Activation of human plasminogen by equimolar levels of streptokinase.

Native Glu-human plasminogen (Mr approximately 92,000 with NH2-terminal glutamic acid) is able to combine directly with streptokinase in an equivalent molar ratio, to yield a stoichiometric complex. The plasminogen moiety in the complex then undergoes streptokinase-induced conformational changes. As a result of such, an active center develops in the plasminogen moiety of the complex. This proteolytically active complex then activates plasminogen in the complex to plasmin and at least two peptide bonds are cleaved in the process. The data presented in this paper reveal that initially an internal peptide bond of plasminogen (in the complex) is cleaved to yield a two-chain, disulfide-linked plasmin molecule. The heavy chain (Mr approximately 67,000 with NH2-terminal glutamic acid) of this plasmin molecule has an identical NH2-terminal amico acid as the native plasminogen. The light chain (Mr approximately 25,000 with NH2-terminal valine) of plasmin is known to be derived from the COOH-terminal portion of the parent plasminogen molecule. A second peptide is then cleaved from the NH2-terminal end of the heavy chain of plasmin producing a proteolytically modified heavy chain (Mr =60.000 with NH2-terminal lysine). This cleavage of the NH2-terminal peptide from the heavy chain of plasmin is shown to be mediated by the dissociated free plasmin present in the activation mixture. Plasmin in the streptokinase-plasmin complex is unable to cleave this NH2-terminal peptide. This same NH2-terminal peptide can also be cleaved from native Glu-plasminogen or from the Glu-plasminogen-streptokinase complex by free plasmin and not by a complex of streptokinase-plasmin. From these studies we conclude (a) in the streptokinase-plasminogen complex, the NH2-terminal peptide need not be released prior to the cleavage of the essential Arg-Val peptide bond which leads to the formation of a two chain plasmin molecule and (b) that this peptide is cleaved from the native plasminogen or from the heavy chain of the initially formed plasmin in the streptokinase complex by free plasmin and not by the plasmin associated with streptokinase. In agreement with this, plasmin associated with streptokinase was unable to cleave the NH2-terminal peptide from the isolated native heavy chain possessing glutamic acid as the NH2-terminal amino acid; whereas free plasmin readily cleaved this peptide from the same isolated Glu-heavy chain.     

The mechanism of activation of rabbit plasminogen by urokinase.

The data presented in this paper show that when rabbit plasminogen is activated to plasmin by urokinase at least two peptide bonds are cleaved in the process. Urokinase first cleaves an internal peptide bond in plasminogen, leading to two-chain disulfide-linked plasmin molecule. The plasmin heavy chain of molecular weight 66,000 to 69,000 possesses an NH2-terminal amino acid sequence identical with the original plasminogen (molecular weight 88,000 to 92,000). The plasmin light chain of molecular weight 24,000 to 26,000 is known to be derived from the COOH-terminal portion of plasminogen. The plasmin generated during the activation of plasminogen is capable, by a feedback process, of cleaving a peptide of molecular weight 6,000 to 8,000 from the NH2 terminus of the heavy chain, producing a proteolytically modified heavy chain of molecular weight 58,000 to 62,000. Plasmin also can cleave this same peptide from the original plasminogen, yielding an altered plasminogen of molecular weight 82,000 to 86,000. This plasmin-altered plasminogen and the plasmin heavy chain derived from it by urokinase activation process NH2-terminal amino acid sequences which are identical with each other and with the plasminolytic product of the original plasmin heavy chain. These studies support a mechanism of activation of plasminogen by urokinase which involves loss of a peptide located on the NH2 terminus of plasminogen. However, these same results show that this NH2-terminal peptide need not be released from rabbit plasminogen prior to the cleavage of the internal peptide bond which leads to the two-chain plasmin molecule. Furthermore, these studies show that urokinase cannot remove this peptide from either the original rabbit plasminogen molecule or from the heavy chain of the initial plasmin formed.     

Amino-acid sequence of the cyanogen-bromide fragment from human plasminogen that forms the linkage between the plasmin chains.

The complete amino acid sequence of a cyanogen bromide fragment (122 residues) obtained from plasminogen is described. This fragment forms the overlap between heavy (A) and light (B) chains of human plasmin. The particular arginyl-valyl bond cleaved in the second step of the activation process is shown to be Arg98-Val99 in this fragment. This site is not very similar to the one in the NH2-terminal part of the molecule (Arg68-Met69). Remarkable homologies with the 'triple loops' ('kringle structures') found in the non-thrombin part of prothrombin are demonstrated. Homologies occurred during evolution of this chain.     

The importance of the preactivation peptide in the two-stage mechanism of human plasminogen activation.

