INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial spores are significantly more resistant to 254-nm UV than are the vegetative forms of the same species (38, 44). Our current understanding is that this high level of UV resistance is due mainly to two interlocking mechanisms. (i) Binding of spore DNA by small, acid-soluble spore proteins of the / class results in an alteration of the helical conformation, and hence the UV photochemistry, of dormant spore DNA to favor production of the unique spore photoproduct (SP) 5-thyminyl-5,6-dihydrothymine (reviewed in references 38-42). (ii) SP formed during exposure of dormant spores to UV is repaired during spore germination by two distinct DNA repair pathways, nucleotide excision repair (NER) and direct reversal of SP to thymine in DNA by an SP-specific enzyme called SP lyase (21-23; reviewed in references 14, 25, and 49).

NER in Bacillus subtilis closely resembles the analogous system which has been well characterized in Escherichia coli (18). NER is a general system capable of detecting and removing a variety of bulky (i.e., helix-distorting) lesions from DNA, including both SP and cyclobutane-type pyrimidine dimers (reviewed in references 14 and 49). SP lyase, in contrast, is specifically dedicated to the in situ monomerization of SP during spore germination (22, 23, 48), but to date little is known of the molecular mechanism of SP repair mediated by SP lyase. SP lyase is encoded by the splB cistron of the splAB operon (13). The splA cistron encodes a small, 9-kDa protein which is not needed for SP lyase activity and which apparently functions in the regulation of splAB operon expression (13, 24).

The results of early indirect experiments suggested that SP lyase was synthesized during either vegetative growth or sporulation and was packaged within the dormant spore (22, 23). Subsequent studies with lacZ fusions to the cloned B. subtilis SP lyase gene splB supported this conclusion; SP lyase is expressed exclusively in the developing forespore compartment during morphological stage III of sporulation as part of the sigma-G regulon of forespore-specific genes (31, 32). Early indirect experiments also indicated that SP disappeared from spore DNA during germination but did not appear in trichloroacetic acid-soluble material, suggesting that SP was not excised from high-molecular-weight DNA but was reversed directly to two thymines in situ (11, 23, 48). The direct reversal of SP by SP lyase is reminiscent of the action of DNA photolyases upon cyclobutane pyrimidine dimers (35), with the exception that SP lyase-mediated repair is light independent (23). Analysis of the deduced amino acid sequence of SP lyase genes cloned from B. subtilis (13) and Bacillus amyloliquefaciens (25) revealed that the two SP lyases indeed share a short stretch of amino acid sequence with members of the DNA photolyase-6-4 photolyase-blue light photoreceptor protein family of enzymes (8, 13, 47), but the potential significance, if any, of this homology is currently unknown.

Inspection of the amino acid sequences of the two cloned SP lyases also revealed that they each contain a total of four cysteine residues, three of which are clustered in an arrangement similar to that seen in certain iron-sulfur (Fe-S) proteins. A search of the protein sequence database with the region of B. subtilis SP lyase encompassing amino acids 80 through 115 indeed revealed substantial homology to Fe-S proteins such as the activating subunits of the anaerobic ribonucleotide reductases (RNR) and pyruvate-formate lyases (PFL) from E. coli, phage T4, Haemophilus influenzae, and Clostridium pasteurianum (24). Iron does appear to be associated with SP lyase activity, as the UV sensitivity of spores of a B. subtilis strain which relies only upon SP lyase for DNA repair is enhanced when this strain is germinated on solid medium lacking iron (24).

