The 1.55 A resolution structure of Nicotiana alata S(F11)-RNase associated with gametophytic self-incompatibility.

The crystal structure of Nicotiana alata (ornamental tobacco) S(F11)-RNase, an S-allelic glycoprotein associated with gametophytic self-incompatibility, was determined by X-ray diffraction at 1.55 A resolution. The protein has a tertiary structure typical of members of the RNase T(2) family as it consists of a variant of the (alpha+beta) fold and has eight helices and seven strands. A heptasaccharide moiety is also present, and amino acid residues that serve as the catalytic acid and base can be assigned to His32 and His91, respectively. Two "hypervariable" regions, known as HVa and HVb, are the proposed sites of S-allele discrimination during the self-incompatibility reaction, and in the S(F11)-RNase these are well separated from the active site. HVa and HVb are composed of a long, positively charged loop followed by a part of an alpha-helix and short, negatively charged alpha-helix, respectively. The S(F11)-RNase structure shows both regions are readily accessible to the solvent and hence could participate in the process of self/non-self discrimination between the S-RNase and an unknown pollen S-gene product(s) upon pollination.

Crystal structure at 1.5-A resolution of Pyrus pyrifolia pistil ribonuclease responsible for gametophytic self-incompatibility.

The crystal structure of the Pyrus pyrifolia pistil ribonuclease (S(3)-RNase) responsible for gametophytic self-incompatibility was determined at 1.5-A resolution. It consists of eight helices and seven beta-strands, and its folding topology is typical of RNase T(2) family enzymes. Based on a structural comparison of S(3)-RNase with RNase Rh, a fungal RNase T(2) family enzyme, the active site residues of S(3)-RNase assigned were His(33) and His(88) as catalysts and Glu(84) and Lys(87) as stabilizers of an intermediate in the transition state. Moreover, amino acid residues that constitute substrate binding sites of the two RNases could be superimposed geometrically. A hypervariable (HV) region that has an S-allele-specific sequence comprises a long loop and short alpha-helix. This region is far from the active site cleft, exposed on the molecule's surface, and positively charged. Four positively selected (PS) regions, in which the number of nonsynonymous substitutions exceeds that of synonymous ones, are located on either side of the active site cleft, and accessible to solvent. These structural features suggest that the HV or PS regions may interact with a pollen S-gene product(s) to recognize self and non-self pollen.

Crystallization and preliminary X-ray crystallographic analysis of S-allelic glycoprotein S(F11)-RNase from Nicotiana alata.

Nicotiana alata S(F11)-RNase is an S-glycoprotein associated with gametophytic self-incompatibility. Crystals of S(F11)-RNase have been grown at room temperature using polyethylene glycol as a precipitant. A crystal diffracted to better than 1.4 A resolution at 100 K at the SPring-8 synchrotron-radiation source, indicating that it is very suitable for high-resolution structure analysis. The crystal belongs to the space group P2(1), with unit-cell parameters a = 65.86 (11), b = 44.73 (5), c = 64.36 (7) A, beta = 90.27 (9) degrees. The asymmetric unit contains two monomers, giving a crystal volume per protein mass (V(M)) of 2.05 A(3) Da(-1) and a solvent content of 39.6% by volume. A full set of X-ray diffraction data was collected to 1.55 A resolution with a completeness of 97.4%. A heavy-atom derivative has been successfully prepared with ethylmercury thiosalicylate (EMTS) and structure analysis is in progress.

Purification, characterization, and primary structure of a novel cell wall hydrolytic amidase, CwhA, from Achromobacter lyticus.

A novel bacteriolytic enzyme CwhA (cell wall hydrolytic amidase) was purified by ion exchange and gel-filtration chromatographies from a commercial bacteriolytic preparation from Achromobacter lyticus. CwhA exhibited optimal pH at 8.5 and lysed CHCl(3)-treated Escherichia coli more efficiently than Micrococcus luteus, Staphylococcus aureus, Enterococcus faecalis, and Pediococcus acidilactici. The enzyme was inhibited by 1,10-phenanthroline strongly and by EDTA to a lesser extent, suggesting that it is probably a metalloenzyme. Amino acid composition and mass spectrometric analyses for the CwhA-derived M. luteus muropeptides revealed that CwhA is N-acetylmuramoyl-L-alanine amidase [EC 3.5.1. 28]. The complete amino acid sequence of CwhA was established by a combination of Edman degradation and mass spectrometry for peptides obtained by Achromobacter protease I (API) digestion and cyanogen bromide (CNBr) cleavage. The enzyme consists of a single polypeptide chain of 177 amino acid residues with one disulfide bond, Cys114-Cys121. CwhA was found to be homologous to N-acetylmuramoyl-L-alanine amidase from bacteriophage T7 (BPT7). Its sequence identity with BPT7 is 35%, but the amino acid residues functioning as zinc ligands in BPT7 are absent in CwhA. These results suggest that CwhA is a new type of N-acetylmuramoyl-L-alanine amidase.

