Evidence that L-(+)-Lactate Dehydrogenase Deficiency …iai.asm.org/content/62/1/60.full.pdf ·...

5
INFECrION AND IMMUNITY, Jan. 1994, p. 60-64 0019-9567/94/$04.00+0 Evidence that L-(+)-Lactate Dehydrogenase Deficiency Is Lethal in Streptococcus mutans JEFFREY D. HILLMAN,`* ANPING CHEN,' MARGARET DUNCAN,2 AND SEOK-WOO LEE' Department of Oral Biology, University of Florida College of Dentistry, Gainesville, Florida 32610,1 and Department of Molecular Genetics, Forsyth Dental Center, Boston, Massachusetts 021152 Received 6 August 1993/Returned for modification 27 September 1993/Accepted 20 October 1993 In order to construct an effector strain for the replacement therapy of dental caries, we wished to combine the properties of low-level acid production and high-level colonization potential in a strain of Streptococcus mutans. To this end, we made a deletion in the lactate dehydrogenase (LDH) gene cloned from the bacteriocin-producing S. mutans strain JH1000. However, we were unable to substitute the mutant for the wild-type allele by transformation with linear DNA fragments. The mutated gene, carried on a suicide vector, was shown by Southern analysis to integrate into the JH1000 chromosome to yield transformants carrying both the wild-type gene and mutated LDH gene. Three spontaneous self-recombinants of one heterodiploid strain were isolated by screening 1,500 colonies for a loss of the tetracycline resistance encoded by the gene used to mark the LDH deletion. In all three cases, Southern analysis showed that a loss of tetracycline resistance was accompanied by a loss of the mutated LDH gene, resulting in restoration of the wild-type genotype. However, screening the same number of colonies for self-recombinants that did not make lactic acid during anaerobic growth in Todd-Hewitt broth failed to identify clones in which the wild-type allele was lost. A second, simpler screening of more than 80,000 colonies grown aerobically on glucose tetrazolium medium to identify low-level-acid-producing colonies was also unsuccessful. These results are interpreted as indicating that LDH deficiency is lethal in S. mutans under the cultivation conditions used in these experiments. The physiological bases for this hypothesis are described. In the past, lactic acid production by fructose-1,6-diphos- phate (FDP)-dependent L-(+)-lactate dehydrogenase (LDH) was shown to be an important virulence factor for Strepto- coccus rattus (11). LDH-deficient mutants were substan- tially less cariogenic than the wild-type organism in both laboratory and rodent models. As a result, LDH deficiency has been proposed as one aspect of a strategy to construct an effector strain for the replacement therapy of dental caries. Following chemical mutagenesis and aerobic incubation on glucose tetrazolium medium, LDH-deficient mutants of S. rattus BHT-2 (5) and Streptococcus cncetus AHT (un- published data) were recovered as red colonies. However, we have been unable to isolate LDH-deficient mutants from any of a large number of laboratory and fresh S. mutans strains, including the bacteriocinogenic strain JH1000 (9), which has colonization properties that make it particularly suitable for use in replacement therapy (8). Abhyankar et al. (1) did succeed in isolating an LDH- deficient mutant from a fresh isolate of S. mutans. This isolate was distinctive in the pattern of fermentation end products it generated during growth in limiting glucose under anaerobic conditions. In particular, it produced relatively high levels of ethanol and acetate and a low level of lactic acid. The authors speculated that a preexisting mutation in an alternate pyruvate-dissimilating pathway was necessary to support viability in the presence of an LDH deficiency. The LDH-deficient S. mutans strain isolated by Abhyan- kar et al. was obtained following mutagenesis with ni- trosoguanidine, which raises concerns about cryptic muta- tions and reversion of the mutation in the LDH gene. * Corresponding author. Mailing address: University of Florida College of Dentistry, Box 100-424, Gainesville, FL 32610. Phone: (904) 392-4370. Fax: (904) 392-3070. Electronic mail address: [email protected]. Furthermore, this strain has no known properties, such as bacteriocin production, that promote its ability to colonize the human oral cavity. To overcome these problems, and to ensure stable and specific mutation of the LDH gene in effector strain construction, we cloned the LDH gene from strain JH1000 into Escherichia coli (7). In vitro disruption of the gene was achieved by deletion of the promoter and a major portion of the protein coding sequence (3). A tetracy- cline resistance gene from S. mutans, which is also ex- pressed in E. coli, was inserted at the deletion site as a marker for selection. Southern analysis demonstrated that strain JH1000 carries a single copy of the LDH gene in its chromosome. Consequently, exchange of the mutated gene for the wild-type gene should give rise to an LDH-deficient mutant in a single step. In this report we describe experiments that show that single-step exchange of the mutated LDH gene for the wild-type LDH gene does not occur via natural transforma- tion. However, when carried on a suicide vector, the mu- tated LDH gene can be introduced into strain JH1000 at a low frequency by natural transformation. The mutated LDH gene became integrated into the chromosome by homolo- gous recombination to produce clones that were heterodip- loid for LDH. Spontaneous recombination events which excised the mutated LDH gene were observed to occur at a relatively high frequency, but none which eliminated the wild-type gene were found, suggesting that LDH deficiency in S. mutans is lethal under the culture conditions employed. MATERIALS AND METHODS Chemicals, media, bacterial strains, and plasmids. Chemi- cals and antibiotics were from Sigma Chemical Co., St. Louis, Mo. Except where indicated, S. mutans strains were grown aerobically overnight in brain heart infusion broth or 60 Vol. 62, No. 1 on June 19, 2018 by guest http://iai.asm.org/ Downloaded from

