The Role of Lipoprotein Processing by Signal Peptidase II in the Gram-positive Eubacterium Bacillus subtilis

Computer-assisted analyses indicate thatBacillus subtilis contains approximately 300 genes for exported proteins with an amino-terminal signal peptide. About 114 of these are lipoproteins, which are retained in the cytoplasmic membrane. We have investigated the importance of lipoprotein processing by signal peptidase II (SPase II) for cellular homeostasis, using cells lacking SPase II. The results show that lipoprotein processing is important for cell viability at low and high temperatures, suggesting that lipoproteins are essential for growth under these conditions. Although certain lipoproteins are required for the development of genetic competence, sporulation, and germination, these developmental processes were not affected in the absence of SPase II. Cells lacking SPase II accumulated lipid-modified precursor and mature-like forms of PrsA, a folding catalyst for secreted proteins. These forms of PrsA seem to have a reduced activity, as the secretion of α-amylase was strongly impaired. Unexpectedly, type I signal peptidases, which process secretory preproteins, were not involved in alternative amino-terminal processing of pre-PrsA in the absence of SPase II. In conclusion, processing of lipoproteins by SPase II in B. subtilis is not strictly required for lipoprotein function, which is surprising as lipoproteins and type II SPases seem to be conserved in all eubacteria.

One of the most commonly used eubacterial sorting (retention) signals for proteins that are exported from the cytoplasm is an amino-terminal lipid-modified cysteine residue (see Refs. 1 and 2). In Gram-positive eubacteria, such as Bacillus subtilis, lipid-modified proteins (lipoproteins) are retained in the cytoplasmic membrane. In Gram-negative eubacteria, such as Escherichia coli, these proteins are retained in the cytoplasmic or the outer membrane; retention in the cytoplasmic membrane depends on the presence of an additional sorting signal in the form of an aspartic acid residue at the ϩ2 position relative to the amino-terminal cysteine residue (see Refs. [3][4][5][6]. Even the organism with the smallest known genome, Mycoplasma genitalium, seems to make use of lipid modification to retain proteins in the cytoplasmic membrane (7). The number of putative lipoprotein-encoding genes per eubacterial genome seems to range from approximately 18 in M. genitalium (http://www.tigr.org/tdb/mdb/mgdb) to approximately 89 in E. coli 1 and 114 in B. subtilis (Table I). Thus, lipoproteins appear to represent about 1-3.5% of the proteome of eubacteria.
Lipoproteins are directed into the general (Sec) pathway for protein secretion by their signal peptides, which show similar structural characteristics as the signal peptides of secretory proteins: a positively charged amino terminus, a hydrophobic core region, and a carboxyl-terminal region containing the cleavage site for signal peptidase (SPase). 2 The major difference between signal peptides of lipoproteins and secretory proteins is the presence of a well conserved "lipobox" of four residues in the former signal peptides, which constitutes the cleavage site for the lipoprotein-specific SPase, also known as SPase II. Invariably, the carboxyl-terminal residue of the lipobox is cysteine, which, upon lipid modification, forms the retention signal of the mature lipoprotein (for details, see Refs. 1 and 8). Modification of this cysteine residue by the diacylglyceryl transferase is a prerequisite for processing of the lipoprotein precursor by SPase II. Processing by SPase II can be inhibited with globomycin, a reversible and noncompetitive peptide inhibitor (2,9,10). In E. coli, the processed lipoprotein is further modified by aminoacylation of the diacylglycerylcysteine amino group (11,12). It is presently not known whether the latter lipid modification step is conserved in all eubacteria. For example, B. subtilis and M. genitalium lack an lnt gene for the lipoprotein aminoacyltransferase. 1 In Gram-negative eubacteria, the outer membrane confines numerous proteins to the periplasm. In Gram-positive eubacteria, which lack an outer membrane, lipid modification of exported proteins prevents their loss into the environment, as these proteins remain anchored to the cytoplasmic membrane. This may explain why B. subtilis contains more putative lipoproteins than E. coli and why, for example, 32 lipoproteins of B. subtilis are homologues of periplasmic high affinity sub-* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ ‡ Supported by a Hungarian State Eötvös Fellowship from the National Scholarships Board (Hungary).
