Conformation of a heptadecapeptide comprising the segment encephalitogenic in rhesus monkey

The 17-residue peptide FKLGGRDSRSGSPMARR derived from myelin basic protein, containing an epitope encephalitogenic in rhesus monkey, has been studied in aqueous solution by high-resolution one- and two-dimensional carbon and proton nuclear magnetic resonance spectroscopy. The resonances of the spectra from both nuclei were assigned with the aid of two-dimensional correlated spectroscopy, pH and solvent titrations, and one-dimensional spin-decoupling techniques and by comparison of the spectra of the heptadecapeptide with those of a phosphorylated form of the peptide, the pentadecapeptide FKLGGRDSRSGSPMA, and the nonapeptide FKLGGRDSR. Amide proton temperature coefficients, coupling constants, 13C- spin-lattice relaxation times, and nuclear Overhauser effect data suggest the existence of three structured regions comprising residues 3-6, 7-12, and 12-14 in the solution conformations of the encephalitogenic heptadecapeptide.     

Localization of the immunity protein-reactive domain in unmodified and chemically modified COOH-terminal peptides of colicin E1.

The region of the colicin E1 polypeptide that interacts with immunity protein has been localized to a 168-residue COOH-terminal peptide. This is the length of a proteolytically generated peptide fragment of colicin E1 against which imm+ function can be demonstrated in osmotically shocked cells. The role of particular amino acids of the COOH-terminal peptide in the expression of the immune phenotype was studied. Chemical modification showed that the two histidine residues (His 427 and His 440) and the single cysteine residue (Cys 505) present in the COOH-terminal peptide were not necessary for the colicin-immunity protein interaction. The immunity protein was localized in the cytoplasmic membrane fraction, consistent with previous work of others on the colicin Ia immunity protein and the prediction from the immunity protein amino acid sequence that it is a hydrophobic protein. The distribution of hydrophobic residues along the immunity polypeptide was calculated.     

Peptide YY. Structure of the precursor and expression in exocrine pancreas

Peptide YY is a 36-residue gastrointestinal hormone which inhibits both pancreatic and gastric secretion. We have isolated a cDNA encoding the peptide YY precursor by screening a rat intestinal lambda gt11 cDNA library with an antiserum directed against the porcine hormone. The nucleotide sequence of the cDNA encodes a 98-residue protein (molecular weight, 11, 121) which has an amino acid sequence identical to that of porcine peptide YY. Rat peptide YY is preceded immediately by a signal sequence and followed by a cleavage-amidation sequence Gly-Lys-Arg plus 31 additional amino acids. Thus the peptide YY precursor is similar in structure to that of two related peptides, pancreatic polypeptide and neuropeptide Y. RNA blot hybridizations reveal that the peptide YY gene is much more actively expressed in pancreas than previously realized. In situ hybridizations localized peptide YY cells exclusively to the exocrine pancreas. The abundance of peptide YY in one of its target organs, the pancreas, suggests a paracrine mechanism for peptide YY in regulating pancreatic enzyme secretion.     

A novel method for producing anti-peptide antibodies. Production of site-specific antibodies to the T cell antigen receptor beta-chain

Peptide antigens used to generate site-specific antibodies to proteins are of interest in the development of vaccines. The need to conjugate them to a carrier protein for optimal immunogenicity results in a number of problems including a possible immune response to the carrier. Here we describe a new method of synthesizing an immunogenic peptide antigen, referred to as multiple antigenic peptide (MAP), which may render the need for a carrier protein obsolete. A 14-residue sequence derived from the human T cell antigen receptor beta-chain constant region was selected, and the peptide was synthesized directly onto a branching lysine core with 8 copies of the 14-residue peptide linked to the core by the COOH-terminal amino acid. The molecular weight of this structure was 13,422 of which only 7% represents the lysine residues of the core. The octameric MAP was highly immunogenic in mice and rabbits, allowing production of polyclonal and monoclonal antibodies. The majority of these antibodies reacted with the peptide in its monomeric form as well as its octameric form. Moreover, the antibodies reacted with the intact beta-chain protein. The antigenic determinants of the peptide that were recognized by the antibodies included continuous determinants and conformational determinants. The NH2-terminal residues of the octameric MAP appeared to be most immunogenic. There were no antibodies to the central lysine core. This method of direct synthesis of a polymeric peptide provides accurate knowledge of the conformation and quantity of the peptide prior to immunization, which is usually not the case when peptides are conjugated to carriers. The method is versatile because the possibility exists to synthesize MAP with 16 or 32 peptide arms or to synthesize polymers containing two different peptides.     

