Subtilis Enzyme Research Articles

The glutamine-utilizing site of Bacillus subtilis glutamine phosphoribosylpyrophosphate amidotransferase.

Reaction of Bacillus subtilis glutamine phosphoribosylpyrophosphate amidotransferase with 6-diazo-5-oxo-L-norleucine resulted in complete loss of its ability to catalyze glutamine-dependent phosphoribosylamine formation and its glutaminase activity, whereas its ability to catalyze ammonia-dependent phosphoribosylamine formation and to hydrolyze phosphoribosylpyrophosphate was increased. The site of reaction with 6-diazo-5-oxo-L-norleucine was the NH2-terminal cysteine residue. The NH2-terminal sequence of the B. subtilis enzyme was homologous with that of the corresponding amidotransferase from Escherichia coli, for which the NH2-terminal cysteine is also essential for glutamine utilization (Tso, J. Y., Hermodson, M. A., and Zalkin, H. (1982) J. Biol. Chem. 257, 3532-3536). The fact that the metal-free E. coli amidotransferase contains a glutamine-utilizing structure that is very similar to that found in B. subtilis amidotransferase, which contains an essential [4Fe-4S] center, indicates that the iron-sulfur center probably plays no role in glutamine utilization.

Structure, function, and regulation of amidophosphoribosyltransferase from prokaryotes

Recent studies on the structure, function and regulation of amidophosphoribosyltransferase from E. coli and B. subtilis are reviewed and these properties compared with those of the enzyme from eukaryotes. The availability of large amounts of stable enzyme from the two microbial sources has facilitated the recent studies. The enzyme subunits from E. coli and B. subtilis are of similar size, 56, 395 and approximately 50,000 respectively. Catalytic properties and patterns for allosteric inhibition are similar but not identical. There are two major differences between these enzymes. In contrast to the enzyme from E. coli, B. subtilis amidophosphoribosyltransferase contains an essential Fe-S center. In addition, the enzyme from B. subtilis but not E. coli is inactivated in stationary phase by an oxygen-dependent mechanism which appears to have regulatory significance. As a consequence of the Fe-S center B. subtilis amidophosphoribosyltransferrae is oxygen-sensitive in vitro. Amidophosphoribosyltransferase from mammalian sources is similar to the B. subtilis enzyme in its oxygen-sensitivity which may result from an Fe-S center. The amino acid sequence of E. coli amidophosphoribosyltransferase was deduced from the DNA sequence of the purF structural gene. The primary translation product contains 504 amino acid residues. Met-1 is removed by processing leaving an NH 2-terminal cysteine residue. The NH 2-terminal cysteine was specifically alkylated by the glutamine affinity analog 6-diazo-5-oxonoreleucine and is thus identified as the cysteine residue involved in formation of the glutaminyl-enzyme covalent intermediate. The mechanism for glutamine utilization appears identical to other glutamine amidotransferases. Sequence homology was not detected in the glutamine amide transfer domains of E. coli anthranilate synthase and amidophosphoribosyltransferse.

Functional homology of Bacillus subtilis methyltransferase II and Escherichia coli cheR protein.

A chemotaxis-related methyltransferase enzyme from Bacillus subtilis has been shown to methylate membrane-bound proteins in Escherichia coli in vitro. The methylated proteins are in the same molecular weight range as authentic E. coli methyl-accepting chemotaxis proteins. It was also shown that wild type E. coli cytoplasmic extract could methylate membrane proteins from B. subtilis in its methyl-accepting chemotaxis protein region. Cytoplasmic extracts from methyltransferase mutants of either species could methylate neither set of methyl-accepting proteins in vitro. The B. subtilis enzyme was incapable of methylating any of a group of soluble eucaryotic proteins. These data suggest functional homology between B subtilis methyltransferase II and E. coli cheR protein (chemotaxis methyltransferase) despite the evolutionary divergence between these two species.

Specificity of promoter site utilization in vitro by bacterial RNA polymerases on Bacillus phage phi 29 DNA. Transcription mapping with exonuclease III.

