Biochimie (1991) 73, 1163-1170 © Soci6t6 franqaise de biochimie et biologie mol6culaire / Elsevier, Paris
1163
Suppression of defective-sporulation phenotypes by mutations in transcription factor genes of Bacillus subtilis C Ng , C Buchanan2,A Leungl, C Gintherl, T Leighton* tDepartment of Biochemistry and Molecular Biology, University of California, Berkeley, California 94720; 2Department of Biological Sciences, Southern Methodist University, Dallas, Texas 75275, USA (Received 20 November 1990; accepted 14 March 1991)
Summary --- Mutations in the Bacillus subtilis major RNA polymerase sigma factor gene (rpoD/crsA47) and a sensory receiver gene (spoOA/rvtA11) are potent intergenic suppressors of several stage 0 spomlation mutations (spoOB, OE, OF & OK). We show here that these suppressors also rescue temperature-sensitive spomlation phenotypes (Spots) caused by mutations in RNA polymerase, ribosomal protein, and protein synthesis elongation factor EF-G genes. The effects of the crsA and rvtA suppressors on RNA polymerase and ribosomal protein spots mutations are similar to those previously described for mutations in another intergenic suppressor gene rev. We have examined the effects of rvtA and crsA mutations on the expression of sporulation-associated membrane proteins, including flagellin and penicillin binding protein 5* (PBP 5*). Both suppressors restored sporu!ation and synthesis of PBP 5* in several spoO mutants. However, only rvtA restored flagellin synthesis in spoO suppressed backgrounds. The membrane protein phenotypes resulting from the presence of crsA or rvtA suppressors in spoO strains suggests that these suppressors function via distinct molecular mechanisms. The rvtA and crsA mutations are also able to block the ability of ethanol to induce spoO phenocopies at concentrations of ethanol which prevent sporulation in wild type cells. The effects of ethanol on sporulation-associated membrane protein synthesis in wild type and suppressor containing strains have been exaniined. Bacillus subtilis I sporulation / intergenic suppression / RNA polymerase / ribosomal proteins / penicillin-binding proteins
Introduction Mutations in over eighty separate genes can cause defective-sporulation phenotypes (Spo-) in Bacillus subtilis [33]. The functional interactions connecting genes which control a single stage of development, or among genes controlling the stage-specific progression of sporulation are not well understood. Intergenie suppression techniques have proven very useful in revealing the higher-order genetic interactions within sporulation regulons [1-4, 25]. Three intergenic suppressor mutations, rev, crsA and rvtA have been isolated which rescue sporulation competence in a variety of sporulation-defective backgrounds [1-4, 25]. rev, Originally isolated as a suppressor of an L17 ribosomal protein spots mutation, is epistatic to Spophenotypes resulting from mutations in RNA polymerase, ribosomal proteins, and protein synthesis elongation factor EF-G genes [4]. Oligosporogenous strains, such as kan-9, and spoOA299 also respond to rev suppression. In addition, this suppressor prevents the
*Correspondence and reprints
induction of Spo- phenocopies in wild type B subtilis by membrane perturbing chemicals such as ethanol or phenethyl alcohol [ 1]. A substantial portion of the rev gene has been cloned and sequenced, however the function of this gene is unknown [32]. Two other defective-sporulation suppressor mutations, crsA and rvtA, have been identified. The rvtA mutation was isolated as a suppressor of spoOF221 [2]. The crsA mutation is a member of a set of lesions in several unlinked genes that confer resistance to catabolite repression of sporulation [5, 6]. Both crsA and rvtA are epistatic to spoOB, spoOE, and spoOF mutations, but have no effect on the sporulation of spoOJ or spoOH strains [1-4, 25]. Like rev, they suppress the induction of Spo- phenocopies in wild type B subtilis by aliphatic alcohols and other membrane active chemicals [1-4]. In addition, crsA47 suppresses spoOK, spollF, spollJ, and spollN279 mutation~ [2, 3, 31]. The proteins encoded by the genes containing crsA47 and rvtA are known: crsA47"'maps within rpoD, the gene encoding the major sigma subunit of B subtilis RNA polymerase [7]; and rvtA maps within the spoOA gene, a twocomponent sensory receiver element controlling entry into the postexponential cell state [8].
1164
C Ng et al
The crsA47 and r v t A l l suppressors provide convenient probes for exploring the genetic structure and interactions controlling the sporulation process. In the studies described here we have analyzed the effects of these suppressors on the defective-sporulation phenotypes caused by mutations in transcription and translation system components. Agents and conditions that alter cell membrane structure can also inhibit sporulation in wild type cells and produce phenocopies similar to the phenotypes of spoO mutants. Treatment of B subtilis with ethanol, which alters membrane phospholipid synthesis and interactions [18, 19]; novobiocin, which increases membrane permeability [20]; and cerulenin, which inhibits fatty acid synthesis [21], block sporulation at stage 0. Correspondingly, other agents which stabilize membrane structure, such as glycerol or Tween 80, physiologically suppress temperature-sensitive sporulation phenotypes [21]. These effects and the abnormal membrane properties of spoO strains [9, 10, 18, 19], suggest that membrane structure and function play an important role in sporulation initiation. In order to further investigate these sporulation related membrane alterations, we have examined the effects of crsA and rvtA mutations on membrane protein synthesis in various spoO backgrounds. We have also studied the effects of sporulation inhibiting concentrations of ethanol on membrane protein synthesis in wild type cells and in crsA and rvtA strains. We have carried out detailed studies of two sporulation-associated membrane proteins, penicillin binding protein (PBP 5*) and flagellin. Addition of penicillin blocks sporulation at two points, stage II (septation) and stage IV (cortex and germ cell wall synthesis) [11, 12, 23]. These sensitive periods likely coincide with synthesis of the asymmetric septum and sporulation specific cortical peptidoglycan. It has been suggested that newly synthesized PBPs may be involved in these synthetic processes, and may be required for sporulation [11]. During sporulation, vegetative PBPs 2B and 3 increase significantly in concentration; PBP 5* (Mr = 40/245) appears; and, in some strains, a second new PBP, designated 4* (molecular weight 60000) also appears. Other penicillin binding proteins, 1, 2A, 4 and 5, decrease in concentration during the sporulation process, but not in sufficient amounts or at a suitable time for them to be precursors of the sporulation-associated proteins [ 11 ]. It is likely that the sporulation-associated changes in PBPs are due to de novo protein synthesis, in that chloramphenicol addition at stage 0 prevents all subsequent sporulation-associated changes in PBPs 2B, 3, 4* and 5* [12]. Flagellin, the primary structural protein of the flagellum, is also a sporulation-associated membrane protein [13]. Since this protein plays no known role in sporulation, it is possible that mutations which affect
sporulation also influence the expression of genes associated with other postexponential systems. Transcription of the flagellin gene (hag) requires RNA polymerase containing the minor sigma factor, o28 [14, 15]. This form of RNA polymerase is active in vegetative cells, and is required for the transcription of o28 promoters [16, 30]. The o28 transcription system is inactivated by spoOA, spoOB, spoOE and spoOF mutations [ 17].
Materials and methods Bacterial strains
The B subtilis strains used in these experiments are listed in table I. The strains were derived from two wild type backgrounds, B subtilis W168, designated CBI, and B subtilis 168, designated RS 1. All strains were given appropriate CB or RS strain numbers consistent with their parental origins. Strain construction
All gene transfer experiments were performed using PBSI mediated transduction. Preparation of donor phage lysates and transduction procedures have been described previously [22]. RNA polymerase and ribosomal antibiotic resistance mutations were introduced into strains containing the cysAl4 marker by selecting for cysteine prototrophy, and screening for cotransduction of antibiotic resistance [22]. Both crsA and ~tA are located within the lys-l-aroD region of the B subtilis chromosome, and were transduced into strains containing the lys-I marker by selecting for lysine prototrophy, and screening transductants for the small, pigmented colony phenotype of crsA47 and rvtAll. The recombinant suppressor phenotypes were confirmed by transductional back-crosses into RS4003 (spoOF221 lys-I pheAl). RNA polymerase and ribosomal protein antibiotic resistance mutations
The antibiotic resistance mutations employed are identical to those previously used to determine the suppression characteristics of rev-4 [4, 22]: erythromycin (cry.l), rifampin (rif-14), streptomycin (strA39), fusidic acid (fus-20), streptovaricin (sty710), neomycin (neo-162), and kanamycin (kan-9, kan-25). Antibiotic concentrations used for marker selection have been described [22]. Growth and sporulation
Media, growth conditions, sporulation in liquid culture, quantitation of spore yields, viable cell counts, and procedures for verification of suppressed and unsuppressed phenotypes have been described [3, 4, 22]. 2 x SG medium supplemented with CaCl2 (1 mM), MnCl2 (0.1 mM), FeSO4 (1 ttM), and glucose (0.1%), was used for all sporulation experiments. Cells were grown in Penassay medium (Difco) for experiments measuring flagellin or vegetative PBP levels. Ethanol (4% v/v) was added to inhibit sporulation as described in [18, 19]. Isolation of membrane proteins
Membrane protein fractions were prepared from sonicaUy disrupted cells as previously described [12, 23].