Suppression and reversal of allergic encephalomyelitis in guinea pigs with a non-encephalitogenic analogue of the tryptophan region of the myelin basic protein.

The administration of synthetic peptide S42 leads to suppression and reversal of experimental allergic encephalomyelitis (EAE) induced in guinea pigs by myelin basic protein. Peptide S42 contains a linear sequence of 21 amino acid residues, H-Phe-Ser-Trp-Gln-Lys-Phe-Ser-Trp-Gln-Lys-Phe-Ser-Trp-Gln-Lys-Phe-Ser-Trp-Gln-Lys-Gly-OH, made up of four repeating unit sequences of H-Phe-Ser-Trp-Gln-Lys-OH in addition to a C-terminal glycine. Injected at relatively high doses, peptide S42 is non-encephalitogenic. It induces delayed-type hypersensitivity which is not followed by EAE, and elicits delayed-type hypersensitivity responses in peptide S42, encephalitogenic trytophan peptide, or BP-challenged animals for either of the three antigens. The repeating unit sequence of peptide S42 is analogous to the encephalitogenic tryptophan region of the BP molecules . The sequence homology is responsible for cellular recognition of this antigen by the skin test assay and suggests in vivo interaction between peptide S42 and EAE-inducing cells leading to suppression and reversal of disease.     

Immunochemical specificity of antisera raised against the synthetic encephalitogenic peptide SH624, residues 59–74 of the myelin basic protein

Synthetic peptide SH624 (SHHPARTAHYGSLPQK), residues 59–74 of human myelin basic protein (MBP) was found to be encephalitogenic in the rabbit. Four antisera raised, against the peptide were employed in a liquid-phase equilibrium competitive radioimmunoassay with a series of synthetic peptide analogs of the region to probe the structural requirements of the B-cell determinant subsumed within SH624. The cross-reactivities of the four antisera with intact MBP were also examined. Immunochemical analyses of the four antisera suggested specificities directed against a conformational determinant dependent upon residues from the more phylogenetically conserved carboxyl C-terminal region, residues 65–74 (TAHYGSLPQK) of the synthetic immunogen. Peptide analogs shorter than SH624 from the C-terminal end showed no cross-reactivity with any of the reagent antisera while analogs shorter from the N-terminal end and including the encephalitogenic sequence TTHYGSLPQK, as well as, HYGSLPQK were reactive under equilibrium competitive conditions. SH624-reactive antibodies, cross-reactive with purified heterologous MBPs from 10 different species were also identified in all four reagent antisera. The results of these experiments support previous investigations demonstrating the accessibility of the encephalitogenic 65–74 region in intact MBP. They also underscore the importance of B-cell recognition of organ specific antigenic determinants with respect to MBP immunology and, in particular, the recognition of autoreactive determinants in the neighborhood of encephalitogenic centers.     

Nicotinamide adenine dinucleotide-specific glutamate dehydrogenase of Neurospora crassa. IX. Isolation and sequences of several large cyanogen bromide peptides.

The isolation and sequences of several large peptides from cyanogen bromide cleavage of the 1030-residue polypeptide chain of the NAD-specific glutamate dehydrogenase of Neurospora crassa are described. One of these is in the 669-residue sequence of the COOH-terminal end of the protein. The remaining peptides have been aligned in two partial sequences in the NH2-terminal portion of the polypeptide chain.     

Fibrinopeptides A and B of Japanese monkey (Macaca fuscata) and patas monkey (Erythrocebus patas): their amino acid sequences, restricted mutations, and a molecular phylogeny for macaques, guenons, and baboons

Amino acid sequences of fibrinopeptides A and B from the macaque, Macaca fuscata (Japanese monkey) and the guenon, Erythrocebus patas (patas monkey) were established. Fibrinopeptides A of the monkeys had a sequence identical with those of baboons: Ala-Asp-Thr-Gly-Glu-Gly-Asp-Phe-Leu-Ala-Glu-Gly-Gly-Gly-Val-Arg. Fibrinopeptides B were 9-residue, "short," peptides with the sequences Asn-Glu-Glu-Ser-Leu-Phe-Ser-Gly-Arg for M. fuscata and Asn-Glu-Glu-Val-Leu-Phe-Gly-Gly-Arg for E. patas. The sequence of the B peptide of M. fuscata differed from that of a close-related species, M. mulatta (rhesus monkey), at a single site, Leu (M.f.)----Pro (M.m.). A single replacement between the B peptides of E. patas and Cercocebus aethiops (green monkey), Val (E.p.)----Gly (C.a.), was detected. A phylogenic relationship of macaques, guenons, and baboons, named Cercopithecinae (Old World monkey), was deduced from the sequence data. A selective rather than random amino acid replacement was observed in the B peptides of these Old World monkeys, suggesting a restricted mutation of their fibrinopeptides during primate evolution.     

