Microcapsules on Streptococcus mutans SerotypesBy Electron Microscopy

ELLA M. GRENIER,* R. H. GRAY, W. J. LOESCHE, and \V. C. EVELAND

The Dental Research Institute, School of Dentistry and Departments of Epidemiology andEnvironmental and Industrial Health, School of Public Health, University of Michigan,Ann Arbor, Michigan 48109, USA

Extracellular microcapsules have been demon-strated on cells of most serotypcs of Strepto-coccus mutans by electron microscopy, usingbacterial strains of the various serotypes andperoxidase labeled or unlabeled immuneserum. A correlation was noted between theamount of capsular substance on the strains ofS mutans examined and degree of antigenicityas expressed by the indirect fluorescent anti-body (FA) title. A serotype d strain was shownto lose both antigenicity as determined by theFA reaction and capsular material as seen byelectron microscopy with repeated in vitropassage. JVhen 10% unheated rabbit serumwas added to the medium, antigenicity andcapsular material were restored.

Antisera produced in rabbits to serotypes ofStreptococcus mutans gave various titers byfluorescent antibody (PA) techniques,1-3 sug-gesting that there are differences in the de-gree of antigenicity of the serotypes. Antiseradirected against the serotypes of S mutans maycross react with S sanguis and S salivarius. Thespecificity of the FA reactions can be increasedby prior absorption of the sera with the crossreacting antigens.4 However, with some S mu-tans antisera, the working titer may be so lowthat cross absorptions will reduce the titer be-yond a useful range. An eriochrome blackcounterstain appears to increase the specificityof the serotypes without the need for cross ab-sorptions.2 Both of these procedures, which en-hance specificity, do not increase the titer, sothat the amount of useful sera is quickly con-sumed. Higher titers may be achieved by im-

Received for publication May 3, 1976.Accepted for publication October 12, 1976.*For reprints: Dr. Ella M. Grenier, Dental Res.

Inst. Sch. of Dent., Sch. of Pub. Health, Univ. of Mich.,Ann Arbor, Mi. 48104.

166

proving the conditions for isolating specificantibody5'6 or by increasing the antigenicity ofthe S mutans vaccine. In the present investi-gation, antigenicity was found to be directlyrelated to the amount of extracellular capsularmaterial on the cells.

Materials and MethodsSTRAINS. The strains of S mutans rep-

resenting the five serotypes' were obtainedfrom Dr. L. A. Thomson (NIDR, Bethesda,Md). These included E49 with a antigen;7BHT with b antigen;8 Fa-1 with b antigen;9GS5 with c antigen;10 Ingbritt with c antigen;"OIHI with d antigen;12 6715 with d antigen;13AHIT with d antigen;8 LM7 with e antigen;14and B2 with e antigen.'5 Strain AHT was orig-inally described as having a antigen.1 However,our strain showed a strong FA reaction withknown anti d sera2 and behaved biochemicallyas a serotype d strain.'6 This strain will bedesignated as AHT-d. Subsequently an AHTstrain was obtained which had been shown by% guanine:cytosine ratio to belong to the ge-netic grouping which contained serotype acells,17 and which did not react with group dantisera.

PREPARATION OF VACCINES, ANTISERA, FACONJUGATES AND STAINING.-The preparationof vaccines and antisera as well as the fluore-scein conjugation and direct FA staining usingeriochrome black as a counterstain were thesame as that described by Grenier et al.2Smears of antigen suspensions (vaccines) weremade on glass slides, air dried and heat fixedfor the indirect FA titrations. Serial dilutionsof unconjugated antisera were added and theslides were incubated in a moist chamber for30 minutes at room temperature. The slideswere washed in two changes of phosphate buf-fered saline (PBS) (pH 7.6), rinsed in dis-

MICROCAPSULES ON MUTANS STEROTYPES

tilled water and air dried. Fluorescein conju-gated goat anti-rabbit globulin* (undiluted)was added, incubated and washed as above.Finally the slides were mounted in bufferedglycerine (one part PBS and nine parts glycer-ine with pH adjusted to 7.6) and read in aZeiss GFL microscope. Appropriate controltests were made using smears of the antigensuspensions, normal rabbit serum (pooled)and the conjugated goat anti-rabbit globulin.These control tests were all negative.