Presence of asparagine-linked N-acetylglucosamine and chitobiose in Pyrus pyrifolia S-RNases associated with gametophytic self-incompatibility.

S-RNases encoded by the S-locus of rosaceous and solanaceous plants discriminate between the S-alleles of pollen in gametophytic self-incompatibility reactions, but it is not clear how. We report the structures of N-glycans attached to each of the N-glycosylation sites of seven S-RNases in Pyrus pyrifolia of the Rosaceae. The structures were identified by chromatographic analysis of pyridylaminated sugar chains prepared from S4-RNase and by liquid chromatography/electrospray ionization-mass spectrometric analysis of the protease digests of reduced and S-carboxymethylated S-RNases. S4-RNase carries various types of sugar chains, including plant-specific ones with beta1-->2-linked xylose and alpha1-->3-linked fucose residues. More than 70% of the total N-glycans of S4-RNase are, however, an N-acetylglucosamine or a chitobiose (GlcNAcbeta1-->4GlcNAc), which has not been found naturally. The N-acetylglucosamine and chitobiose are mainly present at the N-glycosylation sites within the putative recognition sites of the S-RNase, suggesting that these sugar chains may interact with pollen S-product(s).

Bacteriolytic activity and specificity of Achromobacter beta-lytic protease.

Achromobacter beta-lytic protease (blp), one of the bacteriolytic proteases secreted by Achromobacter lyticus, exhibited both peptidase and bacteriolytic activities at alkaline pH. The protease was strongly inhibited by 1,10-phenanthroline, and one zinc atom was detected in the molecule by ion-spray mass spectrometry. The zinc-protease specifically cleaved Gly-X bonds in peptides and possibly possessed subsites S2, S1, S1', and S2' for binding substrate [Schecter, I. and Berger, A. (1967) Biochem. Biophys. Res. Commun. 27, 157-162]. Blp lysed Staphylococcus aureus and Micrococcus luteus cells more efficiently than Achromobacter alpha-lytic protease (alp) and lysozyme, thus being responsible for the high bacteriolytic activity of A. lyticus. In the lysis of bacterial cell walls, blp hydrolyzed both the D-Ala-Gly/Ala bond at the linkage between the peptide subunit and the interpeptide and the Gly-Gly bond in the interpeptide bridge. These results indicate that blp is a highly active bacteriolytic enzyme with a broad bacteriolytic spectrum, which acts primarily by splitting the linkage between the peptide subunit and the interpeptide in the peptidoglycan.

Primary structural features of rosaceous S-RNases associated with gametophytic self-incompatibility.

We isolated cDNA clones encoding five S-RNases (S1-, S3-, S5-, S6-, S7-RNases) from pistils of Pyrus pyrifolia (Japanese pear), a member of the Rosaceae. Their amino acid sequences were aligned with those of other rosaceous S-RNases sequenced so far. A total of 76 conserved amino acid residues were stretched throughout the sequence, but were absent from the 51-66 region which was designated the hypervariable (HV) region. The phylogenetic tree of rosaceous S-RNases showed that S-RNase polymorphism predated the divergence of Pyrus and Malus. Pairwise comparison of these S-RNases detected two highly homologous pairs, P. pyrifolia S1- and S4-RNases (90.0%) and P. pyrifolia S3- and S5-RNases (95.5%). The positions of amino acid substitutions between S1- and S4-RNases were spread over the entire region, but in the pair of S3- and S5-RNases, amino acid substitutions were found in the 21-90 region including the HV region. The substitutions in this restricted region appear to be sufficient to discriminate between S3 and S5 pollen and to trigger the self-incompatible reaction.

Cloning and sequencing of cDNA encoding ribonuclease F1 from Fusarium moniliforme.

