Isolation and sequence analysis of cDNA clones for the small subunit of rabbit calcium-dependent protease

We have isolated and sequenced cDNA clones for the small subunit (30-kDa subunit) of rabbit calcium-dependent protease (Ca2+-protease) using synthesized oligodeoxynucleotide probes based on the partial amino acid sequence of the protein. A nearly full-length cDNA clone containing the total amino acid coding sequence was obtained. From the deduced sequence, the following conclusions about possible functions of the protein are presented. The kDa subunit comprises 266 residues (Mr = 28,238). The N-terminal region (64 residues) is mainly composed of glycine (37 residues) and hydrophobic amino acids and may interact with the cell membrane or an organelle. The sequence of the C-terminal 168 residues is highly homologous to the corresponding C-terminal region of the large subunit (80-kDa subunit) which has been identified as the calcium-binding domain. This region of the 30-kDa subunit contains four E-F hand structures and presumably binds Ca2+, as in the case of the 80-kDa subunit. Thus, the 30-kDa subunit may play important roles in regulating enzyme activity and/or possibly in determining the location of the Ca2+-protease. The marked sequence homology of the C-terminal regions of the two subunits may indicate that the calcium-binding domains have evolved from the same ancestral gene.     

Cloning and nucleotide sequence analysis of the colH gene from Clostridium histolyticum encoding a collagenase and a gelatinase.

The colH gene encoding a collagenase was cloned from Clostridium histolyticum JCM 1403. Nucleotide sequencing showed a major open reading frame encoding a 116-kDa protein of 1,021 amino acid residues. The deduced amino acid sequence contains a putative signal sequence and a zinc metalloprotease consensus sequence, HEXXH. A 116-kDa collagenase and a 98-kDa gelatinase were copurified from culture supernatants of C. histolyticum. While the former degraded both native and denatured collagen, the latter degraded only denatured collagen. Peptide mapping with V8 protease showed that all peptide fragments, except a few minor ones, liberated from the two enzymes coincided with each other. Analysis of the N-terminal amino acid sequence of the two enzymes revealed that their first 24 amino acid residues were identical and coincided with those deduced from the nucleotide sequence. These results indicate that the 98-kDa gelatinase is generated from the 116-kDa collagenase by cleaving off the C-terminal region, which could be responsible for binding or increasing the accessibility of the collagenase to native collagen fibers. The role of the C-terminal region in the functional and evolutional aspects of the collagenase was further studied by comparing the amino acid sequence of the C. histolyticum collagenase with those of three homologous enzymes: the collagenases from Clostridium perfringens and Vibrio alginolyticus and Achromobacter lyticus protease I.     

Purification of twelve cyanogen bromide fragments from bovine plasma fibronectin and the amino acid sequence of eight of them. Overlap evidence aligning two plasmic fragments, internal homology in gelatin-binding region and phosphorylation site near C terminus

Twelve cyanogen bromide fragments (CB1-12) from bovine plasma fibronectin have been isolated and eight of these completely sequenced. Altogether they account for 502 of the total expected 1880 residues in each of the two chains of fibronectin. Four of these fragments (CB1-4) constitute residues 1-289 in fibronectin with CB4 overlapping the N-terminal 29-kDa plasmic fragment to the second plasmic fragment, of 170-kDa in fibronectin. Fragments CB 5-9 are all contained within a 45-kDa gelatin-binding region, which is N-terminal in the 170-kDa fragment. The sequence of two of these five fragments in the 45-kDa fragment (CB7-8) contains two mutually homologous stretches with 57% sequence identity. Another two fragments (CB10-11) are derived from the heparin-binding region of the 170-kDa fragment. CB12 constitutes the C-terminal 13-residue stretch in fibronectin and contains a partly phosphorylated serine residue in the C-terminal sequence: -Arg-Glu-Asp-Ser(P)-Arg-Glu.     

Activation of mammalian DNA methyltransferase by cleavage of a Zn binding regulatory domain.