Ligand-dependent structural changes and limited proteolysis of Escherichia coli phosphofructokinase-2

An isoform (rhesus UGT1A01) orthologus to the human UGT1A1 was cloned and sequenced from female rhesus monkey liver cDNA using primers designed from the human nucleotide sequences. Open reading frame analysis of the PCR-generated product encodes a 533-amino acid protein with a proposed 27-residue signal peptide. Nucleotide sequence comparison of rhesus UGT1A01 to other rhesus UGT1A isoforms detected a single-transition mutation at nucleotide 1520 (T-->C), resulting in a neutral F to S substitution at position 507. Rhesus UGT1A01 was greater than 99 and 95% identical to cynomolgus UGT1A01 and human UGT1A1, respectively. The rhesus UGT1A01 was expressed in HK-293 cells for functional analysis. Catalytic activity of UGT1A01 was determined with 7-hydroxy-4-(trifluoromethyl)-coumarin and more specific human UGT1A1 substrates (1-naphthol, beta-estradiol, 17 alpha-ethinylestradiol, and bilirubin). Expression of UGT1A01 protein was also detected by a Western blot utilizing a polyclonal antibody developed against the human UGT1A family.     

Monkey pepsinogens and pepsins. VII. Analysis of the activation process and determination of the NH2-terminal 60-residue sequence of Japanese monkey progastricsin, and molecular evolution of pepsinogens

Japanese monkey progastricsin was shown to be activated to gastricsin exclusively by a two-step process through an intermediate form. The occurrence of this process was substantiated by the isolation of the intermediate form and released peptides. By NH2-terminal sequence analyses of these protein and peptide species, the amino acid sequence of the 43-residue activation segment (propart) was determined to be as follows: (Formula: see text) The NH2-terminal 26-residue peptide was released first, resulting in generation of the intermediate form. The subsequent release of peptides, residues Nos. 27-40 and 27-43, generated two gastricsins as the final products. This two-step process of activation of Japanese monkey progastricsin is in striking contrast to the one-step activation process occurring exclusively for pepsinogen A of the same monkey species. The course of molecular evolution of pepsinogens including progastricsins was deduced from the amino acid sequences of their activation segments by constructing phylogenic trees. The trees divided pepsinogens into 3 clusters, i.e., pepsinogens A, progastricsins and prochymosin, showing that these three groups diverged from one another very early on in the course of the evolution of pepsinogens.     

Direct expression of recombinant activated human protein C, a serine protease

Human protein C, like other serine proteases, is normally secreted as an inactive zymogen. It is converted to its active form extracellularly by limited proteolysis with the thrombin-thrombomodulin complex. This activation results from the removal of a 12-residue activation peptide from the NH2 terminus of the heavy (COOH-terminal) chain. We report here a successful strategy for the activation of human protein C during post-translational cellular processing, resulting in the secretion of activated protein C from transfected mammalian cells. Deletion of the nucleotides encoding the activation peptide resulted in the expression of a protease with less than 5% of the expected activity. However, the replacement of the activation peptide with an 8-residue sequence (Pro-Arg-Pro-Ser-Arg-Lys-Arg-Arg) involved in the proteolytic processing of the human insulin receptor precursor resulted in the direct expression of fully activated protein C. The mutant protein was shown to be correctly processed by NH2-terminal sequence analysis. This strategy for successful expression of an activated form of protein C may apply to the expression of active forms of other proteases which are naturally expressed as zymogens.     

