Bacillus subtilis - American Society for Microbiology

JOURNAL OF BACTREOLOGY, OCt. 1976, p. 174-181 Copyright X: 1976 American Society for Microbiology

Vol. 128, No. 1 Printed in U.S.A.

Pleiotropic, Extragenic Suppression of dna Mutations in Bacillus subtilis ANTONIO G. SICCARDI,' SILVIA OITOLENGHI, ADRIANA FORTUNATO, AND GIORGIO MAZZA* Istituto di Genetica dell'Universitc di Pavia* and Laboratorio di Genetica Biochimica ed Evoluzionistica del Consiglio Nazionale delle Richerche, 27100 Pavia, Italy Received for publication 14 April 1976

Thermosensitive (dna) mutants of Bacillus subtilis defective in deoxyribonucleic acid replication can be divided into two groups on the basis of their ability to spontaneously yield secondary mutants with an HDS phenotype (thermoinsensitivity and resistance to aryl-azo-pyrimidines) at frequencies higher than 10-8. Such a phenotype is due to alleles at the hds locus (mapping close to cysA), which act as extragenic pleiotropic suppressors. HDS suppressibility has been used as a screening tool to identify new dna strains among uncharacterized temperature-sensitive mutants.

The replication of chromosomal deoxyribonucleic acid (DNA) in Bacillus s-ubtilis depends upon DNA polymerase III (PolIII) (2, 3, 8, 10) and on a number of other functions that have been genetically identified by thermosensitive (ts) mutants at dna loci (1, 4, 11, 12, 14, 16, 22, 25) whose products are involved either in initiation or elongation of DNA chains. Some of these proteins might interact with the PoIIII and/or regulate its catalytic action. Studies on the mode of action of the aryl-azo-pyrimidines 6-(phydroxyphenylazo)-uracil (OHPhN.2 Ura) and 6-(p-hydroxyphenylazo)-2-amino, 4-keto-pyrimidine (OHPhN2 Iso) have established that these act as competitive inhibitors of PolIII (7, 10, 18) and that some mutants resistant to both analogues contain an in vitro resistant enzyme (6, 8). We have found that the reversion frequency of dna mutants to ts+ phenotypes is so high (usually between 10-4 and 0lff) that it is not likely to be accounted for by back-mutation alone; it is also a common finding that most ts+ revertants still carry the original ts allele (whose persistence can be demonstrated by genetic analysis) and are thus due to "suppressor" mutations. We have isolated Dna+ revertants from several different dna strains, selecting simultaneously for temperature insensitivity and arylazo-pyrimidine (OHPhN2 Ura and OHPhN2 Iso) resistance (10 ,uM each; HP resistance), with the aim of restricting our analysis to suppressor mutations in functions directly related to PolIlI and thus, indirectly, to the replication

complex. I Present address: Cattedra di Microbiologia II, FacoltA di Medicina, UniversitA degli Studi di Pavia, Pavia, Italy.

Some of the observed suppression of dna phenotypes might conceivably be exerted through the stabilizing effect of protein-protein interactions within the replication complex itself; i.e., it might represent the effect of secondary mutations in other components of the replication complex. The suppression might be exerted by PoIIII itself or by other proteins interacting with both the polymerase and the affected dna function. Such "suppressors" might be an additional tool in the study of the replication complex composition and function. MATERIALS AND METHODS Bacterial strains and bacteriophages. The origin and the description of the strains used in this work are reported in Table 1. Phage PBS-1 was used for transduction experiments (26). Culture media. Spizizen minimal medium MT (23) was used to prepare competent cells. Medium Y (24), Penassay broth (antibiotic medium no. 3, Difco), and tryptose blood agar base (Difco) were used in transduction experiments. The minimal medium of Davis and Mingioli (9) was used for selection of recombinants. Bouillon-Nutritif Complet (Bio Kar) solidified with agar (nutrient agar) was used to determine cell titer and drug sensitivity. Reagents. Nalidixic acid and actinomycin D were purchased from Serva Feinbiochemica, Heidelberg; OHPhN2 Ura and OHPhN2 Isocytosine were obtained from B. W. Langley (Imperial Chemical Industries, Macclesfield, U.K.); mytomycin C was from Schuchard Corp., Munich; streptomycin was from Squibb, Rome; methyl methane sulfonate was from Eastman Organic Chemicals Corp., Rochester, N.Y.; 5-bromouracil was from Sigma Chemical Co., St. Louis, Mo.; polymyxin B was from Schwarz/ Mann, Orangeburg, N.Y.; and rifampin (rifampicina) was obtained from Lepetit S.p.A., Milan. Transformation. Competent cells were prepared as described by Stewart (24). DNA was prepared 174

