Carboxyl-terminal Amino Acid Sequences Research Articles

The membrane IgM-associated heterodimer on human B cells is a newly defined B cell antigen that contains the protein product of the mb-1 gene.

7/8embrane IgM (mIgM) on human B lymphocytes is noncovalently associated with a disulfide-linked dimer that contains phosphoproteins of 47 and 37 kDa. In this study, the biochemical properties and the identity of these Ag receptor-associated components have been addressed. Both subunits carry N-linked carbohydrate groups. After deglycosylation, the 47-kDa and 37-kDa proteins have similar molecular masses, of about 23 kDa, and relatively acidic but different isoelectric points. The accumulated data, together with a previously performed comparison of tryptic peptides, suggest that the two components are structurally distinct and possibly encoded by different genes. Indeed, a mAb, raised against a synthetic peptide that was made on the basis of the published carboxyl-terminal amino acid sequence of the human mb-1 gene product, specifically reacted with the 47-kDa but not the 37-kDa subunit. None of the established B cell-specific mAb characterized in the Fourth International Workshop on Leukocyte Antigens, including CD24, CD37, and CD72, detect the mIgM-linked heterodimer, which makes it a newly defined human B cell Ag.

Purification, renaturation, and reconstituted protein kinase activity of the Sendai virus large (L) protein: L protein phosphorylates the NP and P proteins in vitro

Sodium dodecyl sulfate-solubilized Sendai virus large (L) protein was highly purified by a one-step procedure, using hydroxylapatite column chromatography. Monoclonal antibodies addressed to the carboxyl-terminal amino acid sequence of the L protein were used for monitoring L protein during purification. By removing sodium dodecyl sulfate from purified L protein, a protein kinase activity was successfully renatured. P and NP proteins served as its substrates. After immunoprecipitation with anti-L antibodies, the immunocomplex already showed protein kinase activity. In the presence of P protein, the NP protein was more highly phosphorylated. The results show that Sendai virus L protein possesses a protein kinase activity phosphorylating the other proteins of the viral nucleocapsid in vitro.

DNA-Binding and Dimerization Domains of Adenosine 3′, 5′- Cyclic Monophosphate-Responsive Protein CREB Reside in the Carboxyl-Terminal 66 Amino Acids

The expression of genes in response to cAMP is mediated by one or more trans-activator proteins, CREBs, that bind to cAMP-responsive enhancers (CREs) of the general motif 5'-TGACGTCA-3'. The carboxyl-terminal amino acid sequences of two isoforms of CREB, CREB-327 and CREB-341, deduced from the cDNAs consist of a positively charged (basic) region adjacent to a leucine zipper motif. Three peptides corresponding to the hypothetical DNA-binding and dimerization domains of CREB-327 were synthesized. A peptide that includes both the basic and leucine zipper domains binds to the CRE specifically. Moreover, this peptide readily forms CRE-binding heterodimers with full-length CREB both synthesized by in vitro cell-free translation and isolated from PC-12 cells, but did not heterodimerize with in vitro translated jun or fos. Two other peptides, either partially or totally lacking the basic region, but containing the intact leucine zipper domain, readily form dimers but do not bind to the CRE. We conclude that the carboxy-terminal basic and leucine zipper regions are necessary and sufficient for specific binding of CREB to the CRE as a homodimer. The leucine zipper domain is responsible for the dimerization, and the basic region confers binding specificity for the CRE. Heterodimerization of CREB-327 does not form heterodimers with jun or fos.

The ligand binding site of the platelet integrin receptor GPIIb-IIIa is proximal to the second calcium binding domain of its alpha subunit.

The extreme carboxyl-terminal amino acid sequence of the gamma chain of fibrinogen is involved in the binding of this adhesive protein to the platelet integrin glycoprotein (GP) IIb-IIIa, and synthetic peptides corresponding to this region inhibit fibrinogen as well as fibronectin and von Willebrand factor binding to platelets. A chemical cross-linking approach was used to characterize the interaction of a 16-amino acid fibrinogen gamma chain peptide with platelets and to localize the site of its binding to GPIIb-IIIa. This peptide became specifically cross-linked to GPIIb, and platelet stimulation selectively enhanced its cross-linking to this alpha subunit. The cross-linking reaction was specifically inhibited by fibrinogen and an Arg-Gly-Asp peptide but not by an unrelated protein or a substituted peptide. Utilizing a combination of immunochemical mapping, enzymatic and chemical digestions, and amino acid sequencing, the cross-linking site of the gamma chain peptide in GPIIb was localized to a stretch of 21 amino acids. The identified region, GPIIb 294-314, contains the second putative calcium binding domain within GPIIb. The primary structure of this region is highly conserved among alpha subunits of other integrin adhesion receptors. These results identify a discrete region of GPIIb that resides in close proximity to a ligand binding site within GPIIb-IIIa. The homologous region may be involved in the functions of other integrin receptors.