The above observations suggest that SP lyase may be an Fe-S protein, and further that class III anaerobic enzymes such as RNR, PFL, or lysine 2,3-aminomutase from C. pasteurianum (reviewed in reference 33) could serve as experimental paradigms for elucidating the molecular mechanism of SP lyase action. An important step towards this goal is the purification and physical characterization of SP lyase and an assay of its activity in vitro. To these ends, the present report describes the engineering of SP lyase containing an amino (N)-terminal histidine tag, its expression in and purification from B. subtilis spores or from an E. coli overexpression system, and preliminary characterization of its properties in vitro.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains, plasmids, and culture conditions. B. subtilis and E. coli strains used in this study are listed in Table 1. Plasmids and cloned fragments are described in Table 2. Media used were Difco sporulation medium (DSM) (37) and Luria-Bertani medium (19). For strains carrying the thyA1 thyB1 markers, thymidine was routinely added to media to a final concentration of 100 µg/ml. When appropriate, antibiotics were added to media at the following final concentrations: chloramphenicol, 3 µg/ml; ampicillin, 50 or 125 µg/ml; tetracycline, 15 µg/ml; erythromycin, 1 µg/ml; lincomycin, 25 µg/ml (the combination of erythromycin and lincomycin is hereafter referred to as MLS). Labelling of spore DNA by growth and sporulation of B. subtilis cultures in medium containing [methyl-³H]thymidine was performed as described previously (45). All liquid cultures were incubated at 37°C with vigorous aeration, and optical density was monitored with a Klett-Summerson colorimeter fitted with a no. 66 (red) filter. Spore production, purification, and germination conditions have all been described previously (27), as have methods for assaying spore UV resistance (12, 13, 27).

Genetic and molecular biology techniques. Preparation of competent E. coli or B. subtilis cells and their transformation with DNA were performed as previously described (6, 34). Large- and small-scale extractions of chromosomal DNA from B. subtilis (9) and plasmid DNA from E. coli (5, 34) were accomplished by standard techniques. Extraction of SP-containing methyl-³H-labelled DNA from spores was previously described (45). Standard techniques were used throughout for enzymatic manipulations and agarose gel electrophoresis of DNA (34), for in vitro mutagenesis (12), and for nucleotide sequencing (13, 36).

Construction of N- and C-terminal SP lyases. SP lyase carrying a C-terminal tag of 6 histidine residues (6×His) was constructed by PCR amplification of the splB gene from plasmid pWN41 (Table 2) with a pair of oligonucleotide primers, 5'-GGTCTAGAGGAAAAGGATGTGGC-3' and 5'-CCGTCGACTTAGTGATGGTGATGGTGATGAGTGAAATATTCAATTTTTGC G-3', followed by a series of cloning steps resulting in plasmid pWN376 (Table 2). SP lyase carrying an N-terminal histidine tag was constructed with a commercial kit (Altered Sites; Promega) by in vitro site-directed mutagenesis of the wild-type splAB operon cloned in plasmid pALTER-1 (plasmid pWN160; Table 2) with the mutagenic oligonucleotide 5'-GGATGTGGCATCATGCACCATCACCATCACCATCAGAACCCATTTGTTCCG-3', resulting in plasmid pWN406. The modified splAB operons were inserted into the amyE locus of B. subtilis by their removal from plasmids pWN376 and pWN406 and ligation into EcoRI-HindIII-cleaved plasmid pDG364, resulting in plasmids pWN378 and pWN413, respectively (Table 2). All final plasmid constructions were confirmed by nucleotide sequencing (36).

SP lyase purification and protein techniques. B. subtilis WN417 (1L) was sporulated for 48 h in DSM containing 3 µg of chloramphenicol and 100 µg of thymidine per ml. Spores were harvested, purified as described previously (27), heat shocked at 80°C for 10 min, and stored at 4°C in water until use. The spore coat layer was removed from purified spores by SDS-urea-dithiothreitol (DTT) treatment as described previously (27), and washed decoated spores were lysed by incubation in sonication buffer (50 mM sodium phosphate buffer [pH, 8.0], 300 mM NaCl, and 10 mM imidazole) containing lysozyme (1 mg/ml final concentration) at 37°C for 30 min. Thereafter, the sample was kept on ice or at 4°C. The lysate was sonicated six times for 10-s intervals to lyse cells and shear DNA and centrifuged (12,100 × g, 20 min), and the cell-free extract was loaded onto a nickel-nitrilotriacetic acid (NTA) agarose column (Superflow, Qiagen Inc.) equilibrated in sonication buffer. The column was routinely washed with sonication buffer containing 20 mM imidazole, and the bound protein was eluted with sonication buffer with 250 mM imidazole, and 0.5-ml fractions were collected. The concentration of protein in each fraction was determined by the Bradford assay (7) (Sigma).