Strain E3113 E3119 SB1084 E3130 E3123 E3142 BD305 PB2346 E3105 E3146 E3102 168TT (dna-1) VUB899 PB2480 PB2481 PB2482 BD288 BD54 (azp-12) PB1424 PB1642 BC26 BC34 PB3409 BD54 PB2510 PB2513 PB2561 PB2562 PB2563 PB2564 PB2565

175

HDS SUPPRESSION IN B. SUBTILIS

VOL. 128, 1976

TABLE 1. List of B. subtilis strains used in this work Genotype ile-1 metB5 dnaAl3 ile-1 metB5 dnaB19 thyAB trpC2 dnaC6 ile-1 metB5 dnaC30 ile-1 metB5 dnaD23 ile-1 metB5 dnaE42 ile-1 metB5 dnaF133 thyAB arg dnaF69 ile-1 metB5 dnaG5 ile-1 metB6 dnaH46 ile-1 metB5 dnaI2 thyAB trpC2 dna-1 thyAB leu-8 dna-8132 hisB2 trpC2 met dna-80 hisB2 trpC2 met dna-81 hisB2 trpC2 met dna-82 ile-1 metB5 polA 1443 mut-1 (ts) ile-1 metB5 leu-8 azp-12 hisB2 trpC2 met polA42 tyrAl hisB2 trpC2 aroB2 pheA12 argA3 ery pyrA26 purA16 cysAl4 leu-8 ile-1 metB5 leu-8 thyAB trpC2 hds-10 ile-1 metB5 hds-13 ile-1 metB5 hds-61 ile-1 metB5 hds-62 ile-1 metB5 hds-63 arg thyAB hds-64 ile-1 metB5 hds-65

according to the method described by Marmur (19). The transformation procedure was as previously described (20). DNA synthesis. DNA synthesis was measured by following the incorporation of [3H]cytidine deoxyribonucleotide into alkali-stable, acid-insoluble material. Samples (0.5 ml) were added to 0.5 ml of 2 N NaOH and left for 18 h at room temperature. The samples were then precipitated by the addition of 100 1.A of trichloroacetic acid (50%) and, after standing at 40C for 30 min, were collected on glass-fiber filter paper (Whatman GF/C) and washed three times with 10-ml portions of cold 2 N HCl and, finally, with 2 ml of cold 95% ethyl alcohol. The radioactivity was determined in a liquid scintillation counter. Genetic mapping. The linkage relationships with markers of known location on the genetic map were determined by PBS-1 transduction, performed according to Hoch et al. (13). The HDS phenotype (hds markers; see Results) was tested on nutrient agar plates containing 10 ,uM each HPUra and HPIsocytosine. Recombinants for auxotrophic markers were selected on minimal medium supplemented with 0.5% glucose and 25 ,ug of the appropriate auxotrophic requirements per ml. After 2 days of incubation at 370C, recombinants were picked, reisolated, and tested for their phenotype. rfm (rifampin-resistant) recombinants were selected by the overlay technique of Kennett and Sueoka (15).