Fluorescein/protein ratios were obtainedfor all of the conjugated antisera,'8 using aGilford spectrophotometer, Model 240. Allwere within the acceptable range of 2 to 10ptg/mg.19

AGGLUTINATION TITRATIONs.-Twenty-fourhour cultures of the various strains of S mutanswere grown in medium containing half strengthBrain Heart infusiont and half strength Spiro-late Brotht to which 0.5% yeast extract wasadded (BHS medium). The cultures were cen-trifuged 10 minutes at 3,000 X g; the super-natant was poured off and the pellet re-suspended and washed once with physiologicsaline. The cultures were finally diluted to theturbidity of a MacFarland No. 5 standard(1.5X108 organisms per millimeter). The sus-pensions were mixed well and used as the slideagglutinating antigens.

Small aliquots of antisera to the S mutansstrains were heated at 56 C each day beforeuse. Doubling dilutions of the antisera weremade in Kahn tubes, starting with 0.2 ml salineand 0.2 ml serum. The antiserum dilutions,0.05 ml each, were pipetted into circled areason the slides. Saline (no antiserum) was pi-petted into one circle to serve as an antigencontrol. Then, 0.25-ml antigen suspension wasadded to each circle of antiserum and to theantigen control. Slides were rotated four min-utes at 160 RPM on a platform shaker. Testswere read immediately in a light microscopeat 100 X magnification.

PEROXIDASE CONJ UGATION.-Sampies ofantisera to strains AHT-d, Fa-1, GS5, LM7, B2as well as goat anti-rabbit serum were conju-gated with peroxidase. The globulin fractionwas precipitated from the antisera in 33% satu-rated ammonium sulfate.20 The precipitate wasdissolved in water (the same amount as theoriginal volume of antiserum) and reprecipitat-ed in 33% saturated ammonium sulfate. The

* Sylvania Company, Millburn, NJ.t Bioquestt Type 1, Sigma Chemical Co., St. Louis, Mo.

final precipitate was dissolved in physiologicsaline (half the original volume of antiserum)and dialyzed ovemight in PBS (pH 7.6) toremove sulfate ions. The protein content ofthe retentate was determined by the Biuretmethod with the spectrophotometer at 540nm.21

Twelve milligrams of horseradish peroxi-dase4 were added to a solution containing 5mg of globulin (protein) in 1 ml of 0.1 M phos-phate buffer (pH 6.8). Then 0.05 ml of 1%glutaraldehyde (coupling agent) was addedwhile shaking and the mixture allowed to standat room temperature for two hours. The con-jugate was dialyzed overnight in two changes ofPBS (pH 7.6), and centrifuged at 10,000 X gfor 30 minutes in a Lourdes refrigerated centri-fuge. Finally the conjugates were filter steril-ized through a Morton ultra-fine sintered glassfilter and stored at 4 C.

PEROXIDASE STAINING.- A sufficient amountof vaccine or culture was placed in a sterileKahn tube to give a pellet after centrifugationof about a half millimeter thick. To eachsample 0.1 ml of appropriate peroxidase con-jugated antiserum was added, mixed and al-lowed to stand 30 minutes at room tempera-ture. The samples were centrifuged at 3000 X gfor 10 minutes; the supematant was pipettedoff and the pellet was washed once with PBS(pH 7.6) and again centrifuged. After pipet-ting off the supernatant fluid, each sample wasstained with 1 ml of a solution containing 5 mgof 3,3'-diaminobenzidine tetrahydrochloridelin 10 ml PBS (pH 7.6) to which 0.01% hydro-gen peroxide (1 ml of 0.1%) was added im-mediately before use.20 The preparations werecentrifuged and washed once as above.

ELECTRON MICROScOPY.-After the perox-idase staining, the bacterial cells were pelletedand immediately resuspended in 3% glutaral-dehyde containing 0.1 M phosphate buffer (pH7.2) for 1-2 hours at room temperature. Thecells were washed three times in 0.1 M phos-phate buffer containing 0.2 M sucrose andstored overnight in this mixture. The follow-ing day the cells were postfixed in 2% osmiumtetroxide in 0.1 M phosphate buffer, de-hydrated in a graded ethanol series, infiltratedwith propylene oxide and Epon and embeddedin Luft's Epon mixture. Sections were madewith a diamond knife on a Reichert Om-U2ultramicrotome. For each preparation, onegrid was poststained with uranyl magnesiumacetate23 and lead citrate24 and one was notpoststained. Grids were examined and photo-