Two cDNA clones encoding ribonuclease F1 (EC 3.1.27.3) have been isolated using a probe prepared by polymerase chain reaction with primers designed on the basis of amino acid sequence of the enzyme. They derived probably from the same gene and contained 393-base pair open reading frame encoding 131 amino acid residues (Mr 13,606) including a putative 25-residue signal peptide. The sequences of 43-base pair 5'-untranslated region and 125-base pair 3'-untranslated region including a poly(A) tail of 25 nucleotides were also elucidated. Homology analyses showed that cDNA for ribonuclease F1 has 65% homology to that for ribonuclease T1 in the coding region. At the preprotein level, they share 53% homology.

Identification of regions in which positive selection may operate in S-RNase of Rosaceae: implication for S-allele-specific recognition sites in S-RNase.

A stylar S-RNase is associated with gametophytic self-incompatibility in the Rosaceae, Solanaceae, and Scrophulariaceae. This S-RNase is responsible for S-allele-specific recognition in the self-incompatible reaction, but how it functions in specific discrimination is not clear. Window analysis of the numbers of synonymous (dS) and non-synonymous (dN) substitutions in rosaceous S-RNases detected four regions with an excess of dN over dS in which positive selection may operate (PS regions). The topology of the secondary structure of the S-RNases predicted by the PHD method is very similar to that of fungal RNase Rh whose tertiary structure is known. When the sequences of S-RNases are aligned with the sequence of RNase Rh based on the predicted secondary structures, the four PS regions correspond to two surface sites on the tertiary structure of RNase Rh. These findings suggest that in S-RNases the PS regions also form two sites and are candidates for the recognition sites for S-allele-specific discrimination.

Purification, staphylolytic activity, and cleavage sites of alpha-lytic protease from Achromobacter lyticus.

Alpha-lytic protease (alp) was purified from a bacteriolytic agent, Achromopeptidase from Achromobacter lyticus M497-1, and has been shown to possess staphylolytic activity. Cleavage sites of this enzyme on the peptidoglycan of Staphylococcus aureus were determined by N-terminal amino acid sequence and amino acid composition analyses. Alp cleaved the N-acetylmuramoyl-L-alanine amide bond, the junction between the polysaccharide and peptide moieties, in addition to the D-Ala-Gly and Gly-Gly peptide bonds, implying that this enzyme recognizes the amino acid of D-configuration at the P1 site and possesses N-acetylmuramoyl-L-alanine amidase activity. However, alp could not cleave the D-Ala-Gly peptide bond in a synthetic peptide, suggesting that this hydrolytic activity of alp is peptidoglycan-specific. The results obtained from different consecutive actions of alp and glycosidase on S. aureus peptidoglycan indicate that the presence of polysaccharide in the peptidoglycan is necessary for the bacteriolytic activity of alp.

A novel D-amino acid-containing peptide, fulyal, coexists with fulicin gene-related peptides in Achatina atria.

Fulicin (Phe-D-Asn-Glu-Phe-Val-NH2) is a neuropeptide from ganglia of the African giant snail (Achatina fulica). Previously, the sequences of nine fulicin gene-related peptides (FGRP-1 to -9) have been predicted from the cDNA encoding the ganglia fulicin precursor and the transcripts have been detectable in the heart. We synthesized twenty peptides related to fulicin and FGRPs containing either an L- or a D-amino acid at position 2 and used them to identify FGRPs in atrial extracts. We identified ten alpha-amidated peptides, including fulicin and confirmed their structures as follows: Tyr-Ala-Glu-Phe-Leu-NH2 (FGRP-9), [D-Ala2]FGRP-9 (fulyal), [L-Asn2]fulicin, fulicin, Ser-Tyr-Asp-Phe-Val-NH2 (FGRP-2), Thr-Tyr-Asp-Phe-Leu-NH2 (FGRP-3), Tyr-Asp-Phe-Ile-NH2 (FGRP-5), Ser-Pro-Tyr-Asp-Phe-Ile-NH2 (FGRP-6), Asn-Tyr-Asp-Phe-Val-NH2 (FGRP-7) and Ser-Pro-Tyr-Asp-Phe-Val-NH2 (FGRP-8). We analyzed the biological activities of synthetic FGRPs using the snail penis retractor muscle. The results revealed that fulyal remarkably potentiated the tetanic contraction at concentrations as low as 10(-12) M. FGRP-9 was about 10,000-fold less potent. Fulyal, like fulicin, seems to undergo preferential maturation to participate in the penis retractor muscle contraction as a neuropeptide containing a D-amino acid.