Transcript of Evidence that L-(+)-Lactate Dehydrogenase Deficiency …iai.asm.org/content/62/1/60.full.pdf ·...

INFECrION AND IMMUNITY, Jan. 1994, p. 60-640019-9567/94/$04.00+0

Evidence that L-(+)-Lactate Dehydrogenase Deficiency IsLethal in Streptococcus mutans

JEFFREY D. HILLMAN,`* ANPING CHEN,' MARGARET DUNCAN,2 AND SEOK-WOO LEE'

Department of Oral Biology, University of Florida College of Dentistry, Gainesville, Florida 32610,1 andDepartment ofMolecular Genetics, Forsyth Dental Center, Boston, Massachusetts 021152

Received 6 August 1993/Returned for modification 27 September 1993/Accepted 20 October 1993

In order to construct an effector strain for the replacement therapy of dental caries, we wished to combinethe properties of low-level acid production and high-level colonization potential in a strain of Streptococcusmutans. To this end, we made a deletion in the lactate dehydrogenase (LDH) gene cloned from thebacteriocin-producing S. mutans strain JH1000. However, we were unable to substitute the mutant for thewild-type allele by transformation with linear DNA fragments. The mutated gene, carried on a suicide vector,was shown by Southern analysis to integrate into the JH1000 chromosome to yield transformants carrying boththe wild-type gene and mutated LDH gene. Three spontaneous self-recombinants of one heterodiploid strainwere isolated by screening 1,500 colonies for a loss of the tetracycline resistance encoded by the gene used tomark the LDH deletion. In all three cases, Southern analysis showed that a loss of tetracycline resistance wasaccompanied by a loss of the mutated LDH gene, resulting in restoration of the wild-type genotype. However,screening the same number of colonies for self-recombinants that did not make lactic acid during anaerobicgrowth in Todd-Hewitt broth failed to identify clones in which the wild-type allele was lost. A second, simplerscreening of more than 80,000 colonies grown aerobically on glucose tetrazolium medium to identifylow-level-acid-producing colonies was also unsuccessful. These results are interpreted as indicating that LDHdeficiency is lethal in S. mutans under the cultivation conditions used in these experiments. The physiologicalbases for this hypothesis are described.