Origin D. Karamata D. Karamata A. T. Ganesan D. Karamata D. Karamata D. Karamata D. Karamata S. Riva (dnaP) D. Karamata D. Karamata D. Karamata N. Sueoka N. Sueoka A. Galizzi A. Galizzi A. Galizzi N. Brown N. Brown A. Galizzi G. Villani J. Copeland J. Copeland U. Canosi N. Brown From SB1084 From E3130 From E3146 From BD305 From E3105 From PB2346 From E3102

Distances between markers are expressed

as per-

centage of recombination, according to the convention: percentage of recombination = (1 - co-transfer) x 100. Drug sensitivity. Drug sensitivity was tested on nutrient agar plates containing increasing concentrations of the indicated chemicals added to melted nutrient agar at 450C. A suspension of bacteria was streaked on the plates, and the minimal inhibitory concentration (MIC) was read after 24 h of incubation at 370C. Determination of reversion frequencies. ts+ reversion frequencies were determined by plating overnight cultures grown at 300C in Penassay broth onto prewarmed nutrient agar plates after the cultures had been washed and concentrated 10-fold. All determinations reported are the average of several

experiments. RESULTS

HDS phenotype. We have found (Table 2) that the dna strains used in this study (see also Fig. 1) can be subdivided into two groups on the basis of their ability to yield colonies at the nonpermissive temperature (460C) on media containing 10 ,uM OHPhN2 Ura and 10 /iM OHPhN2 Iso. For a number of dna strains (group II), HP-resistant ts+ clones arise at a frequency significantly higher than that calcu-

176

SICCARDI ET AL.

TABLE

J. BACTERIOL.

2. Mutant frequencies of dna strains towards temperature insensitivity and HP resistance (OHPhN2 Ura and OHPhN2 Iso, 10 M each) Type of selection

Group

Strain

I

E3113 E3119 E3123 E3142 168TT VUB899 BD288

marker HP resistance at of colo HP resistance at 46°C (no marker 46T (o. ofcolo-350C (no. of colonies 46'C (no. of colonies nies x 106 cells) x 108 cells) x 109 cells) 2 0 dnaA13 5 3 dnaB19 30 0 dnaD23 2 7 0 400 dnaE42 880 0 dna-1 160 8 0 16 dna-8132 20 0 mut-1 25 10 0

II

SB1084 E3130 BD305 PB2346 E3105 E3146 E3102

dnaC6 dnaC30 dnaF133 dnaF69 dnaG5 dnaH46 dnaI2

210 850 12 71 110 210 390

410 2,200 480 540 12 100 79

34 50 100 3,500 40 190 45

N

R,

co

B

FIG. 1. Map location of the hds and dna loci and of other markers relevant to this paper.


VOL. 128, 1976

lated for the coincidence of independent mutational events (i.e., the product of the mutant frequencies obtained by applying separately each kind of selection) and thus probably indicative of a single mutational event. For other dna strains (group I), HP-resistant and temperature-insensitive clones do not arise at detectable frequencies. For the strains dnaF69 (of group II), the mutant frequency towards the double-resistant phenotype is so high that it is similar to the frequency of HP-resistant mutants at permissive temperature and can be evaluated even in the absence of a double selective procedure; more than 50% of the HP-resistant clones selected at 30°C are also ts+, and some 20% of the ts+ revertants selected in the absence of the analogues are HP resistant as well. We propose the code HDS (HP-resistant dna suppression) for the phenotype of the doubleresistant mutants derived from the dna strains of group II (dnaC, dnaG, dnaH, dnaI, and dnaF). To classify the mutations causing an HDS phenotype as extragenic suppressors, we should demonstrate (i) the persistence of the original dna allele in the HDS derivative and (ii) the identity of the HP resistance and suppressor markers. Demonstration of the persistence of dna alleles in HDS derivatives. The loci dnaC, dnaG, and dnaH are linked to purA (1, 14; Fig. 1). Transduction by PBS-1 phage of the pur+ allele of the HDS derivatives of these dna strains into a purA16 recipient shows that a significant fraction of the Pur+ recombinants also become temperature sensitive, indicating that the

