The M Protein Is Dispensable for Maturation of Streptococcal Cysteine Protease SpeB

Abstract

The streptococcal pyrogenic exotoxin B (SpeB) is an important virulence factor of group A streptococci (GAS) with cysteine protease activity. Maturation of SpeB to a proteolytically active form was suggested to be dependent on cell-wall-anchored M1 protein, the major surface protein of GAS (M. Collin and A. Olsén, Mol. Microbiol. 36:1306-1318, 2000). Collin and Olsén showed that mutant GAS strains expressing truncated M protein secrete a conformationally different form of unprocessed SpeB with no proteolytic activity. Alternatively, we hypothesized that a truncated M protein may interfere with processing of this secreted protease, and therefore we tested cysteine protease activity in genetically defined mutant strains that express either no M protein or membrane-anchored M protein with an in-frame deletion of the AB repeat region. Measurements of SpeB activity by cleavage of a substrate n-benzoyl-Pro-Phe-Arg-p-nitroanilide hydrochloride showed that the proteolytic activities in culture supernatants of both mutants were similar to those from the wild-type strain. In addition, Western blot analysis of culture supernatants showed that SpeB expression and processing to a mature form was unaffected by either deletion mutation. Therefore, we conclude that M protein is not required for maturation of the streptococcal cysteine protease SpeB.

The gram-positive Streptococcus pyogenes (group A streptococci [GAS]) is widely considered an important human pathogen which causes a broad variety of diseases. It is commonly associated with mild infections, like tonsillopharyngitis and impetigo, and severe complications, such as necrotizing fasciitis and streptococcal toxic shock syndrome. The latter have increased since the mid 1980s (8, 32, 49). The spectrum of nonsuppurative sequelae includes rheumatic fever, an acute poststreptococcal glomerulonephritis, and, more recently, neuropsychiatric syndromes (39, 48).

GAS produces a variety of surface-bound and secreted virulence factors which are known to contribute to the severity of their infections. Among these, M protein which is anchored to the cell wall by an LPXTG motif (23) plays key roles in bacterial resistance to phagocytosis (28), adherence, invasion of human epithelial cells (13), and microcolony formation in tonsillar tissue (6). The M protein exhibits extensive sequence variations between strains, accounting for over 120 different serotypes of GAS (21). The hypervariable NH2-terminal region of M1 protein, containing unique A and B repeat domains, has been shown to be important for both adherence and entry into epithelial cells (13). GAS that lack emm genes are cleared faster in animal models (1).

The gene encoding the streptococcal cysteine protease (streptococcal pyrogenic exotoxin B), speB, is present in most isolates of GAS and is highly conserved among them (52). Expression of SpeB is controlled by multiple transcriptional regulators, including mga (multiple gene regulator in group A streptococci), ropB (regulation of proteinase), and pel (pleiotropic effect locus) (35, 38, 46). In addition, RopB regulation of SpeB expression is known to be growth phase dependent (40). The oligopeptide and dipeptide permeases, Opp and Dpp, respectively, were suggested to influence speB expression, whereas influence of the two-component system, CsrR-CsrS, is still under discussion (22, 25, 43, 44). Trigger factor, a molecular chaperone and peptidyl-prolyl cis-trans isomerase, is essential for secretion of the protease and its maturation (37, 38). SpeB, like many proteases, is secreted as an inactive precursor of the 40-kDa zymogen form, which is autocatalytically cleaved into an active enzyme of 28 kDa (9). At least six intermediates are produced during this process (18). Recently, analysis of the crystal structure of the zymogen showed that the catalytic domain of SpeB is structurally similar to the papain superfamily of cysteine proteases but differs in the propeptide region (30). In vitro studies have shown that SpeB is able to cleave immunoglobulin G (IgG) (11) and to degrade IgA, IgE, IgM, and IgD (10). In addition, SpeB is able to release bradykinin from its precursor and to activate interleukin-1β (26, 33). SpeB affects proteins displayed on the streptococcal surface as well. For example, SpeB cleaves M protein and the streptococcal C5a peptidase off the bacterial surface (3, 19, 47). However, the contribution of SpeB to human infections is still controversial (12, 15, 24, 27, 31, 50).

Collin and Olsén reported that M1 protein influences folding and maturation of SpeB (9). They and we constructed strains with insertion mutations in emm1 (9, 17). These mutations caused expression of a truncated M1 protein without a cell wall anchor that is in part secreted into the growth media. Both strains had significantly reduced proteolytic activity in culture supernatants and were unable to generate mature 28-kDa enzymes. They suggested that without M protein the protease assumes an aberrant conformation that blocks the active site and prevents maturation into an active protease. This potentially important and novel observation, however, left several unanswered questions with regard to the role of M protein in protease maturation.