Sporulation transcriptionfactor interactions Table I. Bacillus subtilis strains. Strain
CBI RS2 RS700 RS1100 Glu-47 RS701 RS5101 CBI00 CB710 RS710 RS2710 RS2711 RS2712 CB400 CB720 RC720 RS2720 RS2721 RS2722 CB300 CB730 RS730 RS 2730 RS2731 RS2732 CB600 CB740 RS740 RS2740 RS2741 RS2742 CB200 CB750 RS750 RS2750 RS2751 RS2752 CB500 CB2501 RS760 RS2760 RS2761 RS2762 CB520 CB2521 RS770 RS2770 RS2771 RS2772 CBI401 CB1405 RS780 RS2780 RS2781 RS2782 RS2705 IA96 RS4003 IS16 RS5011 ISI7 RS5021 ISI9 RS2112 RS2113 RS5005 RS5006 IS24 IS26 IS28 RS5052
Genotype W168 168 168 cysAl4 lys-I 168 lys-I 168 crsA47 strA 168 crsA47 168 rvtAll W168 rif-14 W168 lys-I rif-14 168 lys-I rif-14 168 rif-14 168 rif-14 crsA47 168 rif-14 rvtAll W 168 ery-I W168 lys-I ery-i 168 lys-I ery-I 168 ery-I 168 ery-i crsA47 168 ery-! rvtAil W 168 strA39 WI68 lys.l strA39 168 lys-I strA39 168 strA39 168 strA39 crsA47 168 strA39 rvtAI!
WI68 fus-20 W!68 lys.! fus.20 168 iys.lfus.20 168fus-20 168~s-20 crsA47 168fus-20 rvtAll W168 sty.710 W168 lys-I sty710 168 iys-I sty-710 168 sty-710 168 stv-710 crsA47 168 sty-710 rvtAll W 168 kan-25 W 168 kan-25 rev-4 168 iys-I kan-25 168 kan-25 168 kan-25 crsA47 168 kan-25 rvtAll W168 nee-162 W168 heel62 rev-4 168 iys-I neo-162 168 neo-162 168 neo-162 crsA47 168 neo-162 ta,tAll W168 strA39 cysAl4 W168 kan.9 strA39 168 lys-I kan-9 168 kan.9 168 kan-9 crsA47 168 kan-9 rvtAll
Refer derivation [27] [1] [21 [6] Glu-47 --~ RS1100 (td) [2] [28l CB710 CB701 CB701 RSSI01 [22]
--> RS700 (td) --> RS710 (td) -* RS710 (td) --> RS710 (td)
CB720 CB701 CB701 RS5101 [29]
--> RS700 (td) -* RS720 (td) - , RS720 (td) - , RS720 (td)
CB730 RS5101 CB701 RS5101 [4]
-~ PS700 (td) -o RS730 (td) --~ RS730 (td) ~ RS730 (td)
CB740 CB701 CB701 RS5101 [27]
--, RS700 (td) --> RS740 (td) --> RS740 (td) --* RS740 (td)
CB750 --> RS700 (td) CB701 --->RS750 (td) CB701 --->RS750 (td) RS5101 --->RS750 (td) [22] [4] CB2501 --> RS700 (td) CB701 --> RS760 (td) CB701 --->RS760 (td) RS5101 --->RS760 (td) [22] [4] CB2521 --~ RS700 (td) CB701 --~ RS770 (td) CB701 --~ RS770 (td) RS5101 --~ RS770 (td) [22] CB510 --->CBI401 (td) CB1405 --~ RS700 (td) CB701 --> RS780 (td) CB701 --> RS780 (td) RS5101 --~ RS780 (td) 168 RS2 --, RS71100 (td) 168 pheAl trpC2 BGSCa 168 spoOF221 lys-I pheAl [2] 168 pheAl trpC2 spoOB136 BGSCa 168 pheAl spoOB136 rvtAl l [2] 168 pheAl trpC2 spoOEl l BGSCa 168 pheAl spoOEl l rvtA!i [2] 168 pheAl trpC2 spoOF221 BGSCa 168 lys-I spoOF221 rvtAl l [2] 168 spoOF221 rvtAl l [2] 168 pheAl spoOF221 rvtAl I [2] 168 pheAl spoOF221 crsA47 [3] 168 pheAl trpC2 spoOHl l6 BGSCa 168 pheAl trpC2 spoOJ87 BGSCa 168 trpC2 spoOKl41 BGSCa 168 spoOKl41 crsA47 [3]
a BGSC is the Bacillus Genetics Stock Center; td indicates transfer by transduction.
1165
One-dimensional SDS polyacrylamide gel electrophoresis analysis of PBPs T h e m e m b r a n e fraction was incubated with [3H] or [14C]benzylpenicillin as described in [12, 23]. The labeled proteins were solubilized and separated by SDS gel electrophoresis. The gel was incubated with a fluorophor solution, and exposed to X-ray film for 1 day.
Results
Suppression of the Spo- phenotype of translation and transcription system mutants The spore yields of cogenic sets of suppressed and unsuppressed strains, containing defects in ribosomal protein, protein synthesis elongation factor EF-G, and RNA polymerase genes are presented in table I. Both crsA47 and retail suppressed the temperature sensitive sporulation phenotypes (Spots) caused by the ery-l, strA39, and kan-25 ribosomal protein mutations, a mutation in elongation factor EF-G fus-20, and two RNA polymerase mutations rif-14 and sty710. At nonpermissive temperatures the various defective-sporulation mutants produced spores at 0.01-0.6% of the wild type level. Introduction of either the crsA47 or retAil suppressors increased the sporulation of these strains to 10-126% of the wild type level. In general, the results were qualitatively similar to the results obtained with the rev-4 suppressed derivatives of these same spo ts mutations [4]. However, the quantitative aspects of suppression varied substantially between crsA47, rvtAll (table II), and rev-4 [1] mutations. Although crsA47 significantlv improved sporulation competence when introduced into the kan-9 strain (RS2780), retail failed to suppress the oligosporogenous phenotype of this strain. The rev-4 mutation does suppress kan-9, increasing sporulation 800-fold. It was not possible to assess the effects of crsA47 on nee-162 sporulation due to the instability of this strain, rvtAll was able to suppress the nee-162 sporulation defect.
Restoration of flagellin synthesis in spoO mutants containing rvtA or crsA suppressors Flagellin, a 37 kDa major membrane protein, is readily identifed on SDS-PAGE gels [13]. The spoOB, OE & OF mutants produce significantly lower amounts of this protein when compared with the wild type strain (fig 1). Flagellin production is restored to wild type levels in rvtA suppressed derivatives of these spoO strains. However, this restoration of flagell i n expression is not seen in crsA suppressed spoO strains (fig 1).
1166
C Ng et al
Table II. Suppression of defective-sporulation phenotypes by crsA47 and rvtAll in 2 x SG medium.