177

dnaC, dnaG, and dnaH alleles are still present. The results in Table 3 also show that the linkage values between the dna loci and purA are altered for unknown reasons in the HDS derivatives, as compared with crosses in which the parental dna strains are used. The persistence of the dnaI allele co-transducible with pheA (14) on the (dnal) HDS derivatives (Table 3) was also demonstrated. dnaF is co-transducible with pyrA (22), but the ts alleles of the HDS derivatives of dnaF strains could not be rescued by co-transduction withpyrA (Table 3). To establish whether these strains were in fact true dnaF+ revertants, we used them as donors in transformation crosses with their dna parents, selecting for ts+ recombinants and then scoring for HP resistance (Table 4). All of the ts+ transformants were also resistant to the analogues, whereas none of the recombinants obtained with control wild-type (WT) DNA were resistant. Another control cross with PB2549 azp-12 (see below) DNA gave the expected frequency of drug-sensitive ts+ recombinants. Whereas the absence of Pyr+ ts recombinants in the crosses described in Table 3 remains unexplained, the results of the transformation back-crosses indicate the absence of a dnaF+ allele (and the nonidentity of the HP resistance marker and azp-12) and might be taken as indirect evidence of the persistance of the dnaF alleles in the corresponding HDS derivatives. Genetic mapping of the HP resistance marker of HDS strains. Clements et al. (6) have described the spontaneous mutant NB841R12, carrying a PoIIII, which is resistant also in vitro to the effects of the two analogues. This

TABLE 3. Transduction rescue of ts alleles from HDS derivatives of dna strains ofgroup II (see Table 2) Fraction ts and selected markers Donor strain

Relevant marker

purAl 6a

SB1084 PB2510 E3130 PB2513 E3105 PB2563 E3146 PB2561 E3102 PB2565 PB2346 PB2564 BD305 PB2562 a PB3409. b BC26. C BC34.

dnaC6 hds-10 (dnaC6) dnaC30 hds-13 (dnaC30) dnaG5 hds-63 (dnaG5) dnaH44 hds-61 (dnaH44) dnal2 hds-65 (dnal2) dnaF69 hds-64 (dnaF69)

dnaF133 hds-62 (dnaFl33)

pheA12"

pyrA26c

95/100a 45/100 95/100 60/100 65/100 32/100 51/100 27/100 80/100 72/100 30/100 0/200 30/100 0/200

178

SICCARDI ET AL.

J. BACTERIOL.

mutation, which identifies the PoIIII structural gene, was indicated by Brown as azp-12 and mapped by transduction between pyrA and recA in the poIC locus, which also contains dnaF markers (Fig. 1) (17; Attolini et al., Mol. Gen. Genet., in press). We have demonstrated that the resistance marker from this strain can be introduced by transduction into dna recipients of both groups I and II of Table 1 without altering the ts phenotype. The HP resistance markers of seven independently isolated HDS mutants (derived from dna strains dnaC6, dnaC30, dnaG5, dnaH46, dnaI2, dnaF133, and dnaF69) were mapped by transduction. The recombinants were scored for HP sensitivity or resistance (Table 5); for comparison, the table also reports the results obtained with strain azp-12. All seven HDS markers map very close to cysA, probably in a single genetic locus. We propose the code hds for the locus defined by these mutations (Fig. 1). Demonstration of the identity of hds alleles and suppressor markers. To demonstrate the identity of hds and suppressor markers, we have introduced the hds markers into their parental dna strains without direct selection for either drug resistance or temperature insenTABrz 4. Indirect demonstration of the persistence of ts alleles in HDS derivatives of strains dnaF133 and dnaF69a

DoNAr

Relevant marker

SB202 BD54 PB2346 PB2562

Fraction HP-resistant and selected marker

dnaF695

dnaFl33c

0/100 80/100 100/100

0/100

azp-12 78/100 hds-64 (dnaF69) hds-62 (dnaF133) 100/100 a By transformation and selection at 46°C for ts+ recombinants. b PB2346. e BD35.