Vol. 56 No. 2 167

168 (CRENIER ET AL

gaiphedl x-itli an- \EI-Cornith 275 eleitron

MAo ROP1t'tGE (11II RII l) STAINING.-Rabbit alveolairlnt Imroph-as t itre otalined bvat tlioitWatioll of -a tii('tod rsttilscied bxMixcl ik. Lioake and 1Faiits. Ne xt Zealandxxiie raibbits xx- gighin' hetwxxeon 2 and(I 3 koxx ci satriu dbv iijiectjottin ( 10 nil of airilito the niarginal ear x in. The tlioraci( axaityx st opi(ned and the tipper part of the tratchbowx as tIaitcted. TlFie limogs, lieart aind tiat bohedxvredisselcted as a singloe iuniit. Th'lirty nll of1alnuiks l)almn(osalt solution (HBSS) conltait-iMU 10 units of Iiepariit por nillilitet xvercslowxlx iiijt' tewd xtitli a svrinii-e throt-ith a 20-(aillge inoedle hexeol ren-iovxed) betxeen txsvotrat heal rini.S. Thlie fluid xsax x ithdraxxn andpltacd iiia large (-e ntrifug(e tuilie. The pro-exx(dLa(s i epeatepcmd aitain and ret ovxered in tlhe

saiiio miinaiiier.Thexl ielc of mni rophag(eet was dclettiniiiod

bv t omiiitiiig the tells iL a henatytonieter. Thevialailitx of the te<ls Nxas determiined hi slS-l)endiiig a11 equal xvoluime of cells in 0.2%trvpan blue andl (counting iti a henacytoiieter.

t Microbiological Associates, Bethesda, Md.

Ii a1ll experinielil'., 0 t)0% of the imirio-phiages ext luded thlCt pan blue die. J)iffer-(eiitial counts m01 e iado on Wright's stainedsiiiears; it wis ftoind that about oftf thee e\ iatcropiaat'(s.

stfaroprtyao's wxr t' suispeindletd in Tissue(UIltnLe MedCCinnu 100# (1)11 adjusted to 7.2xx-ithl 1% sodilu bit aicahonate' cottitailnino 100

iLiul p Icliilliin G and 100 jto ml streptomyxtinand 2O0r9() iioiniial, iniactivxated abbit srurnii to ationtentration of approxiialtelx 5 X 10' cells/nil.

Shaker airl rotloer lube cmltures xxwere usedinitially. Forialin-fixed cclls tof all fixe srol-types tof S inulanms xxere added 0.1 X 107) to

inat rophiages aiid al]txwcd tt inmtubate '3 litursMt 241l otiIs at 37 C.

Later Utultuies wx ore gmroi\\i ill I eiIhitontubes ixitli a 9 X u5 miii. toxetr slip aitldic-blted at 37 C ftr 24 bours for attat liihent antdstalbiliiatioin. lit' next dax tlie nimediun wxas te-iioed and cells xxv(re xxaslhed txxo tion's ixNitliIIBSS andCi fresh mnodiuimi added xx ith 20',noriiial tr inlitune ralthit sernnm. lornialill-fixed cells of all fixe serotypes of S untans xxereadded (9.1 X 10' to llat Ophiao es; and al-

A04

J t8Tto1~~~~~~~~~"

I4,;

a b actv~~~~~~~~~~~[.FIc I-. Electron micrographs of thin sec-

tions of S. omuton,. All prejparations were post-stained with tiran-yl mtagrnesium acetate and leadcitrate. Magnification is X 25 .000. (a) Strain

AJLT-d with no antiserurn. (b) Straiii AHT-dtreated with ATIT-d peroxidase conjugated anti-serturm. (c) Strain AHT-d treated with AHT-dunilabelcd atitiseruini.

J Dent Res February 1,977

VMICROCAPSULES ON MUTANS STEROTYPES 169

lowed to incubate 4 hours at 37 C. The coverslip preparations were washed again in HBSS,fixed in absolute methanol for 10 minutes, airdried and then stained with May-Grunwald-Geisma stain.26 Stained cover slips weremounted with Mounting Medium/ and ex-amined in a light microscope with oil im-mersion 100 X objective. On each preparation100 macrophages were examined for phago-cytosis.

ResultsThe morphology of untreated AHT-d cells

by transmission electron microscopy can beseen in Figure la. There is no evidence of cap-sular or extracellular material around the cells.However, when the same AHT-d cells weretreated with homologous immune serum thathad been conjugated with peroxidase, anelectron-dense layer (approximately 200 nm)was evident (Fig lb). When the AHT-d cellswere treated with unconjugated antiserum(Fig lc), a finer consistency was seen whencompared with the coarse particles of theperoxidase labeling. These electron-denselayers were just as evident in sections whichwere not poststained with uranyl magnesiumacetate and lead citrate. However, this layercould not be demonstrated by India ink prep-arations or by Quellung reaction using light-field or dark-field microscopy.27

The AHT-d peroxidase conjugate did notstain extracellular material on S sanguis cells,strain 10556, or S salivarius cells, strain HHT(Table 1). When AHT-d cells were treatedwith pooled normal rabbit serum followed byperoxidase-labeled goat anti-rabbit serum andpoststained with uranyl acetate and lead ci-

/ Technicon Chemical Co., Inc., Chauncey, NY.

trate, there was no evidence of extracellularstaining (Table 1).