¶ ¶ To whom correspondence should be addressed. Tel.: 31-50-3633079; Fax: 31-50-3632348; E-mail: j.m.van.dijl@farm.rug.nl. strate-binding proteins from Gram-negative eubacteria ( Table  I). Lipoproteins of Gram-positive eubacteria with a known function are involved in a variety of processes, such as the uptake of nutrients, resistance to antibiotics, protein secretion, competence for DNA binding and uptake, sporulation, germination, and bacterial targeting to different substrates, bacteria, and host tissues (see Ref. 13). In addition, a close examination of the putative lipoproteins of B. subtilis, which were identified on the basis of the conserved lipobox, suggests that certain lipoproteins are also involved in oxidative phosphorylation, cell wall biogenesis, and autolysis. The importance of lipoproteins for the homeostasis of Gram-positive eubacteria is underscored by the observation that the most abundant lipoprotein of B. subtilis, PrsA, is essential for the efficient secretion of various proteins and cell viability (14 -17). Notably, no specific function can presently be assigned to the majority of putative lipoproteins of B. subtilis (about 75%; Table I) and other Gram-positive eubacteria.
The present studies were aimed at the evaluation of the importance of B. subtilis lipoproteins for cellular homeostasis in general, and their processing by SPase II in particular. For this purpose, SPase II-depleted cells were used. Unexpectedly, the results show that lipoprotein processing is required for growth at low and high temperatures and the efficient secretion of the ␣-amylase AmyQ (a non-lipoprotein) but not for growth and cell viability at 37°C, development of competence for DNA binding and uptake, sporulation, or spore germination. Table II lists the plasmids and bacterial strains used. TY medium (tryptone/yeast extract) contained Bacto tryptone (1%), Bacto yeast extract (0.5%), and NaCl (1%). S7 media 1 and 3, used for labeling of B. subtilis proteins with [ 35 S]methionine (Amersham Pharmacia Biotech), were prepared as described in Refs. 21 and 22. When required, media for E. coli were supplemented with ampicillin (50 g/ml), erythromycin (100 g/ml), or kanamycin (40 g/ml); media for B. subtilis were supplemented with chloramphenicol (5 g/ml), erythromycin (1 g/ml), kanamycin (10 g/ml), tetracyclin (6 g/ml), spectinomycin (100 g/ml), globomycin (80 g/ml), and/or IPTG (1 mM).

Plasmids, Bacterial Strains, and Media-
DNA Techniques-Procedures for DNA purification, restriction, ligation, agarose gel electrophoresis, and transformation of E. coli were carried out as described in Ref. 30. Enzymes were from Boehringer Mannheim. B. subtilis was transformed as described in Ref. 25. Correct integration of plasmids into the chromosome of B. subtilis was verified by Southern blotting. PCR was carried out with Vent DNA polymerase (New England Biolabs) as described in Ref. 31. pMutin2-MIL (see "Results" for explanation) was constructed by PCR amplification of the 5Ј region of the lsp gene with the primers lsp-m1(5Ј-ATAAGCTTAAC-CGTAAACTGGAGG-3Ј) and lsp-m2 (5Ј-GCGGATCCAAGAAGCCTTT-GTCCC-3Ј) and subsequent cloning in pMutin2. pMutin2-M⌬L was constructed by PCR amplification of an internal fragment of the lsp gene with the primers lsp-m5 (5Ј-ATGTCGACGCATGGGGGATAT-TAG-3Ј) and lsp-m2 and subsequent cloning in pMutin2. To construct pKTH3409, the prsA gene was amplified by PCR with a primer containing a ClaI site and 5Ј sequences of prsA (starting at position Ϫ27 relative to the start codon) and a primer, which adds the sequence CACCATCACCATCACCATTAAGTCGAC to the 3Ј-end of prsA, specifying a hexahistidine tag, stop codon, and SalI cleavage site. The tagged prsA gene was placed under the control of the xylose-inducible xylA promoter of plasmid pSX50, using the ClaI and SalI restriction sites. The resulting plasmid was designated pKTH3409.
