INFECTION AND IMMUNITY, Apr. 1993, p. 1259-12670019-9567/93/041259-09$02.00/0Copyright ©) 1993, American Society for Microbiology

Streptococcus mutans Fructosyltransferase (ft andGlucosyltransferase (gfBC) Operon Fusion

Strains in Continuous CultureDONALD L. WEXLER,"2 MICHAEL C. HUDSON,3 AND ROBERT A. BURNEl.2*

Departments ofDental Research1 and Microbiology and Immunology, 2 University ofRochesterMedical Center, 601 Elmwood Avenue, Rochester, New York 14642, and Department ofBiology, University ofNorth Carolina at Charlotte, Charlotte, North Carolina 282233

Received 28 September 1992/Accepted 18 January 1993

Three glucosyltransferases (GTFs), which catalyze the formation of water-insoluble adherent glucans, andfructosyltransferase (ElF), which synthesizes fructans, are believed to contribute to the pathogenic potentialof Streptococcus mutans. Study of the regulation of expression ofGTF and lTF has been difficult because of thecomplexity and number of exoenzymes produced by this bacterium. By using continuous chemostat culture tocontrol environmental conditions, chloramphenicol acetyltransferase (CAT) operon fusions were utilized tomeasure transcriptional activity of theftf and ggTC gene promoters. Expression of these operon fusions wasdifferentially regulated in response to culture pH and growth rate and during transition states between growthdomains. Furthermore, the addition of sucrose to steady-state cultures resulted in significant increases in CATspecific activities for both fusions. In a few cases, GTF and FTF enzyme specific activities did not parallel thoseof the corresponding CAT fusion activities; this lack of correspondence was likely due to posttranscriptionalevents controlling enzyme secretion and enzyme activity, as well as to the differential expression ofdextranase(s) and fructan hydrolase by S. mutans. These results clearly demonstrate that the extracellularpolymer synthesis machinery of S. mutans is regulated in a complex manner. The use of operon fusions incombination with chemostat culture is a viable approach to analyzing gene expression in S. mutans and will behelpful in defining the molecular mechanisms underlying regulation of expression of virulence attributes underconditions that may more closely mimic those in dental plaque.

The virulence of Streptococcus mutans is dependent onseveral gene products and factors which contribute, tovarying degrees, to the pathogenic potential of this organ-ism. Among these virulence attributes, the ability to synthe-size extracellular polysaccharides from sucrose has beendemonstrated to augment the cariogenicity of oral strepto-cocci (15, 18). S. mutans synthesizes glucans and fructansfrom sucrose via the actions of glucosyltransferases (GTFs)and a fructosyltransferase (FTF), respectively. Glucans ap-pear to serve in multiple capacities, including mediatinginitial adherence to the tooth surface (39, 45), facilitatingbacterial accumulation on smooth surfaces (13, 32), and as areservoir for metabolizable polysaccharides outside of thecell (5). Fructans are believed to function exclusively asextracellular storage compounds (5, 33, 58), but few inves-tigations into other roles for this abundant plaque constituenthave been conducted.

S. mutans produces at least three GTFs. GTF-S is theproduct of the gtfD gene and catalyzes the formation of arelatively water-soluble glucan composed almost entirely ofa(1,6) linkages (23). The GTF-I and GTF-SI enzymes, theproducts of thegtB and gtfC genes, respectively, synthesizepredominantly water-insoluble glucans. These polysaccha-rides are rich in a(1,3) linkages but also contain substantialamounts of a(1,6)-glucan (1, 22). Previous studies (51) haveindicated that the gtfB andgtfC genes are cotranscribed froma single promoter. The gtfD gene is not genetically linked tothe gfBC locus (37). In the rat caries model, strains defec-tive in the synthesis of active gtfBC gene products demon-

* Corresponding author.

strated marked reductions in the ability to elicit smooth-surface caries (36, 44), while inactivation of GTF-S appearedto have little effect on cariogenicity (36).

S. mutans produces a single FTE, which catalyzes thesynthesis of fructans composed predominantly of P(2,1)linkages (3, 11). Plaque fructans accumulate rapidly in vivofollowing ingestion of sucrose (16) and are then hydrolyzedto fructose by fructan hydrolases produced by S. mutans (6,9, 53) and other oral bacteria (49). A significant contributionof FTE, plaque fructans, and fructan hydrolases to virulencehas not been readily demonstrable by using otherwise-isogenic mutants in the rat caries model (30, 57). However,experiments to allow for the manifestation of the full cario-genic potential of organisms which can utilize storagepolysaccharides have not been performed.The use of a chemostat to modulate the growth rate, pH,

and other culture conditions has illustrated the remarkablephenotypic plasticity of the oral streptococci. By altering thegrowth environment, a wide range of phenotypic changes inpolymer synthesis capacities (12, 55, 56), adherence capa-bilities (40), acidogenicity (27), acidurance (2, 20), carbohy-drate transport (19), and glycolytic capacities (7) of S.mutans has been elicited. In particular, it has been demon-strated that the phenotypic expression of the glucan andfructan synthetic capacities of oral streptococci is regulatedin response to multiple environmental variables (24, 34, 50).However, little is known about the expression of individualgene products participating in polysaccharide metabolism.This is largely due to the fact that a variety of factors havecomplicated the study of individual enzymes involved inpolysaccharide synthesis. For example, the antigenic simi-larity between GTFs (31, 43, 48) and the breakdown of GTFs

1259

Vol. 61, No. 4

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

1260 WEXLER ET AL.

(35) and FTF into multiple, active lower-molecular-weightspecies prohibit accurate quantitation of these proteins andtheir activity by immunologic or biochemical methods. Ag-gregation has a clear effect on the kinetic properties of theseenzymes (1), as does surface association (45), so that thedetermination of activity and its correlation with proteinquantity are similarly confounded. GTFs and FTF can befound in cell-associated forms (17), yet little is understoodabout the activities of these enzymes when anchored at orbound to the cell surface. Moreover, the polysaccharidesproduced by the GTFs are heterogeneous and partiallysusceptible to the action of the S. mutans endodextranase,such that it is nearly impossible to discriminate the contri-bution of an individual GTF to the synthesis of the finalglucan end product. Notably, the GTFs and FTF competefor a common substrate, which may influence measurableactivities. Other factors, such as association of lipoteichoicacids, which are found in large quantities in the culturesupernates under particular conditions and can associatewith glucans, may affect the way in which the polysaccha-ride is precipitated, since lipoteichoic acids are amphipathic.Also, the presence of differentially regulated glucan andfructan hydrolases (26, 54), which attack the products of theGTFs and FTF, will influence the amounts of total polysac-charide measured in biochemical reactions. Undoubtedly,other factors, such as proteases or enzyme inhibitors in cellsupernates, could influence measurable activity. Clearlythen, a combination of methods will be necessary to com-pletely dissect the molecular basis for the general observa-tion that the phenotypic expression of polysaccharide syn-thesis is differentially regulated in response to environmentalconditions.