Representative strains of Streptococcusmutans serotypes a, b, c, d, and e were alsotreated with the AHT-d peroxidase conjugatedantiserum and the electron micrographsshowed varying amounts of labeled extracellu-lar substance. AHT-a cells (Fig 2a) showedless extracellular material (ECM) than AHT-dcells (Fig lb). The Fa-1 cells (serotype b)showed very little ECM (Fig 2b); Ingbritt cells(serotype c) seemed to be a mixture of cellswith and without ECM (Fig 2c); OIHI cells(serotype d) showed a considerable amount ofECM on every cell (Fig 2d) and the LM7 cells(serotype e) appeared to have no ECM. Whenthese cells were reacted with peroxidase con-jugated homologous antisera, essentially thesame degree of ECM reactions were seen asthose described with the AHT-d peroxidaseconjugate.

Table 2 shows a comparison of the titersobtained by agglutination tests, direct and in-direct FA tests using homologous cells andantisera, and the relative amounts of ECM asseen in the electron micrographs after stainingwith either the AHT-d peroxidase conjugate,or with homologous peroxidase conjugates.There seemed to be no correlation between theagglutination titers and the FA titers. This wasalso found by Griffin28 using nonspore-forminganaerobes. The indirect FA titers were higherthan the direct with every strain except BHTand LM7. The serotype d strain exhibited afourfold increase with the indirect FA test. Therelative amount of ECM as estimated by stain-ing with the AHT-d peroxidase conjugatedantiserum seemed to parallel direct and in-direct FA titers.

LE 1

SPECIFICITY OF IMMUNE SERUM

Cells

S mutans (AHT-d)S mutans (AHT-d)

Ssanguis (10556)

S salivarius (HHT)

S mutans (AHT-d)

Conjugates

plus

plus

plus

plus

AHT-d immuneserum-peroxidaseAHT-d immuneserum-peroxidaseAHT-d immuneserum-peroxidasepooled normal rabbit serum

and goat anti-rabbit serum-

peroxidase

Capsular Reaction

no capsulecapsule

no capsule

no capsule

no capsule

Vol. 56 No. 2

170 GRSNTIER ETALJ

In suI)sequllet studies, the A IT-d strainthat had been repeatedly passed in BHS brothfailed to give a FA rea tion svith the originalAlIT-d antiserum (Table 3) and exhibited amininmal reaction with the AHT-d peroxidaseconjugate (Fig 3a). This suggested that the

A8

.i..v b

ie.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~n ; * F 0 0a

AlIT-d straini had lost, upon laboratory pas-sage, the antigen(s) that had been present inthe orig-inal vVIi ci x accine anid that ssere re-sp)onsible for both the FA and the ECM re-action. Th-is couild hasve been caused by theroutine sele tion of nonadherent cells by the

WS *'^40t 6.:-41.

.po |IW

^.S, (t5~S11

1.1:.. oe.1..1.,k.

IN.,

X': .,:: .A -"

F:t A~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.ia-

a

w.fc

4

eFim 2. Electron rnicrographs of thin sections

of S mutans. All strains treated with AHT-dperoxidase conj ugated antiserum. Magnificationis X 25,000. (a) Strain AHT-a. Poststained. (b)

Strain Fa-1. Not poststainied. ( ( ) Strain Ingbritt.Poststained. (d) Strain OIHI. Poststained. (e)Straini LM7. Not poststained.

I Deiit Res February 1977

W

4:

b

MICROCAPSULES ON AMUTANS STEROTYPES 171

4 ~CM CM

-

*-F*- j- - 4 ---. .0

Cl-4 0)C

_, _0 _o u _

*>u--4 ..