Internal frameshifts within the mitochondrial genes for cytochrome oxidase subunit II and maxicircle unidentified reading frame 3 of Leishmania tarentolae are corrected by RNA editing: evidence for translation of the edited cytochrome oxidase subunit II mRNA.

The Leishmania tarentolae cytochrome oxidase (EC 1.9.3.1) subunit II (COII) and maxicircle unidentified reading frame 3 (MURF3) mRNAs are edited internally by the addition of four and five uridine residues, respectively, which eliminate -1 and +1 reading frameshifts in the gene sequences. The editing events in COII are conserved in three kinetoplastid species, and those in MURF3 are conserved in two species. A primer extension assay showed that the ratio of edited to unedited RNA differed for each gene: 89% of the COII and 36% of the MURF3 transcripts are edited. Preliminary evidence was obtained for translation of the edited COII transcript into protein: antibodies generated against a synthetic peptide with the predicted carboxyl-terminal amino acid sequence reacted with a polypeptide of the correct molecular weight in immunoblots of a mitochondrial lysate.

Drosophila sn-glycerol-3-phosphate dehydrogenase isozymes are generated by alternate pathways of RNA processing resulting in different carboxyl-terminal amino acid sequences.

Glycerol-3-phosphate dehydrogenase (GPDH, Ec 1.1.1.8) in Drosophila melanogaster consists of a family of three isozymes designated as GPDH-1, 2, and 3 which exhibit a unique temporal and tissue-specific pattern of expression. While each isozyme is encoded by the same structural gene, they differ by the amino acid sequence at the COOH-terminal end, with GPDH-3 having the sequence Asn-His-Pro-Glu-His-Met-COOH and with GPDH-1 extended by the three amino acid sequence Glu-Asn-Leu-COOH. We have isolated both genomic and cDNA clones in order to examine the structure of the 3'-end of this gene and its transcriptional products. This analysis has demonstrated three classes of transcripts, each differing in the 3'-untranslated region and coding for an enzyme with a different COOH-terminal amino acid sequence. Each transcript is shown to arise through the differential expression of three isotype-specific exons at the 3'-end of the gene. We propose a model where the expression of each isotype-specific transcript is controlled through a developmentally regulated process of 3'-end formation and alternate splicing pathways of the pre-mRNA. Furthermore, since each transcript and its cognant isozyme is tissue-specific in expression, this model suggests a role for tissue-specific trans-acting factors in these processing events.

Carboxyl terminal domain of Gs alpha specifies coupling of receptors to stimulation of adenylyl cyclase.

The alpha subunits of Gs and Gi link different sets of hormone receptors to stimulation and inhibition, respectively, of adenylyl cyclase. A chimeric alpha i/alpha s cDNA was constructed that encodes a polypeptide composed of the amino terminal 60% of an alpha i chain and the carboxyl terminal 40% of alpha s. The cDNA was introduced via a retroviral vector into S49 cyc- cells, which lack endogenous alpha s. Although less than half of the hybrid alpha chain is derived from alpha s, its ability to mediate beta-adrenoceptor stimulation of adenylyl cyclase matched that of the normal alpha s polypeptide expressed from the same retroviral vector in cyc- cells. This result indicates that carboxyl terminal amino acid sequences of alpha s contain the structural features that are required for specificity of interactions with the effector enzyme, adenylyl cyclase, as well as with the hormone receptor.

Carboxyl-terminal amino acid sequences of two variant forms of the gamma chain of human plasma fibrinogen.