In order to better understand the modes and mechanismsof regulation of expression ofgfBC andftf, chloramphenicolacetyltransferase (CAT) fusions in conjunction with chemo-stat culture were utilized to monitor the transcriptionalactivity from the promoters of these genes as a function ofculture conditions. The fusions were constructed in such amanner as not to disrupt the normalgtforffgenes (24). Thistechnique allowed for the measurement of transcriptionalactivity from thegfBC andftf promoters and of extracellularpolymer synthesis as a function of growth rate and pH and inresponse to the addition of sucrose to steady-state cultures.(A portion of these data was presented at the 1992nd

General Meeting of the American Society for Microbiology,26-30 May 1992, New Orleans, La. [56a].)

MATERIALS AND METHODS

Bacterial strains and media. S. mutans SMS101 (ff-cat)and SMS102 (gtfBC-cat) have been previously described(24). The chromosomal structures of these strains at theffor gtfBC locus are shown in Fig. 1. These strains weremaintained on brain heart infusion agar (BHI; Difco Labo-ratories, Detroit, Mich.) supplemented with 10 p,g of eryth-romycin per ml. For continuous culture experiments, S.mutans strains were grown in a Bio-Flo III chemostat (NewBrunswick Scientific, Edison, N.J.) in tryptone-yeast extract(TY; Difco) medium (6) supplemented with 10 mM glucoseand 10 ,ug of erythromycin per ml.Growth conditions and preparation of enzymes. One liter of

TY broth was inoculated with 1 ml of a TY starter culture ofeither S. mutans SMS101 or S. mutans SMS102 and grownin the Bio-Flo III at a dilution rate (D) of 0.05 h-1. D isdefined as the amount of medium added as a fraction of thetotal starting vessel volume per unit of time (e.g., D = 0.05

ftfpromoter

V bpCAT

ftf promoter

r\ 8Sbp/

Emr ftfN S. mutensMSSMS101 genome

gtf B/C promoter gtf B/C promoter

S. mutensSMS102 genome

FIG. 1. Chromosomal structure of SMS101(ff-cat) and SMS102(gfBC-cat). Arrows depict direction of transcription. The sectionslabelled CAT, Emr, gtf, andfif are not to scale. For a more detaileddescription of SMS101 and SMS102, see reference 24.

h-1 indicates that a 0.05 vessel volume, 50 ml, was added tothe chemostat vessel, which was held at a constant volumeof 1 liter, per hour). Steady-state growth was assumed whenthe culture was maintained for at least 10 generations undera particular growth condition. Control of culture pH wasaccomplished by the addition of 2.0 M potassium hydroxide.For sucrose induction experiments, a 0.75 M sucrose solu-tion was added to the vessel to yield a final concentration of20 mM sucrose. To assess the stability of the operon fusions,samples of the culture at the end of an experiment werediluted and plated on BHI and on BHI with 10 p,g oferythromycin per ml. Counts (CFU) on BHI and BHI pluserythromycin were compared to evaluate the stability of theEmr marker. Similarly, CAT levels at the beginning and endof each experiment, under identical conditions, were com-pared since lower enzyme activity following long-term cul-ture would be indicative of loss of the CAT fusion.To sample the culture, 50-ml aliquots were removed at

preselected times from the chemostat and immediately cen-trifuged at 8,000 x g for 10 min at 4°C. The supernate wasdialyzed against 10 mM potassium phosphate buffer, pH 6.0,containing 1 mM phenylmethylsulfonyl fluoride (PMSF;Sigma Chemical Co., St. Louis, Mo.) and 0.02% sodiumazide (Sigma). The dialysate was utilized for determiningextracellular enzyme activity. The cell pellet was quicklyfrozen in a dry-ice-ethanol bath and stored at -70°C. Fordetermination of CAT specific activity, the pellet was resus-pended in 1.0 ml of 10 mM Tris-HCl buffer, pH 7.8,containing 10 mM hexanoic acid (Sigma). A one-third vol-ume of glass beads (average diameter, 0.2 mm; BaxterScientific) was added, and the cells were disrupted byhomogenization with a Tissue Tearor cell homogenizer (Bio-spec Products Inc., Bartlesville, Okla.) at maximum speedfor 8 min with intermittent C02-ethanol cooling. The lysateswere centrifuged at 12,000 x g, and the supernate was usedto determine CAT-specific activities (see below).Enzymatic assays. The determination of CAT specific

activities was accomplished by the spectrophotometricmethod of Shaw (47), which measures the rate of acetylationof chloramphenicol. Coincident with the transfer of an acetylgroup from acetyl coenzyme A to chloramphenicol is thereaction of the reduced coenzyme A with 5,5'-dithio-bis-2-nitrobenzoic acid (DTNB). One unit of CAT activity isequivalent to 1 ,umol of chloramphenicol acetylated per min.

INFECT. IMMUN.

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

GENE REGULATION IN S. MUTAANS 1261

TABLE 1. Steady-state levels of CAT specific activity and extracellular synthesis activitya

CAT sp act at: Extracellular synthesis activity at:S. mutans strain pH

0.05 h-1 0.3 h-W 0.05 h-1 0.3 h'

SMS101 (ff-cat)7.0 0.128 ± 0.016 0.256 + 0.002 0.310 ± 0.120 0.063 ± 0.0286.0 0.346 ± 0.017 0.381 ± 0.013 0.030 ± 0.013 0.023 ± 0.010

SMS102 (gtfBC-cat)7.0 0.030 ± 0.002 0.020 ± 0.001 0.158 + 0.047 0.143 ± 0.0186.0 0.049 + 0.007 0.017 ± 0.008 0.144 + 0.013 0.122 ± 0.030

a Values are the averages for one of three separate chemostat runs. Activities are expressed as units per minute per milligram of protein ± the standarddeviations of triplicate samples. Steady-state dilution rates of 0.05 and 0.3 h-1 were reached after cultures were allowed to grow for 10 generations at each dilutionrate.

Protein concentrations were determined by the method ofBradford (4), by using a commercially available kit (Bio-Rad), with bovine serum albumin as the standard.