-l VU _0_ _

;-t

d., e =

E0

0

a

CZ0

a

9 E$ '0

9 *

.0

_ r

C0

a .-

g s

~ O,*

Q9C2Q3

3CZuQc

+:

transfer process, which picked cells suspendedin the medium and left behind those sticking tothe glass walls of the test tubes. A new vaccine,prepared with the repeated passage of AHT-dcells, failed to elicit an antiserum with a satis-factory FA titer, confirming the suppositionthat laboratory passage had resulted in the lossof antigenic determinant(s). As this antigenappeared to be associated with the ECM, thepossibility existed that the ECM could be amicrocapsule.27 Therefore, efforts were madeto enhance capsule formation in this strain. Theaddition of 10% unheated rabbit serum to theBHS medium29 resulted, after three passagesof AHT-d, in the restoration of the FA reaction(Table 3) and the ECM formation as seen inthe electron microscope with the peroxidaseconjugate (Fig 3b). An antiserum, made inrabbits, to the AHT-d restored cells showed adirect FA titer of 1:16. Thus, the unheatedrabbit serum had restored ECM and anti-genicity to the AHT-d repeated passage cells.Heated rabbit serum (56 C for 30 minutes) wasalso tried as a medium supplement but was notas effective as the unheated serum.

A lyophilized sample of our original AHT-d strain was rehydrated and grown in the BHS-rabbit serum medium. This revived strain hada copius amount of capsule (Fig 3c) and gavea strong FA reaction (Table 3). This suggestedthat the low FA titers obtained with the otherserotvpes (Table 2) might be improved bygrowing those cells in rabbit serum. This wasdone using one strain each of the five serotypes.

The possibility that serum protein couldbe adsorbed to the cells thereby presenting newantigenic determinants to the host had to beconsidered. Strains of the five serotypes of Smutans were subcultured three times in BHSmedium with or without the addition of rabbitserum. The cultures were centrifuged and thecells suspended in physiologic saline. The cellswere tested with conjugated antisera made tothe serotypes grown in medium with rabbitserum. Table 4 shows the direct FA reactionsusing the cell preparations and the conjugatedantisera. Fluorescent staining occurred onlywith the serotype homologous cells and conju-gates. Cells grown in the medium with rabbitserum showed stronger FA reactions than thosegrown in medium without the serum supple-ment. The reaction on cells of strain 10449with GS5 conjugate was very weak. There wasno overall nonspecificity caused by the additionof rabbit serum.

The effect of growing the cells in the

Vol. 56 No. 2

En

9D

U

u

92

0H

zU

E4

0

Es4

Ev0

a

92

(A

C2H

H

0

0H

0

0

0

0

04

cH

¢

.0 0

. Ou 0

L.W

-4

r._

0

S._

w

2V._

30

to

to

.2

4

0

V2

172 GRENIER ET AL

TABLE 3

COMPARI.SON OF FA AN-D PLROXIDASE RFAC TION\S WITH CELLSGROXN' IN BRAIN-HEART-SPIROLATF BROTH (BIIS) WITH

AN-D WITHOUT RABBIT SERUM (RS)

AlT-d Cells Grown in

BHSBHS + RSBHS + RS

Labeled Antisera MadeAgainst AHT-Td Original Cells

FA Reaction Peroxidase Reaction

4

4

44

* Repeated-original AHI-d cells passed weekly in BHS.RZestored-repeated passage AHT d cells grown in BHS + tunheated rabbit serurm.Revived original lyophilized AHT-d cells revived in BIUS + Lnheated rabbit serLim.

a

::i.:;. mC:::

Flo 3. Electron micrographs of thin sec-

tions of S mutatis. Magnification is X 25,000.(a) Strain AHT-d (repeated passage) treatedwith AHT-d peroxidase conjugated antiserum

Not poststained. (b) Strain AHT-d (restored

ailtigenicity) treated with AHT-d peroxidaseconjulgated antiserulrni. Poststained. (c) StrainiAHT-d (revived antigenicity) treated with uin-

labeled anlisertLisn tI AHT-d restored cells. Post-stairletd.

CellDesignations*

RepeatedRestoredRevived

J Dent Res February 1977

MICROCAPSULES ON MUTANS STEROTYPES 173

TABLE 4

FA CROSS REACTIONS OF CELLS GROWN IN BRAIN HEART-SPIROLATE BROTHWITH OR WITHOUT RABBIT SERUM AND FLUORESCEIN CONJUGATED

ANTISERA TO SEROTYPES GROWN IN MEDIUM WITH RABBITSERUM ERIOCHROME BLACK COUNTERSTAIN

Conjugated Antisera to Cells Grown in BHS withRabbit Serum

Cells Grownin BHS+

Rabbit serum-E49Fa-1GS510449AHT-dOMZ65B2

S salivarius

No Serum-E49Fa-1GS5AHT-dOMZ65B2

S saliVarius

AHT-d AHT-d-restored revived

Serotype 1:16 1:8

abc

c

dde

abc

dde

Fa-1Undil.

GS5 6715 B2Undil. 1:16 Undil.

