Dental caries results from the interaction ofspecific bacteria with constituents of thediet within dental biofilms known asplaque. Sucrose is considered to be the‘arch criminal’ from the dietary perspec-tive, because it is fermentable, and servesas a substrate for the synthesis of polysac-charides in dental biofilms (3, 27). Inaddition, starches are also an importantsource of fermentable carbohydrates, and

are usually consumed simultaneously orinterspersed with sucrose (24); starch isconsidered non-cariogenic or slightly car-iogenic when used as the sole source ofcarbohydrate in the diet (24). However,combinations of starch and sucrose arepotentially more cariogenic than eithercarbohydrate alone (2, 13, 29).Streptococcus mutans is regarded as the

primary microbial culprit of dental caries;

this bacterium synthesizes extracellularpolysaccharides, mostly glucans, fromsucrose (and may also use starch hydroly-sates as acceptors); it is acidogenic andacid-tolerant, which are critical virulenceproperties involved in the pathogenesisof dental caries in animals and humans(25, 28, 33, 37). Glucans promote theaccumulation of microorganisms onthe tooth surface, and contribute to the

Oral Microbiology Immunology 2008: 23: 206–212Printed in Singapore. All rights reserved

� 2008 The Authors.Journal compilation � 2008 Blackwell Munksgaard

Influences of starch and sucroseon Streptococcus mutansbiofilmsDuarte S, Klein MI, Aires CP, Cury JA, Bowen WH, Koo H. Influences of starch andsucrose on Streptococcus mutans biofilms.Oral Microbiol Immunol 2008: 23: 206–212. � 2008 The Authors. Journal compilation� 2008 Blackwell Munksgaard.

Introduction: The combination of starch and sucrose has been shown to be potentiallymore cariogenic than either alone. The aim of this study was to examine the influence ofstarch and sucrose, alone or in combinations, on formation, polysaccharide composition,gene expression, and acidogenicity of Streptococcus mutans biofilms.Methods: S. mutans UA159 biofilms were formed on saliva-coated hydroxyapatite (sHA)discs in batch culture for 5 days in the presence of 1% (weight/volume) starch, 1%sucrose, 1% starch plus 1% sucrose, 1% starch plus 0.5% fructose plus 0.5% glucose,or 1% sucrose plus 1% glucose.Results: Amylase activity from sHA disks was detected up to 48 h, thereby increasingthe availability of reducing sugars and acidogenicity in the early stages of biofilmdevelopment. S. mutans grown in the presence of sucrose alone or in combinationsformed well-defined and tightly adherent biofilms comprised of mostly water-insolublepolysaccharides (INS); in contrast, the presence of starch or starch + glucose + fructoseresulted in little biofilm formation with minimal amounts of INS. However, thecombination of starch + sucrose produced biofilms with more biomass and acidogenicity,and a higher content of INS than those grown in sucrose or sucrose + glucose (P < 0.05).The INS extracted from biofilms formed in the presence of starch + sucrose displayed ahigher percentage of 3-linked branching (3,4-, 3,6-, and 3,4,6-linked glucose) comparedto those from biofilms grown in sucrose or sucrose + glucose. Furthermore, biofilmsgrown in starch + sucrose expressed significantly higher levels of gtfB messenger RNAthan sucrose-grown or sucrose + glucose-grown biofilms (P < 0.05).Conclusion: The combination of starch and sucrose has profound effects not only on thecomposition and structure of the polysaccharide matrix but also on gene expression ofS. mutans within biofilms, which may enhance the cariogenic potential of dental biofilms.

S. Duarte1, M. I. Klein1,C. P. Aires2, J. A. Cury2, W. H. Bowen1,

H. Koo1

1Eastman Department of Dentistry and Centerfor Oral Biology, University of RochesterMedical Center, Rochester, NY, USA, 2DentalSchool of Piracicaba, State University ofCampinas, Piracicaba, SP, Brazil

Key words: amylase; biofilms; starch;Streptococcus mutans

H. Koo, Eastman Department of Dentistryand Center for Oral Biology, University ofRochester Medical Center, 625 ElmwoodAvenue, Box 683, Rochester, NY 14620,USATel.: +1 585 273 4216;fax: +1 585 276 0190;e-mail: hyun_koo@urmc.rochester.eduAccepted for publication July 13, 2007

establishment of the extracellular polysac-charide (EPS) matrix, which provides bulkand structural integrity for dental biofilms,and serve as a reserve source of energy (3).The formation of EPS matrix by S. mutansinvolves the interaction of at least threeglucosyltransferases (GTFs) and an endo-dextranase, which participate in the syn-thesis and degradation of glucans; theseenzymes are products of the gtfB, gtfC,gtfD, and dexA genes (15, 22). S. mutanssynthesizes glucans directly from sucrose,but not from undigested starch. However,starches can be digested by salivarya-amylases to maltose, maltodextrins, andother oligosaccharides, some of which canbe acceptors during glucan synthesis (14,36).Enzymatically active a-amylase and