The complete amino acid sequence of monkey progastricsin

The complete amino acid sequence of progastricsin from the Japanese monkey (Macaca fuscata) was determined. Progastricsin is composed of 374 residues, including the gastricsin moiety of 331 residues and the activation segment of 43 residues. Upon activation under acidic conditions, progastricsin was converted to gastricsin via the intermediate protein species. NH2-terminal sequence determination of these protein species enabled us to deduce the NH2-terminal 78-residue sequence of progastricsin, including the 43-residue activation segment. The complete sequence of the gastricsin moiety was determined using peptide fragments obtained by several chemical and enzymatic cleavages. The molecular weight of progastricsin was determined to be 40,785. As compared with pepsinogen A of the same monkey species, deletion of 4 residues and insertion of 5 residues were observed. Although monkey progastricsin and pepsinogen A have highly homologous sequences around the two active site aspartyl residues, the homology between these proteins is rather small (49% identity). This indicates that progastricsin diverged from pepsinogen A in the early phase of the evolution of gastric aspartyl proteinases.     

Structure and expression of fibulin-2, a novel extracellular matrix protein with multiple EGF-like repeats and consensus motifs for calcium binding

A new protein, fibulin-2, was predicted from sequence analysis of cDNA clones obtained from a mouse fibroblast library. This protein consists of a 1195-residue polypeptide preceded by a 26-residue signal peptide. The COOH-terminal region of 787 amino acids contained three anaphylatoxin-related segments (domain I), 11 EGF-like repeats (domain II), 10 of which had a consensus motif for calcium-binding, and a 115- residue globular domain III. Except for two additional EGF-like repeats, this COOH-terminal region showed 43% sequence identity with the previously described fibulin-1 (BM-90). The NH2-terminal 408 residues, unique to fibulin-2, showed no sequence homology to other known proteins and presumably form two additional domains that differ in their cysteine content. Recombinant fibulin-2 was produced and secreted by human cell clones as a disulfide-bonded trimer. Rotary shadowing visualized the protein as three 40-45 nm long rods which are connected at one end in a globe-like structure. No significant immunological cross-reaction could be detected between fibulin-1 and fibulin-2. Production of the fibulin-2 was demonstrated by Northern blots and radioimmunoassay in fibroblasts but not in several tumor cell lines. Together with the observation that the serum level of fibulin-2 is 1,000-fold lower than that of fibulin-1, the data indicate that these two isoforms are not always coordinately expressed. This is also suggested by Northern blots of tissue mRNAs and by immunofluorescence localizations using mouse tissues. The latter studies also demonstrated an extracellular localization for fibulin-2 in basement membranes and other connective tissue compartments.     

COOH-terminal requirements for the correct processing of a phosphatidylinositol-glycan anchored membrane protein

Placental alkaline phosphatase (PLAP) is anchored to the plasma membrane by a phosphatidylinositol-glycan (PI-G) moiety. During processing of nascent PLAP, a 29-residue COOH-terminal peptide is cleaved out and the PI-G moiety is attached to the newly created COOH terminus of the mature protein. To investigate the structural requirements of the COOH terminus of the nascent protein for PI-G tailing and anchoring to the plasma membrane, we have transfected COS cells with wild type and mutant forms of cDNA encoding human prepro-PLAP. Utilizing a series of COOH-terminal deletion mutants of prepro-PLAP, it was found that to be PI-G-tailed the newly synthesized protein must possess an uncharged, predominantly hydrophobic amino acid sequence of a minimal length in the COOH-terminal peptide. While forms of prepro-PLAP with 17 consecutive hydrophobic residues in the terminal sequence yielded PI-G-tailed and membrane-bound products, prepro-PLAP mutants with 13 or fewer of such residues yielded hydrophilic proteins that were no longer PI-G-tailed but efficiently secreted into the medium. Studies using cassette mutants demonstrated that the precise amino sequence of the COOH-terminal region could be altered as long as minimal hydrophobicity and length was maintained.     

Activation and maturation mechanisms of boar acrosin zymogen based on the deduced primary structure