Immunologic techniques. SP lyase containing an N-terminal 10×histidine tag was overproduced in an E. coli expression plasmid system (generously provided by C. Kinsland and T. Begley) and purified by nickel-NTA agarose chromatography as described above. Polyclonal goat antiserum was prepared by the following procedure (4). An adult male goat was immunized with 0.5 mg of SP lyase in Ribi Adjuvant System (Ribi ImmunoChem Research Inc.) administered intramuscularly in the hind leg at two sites. A similar booster immunization of 0.5 mg of SP lyase in Ribi Adjuvant was administered 28 days later. Blood was collected on day 42, and antiserum was harvested from the clotted blood. Preimmune serum collected in a similar manner prior to immunization did not demonstrate any reactivity to SP lyase (data not shown). Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting were performed essentially as described by Ausubel et al. (1). Immunostaining of Western blots was performed with rabbit anti-goat immunoglobulin G-peroxidase conjugate (Kirkegaard & Perry Laboratories, Inc.) as the secondary antibody.

Physicochemical techniques. UV-visible spectroscopy of purified SP lyase from the E. coli overexpression system (0.8 mg of protein/ml in 50 mM Tris-HCl [pH 8.0], 300 mM NaCl, 250 mM imidazole) was performed with a Hewlett-Packard Model 8453 spectrophotometer by using the same buffer without protein as a blank.

SP lyase assays. SP lyase assays from crude spore extracts were performed by a modification of the protocol described for RNR by Ollagnier et al. (30). Stock solutions (prepared in 30 mM Tris-HCl [pH 8.0]) of dithionite (100 mM), DTT (50 mM), AdoMet (32 mM), and NADPH (10 mM) were purged of oxygen by bubbling a mixture of 80% N₂-10% H₂-10% CO₂ through them for 20 min. Spores of strain WN417 were decoated and lysed as described above in 30 mM Tris-HCl (pH 8.0), followed by deoxygenation of the cell-free extract on ice for 20 min. Reaction mixtures (200 µl total volume) were assembled on ice in the following order: to cell-free extract (120 µl, containing approximately 200 µg of total protein) were added (final concentration) dithionite (1 mM), DTT (5 mM), NADPH (1 mM), AdoMet (1 mM), and 50,000 cpm of SP-containing spore DNA labelled with [methyl-³H]thymidine. Reaction mixtures were incubated in a 37°C water bath overnight. Samples were precipitated by addition of ice-cold TCA to 10% final concentration and incubation on ice for 30 min. The TCA precipitates were centrifuged, and the pellet was used for SP quantitation as described below.

SP quantitation. Germination samples and in vitro reaction mixtures were precipitated on ice for 30 min with 10% TCA and centrifuged. The pellets were resuspended in 0.5 ml trifluoroacetic acid (TFA), sealed under vacuum in 2.0-ml glass ampules, and hydrolyzed at 155°C for 60 min. TFA was evaporated, and the dried contents in each ampule were resuspended in 100 µl of water and subjected to descending chromatography on Whatman no. 1 paper with 80:12:30 (vol/vol/vol) n-butanol-acetic acid-water. The resulting chromatograms were cut into 1-cm fractions, and SP (R_f, 0.45) and thymine (R_f, 0.6) were quantitated by scintillation counting as described in detail previously (10, 26, 45).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