Experiments described below sought to confirm dependence on M1 protein for SpeB maturation and to define M protein segments required for processing. Genetically defined emm1 mutants that express either no M protein or cell-wall-anchored M1 protein with a deletion of the AB repeat region were constructed and investigated for their capacity to generate mature and active cysteine protease. We also analyzed the effect of a deletion of emm49 as well as emm-like genes mrp and enn on protease activity. Collin and Olsén's results (9) were replicated using the same emm1 insertion mutant, but deletion of the emm1 gene had no effect on protease maturation and activity. Therefore, their interpretation was incorrect: the M protein is not required for maturation of SpeB.

MATERIALS AND METHODS

Bacteria.

GAS M1 strain 90-226 was obtained from the World Health Organization Center for Reference and Research on Streptococci at the University of Minnesota. It was originally isolated from the blood of a patient with sepsis (14). Strain CS101 is a spontaneous streptomycin-resistant derivative of an opacity-factor-positive M49 strain (29). All GAS strains and mutants are summarized in Table 1. In addition, Fig. 1 shows the emm1 gene locus and corresponding M proteins of wild-type and mutant strains. Streptococci were grown at 37°C under 5% CO₂ and 20% O₂ atmosphere. For most experiments, Todd-Hewitt broth (Difco Laboratories, Detroit, Mich.) supplemented with 1% yeast extract (THY) was used as the culture medium. For solid media, 1.5% granulated agar (Difco) was added to THY. After transformation, GAS were grown on blood agar plates (BAP) containing 5% sterile sheep blood. Escherichia coli was grown in Luria-Bertani broth (Difco). Antibiotics were used at the following concentrations: erythromycin, 5 μg/ml (streptococci) and 300 μg/ml (E. coli); spectinomycin, 60 μg/ml (streptococci); and kanamycin, 500 μg/ml (streptococci). Strain 90-226 emm1::Km (17), previously shown to produce inactive SpeB, was used in all experiments as a control to rule out technical differences.

TABLE 1.

Strains used in this study

Strain	Sero- type	Mutation	Phenotype	Reference or source
90-226	M1		Wild type	14
90-226 Δemml	M1	Deletion	M1−	This study
90-226 emml(ΔAB)	M1	Deletion	M1−	This study
90-226 emml::Km	M1	Insertion	M1−	17
90-226 speB::Erm	M1	Insertion	SpeB−	This study
CS101	M49		Wild type	29
MJY1-3	M49	Deletion	Mrp−, M49−, Enn−	29
CS101 speB	M49	Insertion	SpeB−	36

FIG. 1.

The top graphic shows the emm1 gene locus with corresponding M proteins of 90-226, 90-226 emm1::Km, and 90-226 emm1(ΔAB). The bottom graphic shows the mga49 regulon of CS101 and MJY1-3.

Construction of strain 90-226 Δemm1.

Plasmid pG⁺host5::mga-sic was constructed by ligation of the DNA fragment mga-sic into pG⁺host5 digested with NotI and XhoI. The DNA fragment mga-sic, containing sequences corresponding to the C-terminal half of the mga gene contiguous to the N-terminal half of the sic gene with a deletion of the entire emm1 sequence, was generated by overlap PCR amplification of chromosomal DNA from the strain 90-226. Briefly, the mga and sic fragments were amplified by PCR using oligonucleotide primer pairs MGAF (5′-CCC CGC GGC CGC CTA CGC CTT TTT CAT CAC-3′) and MGAR (5′-GTA AGG GCT GCC GCT GTT TTC AGG GTT TAA CTC TAA-3′) as well as SICF (5′-TTA GAG TTA AAC CCT GAA AAC AGC GGC AGC CCT TAC-3′) and SICR (5′-CCC CCC CTC GAG TCT TCA GGC CAA TCT TCA-3′) for mga and sic, respectively. The amplified DNA fragments, which overlap each other at one end, were then used as templates for subsequent PCR to generate DNA fragment mga-sic by using primers MGAF and SICR. The resulting DNA fragment was digested with NotI and XhoI (restriction sites are underlined in primer sequences above) and ligated to pG⁺host5. Plasmid pG⁺host5::mga-sic was used to create the deletion mutant 90-226 Δemm1 by gene replacement as described previously (4). PCR-amplified DNA from one recombinant, including 1.2 kb of mga-sic and the flanking region, was sequenced to confirm deletion of the emm1 gene.