GTFs have been identified in salivarypellicles formed in vitro and in vivo (1,23, 32, 35). Furthermore, starch hydroly-sates produced by salivary a-amylasebound to saliva-coated hydroxyapatite(sHA) increased the synthesis of glucansfrom sucrose by surface-adsorbed GTF B;the hydrolysates also affected the structureand bacterial binding sites of the glucans(36). Moreover, maltose and maltodextrinsfrom starch hydrolysis can be metabolizedinto acids by mutans streptococci (6).Clearly, starch could enhance the cario-genic potential of sucrose, as indicated byprevious in vivo and in situ studies (2, 14,29); the interaction of sucrose and starchthrough GTF enzymes and amylase ad-sorbed on the tooth surface may modulatein situ the development of cariogenicbiofilms by influencing the synthesis ofthe EPS at structural and molecular levels,and the availability of fermentable carbo-hydrates for acid production.Thus, the explanation for the greater

cariogenicity of the dietary combination ofstarch and sucrose may be associated withbiochemical and structural changes in thebiofilms. In this study, we investigatedwhether combinations of starch and su-crose in the presence of surface-adsorbedsalivary amylase and S. mutans, influencebiofilm formation by affecting the synthe-sis and structure of EPS, and expression ofthe gtfB, gtfC, gtfD, and dexA genes usingour sHA disc biofilm model (19).

Materials and methods

Amylase activity of salivary pellicle

Hydroxyapatite discs (Clarkson Chroma-tography Products, Inc., South Williams-port, PA; surface area 2.7 ± 2 cm2) werecoated with filter-sterilized, GTF-free, clar-ified human whole saliva (10, 19). The

levels of amylase in saliva were unaffectedby filtration, as determined experimentallyby immunodetection and direct enzymeassay as described elsewhere (35). ThesHA disc was incubated in ultrafiltered(Amicon 10 kDa molecular weight cut-offmembrane; Millipore Co., Billerica, MA)buffered tryptone yeast-extract broth (pH7.0) containing 1% starch (soluble starch)80% amylopectin and 20% amylose;Sigma Chemical Company, St Louis,MO) at 37�C and 5% CO2 for 5 days;neither bacteria nor saliva was added tothis solution. The 1% starch solution wasreplaced daily until the fifth day of theexperimental period (120 h). Amylaseactivity was determined by measuring theamount of reducing sugars (4) releasedinto the solution at different time-points todetermine whether the surface-adsorbedamylase remain active on the HA surfaceover time.

Biofilm preparation and analysis

Biofilms of S. mutans UA159 (ATCC700610) were formed on sHA discs placedin a vertical position using a disc holder(see Fig. 1) in batch cultures at 37�C in5% CO2 for 5 days (19). The biofilmswere grown in buffered tryptone yeast-extract broth containing: (i) 1% starch, (ii)1% sucrose; (iii) 1% starch + 1% sucrose;(iv) 1% starch + 0.5% glucose + 0.5%fructose; or (v) 1% sucrose + 1% glucose.The culture medium was replaced daily;pH values and amounts of reducing sugarsand total carbohydrates in the mediumwere measured daily after the first 24 h ofincubation. At the end of the experimentalperiod (120-h-old biofilms), the biofilmswere dip-washed three times, and then

gently swirled in physiological saline toremove loosely adherent material. Thebiofilms were placed in 5 ml sterile salinesolution, and the hydroxyapatite surfaceswere gently scraped with a sterile spatulato harvest adherent cells. The removedbiofilms were subjected to sonication usingthree 30-s pulses at an output of 7 W(Branson Sonifier 150; Branson Ultrason-ics, Danbury, CT). The homogenized sus-pension was used for dry weight, totalprotein, and polysaccharide analyses. Forthe dry weight determination, three vol-umes of cold ethanol ()20�C) were addedto 1 ml biofilm suspension, and the result-ing precipitate was collected (10,000 g for10 min at 4�C). The supernatant wasdiscarded, and the pellet was washed twicewith cold ethanol, and then lyophilizedand weighed. Total protein in the biofilmsuspension was determined by acid diges-tion followed by ninhydrin assay (26). Thepolysaccharide composition (extracellularwater-soluble and insoluble, and intracel-lular polysaccharides) was determined bycolorimetric assays as detailed by Kooet al. (18); the polysaccharide content wasexpressed per mg of dry weight or protein.Briefly, an aliquot (4 ml) of the suspensionwas centrifuged at 10,000 g for 10 minat 4�C. The supernatant was collected andthe biofilm pellet was resuspended andwashed in the same volume of water; thisprocedure was repeated twice. All thesupernatants were pooled and threevolumes of cold ethanol were added, andthe resulting precipitate was collected. Theprecipitate, or water-soluble polysaccha-rides, were collected by centrifugation andwashed three times with cold ethanol andresuspended in 1 ml MilliQ H2O; the totalamount of carbohydrate was determined

B

A

Fig. 1. Saliva-coated hydroxyapatite (sHA) biofilm model. (A) sHA discs placed in a verticalposition; (B) biofilms forming in a 24-well plate.

