The subtilisin E gene of Bacillus subtilis is ... - Semantic Scholar

Proc. Natl. Acad. Sci. USA Vol. 81, pp. 1184-1188, February 1984 Genetics

The subtilisin E gene of Bacillus subtilis is transcribed from a cr37 promoter in vivo (serine protease/signal peptide/DNA sequence/Sl nuclease mapping/genetic mapping)

SuI-LAM WONG, CHESTER W. PRICE, DAVID S. GOLDFARBt, AND RoY H. DoI* Department of Biochemistry and Biophysics, University of California, Davis, CA 95616

Communicated by Clinton E. Ballou, November 10, 1983 Table 1. B. subtilis strains Sourcet Strain Genotype* P. Piggot (18) DB2 trpC2 D. Dubnau BD630 (19) DB21 hisH2 leu metB5 pSLW1 -4 DB2 DB24 sprE::cat trpC2 BGSC 1A84 DB25 glyB133 metDl BGSC 1A178 DB29 hprlO trpC2 BGSC 1A180 DB30 hprl6 DB24 s+ DB25 DB32 sprE::cat metDl DB29 DB32 DB33 sprE::cat hprlO DB30 4 DB32 DB34 sprE::cat hprl6 All strains are derivatives of B. subtilis 168, except DB29 and DB30, which are W168 strains. *The symbol sprE: cat refers to the chloramphenicol acetyl transferase gene of pSLW1 integrated at the sprE locus. tBGSC is the Bacillus Genetic Stock Center, Ohio State University. tf, transformation; donor - recipient; c, congression or cotransformation of unlinked markers.

A cloned Bacillus subtilis gene (sprE) exABSTRACT pressed only during the stationary growth phase is shown to encode the subtilisin E protease, an enzyme associated with sporulation. We have determined the DNA sequence of the sprE promoter region and the promoter-proximal half of the structural gene. The sprE gene codes for a putative 29-residue signal peptide and a 77-residue leader peptide preceding the mature subtilisin sequence. By plasmid integration and phage PBS1 transduction, we have mapped the sprE locus between glyB and metD on the B. subtilis chromosome, a region also containing the hyperprotease-producing hpr gene. In vitro the sprE gene is transcribed by the minor form of RNA polymerase containing a 37,000-dalton a factor (a,37). We show by S1 nuclease mapping that sprE transcription initiates at dual start sites both in vitro and in vivo and that the promoter for the downstream site has a characteristic a,37 recognition sequence. We propose that the physiological role of the a0" RNA polymerase is to transcribe a class of genes that are catabolite repressed, that encode extracellular enzymes, or that are expressed only during the stationary phase of growth.

probe plasmid to isolate a cryptic, temporally regulated, B. subtilis promoter. This promoter is recognized in vitro only by the a37-containing form of RNA polymerase and is expressed in vivo only during the stationary growth phase (16). Here we identify the B. subtilis gene controlled by the a'37 promoter by determining the sequence of the promoter-proximal region of the cloned DNA and matching the predicted amino acid sequence with the published sequence of subtilisin (17). We have mapped the locus of the subtilisin gene between glyB and metD on the B. subtilis chromosome, and we show that transcription of this gene initiates at a characteristic J7 promoter sequence both in vitro and in vivo.