Competence and Sporulation-Competence for DNA binding and uptake was determined by transformation with plasmid or chromosomal DNA (28). The efficiency of sporulation was determined by overnight growth in Schaeffer's medium (32), killing of cells with 0.1 volume of chloroform, and subsequent plating.
␤-Galactosidase Activity Assay-Overnight cultures were diluted 100-fold in fresh medium, and samples were taken at hourly intervals for A 600 readings and ␤-galactosidase activity determinations. The assay and the calculation of ␤-galactosidase units (expressed as units per A 600 ) were carried out as described in Ref. 33.
Trypsin Accessibility Assay-The preparation of protoplasts from exponentially growing cells of B. subtilis and the testing of the protease accessibility of membrane proteins was performed as described in Ref. 29.

RESULTS
The lsp Gene for SPase II Is Not Essential for Cell Viability-To determine whether the lsp gene for the SPase II of B. subtilis (35) is essential for growth and cell viability, two lsp mutant strains were constructed, using derivatives of the integration vector pMutin2. B. subtilis M⌬L was constructed by integration of pMutin2-M⌬L within the structural lsp gene, resulting in the disruption of this gene. B. subtilis MIL was constructed by integration of pMutin2-MIL at the 5Ј-end of lsp in such a way that the original lsp promoter region was replaced by the IPTG-inducible Pspac promoter (Fig. 1A). The fact that B. subtilis M⌬L could be obtained shows that SPase II is not essential for cell viability, at least when cells are grown in TY or minimal medium at 37°C. Under these conditions, the growth of B. subtilis M⌬L was only slightly reduced, compared with the parental strain 8G5. Similarly, the growth rate of B. subtilis MIL was slightly reduced in the absence of IPTG, as compared with that of B. subtilis MIL in the presence of IPTG, or the parental strain. Unexpectedly, the disruption of the lsp gene did not inhibit the development of competence for DNA binding and uptake, sporulation, and subsequent spore germination (data not shown), although at least one lipoprotein is required both for competence development and sporulation (OppA) (36), three for sporulation (DppE, also known as DciA, SpoIIIJ, and SpoIVB) (37)(38)(39), and five for germination (GerAC, GerBC, GerKC, GerM, and GerD) (40 -44). Interestingly, SPase II appeared to be essential for growth at 15°C (data not shown) and 48°C. Upon a temperature shift from 37 to 48°C, cells lacking SPase II stopped growing and lysed (Fig. 1B, ⌬lsp). These findings imply that some as yet unidentified lipoproteins of B. subtilis are required for growth at low and high temperatures or that the accumulation of lipoprotein precursors causes cold and heat sensitivity. In what follows, we show that the cold and heat sensitivity of cells lacking SPase II must be due to the malfunction of certain lipoproteins.