Sporulation (sporeslml) Strain
Genotype
Non-permissive (% wt) temperature
Permissive (% wt) temperature
Sporulation ratio suppressed/unsuppressed (nonpermissive temperature)
Suppression ofrif-14and neo-162: nonpermissive temperature 48°C, permissive temperature 30°C RS2705 RS2710 RS2711 RS2712 RS2770 RS2771 RS2772
168 168 rif.14 168 rif-14 crsA47 168 rif-14 rvtAl l 168 neo-162 168 neo-162 crsA47 168 neo-162 rvtAll
9.0 x 10s 1.0 x 10s 6.7 x 108 !.2 x 10s 2.0 × 106 No data 1.8 × 10s
(100) (0,01) (74) (13) (0.2)
2.6 x 108 s (100) 2.8 x 108 s (108)
6700 1200
2.0 × 108 s (77) 105
(20)
Suppression ofery-l, stv-710, and kan-25: nonpermissive temperature 49°C, permissive temperature 300C RS2705 RS2720 RS2721 RS2722 RS2750 RS2751 RS2752
168 168 ery-I i68 ery-I crsA47 168 ery-I rvtAl l 168 sty-710 168 sty-710 crsA47 168 sty-710 rvtAl l
5.0 x 108 3.2 × 106 6.3 x 108 4.8 x 107 1.8 x 106 9.5 × 107 1.5 x 108
(100) (0.6) (126) (10) (0.4) (19) (30)
2.6 x 108 a (100) 6.8 x 108 s (262) 200 15 53 83
Suppression of su'A39 andfus-20: nonpermissive temperature 49°C, permissive temperature 37°C RS2705 RS2730 RS2731 RS2732 RS2740 RS2741 RS2742
168 168 strA39 168 strA39 crsA47 168 strA39 rvtAll
168fus-20 168fus-20 crsA47 168ft, s-20 rvtAll
5.0 x 10s 1.2 x 106 3.4 x 108 8.0 x 107 3.0 x 106 2.8 x 108 1.9 X 10s
(100) (0.2) (68) (16) (0.6) (56) (38)
2.0 × 2.9 × 3.2 × 3.5 x
(100) (15) (160) (18)
1.5 x 103 a (100) 1.5 x !08 s (100) 280 67 2.8 x 108 s (187) 93 63
Suppression of oligosporogenous kan-9: 37°C RS2705 RS2780 RS2781 RS2782
168 168 kan-9 168 kan-9 crsA47 168 kan-9 rvtAll
108 107 10s 107
Restoration of PBP 5* synthesis in spoO mutants containing rvtA or crsA suppressors It has been reported that spoOA and spoOH strains do not produce PBP 5* [11]. Similarly, spoOB, spoOE, and spoOF mutants also fail to produce PBP 5* (data not shown). Introduction of crsA or rvtA into various spoO strains restores PBP 5* production, although often at later times and in lower quantity than was observed in the wild type strain (fig 2).
11 1.2
Effect of ethanol on flagellin and PBP 5* synthesis Ethanol inhibits normal sporulation, synthesis of flagellin (data not shown) and PBP 5* synthesis in wild type cells (fig 2a). Wild type cells arrest sporulation at stage 0 in 2 x SG medium containing 4% ethanol [18, 19]. rvtA and crsA strains are resistant to the sporulation inhibiting effects of ethanol [2-7]. We have investigated the effects of ethanol addition on sporulation, PBP 5* synthesis, and flagellin synthesis
Sporulation transcriptionfactor interactions
1167
Fig 1. SDS gel electrophoresis of membrane-associated proteins from spoO mutants and spoO mutants containing rvtA or crsA. Data are presented for the following strains: (A) IA96 (wild type); (B) RS5101 (168 rvtAll); (C) 1S16 (spoOB136); (D) RS5011 (spoOB136 rvtAll); (E) IS17 (spoOF221); (F) RS5005 (spoOF221 rvtAll); (G) IS24 (spoOHll6); (H) IS26 (spoOJ87); (I) 1S28 (spoOK141); (J) RS701 (crsA47); (K) RS5006 (spoOF221 crsA47); (L) RS5052 (spoOKl41 crsA47); (M) Glu-47 (crsA47 strA); (N) RS2 (168 wild type); (O) IS17 (spoOEll); (P) RS5021 (spoOEll rvtAll). The arrows indicate the position of flagellin. Each lane contains approximately the same amount of protein.
in a cogenic set of strains containing the rvtA mutation. Wild type strain IA96 produces greatly reduced levels of PBP 5* when incubated in the presence of ethanol (fig 2a). A spoOF rvtA strain is able to sporulate and synthesize PBP 5* when grown in the presence of ethanol (fig 2b). Ethanol addition inhibited the postexponential phase synthesis of flagellin in the wild type strain IA96. Ethanol addition also inhibited the synthesis of flagellin in the spoOF rvtA strain, even though PBP 5* synthesis and sporulation occurred.
Discussion Genetic analysis has identified over eighty genes, mutations in which the sporulation process ~s prevented [24, 33]. More than half of these genes have been cloned [33]. The functions of many of these cloned sporulation genes have been identified [33]. However,
little is understood about the functional connections among genes which act either to determine stagespecific development, or to establish the interconnections between groups of genes which allow the progression of development from one stage to the next. Intergenic suppression analysis provides a powerful tool for probing these functional interrelationships and connections [2-4]. The rvtA and crsA suppressor mutations employed in these studies were originally identified as suppressors of spoOB, spoOE, and spoOF spomlation defects [2, 4]. In addition, crsA47 is able to suppress spoOK, spollN279, spolIF and spollJ mutations [25, 31]. The results reported here demonstrate that the crsA47 and rvtAll mutations are also epistatic to Spots phenotypes caused by mutations in RNA polymerase [3-subunit, ribosomal 50S and 30S protein, and protein elongation factor EF-G genes. These data not only expand the epistatic domain of defective-sporulation mutations suppressed by crsA and rvtA, but, more import-
C Ng e t al
1168 4" A
°
I
--
÷
il
Q
Fig 2. Fiuorographs of penicillin binding proteins from (A) strain IA96 (wild type) and (B) spoOF rvtA. The cultures were grown in the absence (-) or presence (+) of ethanol. Ethanol, when present, was added to a concentration of 4% at mid log phase. Samples were removed at mid log phase, at the beginning of the postexponential phase (to), and at 1 to 8 h into the postexponential phase (h to ts). Membrane proteins were separated by electrophoresis, and penicillin binding proteins were identified by labelling with [3H]penicillin.
antly, they suggest a functional relationship between the translational and transcriptional components that affect sporulation efficiency, and the products of the
early sporulation genes spoOB, spoOE, spoOF, spoOK, spollF, spollJ and spollN. With the exception that rvtAll does not suppress the oligosporogenous mutation kan-9, all three intergenie suppressor mutations are epistatic to the same defective-sporulation phenotypes produced by alterations in translational and transcriptional components [4]. However, these domains of epistasis are not precisely congruent since rev is unable to suppress spoOK and spollN279 mutations [4], which are rescued by crsA [25, 31]. The effects of rev on other spoO and spoil mutations have not been analyzed. The same spots transcription and translation system mutations that are suppressed by rev, crsA, and rvtA, are also physiologically suppressed by the addition of compounds such as ribose, fructose, mannose, glycerol, fatty acid esters, or Tween to the growth medium [21, 26]. It has been suggested that the defectivesporulation phenotypes caused by mutations in translational and transcriptional components are the result of subtle metabolic changes in RNA or protein synthesis that alter cellular metabolic balance and principally affect membrane structure and function [21 ]. In these studies we have examined the effects of the rvtA and crsA mutations on the synthesis of two sporulation-associated membrane proteins PBP 5* and flagellin. rvtA suppression of SpoO phenotypes is accompanied by a restoration of the synthesis of flagellin and PBP 5* (figs 1 and 2). crsA suppression of SpoO phenotypes restores only PBP 5* synthesis (figs 1 and 2). Since flagellin has no known role in sporulation, these results suggest that the function of the rvtA mutations in the spoOA gene may extend beyond effects on sporulation-specific pathways. These results also demonstrate a close relationship between expression of PBP 5* and sporulation competence. Flagellin, the product of the hag gene, is transcribed by a minor form of RNA polymerase containing o2s [15-17]. We had shown previously that rvtA was able to restore transcription from o2s promoters in spoO backgrounds [16, 17]. In the same set of experiments the presence of the crsA suppressor did not restore o2s promoter activity concomitant with suppression of spoO sporulation defects [ 17]. One limitation of these previous studies was that the protein products encodea by the 62s promoters analyzed were unknown. The present studies extend and confirm these previous data through analysis of protein expression from the hag gene. The data presented establish that the crsA mutation affects spoO mutant gene expression and developmental competence by a mechanism distinct from that provided by rvtA. The most surprising findings of these studies are the overlapping effects of the rvtA, crsA and rev suppressors on RNA polymerase and ribosomal spots mutations. The results suggest that these lesions, and
Sporulation transcription factor interactions several spoO and spoil mutations all affect a single common aspect of developmental regulation. What are the molecular mechanisms underlying these inter-relationships? Clues are available from the molecular identity and functions encoded by the suppressor genes and their targets. The rvtA mutations occur within the spoOA gene, an environmental response regulator required for the trans-activation of early sporulation genes [34] and the o28 transcription system. The SpoOA protein is expressed in vegetative cells, but is in an inactive state. Protein kinase sensors are activated upon nutrient deprivation, which in turn convert SpoOA into an active state by phosphorylation [34]. SpoOB, SpoOE and SpoOF are required for efficient phosphorylation of SpoOA [32, 34]. We postulate that the rvtA mutations, which occur within the phosphorylated activator domain of SpoOA (Leighton et al, submitted), create a form of SpoOA which is more readily phosphorylated, less readily dephosphorylated, or active in the absence of phosphorylation. Such mutations would bypass the requirement for 'upstream' gene products involved in the activation of SpoOA (ie, SpoOB, SpoOE and SpoOF). The Crs (catabolite resistant sporulation) phenotype of rvtA mutations is also understandable if one presumes that glucose downregulates starvation protein kinase sensor activity such that a critical concentration of SpoOA phosphate is not achieved in wild type cells. However, in rvtA strains the more readily phosphorylated form of SpoOA could reach a critical threshold level of phosphorylation in the presence of diminished protein kinase activity. Other Crs mutations within the N-terminal activator domain of SpoOA have been shown to be more readily phosphorylated by non-cognate protein kinases [34]. The crsA47 mutation occurs within the major RNA polymerase sigma subunit gene rpoD [7]. Recent studies of this mutation (Leighton et al, submitted) have shown that crsA RNA polymerase is defective in the transcription of a 'switch' gene sin, which represses the sporulation pathway, and directs cells into a motile, competent, non-sporulating postexponential cell state [34]. We postulate that the crsA mutation 'channels' cells into the main sporulation pathway and represses alternate non-sporulation pathways. Many of the gene products involved in chemotaxis, motility and competence are protein kinase substrates which would compete with SpoOA in phosphorylation reactions [34]. The repression of these competing substrates could explain the Crs phenotype of crsA, ie more protein kinase flow to SpoOA, and the suppression of 'upstream' spoOB, spoOE, s p o O F and s p o O K mutations, sin is also a transcriptional repressor of several stage II genes (Smith, Leighton et al, submitted), explaining the ability of crsA to suppress certain spoil mutations.