Effects of starch and sucrose on S. mutans biofilms 207

by the phenol–sulfuric acid method (11).The biofilm pellet was dried in a SpeedVac concentrator and used for determina-tion of: (i) extracellular insoluble polysac-charides; and (ii) intracellular iodophilicpolysaccharides. The insoluble polysac-charides were extracted using 1 m NaOH(1 mg biofilm dry weight/0.3 ml of 1 m

NaOH) under agitation for 2 h at 37�C.The supernatant was collected by centri-fugation, and precipitated with three vol-umes of cold ethanol. The precipitate waswashed three times with cold ethanol andresuspended in 1 ml 1 m NaOH; the totalamount of carbohydrate was determinedby the phenol–sulfuric acid method(11). The intracellular iodophilic polysac-charides were extracted with hot 5.3 m

KOH (0.8 mg of biofilm dry weight/mlKOH) and quantified using 0.2% I2/2%KI solution as described by DiPersio et al.(9).

Glycosyl linkage analysis

The extracellular water-soluble and insol-uble polysaccharides were extracted asdescribed above, and dissolved in dimethylsulfoxide (21). For glycosyl linkage anal-ysis, the polysaccharide extracts weremethylated by a modification of themethod of Ciucanu & Kerek (5) followedby combined gas chromatography/massspectrometry (GC/MS) analysis as de-scribed by York et al. (38). The partiallymethylated aldital acetates were analyzedon a 30-m Supelco 2330 bonded phasefused silica capillary column by GC/MSusing a Hewlett Packard 5890 GC inter-faced to a 5970 MSD (mass selectivedetector, electron impact) as detailed else-where (21).

Extraction of RNA and real-time

polymerase chain reaction

The RNA extraction and purification, andreverse transcriptase polymerase chainreaction (PCR) conditions and specificprimers (for gtfB, gtfC, gtfD, and dexA)were similar to those described previously(7, 20). Complementary DNAs (cDNAs)were synthesized using a BioRad iScriptcDNA synthesis kit (Bio-Rad Laborato-ries, Inc., Hercules, CA). To check forDNA contamination, purified total RNAwithout reverse transcriptase served as thenegative control. The resulting cDNA andnegative control were amplified by a MyiQreal-time PCR detection system with iQSYBR Green supermix (Bio-Rad Labora-tories, Inc.) and specific primers. Thecritical threshold cycle (Ct) was defined

as the cycle at which the fluorescencebecomes detectable above the backgroundand is inversely proportional to the loga-rithm of the initial number of templatemolecules. A standard curve was plottedfor each primer set as detailed elsewhere(20). The standard curves were used totransform the Ct values to the relativenumber of cDNA molecules. Relativeexpression was calculated by normalizingeach gene of interest of the biofilms grownin the presence of various carbohydrates tothe 16SrRNA gene (internal control).These values were then compared to thosefrom sucrose-grown biofilms to determinethe change in gene expression.

Statistical analyses

An exploratory data analysis was per-formed to determine the most appropriatestatistical test; the assumptions of equalityof variances and normal distribution oferrors were also checked. The data werethen analyzed using analysis of variance,and the F-test was used to test anydifference among the groups. When sig-nificant differences were detected, pairwisecomparisons were made between all thegroups using Tukey’s method to adjust formultiple comparisons. Triplicates from atleast three separate experiments were con-ducted in each of the assays. Statisticalsoftware JMP version 3.1 (30) was usedto perform the analyses. The level ofsignificance was set at 5%.

Results

The amylase activity of sHA discs inbacteria-free culture medium containing1% starch was measured daily, and theresults are shown in Fig. 2. The salivaryamylase adsorbed on the hydroxyapatitesurface was active 48 h after salivarypellicle formation, although the enzymeactivity declined between 48 and 72 h.The presence of sucrose alone or in

combinations resulted in biofilms display-ing four to seven times more biomass,seven to 15 times more total protein, and

10 to 15 times more total EPS than starchor starch + glucose + fructose-grown bio-films (Table 1). However, biofilms of S.mutans grown with starch in combinationwith sucrose exhibited significantly morebiomass and total amount of EPS than thebiofilms formed in the presence of sucrose,either alone or in combination withglucose (P < 0.05).The total amount (in lg/total biofilm dry

weight) and content (in lg/mg protein) ofextracellular insoluble (INS) and water-soluble (WSP) polysaccharides, and ofintracellular polysaccharides (IPS) in bio-films are shown in Fig. 3. Biofilms formedin the presence of sucrose alone or incombinations displayed a significantlyhigher content of INS than starch orstarch + glucose + fructose-grown bio-films (P < 0.05). The INS content instarch + sucrose-grown biofilms wassignificantly higher than that of thesucrose-grown and sucrose + glucose-grown biofilms (P < 0.05). On the otherhand, starch- or starch + glucose + fruc-tose-grown biofilms were comprised ofmostly WSP, and showed higher content ofthe soluble polysaccharides than biofilmsgrown in sucrose alone or in combinations(P < 0.05). The amount (and content) of