Bacillus subtilis RNA polymerase holoenzyme transcribes phi 29 DNA in vitro producing five major RNA species defined by characteristic electrophoretic mobilities. In addition to these products, Escherichia coli RNA polymerase transcribes phi 29 DNA to yield three RNA species not detected when transcribing with the B. subtilis enzyme under the same optimal reaction conditions for RNA synthesis. Transcriptional analysis of purified restriction fragments and exonuclease III-digested DNA established locations of six promoter and three termination sites defining the eight transcripts. The transcription map shows that E. coli RNA polymerase initiates transcription at three sites not efficiently utilized by the B. subtilis enzyme. However, initiation by the B. subtilis polymerase from at least two of these sites could be detected at E:DNA ratios greater than 10 in the absence of competing promoters. These results indicate that differences between the two polymerases in promoter utilization are not explained by specificity of promoter binding, but represent differences in responding to promoter strength. Transcription of phi 29 DNA and T7 DNA by E. coli core polymerase with either B. subtilis or E. coli sigma subunits results in formation of transcripts identical with those produced by E. coli holoenzyme, suggesting that core polymerase contains elements important in determining relative promoter strength. The efficiency of rifampicin-resistant complex formation on phi 29 and T7 promoters is also dependent upon the source of core polymerase.

L-Alanine dehydrogenase from Thermus thermophilus

A heat-stable l-alanine dehydrogenase was isolated and purified from the extremely thermophilic microorganism, Thermus thermophilus, by affinity chromatography. The enzyme has a molecular weight of 290 000, as determined by the sedimentation equilibrium method, and is composed of six subunits of identical molecular weight as concluded from sodium dodecyl sulphate gel electrophoresis. The enzyme has been characterized in terms of pH- and substrate concentration-dependence of activity, substrate specificity, inhibition of d-alanine and d-cysteine and amino acid composition. The parameters obtained are very similar to those reported for l-alanine dehydrogenase from the mesophilic microorganism, Bacillus subtilis (Yoshida, A. and Freese, E. (1965) Biochim. Biophys. Acta 96, 248–262). The thermal stability of the T. thermophilus enzyme is much greater than that of the B. subtilis enzyme. Activation free energy ( ΔG ‡ ), activation enthalpy ( ΔH ‡ ) and activation entropy ( ΔS ‡ ) values were determined for both the alanine deamination and for the heat inactivation reactions of the thermophilic and mesophilic enzymes. The values obtained for the catalytic reaction were practically equal. However, the two enzymes differed significantly in these parameters determined for the enzyme inactivation, which indicates that the factors ensuring the thermo-resistance of the enzyme from T. thermophilus do not affect enzyme activity.

Regulation of prephenate dehydratase in coryneform species of bacteria by l-phenylalanine and by remote effectors

Corynebacterium glutamicum, Brevibacterium flavum, and B. ammoniagenes utilize the phenylpyruvate branchlet as the sole means of l-phenylalanine biosynthesis. Enzymological aspects of l-phenylalanine biosynthesis were generally similar in the three microbial species selected to represent the coryneform grouping. Thus, the general picture of aromatic biosynthesis in coryneform bacteria is very similar to that found in species of bluegreen algae, i.e., phenylpyruvate branchlet for l-phenylalanine biosynthesis and pretyrosine branchlet for l-tyrosine biosynthesis ( A. Fazel and R. A. Jensen, 1979, J. Bacteriol. 138, 805–815). Prephenate dehydratase is subject to repression control and to feedback inhibition by l-phenylalanine ( K i value range: 35–95 μ m). Extrapathway effectors activated ( l-tyrosine) or inhibited ( l-tryptophan) prephenate dehydratase, reminiscent of the action of remote effectors with Bacillus subtilis prephenate dehydratase. A labile activation of prephenate dehydratase by l-methionine was also found in C. glutamicum. Prephenate dehydratase copurifies with an aromatic aminotransferase which is capable of transaminating phenylpyruvate, 4-hydroxyphenylpyruvate, or prephenate in vitro. The possibility of a multifunctional protein was ruled out, however, by results obtained with l-phenylalanine auxotrophs lacking prephenate dehydratase and a regulatory mutant constitutive for prephenate dehydratase. Other classes of regulatory mutants altered in prephenate dehydratase lost sensitivity to inhibition by l-phenylalanine (and l-tryptophan), or produced an enzyme that was hypersensitive to l-tyrosine activation. Molecular weight estimates of the prephenate dehydratase enzymes isolated from the three species studied ranged from 200,000–260,000. Unlike the B. subtilis enzyme, neither allosteric (positive or negative) effectors nor regulatory mutations altered the molecular weight of prephenate dehydratase. However, the intrinsic activity of prephenate dehydratase and its sensitivity to allosteric effectors was altered substantially following the initial step of partial purification.

Bacillus subtilis deoxyribonucleic acid gyrase.