The gamma chain of human fibrinogen has been shown to be heterogeneous; three forms of various sizes are normally present in plasma. To further characterize the two less prevalent elongated variants, we have purified the three forms of the gamma chain, isolated unique tryptic fragments, and sequenced the variant portions of each chain. The intermediate-sized form (gamma 55) has a sequence identical to the longest variant (gamma 57.5) from residue Val-408 to Pro-423, which is the C terminus of the gamma 55 chain. The gamma 57.5 chain extends an additional four amino acids. The C-terminal amino acid sequence and corresponding nucleotide sequence of the gamma 55 chain show marked similarity with an elongated gamma-chain variant of rat fibrinogen. In addition, a remarkable similarity in both amino acid and nucleotide sequence was noted between the human gamma 55 chain and the C-terminal extension in human and bovine A alpha chain of fibrinogen, which are encoded by the cDNA but are posttranslationally cleaved and are not found in plasma. The findings suggest an evolutionary conservation in the C-terminal amino acid sequence found in both the fibrinogen A alpha and gamma chains in several species.

The primary structure of rat ribosomal protein S8.

The amino acid sequence of rat ribosomal protein S8 was deduced from the sequence of nucleotides in a recombinant cDNA and confirmed from the NH2- and carboxyl-terminal amino acid sequences of the protein. Ribosomal protein S8 contains 207 amino acids (the NH2-terminal methionine is removed after translation of the mRNA) and has a molecular weight of 23,928. Hybridization of the cDNA to digests of nuclear DNA suggests that there are 7-9 copies of the S8 gene. Ribosomal protein S8 contains a possible internal repeat that has 12 or 13 residues, is basic, and occurs 5 times in the protein.

Nucleotide sequence of the lig gene and primary structure of DNA ligase of Escherichia coli.

The DNA ligase of Escherichia coli catalyses the NAD-dependent formation of phosphodiester linkages between 5'-phosphoryl and 3'-hydroxyl groups in DNA. It is essential for DNA replication and repair of damaged DNA strands. We determined the nucleotide sequence of the lig gene of Escherichia coli coding for DNA ligase and flanking regions. The coding frame of the gene was confirmed by the amino acid composition and the amino- and carboxyl-terminal amino acid sequences of the purified ligase. The ligase consists of 671 amino acid residues with a molecular weight of 73,690.

Nucleotide sequence and in vivo expression of the ilvY and ilvC genes in Escherichia coli K12. Transcription from divergent overlapping promoters.

The ilvC gene of Escherichia coli K12 encodes acetohydroxy acid isomeroreductase, the second enzyme in the parallel isoleucine-valine biosynthetic pathway. Previous data have shown that transcription of the ilvC gene is induced by the acetohydroxy acid isomeroreductase substrates, acetohydroxybutyrate or acetolactate, and that this substrate induction of ilvC expression is mediated by a positive activator encoded by the ilvY gene. We report here the isolation and complete nucleotide sequence of the ilvY and ilvC genes. The ilvY and ilvC genes encode polypeptides of Mr 33,200 and 54,000, respectively. In vitro transcription-translation of these gene templates results in the synthesis of gene products of these identical molecular weights. The ilvC gene is transcribed in the same direction as the genes of the adjacent ilvGMEDA operon. The ilvY gene is transcribed in a direction opposite to the ilvC and ilvGMEDA genes. The in vivo transcriptional initiation sites of the ilvY and ilvC genes have been determined by S1 nuclease protection experiments. These transcriptional initiation sites are 45 nucleotides apart, and transcription of the ilvY and ilvC genes is initiated via divergent overlapping promoters. The nucleotide sequence of the ilvY and ilvC promoters and 5'-coding regions of Salmonella typhimurium LT2 have been determined. A comparison of these sequences with E. coli K12 suggests regions important in the promotion, regulation, and translation of the ilvY and ilvC genes. A model is presented in which the ilvY-encoded activator binds to an operator site in the overlapping promoter region and reciprocally regulates the transcription of the ilvY and ilvC genes. The carboxyl-terminal amino acid sequence of threonine deaminase encoded by the ilvA gene of the ilv-GMEDA operon of E. coli K12 has been identified by homology with the previously deduced threonine deaminase amino acid sequence encoded by the ilv1 gene of Saccharomyces cerevisiae. Based on the deduced amino acid sequences of the ilvA and ilvY genes, the translational termination codons for both genes are shown to be separated by 52 nucleotides. The proximity of the ilvA and ilvY genes suggests that the 3'-ends of these transcripts overlap.

Two isozymes of chicken muscle acylphosphatase: purification and properties.