In the past, lactic acid production by fructose-1,6-diphos-phate (FDP)-dependent L-(+)-lactate dehydrogenase (LDH)was shown to be an important virulence factor for Strepto-coccus rattus (11). LDH-deficient mutants were substan-tially less cariogenic than the wild-type organism in bothlaboratory and rodent models. As a result, LDH deficiencyhas been proposed as one aspect of a strategy to construct aneffector strain for the replacement therapy of dental caries.

Following chemical mutagenesis and aerobic incubationon glucose tetrazolium medium, LDH-deficient mutants ofS. rattus BHT-2 (5) and Streptococcus cncetus AHT (un-published data) were recovered as red colonies. However,we have been unable to isolate LDH-deficient mutants fromany of a large number of laboratory and fresh S. mutansstrains, including the bacteriocinogenic strain JH1000 (9),which has colonization properties that make it particularlysuitable for use in replacement therapy (8).Abhyankar et al. (1) did succeed in isolating an LDH-

deficient mutant from a fresh isolate of S. mutans. Thisisolate was distinctive in the pattern of fermentation endproducts it generated during growth in limiting glucose underanaerobic conditions. In particular, it produced relativelyhigh levels of ethanol and acetate and a low level of lacticacid. The authors speculated that a preexisting mutation inan alternate pyruvate-dissimilating pathway was necessaryto support viability in the presence of an LDH deficiency.The LDH-deficient S. mutans strain isolated by Abhyan-

kar et al. was obtained following mutagenesis with ni-trosoguanidine, which raises concerns about cryptic muta-tions and reversion of the mutation in the LDH gene.

* Corresponding author. Mailing address: University of FloridaCollege of Dentistry, Box 100-424, Gainesville, FL 32610. Phone:(904) 392-4370. Fax: (904) 392-3070. Electronic mail address:[email protected].

Furthermore, this strain has no known properties, such asbacteriocin production, that promote its ability to colonizethe human oral cavity. To overcome these problems, and toensure stable and specific mutation of the LDH gene ineffector strain construction, we cloned the LDH gene fromstrain JH1000 into Escherichia coli (7). In vitro disruption ofthe gene was achieved by deletion of the promoter and amajor portion of the protein coding sequence (3). A tetracy-cline resistance gene from S. mutans, which is also ex-pressed in E. coli, was inserted at the deletion site as amarker for selection. Southern analysis demonstrated thatstrain JH1000 carries a single copy of the LDH gene in itschromosome. Consequently, exchange of the mutated genefor the wild-type gene should give rise to an LDH-deficientmutant in a single step.

In this report we describe experiments that show thatsingle-step exchange of the mutated LDH gene for thewild-type LDH gene does not occur via natural transforma-tion. However, when carried on a suicide vector, the mu-tated LDH gene can be introduced into strain JH1000 at alow frequency by natural transformation. The mutated LDHgene became integrated into the chromosome by homolo-gous recombination to produce clones that were heterodip-loid for LDH. Spontaneous recombination events whichexcised the mutated LDH gene were observed to occur at arelatively high frequency, but none which eliminated thewild-type gene were found, suggesting that LDH deficiencyin S. mutans is lethal under the culture conditions employed.

MATERIALS AND METHODS

Chemicals, media, bacterial strains, and plasmids. Chemi-cals and antibiotics were from Sigma Chemical Co., St.Louis, Mo. Except where indicated, S. mutans strains weregrown aerobically overnight in brain heart infusion broth or

60

Vol. 62, No. 1

on June 19, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

LDH DEFICIENCY IS LETHAL IN STREPTOCOCCUS MUTANS 61

Todd-Hewitt broth (Difco Laboratories, Detroit, Mich.) orin candle jars on plates of the same media containing 1.5%agar. E. coli strains were grown overnight with aeration inLuria-Bertani broth or on plates of the same medium con-taining 1.5% agar. Glucose tetrazolium plates were preparedby the method of Miller (14) and contained 1% (wt/vol)glucose. Ampicillin and tetracycline were added to platesand broth at 30 and 10 ,ug/ml, respectively, as required. S.mutans JH1000 (9), JH145 (6), and NG8 (12) and E. coliMC1061 (2) have been described previously. Plasmids usedin this study are described in Results.