+ _-

-- 4 _- _ ___ -_ 4 - -_ ~+__

44

44

_ _ 4 -

_-_ 4 -

_- _ - 3

+

22

presence of various enzymes on the FA reactionand the ECM was also evaluated. TrypsinlT(0.1%, 2 X crystallized) reduced but did noteliminate the FA stain for the restored and re-vived cells, and did not appear to have muchof an effect on the ECM. Amylase,11 lipase**and dextranasett were also tried but had noeffect on the FA reaction. Hyaluronidase(7,500 units, bovine testes+t) reduced the FAreaction but did not alter the peroxidase stain.

Limited studies were done to determinethe phagocytability of the various serotypes.Rabbit alveolar macrophages were grown over-night in Medium 199 and 20% normal rabbitserum. When formalin fixed cells of all fiveserotypes were added (2 to 1) to the washedmacrophages, no phagocytosis occurred exceptwith serotype e. Every intact macrophageshowed phagocytosis of a few cells of serotypee (strain B2 used) when the reaction tookplace in shaker and roller tube cultures (Fig

¶ Nutritional Biochemical Corporation.Nutritional Biochemical Corporation.** Aldrich Chemical Co., Milwaukee, Wi.

tt Sigma Chemical Co., St. Louis, Mo.U Worthington Biochemical Co.

22

22

_ 42

4a). However, when the Leighton tube cultureswere used (Fig 4b), the macrophages were en-gorged with cells. The cover slip preparationsobviously offered a better chance for the con-tact of cells and bacteria. The same resultswere found when homologous immune serawere used in place of the normal rabbit serum,except that 1% of the macrophages showedphagocytosis with the AHT-d antiserum andcells.

Discussion

Strain AHT-d was used extensively in thisstudy because it had a higher working titer inFA tests than the other strains2 and the highestF/P ratio (Table 2). Its antigenicity may inpart be related to the presence of the ECM onthe cells of the original vaccine, i.e., electronmicroscopy of AHT-d revealed ECM whichparalleled the intensity of the FA reaction. TheECM was demonstrable by staining with eitherimmune serum or with peroxidase conjugatedimmune serum. The AHT-d peroxidase con-jugated antiserum stained all serotypes of Smutans to the same extent as did each homolo-

Vol. 56 No. 2

171 GRENIER ET AL

FiG- 4. Photographs of rabbit alveolar inacrophages with pha-ocytized Smutans cells serotyle e. (a) Macrophages in tissue culture medium and normalrabbit serUrm. Strain B2 (ells were added and shaken three hours at 37 C. Mag-nification is X 750. (b) Macropha-cs in tissute culture medium and immutnerabbit serumn grown in, Leightoni tubes. Strain B2 cells were added and incubatedovernight at 37 C. Magnification is X 1,875.

J Dent Res February 1-977

MICROCAPSULES ON MUTANS STEROTYPES

gous conjugated antiserum by electron micro-scopy. However, the AHT-d peroxidase con-jugate did not stain S salivarius or S sanguiscells.

Serial passage of the AHT-d cultures inBHS resulted in the loss of FA reaction and lossof peroxidase reaction. The addition of therabbit serum restored antigenicity as measuredby both reactions. If the increase in both theFA staining and antigenicity of the strepto-coccal cells were due to the absorption of serumproteins on the cells' surface, then antibodymade against the cells would be nonspecificand react with all the serotypes grown in BHSrabbit serum medium. This was not the case.Cells of the various serotypes after growth inBUS rabbit serum medium were not stainedby heterologous antisera (Table 4).

While the addition of unheated rabbitserum did restore antigenicity to the AHT-dcells, the direct FA titer of the conjugated anti-serum was no higher than the original titer hadbeen. The indirect FA test could not be used todetect sensitivity after growing the organismsin rabbit serum, as normal proteins on the cellswould have been detected as well as specificantiserum that had been produced in rabbits.

The rabbit serum may be exposing the Smutans cells to proteins which they normallymeet in vivo either in the saliva30 or in gingivalcrevice fluid.31 These serum proteins maystimulate S mutans to form a protective cap-sule detectable as ECM which enables it tosurvive more successfully in their presence. Incontrast, when the cells are cultured in vitro inthe absence of serum proteins, there is no needfor capsule formation and this surface antigenis no longer synthesized. This might accountfor the low antigenicity of S mutans cultures asjudged by the FA reactions. Agglutinationtiters are not affected by serial passage, indi-cating that these antigens are distinct fromthose measured by the FA reaction. In factthere appears to be an inverse relationship be-tween direct FA titer and the agglutinationtiter (Table 2).

