RNA Polymerase from Sporulating Bacillus subtih PURIFICATION AND PROPERTIES OF A MODIFIED FORM OF THE ENZYME CONTAINING TWO SPORULATION POLYPEPTIDES*

A new form of DNA-dependent RNA polymerase termed enzyme III has been purified from sporulating cells of Bacillus subtilis. In addition to the subunits of core RNA polymerase (/3’, 8, LY, and w), enzyme III contains sporulation-specific polypeptides of 85,000 (P”“) and 27,000 (P”‘) daltons. Pas corresponds to an RNA polymerase-binding protein previously identified by precipitation of RNA polymerase from crude extracts of sporulating cells with antibody directed against core enzyme. Both Pas and PZ’ co-purified with RNA polymerase highly purified by gel filtration, DEAE-cellulose chromatography, phosphocellulose chromatography, and glycerol gradient centrifugation. Enzyme III bound more tightly to phosphocellulose and sedimented more rapidly during zone centrifugation than did RNA polymerase lacking the sporulation polypeptides. RNA polymerase containing P 86 and PZ7 transcribed B. subtilis DNA about 4.5 times more actively than did core RNA polymerase, although both enzymes exhibited similar activities with poly(dA-dT) and phage @e DNA as templates. Enzyme III and core RNA polymerase also differed in their response to increasing concentrations of MgZ+ and KCl.

III contains sporulation-specific polypeptides of 85,000 (P"") and 27,000 (P"') daltons. Pas corresponds to an RNA polymerase-binding protein previously identified by precipitation of RNA polymerase from crude extracts of sporulating cells with antibody directed against core enzyme. Both Pas and PZ' copurified with RNA polymerase highly purified by gel filtration, DEAE-cellulose chromatography, phosphocellulose chromatography, and glycerol gradient centrifugation. Enzyme III bound more tightly to phosphocellulose and sedimented more rapidly during zone centrifugation than did RNA polymerase lacking the sporulation polypeptides. RNA polymerase containing P 86 and PZ7 transcribed B. subtilis DNA about 4.5 times more actively than did core RNA polymerase, although both enzymes exhibited similar activities with poly(dA-dT) and phage @e DNA as templates. Enzyme III and core RNA polymerase also differed in their response to increasing concentrations of MgZ+ and KCl.
The DNA-dependent RNA polymerase of Bacillus subtilis is thought to undergo modifications during the process of spore formation (for a review, see Ref. 1). The onset of sporulation, for instance, is associated with a marked decrease in the activity of the (r subunit of RNA polymerase (2,3). The inhibition of D is a specific event in the process of sporulation since it is prevented in mutants blocked at an early stage of spore formation, including certain mutants of RNA polymerase isolated by resistance to rifampicin (4,5). Extracts of sporulating cells appear to contain a specific component that inhibits (T activity probably by interfering with the binding of u to core RNA polymerase (5)(6)(7).
While lacking (r, RNA polymerase from sporulating cells can be isolated in association with a large sporulation-specific polypeptide (8). This species was first identified by precipitation of RNA polymerase from crude extracts of sporulating cells with antibody directed against core enzyme from vegetative bacteria (8). A reconstitution experiment confirmed that this sporulation polypeptide binds specifically to RNA polymerase. The binding protein first appears early during sporulation and is either absent or present in greatly reduced amounts in enzyme from stationary phase cells of mutants blocked early in spore formation (9). Here we report on the purification and properties of a form of RNA polymerase con- subtilis DNA about 4.5 times more actively than did enzyme II, although both forms of RNA polymerase exhibited similar activities with either poly(dA-dT) or phage @e DNA as templates ( Figs. 1 and 2). This difference in template preference was confirmed under conditions of linear dependence of enzyme activity on protein concentration (Fig. 2). Enzymes II and III also differed in their response to MgZ+. Transcription of the bacterial DNA by enzyme III was stimulated more markedly by increasing concentrations of MgZ+ than was RNA synthesis by enzyme II (Fig. 3).
Another difference between enzymes II and III was in their sensitivity to KCl. Although transcription of B. subtilis DNA by both enzymes was inhibited by KC1 (Fig. 4c), increasing ionic strength stimulated transcription of phage @e DNA by enzyme III but partially inhibited RNA synthesis by enzyme II (Fig. 4~). In contrast, KC1 slightly stimulated transcription of poly(dA-dT) by enzyme II while having little effect on the transcription of the synthetic template by enzyme III (Fig.  46). For comparison, the effect of KC1 on enzyme II was similar to that observed for core RNA polymerase from vegetative bacteria. For example, the addition of 0.15 M KC1 decreased activity with 4e DNA as template by about 50% and stimulated transcription of poly(dA-dT) by about 40%. This finding is consistent with the gel analyses described below that indicate that enzyme II corresponded to core RNA polymerase.
(A more detailed account of the effect of KC1 on RNA synthesis by vegetative core polymerase and holoenzyme has been presented previously (lo).) Slab gel electrophoresis of RNA polymerase in fractions from the phosphocellulose gradient elution revealed polypeptides of 85,000 and 27,000 daltons (and small amounts of a 65,000-dalton species) that co-eluted with the subunits of core RNA polymerase (/3', /3, (Y, and w) in enzyme III (Fractions 20 to 22; Fig. 5). Gel analysis also revealed a contaminating polypeptide of 60,000 daltons that was eluted in fractions throughout the gradient and a 90,000-dalton contaminant that largely eluted after enzyme III. The stoichiometries of the 85,000-and 27,000-dalton species in enzyme III (hereafter referred to as Pas and P*', respectively) were between 0.5 and 1.0 per core RNA polymerase.
Both of these species were absent in enzyme II (fractions 14 to 16; Fig. 5  more rapidly (Fig. 6). (This small difference in sedimentation coefficient was reproducibly observed in several independent experiments.) As evidence of high purity, enzyme activity at this stage of purification was coincident with a peak of protein (17). Fig. 7 compares the subunit structures of glycerol gradient-purified enzymes II and III in the fractions of peak activity with enzyme III in Fraction 21 (Fig. 5) of the phosphocellulose gradient elution. Enzyme II contained p', j3, (Y, and w and appeared to correspond to core RNA polymerase. Glycerol gradient-purified enzyme III contained Pa6 and P*' (Fig. 7) and slab gel analysis of enzyme in fractions across the glycerol gradient (not shown) revealed that these species had co-sedimented with the core subunits of enzyme III. However, the stoichiometries of these sporulation polypeptides had diminished somewhat during sedimentation of enzyme III through the glycerol gradient of high salt concentration.
Glycerol gradient-purified enzyme III transcribed B. subtilis DNA more actively than did the purified enzyme II.
To determine whether Pa5 corresponded to the RNA polymerase-binding protein previously identified by antibody precipitation (8), RNA polymerase was precipitated from a crude extract of radioactively labeled sporulating cells by antibody directed against vegetative core enzyme. The antibody precipitate contained, in addition to core RNA polymerase, a polypeptide with the mobility of Pa6 (Fig. 8). In a previous study, we (8) showed that this species is sporulation-specific; also, a reconstitution experiment indicated that this antibodyprecipitated polypeptide binds to RNA polymerase. Although we had earlier estimated a molecular weight of 70,000, we now calculate a higher molecular weight using the molecular weight markers listed under "Materials and Methods." Since the antibody-precipitated polypeptide co-migrated with Ps5, we assume that these polypeptides are the same species.
Finally, we have noted that the relative amounts of enzymes II and III are apparently characteristic of the stage of sporulation. Phosphocellulose chromatography of RNA polymerase from cells harvested at the 5th hour of spore formation, for instance, yielded only enzymes I and III. This finding could indicate that Ps6 and P*' are most abundant at late times during the sporulation process. Nuclear) and the radioactivity was measured in a scintillation counter. A photograph of me slot containing the glycerol gradientpurified enzyme III is aligned below the radioactive profile. electrophoresis by the procedure of O'Farrell (18) indicated that Pss is absent in crude extracts of vegetative bacteria. Pas from crude extracts of sporulating bacteria streaked to some extent in the isoelectric focusing dimension of the twodimensional electrophoresis but was nevertheless resolved from other bacterial proteins (data not shown). We have not yet ,searched for Pa7 in crude extracts. Third, an earlier "double label" experiment in which Ru polymerase was isolated by antibody precipitation from a mixture of vegetative and sporulating cells separately labeled with two different radioisotopes had indicated that the large sporulation subunit of RNA polymerase is absent in enzyme from vegetative bacteria (8). An analogous double label experiment (not shown) confirms that P*' is also missing from vegetative RNA polymerase.
Phosphocellulose chromatography of RNA polymerase from stationary-phase cells of a mutant blocked at stage 0 of spore formation (SpoOa-5NA; Ref. 19) yielded only a small peak of enzyme III activity (17). Furthermore, two-dimensional electrophoresis (not shown) of crude extracts from these asporogenous bacteria indicated that SpoOa-5NA produced considerably less of Pas than did wild type sporulating cells. Thus, although this RNA polymerase-binding protein is not entirely absent in the asporogenous mutant, its accumulation during stationary phase is severely restricted.