Location of cysteine and cystine residues in S-ribonucleases associated with gametophytic self-incompatibility.

S-Ribonucleases (S-RNases) that cosegregate with S-alleles in the styles of solanaceous and rosaceous plants are associated with gametophytic self-incompatibility (GSI). The amino acid sequences of many S-RNases have been derived from cDNA sequences, but the state of half-cystines has not been clarified. We report the locations of the two free cysteine residues and four disulfide bridges of tobacco S6-RNase and of the four disulfide bridge of Japanese pear S4-RNase. The protein was first S-pyridylethylated at a low pH to selectively modify the free cysteines without thiol-disulfide exchange. The S-pyridylethylated protein (PE-protein) was digested with Achromobacter protease I (API) at pH 6.5 then analyzed by liquid chromatography/electrospray-ionization mass spectrometry (LC/ESI-MS). This analysis showed that tobacco S6-RNase has two free cysteine residues, Cys77 and Cys95, and four disulfide bonds at Cys16-Cys21, Cys45-Cys94, Cys153-Cys182 and Cys165-Cys176. Similarly, four disulfide bonds were identified for pear S4-RNase, which bears no free cysteine, at Cys15-Cys22, Cys48-Cys92, Cys156-Cys195 and Cys172-Cys183. The eight cysteines forming these four disulfide bonds are conserved in all the known S-RNases, indicative that these cross-links are important in stabilizing the tertiary structures of the self-incompatibility-associated glycoproteins in the solanaceous and rosaceous families. The LC/ ESI-MS analysis also provided some structural informations regarding sugar chains attached to the S-RNases.

Crystallization and preliminary X-ray diffraction analysis of two lysinal derivatives of Achromobacter protease I.

Two crystal forms of lysinal derivatives of Achromobacter protease I have been obtained. The first, modified by benzyloxycarbonyl-Val-lysinal crystallizes in the monoclinic space group P2(1) with unit-cell dimensions of a = 39.6, b = 71.2, c = 45.6 A and beta = 98.4 degrees. The second, modified by benzyloxycarbonyl-Leu-Leu-lysinal crystallizes in the orthorhombic space group I222 (or I2(1)2(1)2(1)) with unit-cell dimensions of a = 98.7, b = 102.2 and c = 55.8 A. The space groups and the unit-cell dimensions of the present two lysinal derivatives are different to those of the protease and TLCK- modified one. The space group of the protease is P1 with cell dimensions a = 39.53, b = 40.34, c = 43.92 A, alpha = 114.81, beta = 113.75 and gamma = 74.00 degrees and that of the TLCK-modified one is also P1 with cell dimensions of a = 37.30, b = 42.74, c = 48.02 A, alpha = 120.10, beta = 112.81 and gamma = 68.54 degrees. Diffraction to 1.9 A resolution for the Val-lysinal modified crystal and to 2.2 A resolution for the Leu-Leu-lysinal modified crystal has been observed using a rotating-anode X-ray generator. Full structure determinations of these lysinal-modified protease crystals may lead to an understanding of the molecular basis of enzyme-substrate interactions in the catalytic process of this protease.

Identification and partial amino acid sequences of seven S-RNases associated with self-incompatibility of Japanese pear, Pyrus pyrifolia Nakai.

S-allele-specific proteins (S-proteins) were separated and identified by two-dimensional (2D) gel electrophoresis from the style extract of 14 cultivars of Japanese pear, Pyrus pyrifolia Nakai, which exhibits gametophytic self-incompatibility. These S-proteins were 30-32 kDa basic proteins with putative pIs of 9.6-10.1 and were distinct from the other proteins, which were common for all cultivars examined. Each S-protein was assigned to a given S-genotype based on electrophoretic mobility and the partial amino acid sequence. For S1- to S7-proteins, five different N-terminal amino acid sequences sharing the YFQFTQQY sequence were determined. Since the same N-terminal amino acid sequences were found for both S1- and S7-proteins, and for S3- and S5-proteins, the two S-proteins of each pair were distinguished based on their electrophoretic behavior. The internal amino acid sequences of S2- and S4-proteins, determined for Achromobacter protease I (API) digests, revealed that these proteins are S2- and S4-RNases, respectively. In the cultivar Nijisseiki, these two RNases were expressed from the white bud to mature flower stages when the cultivar acquires and enforces self-incompatibility. Osa-nijisseiki, a self-compatible mutant of Nijisseiki, produced S2-RNase, but did not produce S4-RNase. The absence of S4-RNase was also observed in self-compatible offsprings derived from Osa-Nijisseiki. These results suggest that Japanese pear in the family Rosaceae possesses a gametophytic self-incompatibility system involving an S-RNase, and that a reduction or lack of expression of S4-RNase in the style is responsible for the self-compatibility of Osa-Nijisseiki.