Mammalian DNA (cytosine-5) methyltransferase contains a C-terminal domain that is closely related to bacterial cytosine-5 restriction methyltransferase. This methyltransferase domain is linked to a large N-terminal domain. It is shown here that the N-terminal domain contains a Zn binding site and that the N- and C-terminal domains can be separated by cleavage with trypsin or Staphylococcus aureus protease V8; the protease V8 cleavage site was determined by Edman degradation to lie 10 residues C-terminal of the run of alternating lysyl and glycyl residues which joins the two domains and six residues N-terminal of the first sequence motif conserved between the mammalian and bacterial cytosine methyltransferases. While the intact enzyme had little activity on unmethylated DNA substrates, cleavage between the domains caused a large stimulation of the initial velocity of methylation of unmethylated DNA without substantial change in the rate of methylation of hemimethylated DNA. These findings indicate that the N-terminal domain of DNA methyltransferase ensures the clonal propagation of methylation patterns through inhibition of the de novo activity of the C-terminal domain. Mammalian DNA methyltransferase is likely to have arisen via fusion of a prokaryotic-like restriction methyltransferase and an unrelated DNA binding protein. Stimulation of the de novo activity of DNA methyltransferase by proteolytic cleavage in vivo may contribute to the process of ectopic methylation observed in the DNA of aging animals, tumors and in lines of cultured cells.     

Cloning and Sequence Analysis of a Protease-encoding Gene from the Marine Bacterium Alteromonas sp. Strain O-7

The gene (aprI) encoding alkaline serine protease (AprI; subtilase) from Alteromonas sp. strain O-7 was cloned and sequenced. The nucleotide sequence of aprI has been identified. The deduced amino acid sequence indicated that aprI codes for a precursor of 715 amino acids and the precursor is composed of four regions including a signal peptide, an N-terminal pro-region, a mature protease region and a C-terminal extension region of 215 amino acids as previously described for aprII [H. Tsujibo et al., Gene, 136, 247–251 (1993)]. The amino acid sequence of the mature AprI (AprI-M) showed high sequence homology with those of other class I subtilases. The C-terminal region was characterized by a repeat of 94 amino acids residues, which showed about 50% similarity with those of the C-terminal pro-region of several known proteases from Gram-negative bacteria.     

Synthesis and processing of the transmembrane envelope protein of equine infectious anemia virus.

The transmembrane (TM) envelope protein of lentiviruses, including equine infectious anemia virus (EIAV), is significantly larger than that of other retroviruses and may extend in the C-terminal direction 100 to 200 amino acids beyond the TM domain. This size difference suggests a lentivirus-specific function for the long C-terminal extension. We have investigated the synthesis and processing of the EIAV TM protein by immune precipitation and immunoblotting experiments, by using several envelope-specific peptide antisera. We show that the TM protein in EIAV particles is cleaved by proteolysis to an N-terminal glycosylated 32- to 35-kilodalton (kDa) segment and a C-terminal nonglycosylated 20-kDa segment. The 20-kDa fragment was isolated from virus fractionated by high-pressure liquid chromatography, and its N-terminal amino acid sequence was determined for 13 residues. Together with the known nucleotide sequence, this fixes the cleavage site at a His-Leu bond located 240 amino acids from the N terminus of the TM protein. Since the 32- to 35-kDa fragment and the 20-kDa fragment are not detectable in infected cells, we assume that cleavage occurs in the virus particle and that the viral protease may be responsible. We have also found that some cells producing a tissue-culture-adapted strain of EIAV synthesize a truncated envelope precursor polyprotein. The point of truncation differs slightly in the two cases we have observed but lies just downstream from the membrane-spanning domain, close to the cleavage point described above. In one case, virus producing the truncated envelope protein appeared to be much more infectious than virus producing the full-size protein, suggesting that host cell factors can select for virus on the basis of the C-terminal domain of the TM protein.     

The complete amino-acid sequence of the sweet protein thaumatin I.

The primary structure of the sweet-tasting protein thaumatin has been elucidated. The protein consists of a single polypeptide chain of 207 residues. The sequence of the N-terminal part of the chain was determined by sequenator analysis. As the protein contains only one methionine residue, it was possible to deduce the N-terminal sequence of the C-terminal cyanogen bromide fragment by automatic sequencing of the cyanogen-bromide-cleaved, succinylated protein. To arrive at the sequence of the whole protein tryptic and Staphylococcus protease peptides, together with chymotryptic peptides and a 2-(2-nitrophenylsulfenyl)-3-methyl-3'-bromoindolenine (BNPS-skatole) fragment were also sequenced. Comparing the amino acid sequence of thaumatin with that of the other sweet-tasting protein, monellin, we have located five sets of identical tripeptides. Since immunological cross-reactivity of thaumatin antibodies with monellin has recently been described, one or more of these tripeptides might be part of a common antibody recombination site and possibly be involved in the interaction with the sweet-taste receptor.     