Maximal lsp Transcription in the Exponential Growth Phase-Both in B. subtilis M⌬L and B. subtilis MIL, the transcription of the lacZ gene, present on pMutin2, is directed by the lsp promoter region (Fig. 1A). To study the transcription of the lsp gene, B. subtilis MIL was grown in the presence or absence of IPTG, and samples withdrawn at hourly intervals were assayed for ␤-galactosidase activity. The results showed that the ␤-galactosidase levels increased during exponential growth, reaching a maximum in the transition phase between exponential and postexponential growth. In contrast, the ␤-galactosidase levels were strongly decreased in the postexponen-tial growth phase. This pattern of lsp-lacZ expression was observed when cells were grown in TY or minimal medium, irrespective of the presence of IPTG. However, compared with minimal medium, higher expression levels were detected in TY medium, where, in the postexponential growth phase, ␤-galac-tosidase activity was close to background (Fig. 1C). In summary, these observations indicate that the transcription of lsp is highest in the exponential growth phase and decreases to lower levels in the postexponential growth phase. Thus, it seems that exponentially growing cells produce sufficient Putative lipoprotein signal peptides were identified in two ways. First, the presence of a lipobox was determined by a search for cysteinc residues in putative signal peptides of B. subtilis, which were identified with the SignalP algorithm for the prediction of signal peptides from Gram-positive eubacteria (18). To this purpose, the first 60 residues of the annotated proteins of B. subtilis in the SubtiList database (http://www.pasteur.fr/ Bio/SubtiList.html) were used. Second, putative lipoprotein signal peptides were identified by performing similarity searches in the SubtiList database with signal peptides of known lipoproteins, using the Blast algorithm (19). The highly conserved leucine residue at position Ϫ3 and the strictly conserved cysteine at position ϩ1 relative to the cleavage site for SPase II, are indicated in boldface. Note that aspartic acid residues are absent from position ϩ2. KapB, which seems to be a lipoprotein (20), is not listed in this table because its amino terminus is atypical for signal peptides of lipoproteins due to the presence of two lysine residues in the hydrophobic core region. No other putative lipoproteins of this type were identified.

Signal peptide
Signal peptide

MRFRWVWLFVIMLLLAECQ
a Signals of lipoproteins that are homologues of known periplasmic high affinity substrate binding proteins from Gram-negative eubacteria.
SPase II for lipoprotein processing in the transition and postexponential growth phases.
Alternative Amino-terminal Processing of Pre-PrsA in Cells Lacking SPase II-To examine the effects of the absence of SPase II, the processing of (pre-)PrsA, the major lipoprotein of B. subtilis (15,16) was studied by pulse-chase labeling experiments with B. subtilis M⌬L. As shown in Fig. 2A (⌬lsp), pre-PrsA processing was strongly impaired in the absence of SPase II, and even after a long chase period of 15 min, no mature PrsA was detectable. In contrast, pre-PrsA was rapidly processed to the mature form in the parental strain. As expected, Western blotting experiments showed that B. subtilis M⌬L accumulated pre-PrsA, but surprisingly, this concerned only about 50% of the total PrsA present in the cells. In addition to pre-PrsA, cells of B. subtilis M⌬L also contained mature-like forms of PrsA, which, compared with mature PrsA, had a slightly lower mobility on SDS-PAGE (Fig.  2B, ⌬lsp). This difference in mobility could only be visualized clearly when proteins were separated on long gels (40 cm). In what follows, standard gel systems (15 cm) were used, resulting in a less pronounced separation of mature PrsA (strains containing SPase II) and mature-like forms of PrsA (strains lacking SPase II). As shown in Fig. 2C, the mature-like forms of PrsA were also detected when cells of B. subtilis M⌬L (⌬lsp) or the parental strain (8G5) were grown in the presence of globomycin. Western blotting experiments with a xylose-inducible mutant form of PrsA, containing a carboxylterminal hexahistidine tag, showed that at least one of the mature-like forms of PrsA was cleaved at the amino terminus (Fig. 2D). Taken together, these findings show that in the absence of SPase II, pre-PrsA is subject to alternative aminoterminal processing at a low rate.
It has been shown previously that lipoprotein precursors from which the cleavage site for SPase II was removed by site-directed mutagenesis can be cleaved at alternative sites  in cells lacking SPase II must be (partially) active. This implies that at least one of the latter forms of PrsA is correctly localized at the external surface of the membrane. To determine the topology of pre-PrsA and the mature-like forms of PrsA in the absence of SPase II, protoplasts of B.