Phagocytosis studies were undertaken tofurther demonstrate differences in extra-cellular material in the cells of the S mutansserotypes. Differences were shown withformalin-killed cells, i.e., strains with ECMwere not phagocytized, while the group e strainwhich had no demonstrable ECM was phago-cytized in the presence of normal rabbit serum.

The ECM has several characteristicswhich suggest that it may be a capsule, i.e.. it is

found when the cells are grown in the presenceof serum, it is lost upon serial passage and givesthe cell some protection from phagocytosis.This capsular material appears to be the anti-gen that is stained by the fluorescent and perox-idase conjugated antibody as well as the un-conjugated antibody. This raises the possibilitythat S rnutans can be identified in situ inplaque.

Lai, Listgarten, and Rosan32 have de-scribed a capsular reaction on S sanguis usingimmune serum. In a later report,33 the authorshave shown by absorption that c and d antigensof S sanguis are located in the extracellular ma-terial of this organism. Further they haveshown that the "immunocoating reaction"could be used for the in situ identification of Ssanguis in dental plaque. Cells showing positivereactions were arranged as pyramidal shapedcolonies with the apex at the tooth surface andthe base at the surface facing the oral cavity.

ConclusionsExtracellular microcapsules have been

demonstrated on most serotypes of S mutanswhich might be of significance in the adher-ence, survival, and pathogenicity of this organ-ism. A nongenetic variation was shown to occurin strain AHT-d resulting in the loss of FAstaining and capsular substance, but both wererestored by the addition of unheated rabbitserum to the growth medium. It seemed rea-sonable to assume that other serotypes mightalso lose the same type of antigenicity in thisand other media. This may account for the lowtitered antisera obtained with some strains ofthis organism.

We would especially like to thank Mrs. Roberta K.Brabec for her meticulous work in preparing the largenumber of thin sections used in this study. We alsothank Dr. Generosa G. Morales, Department of Health,Manila, for doing agglutination tests with the variousstrains of Streptococcus mutans. Mr. Samuel Gregorio,graduate student in The University of Michigan Schoolof Public Health, assisted in many aspects of this study.

References1. BRATTHALL, D.: Demonstration of Five

Serological Groups of Streptococcal StrainResembling Streptococcus mutans. OdontRevy 21:143-152, 1970.

2. GRENIER, E.M.; EVELAND, W.C.; andLOESCHE, W.J.: Identification of Strepto-coccus mutans, Serotypes in Dental Plaqueby Fluorescent Antibody Techniques. ArchOral Biol 18: 707-715, 1973.

3. ZINNER, D.D., and JABLON, J.M.: Humanstrept3coccal strains in experimental caries,in Harris, R.S., (ed): The Art and Scienceof Dental Caries Research, New York, Aca-demic Press, 1968, pp. 87-109.

Vol. 56 No. 2 175

176 GRENIER ET AL

4. BRATTHALL, D.: Immunofluorescent Identi-fication of Streptococcus mutans. OdontRevy 23:181-196, 1972.

5. HEBERT, G.A.; PITTMAN, B.; McKINNEY,R.M. and CHERRY, W.B.: The Preparationand Characterization of Fluorescent An-tibody Reagents. P.H.S Manual. U.S. Dept.of HEW, CDC, Atlanta, Georgia 30333,1972.

6. PITTMAN, B.; HEBERT, G.A.; HARRIS, P.P.;McKINNEY, R.M.; THACKER, L.; andCHERRY, W.B.: Preparation and Evaluationof Fluorescent Antibody Reagents for Strep-tococcus mutans. ASM Preprinted Ab-stracts. 74th Annual Meeting. AbstractM134, 1974.

7. FITZGERALD, R.J., and KEYES, P.H.: Dem-onstration of the Etiologic Role of Strepto-cocci in Experimental Caries in the Ham-ster. JADA 61:9-19, 1960.

8. ZINNER, D.D.; JABLON, J.M.; ARAN, A.P.;and SASLAW, M.S.: Experimental Caries In-duced in Animals by Streptococci of HumanOrigin. Proc Soc Expl Biol Med 118:766-770, 1965.

9. FITZGERALD R.J.; JORDAN, H.V.; and STAN-LEY, H.R.: Experimental Caries and Gin-gival Pathologic Changes in the GnotobioticRat. J Dent Res 39:923-935, 1960.

10. GIBBONS, R.J., and FITZGERALD, R.J.: Dex-tran-Induced Agglutination of Streptococcusmutans, and Its Potential Role in the Forma-tion of Microbial Dental Plaque. J Bacteriol98:341-346, 1969.