Construction of strain 90-226 emm1(ΔAB).

Plasmid pG⁺host5::emm1ΔAB, carrying a 486-bp in-frame deletion within emm1, was constructed by subcloning a BamHI/HindIII fragment from the plasmid pPE6 (42) into pG⁺host5 after deleting a BglII/BclI fragment. In addition to the deletion, emm1ΔAB has an insertion of a 17-residue ovalbumin epitope. This insertion has no effect on known M-protein functions (42). The deletion recombinant, 90-226 emm1(ΔAB), was created using this construct by gene replacement (4). The chromosomal AB deletion in the emm1 gene was confirmed by sequencing PCR-amplified DNA from one recombinant. The oligonucleotide primers used for PCR were complementary to flanking sequences of the deleted region of emm1. The expression of a truncated M protein in the mutant strain was confirmed by immunoblotting of cell wall proteins extracted from streptococci (Fig. 2).

FIG. 2.

(a) Expression of an M protein with deletion of the AB repeat on the surface of 90-226 emm1(ΔAB). Bacterial surface proteins were extracted as described in Materials and Methods, separated on SDS-12% PAGE, transferred to a polyvinylidene difluoride membrane, and probed with antibody 10B6. Antibody 10B6 is a monoclonal antibody raised against a 15-amino-acid peptide within the C repeats of M protein. Lanes: 1, 90-226; 2, 90-226 emm1(ΔAB). (b) M1 protein associated with the surface of 90-226 emm1::Km. Bacterial surface proteins were extracted and electrophoresed through 8% polyacrylamide-SDS gels and were transferred to a nitrocellulose membrane. The membrane was incubated with an anti-M antibody and developed with horseradish peroxidase-conjugated secondary antibody and chromogenic substrate. Anti-M1 serum was raised by immunizing rabbits with an M1 peptide corresponding to the N-terminal amino acids 1 to 23 of the mature M1 protein. Lanes: 1, 90-226; 2, 90-226 emm1(ΔAB); 3, 90-226 emm1::Km.

Construction of a 90-226 speB mutant.

Plasmid pUCErm1, carrying a speB knockout construct (2, 36), was introduced into 90-226 by transformation. Transformants were selected on BAP supplemented with erythromycin (5 μg/ml). The plasmid integration into the chromosome was verified by PCR. Culture supernatants of the mutant showed a truncated protein reactive with anti-SpeB serum in immunoblots, which has no cysteine protease activity (data not shown).

Phagocytosis assays.

Human blood phagocytosis assays were performed as described previously (34). Briefly, streptococcal log-phase cultures were diluted in THY to 5 × 10³ CFU/ml. One-hundred microliters of bacterial suspension was mixed with 0.9 ml of fresh human blood or plasma and rotated at 37°C for 3 h. Viable counts were determined by plating diluted samples onto BAP.

Protein preparations and antibodies.

Surface proteins were extracted using a method described by Ji et al. (29). Anti-M1(1-23) serum (13) was raised against a synthetic peptide, corresponding to the N-terminal amino acid residues 1 to 23 of M1 protein, and was a gift from J. Dale, University of Tennessee, Memphis. Antibody 10B6 reacts with a 15-amino-acid sequence within the C repeats of M protein and was kindly provided by V. A. Fischetti, Rockefeller University, New York, N.Y. (23). To prepare supernatant proteins of M1 strains, GAS were grown in THY for 15 h and supernatant was collected. Proteins in supernatants were concentrated fivefold by filtration using Centricon YM-10 filters (Millipore, Billerica, Mass.). M49 strains were grown in THY for 15 h, and bacterial cells were harvested by centrifugation at 4,000 × g for 10 min at 4°C. The supernatant was filtered through a 0.22-μm-pore-size membrane and concentrated by precipitation with 80% ammonium sulfate at room temperature. After dialysis, protein was concentrated twofold using Centriprep YM-10 filters (Millipore). Anti-SpeB serum, as well as purified recombinant SpeB, were kindly provided by P. M. Schlievert, Minneapolis, Minn. (5).

Cysteine protease activity assay.