UV resistance of spores carrying engineered SP lyases expressed at amyE. Construction and characterization of B. subtilis WN356, which lacks NER and in which the splAB operon has been deleted and replaced with an MLS resistance cassette, have been described in detail elsewhere (24). Plasmids pWN378 (encoding SP lyase with a C-terminal histidine tag) and pWN413 (encoding SP lyase with an N-terminal histidine tag) (Table 2) were inserted into the amyE locus by transformation into competent cells of strain WN356, resulting in strains WN390 and WN417, respectively (Table 1). The ability of the two engineered SP lyases to restore UV resistance to spores of strain WN356 was then tested (Fig. 1). It was observed that SP lyase with a C-terminal histidine tag expressed from amyE (strain WN390) conferred only slightly more UV resistance to spores than did insertion of the vector alone into the amyE locus of strain WN356 (Fig. 1). In contrast, strain WN417, which from the amyE locus expressed SP lyase with the N-terminal histidine tag, produced spores whose UV resistance was virtually indistinguishable from that of spores of strain WN386, a strain which carries the wild-type splAB operon at amyE (24) (Fig. 1). This in vivo experiment indicated that SP lyase biological activity was nearly abolished by addition of six histidines at its C-terminus, but was not significantly affected by addition of six histidines at its N-terminus. Therefore, strain WN417 was used in subsequent experiments.

View larger version (17K): [in this window] [in a new window]   FIG. 1.   Spore UV resistance of B. subtilis strains carrying various engineered alleles of SP lyase expressed from the amyE locus. Strains used were WN385 (vector only; open circles), WN386 (wild-type SP lyase; open triangles), WN390 (C-terminal histidine-tagged SP lyase; closed circles), and WN417 (N-terminal histidine-tagged SP lyase; closed triangles).

Repair of SP during spore germination by histidine-tagged SP lyase. To confirm that the enhanced UV resistance observed in spores of strain WN417 (Fig. 1) was indeed due to DNA repair by SP lyase carrying the N-terminal histidine tag, SP repair was assayed during germination of UV-irradiated spores of strain WN417 and its parent strain WN356 (Fig. 2). The DNA in spores of strain WN417 and its parent strain WN356 was labelled by growth and sporulation in DSM supplemented with [methyl-³H]thymidine, spores were irradiated with UV to produce SP, and the kinetics of SP repair during germination was monitored (Fig. 2). It was observed that strain WN356, lacking NER and SP lyase, was unable to remove SP from germinating spore DNA, whereas strain WN417, carrying the N-terminal histidine-tagged SP lyase expressed from amyE, removed approximately 65% of SP from DNA during the first 90 min of germination (Fig. 2). Therefore, it appeared that SP repair during germination, and hence UV resistance, of strain WN417 were due to the activity of N-terminal histidine-tagged SP lyase.

View larger version (20K): [in this window] [in a new window]   FIG. 2.   Kinetics of SP repair during germination of spores of B. subtilis WN356 (open circles) and WN417 (closed circles) in DSM containing 10 mM L-alanine. Before germination, spores were irradiated with UV to obtain the following amounts of SP (expressed as percentage of total thymine): WN356, 1.1% and WN417, 2.1%. Results shown are the averages ± standard deviations of two (WN356) or three (WN417) independent experiments.

Purification of histidine-tagged SP lyase. Dormant spores of strain WN417 were disrupted, the cell debris was removed by centrifugation, and the cell-free extract was passed through a nickel-NTA agarose column. Because SP lyase is present at very low quantities in dormant spores, SP lyase was visualized during purification by Western blot analysis with polyclonal antiserum raised against purified SP lyase from an E. coli overexpression system. Decoated, lysed spores of strain WN417 exhibited a 41-kDa protein which reacted with anti-SP lyase antiserum (Fig. 3, lane 2) and whose mass corresponded very closely to the calculated molecular mass of N-terminal histidine-tagged SP lyase (40,847 Da). The 41-kDa protein was absent from decoated, lysed spores of the parental strain WN356 (Fig. 3, lane 1). The 41-kDa protein band persisted in the cell-free extract after the spore lysate of strain WN417 was clarified by high-speed centrifugation (Fig. 3, lane 3), and the protein bound to a nickel-NTA agarose column, as it was absent from the flow-through fraction at 20 mM imidazole (Fig. 3, lane 4) but was present in the 250 mM imidazole eluate from the column (Fig. 3, lane 5). The 10×His-tagged SP lyase overexpressed in E. coli and isolated by the same procedure was essentially pure, demonstrating a single 43-kDa band on Coomassie blue-stained SDS-PAGE gel which (i) reacted with anti-SP lyase antibody on a Western blot and (ii) when subjected to automated N-terminal sequencing was found to have a sequence that perfectly matched the predicted amino acid sequence of the engineered 10×His-tagged SP lyase (data not shown).

