoculum concentration, which would influ-ence the disruption and distribution of the my-coplasma particles. Although the absolute mini-mal reproductive unit size is difficult to determine,the few CFU that pass through the 100-m, Milli-pore filters are certainly indicative of the minimalsize range of particles in the M. hominis cultures.Chains and clusters of particles approximately 100m,u can be seen in Fig. 7 and 8, evidently just priorto their cleavage from the other five segments.The shadowed electron photomicrographs of theavian and human strains (Fig. 8a and 8b) illus-trate their similar stages of growth as well astheir fragility with the resultant pleomorphicforms. With suitable precautions, the results ofthis technique seem to provide a new dimensionin the visualization of Mycoplasma physicalstructure during stages of development.

DISCUSSION

The results of this investigation indicate thatthe modified sedimentation technique is a suitablemethod for standardizing and identifying M1Iyco-plasma in broth cultures. The development andapplication of this method was based on the recog-nition and enumeration of the individual Myco-plasma particles. In previous studies, Clark et al.

1382 CLARK

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MYCOPLASMA COUNTING AND MORPHOLOGY

FIG. 7. Mycoplasma gallisepticum (5969) broth culture sedimented on electron microscope grids, fixedwith OS04, and shadowed as described. (a) Variable sizes and shapes in the growth cycle stages. (b) Anotherpreparation at X 6,000 magnification showing the stages of growth and the segmented particles.

VOL. 90, 1965 1383

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

FIG. 8. Diluted Mycoplasma broth cultures sedimented on electron microscope grids, fixed with OS04,and shadowed. (a) M. gallisepticum (6969) 24-hr culture showing three basic stages of growth. (b) M. homi-nis, type 2, 72-hr culture showing similar stages of growth along with the distorted shapes evidently producedin preparation.

1384 CLARK

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

MYCOPLASMA COUNTING AND MORPHOLOGY

(1961) developed a new technique for the quali-tative transfer and identification of Mycoplasmacolonies growing on agar. The development of arapid and simplified method for analysis of theindividual Mycoplasma particle now makes itpossible for other laboratories to compare andconfirm Mycoplasma strain differences and simi-larities.

Although only 99.9% of the viable Mycoplasmaparticles are sedimented in 20 min at 27,000 X g,it is sufficiently complete within the limits of thefixation and counting procedures to be acceptable.Counting Mycoplasma particles by the sedimenta-tion technique has the advantage of providing asuitable specimen for both morphological andquantitative analysis by using unaltered spec-imens and direct procedures. Under the con-ditions specified, the yield of sedimented Myco-plasma CFU is greater than 99.9%, which isas good or better than can be obtained from manyphysical and chemical comparative analyses.These results agree with the graded filtrationstudies which indicate that all of the 220 m,u andlarger-sized CFU particles are sedimented, asare only a few of the smaller, minimal-sized CFU.Another possible source of error would come fromparticles with low density and the resultant flo-tation. Such might be the case if the missing"filamentous" forms were fragments of the lipo-protein membrane. However, the insoluble mem-brane fraction has been found to sediment readilyunder these conditions. Because of the manysources of variables and possible experimentalerrors, this direct method of quantitating Myco-plasma is undoubtedly less accurate than the in-direct methods. The extent of Mycoplasma ag-gregation in broth cultures and its influence onCFU and particle counts can be estimated fromthe electron photomicrographs. The apparent in-fluence of aggregation could be attributed to thecolonylike growth of segmented particles withfailure to undergo cleavage (Fig. 7a). Brief ex-posure of some cultures to sonic oscillation resultsin the release of additional CFU, whereas briefexposures in distilled water causes marked loss ofviability with no appreciable change in generalmorphology.The spray technique of particle counting

(Backus and Williams, 1950) would probablyyield more accurate quantitative analysis. How-ever, this method was not attempted because itwould have the disadvantages of being more com-plicated and yielding morphologically alteredpreparations by the dry contact fixation.The Mycoplasma lipoprotein membrane has

stretched our imaginations with about every con-

ceivable pleomorphic form. It now seems advan-tageous to gently and rapidly stop the stretch andfix a uniformly distributed preparation with OS04.The primary advantage of the sedimentationtechnique is based on the Mycoplasma character-istic of strong adhesion to various surfaces whilestill in its original form and prior to completefixation. Human, avian, and rat strains of my-coplasma have been examined by this method.The surprising and yet encouraging results ob-tained indicate that these strains are morpho-logically similar to the first identified strain, 11f.mycoides, characterized by Borrel et al. (1910)and Turner (1935).