Cysteine protease activity was measured using the assay described by North (41). Briefly, 50-μl culture supernatants from M1 strains were incubated with 2 mM dithiothreitol (DTT) for 30 min at 37°C to activate the enzyme. Sixty microliters of 2.5 M n-benzoyl-proline-phenylalanine-arginine-p-nitroanilide hydrochloride (pH 4) (Sigma, St. Louis, Mo.) in distilled water was mixed with 90 μl of 0.1 M phosphate buffer (pH 6) and then added to activated culture supernatant. The change in absorbance was determined at an optical density of 405 nm (OD₄₀₅). Assays were performed in triplicate; protein preparations were tested at least three times. E64 cysteine protease inhibitor was tested in parallel assays to confirm specificity. To measure cysteine protease activity of M49 strains, 7 μg of concentrated supernatant protein was activated with 10 mM DTT for 2 h at 37°C. Protein concentrations were determined by the bicinchoninic acid method (Pierce Chemical Co., Rockford, Ill.), and densitometric analysis of the amount of SpeB protein specifically evaluated band intensities on Western blot membranes by using Scion Image 1.62. Cleavage of n-Bz-Pro-Phe-Arg-p-nitroanilide was measured as described above.

SDS-polyacrylamide gel electrophoresis (PAGE) and Western blot analysis.

Sodium dodecyl sulfate (SDS)—8 or 12% polyacrylamide gels were run using a Bio-Rad Protein Gel kit (Hercules, Calif.). Proteins were either stained with Coomassie brilliant blue or transferred to a nitrocellulose membrane (Bio-Rad). Membranes were blocked with 0.5% gelatin for 1 h, washed three times in Tris-buffered saline, and incubated for 2 h with the antibodies described above. After the membranes were washed, horseradish peroxidase-conjugated goat anti-rabbit IgG or rabbit anti-mouse IgG antibodies (Sigma) were added for 2 h. Blots were developed with 4-chloro-1-naphthol (Bio-Rad).

RESULTS

Characterization of strains 90-226 Δemm1 and 90-226 emm1(ΔAB).

To evaluate the effect of complete and partial deletions of the emm1 gene on resistance to phagocytosis, survival of bacteria in human blood was measured. After 3 h of incubation in fresh human blood, viable counts of both mutants 90-226 Δemm1 and 90-226 emm1(ΔAB) were significantly reduced (∼50-fold), whereas wild-type streptococci increased in numbers by 10- to 30-fold. During the same period, growth rates of wild-type and mutant strains in human plasma were similar. This result confirmed that M protein plays an important role in resistance to phagocytosis and also suggested that this function is in part dependent on the AB sequence of M protein.

Expression of M protein by mutant streptococci was compared to that of wild-type strains by Western blot analysis, using the antibody 10B6 that reacts with a 15-amino-acid sequence within the C repeats (23). 90-226 Δemm1 extracts did not bind the antibody (data not shown), whereas a 31-kDa band was observed in extracts from strain 90-226 emm1(ΔAB) (Fig. 2a, lane 2). Blots of surface protein extracts from 90-226, 90-226 emm1(ΔAB), and 90-226 emm1::Km were also treated with anti-M1(1-23) serum (13). The extract of 90-226 emm1(ΔAB) did not react with this antiserum, confirming the expected deletion (Fig. 2b, lane 2). Extracts from 90-226 emm1::Km, the same strain analyzed by Collin and Olsén (9), revealed variable amounts of M1 protein associated with the bacterial envelope, even though the truncated protein without cell wall anchor and sorting sequences is expected to be secreted into the media (Fig. 2b, lane 3). The molecular size of the band was estimated to be 29 kDa, which is in accordance with the location of transposon insertion.

SpeB expression in culture supernatants.

The migration of SpeB in SDS gels was investigated to evaluate the maturation of the cysteine protease produced by strains carrying the deletions of the emm1 gene. Western blot analysis of culture supernatants from cultures of 90-226, 90-226 Δemm1, and 90-226 emm1(ΔAB) showed mature enzymes that were primarily 28 kDa in size (Fig. 3a). As reported by Collin and Olsén (9), however, extracts of strain 90-226 emm1::Km revealed a different pattern of bands; the 40-kDa band predominated, and no mature 28-kDa enzyme was detected in the supernatants of this strain (Fig. 3a). Protease folding and cleavage of the propeptide is accelerated under reducing conditions (20). Therefore, supernatants were incubated with 10 mM DTT for 30 min, but there was no effect on the size of protease produced by strain 90-226 emm1::Km (Fig. 3a, lanes 5 and 6). Hence, Western blot analysis of strain 90-226 and these mutants suggested that cell-wall-anchored M1 protein was not required for the cleavage process that results in an active protease.