Bacillus subtilis has a major form of RNA polymerase analogous to the Escherichia coli enzyme (1) and at least four minor holoenzyme forms of unknown physiological function (2-5). These polymerase forms differ by their associated a factors, each of which confers a characteristic promoter specificity on the holoenzyme in vitro. Losick and Pero (6) have suggested that the multiple holoenzyme forms control sporulation gene expression in vivo by a cascade mechanism, in which polymerase activities of changing promoter specificity are sequentially induced. Evidence to support this model remains limited and as yet none of the minor polymerase forms has been shown to transcribe a gene of known function. We show here that the gene encoding the extracellular subtilisin protease, sprE, is transcribed by the RNA polymerase containing a 37,000-dalton a" factor (a37). The subtilisin protease, also known as serine protease, alkaline protease, and bacillopeptidase E, is one of three extracellular proteases produced at the end of vegetative growth (7). This protease has been circumstantially implicated in sporulation control (8) and its expression is affected by the same physiological conditions or genetic lesions that affect the sporulation process. Subtilisin expression is repressed by glucose (9) and blocked by several spoO sporulation mutations (8). Expression is enhanced by a number of loci (10), including the catA (11), hpr (12), and scoC (13) mutations, which may comprise a single locus (13, 14). But the subtilisin structural gene itself has proven resistant to conventional genetic analysis (15) and, in the absence of sprE mutations, the developmental role of the protease remains

MATERIALS AND METHODS Bacterial Strains. The bacterial strains used in this study are listed in Table 1. DNA Sequence Analysis. The strategy for Maxam-Gilbert analysis (20) is shown in Fig. 1. Isolation of in Vivo and in Vitro RNA Transcripts for Nuclease S1 Mapping. RNA was isolated from 10 ml of sporulating cells [harvested at stage T3, 3 hr after the end of logarithmic growth in 2x SG medium (16)] by the procedure of Gilman and Chamberlin (23), except that we sonicated the cells 15 times and used proteinase K and proteinase K-treated DNAse at 200 rather than 100 ,ug/ml. We carried out in vitro transcription as before (16) but removed the DNA template by digestion with proteinase K-treated DNAse I. S1 Nuclease Mapping Using a Single-Stranded Hybridization Probe. The 1250-base-pair (bp) HindIII fragment was digested with Hpa II to yield a 370-base-pair fragment that carried the promoter region of the subtilisin gene (Fig. 1). This fragment was purified, end labeled, and strand separated as

unclear. We report elsewhere (16) the use of the pGR71 expression

Abbreviations: bp, base pair(s); Cmr, chloramphenicol resistant(ce). tPresent address: Department of Structural Biology, Stanford University Medical School, Stanford, CA 94305. *To whom reprint requests should be addressed.

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.

1184

Proc. Natl. Acad. Sci. USA 81 (1984)

Genetics: Wong et aL 200

0

800

600

400

1000

5

1185

1200 bp

*

CAT _

_

2 3

>

4 FIG. 1. Restriction map of the 1250-bp HindIlI fragment. The heavy arrow shows the location of the temporally regulated a-" promoter and the direction of transcription toward the chloramphenicol acetyl transferase (CAT) gene of pGR71S (16). The light arrows indicate the direction and extent of DNA sequence analysis on the following fragments: 1, 1250-bp HindIII; 2, 700-bp Hae III; 3, 700-bp Hinfl/HindIII; 4, 350-bp Dde I/Hpa I. These fragments were subcloned into the polylinker of pUC9 (21) and sequences were determined by the Maxam-Gilbert method (20) using the procedure of Ruther et al. (22). Fragment 5 is the 370-nucleotide Hpa II probe used for S1 mapping, 5' end labeled as shown by the asterisk.

tion (26), and for transformation (27). Auxotrophic markers were selected or scored on minimal glucose medium (28) supplemented with the appropriate amino acids and the sprE: :cat marker was selected on tryptose blood agar base plates (Difco) containing chloramphenicol at 5 ,g/ml. We

described by Maxam and Gilbert (20). S1 nuclease mapping was carried out by the Berk and Sharp procedure (24) with modifications described by Gilman and Chamberlin (23). Genetic Methods. We followed published methods for making PBS1 transducing lysates (25), for PBS1 transduc-

-52

TGATATACCTAAATAGAGATAAAATCATCTCAAAAAAATGGGTCTACTAAAATATTATTCCATCTAT HO-UC

CACUA G---

uuuoouoC "-c "-35" TACAATAAATTOACAGAATAGTCTTTTAAGTAAGTCTACTCTGAATTTTTTTAAAAGGAGAGGGTAA

n"-10"f

".-35"