If filaments are a characteristic morphologicalstage in growth development of llIycoplasma, asreported by some investigators (Freundt, 1958),they apparently are not produced in the culturemedia used in this laboratory. A few filamentshave been observed occasionally in duplicatespecimens; however, they apparently were arti-ficial (Fig. 3a). On one occasion, filamentousforms were found contaminating the distilledwater reservoir. As yet, no naturally occurringfilaments have been observed in the sedimentedcultures fixed with OS04, nor in the Mycoplasmacolonies fixed with hot water after adhesion on aglass slide. Consequently, it is difficult to accountfor an intermediate or terminal filamentous stageof endomycelial development with the definitestages of coccoidal growth observed in the brothcultures.The conclusions drawn from this study essen-

tially confirm the basic stages of growth reportedby Klieneberger-Nobel (1962). In addition, ourstudies indicate that this unique biological growthsystem is the orderly development of a constantnumber of segmented particles, primarily six, inthe final stages of growth. It is anticipated thatfuture investigations will indicate whether thesegmentation number is a species characteristic(M. laidlawii), dependent upon culture condi-tions, or is due to the geometric physical resist-ance of the membrane. Any explanation of theconditions directing Mycoplasma reproductionshould also include possible mechanisms of thehexadic fission with the production of minimalreproductive units of 1.2 X 10-10 ,ug mass and asmall genetic coding DNA molecule (approxi-mately 4 X 106).

This study again raises the old questionwhether Mycoplasmatales has filamentous or coc-coidal morphology. If their fundamental morph-ology is coccoidal, as this study indicates, the ap-propriatness of the generic term Mycoplasmatalesis certainly challenged.

VOL. 90, 1965 1385'

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from

J. BACTERIOL.

ACKNOWLEDGMENTSThe author wishes to express his thanks to J. S.

Bailey for providing the Mycoplasma strains andto R. C. Fowler and D. W. Coble for their assist-ance in obtaining the electron micrographs.

This investigation was supported by PublicHealth Service research grant AI-04071 from theNational Institute of Allergy and InfectiousDiseases.

LITERATURE CITEDBACKUS, R. C., AND R. C. WILLIAMS. 1950. Use of

spraying methods and volatile suspending me-dia in the preparation of specimens for electronmicroscopy. J. Appl. Phys. 21:11-15.

BORREL, A., E. DUJARDIN-BEAUMETZ, JEANTET,AND JOUAN. 1910. Le microbe de 1 peripneu-monie. Ann. Inst. Pasteur 24:168-179.

CLARK, H. W., R. C. FOWLER, AND T. McP.BROWN. 1961. Preparation of pleuropneumonia-like organisms for microscopic study. J. Bac-teriol. 81:500-502.

DIENES, L. 1945. Morphology and nature of thepleuropneumonia group of organisms. J. Bac-teriol. 50:441-458.

FREUNDT, E. A. 1958. The Mycoplasmataceae.Munksgaard, Copenhagen.

KLIENEBERGER-NOBEL, E. 1962. Pleuropneumo-nia-like organisms (PPLO) Mycoplasmata-ceae. Academic Press, Inc., New York.

LIEBERMEISTER, K. 1960. Morphology of the PPLOand L-forms of Proteus. Ann. N.Y. Acad. Sci.79:326-343.

MOROWITZ, H. J., M. E. TOURTELLOTTE, W. R.GUILD, E. CASTRO, C. WOESE, AND R. C. CLE-VERDON. 1962. The chemical composition andsubmicroscopic morphology of Mycoplasmagallisepticum, Avian PPLO 5969. J. Mol. Biol.4:93-103.

ORSKOV, J. 1927. etude sur la morphologie duvirus p6ripneumonique. Ann. Inst. Pasteur41:473-482.

SHARP, D. G. 1958. Sedimentation counting of par-ticles via electron microscopy. Intern. Kong.Elektronenmikroskopie, 4, Berlin, Verhandl., p.542-548.

TURNER, A. W. 1935. A study on the morphologyand life cycles of the organism of pleuropneu-monia contagiosa boum (Borrelomyces peri-pneumoniae nov. gen.) by observations in theliving state under dark-ground illumination.J. Pathol. Bacteriol. 41:1-32.

WEIBULL, C., AND B. LUNDIN. 1962. Size and shapeof pleuropneumonia-like organisms grown inliquid media. J. Bacteriol. 84:513-519.

1386 CLARK

on March 22, 2021 by guest

http://jb.asm.org/

Dow

nloaded from