subtilis 8G5 M⌬L were incubated with trypsin. Like the mature PrsA of B. subtilis 8G5, pre-PrsA and the mature-like forms of PrsA of B. subtilis 8G5 M⌬L were associated with protoplasts and accessible to trypsin (Fig. 3A, ⌬lsp). In contrast, the cytosolic protein GroEL was only accessible to trypsin when the protoplasts were lysed with Triton X-100 (Fig. 3B). Because the latter findings show that pre-PrsA and the mature-like forms of PrsA are correctly localized in cells lacking SPase II, we also addressed the question whether these forms are lipid-modified. To this purpose, palmitic acid labeling experiments were performed with strains lacking SPase II or, as a control, the diacylglyceryl transferase, specified by the lgt gene (i.e. prs-11) (14). 4 Cells of the latter strain (⌬lgt) accumulated non-lipomodified pre-PrsA, whereas the strain lacking SPase II (⌬lsp) accumulated lipomodified pre-PrsA and mature-like PrsA (Fig. 4, A and B). In conclusion, these data show that, in the absence of SPase II, the precursor and mature-like forms of PrsA are lipid-modified, displaying a similar membrane topology as the mature PrsA in SPase II-proficient cells. Thus, all of these forms might, in principle, be active and responsible for the viability of cells lacking SPase II.
Lsp Is Required for the Efficient Processing and Secretion of the Non-lipoprotein Pre-AmyQ-It was previously shown that PrsA sets a limit for high level secretion of the Bacillus amyloliquefaciens ␣-amylase AmyQ and that it is required for the folding of AmyQ into a protease-resistant conformation (16). To examine the effect of SPase II depletion on PrsA activity, AmyQ secretion was monitored in B. subtilis 8G5 M⌬L (Fig. 5, ⌬lsp) and 8G5 MIL (Fig. 5, Ilsp). To this purpose, both strains were transformed with pKTH10, which contains the amyQ gene (24). Next, the secretion of AmyQ was analyzed by Western blotting. As shown in Fig. 5, the accumulation of pre-PrsA and mature-like forms of PrsA in cells depleted of SPase II (Fig. 5A, ⌬lsp and Ilsp in the absence of IPTG) was paralleled by the secretion of about 5-fold reduced amounts of mature AmyQ into the growth medium (Fig. 5C). The latter observation is diagnostic for reduced levels of PrsA activity. In addition, SPase II-depleted cells accumulated increased levels of pre-AmyQ (Fig. 5B), which is atypical for PrsA mutants (14). As shown by pulse-chase labeling experiments, the rate of pre-AmyQ processing by type I SPase(s) was slightly (but reproducibly) reduced in cells lacking SPase II (Fig. 5D). As mutations in PrsA do not cause the accumulation of pre-AmyQ, this effect of the absence of SPase II must be attributed to the accumulation of lipoprotein precursors or the malfunction of an as yet unidentified lipoprotein. The reason why the kinetic effects of the absence of SPase II on pre-AmyQ processing (Fig.  5D) are mild in comparison to the strong accumulation of pre-AmyQ at steady state (Fig. 5B) is not clear. One explanation could be that the growth conditions in both types of experiments are not identical.