+15

Met Arg Ser Lys Lys Leu Trp Ile Ser Leu Leu Phe Ala Leu Thr Leu AGA GTG AGA AGC AAA AAA TTG TGG ATC AGC TTG TTG TTT GCG TTA ACG TTA 10

+66

Ile Phe Thr Met Ala Phe Ser Asn Met Ser Ala Gln Ala Ala Gly Lys Ser ATC TTT ACG ATG GCA TTC AGC AAC ATG TCT GCG CAG GCT GCC GGA AAA AGC 30 20

+117

Ser Thr Glu Lys Lys Tyr Ile Val Gly Phe Lys Gln Thr Met Ser Ala Met AGT ACA GAA AAG AAA TAC ATT GTC GGA TTT AAA CAG ACA ATG AGT GCC ATG 50 40

+168

Ser Ser Ala Lys Lys Lys Asp Val Ile Ser Glu Lys Gly Gly Lys Val Gln AGT TCC GCC AAG AAA AAG GAT GTT ATT TCT GAA AAA GGC GGA AAG GTT CAA

+219

60 Lys Gln Phe Lys Tyr Val Asn Ala Ala Ala Ala Thr Leu Asp Glu Lys Ala AAG CAA TTT AAG TAT GTT AAC GCG GCC GCA GCA ACA TTG GAT GAA AAA GCT

+270

80

70

Val Lys Glu Leu Lys Lys Asp Pro Ser Val Ala Tyr Val Glu Glu Asp His GTA AAA GAA TTG AAA AAA GAT CCG AGC GTT GCA TAT GTG GAA GAA GAT CAT 100 90

+321

Ile Ala His Glu Tyr Ala Gln Ser Val Pro Tyr Gly Ile Ser Gln Ile Lys ATT GCA CAT GAA TAT GCG CAA TCT GTT CCT TAT GGC ATT TCT CAA ATT AAA

+372

110

Ala Pro Ala Leu His Ser Gln Gly Tyr Thr Gly Ser Asn Val Lys Val Ala GCG CCG GCT CTT CAC TCT CAA GGC TAC ACA GGC TCT AAC GTA AAA GTA GCT 130 120 Val Ile Asp Ser Gly Ile Asp Ser Ser His Pro Asp Leu Asn Val Arg Gly GTT ATC GAC AGC GGA ATT GAC TCT TCT CAT CCT GAC TTA AAC GTC AGA GGC 150 140 Gly Ala Ser GGA GCA AGC

+423

+474 +483

FIG. 2. DNA sequence of the subtilisin gene. The deduced amino acid sequence is given above the base sequence for the nontranscribed strand. The NH2-terminal alanine of the mature subtilisin is underlined (residue 107). The probable tandem transcription initiation sites P1 and P2 are shown by the open and closed arrows at nucleotides -15 and + 1, respectively. The putative -35 and -10 regions of the P1 promoter are overlined and those of the P2 (a37) promoter are underlined. A region of dyad symmetry overlapping the promoter region is indicated by the dashed line, with the center of symmetry marked by the dot at nucleotide -13. We show the ribosomal binding site aligned with the 3' end of the B. subtilis 16S RNA.

1186

Genetics: Wong et aL

scored the hpr marker on tryptose blood containing 1% milk solids.