sitivity. This aim was achieved by selecting rfm derivatives of the HDS strains; since rfm and hds are linked, the rfm hds strains can be used as donors in transduction crosses. The rifampin-resistant recombinants are then screened for their analogue resistance and temperature insensitivity. In no case were the HDS and ts+ phenotypes dissociated, indicating that hds alleles are responsible for the whole HDS phenotype. By using the same technique, we have also introduced the hds-14 allele of strain PB2414 (a derivative of dnaC30) into all the ts strains listed in Table 1, and we have checked the recombinants for their ts and HP resistance phenotypes: all the strains of group II (that can give rise to HDS derivatives) can be suppressed by the hds-14 allele, whereas the strains of group I can not. Analogous results have been obtained by introducing other hds alleles into the dna strains. In the course of the experiments, we have observed that all the ts+ recombinants grow well on HP-containing media at 46°C, but several of them grow poorly on the same media at 35°C, a phenotype never observed on the HDS derivatives of direct isolation and of which the meaning is unclear. The results, however, are clear-cut, since in no case are the ts+ and HP resistance phenotypes dissociated. The evidence described above indicates that each hds allele can act as a suppressor of all dna mutations of group II; thus, the suppression is neither allele nor locus specific. Comparison of HDS suppression and a specific "phenotypic correction." The relative lack of specificity of HDS suppression has prompted us to compare it with two kinds of "phenotypic correction" active on ts phenotypes, i.e., the ones exerted by high concentrations of NaCl and sucrose (5, 11). The effect of 2% NaCl or 20% sucrose in the phenotype of ts mutants is seen as an ionic strength-dependent

TABLE 5. Transduction mapping of HDS markers Donor strain

Fraction HP-resistant and selected marker

Relevant marker

pyrA26 BD54 PB2510 PB2513 PB2563 PB2561 PB2564 PB2562 PB2565 a PB3373. b PB3409. c PB3409.

azp-12 hds-10 (dnaC6) hds-13 (dnaC30) hds-63 (dnaG5) hds-61 (dnaH46) hds-64 (dnaF69) hds-62 (dnaF133) hds-65 (dnaI2)

25/lOOa 0/200 0/200 0/200 0/200 0/200 0/200 0/200

cysAl4b 0/100 85/100 80/100 75/100 83/100 82/100 79/100 80/100

purAl6S

0/100 31/100 38/100 35/100 40/100 38/100 36/100 32/100

179


VOL. 128, 1976

(or an osmotic pressure-dependent) stabiliza- teristics, even at the permissive temperature, tion of mutant proteins and is strictly allele and that the hds allele suppresses most of these specific; in Escherichia coli, Ricard and Hir- differences, restoring a pseudo-WT phenotype. We have considered the possibility that the ota (21) reported that some dnaB mutations are hds gene product might be a membrane composalt correctable and others are not. We have tested the effect of 2% NaCl and 20% nent (or the regulator of a membrane component); we have thus designed experiments to sucrose on dna strains of both groups I and II (see Table 2); the results (Table 6) reveal no check whether hds derivatives have an altered correlation between these kinds of phenotypic membrane stability if treated with the membrane-active agents polymixin B and sodium correction and HDS suppressibility. Further phenotypic analysis of (dna) hds dodecyl sulfate. No difference in the MICs of strains. The MIC of OHPhN2 Ura for HDS the two agents is observed between WT and hds strains is somewhat variable (4 to 10 ,ug/ml) in strains both at 35 and 460C; at the higher temthe various isolates and is not very high (com- perature the membrane-active agents have pared to the MIC for WT strains, i.e., 1 to 3 gg/ much lower MICs, but no differential effect is observed. ml) in any of the cases studied. This negative evidence does not, however, The MIC of a number of antibacterial agents (macromolecular synthesis inhibitors, ana- rule out that the hds gene product might be logues, .lkylating agents, DNA binding involved in the membrane attachment of the agents, surface-active compounds) was then replication complex; a number of other deterevaluated on otherwise isogenic WT, dna, and gents or different experimental approaches would be required. (dna) hds strains. Isolation of "new" HDS-suppressible dna The results reported in Table 7 show that dna strains differ from WT in a number of charac- strains. The observation that HDS suppressors are not locus specific prompted us to look for TABLz 6. Comparison of HDS suppression and other HDS-suppressible dna mutations. Fortyfive previously uncharacterized ts strains (iso'phenotypic correction" by salt or sugar lated in this laboratory by A. Galizzi) were Growth at 46°C in for their ability to yield, at high frequentested sUp HDS presence of: Relevant doubly resistant (HDS-like) spontaneous cies, pressibilStran marker mutants. Three such strains (PB2480, PB2481, ity Su2% NaCl 20% crose and PB2482) were found and demonstrated, by + + E3113 dnaA13 precursor incorporation experiments (Fig. 2), to dnaBI9 E3119 be dna mutants of the immediate-stop type. E3123 dnaD23 Thus, the HDS phenotype is also a powerful E3142 dnaE42 tool for indirect selection of dna mutants 168TT dna-i + SB1084 dnaC6 among uncharacterized ts strains. + + E3130 dnaC30 The new dna mutations, dna-80, dna-81, + + + E3105 dnaG5 and dna-82, have not yet been mapped but + dnaH46 E3146 differ in their map location from dnaC, dnaG, + + + E3102 dnal2 + dnaF69 PB2346 dnaH, and dnaI (i.e., all the other immediate+ BD305 dnaFl33 stop dna mutants capable of yielding HDS deTABLE 7. MICs of several antimicrobial compounds MIC