Linker polypeptides of the phycobilisome from the cyanobacterium Mastigocladus laminosus: amino-acid sequences and relationships

Three linker polypeptides of the phycobilisome from the cyanobacterium Mastigocladus laminosus were isolated: A 8.9-kDa polypeptide, L8.9R(C), which is probably associated with C-phycocyanin, a 34.5-kDa polypeptide, L34.5,PCR, which forms a complex with C-phycocyanin, and a 34.5-kDa polypeptide, L34.5,PECR, which is linked to phycoerythrocyanin. The complete amino-acid sequence (80 residues) of the L8.9R(C) polypeptide was determined as well as the N-terminal 44 residues of both L34.5R polypeptides and the 114 C-terminal residues of L34.5,PECR. L8.9R(C) is homologous to L8.9C (Füglistaller et al. (1984) Hoppe-Seyler's Z. Physiol. Chem. 365, 1085-1096) and to the C-terminal sequence of L34.5,PECR. The N-terminal sequences of L34.5,PECR and L34.5,PCR exhibit 34% homology. The 44 N-terminal residues of L34.5,PECR are related to the beta-subunit of phycoerythrocyanin (23% homology), while the C-terminal sequence of L34.5,PECR is more related to alpha PEC (21% homology within 60 residues). This suggests that the 30-kDa-linker polypeptide family originates from a fusion of the alpha- and beta-subunit genes and the corresponding intercistronic DNA sequence, as might have arisen through mutation in the stop-codon of the beta-subunit gene. Hence, all polypeptides of the phycobilisome (including perhaps the anchor polypeptide) may be derived from an early ancestor phycobiliprotein subunit, which itself is also related to myoglobin (Schirmer et al. (1985) J. Mol. Biol. 184, 251-277).     

Cloning and Characterization of the cDNA Encoding a Novel Brain-Specific 14-kDa Protein

A new acidic protein with a molecular weight of 14,000 was purified from rat brain, in which it was specifically expressed, and partially sequenced by protein sequencing. On the basis of results obtained from the amino acid sequences, mixed oligonucleotides were synthesized and used as probes to clone a cDNA from a rat brain cDNA library. The cloned cDNA provided the full-length sequence of the 14-kDa protein. Northern blot hybridization using total RNA from several tissues of the rat provided evidence that the 14-kDa protein was expressed specifically in rat brain. Transfection of this cDNA into mammalian cells resulted in expression of the 14-kDa protein. The amino acid sequence predicted from the cDNA of the rat brain 14-kDa protein contained 137 amino acid residues. A hydropathy profile revealed a hydrophobic domain (amino acids 60-80) flanked by highly hydrophilic stretches on both sides. Whereas the N-terminal region of the 14-kDa protein contained four repeating motifs, EKTKEGV, the C-terminal domain was rich in glutamic acid and proline. A computer search of the amino acid sequence of the 14-kDa protein indicated no homology to any other protein reported so far.     

Characterization of nicking of the nontoxic-nonhemagglutinin components of Clostridium botulinum types C and D progenitor toxin

Clostridium botulinum C and D strains produce two types of progenitor toxins, M and L. Previously we reported that a 130-kDa nontoxic-nonhemagglutinin (NTNHA) component of the M toxin produced by type D strain CB16 was nicked at a unique site, leading to a 15-kDa N-terminal fragment and a 115-kDa C-terminal fragment. In this study, we identified the amino acid sequences around the nicking sites in the NTNHAs of the M toxins produced by C. botulinum type C and D strains by analysis of their C-terminal and N-terminal sequences and mass spectrometry. The C-terminus of the 15-kDa fragments was identified as Lys127 from these strains, indicating that a bacterial trypsin-like protease is responsible for the nicking. The 115-kDa fragment had mixtures of three different N-terminal amino acid sequences beginning with Leu135, Val139, and Ser141, indicating that 7–13 amino acid residues were deleted from the nicking site. The sequence beginning with Leu135 would also suggest cleavage by a trypsin-like protease, while the other two N-terminal amino acid sequences beginning with Val139 and Ser141 would imply proteolysis by an unknown protease. The nicked NTNHA forms a binary complex of two fragments that could not be separated without sodium dodecyl sulfate.     

Biotin protein ligase from Saccharomyces cerevisiae. The N-terminal domain is required for complete activity.