1169
Conclusion The results of the described intergenic suppression studies suggest that RNA polymerase core subunit and ribosomal protein genes are intimately involved in the starvation-activated response system which 'resets' the global pattern of gene expression during the transition from vegetative to postexponential phase. The fact that mutations in these genes are suppressible by spoOA and rpoD mutations suggests that transcription and translation system components have previously unexpected roles in developmental sensory transduction.
Acknowledgments We are grateful to HG Wittmann for his support and encouragement during these studies. This research was supported by grants from the National Science Foundation (ECE86-13227, TL) and the National lqstitutes of Health (GM-43564, CB).
References 1 Sharrock RA, Leighton T (1982) Suppression of defective-sporulation phenotypes by the Bacillus subtilis mutation rev-4. Mol Gen Genet 186, 432-438 2 SharrockRA, Rubinstein S, Chan M, Leighton T (1984) Intergenic suppression of spoO phenotypes by the Bacillus subtilis mutation iwtA. Mol Gen Genet 194, 260-264 3 LeungA, Rubinstein S, Yang C, Jing-Wei L, Leighton T (1985) Suppression of defective-sporulation phenotypes by mutations in the major sigma factor gene (rpoD) of Bacillus subtilis. Moi Gen Genet 201, 96-98 4 Sharrock RA, Leighton T (1981) lntergenic suppressors of temperature-sensitive sporulation in Bacillus subtilis are allele non-specific. Mol Gen Genet 183,532-537 5 Takahashi I (1979) Catabolite repression-resistant mutants of Bacillus subtilis. Can J Microbio125, ~1283-1287 6 Sun D, Takahashi 1 (1982) Genetic mapping of cataboliteresistant mutants of Bacillus subtilis. Can J Microbiol 28, 1242-125 I 7 Sun D, Takahashi I (1984) A catabolite-resistant mutation is localized in the rpo operon of Bacillus subtilis. Can J Microbio130, 423-429 8 Price CW, Doi RH (1985) Genetic mapping of rpoD implicates the major sigma factor of Bacillus subtilis RNA polymerase in the control of sporulation initiation. Mol Gen Genet 201, 88-95 9 Rhaese HJ, Reichenbach A, Stamminger G (1984) Changes in membrane-associated proteins during sporulation in Bacillus subtilis. J Biochem 139, 593--603 10 Andreoli AJ, Kao M, Chui R, Cabrera J, Wong SKS (1981) Two-dimensional polyacrylamidegel electrophoresis analysis of cytoplasmic and membrane proteins from sporulating cells and forespore and mother-cell compartments of Bacillus subtilis. In: Sporulation and Germination (Levinson HS, Sonenshein AL, Tipper DJ, eds) American Society for Microbiology, Washington, 168 p
1170
C Ng et
11 Sowell MO, Buchanan CE (1983) Changes in penicillinbinding proteins during sporulation of Bacillus subtilis. J Bacteriol 153, 1331-1337 12 Buchanan CE, Sowell MO (1983) Stability and synthesis of penicillin-binding proteins during sporulation. J Bacteriol 156, 545-551 13 Aubert E, Klier A, Rapoport G (1985) Cloning and expression in Escherichia coli of the regulatory sacU germ from Bacillus subtilis. J Bacteriol 161, 1182-1187 14 Helmann JD, Chamberlin MJ (1988) Structure and function of bacterial sigma factors. Annu Rev Biochem 57, 839-872 15 Mirel DB, Chamberlin MJ (1989) The Bacillus subtilis flagellin gene (hag) is transcribed by the ~-28 form of RNA polymerase. J Bacteriol 171, 3095-3101 16 Gilman MZ, Glenn JS, Singer VL, Chamberlin MJ (1984) Isolation of sigma-28-specific promoters for Bacillus subtilis DNA. Gene 32, 11-20 17 Chamberlin MJ, Briat JF, Singer VL, Glenn JS, Helmann J, Leung A, Moorefield MB, Gilman MZ (1985) Isolation of genomic segments from Bacillus subtilis that contain promoters for a2s RNA polymerase, and studies on regulation of their transcription in Bacillus subtilis and Escherichia coli. In: Molecular Biology of Microbial Differentiation (Hoeh J, Setlow P, eds) American Society for Microbiology, Washington, 143 p 18 Bohin JP, Rigomier D, Sehaeffer P (1976) Ethanol sensitivity of sporulation in Bacillus subtilis: a new tool for the analysis of sporulation process. J Bacteriol 127, 934-940 19 Rigomier D, Bohin JP, Lubochinsky B (1980) Effects of ethanol and methanol on lipid metabolism in Bacillus subtilis. J Gen Microbiol 121,139-149 20 Takahashi I, MacKenzie LW (la~'~...,,,.,Effects of various inhibitory agents on sporulation of Bacillus subtilis. Can J Microbio128, 80-86 21 Wayne RR, Price CW, Leighton T (1981) Physiological suppression of the temperature-sensitive sporulation defect in a Bacillus subtilis RNA polymerase mutant. Mol Gen Genet 183, 544-549 22 Sharrock RA, Leighton T, Wittmann HG (1981) Macrolide and amino-glycoside antibiotic resistance mutations in the Bacillus subtilis ribosome resulting in temperature-sensitive sporulation. Mol Gen Genet 183, 538-543
al
23 24 25
26
27 28 29 30 31
32
33
34
Buchanan CE (1979) Altered membrane proteins in a minicell-producing mutants of Bacillus subtilis. J Bacteriol 139, 305-307 Piggot PJ, Coote JC (1976) Genetic aspects of bacterial endospore formation. Bacteriol Rev 40, 908-962 Leung A, Ng D, Yang C, Rubinstein S, Deede C, Putnam W, Leighton T (1985) The Bacillus subtilis spoO regnlon: intergenic suppression analysis of initiation control. In: Molecular Biology of Microbial Differentiation (Hoch J, Setlow P, eds) American Society for Microbiology, Washington, 176 p Wayne RR, Leighton T (1981) Physiological suppression of Bacillus subtilis conditional sporulation phenotypes: RNA polymerase and ribosomal mutations. Mol Gen Genet 183, 550-552 Leighton TJ (1977) New types of RNA polymerase mutations causing temperature-sensitive sporulation in Bacillas subtilis. J Bioi Chem 252, 268-272 Leighton TJ (1973) An RNA polymerase mutation causing temperature-sensitive sporulation in Bacillus subtilis. Proc Natl Acad Sci USA 70, 1179-1183 Leighton TJ (1974) Sporulation-specific translational discrimination in Bacillus subtilis. J Mol Bio186, 855-863 Gilman M, Chamberlin M (1983) Development and genetic regulation of Bacillus subtilis genes transcribed by a 28RNA polymerase. Cell 35, 285-293 Stragier P (1989) Temporal and spatial control of gene expression during sporulation: from facts to speculation. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 243 p Trach K, Chapman JW, Piggot P, LeCoq D, Hoch JA (1988) Complete sequence and transcriptional analysis of the spoOF region of the Bacillus subtUis chromosome. J Bacteriol 170, 4194-4208 Piggot PJ (1989) Revised genetic map of Bacillus subtilis 168. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 1 p Smith I (1989) Initiation of sporulation. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 185 p
1163
Suppression of defective-sporulation phenotypes by mutations in transcription factor genes of Bacillus subtilis C Ng , C Buchanan2,A Leungl, C Gintherl, T Leighton* tDepartment of Biochemistry and Molecular Biology, University of California, Berkeley, California 94720; 2Department of Biological Sciences, Southern Methodist University, Dallas, Texas 75275, USA (Received 20 November 1990; accepted 14 March 1991)
Summary --- Mutations in the Bacillus subtilis major RNA polymerase sigma factor gene (rpoD/crsA47) and a sensory receiver gene (spoOA/rvtA11) are potent intergenic suppressors of several stage 0 spomlation mutations (spoOB, OE, OF & OK). We show here that these suppressors also rescue temperature-sensitive spomlation phenotypes (Spots) caused by mutations in RNA polymerase, ribosomal protein, and protein synthesis elongation factor EF-G genes. The effects of the crsA and rvtA suppressors on RNA polymerase and ribosomal protein spots mutations are similar to those previously described for mutations in another intergenic suppressor gene rev. We have examined the effects of rvtA and crsA mutations on the expression of sporulation-associated membrane proteins, including flagellin and penicillin binding protein 5* (PBP 5*). Both suppressors restored sporu!ation and synthesis of PBP 5* in several spoO mutants. However, only rvtA restored flagellin synthesis in spoO suppressed backgrounds. The membrane protein phenotypes resulting from the presence of crsA or rvtA suppressors in spoO strains suggests that these suppressors function via distinct molecular mechanisms. The rvtA and crsA mutations are also able to block the ability of ethanol to induce spoO phenocopies at concentrations of ethanol which prevent sporulation in wild type cells. The effects of ethanol on sporulation-associated membrane protein synthesis in wild type and suppressor containing strains have been exaniined. Bacillus subtilis I sporulation / intergenic suppression / RNA polymerase / ribosomal proteins / penicillin-binding proteins