To determine whether the accumulation of pre-AmyQ in SPase II-depleted cells reflects a reduced rate of translocation of pre-AmyQ, which might be caused by the accumulation of lipoprotein precursors, we made use of an AmyQ variant (AmyQ-PSBT) (25) containing the biotinylation domain of the pyruvate decarboxylase of Propionibacterium shermanii (48). The rationale of this experiment is that pre-AmyQ-PSBT can only be biotinylated by the cytoplasmic biotin-ligase if the PSBT domain folds into its native three-dimensional structure in the cytoplasm. This will only happen if the rate of translocation of pre-AmyQ-PSBT across the membrane is significantly reduced. As shown in Fig. 6, cells lacking SPase II(⌬lsp) did not accumulate biotinylated pre-AmyQ-PSBT, irrespective of the growth temperature (15, 37, or 48°C), although AmyQ-PSBT was produced under these conditions (data not shown). In contrast, B. subtilis cells with a disrupted secDF gene (B. subtlis MIF) (29) accumulated biotinylated forms of AmyQ-PSBT at all growth temperatures tested (Fig. 6, ⌬secDF), whereas cells lacking SipS and SipT but expressing the temperature-sensitive SipS-D146A protein (Fig. 6, ⌬STxS-D146A) (25) accumulated biotinylated forms of AmyQ-PSBT only at 48°C. These observations indicate that the accumulation of pre-AmyQ in SPase II-depleted cells is not due to impaired translocation across the membrane and that the cold sensitivity of these cells is not related to protein translocation defects. Instead, the accumulation of pre-AmyQ must be attributed to the malfunction of certain lipoproteins in the absence of SPase II. DISCUSSION We have previously shown that five paralogous type I SPases are involved in the processing of secretory precursor proteins in B. subtilis (25,47,49,50). Two of these, designated SipS and SipT, are of major importance for protein secretion, and cells depleted of both SipS and SipT stop growing and lyse. The other type I SPases (SipU, SipV, and SipW) are of minor importance for protein secretion and viability. Thus, B. subtilis is representative for Gram-positive eubacteria, archaea, and eukaryotes, many of which contain paralogous type I SPases (25). In contrast to the type I SPases, B. subtilis (35,51) and other eubacteria seem to contain only one gene for SPase II, 5 whereas type II SPases appear to be absent from archaea and eukaryotes. 1 Here, we document four unexpected observations with respect to SPase II function in B. subtilis. First, unlike the SPase II of E. coli, the SPase II of B. subtilis is not essential for 5 The product of the yaaT gene, which has been annotated as a putative SPase II-encoding gene of B. subtilis (51), does not show sequence similarity to known type II SPases, and, moreover, is predicted to be a soluble cytoplasmic protein (see Footnote 1). This makes a role of the YaaT protein in lipoprotein processing highly unlikely. growth and viability. Second, the absence of SPase II resulted in cold and heat sensitivity. Third, SPase II is not required for the development of genetic competence, sporulation, and spore germination although at least eight known lipoproteins are important for these processes. Fourth, the secretion of the non-lipoprotein AmyQ was severely reduced in the absence of SPase II.
Lipoprotein processing by SPase II in E. coli can be inhibited by globomycin. Consistent with the observation that SPase II is essential for cell viability of E. coli (52), globomycin can serve as an antibiotic for E. coli and other Gram-negative eubacteria (9). Despite the fact that globomycin can inhibit the processing of certain lipoprotein precursors in B. subtilis (Ref. 53 and this paper), this antibiotic displays no cytotoxic activity against B. subtilis and other Gram-positive eubacteria (9). The present results show that this is most likely due to the fact that the SPase II is not essential for cell viability of B. subtilis and not to degradation or modification of globomycin. One reason why SPase II is more important for cell viability of E. coli than for cell viability of B. subtilis may be that most lipoproteins of E. coli are sorted to the outer membrane. Of the 89 (putative) lipoproteins of E. coli, which we have recently identified by computer-assisted analyses, only 8 have an aspartic acid residue at the ϩ2 position for retention in the cytoplasmic membrane (data not shown), suggesting that the other 81 lipoproteins are sorted to the outer membrane. Thus, the lethality of lsp mutations in E. coli may be attributed to the depletion of lipoproteins from the outer membrane. Furthermore, even though B. subtilis has more lipoprotein-encoding genes than E. coli, the latter organism may produce more lipoprotein molecules per cell, and therefore, the lethality of lsp mutations in E. coli could also be due to the accumulation of lipoprotein precursors in the cytoplasmic membrane.
Notably, the lipoproteins of B. subtilis (Table I) and other Gram-positive eubacteria (13) seem to lack aspartic acid residues at the ϩ2 position. The latter observation suggests that this aspartic acid residue has specifically evolved as a retention signal for lipoproteins of Gram-negative eubacteria. Furthermore, the lack of lnt genes from B. subtilis and M. genitalium suggests that the lipoproteins of these organisms are not aminoacylated. Consistent with the latter hypothesis, a macrophage-stimulating lipoprotein with a non-acylated amino terminus has been isolated from Mycoplasma fermentans (54). The latter observations raise the intriguing question of whether aminoacylation has a particular function in Gram-negative eubacteria; for example, in the sorting of lipoproteins.