We have isolated cDNA clones encoding boar acrosin, a serine protease participating in the initial stage of fertilization, from boar testis lambda gt11 cDNA libraries. Nucleotide sequencing of the overlapping clones indicates that the composite cDNA inserts contain 1,391 base pairs coding for a 5'-untranslated region, an open reading frame, a stop codon, a 3'-untranslated region, and a poly(A)+ tail. A polyadenylation signal, AATAAA, is located 33 bases upstream from the start of the poly(A)+ tail. The amino acid sequence deduced from the cDNAs shows that boar acrosin is initially synthesized as a prepro-protein with a 16-residue signal peptide at the NH2 terminus. This signal sequence is followed by a 399-residue sequence corresponding to the acrosin zymogen. COOH-terminal sequence analysis of boar sperm 55-kDa proacrosin and its processed forms indicates that the mature acrosin molecule contains 322 amino acid residues in two polypeptide chains, a 23-residue light chain and a 299-residue heavy chain, with a combined molecular mass of 35,735 Da, and that the 55-kDa proacrosin molecule has 14-, 18-, and 43-residue segments as COOH-terminal extensions that are removed during proacrosin maturation. The COOH-terminal 43-residue segment is rich in proline residues, including an unusual repeat of 23 consecutive prolines. The deduced amino acid sequence of boar acrosin shows a high degree of identity with major portions of other serine proteases, including the active site region and the location of cysteine residues. We conclude that boar acrosin is synthesized as a single-chain polypeptide with the regions corresponding to the light and heavy chains covalently connected by two disulfide bonds, and that the single-chain molecule is autoactivated by cleavage of the Arg23-Val24 bond after removal of the COOH-terminal 14-residue segment, resulting in the formation of the light and heavy chains. This two-chain molecule is then converted to the mature enzyme by removal of the COOH-terminal 18- and 43-residue segments.     

Structure of the tetragonal surface virulence array protein and gene of Aeromonas salmonicida

The paracrystalline surface protein array of the pathogenic bacterium Aeromonas salmonicida is a primary virulence factor with novel binding capabilities. The species-specific structural gene (vapA) for this array protein (A-protein) was cloned into lambda gt11 but was unstable when expressed in Escherichia coli, undergoing an 816-base pair deletion due to a 21-base pair direct repeat within the gene. However, the gene was stable in cosmid pLA2917 as long as expression was poor. A-protein was located in the cytoplasmic, inner membrane and periplasmic fractions in E. coli. The DNA sequence revealed a 1,506-base pair open reading frame encoding a protein consisting of a 21-amino acid signal peptide, and a 481-residue 50,778 molecular weight protein containing considerable secondary structure. When assembled into a paracrystalline protein array on Aeromonas the cell surface A-protein was totally refractile to cleavage by trypsin, but became trypsin sensitive when disassembled. Trypsin cleavage of the isolated protein provided evidence that both the NH2- and COOH-terminal regions form distinct structural domains, consistent with three-dimensional ultrastructural evidence. The NH2-terminal 274-residue domain remained refractile to trypsin activity. This segment connects by a trypsin and CNBr-sensitive 78-residue linker region to a COOH-terminal 129-residue fragment which could apparently refold into a partially trypsin-resistant structure after cleavage at residue 323.     

Novel organization and processing of the guinea pig pancreatic polypeptide precursor

Studies on New World hystricomorph rodents have revealed interesting structural divergences in the peptide hormones of the islets of Langerhans, particularly with respect to insulin and glucagon. Herein we report the isolation and sequencing of a cDNA encoding the precursor of pancreatic polypeptide (PP) from a guinea pig pancreas cDNA library. The 126-residue precursor sequence is predicted to include a 26-residue NH2-terminal signal peptide followed by the 36-amino acid PP hormonal sequence and a large COOH-terminal extension. The sequence identity between guinea pig and human PP is 89% (32/36 residues), and the predicted sequence is in agreement with that reported by Eng et al. (Eng, J., Huang, C.-G., Pan, Y.-C. E., Hulmes, J. D., and Yalow, R. S. (1987) Peptides 8, 165-168). In contrast, the icosapeptide domain in the guinea pig precursor exhibits only 40% (8/20) identity with the corresponding human precursor domain, and the COOH-terminal extension differs greatly in both sequence and size. The guinea pig precursor lacks the monobasic processing site (Pro-Arg) found at the COOH terminus of the icosapeptide domain in human, ovine, canine, and feline proPP. An icosapeptide is thus not likely to be liberated as such from this precursor. Of particular interest in guinea pig proPP is the substitution of serine for arginine at the dibasic amino acid processing site on the COOH-terminal side of the PP domain. Results of radioimmunoassays of gel-filtered protein fractions from a guinea pig pancreas extract indicate that efficient proteolytic cleavage takes place at this Lys-Ser site and that mature guinea pig PP is normally carboxyamidated.     

Determination of the gene sequence and the molecular structure of the enterococcal peptide antibiotic AS-48.