Introduction Mutations in over eighty separate genes can cause defective-sporulation phenotypes (Spo-) in Bacillus subtilis [33]. The functional interactions connecting genes which control a single stage of development, or among genes controlling the stage-specific progression of sporulation are not well understood. Intergenie suppression techniques have proven very useful in revealing the higher-order genetic interactions within sporulation regulons [1-4, 25]. Three intergenic suppressor mutations, rev, crsA and rvtA have been isolated which rescue sporulation competence in a variety of sporulation-defective backgrounds [1-4, 25]. rev, Originally isolated as a suppressor of an L17 ribosomal protein spots mutation, is epistatic to Spophenotypes resulting from mutations in RNA polymerase, ribosomal proteins, and protein synthesis elongation factor EF-G genes [4]. Oligosporogenous strains, such as kan-9, and spoOA299 also respond to rev suppression. In addition, this suppressor prevents the
*Correspondence and reprints
induction of Spo- phenocopies in wild type B subtilis by membrane perturbing chemicals such as ethanol or phenethyl alcohol [ 1]. A substantial portion of the rev gene has been cloned and sequenced, however the function of this gene is unknown [32]. Two other defective-sporulation suppressor mutations, crsA and rvtA, have been identified. The rvtA mutation was isolated as a suppressor of spoOF221 [2]. The crsA mutation is a member of a set of lesions in several unlinked genes that confer resistance to catabolite repression of sporulation [5, 6]. Both crsA and rvtA are epistatic to spoOB, spoOE, and spoOF mutations, but have no effect on the sporulation of spoOJ or spoOH strains [1-4, 25]. Like rev, they suppress the induction of Spo- phenocopies in wild type B subtilis by aliphatic alcohols and other membrane active chemicals [1-4]. In addition, crsA47 suppresses spoOK, spollF, spollJ, and spollN279 mutation~ [2, 3, 31]. The proteins encoded by the genes containing crsA47 and rvtA are known: crsA47"'maps within rpoD, the gene encoding the major sigma subunit of B subtilis RNA polymerase [7]; and rvtA maps within the spoOA gene, a twocomponent sensory receiver element controlling entry into the postexponential cell state [8].
1164
C Ng et al
The crsA47 and r v t A l l suppressors provide convenient probes for exploring the genetic structure and interactions controlling the sporulation process. In the studies described here we have analyzed the effects of these suppressors on the defective-sporulation phenotypes caused by mutations in transcription and translation system components. Agents and conditions that alter cell membrane structure can also inhibit sporulation in wild type cells and produce phenocopies similar to the phenotypes of spoO mutants. Treatment of B subtilis with ethanol, which alters membrane phospholipid synthesis and interactions [18, 19]; novobiocin, which increases membrane permeability [20]; and cerulenin, which inhibits fatty acid synthesis [21], block sporulation at stage 0. Correspondingly, other agents which stabilize membrane structure, such as glycerol or Tween 80, physiologically suppress temperature-sensitive sporulation phenotypes [21]. These effects and the abnormal membrane properties of spoO strains [9, 10, 18, 19], suggest that membrane structure and function play an important role in sporulation initiation. In order to further investigate these sporulation related membrane alterations, we have examined the effects of crsA and rvtA mutations on membrane protein synthesis in various spoO backgrounds. We have also studied the effects of sporulation inhibiting concentrations of ethanol on membrane protein synthesis in wild type cells and in crsA and rvtA strains. We have carried out detailed studies of two sporulation-associated membrane proteins, penicillin binding protein (PBP 5*) and flagellin. Addition of penicillin blocks sporulation at two points, stage II (septation) and stage IV (cortex and germ cell wall synthesis) [11, 12, 23]. These sensitive periods likely coincide with synthesis of the asymmetric septum and sporulation specific cortical peptidoglycan. It has been suggested that newly synthesized PBPs may be involved in these synthetic processes, and may be required for sporulation [11]. During sporulation, vegetative PBPs 2B and 3 increase significantly in concentration; PBP 5* (Mr = 40/245) appears; and, in some strains, a second new PBP, designated 4* (molecular weight 60000) also appears. Other penicillin binding proteins, 1, 2A, 4 and 5, decrease in concentration during the sporulation process, but not in sufficient amounts or at a suitable time for them to be precursors of the sporulation-associated proteins [ 11 ]. It is likely that the sporulation-associated changes in PBPs are due to de novo protein synthesis, in that chloramphenicol addition at stage 0 prevents all subsequent sporulation-associated changes in PBPs 2B, 3, 4* and 5* [12]. Flagellin, the primary structural protein of the flagellum, is also a sporulation-associated membrane protein [13]. Since this protein plays no known role in sporulation, it is possible that mutations which affect
sporulation also influence the expression of genes associated with other postexponential systems. Transcription of the flagellin gene (hag) requires RNA polymerase containing the minor sigma factor, o28 [14, 15]. This form of RNA polymerase is active in vegetative cells, and is required for the transcription of o28 promoters [16, 30]. The o28 transcription system is inactivated by spoOA, spoOB, spoOE and spoOF mutations [ 17].
Materials and methods Bacterial strains
The B subtilis strains used in these experiments are listed in table I. The strains were derived from two wild type backgrounds, B subtilis W168, designated CBI, and B subtilis 168, designated RS 1. All strains were given appropriate CB or RS strain numbers consistent with their parental origins. Strain construction
All gene transfer experiments were performed using PBSI mediated transduction. Preparation of donor phage lysates and transduction procedures have been described previously [22]. RNA polymerase and ribosomal antibiotic resistance mutations were introduced into strains containing the cysAl4 marker by selecting for cysteine prototrophy, and screening for cotransduction of antibiotic resistance [22]. Both crsA and ~tA are located within the lys-l-aroD region of the B subtilis chromosome, and were transduced into strains containing the lys-I marker by selecting for lysine prototrophy, and screening transductants for the small, pigmented colony phenotype of crsA47 and rvtAll. The recombinant suppressor phenotypes were confirmed by transductional back-crosses into RS4003 (spoOF221 lys-I pheAl). RNA polymerase and ribosomal protein antibiotic resistance mutations
The antibiotic resistance mutations employed are identical to those previously used to determine the suppression characteristics of rev-4 [4, 22]: erythromycin (cry.l), rifampin (rif-14), streptomycin (strA39), fusidic acid (fus-20), streptovaricin (sty710), neomycin (neo-162), and kanamycin (kan-9, kan-25). Antibiotic concentrations used for marker selection have been described [22]. Growth and sporulation
Media, growth conditions, sporulation in liquid culture, quantitation of spore yields, viable cell counts, and procedures for verification of suppressed and unsuppressed phenotypes have been described [3, 4, 22]. 2 x SG medium supplemented with CaCl2 (1 mM), MnCl2 (0.1 mM), FeSO4 (1 ttM), and glucose (0.1%), was used for all sporulation experiments. Cells were grown in Penassay medium (Difco) for experiments measuring flagellin or vegetative PBP levels. Ethanol (4% v/v) was added to inhibit sporulation as described in [18, 19]. Isolation of membrane proteins
Membrane protein fractions were prepared from sonicaUy disrupted cells as previously described [12, 23].