Culture supernates from SMS101 and SMS102 treated asdescribed above were assayed for FTF and GTF activity bya modification of the method of Robrish et al. (38) andGermaine et al. (14), respectively. Briefly, for FIF, theassay determined the incorporation of the fructose moietyfrom [fructose-1-3H(N)J-sucrose into methanol-insolublepolysaccharides. The enzyme preparations were incubatedwith an equal volume of the substrate mixture containing 10mM potassium phosphate buffer (pH 6.0), 0.02% sodiumazide, sucrose (200 mM), and 100 mM [fructose-1-3H(N)]-sucrose. After incubation for 4 h at 37°C, 1 ml of ice-coldmethanol was added. The precipitated polysaccharides werewashed with ice-cold methanol in a filter manifold (MilliporeCorporation, Bedford, Mass.) onto a 2.5-cm-diameter glassfiber filter (Whatman). The filters were dried and immersedin scintillation fluid (Ecoscint A; National Diagnostics, Man-ville, N.J.). The amount of radiolabelled sugar incorporatedinto the polysaccharides was measured by scintillationcounting. GTF assays were performed essentially as de-scribed for FTF, except that [glucose-14C(U)J-sucrose, 40,M dextran 9000, and 0.04 M Imidazole-HCl buffer (pH 6.5)were used in the substrate mixture instead of [fructose-l-3H(N)]-sucrose and phosphate buffer. One unit of GTF orFTF activity was defined as the amount of enzyme necessaryto incorporate 1 ,umol of hexose into methanol-insolublematerial in 1 min. All assays were performed in triplicate. Alldata were collected from three separate chemostat runs andare the averages of the results of triplicate assays. Proteinconcentrations were determined by the method of Bradford(4) with bovine serum albumin as the standard.

RESULTS

Steady-state growth experiments. Growth rate and pH havebeen previously demonstrated to modulate glucan and fruc-tan synthetic capabilities of S. mutans growing at steadystate (28, 56). To determine the influence that growth at low(0.05 h-i; doubling time, 13.8 h) and high (0.3 h-1; doublingtime, 2.3 h) dilution rates and environmental pH (pH = 6.0or 7.0) had on transcription from the gtf and ftf promotersand on exoenzyme activity, cultures were grown to steadystate at each set of conditions. The vessel was sampled, andCAT, GTF, and FTF specific activities were determined(Table 1). Since the gtfBC and ftf structural genes were notdisrupted in the construction of the fusion strains, it waspossible to determine FTF or GTF activity. In all cases, atsteady state, the organisms were growing under glucose-

limiting conditions. No residual glucose was detected in theculture supernates by using a glucose oxidase reaction(Sigma Chemical Co.) (data not shown), indicating that lessthan 5 ,M glucose remained in the vessel.For SMS101(ft-cat), transcription from the ftf promoter

as determined by CAT-specific activity (CAT-ftf) was notsignificantly different between dilution rates of 0.05 and 0.3h-1 when sampled at pH 6.0. However, when sampled at pH7.0, there was a twofold increase in CAT-ftf noted for cellsgrown at the faster growth rate. In all cases, CAT-ftf washigher in cells grown at pH 6.0 than at pH 7.0. No differenceswere observed in FTF activity as a function of the growthrate at pH 6.0. However, at pH 7.0 and D = 0.3 h-1,measurable extracellular FTF specific activity was somefivefold lower than that recorded for cells grown at pH 7.0andD = 0.05 h-1, despite the fact that CAT-fifwas doubledat the faster growth rate. Likewise, FTF specific activity wassignificantly higher in culture supernates of pH 7.0-growncells than in those of pH 6.0-grown cells, under circum-stances at which transcription from the fif promoter wasmuch greater for cells cultured at pH 6.0. Clearly, the culturepH and the specific growth rates of the organisms influencedtranscription rates from the ftf promoter, as well as pheno-typic expression of extracellular FTF activity. Further, theinfluence of growth rate on the cells occurred in a pH-dependent fashion.

In SMS102 (gtfB-cat), transcription from the gtfBC pro-moter as measured by CAT-specific activity (CAT-gtj) wasconsistently lower for cells grown at a dilution rate of 0.3 h-1than of 0.05 h-1 regardless of the culture pH. ExtracellularGTF specific activity differed only slightly between the twosteady-state levels and between pH values. Under all steady-state conditions, extracellular GTF specific activity was notsignificantly diminished at the higher dilution rates despitemeasurable decreases in CAT-gtf.CAT fusion and extracellular enzyme activities during tran-

sition states between steady-state growth domains. The levelsof CAT and polymer synthesis activities from S. mutansSMS101 and SMS102 were determined during a shift up to ahigh growth rate (steady-state D = 0.05 h-1 to D = 0.4 h-1)and a shift down to a low growth rate (steady-state D = 0.3h-1 to D = 0.05 h-1) at a constant pH value of either 6.0 or7.0. Cultures could not be maintained without washout at D= 0.4 h-1, likely because of the glucose limitation. Re-sponses by the organisms to alterations in the dilution ratesoccurred for both fusion strains, and the patterns of CAT-and exoenzyme-specific activities observed for each strainduring the transition states differed as a function of theculture pH. It should be noted that during the shift in growth

VOL. 61, 1993

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

1262 WEXLER ET AL.

A

0

B

U-I-

C-

C.)

0

Time (h)Time (h)

LL

2

FIG. 2. CAT and extracellular FTF specific activities as a measure of the expression of the S. mutansfgene at pH 7.0 during transitionperiods between different growth rates. (A) Shift from a 0.05 h-' steady-state growth rate to 0.4 h-1 at arrow; (B) shift from a 0.3 h-1steady-state growth rate to 0.05 h-1 at arrow. Values are the averages of triplicate assays from one of three independent chemostat runs andare expressed as specific activity (units per milligram of Bradford protein) as described in Materials and Methods. Error bars representstandard deviations.

rate from D = 0.05 h-1 to 0.4 h-', the cultures were notglucose limited.At pH 7.0, CAT-ftf increased dramatically with the in-

crease in the dilution rate in the vessel (Fig. 2A) afterapproximately 1 h following the shift. Extracellular FTFactivity, on the other hand, decreased dramatically duringthis period (Fig. 2A), consistent with the lower activity notedfor cells growing at steady state at pH 7.0, D = 0.3 h-1compared withD = 0.05 h'- (Table 1). It seems unlikely thatthis phenomenon was simply due to enzyme washout, sincespecific activities and not units of activity per milliliter wererecorded. Also, there was not sufficient dilution of theculture during the transition to account for the decrease inspecific activity. Moreover, the decrease occurred in adiscontinuous manner, with most of the reduction during thefirst hour of the transition.