Catalytically active biotin protein ligase from Saccharomyces cerevisiae (EC 6.3.4.15) was overexpressed in Escherichia coli and purified to near homogeneity in three steps. Kinetic analysis demonstrated that the substrates ATP, biotin, and the biotin-accepting protein bind in an ordered manner in the reaction mechanism. Treatment with any of three proteases of differing specificity in vitro revealed that the sequence between residues 240 and 260 was extremely sensitive to proteolysis, suggesting that it forms an exposed linker between an N-terminal 27-kDa domain and the C-terminal 50-kDa domain containing the active site. The protease susceptibility of this linker region was considerably reduced in the presence of ATP and biotin. A second protease-sensitive sequence, located in the presumptive catalytic site, was protected against digestion by the substrates. Expression of N-terminally truncated variants of the yeast enzyme failed to complement E. coli strains defective in biotin protein ligase activity. In vitro assays performed with purified N-terminally truncated enzyme revealed that removal of the N-terminal domain reduced BPL activity by greater than 3500-fold. Our data indicate that both the N-terminal domain and the C-terminal domain containing the active site are necessary for complete catalytic function.     

Insertion of the mannitol permease into the membrane of Escherichia coli. Possible involvement of an N-terminal amphiphilic sequence

The in vivo membrane assembly of the mannitol permease, the mannitol Enzyme II (IImtl) of the Escherichia coli phosphotransferase system, has been studied employing molecular genetic approaches. Removal of the N-terminal amphiphilic leader of the permease and replacement with a short hydrophobic sequence resulted in an inactive protein unable to transport mannitol into the cell or catalyze either phosphoenol-pyruvate-dependent or mannitol 1-phosphate-dependent mannitol phosphorylation in vitro. The altered protein (68 kDa) was quantitatively cleaved by an endogenous protease to a membrane-associated 39-kDa fragment and a soluble 28-kDa fragment as revealed by Western blot analyses. Overproduction of the wild-type plasmid-encoded protein also led to cleavage, but repression of the synthesis of the plasmid-encoded enzyme by inclusion of glucose in the growth medium prevented cleavage. Several mtlA-phoA gene fusions encoding fused proteins with N-terminal regions derived from the mannitol permease and C-terminal regions derived from the mature portion of alkaline phosphatase were constructed. In the first fusion protein, F13, the N-terminal 13-aminoacyl residue amphiphilic leader sequence of the mannitol permease replaced the hydrophobic leader sequence of alkaline phosphatase. The resultant fusion protein was inefficiently translocated across the cytoplasmic membrane and became peripherally associated with both the inner and outer membranes, presumably via the noncleavable N-terminal amphiphilic sequence. The second fusion protein, F53, in which the N-terminal 53 residues of the mannitol permease were fused to alkaline phosphatase, was efficiently translocated across the cytoplasmic membrane and was largely found anchored to the inner membrane with the catalytic domain of alkaline phosphatase facing the periplasm. This 53-aminoacyl residue sequence included the amphiphilic leader sequence and a single hydrophobic, potentially transmembrane, segment. Analyses of other MtlA-PhoA fusion proteins led to the suggestion that internal amphiphilic segments may function to facilitate initiation of polypeptide trans-membrane translocation. The dependence of IImtl insertion on the N-terminal amphiphilic leader sequence was substantiated employing site-specific mutagenesis. The N-terminal sequence of the native permease is Met-Ser-Ser-Asp-Ile-Lys-Ile-Lys-Val-Gln-Ser-Phe-Gly.... The following point mutants were isolated, sequenced, and examined regarding the effects of the mutations on insertion of IImtl into the membrane: 1) S3P; 2) D4P; 3) D4L; 4) D4R; 5) D4H; 6) I5N; 7) K6P; and 8) K8P.(ABSTRACT TRUNCATED AT 400 WORDS)     

Dichain structure of botulinum neurotoxin: identification of cleavage sites in types C, D, and F neurotoxin molecules