Sporulation transcriptionfactor interactions Table I. Bacillus subtilis strains. Strain
CBI RS2 RS700 RS1100 Glu-47 RS701 RS5101 CBI00 CB710 RS710 RS2710 RS2711 RS2712 CB400 CB720 RC720 RS2720 RS2721 RS2722 CB300 CB730 RS730 RS 2730 RS2731 RS2732 CB600 CB740 RS740 RS2740 RS2741 RS2742 CB200 CB750 RS750 RS2750 RS2751 RS2752 CB500 CB2501 RS760 RS2760 RS2761 RS2762 CB520 CB2521 RS770 RS2770 RS2771 RS2772 CBI401 CB1405 RS780 RS2780 RS2781 RS2782 RS2705 IA96 RS4003 IS16 RS5011 ISI7 RS5021 ISI9 RS2112 RS2113 RS5005 RS5006 IS24 IS26 IS28 RS5052
Genotype W168 168 168 cysAl4 lys-I 168 lys-I 168 crsA47 strA 168 crsA47 168 rvtAll W168 rif-14 W168 lys-I rif-14 168 lys-I rif-14 168 rif-14 168 rif-14 crsA47 168 rif-14 rvtAll W 168 ery-I W168 lys-I ery-i 168 lys-I ery-I 168 ery-I 168 ery-i crsA47 168 ery-! rvtAil W 168 strA39 WI68 lys.l strA39 168 lys-I strA39 168 strA39 168 strA39 crsA47 168 strA39 rvtAI!
WI68 fus-20 W!68 lys.! fus.20 168 iys.lfus.20 168fus-20 168~s-20 crsA47 168fus-20 rvtAll W168 sty.710 W168 lys-I sty710 168 iys-I sty-710 168 sty-710 168 stv-710 crsA47 168 sty-710 rvtAll W 168 kan-25 W 168 kan-25 rev-4 168 iys-I kan-25 168 kan-25 168 kan-25 crsA47 168 kan-25 rvtAll W168 nee-162 W168 heel62 rev-4 168 iys-I neo-162 168 neo-162 168 neo-162 crsA47 168 neo-162 ta,tAll W168 strA39 cysAl4 W168 kan.9 strA39 168 lys-I kan-9 168 kan.9 168 kan-9 crsA47 168 kan-9 rvtAll
Refer derivation [27] [1] [21 [6] Glu-47 --~ RS1100 (td) [2] [28l CB710 CB701 CB701 RSSI01 [22]
--> RS700 (td) --> RS710 (td) -* RS710 (td) --> RS710 (td)
CB720 CB701 CB701 RS5101 [29]
--> RS700 (td) -* RS720 (td) - , RS720 (td) - , RS720 (td)
CB730 RS5101 CB701 RS5101 [4]
-~ PS700 (td) -o RS730 (td) --~ RS730 (td) ~ RS730 (td)
CB740 CB701 CB701 RS5101 [27]
--, RS700 (td) --> RS740 (td) --> RS740 (td) --* RS740 (td)
CB750 --> RS700 (td) CB701 --->RS750 (td) CB701 --->RS750 (td) RS5101 --->RS750 (td) [22] [4] CB2501 --> RS700 (td) CB701 --> RS760 (td) CB701 --->RS760 (td) RS5101 --->RS760 (td) [22] [4] CB2521 --~ RS700 (td) CB701 --~ RS770 (td) CB701 --~ RS770 (td) RS5101 --~ RS770 (td) [22] CB510 --->CBI401 (td) CB1405 --~ RS700 (td) CB701 --> RS780 (td) CB701 --> RS780 (td) RS5101 --~ RS780 (td) 168 RS2 --, RS71100 (td) 168 pheAl trpC2 BGSCa 168 spoOF221 lys-I pheAl [2] 168 pheAl trpC2 spoOB136 BGSCa 168 pheAl spoOB136 rvtAl l [2] 168 pheAl trpC2 spoOEl l BGSCa 168 pheAl spoOEl l rvtA!i [2] 168 pheAl trpC2 spoOF221 BGSCa 168 lys-I spoOF221 rvtAl l [2] 168 spoOF221 rvtAl l [2] 168 pheAl spoOF221 rvtAl I [2] 168 pheAl spoOF221 crsA47 [3] 168 pheAl trpC2 spoOHl l6 BGSCa 168 pheAl trpC2 spoOJ87 BGSCa 168 trpC2 spoOKl41 BGSCa 168 spoOKl41 crsA47 [3]
a BGSC is the Bacillus Genetics Stock Center; td indicates transfer by transduction.
1165
One-dimensional SDS polyacrylamide gel electrophoresis analysis of PBPs T h e m e m b r a n e fraction was incubated with [3H] or [14C]benzylpenicillin as described in [12, 23]. The labeled proteins were solubilized and separated by SDS gel electrophoresis. The gel was incubated with a fluorophor solution, and exposed to X-ray film for 1 day.
Results
Suppression of the Spo- phenotype of translation and transcription system mutants The spore yields of cogenic sets of suppressed and unsuppressed strains, containing defects in ribosomal protein, protein synthesis elongation factor EF-G, and RNA polymerase genes are presented in table I. Both crsA47 and retail suppressed the temperature sensitive sporulation phenotypes (Spots) caused by the ery-l, strA39, and kan-25 ribosomal protein mutations, a mutation in elongation factor EF-G fus-20, and two RNA polymerase mutations rif-14 and sty710. At nonpermissive temperatures the various defective-sporulation mutants produced spores at 0.01-0.6% of the wild type level. Introduction of either the crsA47 or retAil suppressors increased the sporulation of these strains to 10-126% of the wild type level. In general, the results were qualitatively similar to the results obtained with the rev-4 suppressed derivatives of these same spo ts mutations [4]. However, the quantitative aspects of suppression varied substantially between crsA47, rvtAll (table II), and rev-4 [1] mutations. Although crsA47 significantlv improved sporulation competence when introduced into the kan-9 strain (RS2780), retail failed to suppress the oligosporogenous phenotype of this strain. The rev-4 mutation does suppress kan-9, increasing sporulation 800-fold. It was not possible to assess the effects of crsA47 on nee-162 sporulation due to the instability of this strain, rvtAll was able to suppress the nee-162 sporulation defect.
Restoration of flagellin synthesis in spoO mutants containing rvtA or crsA suppressors Flagellin, a 37 kDa major membrane protein, is readily identifed on SDS-PAGE gels [13]. The spoOB, OE & OF mutants produce significantly lower amounts of this protein when compared with the wild type strain (fig 1). Flagellin production is restored to wild type levels in rvtA suppressed derivatives of these spoO strains. However, this restoration of flagell i n expression is not seen in crsA suppressed spoO strains (fig 1).
1166
C Ng et al
Table II. Suppression of defective-sporulation phenotypes by crsA47 and rvtAll in 2 x SG medium.
Sporulation (sporeslml) Strain
Genotype
Non-permissive (% wt) temperature
Permissive (% wt) temperature
Sporulation ratio suppressed/unsuppressed (nonpermissive temperature)
Suppression ofrif-14and neo-162: nonpermissive temperature 48°C, permissive temperature 30°C RS2705 RS2710 RS2711 RS2712 RS2770 RS2771 RS2772
168 168 rif.14 168 rif-14 crsA47 168 rif-14 rvtAl l 168 neo-162 168 neo-162 crsA47 168 neo-162 rvtAll
9.0 x 10s 1.0 x 10s 6.7 x 108 !.2 x 10s 2.0 × 106 No data 1.8 × 10s
(100) (0,01) (74) (13) (0.2)
2.6 x 108 s (100) 2.8 x 108 s (108)
6700 1200
2.0 × 108 s (77) 105
(20)
Suppression ofery-l, stv-710, and kan-25: nonpermissive temperature 49°C, permissive temperature 300C RS2705 RS2720 RS2721 RS2722 RS2750 RS2751 RS2752
168 168 ery-I i68 ery-I crsA47 168 ery-I rvtAl l 168 sty-710 168 sty-710 crsA47 168 sty-710 rvtAl l
5.0 x 108 3.2 × 106 6.3 x 108 4.8 x 107 1.8 x 106 9.5 × 107 1.5 x 108
(100) (0.6) (126) (10) (0.4) (19) (30)
2.6 x 108 a (100) 6.8 x 108 s (262) 200 15 53 83
Suppression of su'A39 andfus-20: nonpermissive temperature 49°C, permissive temperature 37°C RS2705 RS2730 RS2731 RS2732 RS2740 RS2741 RS2742
168 168 strA39 168 strA39 crsA47 168 strA39 rvtAll
168fus-20 168fus-20 crsA47 168ft, s-20 rvtAll
5.0 x 10s 1.2 x 106 3.4 x 108 8.0 x 107 3.0 x 106 2.8 x 108 1.9 X 10s
(100) (0.2) (68) (16) (0.6) (56) (38)
2.0 × 2.9 × 3.2 × 3.5 x
(100) (15) (160) (18)
1.5 x 103 a (100) 1.5 x !08 s (100) 280 67 2.8 x 108 s (187) 93 63
Suppression of oligosporogenous kan-9: 37°C RS2705 RS2780 RS2781 RS2782
168 168 kan-9 168 kan-9 crsA47 168 kan-9 rvtAll
108 107 10s 107
Restoration of PBP 5* synthesis in spoO mutants containing rvtA or crsA suppressors It has been reported that spoOA and spoOH strains do not produce PBP 5* [11]. Similarly, spoOB, spoOE, and spoOF mutants also fail to produce PBP 5* (data not shown). Introduction of crsA or rvtA into various spoO strains restores PBP 5* production, although often at later times and in lower quantity than was observed in the wild type strain (fig 2).