Following reequilibration of the vessel and establishmentof steady-state growth at D = 0.3 h-1 (10 generations), thedilution rate was shifted from 0.3 h-1 to 0.05 h-1 (Fig. 2B).Following the shift down, an increase in CAT-ftf and extra-

A

I-

0.

cellular FTF specific activities was noted, with a subsequentdecline in both activities after 30 min.When these experiments were performed with SMS101 at

pH 6.0 (Fig. 3), a pattern different from that for pH 7.0 wasobserved. Following shift up (Fig. 3A), a decrease in CAT-ftfand a small decrease in extracellular FTF specific activitywere noted. After the vessel had equilibrated and cells hadbeen cultured for 10 generations at D = 0.3 h-1 and pH 6.0,the dilution rate was reduced to 0.05 h-1 and the vessel wassampled (Fig. 3B). In this case, slight decreases in CAT-ftfand FTF specific activities were observed, and no additionaldecrease following 1 h at the new dilution rate was seen.The identical manipulations of the growth conditions for a

transition-state culture were performed with S. mutansSMS102(gtC-cat), and CAT-gtf and extracellular GTFspecific activities were measured (Fig. 4 and 5). At pH 7.0,an increase in GTF-specific activity was noted upon shift up,followed by a decline in this activity within 1 h (Fig. 4A).This increase in GTF activity was not accompanied by acoincident increase in CAT-gtf, but rather a steady decline in

B

L C<0-.

0.4

0.3-A* CAT

-0---U- FTF

pH-6.0

0.2 _

0.1

0.0 A I I

O0 1 2.

3

0.15

U-0.10 1-.

U.

0.05

Time (h) Time (h)FIG. 3. CAT and extracellular FTF specific activities as a measure of the expression of the S. mutansfi gene at pH 6.0 during transition

periods between different growth rates. (A) Shift from a 0.05 h-1 steady-state growth rate to 0.4 h-' at arrow; (B) shift from a 0.3 h-1steady-state growth rate to 0.05 h-1 at arrow. Values are expressed as described in the legend to Fig. 2.

INFECT. IMMUN.

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

GENE REGULATION IN S. MUTANS 1263

A

Cl-

B o.lo

IL I

a 00.05

0.00 i2

10.1

3

U-

a-

Time (h) Time (h)FIG. 4. CAT and extracellular GTF activities as a measure of the expression of the S. mutans gtfBC gene at pH 7.0 during transition

periods between different growth rates. (A) Shift from a 0.05 h-1 steady-state growth rate to 0.4 h-1 at arrow; (B) shift from a 0.3 h-steady-state growth rate to 0.05 h-1 at arrow. Values are expressed as described in the legend to Fig. 2.

CAT-gtf activity was noted during the shift up. After steadystate was achieved at D = 0.3 h-1, the dilution rate wasadjusted to 0.05 h-1 (Fig. 4B). This shift was accompaniedby only slight changes in CAT-gtf and glucan synthesisactivities.At pH 6.0, a decline in both CAT-gtf and GTF specific

activities was noted when the dilution rate was increased(Fig. SA). Alternatively, adjustment of the dilution rate from0.3 to 0.05 h-' was accompanied by an increase in assayableGTF specific activity but with little change in the expressionfrom the CAT fusion (Fig. SB). GTF specific activity de-clined to steady-state levels within 2 h of the shift to 0.05h-1. Again, the values measured for CAT and GTF ap-proached those recorded for steady state-grown cells fairlyrapidly (Table 1).

Responses of low-growth-rate cultures to sucrose. The ef-fects of the addition of sucrose to cultures of glucose-limitedS. mutans growing at a low growth rate on transcriptionalfusion activity and on polymer synthesis were examined asdetailed in Materials and Methods. The addition of sucrose(final concentration, 20 mM) to S. mutans SMS1O1(ftf-cat)resulted in a rapid increase in CAT-ftf of approximately 50%for 40 min after the introduction of sucrose (Fig. 6). Thisincrease was followed by a slow decline in CAT-ftf. In

A

I-

0.10

0.05

2

contrast, only a slight increase in FTF specific activityoccurred for 30 min after the sucrose addition, followed by adecline in the assayable activity.The baseline levels of CAT-gtf were significantly lower

than those expressed with theftf-cat fusion strain, consistentwith observations for steady-state low-growth-rate cultures(Table 1). The addition of sucrose to slowly growing glucose-limited SMS102(gtfBC-cat) cultures resulted in a threefoldincrease in expression from thegtfBC promoter as measuredby CAT-gtf (Fig. 7). There was a phenotypic lag in expres-sion of extracellular GTE specific activity after the sucroseaddition for at least 10 min, during which there was actuallya slight but consistently repeatable decrease in assayableglucan synthesis, perhaps reflecting the aggregation of anexisting supernatant enzyme by nascent dextrans. The lagperiod was followed by an increase of approximately 40% inextracellular GTF activity over 90-min. Unlike that for FTF,no decrease was seen in extracellular GTF specific activityduring these experiments.

Stability of CAT fusions. A single chemostat run for eachfusion strain lasted 4 to 5 weeks, approximately 60 to 80generations. The effect that long-term continuous culturewould have on the stability of the fusions was not known,even though erythromycin was included in the culture me-

B 0.3

0.150.2

0.10 1 oC

0.1

0.05

0.00.003

Time (h) Time (h)

FIG. 5. CAT and extracellular GTE specific activities as a measure of the expression of the S. mutans gtfBC gene at pH 6.0 duringtransition periods between different growth rates. (A) Shift from a 0.05 h-' steady-state growth rate to 0.4 h-1 at arrow; (B) shift from a 0.3h-1 steady-state growth rate to 0.05 h-1 at arrow. Values are expressed as described in the legend to Fig. 2.

I pH-7.0

-@ - CAT

-0- GTF

+ '+~~~~

I pH-6.0* CAT

GTF:~A I I

VOL. 61, 1993

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

1264 WEXLER ET AL.

I~-

0.4-

0.3-

0.2

0.1

T01 20 40 80 100

Time(m in.)

FIG. 6. CAT and extracellular FTF specific activities as a mea-sure of the expression of the S. mutans f gene after sucrose pulseof 0.75 M sucrose for a final sucrose concentration of 20 mM. Theaddition of Sucrose is indicated by the arrow. Values are expressedas described in the legend to Fig. 2.

dium for the duration of the experiments. Following thecompletion of a chemostat run, organisms were recoveredfrom the vessel, briefly sonicated, and plated on BHI andBHI plus erythromycin Em to ascertain whether the fusionswere lost at an appreciable frequency from the chromosomesof the test strains. Lower recoveries of organisms on eryth-romycin-containing medium would reflect loss of the inte-grated plasmid harboring the fusion and proximal promoter.There were no differences between the number of CFU oforganisms on BHI and on BHI plus erythromycin (data notshown), suggesting that the marker was not lost at a appre-ciable frequency. The retention of CAT and erythromycinmarkers in the test strains as determined by assay andplating, respectively, was also considered to be indicative ofstable maintenance of the fusion. Most notably, CAT ex-

pression levels from cells obtained from the final samplingsof each experiment were comparable to those observed atthe first steady-state time point and to those recorded by

0.20 0.15pH-6-0

0.15

0.10

0.10

-l/* CAT

GTF0.05

0.05

0.00 * ' * ' * * . . 0.000*ooOi- 20 40 60 80 1 000

Time (min.)FIG. 7. CAT and extracellular GTF specific activities as a mea-

sure of the expression of the S. mutans gtJTC gene after sucrose

pulse of 0.75 M sucrose for a final sucrose concentration of 20 mM.Values are as expressed in Fig. 2. The addition of sucrose isindicated by the arrow.