Bacillus subtilis 168 was shown to contain a deoxyribonucleic acid (DNA) gyrase activity which closely resembled those of the enzymes isolated from Escherichia coli and Micrococcus luteus in its enzymatic requirements, substrate specificity, and sensitivity to several antibiotics. The enzyme was purified from the wild type and nalidixic acid-resistant and novobiocin-resistant mutants of B. subtilis and was functionally characterized in vitro. The genetic loci nalA and novA but not novB were shown to code for portions of the functional gyrase. Enzyme from the antibiotic-resistant mutants was resistant to the drug in vitro. The most striking observation was the remarkable similarity between the B. subtilis enzyme and other DNA gyrases, especially with respect to the oxolinic acid-induced DNA cleavage in the presence of sodium dodecyl sulfate. All of the enzymes appeared to possess the same specificity of cutting sites regardless of the source or type of DNA used. This result implies that gyrase binding to DNA is highly specific.

The nature of transcription selectivity of bacteriophage SPO1-modified RNA polymerase.

Escherichia coli and Bacillus subtilis RNA polymerase have almost identical transcription specificities on bacteriophage SPO1 DNA when assayed in a coupled transcription-translation cell free system. SPO1-modified B. subtilis RNA polymerase has altered transcription specificity. It is shown that rifampicin-inhibited E. coli RNA polymerase can completely block transcription of SPO1 DNA by rifampicin resistant B. subtilis enzyme, whereas it has no effect on transcription by SPO1-modified B. subtilis RNA polymerase. We conclude that the new transcription by SPO1-modified RNA polymerase results from newly recognized promoters, rather than by elongation of transcripts which could also be made by B. subtilis vegetative RNA polymerase.

Utilization of promoter and terminator sites on bacteriophage T7 DNA by RNA polymerases from a variety of bacterial orders

The transcriptional properties of bacterial RNA polymerases purified from seven different species and representing a variety of bacterial orders have been studied using the well characterized DNA from phage T7 as template. The subunit composition of the different preparations suggests that each RNA polymerase holoenzyme has a promoter structure (betabeta'alpha2sigma) similar to that of the well studied E. coli and B. subtilis enzymes. Each enzyme utilizes DNA from bacteriophage T7 as an effective template for RNA synthesis, although all preparations contain a substantial fraction of inactive enzyme. Electrophoretic analysis of the RNA products made with the different RNA polymerases in vitro using T7 DNA (deletion mutant deltaD111) as template reveals that with minor exceptions, all of the heterologous RNA polymerases utilize the same collection of promoter sites on T7 used by the E. coli host enzyme, and only those promoter sites. The T7 early terminator is also efficiently utilized by each enzyme. Since the different bacterial species from which the RNA polymerases were derived are genetically quite distant, it appears that there is a structural element in the promoter which governs its recognition and which is universally recognized among RNA polymerases of different bacterial species. While the different bacterial RNA polymerases generally utilize the same set of T7 promoter sites, the efficiency of utilization of the different promoters varies considerably for different RNA polymerases and for different reaction conditions with the same RNA polymerase. Hence although the several T7 promoter sites share the ability to be recognized by bacterial RNA polymerases, each shows a unique pattern of utilization and therefore must possess a unique element of promoter structure as well. It has previously been shown (Stahl and Chamberlin, 1977) that T7 promoters A1, C, D and E interact differently with E. coli RNA polymerase as judged by the properties of complexes formed between each promoter and the latter enzyme. Since competition takes place among different promoter sites on a template and since these sites can differ functionally, small changes in reaction conditions or in the structure of the RNA polymerase can lead to significant changes in the rate of utilization of different promoter sites even when these promoter sites share common elements. Because the T7 promoters and terminator are utilized efficiently by such a wide range of RNA polymerases and because each of the several T7 promoters possesses unique properties which govern its utilization by RNA polymerase, analysis of the transcripts formed on a T7 DNA template provides a simple and rapid procedure for detecting and analyzing alterations in bacterial RNA polymerases which affect promoter or terminator recognition or utilization.