Molecular cloning and nucleotide sequences of cDNAs encoding S-allele specific stylar RNases in a self-incompatible cultivar and its self-compatible mutant of Japanese pear, Pyrus pyrifolia Nakai.

The genes encoding three RNases were cloned from the style of a self-incompatible cultivar, Nijisseiki (S2S4), and its self-compatible mutant, Osa-Nijisseiki (S2S4sm, sm means stylar part mutant), of Japanese pear. For Nijisseiki, cDNAs coding for two S-RNase (S2-RNase and S4-RNase) and an RNase unrelated to self-incompatibility (non-S-RNase) were cloned from the stylar cDNA library. The cDNAs coding for S2-RNase, S4-RNase, and non-S-RNase include 678-, 684-, and 681-bp open reading frames, respectively. Their deduced amino acid sequences were composed of signal peptides and mature RNases (201-203 residues) which were verified by partial amino acid sequencing. The primary structures of mature proteins revealed that these RNases are of the RNase T2 type; only the two S-RNases have several potential N-glycosylation sites and 60% of their amino acid residues are identical, compared with 25% sequence identity with the non-S-RNase. Such a distinct difference in the primary structures between S-RNases and non-S-RNase has not previously been reported and may be a feature typical of S-RNases in the family Rosaceae. Similar experiments were performed for Osa-Nijisseiki. The cDNAs coding for S2-RNase and non-S-RNase were similarly cloned from the stylar cDNA library. However, the cDNA coding for S4-RNase was neither amplified by PCR nor cloned from the library, suggesting that the mutation of self-incompatible Nijisseiki to self-compatible Osa-Nijisseiki is due to a failure of expression of S4-RNase. These results lead to the idea that Osa-Nijiisseiki is a variant of Nijisseiki in which the S4-allelic gene in the S-locus is exclusively mutated or deleted, causing severely impaired or suppressed expression of its gene product, S4-RNase, at the style.

The nucleotide sequence of human acylamino acid-releasing enzyme.

The nucleotide sequence of a cDNA coding for the human acylamino acid-releasing enzyme (AARE, also known as acylpeptide hydrolase) [EC 3.4.19.1] subunit has been determined. The amino acid sequence of human AARE subunit deduced from its cDNA nucleotide sequence showed a high degree of identity (91.5%) with both the corresponding proteins from the pig and the rat. The AARE cDNA shows 99.2% identity with a 3.3 kb cDNA transcribed from a locus (DNF15S2) on the short arm of human chromosome 3, whose deletion is associated with small cell lung cancer, taking into consideration that the sequence of the 3.3-kb cDNA previously reported was caused by misreading.

Identification of histidine 31 and cysteine 95 in the active site of self-incompatibility associated S6-RNase in Nicotiana alata.

S-RNase is associated with the gametophytic self-incompatibility of flowering plants in Solanaceae and, on the basis of sequence homology, belongs to the RNase T2 family. To identify the active site residues in S-RNase, Nicotiana alata S6-RNase was studied by chemical modification. S6-RNase was inactivated with iodoacetic acid under conditions similar to those used for the inactivation of RNase T2. No inactivation took place in the presence of 2'GMP. Analysis of carboxymethylated S6-RNase revealed that the S-carboxymethylation of Cys95 caused inactivation of the enzyme and that the two histidine residues corresponding to two essential histidine residues of RNase T2 remained intact. Treatment of S6-RNase with diethyl pyrocarbonate (DEPC) resulted in loss of enzyme activity, and the enzyme was protected from inactivation in the presence of 2'GMP. The ethoxycarbonylated residues in DEPC-inactivated S6-RNase were analyzed by mass spectrometry, which also provided structural information on sugar moieties attached to Asn27 and Asn37. His31 was modified with DEPC in the absence of 2'GMP and was not modified in its presence. His31 and His91 are conserved in all members of the RNase T2 family sequenced so far, but Cys95 is not conserved in all known Solanaceae S-RNases. These results suggest that His31, possibly together with His91, corresponding to His115 at the active site of RNase T2, is essential to the function of S6-RNase, but Cys95 is not essential though its S-carboxymethylation causes inactivation.