Hudson and Curtiss (24) in batch culture. Had a largeproportion of the cells lost the fusion during the course of theexperiments, then a significant decrease in total CAT spe-cific activity would be predicted. This was never observed.

DISCUSSION

Microorganisms in the human mouth are subjected toradical and periodic changes in their environment. Becauseof intermittent eating patterns, vast differences in salivaquantity and composition associated with diurnal rhythms,mechanical forces, and specific and nonspecific host de-fenses, bacteria in plaque are continuously forced to adaptrapidly and efficiently to constantly changing conditions. S.mutans can persistently colonize a human host and, incertain instances, initiate dental caries (17) despite theseenvironmental insults. The ability of S. mutans to modulatethe expression of a variety of virulence determinants, includ-ing glucan and fructan metabolic machinery (26, 56), sugartransport systems (19), adhesins (42, 45), and aciduricity (2,20), must be essential for the survival of this bacterium in theoral cavity. In order to understand how this organismpersists and initiates disease and to devise novel and effec-tive strategies to interfere with the disease process, it will benecessary to describe the phenotypic capabilities of thisbacterium and to understand the molecular strategies under-lying differential expression of its virulence factors.For a long time, it was felt that the expression of the GTFs

and FTF of S. mutans was constitutive and was not gov-erned by complex regulatory circuits. However, studiesperformed by Walker et al. (56), Keevil et al. (28), andJacques et al. (26) have clearly demonstrated that the pro-duction of exoenzymes of oral streptococci participating inthe synthesis or breakdown of complex polysaccharides ismarkedly influenced by environmental conditions. Thesestudies examined total glucan synthesis or structure but didnot quantify individual enzymes. The data derived fromexamining steady-state cultures for GTF and FTF enzymeactivity and CAT-gtf and -ft specific activities indicateseveral key points regarding the regulatory signals control-ling the expression of the gtf andftf genes.Environmental pH and the specific growth rate of the

organisms dramatically influence transcriptional initiationfrom the ftf promoter and fructan synthesis capabilities.However, the influence of growth rate on ftf transcriptionoccurs in a pH-dependent fashion. At pH 6.0, growth rateappears to have little influence on CAT-ftf activity or onexpression of extracellular FTF activity. On the other hand,the growth rate at pH 7.0 influences transcriptional activityand the amount of enzyme activity which can be determinedin culture supernates. In this case, FTF activity and CAT-ftdid not correlate well. In large part, this was due to thepresence of greater amounts of total extracellular protein infaster-growth-rate cultures (data not shown), which resultedin an apparent decrease in FTF specific activity. However,posttranscriptional or posttranslational events may also con-trol phenotypic expression of extracellular FTF, such as therelease of cell-bound FTF (25). Alternatively, increases inexpression of fructan hydrolase at pH 7.0 and D = 0.3 h-may have resulted in the degradation of product fructans,although, in our experience, fructanase appears to have verylittle influence on FTF activity determined in vitro (57). Theinfluence of pH and growth rate on Nf gene transcription isclear, but a more detailed study will be necessary to ascer-tain the molecular basis for the posttranscriptional control ofFTF expression.

| pH-6.0

6~A*l- CAT

-0-- FrF

INFECT. IMMUN.

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

GENE REGULATION IN S. MUTANS 1265

Under steady-state conditions, transcriptional activityfrom gtfBC-cat or total glucan synthesis is modestly influ-enced by the culture pH values utilized in this study or thegrowth rate of the organisms. Decreases in CAT-gtf activityare associated with increased growth rate. Although enzymeactivity changes mimic changes in CAT-gtf, decreases asso-ciated with increased growth rate are not reflected in con-comitant decreases of an identical magnitude in total GTFactivity. This suggests that GTF-S activity (gtD) may beregulated in a manner opposite that of gfBC. Alternatively,dextranase may be regulated in a manner consistent withexpression of GtfB and GtfC, and thus decreased productbreakdown in reactions with supernatants from highergrowth rates would be consistent. A final possibility, whichwill require much more-extensive investigation is the controlof expression of the GtfC or GtfB proteins by posttranscrip-tional events, particularly at the translational level (22).Changes in the nutrient feed rate appear to elicit major

changes in transcriptional activity and extracellular enzymeactivity. In the oral cavity, the availability of nutrients todental plaque varies considerably as a result of dietary intakeof foodstuffs. For example, glucose concentrations in plaquecan quickly change from 10 ,M to 10 mM with ingestion ofsweetened meals (8). Clearly, S. mutans has evolved mech-anisms to adapt to these and other environmental challenges.By increasing and decreasing the nutrient feed rate in thechemostat, organisms are faced with new growth parametersand are forced to respond to these changes by altering theirphysiology through modulation of gene expression patternsand allosteric control of enzyme reactions (10). The dataderived from experiments examining transition-state expres-sion of GTF and FEF plainly demonstrate that S. mutans iscapable of responding rapidly to these new conditions byaltering transcription from both thegfBC andftfpromoters.Moreover, the responses to the shift, as under steady-stateconditions, occurred in a pH-dependent fashion, which mayallow for differential expression of these gene productsunder ecologically disparate situations, e.g., those faced asan early colonizer of the tooth surface versus those whichoccur in mature plaque composed of a strongly acidogenic-aciduric microflora. These data strongly support that there isa hierarchy of control for expression of FTF and GTF,probably involving multiple variables. Another importantpoint to be extrapolated from the transition data is that thelevels of transcription from the fusions and the levels ofenzyme activity return to the respective steady-state levelsrelatively quickly, usually within 2 h. This observationindicates that regulation of expression, and not selection ofsubpopulations within the chemostat which may be bettersuited to growth under specific conditions, is responsible forthe differences noted in CAT and extracellular enzymeactivity.

It should be noted that transition states, as affected byalteration in the dilution rate of the chemostat vessel, do notsimply reflect changing the growth rate of the organisms.With a change in the growth rate, there are changes in thelimiting nutrient (i.e., glucose limited to glucose excess),changes in potassium ion concentration, and a change in theconcentration of available amino acids, vitamins, and othernutrients. Thus, the signals received by the cell are likelycomplex and involve overlapping regulons. In our experi-ments, a shift of the cells to a high growth rate resulted in aglucose-excess condition in the vessel during the transition,which is known to influence levels of phosphorylated com-ponents of the PTS system (52), activity and levels ofglycolytic enzymes (7), and other regulatory circuits which

could directly or indirectly affect GTF and FTF transcrip-tion, translation, and secretion. The use of transcriptionalfusions in conjunction with mutants, such as E"g'-defectiveS. mutans (21), with strains carrying mutations in the genesencoding the putative DNA-binding proteins associated withtheftfgene (44), or the use of strains with targeted mutationsin the promoter regions of ff or gfBC should facilitatedefinition of the cis- and trans-acting regulatory componentsgoverning the differential expression of these genes in S.mutans in response to environmental stimuli.