Botulinum neurotoxin (NT) is synthesized by Clostridium botulinum as about a 150-kDa single-chain polypeptide. Posttranslational modification by bacterial or exogenous proteases yielded dichain structure which formed a disulfide loop connecting a 50-kDa light chain (Lc) and 100-kDa heavy chain (Hc). We determined amino acid sequences around cleavage sites in the loop region of botulinum NTs produced by type C strain Stockholm, type D strain CB16, and type F strain Oslo by analysis of the C-terminal sequence of Lc and the N-terminal sequence of Hc. Cleavage was found at one or two sites at Arg444/Ser445 and Lys449/Thr450 for type C, and Lys442/Asn443 and Arg445/Asp446 for type D, respectively. In culture fluid of mildly proteolytic strains of type C and D, therefore, NT exists as a mixture of at least three forms of nicked dichain molecules. The NT of type F proteolytic strain Oslo showed the Arg435 as a C-terminal residue of Lc and Ala440 as an N-terminal residue of Hc, indicating that the bacterial protease cuts twice (Arg435/Lys436 and Lys439/Ala440), with excision of four amino acid residues. The location of cleavage and number of amino acid residue excisions in the loop region could be explained by the degree of exposure of amino acid residues on the surface of the molecule, which was predicted as surface probability from the amino acid sequence. In addition, the observed correlation may also be adapted to the cleavage sites of the other botulinum toxin types, A, B, E, and G.     

Synthesis of an immunopathogenic fusion protein derived from a bovine interphotoreceptor retinoid-binding protein cDNA clone

We have extended the cDNA sequence of bovine interphotoreceptor retinoid-binding protein (IRBP) and subcloned one of the sequenced cDNA fragments into an expression vector. The nucleotide (nt) sequences of four bovine IRBP cDNA clones have been determined. These sequences when assembled cover the 3' proximal 3629 nt of the IRBP mRNA and encode the C-terminal 551 amino acids (aa) of IRBP. This cDNA sequence validates the intron: exon boundaries predicted from the gene. A 2-kb EcoRI insert from lambda IRBP2, one of the clones sequenced, encoding the C-terminal 136 aa of IRBP was subcloned into the expression vector pWR590-1. Escherichia coli carrying this plasmid construction, pXS590-IRBP, produced a fusion protein containing 583 N-terminal aa of beta-galactosidase, three linker aa residues, 136 C-terminal aa of IRBP and possibly a number of additional C-terminal residues due to suppressed termination. This 86-kDa fusion protein, purified by detergent/chaotrope extraction followed by reverse-phase high-performance liquid chromatography, cross-reacted with anti-bovine IRBP on Western blots. This protein induced an experimental autoimmune uveo-retinitis and experimental autoimmune pinealitis in Lewis rats indistinguishable from that induced by authentic bovine IRBP. Thus, it is evident that biological activity of this region of IRBP, as manifested by immuno-pathogenicity, is retained by the fusion protein.     

Localisation of light chain and actin binding sites on myosin

A gel overlay technique has been used to identify a region of the myosin S-1 heavy chain that binds myosin light chains (regulatory and essential) and actin. The 125I-labelled myosin light chains and actin bound to intact vertebrate skeletal or smooth muscle myosin, S-1 prepared from these myosins and the C-terminal tryptic fragments from them (i.e. the 20-kDa or 24-kDa fragments of skeletal muscle myosin chymotryptic or Mg2+/papain S-1 respectively). MgATP abolished actin binding to myosin and to S-1 but had no effect on binding to the C-terminal tryptic fragments of S-1. The light chains and actin appeared to bind to specific and distinct regions on the S-1 heavy chain, as there was no marked competition in gel overlay experiments in the presence of 50-100 molar excess of unlabelled competing protein. The skeletal muscle C-terminal 24-kDa fragment was isolated from a tryptic digest of Mg2+/papain S-1 by CM-cellulose chromatography, in the presence of 8 M urea. This fragment was characterised by retention of the specific label (1,5-I-AEDANS) on the SH1 thiol residue, by its amino acid composition, and by N-terminal and C-terminal sequence analyses. Electron microscopical examination of this S-1 C-terminal fragment revealed that: it had a strong tendency to form aggregates with itself, appearing as small 'segment-like' structures that formed larger aggregates, and it bound actin, apparently bundling and severing actin filaments. Further digestion of this 24-kDa fragment with Staphylococcus aureus V-8 protease produced a 10-12-kDa peptide, which retained the ability to bind light chains and actin in gel overlay experiments. This 10-12-kDa peptide was derived from the region between the SH1 thiol residue and the C-terminus of S-1. It was further shown that the C-terminal portion, but not the N-terminal portion, of the DTNB regulatory light chain bound this heavy chain region. Although at present nothing can be said about the three-dimensional arrangement of the binding sites for the two kinds of light chain (regulatory and essential) and actin in S-1, it appears that these sites are all located within a length of the S-1 heavy chain of about 100 amino acid residues.     