The structural gene of the enterococcal peptide antibiotic AS-48 (as-48) has been identified and cloned by using two degenerate 17-mer DNA oligonucleotides on the basis of the amino acid sequences of two peptides obtained by digestion of the antibiotic with Glu-C endoproteinase. That as-48 gene codes for a 105-amino-acid prepeptide, giving rise to a 70-amino-acid mature protein. Comparative analysis demonstrated that the 16-amino-acid sequence of one of the AS-48 Glu-C peptides, designated V8-5, was composed of a 12-amino-acid sequence corresponding to the C-terminal end sequence (from isoleucine +59 to tryptophan +70 [I+59 to W+70]) of the prepeptide and terminated in four residues forming the N terminus (M+1 to E+4) of a putative AS-48 propeptide. These data, combined with the characteristics of the gene sequence, strongly suggested that the antibiotic peptide was a 70-residue cyclic molecule. We propose that the AS-48 translated primary product is very likely submitted to a posttranslational modification during secretion (i) by an atypical or a typical signal peptidase that cleaves off a 35-residue or shorter signal peptide, respectively, from the prepeptide molecule and (ii) by the linkage of the methionine residue (M+1) to the C-terminal tryptophan residue (W+70) to obtain the cyclic peptide (a tail-head linkage).     

Limited Digestion of Guinea Pig Myelin Basic Protein and Its Carboxy-Terminal Fragment (Residues 89–169) with Staphylococcus aureus V8 Protease

Staphylococcus aureus V8 protease has been reported to have a strict specificity for cleavage of the Glu-X bond in ammonium bicarbonate (pH 7.9). With myelin basic protein and one of its major peptic fragments (residues 89-169) as substrates, selective cleavage of Asp(32)-Thr(33), Asp(37)-Ser(38), and Glu(118-Gly(119) bonds was observed, as well as the unusual cleavage of the Gly(127)-Gly(128) bond. The Asp-Glu and Glu-Asn bonds in the sequence of Gln-Asp-Glu-Asn-Pro(81-84) were resistant to V8 protease attack. The following peptides were identified as products of limited cleavage of basic protein by V8 protease: (1-32), (1-37), (33-169), (38-169), (33-118), (38-118), (33-127), (38-127), (119-169), and (128-169). Cleavage of the peptic peptide (89-169) yielded fragments (89-118), (89-127), (119-169), and (128-169). All peptides were identified by amino acid analysis, as well as NH2- and COOH-terminal analyses. Time course studies with basic protein showed that V8 protease initially attacked the bonds between Asp(32) and Thr(33) and Asp(37) and Ser(38). With peptide (89-169) the initial cleavage was between Glu(118) and Gly(119). Peptides (89-118) and (89-127) were encephalitogenic in the Lewis rat. The activity of these peptides in the rat confirms the presence of a minor encephalitogenic site in guinea pig basic protein. Peptide (89-127) was encephalitogenic in the guinea pig, as expected, because it contains the intact encephalitogenic site. V8 protease digestion of basic protein yields some interesting new fragments, not previously available for biologic studies.     

Experimental allergic encephalomyelitis: sequestered encephalitogenic determinant in the bovine myelin basic protein.

The presence of a sequestered encephalitogenic determinant for Lewis rats in the bovine myelin BP was demonstrated with synthetic peptide sequences prepared in our laboratory by the Merrifield solidphase method. The sequence of the encephalitogenic determinant (residues 75-84) from bovine BP (peptide S6), H-Ala-Gln-Gly-His-Arg-Pro-Gln-Asp-Glu-Asn-OH, is similar but not identical to the sequence reported for the guinea pig BP (peptide S53), H-Ser-Gln-(–)-(–)-Arg-Ser-Gln-Asp-Glu-Asn-OH. The presence or the absence of Gly-His from the sequence of either the bovine or the guinea-pig determinants did not alter their encephalitogenic potencies; however, the presence of Gly-His at positions 77 and 78 together with H-Gly-Ser-Leu-Pro-Gln-Lys- (residues 69-74) at the N-terminal end of the bovine determinant destroyed its encephalitogenic potency. In contrast to the absence of Gly-His from the potent encephalitogenic guinea-pig BP, guinea-pig fragment 44-89, and synthetic peptide S49, its presence in the bovine sequence prevents recognition of this determinant and renders the parent bovine BP, bovine fragment 44-89, and synthetic peptide S8 (residues 69-84) relatively non-encephalitogenic. The results of this study suggest that intramolecular interactions occur between adjacent amino acids, conferring secondary or tertiary structures upon this region of the bovine BP which renders the encephalitogenic determinant inaccessible for recognition by the host animal. The presence of sequestered disease-inducing determinants needs to be considered in choosing a particular BP for therapeutic use.