11 1.2
Effect of ethanol on flagellin and PBP 5* synthesis Ethanol inhibits normal sporulation, synthesis of flagellin (data not shown) and PBP 5* synthesis in wild type cells (fig 2a). Wild type cells arrest sporulation at stage 0 in 2 x SG medium containing 4% ethanol [18, 19]. rvtA and crsA strains are resistant to the sporulation inhibiting effects of ethanol [2-7]. We have investigated the effects of ethanol addition on sporulation, PBP 5* synthesis, and flagellin synthesis
Sporulation transcriptionfactor interactions
1167
Fig 1. SDS gel electrophoresis of membrane-associated proteins from spoO mutants and spoO mutants containing rvtA or crsA. Data are presented for the following strains: (A) IA96 (wild type); (B) RS5101 (168 rvtAll); (C) 1S16 (spoOB136); (D) RS5011 (spoOB136 rvtAll); (E) IS17 (spoOF221); (F) RS5005 (spoOF221 rvtAll); (G) IS24 (spoOHll6); (H) IS26 (spoOJ87); (I) 1S28 (spoOK141); (J) RS701 (crsA47); (K) RS5006 (spoOF221 crsA47); (L) RS5052 (spoOKl41 crsA47); (M) Glu-47 (crsA47 strA); (N) RS2 (168 wild type); (O) IS17 (spoOEll); (P) RS5021 (spoOEll rvtAll). The arrows indicate the position of flagellin. Each lane contains approximately the same amount of protein.
in a cogenic set of strains containing the rvtA mutation. Wild type strain IA96 produces greatly reduced levels of PBP 5* when incubated in the presence of ethanol (fig 2a). A spoOF rvtA strain is able to sporulate and synthesize PBP 5* when grown in the presence of ethanol (fig 2b). Ethanol addition inhibited the postexponential phase synthesis of flagellin in the wild type strain IA96. Ethanol addition also inhibited the synthesis of flagellin in the spoOF rvtA strain, even though PBP 5* synthesis and sporulation occurred.
Discussion Genetic analysis has identified over eighty genes, mutations in which the sporulation process ~s prevented [24, 33]. More than half of these genes have been cloned [33]. The functions of many of these cloned sporulation genes have been identified [33]. However,
little is understood about the functional connections among genes which act either to determine stagespecific development, or to establish the interconnections between groups of genes which allow the progression of development from one stage to the next. Intergenic suppression analysis provides a powerful tool for probing these functional interrelationships and connections [2-4]. The rvtA and crsA suppressor mutations employed in these studies were originally identified as suppressors of spoOB, spoOE, and spoOF spomlation defects [2, 4]. In addition, crsA47 is able to suppress spoOK, spollN279, spolIF and spollJ mutations [25, 31]. The results reported here demonstrate that the crsA47 and rvtAll mutations are also epistatic to Spots phenotypes caused by mutations in RNA polymerase [3-subunit, ribosomal 50S and 30S protein, and protein elongation factor EF-G genes. These data not only expand the epistatic domain of defective-sporulation mutations suppressed by crsA and rvtA, but, more import-
C Ng e t al
1168 4" A
°
I
--
÷
il
Q
Fig 2. Fiuorographs of penicillin binding proteins from (A) strain IA96 (wild type) and (B) spoOF rvtA. The cultures were grown in the absence (-) or presence (+) of ethanol. Ethanol, when present, was added to a concentration of 4% at mid log phase. Samples were removed at mid log phase, at the beginning of the postexponential phase (to), and at 1 to 8 h into the postexponential phase (h to ts). Membrane proteins were separated by electrophoresis, and penicillin binding proteins were identified by labelling with [3H]penicillin.
antly, they suggest a functional relationship between the translational and transcriptional components that affect sporulation efficiency, and the products of the
early sporulation genes spoOB, spoOE, spoOF, spoOK, spollF, spollJ and spollN. With the exception that rvtAll does not suppress the oligosporogenous mutation kan-9, all three intergenie suppressor mutations are epistatic to the same defective-sporulation phenotypes produced by alterations in translational and transcriptional components [4]. However, these domains of epistasis are not precisely congruent since rev is unable to suppress spoOK and spollN279 mutations [4], which are rescued by crsA [25, 31]. The effects of rev on other spoO and spoil mutations have not been analyzed. The same spots transcription and translation system mutations that are suppressed by rev, crsA, and rvtA, are also physiologically suppressed by the addition of compounds such as ribose, fructose, mannose, glycerol, fatty acid esters, or Tween to the growth medium [21, 26]. It has been suggested that the defectivesporulation phenotypes caused by mutations in translational and transcriptional components are the result of subtle metabolic changes in RNA or protein synthesis that alter cellular metabolic balance and principally affect membrane structure and function [21 ]. In these studies we have examined the effects of the rvtA and crsA mutations on the synthesis of two sporulation-associated membrane proteins PBP 5* and flagellin. rvtA suppression of SpoO phenotypes is accompanied by a restoration of the synthesis of flagellin and PBP 5* (figs 1 and 2). crsA suppression of SpoO phenotypes restores only PBP 5* synthesis (figs 1 and 2). Since flagellin has no known role in sporulation, these results suggest that the function of the rvtA mutations in the spoOA gene may extend beyond effects on sporulation-specific pathways. These results also demonstrate a close relationship between expression of PBP 5* and sporulation competence. Flagellin, the product of the hag gene, is transcribed by a minor form of RNA polymerase containing o2s [15-17]. We had shown previously that rvtA was able to restore transcription from o2s promoters in spoO backgrounds [16, 17]. In the same set of experiments the presence of the crsA suppressor did not restore o2s promoter activity concomitant with suppression of spoO sporulation defects [ 17]. One limitation of these previous studies was that the protein products encodea by the 62s promoters analyzed were unknown. The present studies extend and confirm these previous data through analysis of protein expression from the hag gene. The data presented establish that the crsA mutation affects spoO mutant gene expression and developmental competence by a mechanism distinct from that provided by rvtA. The most surprising findings of these studies are the overlapping effects of the rvtA, crsA and rev suppressors on RNA polymerase and ribosomal spots mutations. The results suggest that these lesions, and
Sporulation transcription factor interactions several spoO and spoil mutations all affect a single common aspect of developmental regulation. What are the molecular mechanisms underlying these inter-relationships? Clues are available from the molecular identity and functions encoded by the suppressor genes and their targets. The rvtA mutations occur within the spoOA gene, an environmental response regulator required for the trans-activation of early sporulation genes [34] and the o28 transcription system. The SpoOA protein is expressed in vegetative cells, but is in an inactive state. Protein kinase sensors are activated upon nutrient deprivation, which in turn convert SpoOA into an active state by phosphorylation [34]. SpoOB, SpoOE and SpoOF are required for efficient phosphorylation of SpoOA [32, 34]. We postulate that the rvtA mutations, which occur within the phosphorylated activator domain of SpoOA (Leighton et al, submitted), create a form of SpoOA which is more readily phosphorylated, less readily dephosphorylated, or active in the absence of phosphorylation. Such mutations would bypass the requirement for 'upstream' gene products involved in the activation of SpoOA (ie, SpoOB, SpoOE and SpoOF). The Crs (catabolite resistant sporulation) phenotype of rvtA mutations is also understandable if one presumes that glucose downregulates starvation protein kinase sensor activity such that a critical concentration of SpoOA phosphate is not achieved in wild type cells. However, in rvtA strains the more readily phosphorylated form of SpoOA could reach a critical threshold level of phosphorylation in the presence of diminished protein kinase activity. Other Crs mutations within the N-terminal activator domain of SpoOA have been shown to be more readily phosphorylated by non-cognate protein kinases [34]. The crsA47 mutation occurs within the major RNA polymerase sigma subunit gene rpoD [7]. Recent studies of this mutation (Leighton et al, submitted) have shown that crsA RNA polymerase is defective in the transcription of a 'switch' gene sin, which represses the sporulation pathway, and directs cells into a motile, competent, non-sporulating postexponential cell state [34]. We postulate that the crsA mutation 'channels' cells into the main sporulation pathway and represses alternate non-sporulation pathways. Many of the gene products involved in chemotaxis, motility and competence are protein kinase substrates which would compete with SpoOA in phosphorylation reactions [34]. The repression of these competing substrates could explain the Crs phenotype of crsA, ie more protein kinase flow to SpoOA, and the suppression of 'upstream' spoOB, spoOE, s p o O F and s p o O K mutations, sin is also a transcriptional repressor of several stage II genes (Smith, Leighton et al, submitted), explaining the ability of crsA to suppress certain spoil mutations.