View larger version (39K): [in this window] [in a new window]   FIG. 3.   SDS-10% PAGE-Western blot analysis of SP lyase purification. The 43- and 30-kDa protein molecular mass markers were visualized by staining with Ponceau S. Lane 1, lysed spores of strain WN356; lane 2, lysed spores of strain WN417; lane 3, cell-free extract of strain WN417 after clarification by high-speed centrifugation; lane 4, flow-through from nickel-NTA column; lane 5, eluate from nickel-NTA column with 250 mM imidazole.

Purified SP lyase contains iron and sulfur. During purification of the 10×His-tagged SP lyase overexpressed in E. coli, it was noted that as the cell-free extract was allowed to flow through the nickel-NTA agarose column a reddish-brown band appeared at the top of the column, which remained after washing the column with sonication buffer containing 20 mM imidazole (reference 2 and our unpublished results). Upon application of a linear 25 to 500 mM imidazole gradient, the colored band eluted at approximately 120 mM imidazole, along with SP lyase (data not shown). This observation, along with the characteristic smell of sulfide upon treatment of the purified SP lyase with acid, was consistent with SP lyase containing an Fe-S cluster. As we were unable to purify sufficient quantities of SP lyase from dormant B. subtilis spores to perform chemical analyses, we assayed the purified 10×His-tagged SP lyase obtained from the E. coli overexpression system for iron, acid-labile sulfide, and flavin content (28) (Table 3). The SP lyase purified from E. coli was found to contain 1.03 ± 0.26 mol of iron and 1.55 ± 0.04 mol of acid-labile sulfide per mol of protein (Table 3). As no special care had been taken to exclude oxygen from these enzyme preparations, the above values are probably underestimates of the in vivo stoichiometry of Fe and S. No flavin cofactors were detected by chemical assay of the SP lyase overexpressed from E. coli (data not shown).

UV-visible absorption spectrum of SP lyase. UV-visible absorption spectroscopy was performed on the 10×His-tagged SP lyase purified from E. coli in order to gain information regarding unusual structural features or the potential presence of UV-absorbing cofactors. The reddish-brown form of SP lyase obtained directly from the nickel-NTA agarose column exhibited a spectrum characteristic of an Fe-S protein (11a, 17, 45a), with peaks centered at 340, 400, and 472 nm (Fig. 4). Absorption peaks characteristic of potential chromophores, such as flavins, expected to be visualized at 377 and 452 nm (28) were not detected (Fig. 4). Upon addition of 1 mM (final concentration) dithionite to the purified SP lyase, the reddish-brown color immediately disappeared and the UV-visible absorption spectrum of the dithionite-treated SP lyase (30 min on ice) exhibited a dramatic shift (Fig. 4; see Discussion).

View larger version (18K): [in this window] [in a new window]   FIG. 4.   UV-visible absorption spectrum of 10×His-tagged SP lyase (0.8 mg/ml) purified from the E. coli overexpression system before (thick line) and after (thin line) treatment with 1 mM dithionite (0°C, 30 min).

In vitro assay of SP lyase activity. Numerous previous attempts to detect SP lyase activity in vitro have been unsuccessful; however, the homology observed between SP lyase and the anaerobic RNR and PFL enzymes (24), coupled with the spectroscopic data indicating that the Fe-S cluster in SP lyase was partially oxidized upon purification (Fig. 4), presented the possibility that SP lyase would require at least reducing, if not anoxic, conditions for its activity. To assay SP lyase in vitro, we therefore modified conditions which were previously used to achieve in vitro activity of the anaerobic RNR of E. coli (29, 30). SP lyase activity was detected in vitro with cell-free spore extracts from strain WN417, encoding the N-terminal 6×His-tagged SP lyase (Fig. 5), but not from the parental strain WN356 treated in an identical manner (data not shown). When cell-free extracts prepared from strain WN417 were incubated under reducing conditions with SP-containing DNA, a paradoxical increase in SP was observed (Fig. 5), apparently due to enhanced TCA precipitation of small SP-containing DNA fragments bound by SP lyase (data not shown). Addition of either AdoMet or NADPH to the reaction mixture enabled the WN417 cell-free extract to correct approximately 20 and 25% of the SP present in the substrate DNA, respectively (Fig. 5). The two cofactors added in combination also stimulated SP repair activity in vitro, but not as effectively as either added alone (Fig. 5).