Brain micro glutamic acid-rich protein is the C-terminal endpiece of the neurofilament 68-kDa protein as determined by the primary sequence

The amino acid sequence of bovine brain micro glutamic acid-rich protein was determined by analysis of tryptic and Trimeresurus flavoviridis protease peptides of the molecule. The protein comprised 82 amino acid residues and has an Mr of 8992. The established sequence was highly homologous (90% identity) to the sequence of C-terminal 82 residues of the neurofilament 68-kDa protein from porcine spinal cord; there are differences of 8 residues which could be species-specific amino acid substitutions. This indicates that the micro glutamic acid-rich protein may arise by a restricted proteolysis of the neurofilament 68-kDa protein, with the break occuring toward the C-terminus.     

Gene identification and molecular characterization of solvent stable protease from a moderately haloalkaliphilic bacterium, Geomicrobium sp. EMB2

Cloning and characterization of the gene encoding a solvent-tolerant protease from the haloalkaliphilic bacterium Geomicrobium sp. EMB2 are described. Primers designed based on the N-terminal amino acid sequence of the purified EMB2 protease helped in the amplification of a 1,505-bp open reading frame that had a coding potential of a 42.7-kDa polypeptide. The deduced EMB2 protein contained a 35.4-kDa mature protein of 311 residues, with a high proportion of acidic amino acid residues. Phylogenetic analysis placed the EMB2 gene close to a known serine protease from Bacillus clausii KSM-K16. Primary sequence analysis indicated a hydrophobic inclination of the protein; and the 3D structure modeling elucidated a relatively higher percentage of small (glycine, alanine, and valine) and borderline (serine and threonine) hydrophobic residues on its surface. The structure analysis also highlighted enrichment of acidic residues at the cost of basic residues. The study indicated that solvent and salt stabilities in Geomicrobium sp. protease may be accorded to different structural features; that is, the presence of a number of small hydrophobic amino acid residues on the surface and a higher content of acidic amino acid residues, respectively.     

The amino acid sequence of the enterotoxin from Clostridium perfringens type A

The amino acid sequence of the enterotoxin from Clostridium perfringens type A was determined by analysis of peptides derived from the protein by digestion with trypsin chymotrypsin, thermolysin, pepsin, a lysine-specific protease. S. aureus V8 protease and a proline-specific protease, and fragments generated by cleavage with cyanogen bromide or by dilute acetic acid in 7 M guanidine HCl. The sequence which is complete except for the definite order of 3 small peptides between residues 88 and 103 consists of 309 amino acids and contains a correction to our preliminary announcement [(1984) FEMS Symp. 24, 329-330].     

Full-length cDNA of human cathepsin F predicts the presence of a cystatin domain at the N-terminus of the cysteine protease zymogen

A novel human cDNA encoding a cysteine protease of the papain family named cathepsin F is reported. The mature part of the predicted protease precursor displays between 26% and 42% identity to other human cysteine proteases while the proregion is unique by means of length and sequence. The very long proregion of the cathepsin F precursor (251 amino acid residues) can be divided into three regions: a C-terminal domain similar to the pro-segment of cathepsin L-like enzymes, a 50 residue flexible linker peptide, and an N-terminal domain predicted to adopt a cystatin-like fold. Cathepsin F would therefore be the first cysteine protease zymogen containing a cystatin-like domain.     

Sequence analysis of peptide fragments from the intrinsic membrane protein of calf lens fibers MP26 and its natural maturation product MP22

Calf lens fiber plasma membranes, containing only the intrinsic membrane protein MP26 and its maturation product MP22 were treated with proteolytic enzymes such as trypsin, protease V8 from S. aureus or with chemical agents as CNBr in formic acid. The cleavage products, purified by electrophoresis, were analysed for their amino acid composition and N-terminal sequences. Proteolysis gave rise to peptides which were mainly shortened at the C-terminal end of the molecules. While the V8 protease produced a fragment with a similar N-terminal sequence as the maturation product MP22, trypsin yielded another cleavage product. Chemical hydrolysis yielded large fragments (11-15 kDa) with hydrophobic N-terminal sequences. Our results suggest that MP26 is characterised by an N-terminal signal sequence and possesses other hydrophobic domains which could function as untranslocated insertion sequences.