Two acylphosphatases, named Ch1 and Ch2, have been purified from chicken skeletal muscle. The molecular weights were determined to be 11,900 and 12,000 for Ch1 and Ch2, respectively, by sedimentation equilibrium. In the amino acid compositions of Ch1 and Ch2, two residues of histidine were contained in Ch2, but none in Ch1, and one residue of cysteine was contained in Ch1, but none in Ch2. There were 11 lysines and 6 arginines in Ch1, whereas there were 6 lysines and 11 arginines in Ch2. In addition, the contents of methionine, serine, glutamic acid and glutamine, and alanine were considerably different between Ch1 and Ch2. There were also differences in the peptide maps and carboxyl-terminal amino acid sequences (-Ser-Thr-Arg-Tyr-COOH for Ch1, and -Phe-Thr-Ile-Arg-Lys-COOH for Ch2). In the double immunodiffusion, Ch2 did not form a precipitin line with the rabbit anti-Ch1 antiserum. These results indicate that Ch1 and Ch2 are different, genetically specified isozymes of acylphosphatase of chicken skeletal muscle. Ch2 is considered to be a new type of acylphosphatase from skeletal muscle.

A “Carboxyl Terminal” Clinical R Adioim Muno Assay for Parathyroid Hormone with Apparent Recognition Preference for the Intact Hormone

A radioimmunoassay for Parathyroid Hormone which is used in a clinical setting was characterized by immunoreactivity with various synthetic fragments of the hormone, serum parathyroid hormone response to oral calcium intake in normocalcemic calcium stone-formers, and ability to detect fragments of parathyroid hormone secreted by abnormal human parathyroid tissue in vitro. Although almost all of the recognition sites for the antiserum were within the 53-84 carboxyl terminal amino acid sequence of the hormone, the radioimmunoassay mainly detected the "intact" hormone rather than the carboxyl terminal fragment(s) which most "carboxyl-terminal" assays of parathyroid hormone are claimed to preferentially detect. Differences in tertiary structure between the intact hormone and its fragments probably account for the relative inability of this antiserum to detect the carboxyl terminal fragment(s).

Developmental variations in the splicing pattern of transcripts from the Drosophila gene encoding myosin alkali light chain result in different carboxyl-terminal amino acid sequences.

The total sequence of the Drosophila melanogaster gene encoding the myosin light chain dissociated by alkali (MLC-ALK) has been determined. By sequence comparisons with an MLC-ALK cDNA clone and by S1 nuclease analyses, the pattern of introns and exons within the gene has been deduced. There are multiple polyadenylylation signals that can account for most of the observed heterogeneity in the lengths of mRNAs. In the 3' half of the gene, there are two alternative splicing patterns which result in mRNAs that translate to give proteins with two alternative 14 amino acid carboxyl-terminal sequences. There is developmental regulation of the selection of the above splicing sites. One splicing pattern produces an mRNA that translates into a protein used for both larval and adult musculature, whereas the other splicing pattern is used for the latter stage only.

The nucleotide binding site detected by affinity labeling in the large T proteins of polyoma and SV40 viruses is distinct from their ATPase catalytic site.

Binding of nucleotides to a specific site of the large T proteins of polyoma and SV40 viruses was demonstrated by covalent affinity labeling with periodate-oxidized [alpha-32P]ATP (oxATP) (Clertant, P., and Cuzin, F. (1982) J. Biol. Chem. 257, 6300-6305). This site appears different from the catalytic site of these proteins for ATP hydrolysis: (i) nucleotide binding and ATPase activities exhibited different ionic requirements and kinetic parameters; (ii) different antibodies directed against the polyoma large T protein either completely inhibited ATPase activity and not affinity labeling, or vice versa; (iii) a truncated form of polyoma large T, with its carboxyl-terminal third deleted, does not bind oxATP but exhibits normal ATPase activity; (iv) conversely, a "super T" SV40 protein, resulting from a duplication within the coding region of large T, was efficiently labeled with oxATP, although it lacks detectable ATPase activity. Cyanogen bromide cleavage after affinity labeling mapped the nucleotide binding site of the polyoma and SV40 large T proteins within a carboxyl-terminal amino acid sequence highly homologous between the two polypeptides. A survey of the phenotypes of the known mutations in these multifunctional proteins suggests that their ATPase and nucleotide-binding activities, although distinct, might both be required to ensure crucial steps in the lytic cycle.

Synthetic peptides corresponding to keratin subunits elicit highly specific antibodies.