View larger version (21K): [in this window] [in a new window]   FIG. 5.   In vitro assay of SP lyase activity from cell-free extracts of strain WN417 spores. Assays were performed as described in Materials and Methods, and results are reported as averages ± standard errors of the means (n = 4). Differences in SP level in reaction mixtures to which cofactors were added compared to the control reaction mixture with extract added alone were significant at P = 0.05 (*), P = 0.01 (**), or P = 0.001 (***) as determined by analysis of variance.

View larger version (22K): [in this window] [in a new window]   FIG. 6.   In vitro repair of SP by 6×His-tagged SP lyase purified from B. subtilis spores as a function of enzyme added in the absence (open circles) or presence (closed circles) of 2 mM AdoMet. Data are reported as averages ± standard errors of the means (n = 2).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Repair of UV-induced DNA damage by SP lyase during spore germination is an important determinant of bacterial spore UV resistance (reviewed in references 14, 25, 42, and 49). We have shown in this report that SP lyase from B. subtilis is an Fe-S protein which requires reducing conditions and AdoMet for its activity, in a fashion similar to enzymes such as class III anaerobic RNR and to PFL.

The 6×His-tagged SP lyase was expressed in B. subtilis from the amyE locus by using its native transcriptional and translational signals; therefore, its level in the dormant spore is probably similar to that of native SP lyase. Calculation of the yield of the purified 6×His-tagged SP lyase from spores (approximately 1 µg of protein per 1.25 × 10¹¹ spores) corresponds to approximately 100 to 200 subunit copies per spore. Because SP lyase is present at such a low concentration in spores, some of the physicochemical data reported here were derived from the splB gene product overexpressed in E. coli. Chemical extraction of flavin and spectral analysis (28) of the 10×His-tagged SP lyase purified from E. coli failed to detect flavin-containing compounds, consistent with the absence of absorption peaks characteristic of flavins in its UV-visible spectrum (Fig. 4). Chemical analysis of iron and acid-labile sulfide content (28) of the 10×His-tagged SP lyase purified from the E. coli overexpression system indicated 1.03 mol of iron and 1.60 mol of sulfide per mol of SP lyase subunit (Table 3), strongly suggesting that SP lyase contains an Fe-S center. The observed iron and sulfur stoichiometry of the purified 10×His-tagged SP lyase overexpressed in E. coli is probably an underestimate of the in vivo stoichiometry, as the protein was isolated under nonreducing conditions. The UV-visible spectrum of the 10×His-tagged SP lyase from E. coli (Fig. 4) is also characteristic of a protein containing an Fe-S cluster; the presence of absorption peaks in the SP lyase spectrum at 340, 400, and 472 nm are reminiscent of 2Fe-2S proteins such as biotin synthase (11a) and ferredoxin (45a). The dramatic shift in UV-visible spectrum of SP lyase upon reduction with dithionite is similar to that seen in 2Fe-2S to 4Fe-4S cluster conversion upon dithionite reduction of biotin synthase (11a). While definitive determination of which type of Fe-S cluster is contained by SP lyase awaits further characterization, the spectroscopic and chemical analysis data are consistent with the previous observation that the amino acid sequence of SP lyase contains a constellation of cysteine residues arranged in a manner highly homologous to the regions of the anaerobic RNR and PFL proteins which participate in 4Fe-4S cluster formation (24).