Proc. Natl. Acad. Sci. USA 81 (1984) agar

base plates

a

b

c C

RESULTS Nucleotide Sequence of the Subtilisin Gene. Previously we used the pGR71 expression probe plasmid to isolate a 1250bp HindIII fragment containing a promoter expressed only during the stationary growth phase. The promoter is transcribed in vitro exclusively by the form of RNA polymerase containing 37 (16). As shown in Fig. 1, we located the ur37 promoter by runoff transcription to a region 400 bp from the point of fusion of B. subtilis DNA with the pGR71 CATgene. We determined the DNA sequence of the promoter-proximal part of the gene by the Maxam-Gilbert method (20), using the strategy shown in Fig. 1. The 602-nucleotide sequence is shown in Fig. 2. Only one open reading frame is available, ending at the HindIII site at nucleotide 483, the site of fusion to the pGR71 CAT gene. The deduced amino acid sequence of this open reading frame is given above the nucleotide sequence. Beginning with the alanine residue at position 107 and continuing to the site of fusion after the serine residue at position 155, the amino acid sequence is identical in 46 of 49 residues to the published NH2-terminal sequence of subtilisin BPN' and matches the subtilisin Carlsberg sequence in 33 residues (17). In two of the three differences between subtilisin BPN' and our sequence, at residues 114 and 149, we find the isoleucine and asparagine residues of the Carlsberg enzyme, whereas the third difference, an arginine at residue 151, appears variable among the three sequences. These results are consistent with the immunological data of Hageman and Carlton (29), who showed that the subtilisin E of B. subtilis 168 is very similar to the subtilisin BPN' of Bacillus amyloliquefaciens (subtilisin B) and less so to Carlsberg (subtilisin A). Thus we have identified the temporally expressed gene controlled by the &r37 promoter as the subtilisin E gene (sprE), encoding the extracellular alkaline seine protease of B. subtilis. Transcriptional Start Sites and Regulatory Regions of the sprE Gene. We located both the in vitro and in vivo transcription start sites of sprE by S1 nuclease mapping (23, 24). Fig. 3 shows the results obtained by using either the in vitro transcript of the ,37 RNA polymerase (Eu37) or in vivo mRNA isolated from T3 sporulating cells. We chose the +1 sites based on transcription initiation with the nearest adenosine, but this assignment is tentative until confirmed by mRNA sequence analysis. In vitro there are two start sites for sprE transcription, separated by 15 nucleotides. Transcription in vivo initiates at or near these same two start sites, and the downstream site (P2) has a characteristic v37 recognition sequence. The sprE P2 ,37 promoter and the two ou7 promoters studied by Moran et al. (31, 32) are compared in Table 2. Comparing only the presumed conserved bases, the -35 region of our sprE P2 promoter matches the consensus sequence in 5 out of 6 and the -10 region in 8 of 10. These two regions are separated by 15 bases in the sprE v.37 promoter. Tandem initiation sites have been observed for the Bacillus thuringiensis crystal protein gene (33) and the spoVG gene (32). The upstream site of spoVG is transcribed by the

A G A A

T T

o

C

>

..

A G A T I:

low

Ap em*

G Al

-~ GI A C i,

T

Al

T

A A A A

A A'

A

T 4 T T T C C

T C

FIG. 3. High-resolution S1 nuclease mapping of in vivo and in vitro transcripts from the tandem subtilisin promoters. In vivo RNA isolated from sporulating cells (T3) of B. subtilis DB21 (lane a) or RNA transcribed in vitro by the purified &r37 polymerase with the 1250-bp HindI1 fragment as template (lane b) were hybridized with the 5' endlabeled single-stranded Hpa II probe shown in Fig. 1. After S1 nuclease digestion, the protected DNA fragments were analyzed by electrophoresis on an 8% acrylamide/8 M urea sequencing gel. Lane c shows the A+G sequence ladder of the end-labeled probe. The DNA sequence shown on the right corresponds to the transcribed strand of the subtilisin gene. The vertical lines indicate the initiation regions. Open and closed arrows on the right show the probable in vivo transcription initiation sites for the P1 and P2 promoters, respectively, corrected by two bases from the apparent length of the Si-protected fragment (30).