Recent experimentation with batch-grown S. mutans cellshas indicated that the synthesis of GTFs and F-TE is regu-lated by the addition of sucrose, and twofold increases inenzyme synthesis have been recorded (24). Our data with thegfBC-cat strain indicate that there is at least a threefoldincrease in activity from thegtfBC promoter when sucrose isadded to slowly growing glucose-limited S. mutans cells.This observation may be significant in terms of the responseof S. mutans to sucrose in dental plaque and may reflectmechanisms which have evolved in this bacterium to adaptquickly to the addition of sucrose in order to synthesize largequantities of water-insoluble glucans.The fact that the large increase in CAT-gtf was not

reflected in a similar magnitude of increase in the amount ofglucan synthesized by culture supernates of cells induced bysucrose may be a result of multiple variables affecting theamounts of glucans synthesized during in vitro assays. Oneexplanation for the disparity between CAT-gtf and GTFspecific activities may be that in this study, only extracellu-lar enzyme was assayed. It is known that when sucrose ispresent in the culture medium, a large increase in theproportion of GTF is found in a cell-associated form, pre-sumably as a result of a complex of glucan-binding proteins,glucans, and GTF developing at the cell surface (17, 41, 50).Alternatively, the proportionately smaller increase in GTFspecific activity to CAT-g may reflect down-regulation ofthe gtfD gene or other posttranscriptional events controllingGTF secretion or activity, which could result in a lower netyield of glucans. The attack of product glucans at a(1,6)linkages by endodextranase (54) may also have reduced theamount of glucan measured as precipitable material.The data obtained in this study indicate that the operon

fusions were stably maintained throughout the course of theexperiments. The fusions were constructed by juxtaposing afragment containing the 5' region of the gtfBC gene or ftfgenes to a promoterless cat gene and integrating this con-struction by an insertion-duplication mechanism into the S.mutans UA130 chromosome. Because of the mechanism bywhich the foreign DNA was inserted, it was possible that thefusion could be excised from the chromosome and be lostfrom the strain by segregation. The inclusion of erythromy-cin in the culture medium provided selective pressure tomaintain the construction in S. mutans. Although the pres-ence of erythromycin in the medium might have a minoreffect on cell physiology, the experiments were confined tointernal comparison of single strains under differing condi-tions. It is also noteworthy that the presence of an erythro-mycin determinant in S. mutans does not affect manyphysiologically measurable parameters in vitro (29, 57), nordoes it seem to affect the ability of strains to colonize andelicit caries (23, 46, 57). Therefore, the use of erythromycindoes not seem to be a disadvantage. Perhaps more impor-tantly, insertion by Campbell model recombination, as wasdone for the strains utilized in this study, allows for con-structing strains without insertionally inactivating the parentgene. During the course of these experiments, the fusion

VOL. 61, 1993

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

1266 WEXLER ET AL.

strains likely had the duplicated promoter element in unit ornear unit copy per chromosomal equivalent. Therefore, it isunlikely that the additional promoter element(s) had an im-pact on gene expression through titration of positive ornegative trans-acting regulatory elements. The proximity ofthe CAT fusions to the cognate gfBC andftfpromoters mayhave had some influence on transcription from these geneswhich was not readily apparent. Unfortunately, the alterna-tive to these experiments would be to construct merodiploidsin a recombination-deficient background or to integrate thefusion by double-crossover recombination, thereby inactivat-ing the parent gene. For the former, such strains wouldpotentially have inherent problems arising from titration oftrans-acting regulatory elements, effects on cell physiologyfrom harboring plasmid DNA, and physiologic defects arisingfrom deficiency in the RecA-like enzyme of S. mutans, all ofwhich could render these strains less fit than the wild type.For the second alternative, insertion by a double-crossovermechanism could inactivate the enzyme being studied orcreate polar effects which would prohibit assaying the geneproduct or otherwise confound the data. Therefore, the use ofinsertion-duplication seems a reasonable approach to estab-lishing operon or gene fusions in strains of S. mutans.

In vivo, S. mutans is subjected to wide variations in sugarconcentration and pH, and it will be of interest to study theeffects of lower pH values and alternative carbon sources orconcentrations on gene expression. Clearly, the gene fusion-chemostat technique presented here is applicable to studiesexamining a wide variety of culture conditions. Notwithstand-ing, the principal purposes of this investigation were todetermine (i) whether gene fusions could be utilized stably incontinuous culture to study transcriptional regulation of geneexpression in S. mutans and potentially in other oral strepto-cocci, (ii) whether transcription of GTFs and FTF wasregulated by pH and/or growth rate, and (iii) whether GTFand FTF synthesis was inducible by sucrose in cells whichwere C limited and growing slowly, as in dental plaque. ThepH values and growth rates chosen here were selected on thebasis that pH 6.0 and 7.0 are physiologically relevant valuesfor the oral streptococci. Also, previous chemostat studieshave successfully demonstrated phenotypic changes in S.mutans in response to environmental stimuli utilizing similarculture conditions. This study demonstrated the utility of thetechnology and the influence of pH, growth rate, and sucroseon GTF and FTF expression and can now be expanded toexperiments designed to study the influence of extreme pHvalues, carbon sources, and antimicrobial agents or othercompounds on gene expression. It also seems feasible to usemixed cultures of oral bacteria to ascertain what influencecoculture of organisms has on gene expression.

Reporter gene fusions in conjunction with continuous cul-ture represent a novel approach in the investigation of generegulation in the oral streptococci. The data presented heredemonstrate that many of the observed phenotypic changesresult from differential control of transcription of GTF andFTF and that posttranslational events are key control pointsin expression of GTF and FTF activities. Since the ability ofS. mutans to modulate expression of virulence determinantsunder various environmental conditions is likely crucial to theorganism's survival in the oral cavity, the dissection ofregulation of individual genes in response to conditions com-monly encountered in the mouth will be essential if thepathogenic strategies of this organism are to be understood.Likewise, such studies should provide insight into generalmolecular mechanisms governing pH, growth rate, and car-bon source control of gene expression in prokaryotes. More-

over, this technique will be broadly applicable to a variety ofvirulence-related genes in oral streptococci and other oralpathogens. In combination with biochemical and immuno-logic techniques and experiments examining mRNA levels,analysis of the regulation of transcription of virulence deter-minants of oral bacteria should result in a more completepicture of the strategies utilized by oral pathogens to persist,to emerge within a plaque community, and to initiate disease.