Genetic methods. Restriction enzymes and T4 ligase wereobtained from New England BioLabs, Inc. (Beverly, Mass.),and were used with the buffer supplied, according to theconditions specified by the manufacturer.Chromosomal DNA was isolated from S. mutans as pre-

viously described (7). Plasmid DNA was purified by dye-buoyant density centrifugation (13). DNA fragments fromrestriction enzyme digestions were excised from agarosegels (0.6 to 1.0%) and purified by electroelution as describedby Maniatis et al. (13).

Southern hybridizations were performed with the En-hanced Chemiluminescence Gene Detection System fromAmersham International PLC (Amersham, England). Thesupplier's protocols were used without modification forhybridization and for DNA probe preparation. Nitrocellu-lose filters were exposed to Hyperfilm-ECL (Amersham).

Transformation of S. mutans JH1000 and NG8 was per-formed by the methods of Lee et al. (12).

Screening for lactic acid production. Isolated colonies of S.mutans clone JDM4 were picked with sterile toothpicks andinoculated into microtiter wells containing 100 ,ul of Todd-Hewitt broth plus tetracycline. The plates were incubatedanaerobically (85% nitrogen, 10% carbon dioxide, and 5%hydrogen) for 3 days at 37°C. Lactic acid production wasdetermined semiquantitatively by placing a 5-,ul sample fromeach well into a fresh microtiter well and adding 5 ,ul of acocktail containing 0.25 M potassium phosphate buffer (pH7.4), 20 mg of NAD per ml, 2 mg of nitroblue tetrazolium perml, 2 mg of phenazine methosulfate per ml, and 50 U ofL-(+)-lactate dehydrogenase from rabbit muscle (SigmaChemical Co.) per ml. The microtiter plates were incubatedfor 20 min at 37°C. The color of each well was compared withthose of standards consisting of Todd-Hewitt broth plus 0, 5,10, 20, or 40 mM L-(+)-lactic acid (Sigma Chemical Co.)which were treated as described above. Microtiter wellcultures of strain JH145 (ldh) and the wild-type strain,JH1000, were tested as negative and positive controls,respectively.LDH assay. Cell extracts of S. mutans strains were pre-

pared from 100-ml overnight cultures grown in Todd-Hewittbroth supplemented with 0.5% (wt/vol) glucose. Cells werewashed and resuspended in 1 ml of 0.1 M potassium phos-phate buffer (pH 6.2) and ruptured with a French press at14,000 lb/in2. The extract was clarified by centrifugation at14,000 x g for 30 min at 4°C. LDH activity was assayed asthe pyruvate-dependent oxidation of NADH with and with-out FDP as previously described (7). The values presentedare the level of FDP-dependent NADH oxidation correctedby subtracting the level of FDP-independent NADH oxida-tion. In all cases, the latter represented <5% of the totalactivity. The protein concentrations of cell extracts weredetermined with the BCA protein assay kit (Pierce ChemicalCo., Rockford, Ill.).

RESULTS

In a previous study, a pBR322-based plasmid, plO-5,which carries the wild-type LDH gene from S. mutansJH1000 (7) was isolated. The LDH gene was inactivated (3)by deletion of the promoter and approximately two-thirds ofthe protein coding sequence, followed by insertion of thetetracycline resistance gene from pVA981. In E. coli strainstransformed with the resulting plasmid, p3-5, the LDH genewas completely disabled, as evidenced by their growthphenotype on glucose tetrazolium medium and the absenceof enzyme activity in cell extracts. Plasmid p3-16, which wasused in the studies described here, is identical to p3-5 exceptthat the tetracycline resistance gene is inserted in the oppo-site orientation.