-

10

20

30

40

50

60

70

2 Time (h) after salivary pellicle formation

µg o

f re

duci

ng s

ugar

s re

leas

ed /

min

Medium replacement

4 6 8 24 48 72 96 120

Fig. 2. Amylase activity indicated by theamount of reducing sugars released into themedium during incubation of sHA in 1% starch.The amylase activities at 2, 4, and 6 h aftersalivary pellicle formation were not significantlydifferent from each other (n = 12; P > 0.05,anova, comparison for all pairs using Tukeytest).

Table 1. Biomass (dry-weight), total amount of protein, and EPS in Streptococcus mutans UA159biofilms formed in the presence of starch and sucrose, alone or in combinations

Experimentalgroups

Dry-weight(mg)

Total amountof protein (mg)

Total amount ofEPS (lg)

Starch 0.75 (0.27)1 0.1 (0.08)1 88.98 (17.84)1

Starch + sucrose 6.25 (0.69)2 1.3 (0.12)2 1747.99 (146.62)2

Sucrose 5.50 (0.45)3 1.5 (0.32)3 1411.28 (256.45)3

Starch + glucose + fructose 1.25 (0.42)1 0.2 (0.12)1 126.37 (16.58)4

Sucrose + glucose 3.92 (0.92)4 1.4 (0.21)2 850.31 (190.79)5

Values (SD, n = 12) in the same column followed by the same superscript numbers are notsignificantly different from each other (P > 0.05, anova, comparison for all pairs using Tukey test).

208 Duarte et al.

IPS in biofilms formed with sucrose aloneor in combinations was significantly higherthan in starch-grown biofilms (P < 0.05).The activity of surface-adsorbed amy-

lase on starch could increase the levels offermentable carbohydrates in the culturemedium, and thereby enhance the acidproduction by biofilms. Therefore, thereducing sugars and total carbohydratelevels, and the pH of the culture mediumsurrounding the biofilms were determineddaily.The pH of the culture medium was

measured at various time-points duringeach 24 h incubation period, and the pH-drop curves are illustrated in Fig. 4.Biofilms grown in sucrose, alone or incombinations, rapidly lowered the culture

pH to values below 4.5 each day ofgrowth. Biofilms formed in the presenceof starch + sucrose displayed the highestrate of acid production during the first 8 hof incubation of early-formed biofilms(between 24 and 48 h) showing signifi-cantly lower pH values than starch (t28h,t32h, t36h, and t48h), starch + glu-cose + fructose (t28h, t32h, and t36h), su-crose (t28h and t32h), and sucrose + glucose(t28h) -grown biofilms (P < 0.05). Biofilmsformed in starch also lowered the culturepH to 5.3 in the first 48 h. The rate of acidproduction was slowed after 72 h eventhough fresh medium was added daily,which is consistent with a decline ofamylase activity after 48 h of incubation.Furthermore, biofilms of S. mutans grown

in starch, either alone or in combinations,displayed elevated levels of reducing sug-ars in the earlier stages (24–48 h) ofbiofilm formation (data not shown), whichagrees well with the amylase activity data(Fig. 2) and the pH drop curves (Fig. 4).In an approach to determine whether

polysaccharide matrices formed in thepresence of different carbohydrates haddistinctive structures, the type of glycosyllinkages in WSP and INS extracted fromthe biofilms were determined (Table 2).Major structural differences were observedin WSP and INS from biofilms grown withstarch or sucrose, alone or in combina-tions. Soluble polysaccharides fromstarch + sucrose-grown biofilms displayedhigher percentages of 4-linked glucose and

0

200

400

600

800

1000

1200

1400

1600

µg o

f po

lysa

ccha

ride

/ m

g of

pro

tein

IPS INS WSP

0

200

400

600

800

1000

1200

1400

1600

Starch Starch + Glucose + Fructose

Sucrose Sucrose + Glucose

Starch + Sucrose Starch Starch + Glucose + Fructose

Sucrose Sucrose + Glucose

Starch + Sucrose

Tot

al a

mou

nt o

f po

lysa

ccha

ride

(µg

)

IPS INS WSP

Total amount Expressed per mg of protein

a

b

a

c c

b b

a a

b b

c c

d d,e e

Fig. 3. Total amount (in lg/total biofilm dry-weight) and content (expressed per mg of protein) of intracellular (IPS) and extracellular water-soluble(WSP) and insoluble (INS) polysaccharides in Streptococcus mutans UA159 biofilms formed in the presence of starch and sucrose, alone or incombinations. Values (SD, n = 12) for each type of polysaccharides marked by the same letters are not significantly different from each other (P > 0.05,anova, comparison for all pairs using Tukey test).