Complete amino acid sequence of a zinc metalloendoprotease from Streptomyces caespitosus.

We determined the complete amino acid sequence of a zinc metalloendoprotease from Streptomyces caespitosus (ScNP). Peptide fragments obtained by digestion of Rcm-ScNP with trypsin, ScNP and endoproteinase Asp-N were purified by reverse-phase HPLC and their amino acids were analyzed using an automatic sequencer. ScNP consisted of a single polypeptide chain of 132 amino acid residues with one disulfide bond between residues 99 and 112 (M(r) 14376). Thus, the number of amino acid residues determined for this enzyme is much lower than the number of residues previously reported for metalloendoproteases. The amino acid sequence indicated that although ScNP has the zinc-binding motif. His-Glu-Xaa-Xaa-His, which is found at the active site of most zinc metalloendoproteases, it does not share overall significant similarity to the sequences of other zinc metalloendoproteases. Moreover, an analysis of the X-ray structure of ScNP at 0.2-nm resolution (Kirisu et al., unpublished results) revealed that Asp93, together with two histidine residues in the zinc-binding motif (His83 and His87) and a water molecule, is a zinc ligand. We propose that ScNP, which bears the HEXXHXXGXXD motif, represents a novel subfamily of zinc-containing metalloendoproteases.

Complete covalent structure of porcine liver acylamino acid-releasing enzyme and identification of its active site serine residue.

The complete covalent structure of porcine liver acylamino acid-releasing enzyme (AARE) [EC3.4.19.1], which catalyzes the hydrolysis of an N-terminally acylated peptide to release an N-acylamino acid, has been established. On basis of the amino acid sequence deduced from the cDNA sequence of porcine liver AARE [Mitta, M. et al. (1989) J. Biochem. 106, 548-555], sequence determination has been achieved by automated Edman degradation of peptides generated by chemical or enzymatic cleavages of the reduced and S-carboxymethylated protein. Ion-spray mass spectrometry was also successfully used to confirm the amino acid sequences of the peptides determined above and to elucidate both the N-terminal blocking group and the status of half-cystine residues of this protein. The protein consists of 732 amino acid residues, and the N-terminal methionine residue is blocked by an acetyl group. All of 18 half-cystine residues of this protein were proved to exist as cysteine residues. A serine residue reactive with diisopropyl fluorophosphate (DFP) was also identified as Ser587 by preparation of the AARE labeled with tritiated DFP followed by isolation and sequence analysis of a radioactive peptide obtained from its endoproteinase Asp-N digest.

Molecular cloning, nucleotide sequence, and expression of the gene encoding a trypsin-like protease from Streptomyces erythraeus.

Streptomyces erythraeus produces an extracellular mammalian-type serine protease bearing trypsin-like substrate specificity. The gene encoding the protease was cloned and sequenced as an initial step for investigating its structure-function relationship by site-specific mutagenesis. The cloned gene is composed of an 816-bp open reading frame encoding 272 amino acid residues, suggesting that it is synthesized as a precursor protein containing a 42-residue prepropeptide. In the N-terminal prepropeptide portion, the tract of 30 residues from the initiator methionine has a typical signal sequence for Streptomyces and the remaining 12 residues are thought to comprise a propeptide. The cloned gene was replaced downstream of a strong promoter in a high expression plasmid, pSEV2, and expressed in Streptomyces lividans TK24. The gene product was secreted extracellularly and identified as an inactive precursor which consists of the mature enzyme and the 12-residue N-terminally extended peptide chain. The precursor protein was converted to a fully active mature form by limited proteolysis with alpha-chymotrypsin at the Phe-(-1)-Ile-1 bond. Protein sequence analysis revealed that, except for the C-terminal three residues, recombinant SET is identical with the native enzyme.