By using a modification of conditions used for assaying RNR activity (29; see Materials and Methods), SP lyase activity was detected in a cell-free extract prepared from dormant spores of strain WN417 (Fig. 5). This represents the first in vitro demonstration of SP lyase enzymatic activity, a significant first step toward characterizing the enzymatic mechanism of SP lyase in detail. Deoxygenated cell-free extract from strain WN417 dormant spores could not by itself repair SP in vitro, but activity was obtained by addition of either AdoMet or NADPH (Fig. 5). That AdoMet stimulated SP lyase activity seems reasonable, based upon the analogy with RNR and PFL, both of which require AdoMet for activity (29). Stimulation of SP lyase activity by NADPH addition to the cell-free extract from strain WN417 dormant spores may indicate that the reduced form of NADPH could be involved in a redox system used in vivo for regeneration of active-site cysteines, as seen in other RNRs (33). Interestingly, although dormant spores contain amounts of pyridine nucleotides comparable to vegetative cells, almost none is in the reduced form (43).

Chemical reduction of the 6×His-tagged SP lyase purified from B. subtilis spores with formate resulted in activation of the enzyme such that it required only AdoMet for activity (Fig. 6). However, we were unable to detect activity from the 10×His-tagged SP lyase purified from the E. coli overexpression system pretreated identically. This result suggests either that the loss or inactivation of some essential factor(s) occurred during purification of the E. coli enzyme or that the cloned splB locus by itself expressed in E. coli does not encode the entire SP lyase holoenzyme. By analogy to RNR or PFL, what we call SP lyase encoded by splB would correspond only to the ₂ activating subunit of an ₂₂ tetrameric holoenzyme. In anaerobic RNR and in PFL, the activating subunits are homodimers which associate through a 4Fe-4S cluster (30); separate genetic evidence suggests that SP lyase may also be active as a homodimer (24). The possibility therefore presents itself that while the B. subtilis WN417 spore contains the putative SP lyase holoenzyme, the E. coli system expresses only the splB-encoded activating subunit. Experiments are currently under way to identify the putative second subunit.

Fe-S clusters have been shown to be ubiquitous and important determinants of protein structure and enzymatic activity (reviewed in reference 3) and have been shown to operate by such diverse mechanisms as stabilization of the DNA binding site, as in the endonuclease III of E. coli (46); behaving as a sensor of oxidative stress by changes in oxidation state, as in the SoxR protein of E. coli (15, 16); or regulation of DNA binding through oxidative assembly-disassembly of an Fe-S cluster holding together two identical subunits, as in the E. coli anaerobic transcriptional activator FNR (3, 17). Because SP lyase was found by an amino acid sequence homology search to resemble the activating subunits of anaerobic RNR and PFL (24), these enzymes are being used as models for probing in vitro SP lyase activity. The enzymology of RNR and PFL is rather complex. Both enzymes are active as ₂₂ tetramers (29). Enzymatic activity is dependent upon generation of an oxygen-labile glycyl free radical in the larger ₂ catalytic subunit. The smaller ₂ activating subunit dimerizes through an oxygen-labile 4Fe-4S cluster which in its reduced form participates in splitting AdoMet into methionine and a 5'-deoxyadenosyl radical which then generates the glycyl radical in the ₂ catalytic subunit by abstracting a hydrogen atom from the -carbon of glycine. AdoMet is thus required for activity (20). In AdoMet-requiring enzymes which cleave C---H or C==C bonds, such as PFL, anaerobic RNR, lysine 2,3-aminomutase, biotin synthase, and lipoic acid synthase, a common theme appears to be oxidative degradation of subunit-bridging 4Fe-4S clusters during purification to 2Fe-2S clusters (11a). The 4Fe-4S cluster in the ₂ subunit of RNR can be reduced in vitro by treatment with either (i) flavodoxin, flavodoxin reductase, and NADPH; (ii) photochemically reduced 5-deazaflavin; or (iii) dithionite (29). By analogy to these enzymes, the Fe-S cluster in SP lyase also appears to share these properties (Fig. 4). The results reported above with SP lyase are important steps towards determination of the physical organization and reaction mechanism of this unique DNA repair enzyme.