Plasmid p3-16 was digested with SpeI, and a 5.8-kbfragment containing the mutated LDH gene plus 1.0 kb of 5'flanking DNA and 0.7 kb of 3' flanking DNA was recoveredby agarose gel electrophoresis. When 0.5 to 5 pug of thefragment was used to transform JH1000 to tetracyclineresistance, no transformants were obtained. The readilytransformable S. mutans strain NG8 also failed to yieldtetracycline-resistant clones when treated in the same fash-ion.Because of our inability to create the LDH-deficient clone

directly by allelic exchange, we attempted a two-step pro-cess involving the creation of a heterodiploid intermediate.First, since S. mutans is not naturally resistant to ampicillin,p3-16 was partially digested with DraI to remove a 0.7-kbfragment containing the ampicillin resistance gene. Theresulting 10.8-kb fragment was gel purified, religated, andused to transform E. coli MC1061 with selection for tetracy-cline resistance. Colonies which arose following overnightincubation were purified on the selective medium and testedfor ampicillin sensitivity by replica patching. Plasmid DNAfrom four ampicillin-sensitive, tetracycline-resistant cloneswas mapped with restriction enzymes to confirm the pre-dicted deletion of the ampicillin gene.One confirmed plasmid, pDM4, was used to transform

JH1000 (Fig. 1). By using 2 pug of cesium chloride-purifiedDNA, three clones (JDM1, -2, and -3) were obtained follow-ing anaerobic incubation for 2 days on selective mediumcontaining tetracycline. One additional clone (JDM4) wasobtained from aerobically incubated plates. More than 600colonies were obtained when the naturally transformable S.mutans strain NG8 was used as a recipient.

Southern blot analysis of SpeI-digested chromosomalDNA from JDM2, -3, and -4 was performed. When an

SpeI-HpaI fragment of plO-5 containing a portion of theLDH gene which was not deleted in p3-16 was used as a

probe, two bands were observed (Fig. 2B). The size of thelower band (3.3 kb) agreed with the size predicted from thepreviously reported (7) restriction map for the wild-typegene. Only the lower band was observed in DNA fromJH1000, further indicating that this band contained thewild-type gene. Furthermore, in Spel digests of chromo-somal DNA of JDM2, -3, and -4, only this band was reactivewith an HpaI fragment of plO-5 that spanned the portion ofthe LDH gene that was deleted in p3-16 (Fig. 2C). The upperband contained a fragment with the mutated LDH gene, as

evidenced by its reaction when a HinclI fragment of pVA981containing the tetracycline resistance gene was used as a

probe (Fig. 2A). The size of this band (5.0 kb) also agreeswith the prediction based on the size of the wild-typefragment (3.3 kb) minus the size of the deletion (1.7 kb) plusthe size of the pVA981 insert (3.5 kb). These results are

VOL. 62, 1994

on June 19, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

62 HILLMAN ET AL.

x

+S (H) S

JH1 000 chromosomal DNA

S H H H S

==Probe 2 Probe 3Probe 2 Probe 1

I I I l I l0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

KilobasesFIG. 1. ldh heterodiploid construction by transformation of JH1000 with pDM4. Symbols: E, S. mutans DNA; EZa, ldh gene; E , Tetr

gene of pVA981; _, pBR322 DNA. Abbreviations: S, SpeI; H, HpaI. Probes 1, 2, and 3 were used in the blots shown in Fig. 2 and 3.