1169
Conclusion The results of the described intergenic suppression studies suggest that RNA polymerase core subunit and ribosomal protein genes are intimately involved in the starvation-activated response system which 'resets' the global pattern of gene expression during the transition from vegetative to postexponential phase. The fact that mutations in these genes are suppressible by spoOA and rpoD mutations suggests that transcription and translation system components have previously unexpected roles in developmental sensory transduction.
Acknowledgments We are grateful to HG Wittmann for his support and encouragement during these studies. This research was supported by grants from the National Science Foundation (ECE86-13227, TL) and the National lqstitutes of Health (GM-43564, CB).
References 1 Sharrock RA, Leighton T (1982) Suppression of defective-sporulation phenotypes by the Bacillus subtilis mutation rev-4. Mol Gen Genet 186, 432-438 2 SharrockRA, Rubinstein S, Chan M, Leighton T (1984) Intergenic suppression of spoO phenotypes by the Bacillus subtilis mutation iwtA. Mol Gen Genet 194, 260-264 3 LeungA, Rubinstein S, Yang C, Jing-Wei L, Leighton T (1985) Suppression of defective-sporulation phenotypes by mutations in the major sigma factor gene (rpoD) of Bacillus subtilis. Moi Gen Genet 201, 96-98 4 Sharrock RA, Leighton T (1981) lntergenic suppressors of temperature-sensitive sporulation in Bacillus subtilis are allele non-specific. Mol Gen Genet 183,532-537 5 Takahashi I (1979) Catabolite repression-resistant mutants of Bacillus subtilis. Can J Microbio125, ~1283-1287 6 Sun D, Takahashi 1 (1982) Genetic mapping of cataboliteresistant mutants of Bacillus subtilis. Can J Microbiol 28, 1242-125 I 7 Sun D, Takahashi I (1984) A catabolite-resistant mutation is localized in the rpo operon of Bacillus subtilis. Can J Microbio130, 423-429 8 Price CW, Doi RH (1985) Genetic mapping of rpoD implicates the major sigma factor of Bacillus subtilis RNA polymerase in the control of sporulation initiation. Mol Gen Genet 201, 88-95 9 Rhaese HJ, Reichenbach A, Stamminger G (1984) Changes in membrane-associated proteins during sporulation in Bacillus subtilis. J Biochem 139, 593--603 10 Andreoli AJ, Kao M, Chui R, Cabrera J, Wong SKS (1981) Two-dimensional polyacrylamidegel electrophoresis analysis of cytoplasmic and membrane proteins from sporulating cells and forespore and mother-cell compartments of Bacillus subtilis. In: Sporulation and Germination (Levinson HS, Sonenshein AL, Tipper DJ, eds) American Society for Microbiology, Washington, 168 p
1170
C Ng et
11 Sowell MO, Buchanan CE (1983) Changes in penicillinbinding proteins during sporulation of Bacillus subtilis. J Bacteriol 153, 1331-1337 12 Buchanan CE, Sowell MO (1983) Stability and synthesis of penicillin-binding proteins during sporulation. J Bacteriol 156, 545-551 13 Aubert E, Klier A, Rapoport G (1985) Cloning and expression in Escherichia coli of the regulatory sacU germ from Bacillus subtilis. J Bacteriol 161, 1182-1187 14 Helmann JD, Chamberlin MJ (1988) Structure and function of bacterial sigma factors. Annu Rev Biochem 57, 839-872 15 Mirel DB, Chamberlin MJ (1989) The Bacillus subtilis flagellin gene (hag) is transcribed by the ~-28 form of RNA polymerase. J Bacteriol 171, 3095-3101 16 Gilman MZ, Glenn JS, Singer VL, Chamberlin MJ (1984) Isolation of sigma-28-specific promoters for Bacillus subtilis DNA. Gene 32, 11-20 17 Chamberlin MJ, Briat JF, Singer VL, Glenn JS, Helmann J, Leung A, Moorefield MB, Gilman MZ (1985) Isolation of genomic segments from Bacillus subtilis that contain promoters for a2s RNA polymerase, and studies on regulation of their transcription in Bacillus subtilis and Escherichia coli. In: Molecular Biology of Microbial Differentiation (Hoeh J, Setlow P, eds) American Society for Microbiology, Washington, 143 p 18 Bohin JP, Rigomier D, Sehaeffer P (1976) Ethanol sensitivity of sporulation in Bacillus subtilis: a new tool for the analysis of sporulation process. J Bacteriol 127, 934-940 19 Rigomier D, Bohin JP, Lubochinsky B (1980) Effects of ethanol and methanol on lipid metabolism in Bacillus subtilis. J Gen Microbiol 121,139-149 20 Takahashi I, MacKenzie LW (la~'~...,,,.,Effects of various inhibitory agents on sporulation of Bacillus subtilis. Can J Microbio128, 80-86 21 Wayne RR, Price CW, Leighton T (1981) Physiological suppression of the temperature-sensitive sporulation defect in a Bacillus subtilis RNA polymerase mutant. Mol Gen Genet 183, 544-549 22 Sharrock RA, Leighton T, Wittmann HG (1981) Macrolide and amino-glycoside antibiotic resistance mutations in the Bacillus subtilis ribosome resulting in temperature-sensitive sporulation. Mol Gen Genet 183, 538-543
al
23 24 25
26
27 28 29 30 31
32
33
34
Buchanan CE (1979) Altered membrane proteins in a minicell-producing mutants of Bacillus subtilis. J Bacteriol 139, 305-307 Piggot PJ, Coote JC (1976) Genetic aspects of bacterial endospore formation. Bacteriol Rev 40, 908-962 Leung A, Ng D, Yang C, Rubinstein S, Deede C, Putnam W, Leighton T (1985) The Bacillus subtilis spoO regnlon: intergenic suppression analysis of initiation control. In: Molecular Biology of Microbial Differentiation (Hoch J, Setlow P, eds) American Society for Microbiology, Washington, 176 p Wayne RR, Leighton T (1981) Physiological suppression of Bacillus subtilis conditional sporulation phenotypes: RNA polymerase and ribosomal mutations. Mol Gen Genet 183, 550-552 Leighton TJ (1977) New types of RNA polymerase mutations causing temperature-sensitive sporulation in Bacillas subtilis. J Bioi Chem 252, 268-272 Leighton TJ (1973) An RNA polymerase mutation causing temperature-sensitive sporulation in Bacillus subtilis. Proc Natl Acad Sci USA 70, 1179-1183 Leighton TJ (1974) Sporulation-specific translational discrimination in Bacillus subtilis. J Mol Bio186, 855-863 Gilman M, Chamberlin M (1983) Development and genetic regulation of Bacillus subtilis genes transcribed by a 28RNA polymerase. Cell 35, 285-293 Stragier P (1989) Temporal and spatial control of gene expression during sporulation: from facts to speculation. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 243 p Trach K, Chapman JW, Piggot P, LeCoq D, Hoch JA (1988) Complete sequence and transcriptional analysis of the spoOF region of the Bacillus subtUis chromosome. J Bacteriol 170, 4194-4208 Piggot PJ (1989) Revised genetic map of Bacillus subtilis 168. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 1 p Smith I (1989) Initiation of sporulation. In: Regulation of Procaryotic Development (Smith I, Slepecky RA, Setlow P, eds) American Society for Microbiology, Washington, 185 p