DISCUSSION
This paper describes the isolation of a form of RNA polymerase (enzyme III) that contains, in addition to the subunits of core enzyme, sporulation-specific polypeptides Pas and P2'. Both of these species appear to be tightly associated with enzyme III since they co-purified with enzyme purified to apparent homogeneity by conventional purification pro-cedures. RNA polymerase containing the sporulation polypeptides differed from core enzyme (enzyme II) in both purification and transcriptional properties. Enzyme III eluted at a higher salt concentration during phosphocellulose chromatography and sedimented more rapidly during zone centrifugation than did enzyme lacking the sporulation polypeptides. Enzyme III also transcribed B. subtilis DNA more actively than did enzyme II and RNA synthesis by each of these enzymes was affected differently by MgZ+ and by KCl.
Subsequent to our isolation of an RNA polymerase-binding protein from sporulating cells (8), Nishimoto and Takahashi (20) reported the isolation of a form of B. subtilis RNA polymerase containing a sporulation-specific polypeptide of 95,000 daltons. The purification properties of this form of RNA polymerase are similar to those described here for enzyme III and, despite the discrepancy in molecular weight, their sporulation-specific subunit of RNA polymerase probably corresponds to the species described by this laboratory.
Recently, Fukuda et al. (21) have reported small polypeptides of 27,000 and about 20,000 daltons associated with RNA polymerase from sporulating B. subtilis.
B. subtilis RNA polymerase also acquires additional subunits following infection of vegetative bacteria with phages SPOl (16,22,23) and SP82 (24). Phage SPOl, for instance, induces several polypeptides that are associated with RNA polymerase isolated either by antibody precipitation or by conventional purification procedures (16,22,23). One of these polypeptides is now known to be coded by an SPOl regulatory gene' and a form of purified RNA polymerase containing this phage-specified protein and lacking u directs the specific transcription of SPOl "middle" genes in vitro (16,23). Although it is not known whether Pss or PST direct specific gene transcription, it is tempting to speculate that gene expression during spore formation might also be controlled by proteins that interact directly with RNA polymerase.