Stabilization of proteolytic enzymes in solution

AbstractThe stability of the neutral and alkaline proteases in a Bacillus subtilis enzyme mixture was studied in aqueous solutions at room temperature. Stabilization of the proteases in solution for periods up to 25 days was achieved by the addition of various protein preparations including casein and soya protein. The degree of stabilization by casein was concentration dependent to about 2% protein. The instability of the neutral protease in solutions of the B. subtilis enzyme mixture was shown to be due primarily to proteolysis by the alkaline protease since the diisopropylfluorophosphate‐treated enzyme was quite stable. Formulation of such enzyme solutions at low pH gave greater stability as did solutions containing an alkaline protease inhibitor from potatoes. A Conceptual approach to the formulation of enzyme solutions containing proteolytic enzyme to ensure maximum stability is proposed.

Subunit contacts of the rifamycin binding site of RNA polymerase (B. subtilis)

The radiolabeled RNA polymerase inhibitors 3-(2-bromoacetamidoethyl)-thiorifamycin and 3-(2-acetamidoethyl)-thiorifamycin quinone have been prepared and covalently attached to the B. subtilis enzyme under mild conditions. Analysis of the subunits labeled indicates that regions of subunits sigma, beta, and beta-prime lie within about 7Å of the 3-position of rifamycin bound to the enzyme, while subunits alpha , beta, and beta-prime lie near the aromatic rings. The results of this work imply that the rifamycin binding site lies near the center of an arrangement involving at least four of the subunits of RNA polymerase. This idea is supported by the results of other studies involving cross-linking of subunits and also rifamycin binding to subunit complexes.

Hydroxylaminolysis of penicillin binding componenets is enzymatically catalyzed.

The hydroxylaminolysis of the penicilloyl moiety from [14C]penicillin G binding component (PBC) complexes of the Bacillus subtilis D-alanine carboxypeptidase and of the mixture of PBC's of Staphylococcus aureus was inhibited by denaturation of the complexes by heat (55 degrees), detergent (1% sodium dodecyl sulfate), or trichloroacetic acid. The kinetics of inhibition by denaturation were comparable to those of the inhibition of [14C]penicillin G binding to the PBC's and of carboxypeptidase activity of the B. subtilis enzyme under identical denaturing conditions. These data establish that the hydroxylaminolysis is an enzymatically catalyzed process suggesting that penicillin G is bound to an enzymatically active site. Treatment of the denatured [14C]penicillin G-carboxypeptidase complex with sodium borohydride or at pH 12 resulted in the release of the penicilloyl moiety. These results are consistent with a carboxylic ester bond for the penicilloyl-PBC instead of a thiolester linkage as was initially presumed.

Dihydrodipicolinate Reductases from <italic>Bacillus cereus</italic> and <italic>Bacillus megaterium</italic>

Dyhydrodipicolinate reductases were purified 100-fold from crude extracts of B. cereus and B. megaterium and their properties were compared with those of the reductase from B. subtilis. The molecular weights of the reductases of B. cereus and B. megaterium were fount to be 155,000 and 150,000, respectively. These reductases were shown to be free of flavin, unlike the B. subtilis enzyme, which contains flavin. Both NADPH and NADH acted as coenzymes for these two reductases. NADPH being three or four times more effective than NADH. The Km values for NADPH and dihydrodipicolinate were 8 micrometer and 62 micrometer, respectively, with B. cereus reductase, and 13 micrometer and 59 micrometer with B. megaterium reductase. The pH optima of the enzymes from B. cereus and B. megaterium were pH 7.4 and 7.2, respectively. The reductases were inhibited by dipicolinate noncompetitively with respect to dihydrodipicolinate and the Ki values were 85 micrometer and 140 micrometer, respectively. Lysine and diaminopimelate were not inhibitory. The properties of the reductases from B. cereus and B. megaterium were similar, but they differed considerably from those of the B. subtilis enzyme. However, all three Bacillus reductases were markedly inhibited by dipicolinate, unlike the enzyme from E. coli.

Sodium effect of growth on aspartate and genetic analysis of a Bacillus subtilis mutant with high aspartase activity

Most strains of Bacillus subtilis, dervied from the 168 (Marburg) strain, grow slowly on aspartate as sole carbon source. We isolated a mutant (aspH) that grows rapidly on aspartate because it produces aspartase constitutively. Thus, aspartase is needed for rapid growth on aspartate, whereas aspartate-alpha-ketoglutarate aminotransferase is not needed, as was demonstrated by a mutant lacking that enzyme activity. By two--and three-factor crosses using PBSl transduction, the aspH mutation was located between the aroD and the lys markers of the genetic map. Although sodium ions do not affect growth on glucose or L-malate, they specifically stimulate growth on aspartate in both the parent and the aspH mutant strains. Enzyme activities of crude aspartase and fumarase and of purified aspartase do not increase in the presence of sodium. These results show that stimulation by sodium involves some reaction other than the enzymes catabolizing aspartate. The ease of purification from the aspH strain and the stability of aspartase suggest that the B. subtilis enzyme is particularly useful for aspartate determinations.