a37 RNA polymerase, whereas the downstream initiation was shown by Johnson et al. (5) to result from an RNA polymerase containing a 32,000-dalton sigma factor (0.32), present as a minor contaminant in their Eu37 preparation. The

upstream (P1) promoter in the sprE gene matches the suggested u32 promoter sequences well at the presumed -35 region but only somewhat at the -10 (5). We cannot yet tell whether the in vitro initiation at P1 is due to Eu37, Eu32, or an as yet uncharacterized form of polymerase that copurifies with our Eu37 enzyme. Two other regions that may affect sprE expression are shown in Fig. 2. A 35-bp region of dyad symmetry overlaps the promoter region as indicated, and the ribosome binding site on the mRNA has the sequence pppA-A-A-A-G-G-A-GA-G-G-G-U, which has a calculated binding energy with the B. subtilis 16S rRNA of -13.5 kcal/mol (1 cal = 4.18 J). The sprE Gene Encodes a Prepropeptide. We chose GUG as the initiation codon based on the presence of the ShineDalgarno sequence immediately upstream. The deduced sequence of the subtilisin shows a putative signal peptide (residues 1-29) followed by a propeptide of 77 amino acids (residues 30-106) preceding the 49 residues of the NH2-terminal portion of the mature subtilisin enzyme (residues 107-155). The 29-residue signal peptide has a short positively charged

Table 2. Comparison of &," promoter sequences from three B. subtilis genes Gene -35 Spacer -10 sprE (P2) A-G-T-C-T-T-T T-A-A-G-T-A-A-G-T -C-T-A-C-T -C T-G-A-A-T-T-T-T-T -T ctc A-G-G-T-T-T-A A-A-T-C-C-T-T-A-T -C-G-T-T-A-T-G G-G-T-A-T-T-G-T-T -T A-G-G-A-T-T-T C-A-G-A-A-A-A-A-A-T-C-G-T spoVG (P1) G-G-A-A-T-T-G-A-T -A A-G-G - T-T-T +- 13-16 bp -G-G-A-A-T-T-G-T-T -T The ctc and spoVG sequences are from Moran et al. (31, 32). Bases common to all three sequences are indicated by bold-faced type in the -35 and -10 consensus sequences shown.

Proc. Natl. Acad. Sci. USA 81 (1984)

Genetics: Wong et aL Table 3. Four-factor transductional locus Recipient classes Selection Gly Hpr Cmr Met 1 1 1 1 Gly' 0 1 1 1 1 0 1 1 0 0 1 1 1 1 0 1 0 1 0 1 1 0 0 1 0 0 0 1

Cmr

1 0 1 0 1 0 1 0

1 1 1 1 1 1 1 1

1 1 0 0 1 1 0 0

1 1 1 1 0 0 0 0

crosses to map the sprE

No. 79 80 1 30 1 0 0 1

Order implied

(hpr-glyB)-sprE-metD

(hpr-glyB)-sprE-metD

66 81 0 2 1 1 29 12

91 (hpr-glyB)-sprE-metD 1 1 1 1 1 1 0 1 1 2 1 1 0 1 0 1 0 0 1 1 1 1 1 0 1 1 0 1 0 29 1 1 0 0 67 1 0 0 0 Donor: DB33 (sprE::cat hprlO); recipient: DB25 (glyB133 metDl). The symbols 1 and 0 refer to donor and recipient phenotypes, respectively.

Met+

region followed by a long sequence of hydrophobic amino acid residues and ends with the Ala-Gln-Ala-Ala sequence typically recognized by signal peptidase (34). The long highly charged propeptide (35% charged residues) preceding the mature subtilisin sequence is uncommon among prokaryotic enzymes. The propeptide-which may regulate subtilisin activity-must be cleaved after the tyrosine residue at position 106 to yield the NH2-terminal alanine on the mature enzyme

(17).