FIG. 3.

(a) SpeB in culture supernatants of 90-226 and isogenic mutants. Culture supernatant from bacteria grown in THY overnight was collected and concentrated as described in Materials and Methods. Samples were then separated on SDS-12% PAGE, blotted to a membrane, and probed with antibody against mature cysteine proteinase. Lanes: 1, purified mature SpeB protein; 2, 90-226; 3, 90-226 emm1(ΔAB); 4, 90-226 Δemm1; 5, 90-226 emm1::Km; 6, 90-226 emm1::Km (30 min DTT). (b) SpeB in culture supernatants of CS101 and MJY1-3. Culture supernatants from bacteria grown in THY for 15 h were collected and concentrated by precipitation with ammonium sulfate. Samples were then incubated at 37°C for 2 h in the presence or absence of 10 mM DTT, run on SDS-12% PAGE, and probed with antibody against mature cysteine proteinase. Lanes: 1, purified mature SpeB protein; 2, CS101 (no DTT); 3, CS101 (2 h DTT); 4, MJY1-3 (no DTT); 5, MJY1-3 (2 h DTT).

The above results were confirmed using another GAS serotype. Supernatant proteins from the M49 strain CS101 and its isogenic mutant MJY1-3 were analyzed. Strain MJY1-3 lacks the emm49, mrp, and enn genes, thus, no M or M-like proteins are expressed on its surface (29). In both strains the 40-kDa zymogen and several intermediates were detected when analyzed without activation by DTT (Fig. 3b, lane 2 and 4). After 2 h of incubation with 10 mM DTT, however, CS101 and MJY1-3 extracts contained only the 28-kDa band, confirming that zymogen and intermediates were processed to mature enzyme (Fig. 3b, lane 3 and 5). This experiment again confirms that M protein is not required for maturation of the cysteine protease, SpeB.

SpeB activity in culture supernatants.

The above immunoblot experiments suggested that M⁻ strains 90-226 Δemm1 and 90-226 emm1(ΔAB) and MJY1-3 should produce enzymatically active SpeB protease activity at levels comparable to that of the wild-type culture. Strain 90-226 and its mutants were grown in THY for 15 h, the filtered supernatants were activated with DTT, and then they were incubated with the chromogenic substrate n-Bz-Pro-Phe-Arg-p-nitroanilide, as described in Materials and Methods. After 2 and 6 h of incubation, relative amounts of cleaved substrate were determined spectrophotometrically. Mutant strains 90-226 Δemm1 and 90-226 emm1(ΔAB) produced slightly increased levels of protease activity compared to that of the wild type (Fig. 4a, columns 2 and 3). Strain 90-226 emm1::Km, as previously reported, failed to cleave the substrate (column 4). Therefore, the implications of the Western blot analyses confirmed that the deletion of M1 did not decrease the proteolytic activity of supernatant proteins. This conclusion was also supported by proteolytic activity assays of supernatant proteins from strains CS101 and MJY1-3. Proteolysis assays were performed as described above for M1 strains. After 24 h of incubation with n-Bz-Pro-Phe-Arg-p-nitroanilide, supernatant proteins from CS101 and MJY1-3 strains showed similar levels of proteolytic activity (Fig. 4b). Overall, the amount of protease activity produced by these strains was much lower than that of M1 strain 90-226, requiring longer incubation times to produce measurable cleavage of substrate. Nevertheless, Western immunoblot studies were again confirmed: deletion of emm49, mrp, and enn did not affect the activity of SpeB. Taken together, these results show that maturation of the streptococcal cysteine protease SpeB to an active 28-kDa enzyme does not require M protein.

FIG. 4.

(a) Relative proteolytic activity in culture supernatants of M1 strain 90-226 and isogenic mutants. Culture supernatants were activated with a reducing agent, and proteolytic cleavage of n-Bz-Pro-Phe-Arg-p-nitroanilide was measured as absorbancy at 405 nm after incubation at 37°C for 2 h (left) or 6 h (right). Bars: 1, 90-226; 2, 90-226 Δemm1; 3, 90-226 emm1(ΔAB); 4, 90-226 emm1::Km. Column 4 is too small to be visible in the figure. (b) Proteolytic activity of culture supernatant proteins of M49 strain CS101 and isogenic mutant. Culture supernatants were concentrated by precipitation with ammonium sulfate. After dialysis, samples were further concentrated using Centriprep YM-10 and then activated with a reducing agent. Proteolytic cleavage of n-Bz-Pro-Phe-Arg-p-nitroanilide was measured as absorbancy at 405 nm after incubation at 37°C for 6 h (left) or 24 h (right). Bars: 1, CS101; 2, MJY1-3.