3.5

4

4.5

5

5.5

6

6.5

7

7.5

24 32 40 48 56 64 72 80 88 96 104 112 120

Time of biofilm formation (h)

pH

Starch Starch + Sucrose Sucrose Starch + Glucose + Fructose Sucrose + Glucose

Medium replacement

Fig. 4. pH measurements in the culture medium during Streptococcus mutans biofilm formation. The medium was replaced daily with fresh medium. ThepH values (n = 12) were determined after 4, 8, 12, and 24 h of incubation; the reducing sugars and total carbohydrate levels (n = 12) were measured aftera 24-h period of incubation for each day.

Effects of starch and sucrose on S. mutans biofilms 209

less 3-, 6-, and 3,6-linked glucose thanthose from biofilms grown in sucrose orsucrose + glucose. In contrast, the INSfrom starch + sucrose biofilms showedhigher levels of 3-linked branching (3,4-,3,6-, and 3,4,6-linked glucose) and con-siderably less 4-linked glucose than thesucrose-based biofilms. The WSP ofstarch-grown and starch + glucose + fruc-tose-grown biofilms were comprised pre-dominantly of 4-linked glucose.Lastly, the expression of gtfB, gtfC,

gtfD, and dexA in S. mutans biofilmsgrown in starch or sucrose, alone or incombinations, was determined by real-timereverse transcription PCR. Overall, theexpression of gtfB messenger RNA(mRNA) in biofilms formed withstarch + sucrose was significantly in-creased (25–40%) whereas gtfD mRNAlevels were decreased (20–30%) whencompared with sucrose and sucrose + glu-cose-grown biofilms (P < 0.05); gtfC anddexA expression was also decreased instarch + sucrose-grown biofilms but thedifferences were not statistically signifi-cant (P > 0.05). The gene expression ofstarch- and starch + glucose + fructose-grown biofilms was not determined be-cause of minimal biofilm formation (andpoor RNA yield).

Discussion

The results of this study showed that thecombination of starch and sucrose exposedto surface-adsorbed salivary amylase and S.mutans clearly influenced the formationand acidogenicity of biofilms by at leastfour routes: (i) enhanced the total biomassand the content of extracellular insolublepolysaccharides, (ii) synthesized a structur-ally distinctive EPS matrix, (iii) enhancedacid production in the early stages ofbiofilm formation, and (iv) affected theexpression of specific genes involved inEPS matrix formation (e.g. gtfB). Ourmonospecies biofilm model is advanta-geous in examining specific actions of

carbohydrates on S. mutans physiologyand genetics, especially on the glucan-mediated processes involved in the forma-tion of the polysaccharide matrix inbiofilm, although it does not mimic thecomplex microbial community found indental plaque.S. mutans growing in the presence of

sucrose alone or in combinations formed awell-defined, firmly adherent, and highlyacidogenic biofilm on the surface of sHAcomprised mostly of insoluble polysaccha-rides containing 3-, 4-, and 6-linked glu-cose; which agrees well with the glycosyllinkage profile of insoluble glucans syn-thesized by surface-adsorbed streptococcalglucosyltransferases (17, 21). In contrast,the presence of starch alone or in combi-nation with glucose and fructose resultedin little (and loosely attached) biofilmformation displaying predominantly solu-ble polysaccharides with 1,4-linked glu-cose, which suggests that starch and itshydrolysates might be incorporated on tothe sHA surface (36). The inability ofS. mutans to form adherent and establishedbiofilms on the surface of sHA in thepresence of starch or starch + glu-cose + fructose may be related to a lackof insoluble polysaccharide synthesis be-cause insoluble glucans are essential inproviding structural integrity and bulk tobiofilms (3). However, the combination ofstarch and sucrose enhanced the acidoge-nicity of early-formed biofilms (up to 48 h,when salivary amylase is active), and moreimportantly increased the production ofinsoluble EPS by S. mutans within bio-films when compared to sucrose alone orsucrose with glucose. The starch hydroly-sates released by the action of surface-adsorbed amylase combined with sucrosein the medium enhanced the extracellularand intracellular sugar metabolism byS. mutans at the pellicle–biofilm interfaceby providing oligosaccharides to serve asacceptors in glucan synthesis by glucos-yltransferases (14, 36), and fermentablecarbohydrates for acid production (6).