Strain

DB54 E3130 PB2513 PB2346 PB2564 E3105 PB2563 E3146 PB2561 E3102 PB2564

Relevant

marker

Parental dnaC30 hds-13 dna.F69 hds-64 dnaG5 hds-63 dnaH46 hds-61 dnal2 hds-65

5Bromouracil

Methyl methane sulfonate

200 200 200 150 200 200 200 200 200 200 200

400 500 500 200 400 200 200 200 500 200 500

(Ag/ml)

Actinomycin D Streptomycin

0.1 0.1 0.1 0.1 0.1 0.1 0.5 1.0 0.1 0.5 0.1

50 100 100 100 100 100 100 200 100 100 100

Nalidixic acid

2 2 2 4

4 4

2 8 8 8 8

180

SICCARDI ET AL.

J. BACTERIOL.

phenotype is compatible with the lack of DNA polymerase I; the three markers can coexist in viable strains, and the PollI of one of these recombinants was shown to be as sensitive to the analogues as that of the WT.

0%

0.

90 60 30 TIME (MIN) FIG. 2. DNA synthesis during a temperature shift from 35 to 46°C in the HDS-suppressible dna mutants PB2480 (0), PB2481 (0), and PB-2482 (0), and in their ts+ parent PB1424 (*).

0

rivatives), as demonstrated by the absence of linkage in transduction crosses within the purA-cysA and argA-pheA regions. The hds markers of the HDS strains derived from dna80, dna-81, and dna-82 map near cysA with all the other hds markers. DNA PoIIII studies in hds mutants. The PolIll of B. subtilis is composed of a single subunit (6) whose structural gene is defined by the azp-12 locus. Since hds mutations map elsewhere, hds mutants are likely to carry a WT form of DNA PolIlI. The PolIII of strain PB2513 (hds-13) obtained from dnaC30 is as sensitive to the analogues in vitro as that of the WT (N. Brown, personal communication). We have tested several other HDS derivatives and their dna parental strains by means of an assay (Ciarrocchi et al., submitted for publication) that can evaluate, in crude extracts, the total DNA-synthesizing activity and its HPUra-sensitive fraction. In all cases, the HPUra-sensitive fraction was approximately the same, indicating that PolIlI is in its WT form in all HDS derivatives. We have also constructed a polA hds (dnaC) strain to check whether the HDS