DISCUSSION

A variety of activities have been attributed to the streptococcal cysteine protease, or SpeB, which may influence the virulence of GAS. These included activation of apoptosis, activation of interleukin-1 precursor, superantigenicity, and binding of fibronectin. Expression of SpeB is globally regulated and growth phase dependent. SpeB was found to be primarily produced after periods of growth longer than 8 h (7, 18, 45). Although posttranslational activation of the enzyme has long been known to depend on sulfhydryl reduction of the protein, Collin and Olsén more recently reported that in vitro maturation of SpeB expressed by GAS was also dependent on cell-wall-anchored M1 protein (9). They suggested that the zymogen form of SpeB, produced by M⁻ mutant strains, was folded in an aberrant manner (9) which physically blocked the active site of the protease. Their experiments showed that addition of M protein to growth medium did not influence cleavage of the zymogen, but they did not question whether mutation interfered with processing or activation of other surface-bound or -secreted streptococcal proteins. Their conclusion that maturation of SpeB depended on M protein was primarily based on two M⁻ strains with plasmid insertion mutations in the emm1 gene. One mutant, strain 90-226 emm1::Km, was constructed in our laboratory and was, therefore, available to us for further study. We investigated the alternative hypothesis that truncated fragments of M1 protein produced by insertion mutants could interfere with folding and/or secretion of the protease.

To test the above hypothesis, we constructed two strains that contain complete or partial deletion of emm1. Neither of these strains had decreased cysteine protease activity. In fact, the activity had increased compared to that of the wild type. Western blot analysis clearly revealed that despite the mutation of M1 protein, both strains were able to generate a 28-kDa enzyme. We also evaluated the effect of a mutation in an M49 strain MJY1-3 (29) that deleted all M-like surface proteins encoded by emm49, mrp, and enn genes on secretion and maturation of the cysteine protease. These experiments confirmed studies with the serotype M1 strain: deletion of M49 and M-like proteins did not affect the translocation or activation of the cysteine protease. It is very unlikely, therefore, that cell-wall-anchored M protein influences the maturation of SpeB. It seems plausible that truncated M1 protein expressed in insertional mutant strains dramatically changes the transport and processing machinery of GAS. It could interfere with processes involved in protein export across the cell wall, for example, which ultimately inhibit a correct secretion of the cysteine protease.

Translocation of secreted proteins across the cytoplasmic membrane and the cell wall for gram-positive bacteria has been studied extensively in Bacillus subtilis (51). The genome of GAS reveals the sequence of proteins with significant homologies to SecA and SecY proteins, two essential components of the cytoplasmic membrane translocation apparatus, as well as the signal peptidase SipS. Thus, it seems likely that secretion mechanisms in GAS are similar to those in bacilli. We postulate that secreted malformed M protein interacts with this apparatus to disrupt appropriate folding of secreted proteins. We observed various amounts of truncated M protein in cell surface extracts from strain 90-226 emm1::Km, one of the strains examined previously (9). The nature of this interaction is unknown, but it is likely that the truncated M1 protein is trapped either in the membrane or the peptidoglycan cell wall. Dieye et al. (16) reported a similar phenomenon for Lactococcus lactis. Approximately 50% of a mutated protein was found associated with the cell wall, rather than being secreted into the growth media as expected. The interaction between the secreted molecule and the cell wall was suggested to involve electrostatic interactions (16). Collin and Olsén analyzed two different M⁻ strains with plasmid insertions in the emm1 gene (9). Both failed to produce proteolytically active SpeB. During translocation, M1 could accumulate in a similar way and, for that reason, interfere with the secretion of other proteins.

In conclusion, we reproduced Collin and Olsén's results (9) that insertion mutations in the emm1 gene impede maturation of SpeB into an active protease. However, complete or partial deletion of M1 protein did not negatively affect protease activity or prevent conformational changes required for autocleavage of the 40-kDa zymogen into the smaller 28-kDa mature enzyme. Therefore, the results of Collin and Olsén were correct, but their interpretation failed to consider the potential negative effect of M-protein fragments produced by the mutants they studied. We conclude that M protein is not required for secretion, folding, or autocatalytic activation of the cysteine protease.