Mapping the Chromosomal Locus of sprE by Plasmid Integration. Several B. subtilis genes have been mapped by inserting a plasmid-encoded drug-resistance marker at the chromosomal locus of a cloned gene (see ref. 35). We moved

met D

hprglyB sprE '

7(4)_ 5

22(6) -

1

c

~~~58(24)

0(19)

l36(1)

e1 1 51 (19) lI.-

Genetic map of the sprE region of the B. subtilis chromoconstructed from the four-factor cross shown in Table 3. The tails indicate the selected marker, and genetic distances are shown as 100 minus the percentage cotransduction and in parentheses as t x 100 (14). Reciprocal recombination values between sprE::cat and either glyB or metD were unequal, with the genetic distance appearing greater when sprE::cat was the selected marker. hpr and glyB were 99% linked and the order with respect to one another is tentative (see text), so we have represented these markers as a single locus. Selection from this locus was for Gly'. FIG. 4.

some, arrow

1187

the 1250-bp HindIII fragment containing the sprE promoter to the integrative plasmid pCP115, a chloramphenicol-resistant (Cmr) derivative of pCP112 (36) that lacks the Sal I site at nucleotide 2566. When we used the resultant plasmid pSLW1 to transform strain DB2 to Cmr, pSLW1 integrated into the B. subtilis chromosome at the site of homology provided by the cloned sprE fragment, as we confirmed by Southern blot analysis of strain DB24 (data not shown). Using a PBS1 lysate of DB24 to transduce to prototrophy each of the mapping kit strains of Dedonder (37), we found the Cmr marker 77% linked to glyB, lying at about 900 on the B. subtilis genetic map (14). For multifactor crosses we constructed the DB33 donor strain by moving the sprE::cat locus and the hprlO allele into the DB25 genetic background (Table 1). Data from the cross shown in Table 3 were used to construct the genetic map in Fig. 4. glyB and hprlO were tightly linked and could not be ordered with respect to each other. Thus the order suggested by this cross is (hpr-glyB)-sprE::cat-metD. A second cross, using a different donor strain DB34 (sprE::cat hprl6) and DB25 as recipient, gave essentially the same linkages as the DB33 cross shown in Table 3 and established the order hprglyB-sprE::cat-metD (data not shown). We tentatively assign this gene order in Fig. 4.

DISCUSSION We have shown that the subtilisin E gene is transcribed from tandem promoters both in vivo and in vitro and that the downstream promoter has a characteristic a37 recognition sequence. These results strongly suggest that the subtilisin gene is transcribed by the oJ3 RNA polymerase in vivo, but this can be confirmed only when a mutation in the 0a37 structural gene becomes available. Two other transcriptional units, ctc and spoVG, are expressed from o-37 promoters (2, 31, 32), but their functions are unknown. Thus subtilisin is the only recognized gene product whose expression is dependent on a minor RNA polymerase. What features do these three genes share that might suggest the physiological role of the o(37 RNA polymerase? All three are expressed only in the stationary growth phase. The ctc gene may be the same as spoVC (38) and thus may affect development, and a spoVG deletion mutation apparently does affect sporulation (39). But the results of Millet et al. (15) suggest that the subtilisin protease is not essential for sporulation. Expression of sprE and spoVG, but not ctc, is prevented by certain of the early blocked SpoO sporulation mutations (8, 40). There is no evidence to suggest whether the spoO loci encode positive effectors that regulate spoVG and subtilisin expression directly, as suggested by Moran et al. (31), or whether spoO mutations act indirectly by