DISCUSSION The data reported in this paper suggest that extragenic suppression accounts for the very high ts+ reversion rate of several dna mutants in B. subtilis. We have identified the existence of a suppressor locus, named hds, whose alleles suppress a number of dna mutations (group II in Table 2) and confer resistance to aryl-azo-pyrimidines (analogues known to act upon DNA PoIIII); these two properties define the HDS phenotype. To classify hds mutations as extragenic suppressors, we have demonstrated the persistence of the original dna allele in the HDS derivatives and the identity of the HP resistance and suppressor markers. hds markers have been introduced by transduction into all the available dna strains; the results are in full agreement with spontaneous mutation data, since only dna strains of group II (see Table 2) are HDS suppressible. All the findings taken together clearly support the hypothesis that hds alleles are pleiotropic, extragenic suppressors of a class of dna mutants. Some of the results have nevertheless raised some perplexities that are worth discussing. First, the linkage of dnaC, dnaG, and dnaH to purA is reduced in HDS derivatives. This abnormality might represent some kind of negative interference or an abnormally frequent cotransduction of the suppressor marker (located nearby). Second, dnaF markers cannot be rescued by pyrA transduction from HDS derivatives, in spite of the fact that their HP resistance marker is located in the hds locus and HDS derivatives can be reconstructed by introducing an hds marker into dnaF strains by means of transduction, selecting neither for thermosensitivity nor for HP resistance. In this case no co-transduction of the suppressor and ts alleles can be invoked, and we cannot offer any explanation for the phenomenon. The result that all ts+ transformants of an hds (dnaF) x dnaF cross are also analogue resistant is indirect evidence that no dnaF+ allele is present in the HDS derivatives (since in this case we would expect some 50% ts+ HP-sensitive transformants). HDS suppression is neither allele nor locus specific and is ineffective on most "initiation" dna mutations and on a large sample of uncharacterized ts mutations; it is unrelated to

VOL. 128, 1976

other kinds of "phenotypic correction" and does not significantly alter cell permeability and membrane stability. The possibility that the HDS phenotype might be due to a different membrane composition (related to the replication complex binding to the membrane) has to be taken into consideration, but the indirect experiments designed to check this hypothesis (permeability and stability toward membrane-active agents) are clearly inconclusive. The extreme pleiotropic and the essential phenotypic identity of all hds alleles makes it likely that their effect is the lack of a function, rather than the specific modification of a protein sequence; with this hypothesis, the protein missing in HDS derivatives would be a "physiological destabilizer" of the DNA replication complex or, alternatively, the regulator of a "physiological stabilizer" of the complex that would thus be derepressed and present in higher amount in the HDS derivatives. HDS suppressibility has been successfully used as a tool to identify new dna mutants among uncharacterized ts strains.


9. 10.

11. 12.

13.

14.

15.

16. 17.

ACKNOWLEDGMENTS We wish to thank N. Brown and A. Falaschi for their interest and cooperation, A. Maragia and A. Clivio for performing some of the experiments, G. Alloni for excellent technical help, and S. Jayakar for correcting the manuscript. 1. 2. 3. 4.

5. 6.

7. 8.

LITERATURE CITED Andersen, J. J., and A. T. Ganesan. 1975. Temperaturesensitive deoxyribonucleic acid replication in a dnaC mutant ofBacillus subtilis. J. Bacteriol. 121:173-183. Bazil, G. W., and J. D. Gross. 1972. Effect of 6-(phydroxyphenyl)-azouracil on B. 8ubtilis DNA polymerases. Nature (London) New Biol. 240:82-83. Bazil, G. W., and J. D. Gross. 1973. Mutagenic DNA polymerase in B. subtili8. Nature (London) New Biol. 243:241-243. Bazil, G. W., and Y. Retief 1969. Temperature-sensitive DNA synthesis in a mutant of Bacillus subtilis. J. Gen. Microbiol. 56:87-97. Bilsky, A. Z., and J. B. Anrstrong. 1973. Osmotic reversal of temperature sensitivity in Escherichia coli. J. Bacteriol. 113:76-81. Clements, J. E., J. D'Ambrosio, and N. C. Brown. 1975. Inhibition of Bacillus subtilis deoxyribonucleic acid polymerase III by phenylhydrazinopyrimidines: demonstration of a drug-induced deoxyribonucleic acidenzyme complex. J. Biol. Chem. 250:522-526. Coulter, C. L., and N. R. Cozzarelli. 1975. Crystal structure of an inhibitor and a model for inhibition of replicative DNA synthesis. J. Mol. Biol. 91:329-344. Cozzarelli, N. R., and R. L. Low. 1973. Mutational alteration of Bacillus 8ubtilis DNA polymerase III to hydroxyphenylazopyrimidine resistance: polymerase

18.