These effects would certainly increase thebiofilm accumulation on the tooth surfaceand accelerate the breakdown of microbialhomeostasis in dental plaque (3, 27).Interestingly, the addition of excess glu-cose in the sucrose medium resulted in lessbiomass and EPS content than biofilmsformed with sucrose alone; an observationconsistent with previous studies showingthat S. mutans growing with glucosein excess diminished the synthesis ofextracellular polysaccharides andrepressed the sugar uptake by the phos-photransferase system (12, 16). The intra-cellular polysaccharide accumulation wasnot markedly affected whether biofilmswere grown in the presence of sucrosealone or in combinations, although the IPScontent in starch + sucrose-grown andsucrose + glucose-grown biofilms wasslightly higher than in sucrose-grownbiofilms.Furthermore, biofilms grown in the

presence of starch + sucrose resulted in astructurally distinct EPS matrix whencompared to those formed in sucrose orsucrose + glucose. It is noteworthy thatthe presence of starch + sucrose resultedin insoluble polysaccharides comprisedpredominantly of 1 fi 3 and 1 fi 6linkages, and higher percentages of branchpoints from 3,4-, 3,6-, and 3,4,6-linkedglucose than those from sucrose- orsucrose + glucose-grown biofilms. It isapparent that the presence of oligosaccha-rides from starch hydrolysis is contributingto the insolubilization of exopolysaccha-ride matrix in starch + sucrose-grown bio-films by (i) enhancing the content ofinsoluble polysaccharides containing (ii)a higher percentage of insoluble 3-linkedbranching. The higher content of insolublepolysaccharides in starch + sucrose-grownbiofilm matrix can be explained byprevious observations that starch hydroly-sates in combination with sucrose in-creased the synthesis of insolubleglucans, and also affected the structure ofglucans synthesized by surface-adsorbed

Table 2. Percentage of glycosyl linkages of water-soluble (WSP) and insoluble (INS) polysaccharides extracted from Streptococcus mutans biofilmsgrown in the presence of starch and sucrose, alone or in combinations

Glycosyl residue

Starch Starch + sucrose SucroseStarch +glucose + fructose Sucrose + glucose

WSP INS WSP INS WSP INS WSP INS WSP INS

3-linked glucose ) n/d ± +++ ++ +++ ) n/d ++ +++4-linked glucose +++++ n/d +++++ + +++ +++ +++++ n/d ++ ++++6-linked glucose ) n/d + +++ +++ ++ ) n/d ++ ++3,4-linked glucose ) n/d ) ± ) ) ) n/d ) )3,6-linked glucose ) n/d ± ++ ++ + ) n/d + +3,4,6-linked glucose ) n/d ) ± ) ) ) n/d ) )

n/d, not determined.), 0–1%; ±, 1–4%; +, 5–9%; ++, 10–19%; +++, 20–29%; ++++, 30–59%; +++++, ‡ 60% or more.

210 Duarte et al.

GTF B, resulting in enhanced S. mutansbinding compared to those formed withsucrose alone (21, 36); this enhancementmay be associated with changes in thebinding sites of the modified glucans (21,36). Clearly, the influence of the oligosac-charides on the GTF B activity plays acritical role in changing the physical andbiochemical properties of the biofilmsmatrix, and thereby influencing its cario-genic properties. However, in biofilms, theformation and maturation of the EPSmatrix is a result of a dynamic interactionof all of the three GTFs acting in concertand is influenced by an endodextranaseproduced simultaneously by S. mutans (15,17, 22). The presence of starch in combi-nation with sucrose may be modulating allof the enzymes responsible for synthesisand degradation concomitantly, resultingin a structurally distinct matrix. Furtherstudies shall elucidate how starch andsucrose influence the synthesis and degra-dation of glucans concomitantly during theEPS matrix development at molecular andstructural levels.A recent in situ study showed that a

combination of 2% starch + 10% sucrosewas potentially more cariogenic than 10%sucrose alone, despite the total amounts ofEPS in the matrices of the biofilms beingsimilar to each other (29). Although higherlevels of acidogenic and aciduric bacteria,such as Lactobacilli, were found in thebiofilms, the enhanced cariogenicity ofstarch + sucrose may be also explained bythe structural differences of the EPS matrixbetween starch + sucrose-grown biofilmsand those formed in sucrose alone. Thestructural changes in the matrix may affectthe diffusion properties, bacterial bindingsites, physical integrity, and architecture ofthe biofilms (8, 21, 34, 36, 39). It istherefore feasible that such changes in theEPS matrix of biofilms could modulate thepathogenesis of dental caries, substantiat-ing the concept that there is a starchhydrolysate contribution to the formationof cariogenic dental plaque. The exactmechanisms by which the structuralchanges enhance the virulence of thebiofilms need further elucidation.Lastly, we examined the expression

profile of the genes encoding the synthesis(gtfB, gtfC, gtfD) and degradation (dexA)of glucans by S. mutans within biofilms,in an attempt to explain the structuraldifferences of the EPS matrix observedbetween starch + sucrose-grown and su-crose-grown biofilms. Our data indicatethat biofilms formed in starch + sucroseexpressed significantly higher levels ofgtfB mRNA and less gtfD mRNA than