consistent with the hypothesis that the transformants aroseas a result of Campbell-type recombination and are het-erodiploid for the LDH gene. The wild-type and mutantgenes are located on the chromosome, separated by vectorDNA.Recombination between the mutated and wild-type LDH

genes in heterodiploid transformants should lead to wild-type revertants and LDH-deficient mutants at approximatelyequal frequencies. To demonstrate that these revertants andmutants occur and to determine the frequencies at whichthey occur, we first sought self-recombinants in which themutated LDH gene was eliminated. Three overnight culturesof JDM4 were grown in Todd-Hewitt broth without tetracy-cline, decimally diluted, and spread on brain heart infusionagar. After 2 days of anaerobic incubation, 500 colonies fromeach culture were replica patched onto brain heart infusionagar with or without tetracycline. After overnight incuba-tion, three tetracycline-sensitive clones were identified, onefrom each culture. An in vitro assay of crude cell extractsshowed wild-type levels ofLDH activity for all three clones,as they oxidized 19.13, 18.66, and 18.74 ,umol of NADH permin per mg of protein, compared with 19.26 and 18.39,umol/min/mg for controls JH1000 and JDM4, respectively.Southern blot analysis confirmed that each of these cloneshad a single LDH gene that migrated coincidently with thewild-type gene in JH1000 (Fig. 3A) and that none of themcontained DNA homologous to that of the tetracyclineresistance gene of pVA981 (Fig. 3B). These results areconsistent with the hypothesis that homologous recombina-tion between the mutated and wild-type LDH genes in JDM4had occurred and eliminated the mutated gene in each of thetetracycline-sensitive clones.

_-

*f ,S

Bt:J:"_|fK if 0000 0 *d

14.0

aI0E. .

.i.

1 2 3 4 5 6 7 1 2 3 4 5 6 7 1 2 3 4 5 67

FIG. 2. Southern analysis of chromosomal DNA from JH1000transformed with pDM4 and controls. Lane 1, lambda DNA di-gested with HindIII; lane 2, Tetr gene from pVA981; lanes 3, 4, and5, chromosomal DNA of JDM2, JDM3, and JDM4, respectively,digested with SpeI; lane 6, chromosomal DNA of JH1000 digestedwith SpeI; lane 7, plasmid p10-5 containing the wild-type Idh genedigested with HpaI. The blot was probed with the Tetr gene ofpVA981 (Fig. 1, probe 1) (A), an SpeI-HpaI fragment of plO-5containing a portion of the LDH gene which was not deleted inpDM4 (Fig. 1, probe 2) (B), and an HpaI fragment of plO-5 that wasdeleted in pDM4 (Fig. 1, probe 3) (C).

I A rZZZZZA

INFECT. IMMUN.

on June 19, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

LDH DEFICIENCY IS LETHAL IN STREPTOCOCCUS MUTANS 63

a

1 2 3 4 5 1 2 3 4 5FIG. 3. Southern blot analysis of chromosomal DNA from a

spontaneous tetracycline-sensitive recombinant and controls. Lane1, lambda DNA digested with HindIll; lane 2, Tetr gene frompVA981; lane 3, chromosomal DNA of a spontaneous tetracycline-sensitive recombinant digested with SpeI; lane 4, chromosomalDNA of JH1000 digested with SpeI; lane 5, chromosomal DNA ofJDM4 digested with SpeI. The blot was probed with an HpaIfragment of plO-5 containing a portion of the LDH gene which was

deleted in pDM4 (Fig. 1, probe 3) (A) and reprobed with the Tetrgene of pVA981 (Fig. 1, probe 1) (B).

A similar experiment was performed to isolate self-recom-binants in which the wild-type LDH gene was eliminatedfrom the heterodiploid. Five hundred colonies each fromthree JDM4 cultures were analyzed for the ability to producelactic acid during anaerobic growth in Todd-Hewitt broth as

described in Materials and Methods. A marked difference inlactate production between the LDH-deficient negative con-

trol strain, JH145 (<5 mM), and the wild-type positivecontrol strain, JH1000 (>40 mM), was observed. Withoutexception, the 1,500 JDM4 isolates tested in culture resem-

bled JH1000 with regard to lactic acid content, as indicatedby the amount of colored formazan produced.