19. 20.

21.

22.

23. 24.

25.

26.

181

III is necessary for DNA replication. Biochem. Biophys. Res. Commun. 51:151-157. Davis, B. D., and E. S. Mingioli. 1950. Mutants of Escherichia coli requiring methionine or vitamin B,2. J. Bacteriol. 60:17-28. Gass, K. B., R. L. Low, and N. R. Cozzarelli. 1973. Inhibition of a DNA polymerase from Bacillus subtilis by hydroxyphenylazopyrimidines. Proc. Natl. Acad. Sci. U.S.A. 70:103-107. Gross, J. D. 1972. DNA replication in bacteria. Curr. Top. Microbiol. Immunol. 57:39-74. Hara, H., and H. Yoshikawa. 1973. Asymmetric bidirectional replication of Bacillus subtilis chromosome. Nature (London) New Biol. 244:200-203. Hoch, J. A., M. Barat, and C. Anagnostopoulos. 1967. Transformation and transduction in recombinationdefective mutants of Bacillus subtilis. J. Bacteriol. 93:1925-1937. Karamata, D., and J. D. Gross. 1970. Isolation and genetic analysis of temperature-sensitive mutants of B. subtilis defective DNA synthesis. Mol. Gen. Genet. 108:277-287. Kennett, R. H., and N. Sueoka. 1971. Gene expression during outgrowth of Bacillus subtilis spores. The relationship between gene order on the chromosome and temporal sequence of enzyme synthesis. J. Mol. Biol. 60:3144. Laurent, S. J., and S. F. Vannier. 1973. Temperaturesensitive initiation of chromosome replication in a mutant of Bacillus subtilis. J. Bacteriol. 114:474-484. Love, E., J. D'Ambrosio, and N. C. Brown. 1976. Mapping of the gene specifying DNA polymerase III of Bacillus subtilis. Mol. Gen. Genet. 144:313-321. Mackenzie, J. M., M. M. Neville, G. E. Wright, and N. C. Brown. 1973. Hydroxyphenylazo-pyrimidines: characterization of the active form and their inhibitory action on a DNA polymerase from Bacillus subtilis. Proc. Natl. Acad. Sci. U.S.A. 70:512-516. Marmur, J. 1961. A procedure for the isolation of deoxyribonucleic acid from micro-organisms. J. Mol. Biol. 3:208-218. Mazza, G., H. M. Eisenstark, M. C. Serra, and M. Polsinelli. 1972. Effect of caffeine on the recombination process of Bacillus subtilis. Mol. Gen. Genet. 115:73-79. Ricard, M., and Y. Hirota. 1969. Effect des sels sur le processus de division cellulaire d'Escherichia coli. C. R. Acad. Sci. 168:1335-1338. Riva, S., C. Van Sluis, G. Mastromei, C. Attolini, G. Mazza, M. Polsinelli, and A. Falaschi. 1975. A new mutant of Bacillus subtilis altered in the initiation of chromosome replication. Mol. Gen. Genet. 137:185202. Spizizen, J. 1958. Transformation of biochemically deficient strains of Bacillus subtilis by deoxyribonucleate. Proc. Natl. Acad. Sci. U.S.A. 44:1072-1078. Stewart, C. R. 1969. Physical heterogeneity among Bacillus subtilis deoxyribonucleic acid molecules carrying particular genetics markers. J. Bacteriol. 98:1239-1247. White, K., and N. Sueoka. 1973. Temperature-sensitive DNA synthesis mutants of Bacillus subtilis. Appendix: theory of density transfer for symmetric chromosome replication. Genetics 73:185-214. Yamagishi, H., and I. Takahashi. 1968. Transducing particles of PBS-i1. Virology 36:639-645.