those formed in sucrose. It is noteworthythat expression of gtfC and dexA wasdecreased in starch + sucrose biofilms(although the differences were not statisti-cally significant), indicating an overalleffect of induction of gtfB. This observa-tion could explain, in part, the differencesobserved in the structure of the EPSmatrix, e.g. higher percentage of 3-linkedbranching in insoluble polysaccharide andless 6-linked glucose in the soluble poly-saccharide matrix of starch + sucrosebiofilms. Furthermore, gtfB is a criticalvirulence gene associated with the patho-genesis of dental caries (37); S. mutanstreated with therapeutic agents that repressthe expression of gtfB, or mutant strain ofthis organism defective in gtfB, are far lesscariogenic than untreated or parent strainsin vivo (19, 20, 37). Thus, the presence ofa combination of starch and sucrose wouldresult in a more virulent (cariogenic)biofilm. We are currently pursuing detailedgene expression profiling at differentstages of biofilm formation to betterunderstand the molecular mechanismsinvolved in the EPS synthesis in thepresence of starch and sucrose.Our data offer, in part at least, an

explanation for why starch and sucrosecombinations are potentially more cario-genic than either alone; and furthermoreillustrate that composition of the diet caninfluence the virulence traits of the oralpathogen S. mutans. Clearly, surface-adsorbed a-amylase may have an addi-tional role in dental biofilm formationother than promoting specific bacterialadhesion (31) by contributing directly withthe synthesis of a structurally distinctextracellular polysaccharide matrix andby enhancing the expression of gtfB.Further studies using additional microor-ganisms that bind amylase in a multispe-cies biofilm model shall elucidate evenfurther the role of starch and sucrose in thevirulence of cariogenic biofilms.

Acknowledgment

The authors thank FAPESP (04/00688-3)from whom the 3rd author received ascholarship.

References

1. Al-Hashimi I, Levine MJ. Characterizationof in vivo salivary-derived enamel pellicle.Arch Oral Biol 1989: 34: 289–295.

2. Bowen WH, Amsbaugh SM, Monell-Torrens S, Brunelle J, Kuzmiak-Jones H,Cole MF. A method to assess cariogenicpotential of foodstuffs. J Am Dent Assoc1980: 100: 677–681.

3. Bowen WH. Do we need to be concernedabout dental caries in the coming millen-nium? J Am Dent Assoc 2002: 133: 1405–1407.

4. Chaplin MF. Monosaccharides. In: ChaplinMF, Kennedy JF, ed. Carbohydrate analysis– a practical approach. Oxford: IRL Press,1986: 1–36.

5. Ciucanu I, Kerek F. A simple and rapidmethod for the permethylation of carbohy-drates. Carbohydr Res 1984: 131: 209–217.

6. Clarkson CH, Krell D, Wefel JS, Crall J,Feagin FF. In vitro caries-like lesion pro-duction by Streptococcus mutans and Acti-nomyces viscosus using sucrose and starch.J Dent Res 1987: 66: 795–798.

7. Cury JA, Koo H. Extraction and purifica-tion of total RNA from Streptococcusmutans biofilms. Anal Biochem 2007:365: 208–214.

8. Dibdin GH, Shellis RP. Physical and bio-chemical studies of Streptococcus mutanssediments suggest new factors linking thecariogenicity of plaque with its extracellularpolysaccharide content. J Dent Res 1988:67: 890–895.

9. DiPersio JR, Mattingly SJ, Higgins ML,Shockman GD. Measurement of intracellu-lar iodophilic polysaccharide in two cario-genic strains of Streptococcus mutans bycytochemical and chemical methods. InfectImmun 1974: 10: 597–604.

10. Duarte S, Gregoire S, Singh AP et al.Inhibitory effects of cranberry polyphenolson formation and acidogenicity of Strepto-coccus mutans biofilms. FEMS MicrobiolLett 2006: 257: 50–56.

11. Dubois M, Gilles KA, Hamilton JK, RebersPA, Smith F. Colorimetric method fordetermination of sugars and related sub-stances. Anal Chem 1956: 28: 350–356.

12. Ellwood DC, Phipps PJ, Hamilton IR.Effect of growth rate and glucose concen-tration on the activity of the phosphoenol-pyruvate phosphotransferase system inStreptococcus mutans Ingbritt grown incontinuous culture. Infect Immun 1979:23: 224–231.

13. Firestone AR, Shmid R, Muhlemann HR.Cariogenic effects of cooked wheat starchalone or with sucrose and frequency-controlled feedings in rats. Arch Oral Biol1982: 27: 759–763.

14. Fu DT, Robyt JF. Maltodextrin acceptorreactions of Streptococcus mutans 6715glucosyltransferases. Carbohydr Res 1991:217: 201–211.

15. Guggenheim B, Burckhardt JJ. Isolationand properties of a dextranase from Strep-tococcus mutans OMZ 176. Helv OdontolActa 1974: 18: 101–113.

16. Hamilton IR, Phipps PJ, Ellwood DC.Effect of growth rate and glucose concen-tration on the biochemical properties ofStreptococcus mutans Ingbritt in continuousculture. Infect Immun 1979: 26: 861–869.