ACKNOWLEDGMENTS

We thank R. E. Marquis for critical assessment of the manuscript.We acknowledge the technical assistance of Jana E. C. Penders andthank Hilary Fraser for assistance in the preparation of the manuscript.

This study was supported by Public Health Service grantsDE09878, DE07003, and DE07165.

REFERENCES1. Aold, H., T. Shiroza, M. Hayakawa, S. Sato, and H. K.

Kuramitsu. 1986. Cloning of a Streptococcus mutans glucosyl-transferase gene coding for insoluble glucan synthesis. Infect.Immun. 53:587-594.

2. Belli, W. A., and R. E. Marquis. 1991. Adaptation of Strepto-coccus mutans and Enterococcus hirae to acid stress in contin-uous culture. Appl. Environ. Microbiol. 57:1134-1138.

3. Birkhed, D., K.-G. Rosell, and K. Granath. 1979. Structure ofextracellular water-soluble polysaccharides synthesized fromsucrose by oral strains of Streptococcus mutans, Streptococcussalivarius, Streptococcus sanguis, and Actinomyces viscosus.Arch. Oral Biol. 24:53-61.

4. Bradford, M. M. 1976. A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal. Biochem. 72:248-254.

5. Burne, R. A. Oral ecological disasters: the role of short-termextracellular storage polysaccharides. In W. H. Bowen andL. A. Tabak (ed.), Cariology of the Nineties Conference, inpress. University of Rochester Press, Rochester, N.Y.

6. Burne, R. A., K. Schilling, W. H. Bowen, and R. E. Yasbin.1987. Expression, purification, and characterization of an exo-,-D-fructosidase of Streptococcus mutans. J. Bacteriol. 169:4507-4517.

7. Carlsson, J., and C. J. Griffith. 1974. Fermentation products andbacterial yields in glucose-limited and nitrogen-limited culturesof streptococci. Arch. Oral Biol. 19:1105-1109.

8. Carlsson, J., and B. Sundstrom. 1968. Variations in compositionof early dental plaque following ingestion of sucrose and glu-cose. Odontol. Rev. 19:161-169.

9. DaCosta, T., and R. J. Gibbons. 1968. Hydrolysis of levan byhuman plaque streptococci. Arch. Oral Biol. 13:609-617.

10. Dykhuizen, D. E., and D. L. Hartl. 1983. Selection in chemo-stats. Microbiol. Rev. 47:150-168.

11. Ebisu, S., K. Keiiro, S. Kotani, and A. Misaka. 1975. Structuraldifferences in the fructans elaborated by Streptococcus mutansand Streptococcus salivarius. J. Biochem. 78:879-887.

12. Ellwood, D. C. 1976. Chemostat studies of oral bacteria. Micro-biol. Abstr. 3(Spec. Suppl.):785-798.

13. Freedman, M. L., and J. M. Tanzer. 1982. Use of mutants tostudy the glucan-associated pathophysiology of Streptococcusmutans, p. 186-190. In D. Schlessinger (ed.), Microbiology-1982. American Society for Microbiology, Washington, D.C.

14. Germaine, G. R., C. F. Schachtele, and A. M. Chludzinski. 1974.Rapid filter paper assay for the dextransucrase activity ofStreptococcus mutans. J. Dent. Res. 53:1355-1360.

15. Gibbons, R. J., and J. van Houte. 1975. Bacterial adherence inoral microbial ecology. Annu. Rev. Microbiol. 29:19-44.

16. Gold, W., F. B. Preston, M. C. Lache, and H. Blechman. 1974.Production of levan and dextran in plaque in vivo. J. Dent. Res.53:442-449.

17. Hamada, S., and H. D. Slade. 1978. Effect of sucrose in culturemedia on the location of glucosyltransferase of Streptococcusmutans and cell adherence to glass surfaces. Infect. Immun.20:592-599.

INFECT. IMMUN.

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from

GENE REGULATION IN S. MUTANS 1267

18. Hamada, S., and H. D. Slade. 1980. Biology, immunology, andcariogenicity of Streptococcus mutans. Microbiol. Rev. 44:331-384.

19. Hamilton, I. R. 1987. Effect of changing environment on sugartransport and metabolism by oral bacteria, p. 94-133. In J.Reizer and A. Peterkofsky (ed.), Sugar transport and metabo-lism by gram-positive bacteria. Ellis Horwood, Chichester,England.

20. Hamilton, I. R., and N. D. Buckley. 1991. Adaptation byStreptococcus mutans to acid tolerance. Oral Microbiol. Immu-nol. 6:65-71.

21. Hamilton, I. R., and E. J. St. Martin. 1982. Evidence for theinvolvement of proton motive force in the transport of glucoseby a mutant of Streptococcus mutans strain DR0001 defective inglucose-phosphoenolpyruvate phosphotransferase activity. In-fect. Immun. 36:567-575.

22. Hanada, N., and H. K. Kuramitsu. 1988. Isolation and charac-terization of the Streptococcus mutans gtfC gene, coding forsynthesis of both soluble and insoluble glucans. Infect. Immun.56:1999-2005.

23. Hanada, N., and H. K. Kuramitsu. 1989. Isolation and character-ization of the Streptococcus mutans gtJD gene, coding for primer-dependent soluble glucan synthesis. Infect. Immun. 57:2079-2085.

24. Hudson, M. C., and R. Curtiss III. 1990. Regulation of expres-sion of Streptococcus mutans genes important to virulence.Infect. Immun. 58:464-470.

25. Jacques, N., and C. Wittenberger. 1981. Inactivation of cell-associated fructosyltransferase in Streptococcus salivarius. J.Bacteriol. 148:912-918.

26. Jacques, N. J., J. G. Morrey-Jones, and G. J. Walker. 1985.Inducible and constitutive production of fructanase in batch andcontinuous culture of Streptococcus mutans. J. Gen. Microbiol.131:1625-1633.

27. Keevil, C. W., P. D. Marsh, and D. C. Ellwood. 1984. Regulationof glucose metabolism in oral streptococci through independentpathways of glucose 6-phosphate and glucose 1-phosphate. J.Bacteriol. 157:560-567.

28. Keevil, C. W., A. A. West, P. D. Marsh, and D. C. Ellwood.1983. Batch versus continuous culture studies of glucosyltrans-ferase synthesis in oral streptococci. Chem. Senses 1983(Spec.Suppl.):189-200.

29. Kuramitsu, H. K. 1987. Utilization of a mini-mu transposon toconstruct defined mutants in Streptococcus mutans. Mol. Mi-crobiol. 1:229-231.

30. Kuramitsu, H. K. (University of Texas at San Antonio). 1992.Personal communication.

31. Kuramitsu, H., and L. Ingersoll. 1976. Immunological relation-ships between glucosyltransferases from Streptococcus mutansserotypes. Infect. Immun. 14:636-644.