In the previous experiment, we failed to find a self-recombinant in which the wild-type LDH gene was elimi-nated. A simpler screening method was employed in an

attempt to isolate this recombinant. Five overnight culturesof JDM4 were diluted to give ca. 3,000 CFU/ml. Samples(0.1 ml) were spread on glucose tetrazolium plates andincubated aerobically for 2 days. More than 80,000 coloniesarose and were screened for the appearance of red colora-tion, which is indicative of LDH deficiency. None were

found to have this coloration.

DISCUSSION

Since it was known that S. mutans JH1000 has a singlechromosomal copy of the gene encoding the FDP-dependentL-(+)-LDH activity (3), we had hoped to replace this gene

with a mutated LDH gene in a single step. However,transformation with a linear DNA fragment containing themutated LDH gene, which was expected to replace thewild-type gene in a double-crossover event, was unsuccess-

ful. The fact that the naturally transformable S. mutans

strain NG8 was also untransformable with this fragmentprovided preliminary evidence that an LDH deficiency islethal in S. mutans under the cultivation conditions em-ployed for selection of transformants.We were able to introduce the mutated LDH gene as part

of a circular DNA molecule which required a single-cross-over event to create heterodiploid transformants. The resultsof Southern blot analysis are consistent with the hypothesisthat the mutated and wild-type genes are situated on thechromosome, separated by vector DNA.One of the heterodiploids was found to undergo sponta-

neous homologous recombination which eliminated the mu-tated LDH gene. This event occurred at a frequency of1/500. The alternative outcome of this recombination event,elimination of the wild-type LDH gene, was not observed in1,500 colonies of the heterodiploid screened for decreasedlactic acid production. Although anaerobic cultivation con-ditions are more likely to be compatible with LDH defi-ciency (see below), we also performed screening whichused tetrazolium medium (which works only under aerobicconditions) to identify LDH-deficient self-recombinants.Screening of more than 80,000 colonies failed to identify asingle colony for which red coloration gave an indication ofdecreased acid production. These results provide additionalsupport for the hypothesis that LDH deficiency is lethal in S.mutans under the cultivation conditions used in the screen-ing.The reliance of S. rattus LDH mutants on acetoin produc-

tion as a major route of pyruvate metabolism, especiallyunder conditions of aerobic growth (6), can be interpreted toindicate that LDH deficiency may impose constraints on themaintenance of cell homeostasis. One possible constraint isindicated by the fact that the biosynthesis of acetoin has noobvious advantage over that of pyruvate itself as an endproduct of metabolism in terms of energetics or balance ofoxidation and reduction. Therefore, acetoin may be formedbecause pyruvate excretion is rate limiting and pyruvate maybe toxic at high intracellular concentrations. Formation ofacetoin from two molecules of pyruvate represents an effi-cient means for the mutant to reduce intracellular levels ofpyruvate.A second possible explanation for the lethal effect of LDH

deficiency may be found in the fact that the conversion ofpyruvate to lactate serves as a major sink for electronsgenerated by glyceraldehyde-3-phosphate dehydrogenase.Anaerobically growing LDH mutants of S. rattus are able tocompensate for their metabolic deficiency by increasing theirproduction of ethanol (6). Under aerobic conditions, thispathway is less utilized, so other sinks for electron disposalmust be employed. Higuchi (4) demonstrated that the abilityof S. rattus to grow on mannitol in the presence of oxygencorrelated extremely well with high levels ofNADH oxidaseand superoxide dismutase.Most strains of S. mutans appear to have less flexibility for

dealing with excess reducing equivalents generated by gly-colysis, particularly under aerobic conditions. It was shown(4) that the growth of two of three S. mutans strains onmannitol was severely inhibited by oxygen, and this corre-lated with a relative lack of NADH oxidase and superoxidedismutase activities. If the addition of a single excess reduc-ing equivalent per molecule of substrate utilized, as in thecase of mannitol consumption, has toxic effects on wild-typeS. mutans, an excess of two electrons generated by glucosemetabolism in an LDH-deficient mutant would likely havelethal consequences.