Cold sensitivity seems to be a general property of E. coli (55) and B. subtilis (29) strains, which are defective in protein translocation via the Sec-machinery. Thus, the cold sensitivity of B. subtilis lsp mutants might reflect a general defect in protein translocation, which could be caused primarily by the accumulation of lipoprotein precursors. However, this possibility is unlikely, as SPase II-depleted cells did not show a translocation defect for AmyQ-PSBT, irrespective of the growth temperature. Instead, our results indicate that the observed cold and heat sensitivity of B. subtilis mutants lacking SPase II is caused by the malfunction of certain lipoproteins, which are required for cell viability at low and high temperatures.
The observation that the development of competence, sporulation, and germination are not affected by the absence of SPase II shows that the precursors, or (putative) alternatively processed forms of the lipoproteins required for these primitive developmental processes are active. Similarly, the fact that the strain lacking SPase II is viable at 37°C shows that pre-PrsA and/or the mature-like forms of PrsA are active, because PrsA is essential for cell viability (16). Nevertheless, as indicated by the reduced secretion of AmyQ in lsp mutants B. subtilis, lipoprotein processing is important for the full functionality of lipoproteins, such as PrsA. Our current working model for the effects of SPase II limitation on the processing of pre-PrsA and the secretion of AmyQ is shown in Fig. 7. Although SPase II-depleted cells contain similar amounts of PrsA protein as wild-type cells, processing by SPase II seems to be important for the stable maintenance of certain other lipoproteins, 6 suggesting that processing by SPase II is required to protect these proteins against proteolytic degradation at the membrane-cell wall interface.
The fact that pre-PrsA is subject to alternative processing in the absence of SPase II is reminiscent of the previously reported observation that mutants of the penicillinase PenP of Bacillus licheniformis, which lack the cysteine residue at the ϩ1 position, are subject to alternative processing. As these mutant proteins are not lipid-modified and contain putative SPase I cleavage sites, type I SPases have been invoked in their processing (57,58). In contrast, our results indicate that type I SPases are not involved in the alternative amino-terminal processing of pre-PrsA, which is consistent with the fact that this precursor is lipid-modified. Surprisingly, the nonmodified pre-PrsA produced by the lgt mutant is not processed, even though putative SPase I cleavage sites are present in this precursor, 1 suggesting that it contains an as yet unidentified "SPase I avoidance signal." Important challenges for future research are the identification of the protease(s) involved in the alternative processing of pre-PrsA and the SPase I avoidance signal in this precursor. 6 J. Bengtsson and L. Hederstedt, personal communication.  (25). Cells were grown overnight in TY medium, 100-fold diluted in fresh TY medium, and grown until the transition phase between exponential and postexponential growth at 37°C. Next, aliquots were incubated for 3 h at 15, 37, or 48°C. Cells were collected by centrifugation, and biotinylated (pre-)AmyQ-PSBT was visualized by SDS-PAGE and Western blotting using a streptavidin-horseradish peroxidase. p, pre-AmyQ-PSBT; m, mature AmyQ-PSBT.
Finally, the reason why SPase II-depleted cells accumulate pre-AmyQ is presently not completely clear. First, as shown with AmyQ-PSBT, the rate of AmyQ translocation in these cells is not detectably affected. Second, as PrsA mutants do not accumulate pre-AmyQ (14), this effect can not be attributed to PrsA malfunction. Consequently, the accumulation of pre-AmyQ must be due to the malfunction of at least one as yet unknown lipoprotein, which affects the stability or processing of pre-AmyQ. This hypothesis is supported by our observation that the half-life of pre-AmyQ and at least one other precursor, pre(A13i)-␤-lactamase (50) (data not shown) is slightly increased in the absence of SPase II. We are presently investigating which of the putative lipoproteins of B. subtilis could be responsible for this effect.