32. Larrimore, S. H., H. Murchison, T. Shiota, S. M. Michalek, andR. Curtiss HI. 1983. In vitro and in vivo complementation ofStreptococcus mutans mutants defective in adherence. Infect.Immun. 42:558-566.

33. Manly, R. S., and D. T. Richardson. 1968. Metabolism of levanby oral samples. J. Dent. Res. 47:1080-1086.

34. Milward, C., and N. Jacques. 1990. Secretion of fructosyltrans-ferase by Streptococcus salivanus involves the sucrose-dependentrelease of the cell-bound form. J. Gen. Microbiol. 136:165-169.

35. Mukasa, H., H. Tsumori, and A. Shimamura. 1985. Isolationand characterization of an extracellular glucosyltransferase syn-thesizing insoluble glucan from Streptococcus mutans serotypec. Infect. Immun. 49:790-796.

36. Munro, C., S. M. Michalek, and F. L. Macrina. 1991. Carioge-nicity of Streptococcus mutans V403 glucosyltransferase andfructosyltransferase mutants constructed by allelic exchange.Infect. Immun. 59:2316-2323.

37. Perry, D., and H. K. Kuramitsu. 1990. Linkage of sucrose-metabolizing genes in Streptococcus mutans. Infect. Immun.58:3462-3464.

38. Robrish, S. A., W. Reid, and M. Krichevsky. 1972. Distributionof enzymes forming polysaccharides from sucrose and thecomposition of extracellular polysaccharide synthesized byStreptococcus mutans. Appl. Microbiol. 24:184-190.

39. Rolla, G., J. E. Ciardi, K. E. Eggen, W. H. Bowen, and J.Afseth. 1983. Free glucosyl- and fructosyltransferase in humansaliva and adsorption of these enzymes to teeth in vivo. Chem.Senses 1983(Spec. Suppl.):21-29.

40. Rosan, B., B. Appelbaum, L. K. Campbell, K. W. Knox, andA. J. Wicken. 1982. Chemostat studies of the effect of environ-mental control on Streptococcus sanguis adherence to hy-droxyapatite. Infect. Immun. 35:64-70.

41. Russell, R. R. B. 1979. Wall-associated protein antigens ofStreptococcus mutans. J. Gen. Microbiol. 114:109-115.

42. Russell, R. R. B., E. Abdulla, M. L. Gilpin, and K. Smith. 1986.Characterization of Streptococcus mutans surface antigens, p.61-70. In S. D. Hamada, L. Menaker, H. Kiyono, S. Michalek,and J. R. McGhee (ed.), Cellular, molecular, and clinical aspects ofStreptococcus mutans. Elsevier Science Publishers, Amsterdam.

43. Russell, R. R. B., and M. L. Gilpin. 1987. Identification ofvirulence components of mutans streptococci, p. 201-204. InJ. J. Ferretti and R. Curtiss III (ed.), Streptococcal genetics.American Society for Microbiology, Washington, D.C.

44. Sato, S., S. Ueda, and H. K. Kuramitsu. 1987. Construction ofStreptococcus mutans glucosyltransferase mutants utilizing acloned gene fragment. FEMS Microbiol. Lett. 48:207-210.

45. Schilling, K. M., and W. H. Bowen. 1992. Glucans synthesized insitu in experimental salivary pellicle function as specific bindingsites for Streptococcus mutans. Infect. Immun. 60:284-295.

46. Schroeder, V. A., S. M. Michalek, and F. L. Macrina. 1989.Biochemical characterization and evaluation of virulence of afructosyltransferase-deficient mutant of Streptococcus mutansV403. Infect. Immun. 57:3560-3569.

47. Shaw, W.V. 1979. Chloramphenicol acetyltransferase from chlor-amphenicol-resistant bacteria. Methods Enzymol. 43:737-755.

48. Smith, D. J., and M. A. Taubman. 1977. Antigenic relatednessof glucosyltransferase enzymes from Streptococcus mutans.Infect. Immun. 15:91-103.

49. Takamori, K., F. Mizuno, A. Yamamoto, Y. Etoh, M. Taka-hashi, and N. Takahashi. 1985. Preliminary studies of fructan-hydrolyzing bacteria from human dental plaque. Microbiol.Immunol. 29:359-363.

50. Townsend-Lawman, P., and A. S. Bleiweis. 1991. Multilevelcontrol of extracellular sucrose metabolism in Streptococcussalivarius by sucrose. J. Gen. Microbiol. 137:5-13.

51. Ueda, S., T. Shiroza, and H. K. Kuramitsu. 1988. Sequenceanalysis of the gtf C gene from Streptococcus mutans GS-5.Gene 69:101-109.

52. Vadeboncoeur, C., L. Thibault, S. Neron, H. Halvorson, and I.Hamilton. 1987. Effect of growth conditions on the level of thecomponents of the phosphoenolpyruvate:sugar phosphotrans-ferase system in Streptococcus mutans and Streptococcus so-brinus grown in continuous culture. J. Bacteriol. 169:5686-5691.

53. van Houte, J., and H. M. Jansen. 1968. Levan degradation bystreptococci isolated from human dental plaque. Arch. OralBiol. 13:827-830.

54. Walker, G. J., M. D. Hare, and J. G. Morrey-Jones. 1982. Effectof variation in growth conditions on endo-dextranase produc-tion by Streptococcus mutans. Carbohydr. Res. 107:111-122.

55. Walker, G. J., V. L. Jacques, and E. Fiala-Beer. 1992. Influenceof the culture medium on the synthesis of a-D-glucans byStreptococcus cricetus AHT. Carbohydr. Res. 227:1-17.

56. Walker, G. J., J. G. Morrey-Jones, S. Svensson, and C. Taylor.1983. Effect of variation in growth conditions on the activity ofD-glucosyltransferases on the synthesis of a-D-glucans by Strep-tococcus mutans OMZ176. Chem. Senses 1983(Spec. Suppl.):179-187.

56a.Wexler, D. L., and R. A. Burne. 1992. Regulation of expressionof genes contributing to virulence of Streptococcus mutans, p.205, H-136. Abstr. 92nd Gen. Meet. Am. Soc. Microbiol. 1992.American Society for Microbiology, Washington, D.C.

57. Wexler, D. L., J. E. C. Penders, W. H. Bowen, and R. A. Burne.1992. Characteristics and cariogenicity of a fructanase-defectiveStreptococcus mutans strain. Infect. Immun. 60:3673-3681.

58. Wood, J. M. 1967. The amount, distribution, and metabolism ofsoluble polysacharides in human dental plaque. Arch. Oral Biol.12:849-858.

VOL. 61, 1993

on January 11, 2021 by guesthttp://iai.asm

.org/D

ownloaded from