17. Hayacibara MF, Koo H, Vacca-Smith AMet al. The influence of a novel propolis onmutans streptococci biofilms and cariesdevelopment in rats. Arch Oral Biol 2006:51: 15–22.

18. Koo H, Hayacibara MF, Schobel BD et al.Inhibition of Streptococcus mutans biofilmaccumulation and polysaccharide produc-

Effects of starch and sucrose on S. mutans biofilms 211

tion by apigenin and tt-farnesol. J Antimic-rob Chemother 2003: 52: 782–789.

19. Koo H, Schobel B, Scott-Anne K et al.Apigenin and tt-farnesol with fluorideeffects on Streptococcus mutans biofilmsand dental caries. J Dent Res 2005: 84:1016–1020.

20. Koo H, Seils J, Abranches J, Burne RA,Bowen WH, Quivey RG Jr. Influence ofapigenin on gtf gene expression in Strepto-coccus mutans UA159. Antimicrob AgentsChemother 2006: 50: 542–546.

21. Kopec LM, Vacca-Smith AM, BowenWH. Structural aspects of glucans formedin solution and on the surface of hydro-xyapatite. Glycobiology 1997: 7: 929–934.

22. Kuramitsu HK. Molecular genetic analysisof the virulence of oral bacterial pathogens:an historical perspective. Crit Rev Oral BiolMed 2003: 14: 331–344.

23. Lamkin MS, Arancillo AA, OpennheimFG. Temporal and compositional charac-teristics of salivary protein adsorption tohydroxyapatite. J Dent Res 1996: 75: 803–808.

24. Lingstrom P, Van Houte J, Kashket S. Foodstarches and dental caries. Crit Rev OralBiol Med 2000: 11: 366–380.

25. Loesche WJ. Role of Streptococcus mutansin human dental decay. Microbiol Rev1986: 50: 353–380.

26. Moore S, Stein WH. A modified ninhydrinreagent for the photometric determinationof amino acids and related compounds. JBiol Chem 1954: 211: 907–913.

27. Marsh PD. Are dental diseases examples ofecological catastrophes? Microbiology2003: 149: 279–294.

28. Paes Leme AF, Koo H, Bellato CM, BediG, Cury JA. The role of sucrose in cario-genic dental biofilm formation – newinsight. J Dent Res 2006: 85: 878–887.

29. Ribeiro CC, Tabchoury CP, Del Bel CuryAA, Tenuta LM, Rosalen PL, Cury JA.Effect of starch on the cariogenic potentialof sucrose. Br J Nutr 2005: 94: 44–50.

30. SAS Institute Inc. (ed.) JMP.� User’sGuide: Version 2 JMP. Cary, NJ: SASInstitute, 1989.

31. Scannapieco FA, Torres G, Levine MJ.Salivary alpha-amylase: role in dentalplaque and caries formation. Crit Rev OralBiol Med 1993: 4: 301–307.

32. Scheie AA, Eggen KH, Rolla G. Glucosyl-transferase activity in human in vivo formedenamel pellicle and in whole saliva. ScandJ Dent Res 1987: 95: 212–215.

33. Tanzer JM, Freedman ML, Fitzgerald RJ.Virulence of mutants defective in glucosyl-transferase, dextran-mediated aggregation,or dextranase activity. In: Mergenhagen SE,Rosan B ed. Molecular basis of oralmicrobial adhesion. Washington, DC:

American Society of Microbiology, 1985:204–211.

34. Thurnheer T, Gmur R, Shapiro S, Guggen-heim B. Mass transport of macromoleculeswithin an in vitro model of supragingivalplaque. Appl Environ Microbiol 2003: 69:1702–1709.

35. Vacca-Smith AM, Bowen WH. In situstudies of pellicle formation on hydroxy-apatite discs. Arch Oral Biol 2000: 45:277–291.

36. Vacca-Smith AM, Venkitaraman AR, Qui-vey RG Jr, Bowen WH. Interactions ofstreptococcal glucosyltransferases withalpha-amylase and starch on the surface ofsaliva-coated hydroxyapatite. Arch OralBiol 1996: 41: 291–298.

37. Yamashita Y, Bowen WH, Burne RA,Kuramitsu HK. Role of the Streptococcusmutans gtf genes in caries induction in thespecific-pathogen-free rat model. InfectImmun 1993: 61: 3811–3817.

38. York WS, Darvill AG, McNeil M, Steven-son TT, Albersham P. Isolation and charac-terization of plant cell walls and cell wallcomponents. Methods Enzymol 1985: 118:3–40.

39. Zero DT, Fu J, Anne KM, Cassata S,McCormack SM, Gwinner LM. An im-proved intra-oral enamel demineralizationtest model for the study of dental caries.J Dent Res 1992: 71: 871–878.

212 Duarte et al.