Like other streptococci, S. rattus and S. mutans appear to

VOL. 62, 1994

.5. -. .' 1.11,., I

on June 19, 2018 by guesthttp://iai.asm

.org/D

ownloaded from

64 HILLMAN ET AL.

share the same pathways for pyruvate metabolism, as indi-cated by the similarity in their end products under variousculture conditions. It is likely, therefore, that S. rattus isable to accommodate a deficiency in LDH while S. mutans isnot because the species differ somehow in their regulation ofthe enzyme activities that catalyze these pathways. Giventhe flexibility afforded by multiple pathways for pyruvatemetabolism (10), we predict that culture conditions whichwill be permissive for LDH deficiency in S. mutans will beable to be found. Experiments are under way to test thisprediction.

ACKNOWLEDGMENT

This work was supported by Public Health Service grantDE04529 from the National Institute of Dental Research.

REFERENCES1. Abhyankar, S., H. J. Sandham, and K. H. Chan. 1985. Serotype

c Streptococcus mutans mutable to lactate dehydrogenase de-ficiency. J. Dent. Res. 64:1267-1271.

2. Casadaban, M. J., and S. N. Cohen. 1980. Analysis of genecontrol signals by DNA fusion and cloning in Escherichia coli.J. Mol. Biol. 138:179-207.

3. Duncan, M. J., and J. D. Hillman. 1991. DNA sequence and invitro mutagenesis of the gene encoding the fructose-1,6-diphos-phate-dependent L-(+)-lactate dehydrogenase of Streptococcusmutans. Infect. Immun. 59:3930-3934.

4. Higuchi, M. 1984. The effect of oxygen on the growth andmannitol fermentation of Streptococcus mutans. J. Gen. Micro-biol. 130:1819-1826.

5. Hillman, J. D. 1978. Lactate dehydrogenase mutants of Strep-

tococcus mutans: isolation and preliminary characterization.Infect. Immun. 21:206-212.

6. Hillman, J. D., S. W. Andrews, and A. L. Dzuback. 1987.Acetoin production by wild-type strains and a lactate dehydro-genase-deficient mutant of Streptococcus mutans. Infect. Im-mun. 55:1399-1402.

7. Hillman, J. D., M. J. Duncan, and K. P. Stashenko. 1990.Cloning and expression of the gene encoding the fructose-1,6-diphosphate-dependent L-(+)-lactate dehydrogenase of Strepto-coccus mutans. Infect. Immun. 58:1290-1295.

8. Hillman, J. D., A. L. Dzuback, and S. W. Andrews. 1987.Colonization of the human oral cavity by a Streptococcusmutans mutant producing increased bacteriocin. J. Dent. Res.66:1092-1094.

9. Hillman, J. D., K. P. Johnson, and B. I. Yaphe. 1984. Isolationof a Streptococcus mutans strain producing a novel bacteriocin.Infect. Immun. 44:141-144.

10. Iwami, Y. 1988. Influence of environment on glycolysis instreptococci. Kitasato Arch. Exp. Med. 61:1-20.

11. Johnson, C. P., S. M. Gross, and J. D. Hillman. 1980. Cariogenicpotential in vitro in man and in vivo in the rat of lactatedehydrogenase mutants of Streptococcus mutans. Arch. OralBiol. 25:707-713.

12. Lee, S. F., A. Progulske-Fox, G. W. Erdos, D. A. Piacentini,G. Y. Ayakawa, P. J. Crowley, and A. S. Bleiweis. 1989.Construction and characterization of isogenic mutants of Strep-tococcus mutans deficient in major surface protein antigen P1(I/II). Infect. Immun. 57:3306-3313.

13. Maniatis, T., E. F. Fritsch, and J. Sambrook 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

14. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

INFECT. IMMUN.

on June 19, 2018 by guesthttp://iai.asm

.org/D

ownloaded from