disturbing stationary phase physiology, as has been proposed for other sporulation mutations (41). It is striking that sprE and spoVG, genes whose expression requires some SpoO functions, are both transcribed from tandem promoters. Johnson et al. (5) have discussed the possibility that in both B. subtilis and E. coli, genes under complex regulation may commonly have dual promoters. The subtilisin protease has two additional characteristics that suggest other roles for the 37 polymerase. It is an extracellular enzyme, and its expression is catabolite controlled. This suggests the possibility that certain classes of genes have specific types of promoters regulated by nutritional conditions and that other growth phase regulated genes, such as extracellular nucleases, proteases, phosphatases, and carbohydrate-degrading enzymes, are also37transcribed from or37 or other minor promoters. Because a is present during logarithmic growth as well as early stationary phase (2) and the subtilisin gene is expressed only after logarithmic of regulation must also growth ceases (16), another level the critical overthis In regard, control protease expression.

1188

Genetics: Wong et aL

lapping of the subtilisin promoter and the 35-bp region of dyad symmetry is suggestive of a binding site for a regulatory protein, since such sites frequently contain regions of dyad symmetry (42). The conventional genetic analysis of sprE has been difficult (15). By a combination of recombinant DNA and classical genetic methods we have unambiguously located sprE between glyB and metD on the B. subtilis genetic map (14). This position is distantifrom the locus between aroD and lysi reported by Leighton et al. (43) for their tsS mutation, which caused both sporulation and the subtilisin protease to become temperature sensitive. Millet et al. (15) suggested that tsS, like their similar tsl9, is comprised of two separate mutations, one affecting sporulation and the other, the subtilisin protease. Although the tsS allele is clearly not the structural gene for the subtilisin protease, the ts5 product may interact with the protease and thus render it temperature sensitive, a possibility recognized by Leighton and his colleagues (43). In Fig. 4 we have adjusted the marker order and genetic distances in the glyB region (14) and have found that sprE maps near hpr, another locus that affects protease activity (12). The hpr locus might include catA and scoC (13, 14), loci that increase protease expression and render it and sporulation less sensitive to glucose repression. The molecular basis of these effects is unknown. Our suggestion of a specific role of a37 RNA polymerase-transcription of the subtilisin E gene-expands the possible roles for the other minor RNA polymerase holoenzymes. Are they involved only narrowly in the transcription of sporulation-specific genes or perhaps more broadly in controlling other aspects of stationary cell physiology and the critical transition from logarithmic growth? These aspects could include the expression of catabolite-repressed genes, genes regulated by growth phase, and genes encoding extracellular enzymes as well as genes directly involved in the sporulation process. Note Added in Proof. Since this paper was submitted, Wells et al. (44) have reported the DNA sequence of the subtilisin BPN' gene from B. amyloliquefaciens. This gene encodes prepropeptide and amino-terminal regions similar to the subtilisin E gene described here, but the two genes differ substantially in their putative control regions. This research was supported in part by National Science Foundation Grant PCM 7924872, National Institute of General Medical Sciences Grant GM 19673, and the University of California.