Inactivation of bacterial D-amino acid transaminases by the olefinic amino acid D-vinylglycine.

D-Vinylglycine (2-amino-3-butenoate) functions as a transamination substrate and irreversible inactivator of the homogeneous pyridoxal phosphate-dependent D-amino acid transaminases from Bacillus subtilis and Bacillus sphaericus. In the absence of alpha-ketoglutarate as co-substrate, vinyl-glycine causes little if any inactivation of either enzyme; in the presence of excess alpha-ketoglutarate, both enzymes are inactivated with pseudo-first order kinetics. The limiting rate constant for inactivation of the B. sphaericus enzyme is 1.9 min-1, for the B. subilis enzyme it is 0.36 min-1. The number of catalytic events before inactivation is about 450 for the B. sphaericus enzyme and about 800 for the B. subtilis enzyme; that is, about 0.2% inactivation in each catalytic cycle for the former enzyme and 0.15% for the latter. Comparisons are made with the L-aspartate amino-transferase from pig heart which is inactivated completely in one catalytic cycle and the L-alanine aminotransferase which is not inactivated in many cycles. Comparisons are also made between the likely mode of D-transaminase inactivation produced by vinylglycine and the mode of inactivation induced by beta-chloro-D-alanine.

Comparative studies of RNA polymerase subunits from various bacteria

The molecular structure of RNA polymerases from Escherichia coli, Salmonella typhimurium, Salmonella anatum,serratia marcescens, Aerobacter aerogens, Proteus mirabilis and Bacillus subtilis were compared based on:i) inhibition of the enzyme activity by treatment with antibodies against E. coli RNA polymerase subunits;ii) analysis of antibody precipitates by sodium ododecyl sulfatepolyacrylamide gel electrophoresis; and iii) analysis of antibody precipitates by urea-isoelectrofocusing followed by sodium dodecyl sulfate-slab gel electrophoresis in the second dimension. All the bacterial RNA polymerases examined cross-react equally with anti-E. COLI HOLOPOLYMERASE BUT EXHIbit different extents of cross-reaction with antibodies against individual subunits. Except for B. subtilis RNA polymerase, the molecular weight and isoelectric point of the enzyme subunits are close to those of E. coli polymerase. However, minor difference were found at least within the resolution of the techniques employed:S. anatum polymerase has sigma subunit larger than E. coli sigma subunit; P. mirabilis enzyme has sigma subunit larger in size and more acidic in charge, and alpha subunit smaller and more basic than corresponding E. coli subunits. The electrophoretic map of B. subtilis enzyme subunits is completely different from that of E. coli enzyme.

Diagnostic tests in the skin and serum of workers sensitized to Bacillus subtilis enzymes*†‡

Two allergen pools of commercial detergent enzymes were prepared as skin test reagents: (1) Carlsberg type, composed of three products containing subtilopeptidase A, and (2) BPN type, composed of two products containing subtilopeptidase B and alpha-amylase. In 100 non-exposed controls a reaction suggesting primary irritancy was found at protein concentrations greater than 1 microng/ml intradermally or 1 mg/ml by prick test. Intradermally at 10 microng/ml weals were accompained by less pronounced flare reactions than observed in specifically sensitized enzyme workers. At 100 micronh/ml the reactions were like strong specific reactions. Galse positive prick test reactions occurred irregularly at 10 mg/ml. In 100 sensitized enzyme workers, reactions were elicited at concentrations from 1-0 to 10(-5) microng/ml intradermally and from 1000 to 1 micron by prick test. Intradermal and prick tests correlated well (r=0-84, P less than 0-001). Ratings of symptom severity upon exposure obtained from questionnaires were significantly correlated with skin test reactivity (P less than 0-01). RAST performed on sera collected simultaneously also correlated significantly with symptom scores. PCA tests in monkeys were less sensitive. Standardized test reagents allow diagnostic skin testing by either intradermal or prick test in B. subtilis enzyme sensitive patients. A clear distinction between primary irritant reactions and true sensitization was made on the basis of the concentration required to elicit a reaction.