The presence of common antigenic determinants within all keratin proteins has made difficult the production of antisera which are monospecific for individual keratin subunits. Synthetic peptides corresponding to the carboxyl-terminal amino acid sequences of the mouse 59- and 67-kilodalton keratins were used to produce antibodies which were highly specific for these keratin subunits. This method of antibody production was chosen after examination of amino acid sequence data (which were deduced from the nucleotide sequence of cDNA clones for these and other mouse keratins) revealed that the carboxyl-terminal amino acid sequences of various keratins were unique. Indirect immunofluorescence staining of newborn-mouse skin with these antisera demonstrated that the 59- and 67-kilodalton keratins were only present within the differentiated cells of the epidermis (the suprabasal layers) and not in the undifferentiated cells (the basal layer). These results are consistent with our previous work concerning the expression of these keratin genes at the messenger RNA level (Roop, D. R., Hawley-Nelson, P., Cheng, C. K., and Yuspa, S. H. (1983) Proc. Natl. Acad. Sci. U.S.A. 80, 716-720). Data obtained with the antisera directed against the 67-kilodalton keratin also indicated that the carboxyl-terminal sequences of this subunit were not present in the nonliving layer (the stratum corneum) of the epidermis. This approach should be useful for the production of antisera specific for other keratin subunits as additional sequence information becomes available.

Isolation of human complex-forming glycoprotein, heterogeneous in charge (protein HC), and its IgA complex from plasma. Physiochemical and immunochemical properties, normal plasma concentration.

Human complex-forming glycoprotein, heterogeneous in charge (protein HC) has previously been isolated from urine and immunochemically shown to be present in low and high molecular weight forms in blood plasma (Tejler, L., and Grubb, A. O. (1976) Biochim. Biophys. Acta 439, 82-94). In the present work, the major low and high molecular weight forms of the protein were isolated from plasma by immunosorption followed by gel chromatography. The plasma low molecular weight protein HC and the urinary protein had similar, if not identical, molecular weight, amino acid composition, NH2-terminal and carboxyl-terminal amino acid sequences and electrophoretic mobility. The low molecular weight plasma protein HC carried a yellow chromophore like the urinary protein, but its molar extinction coefficient at 280 nm was lower and its charge heterogeneity less pronounced than that of urinary protein HC. The plasma high molecular weight protein HC had a hydrodynamic volume which was greater than that of monomeric IgA but smaller than that of dimeric IgA. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the isolated high molecular weight protein followed by electrophoretic blotting and immunochemical analysis demonstrated that the protein contained four polypeptide chains: two light immunoglobulin chains (Mr = 23,000), one IgA alpha-chain (Mr = 54,000), and one chain with Mr approximately 90,000 which carried both alpha-chain and protein HC antigenic determinants. Whether the protein HC X IgA complex is a functionally significant part of the humoral immune system cannot be decided without further experimentation, but the complex was found to be completely absent from the blood plasma of patients with a selective deficiency of IgA-secreting immunocytes. The isolated low and high molecular weight plasma protein HC components were used as standard proteins in the construction of a quantitative crossed immunoelectrophoretic assay for the simultaneous quantitation of the two major protein HC components in blood plasma. The plasma concentrations of the low and high molecular weight protein HC components were measured by this method in 13 healthy Caucasians. The results for the low molecular weight protein HC were: mean, 20.3 mg/liter, S.D., 3.2 mg/liter, range, 13.6-26.0 mg/liter; and for the protein HC X IgA complex: mean, 293 mg/liter, S.D., 176 mg/liter, range, 36-620 mg/liter.

Human protein HC displays variability in its carboxyl-terminal amino acid sequence