1. Shorenstein, R. G. & Losick, R. (1973) J. Biol. Chem. 248, 6170-6173. 2. Haldenwang, W. G. & Losick, R. (1980) Proc. Natl. Acad. Sci. USA 77, 7000-7004. 3. Haldenwang, W. G., Lang, N. & Losick, R. (1981) Cell 23, 615-629. 4. Wiggs, J. L., Gilman, M. A. & Chamberlin, M. J. (1981) Proc. Natl. Acad. Sci. USA 78, 2762-2766. 5. Johnson, W. C., Moran, C. P. & Losick, R. (1983) Nature (London) 302, 800-804. 6. Losick, R. & Pero, J. (1981) Cell 25, 582-584. 7. Roitsch, C. A. & Hageman, J. H. (1983) J. Bacteriol. 155, 145152. 8. Piggot, P. J. & Coote, J. G. (1976) Bacteriol. Rev. 40, 908-962. 9. Dod, B. & Balassa, G. (1978) Mol. Gen. Genet. 163, 57-63.

Proc. NatL. Acad. Sci. USA 81 (1984) 10. Debabov, V. G. (1982) in The Molecular Biology ofthe Bacilli, ed. Dubnau, D. A. (Academic, New York), pp. 331-370. 11. Ito, J. & Spizizen, J. (1973) Colloq. Int. Cen. Natl. Res. Sci. 227, 81-82. 12. Higerd, T. B., Hoch, J. A. & Spizizen, J. (1972) J. Bacteriol. 112, 1026-1028. 13. Milhaud, P., Balassa, G. & Zucca, J. (1978) Mol. Gen. Genet. 163, 35-44. 14. Henner, D. J. & Hoch, J. A. (1980) Microbiol. Rev. 44, 57-82. 15. Millet, J., Larribe, M. & Aubert, J.-P. (1976) Biochimie 58, 109-117. 16. Goldfarb, D. S., Wong, S.-L., Kudo, T. & Doi, R. H. (1983) Mol. Gen. Genet. 191, 319-325. 17. Smith, E. L., DeLange, R. J., Evans, W. H., Landon, M. & Markland, F. S. (1968) J. Biol. Chem. 243, 2184-2191. 18. Piggot, P. (1973) J. Bacteriol. 114, 1241-1253. 19. Gryczan, T. J., Hahn, J., Contente, S. & Dubnau, D. (1982) J. Bacteriol. 152, 722-735. 20. Maxam, A. M. & Gilbert, W. (1980) Methods Enzymol. 65, 499-560. 21. Vieira, J. & Messing, J. (1982) Gene 19, 259-268. 22. Ruther, U., Koenen, M., Otto, K. & Muller-Hill, B. (1981) Nucleic Acids Res. 9, 4087-4098. 23. Gilman, M. Z. & Chamberlin, M. J. (1983) Cell 35, 285-293. 24. Berk, A. J. & Sharp, P. A. (1977) Cell 12, 721-731. 25. Karamata, D. & Gross, J. D. (1970) Mol. Gen. Genet. 108, 277-287. 26. Haworth, S. R. & Brown, L. R. (1973) J. Bacteriol. 114, 103113. 27. Boylan, B. J., Mendelson, N. H., Brooks, D. & Young, F. E. (1972) J. Bacteriol. 110, 281-290. 28. Anagnostopoulos, C. & Spizizen, J. (1961) J. Bacteriol. 81, 741-746. 29. Hageman, J. H. & Carlton, B. C. (1970) Arch. Biochem. Biophys. 139, 67-79. 30. Hentschel, C., Irminger, J.-C., Bucher, P. & Birnstiel, M. L. (1980) Nature (London) 285, 147-151. 31. Moran, C. P., Jr., Lang, N. & Losick, R. (1981) Nucleic Acids Res. 9, 5979-5990. 32. Moran, C. P., Jr., Lang, N., Banner, C. D. B., Haldenwang, W. G. & Losick, R. (1981) Cell 25, 783-791. 33. Wong, H. C., Schnepf, H. E. & Whiteley, H. R. (1983) J. Biol. Chem. 258, 1960-1967. 34. Perlman, D. & Halvorson, H. 0. (1983) J. Mol. Biol. 167, 391409. 35. Ferrari, F. A., Ferrari, E. & Hoch, J. A. (1982) J. Bacteriol. 152, 780-785. 36. Price, C. W., Gitt, M. A. & Doi, R. H. (1983) Proc. Natl. Acad. Sci. USA 80, 4074-4078. 37. Dedonder, R. A., Lepesant, J. A., Lepesant-Kejzlarova, J., Billaut, A., Steinmetz, M. & Kunst, F. (1977) Appl. Environ. Microbiol. 33, 989-993. 38. Moran, C. P., Losick, R. & Sonenshein, A. L. (1980) J. Bacteriol. 142, 331-334. 39. Rosenbluh, A., Banner, C. D. B., Losick, R. & Fitz-James, P. C. (1981) J. Bacteriol. 148, 341-351. 40. Ollington, J. F., Haldenwang, W. G., Huynh, T. V. & Losick, R. (1981) J. Bacteriol. 147, 432-442. 41. Wayne, R. R., Price, C. W. & Leighton, T. (1981) Mol. Gen. Genet. 183, 544-549. 42. Rosenberg, M. & Court, D. (1979) Annu. Rev. Genet. 13, 319353. 43. Leighton, T. J., Doi, R. H., Warren, R. A. J. & Kelln, R. A. (1973) J. Mol. Biol. 76, 103-122. 44. Wells, J. A., Ferrari, E., Henner, D. J., Estell, D. A. & Chen, E. Y. (1983) Nucleic Acids Res. 11, 7911-7925.