11. KRASSE, B.: Human Streptococci and Ex-perimental Caries in Hamsters. Arch OralBiol 11:429-436, 1966.

12. CARLSSON, J.; S6DERHOLM, G.; and ALM-FELDT, I.: Prevalence of Streptococcus san-guis and Streptococcus mutans in the Mouthof Persons Wearing Full Dentures. ArchOral Biol 14: 243-249, 1969.

13. FITZGERALD, R.J.; KEYES, P.H.; STOUDT,T.H.; and SPINELL, D.M.: The Effects of aDextranase Preparation on Plaque andCaries in Hamsters, a Preliminary Report.JADA 76:301-304, 1968.

14. GIBBONS, R.J.; BERMAN, K.S.; KNOETTNER,R.; and KAPSIMALIS, B.: Dental Caries andAlveolar Bone Loss in Gnotobiotic Rats In-fected with Capsule Forming Streptococciof Human Origin. Arch Oral Biol 11:549-560, 1966.

15. EDWARDSSON, S.: Characteristics of Caries-Inducing Human Streptococci ResemblingStreptococcus mutans. Arch Oral Biol 13:637-646, 1968.

16. SHKLAIR, I, and KEENE, H.: A BiochemicalScheme for the Separation of the Five Va-rieties of Streptococcus mutans. Arch OralBiol 19:1079-1081, 1974.

17. COYKENDALL, A.L.: Four types of Strepto-coccus mutans Based on their Genetic Anti-

genic and Biochemical Characteristics. J GenMicrobiol 83: 327-338, 1974.

18. WELLS, A.F.; MILLER, C.F.; and NADEL,M.D.: Rapid Fluorescein and Protein AssayMethod for Fluorescent-Antibody Conju-gates. Appl Microbiol 14(2): 271-275,1966.

19. GOLDMAN, M.: Fluorescent Antibody Meth-ods, New York, Academic Press, 1968, p.126.

20. PETTS, V., and ROITT, I.M.: PeroxidaseConjugates for Demonstration of Tissue An-tibodies: Evaluation of the Technique. ClinExpl Immunol 9:407-418, 1971.

2 1. GORNALL, A.G.; BARDAWILL, C.J.; andDAVIS, M.M.: Determination of Serum Pro-teins by Means of the Biuret Reaction. JBiol Chem 177: 751-766,, 1949.

22. AVRAMEAS, S.: Coupling of Enzymes to Pro-teins with Glutaraldehyde. Immunochemis-try 6:43-52. 1959.

23. FRASCA, J.M., and PARKS, V.R.: A RoutineTechnique for Double Staining UltrathinSections Using Uranyl and Lead Salts. JCell Biol 25:157-161, 1965.

24. REYNOLDS, E.S.: The Use of Lead Citrateof High pH as an Electron Opaque Stain inElectron Microscopy. I Cell Biol 17:208-212, 1963.

25. MYRVIK, O.N.; LEAKE, E.A.; and FARISS,B.: Studies on Pulmonary Alveolar Macro-phages from the Normal Rabbit: A Tech-nique to Procure Them in a High State ofPurity. J Immunol 86:128-136, 1961.

26. MERCHANT, D.J.; KAHN, R.H.; andMURPHY, W.H.; JR.: Handbook of Cell andOrgan Culture. 2nd. Ed. Burgess PublishingCo., Minneapolis, Minn., 1969.

27. WILKINSON, J.F.: The Extracellular Poly-saccharides of Bacteria. Bacteriol Revs 22:46-73, 1958.

28. GRIFFIN, M.: Fluorescent Antibody Tech-niques in the Identification of the Gram-Negative Nonspore-Forming Anaerobes.Health Lab Science 7: 78-83, 1975.

29. HANSON, A.W., and CANNEFAX, G.R.: Rehy-dration of Lyophilized Spirochaetal Cultures.J Bacteriol 88:811, 1964.

30. LEHNER, T.; CARDWELL, T.; and CLARRY,E.: Immunoglobulins in Saliva and Serumin Dental Caries. The Lancet 1294-1297,1967.

31. WEINSTEIN, E., and MANDEL, I.D.: TheFluid of the Gingival Sulcus. Periodontics 2:149-153, 1964.

32. LAI, C.; LISTGARTEN, M.: and ROSAN, B.:Serology of Streptococcus sanguis: Locali-zation of Antigens with Unlabeled Antisera.Infect Immunity 8:475-481, 1973.

33. ROSAN; B.; LAI, C.H.; and LISTGARTEN,M.A.: Streptococcus sanguis: A Model inthe Application of Immunochemical Analy-sis for the Localization of Bacteria in DentalPlaque. J Dent Res (In Press), 1976.

J Dent Res February 1977