1. INTRODUCTION Protein HC (human complex-forming glycoprotein, heterogeneous in charge) is a glycoprotein present in human biological fluids [ 11. It carries yellow- brown chromophores and forms complexes with IgA and albumin to a considerable extent [l]. Although it gives a single band on SDS-polyacryl- amide gel electrophoresis, it shows an extensive charge heterogeneity on isoelectric focussing and on agarose gel electrophoresis which does not decrease after desialylation [ 11. The complete amino acid sequence of protein HC isolated from the urine of an individual was reported in [2]; no evidence for sequence variability of the single polypeptide chain of protein HC was found [3]. This work was under- taken to investigate if the polypeptide chain of protein HC isolated from a pool of urine from several individuals displayed any variability. The polypeptide chain of the protein HC (pool) showed a unique NH2-terminal amino acid sequence up to residue 35, which is identical to the one reported for protein HC (single). However, two different C-ter- minal CNBr fragments were isolated from protein HC (pool). One had the same C-terminal amino acid sequence as reported for protein HC (single) and the other as reported for cut-microglobulin 141, a protein closely related to protein HC. 2. MATERIALS AND METHODS Protein HC was isolated as in [l] from the urine of an individual (J.L.) and from a pool of urine from several individuals. The 2 protein HC preparations are referred to as protein HC (single) and protein HC (pool), respectively. CN7 is the C-terminal CNBr fragment isolated from protein HC (single) as in [3]. Complete reduction and S-carboxy- methylation of the protein preparations was performed in the presence of 6 M guanidinium-HCl [ 11. Hy- drolyses were carried out at 110°C for 20 h using 200 4 5.7 M HCl with 0.02% 2-mercaptoethanol. The analyses were performed using a Beckman 12 l-MB analyzer. Automated amino acid sequence deter- minations by the procedure in [5] were performed on a Beckman 890 C Sequencer using the program in [6]. The phenylthiohydantoin amino acids were identified by high-performance liquid chromato- graphy and in some instances by amino acid analy- sis after back hydrolysis 171. Reduced and [ “C]al- kylated protein HC was dissolved in 70% (v/v) for- mic acid to 20 mg/ml and treated with CNBr with a protein-CNBr ratio of 1:25 (w/w) for 24 h at room temperature [S]. Protein HC (150 pg) or 5 nmol CNBr fragments thereof was used for each of the following carboxypeptidase treatments. Digestion with carboxypeptidase A was carried out in 0.2 M N-methylmorpholine acetate buffer (pH 8.2) for 5 h at 37°C at an enzyme-substrate ratio of 1:lO (w/w) [9]. Digestion with carboxypeptidase B was performed in 0.2 M N-methylmorpholine acetate buffer (pH 8.2) for 16 h at 37°C at an enzyme

Amino- and carboxyl-terminal sequences of proteins coded by gag gene of endogenous baboon and cat type c viruses

Abstract The amino- and carboxyl-terminal amino acid sequences of proteins (p10, p12, p15, and p30) coded by the gag gene of endogenous baboon and eat type C viruses were determined. Among these proteins, p12 from both viruses appears to have a blocked amino terminus. Proline was found to be the common amino end of both p30s and of cat p15, while alanine is the amino terminus of baboon p15 and both p10s. The amino-terminal sequences of the p30s differ at only one of the first 27 residues, while the p15s and p10s are distinctive but show substantial homology. Leucine was found to be the carboxyl-terminal amino acid for p10, p12, and p30 from both viruses, while phenylalanine is the carboxyl-terminus of both p15s. Based on these amino acid sequences and the previously proposed H-p12-p15-p30-p10-C00H polypeptide sequence in the precursor polyprotein, a model (similar to that suggested for mouse type C viruses) for proteolytic cleavage sites involved in the post translational processing of the precursor coded for by the gag vene is Dronosed.

Importance of the carboxyl-terminal four amino acid residues in the inhibitory activity of Streptomyces subtilisin inhibitor (with a revision of its carboxyl-terminal sequence).

The carboxyl-terminal amino acid sequence of Streptomyces subtilisin inhibitor (SSI) was reinvestigated by analysis of the amino acid sequences of the thermolysin peptides from the C-terminal decapeptide, and the sequence -Val110-Ala-Phe-Phe113, which was reported in J. Biochem. 76, 1191-1209 (1974), was revised to -Val110-Phe-Ala-Phe113. Carboxypeptidase A digestion of SSI resulted in loss of the inhibitory activity in parallel with the release of the carboxyl-terminal four amino acid residues. The resulting modified inhibitor, des(Val110-Phe113)-SSI, possessed almost full inhibitory activity against subtilisin BPN' when the inhibitor was incubated with the enzyme in amounts less than one mol of enzyme per mol of the inhibitor. However, no inhibitor activity was observed when the molar ratio of the inhibitor to the enzyme was less than one. This phenomenon suggests that the carboxyl-terminal four amino acid residues might play an important role in the maintenance of the three-dimensional structure of SSI, which resists the action of the proteinase. The addition of more than 30-fold molar excess of SSI-(104-113)-decapeptide (C-terminal decapeptide of SSI) to the modified inhibitor resulted in refolding of the polypeptide chain, rendering it immune from proteolytic digestion.