Antigenic reactivity of inactivated Bacillus subtilis enzyme preparations

The Bacillus subtilis preparations to which detergent industry workers have been exposed contain a number of enzymes including amylase and the alkaline proteinases, SPA, and SPB. Proteinase-dependent properties of these preparations are a potential source of confusion in some immunologic tests, so that the use of inactivated preparations may be preferable. It was shown that a selective chemical blockage of the active site of SPA is associated with a preservation of antigenic reactivity; however, this preparation has not been used in humans. Heat inactivation of the enzyme activities of SPA- and SPB-containing preparations led to a corresponding loss of antigenic reactivity. The macroglobulin inhibitors of rabbit serum represent an example of naturally occurring proteinase inhibitors. SPA mixed with preparations of these serum inhibitors in inhibitor excess suffered complete or almost complete loss of antigenic reactivity. However, when reaction mixtures were formed in enzyme excess, there was at least partial preservation of the antigenic reactivity of macroglobulin-associated SPA. These properties of inhibited B. subtilis enzyme preparations require consideration in their use in immunologic procedures.

Teichoic acid synthesis in Bacillus stearothermophilus.

1. Particulate enzyme preparations obtained from Bacillus stearothermophilus B65 by digestion with lysozyme were shown to catalyse teichoic acid synthesis. With CDP-glycerol as sole substrate the preparations synthesized 1,3-poly(glycerol phosphate). It was characterized by alkaline hydrolysis, by glucosylation to the alkali-stable 2-glucosyl-1,3-poly(glycerol phosphate) with excess of UDP-glucose and a Bacillus subtilis Marburg enzyme system, by degradation of this latter product with 60%HF and periodate oxidation of the resulting glucosylglycerol. The specificity of the B. subtilis system previously reported (Glaser & Burger, 1964), was confirmed in the present work. 2. Pulse-labelling experiments, followed by periodate oxidation of the product and isolation of formaldehyde from the glycerol terminus of the polymer, showed that the B. stearothermophilus enzyme system transferred glycerol phosphate units to the glycerol end of the chain. The transfer reaction was irreversible. It was not determined if these poly(glycerol phosphate) chains were synthesized de novo, but it was shown that the newly synthesized oligomers were bound to much larger molecules. 3. When the B. stearothermophilus enzyme system was supplied with both CDP-glycerol and UDP-glucose, 1-glucosyl-2,3-poly(glycerol phosphate) was synthesized in addition to the 1,3-isomer. The former polymer was characterized by acid and alkaline hydrolysis, degradation with HF and periodate oxidation of the resulting glucosylglycerol, and periodate oxidation of the intact polymer followed by mild acid hydrolysis. This latter procedure removed the glucose substituents without disrupting the poly(glycerol phosphate) chain. 4. The poly(glycerol phosphate) isomers were distinguished by glucosylation with the B. subtilis enzymes and alkaline hydrolysis, the 2,3-isomer remaining alkali-labile. The proportion of 2,3-poly(glycerol phosphate) in the product increased with increasing amounts of UDP-glucose in the incubation mixture, but the total glycerol phosphate incorporated into products remained constant. It is suggested that the synthetic pathways of the two poly(glycerol phosphate) species may share a rate-limiting step.

Further Genetic and Enzymological Characterization of the Three Bacillus subtilis Deoxyribonucleic Acid Polymerases

Bacillus subtilis has three DNA polymerases. Polymerase III is necessary for DNA replication while polymerase I is involved in DNA repair. Four independent mutations leading to loss of polymerase I activity were mapped by transformation between mdh and phoP. Since one of these mutant strains produces a heat-labile polymerase this locus is likely the polymerase I structural gene. Some polymerase III mutants fail to grow at elevated temperatures and have a greatly increased mutation rate. The structural genes for polymerases I and III are unlinked. Purified preparations of the three B. subtilis DNA polymerases have been characterized. Polymerases II and III catalyze a high rate of synthesis in the absence of a full deoxyribonucleoside triphosphate complement, the mechanism of which is probably a truncation of the normal synthetic process. Arabinosylcytosine triphosphate competitively inhibits polymerases II and III as well as DNA synthesis in toluene-treated cells. This analogue is incorporated by these polymerases. Single-stranded DNA is a potent inhibitor of polymerases II and III but not I. The B. subtilis polymerases are readily separable by column chromatography and respond differently to high ionic strength, heat, sulfhydryl reagents, arylazopyrimidines, and various templates. The differences and similarities of the three B. subtilis enzymes relative to the correspondingly numbered Escherichia coli enzymes are discussed.