The two stages in the activation of human plasminogen by urokinase have been examined kinetically in order to evaluate the significance of each stage in the activation process. The cleavage of the preactivation peptide from the NH2 terminus of native plasminogen (NH2-terminal glutamic acid) is clearly catalyzed by urokinase and is the rate-limiting first step in activation (Stage 1); this reaction is 20-fold slower than the conversion of the intermediate plasminogen (NH2-terminal lysine) to plasmin (Stage 2). Both lysine and its analogoue, epsilon-aminocaproic acid, exert two effects on the activation of native plasminogen. At low concentrations of these agents, activation is greatly accelerated. Analysis of activation in the presence and absence of these agents by sodium dodecyl sulfate gel electrophoresis indicates that the activation pathway is the same in both cases with the formation of a transient intermediate plasminogen; only the kinetics of proteolysis are altered. This enhancement in the rate of activation results solely from acceleration of the Stage 1 reaction; Stage 2 is essentially unaffected at low concentrations. Stage 1 is maximally enhanced (75-fold) at either 0.0025 M epsilon-aminocaproic acid or 0.025 M lysine and occurs 4 times more rapidly than Stage 2, which becomes the rate-limiting step at these concentrations. Plasmin also cleaves the preactivation peptide from native plasminogen and this reaction rate is enhanced by the same concentrations of lysine and epsilon-aminocaproic acid. These data suggest that lysine and epsilon-aminocaproic acid, which are known to bind to plasminogen and significantly alter its conformation, may thereby enhance preactivation peptide cleavage and consequently, plasminogen activation. At high concentrations, both Stages 1 and 2 are similarly inhibited by these agents, which suggests that this effect may be exerted by the direct inhibition of urokinase. The relative rates of preactivation peptide cleavage by the enzymes urokinase, plasmin, thrombin, and ancrod were also determined. Urokinase is 10 times more effective than plasmin in catalyzing this reaction and 1.8 X 10(4) times more effective than thrombin, while ancrod does not exert an effect. No plasmin is formed by either thrombin or ancrod.     

Primary structure of human fibrinogen and fibrin. Isolation and partial characterization of chains of fragment D.

Fragment D has been isolated as an apparently single molecular weight species (molecular weight about 100,000) from plasmin digests of humman fibrinogen, using a combination of affinity chromatography on insolubilized "fibrin monomer" and gel filtration. This fragment consists of three chains with molecular weights of 15,000 (Dbeta), 42,500 (Dgamma1) or 39,500 (Dgamma2), and 14,000 (Dalpha) held together by disulfide bonds. The S-carboxymethyl derivatives of the chains have been separated by gel filtration and ion exchange chromatography, and their identity has been confirmed by peptide mapping and immunological analysis. The chain with a molecular weight of 45,000 is a fragment of the Bbeta chain of fibrinogen. The chain derived from the gamma chain of fibrinogen occurred in two molecular forms having molecular weight 42,500 and 39,500. The chain derivative with molecular weight 14,000 is most likely derived from the Aalpha chain of fibrinogen. The chains were characterized by NH2-terminal sequence analysis, amino acid composition, and carbohydrate staining. The two molecular analysis, amino acid composition, and carbohydrate staining. The two molecular forms of the gamma chain appeared to be identical except for an NH2-terminal peptide extension of 23 amino acid residues in the longer chain. The latter has sequences in common with the COOH-terminal part of the gamma chain of the NH2-terminal disulfide knot (BROMBACK, B., BRONDAHL, N. J., HESSEL, B., IWANAGA, S., and WALLEN, P. (1973) J. Biol. Chem. 248, 5806-5820); its NH2-terminal residue being Ala-63 of the gamma chain of fibrinogen.     

On the primary structure of human plasminogen and plasmin. Purification and characterization of cyanogen-bromide fragments.

Most of the cyanogen bromide fragments obtained from human plasminogen and plasmin have been purified using combinations of gel filtration and ion-exchange chromatography. The purified fragments have been characterized by molecular weight determination (dodecyl sulphate electrophoresis), amino acid analysis, carbohydrate analysis and direct NH2-terminal amino acid sequence determination. Since some of the purified fragments were compounds with uncompletely cleaved methionyl bonds it was possible to clarify the organization of most of the cyanogen bromide fragments in the plasminogen molecule. The fragment containing the arginyl-valyl bond cleaved during the second step of the activation process is further identified. It is also shown that the microheterogeneity that normally exists in human plasminogen probably has its origin in several sites. One such site is situated in the light (B) chain of plasmin, while another is situated in the carboxyterminal part of the heavy (A) chain. Neither of these sites seems to contain sialic acid.     

Molecular cloning of the high affinity calcium-binding protein (calreticulin) of skeletal muscle sarcoplasmic reticulum

A cDNA clone encoding the high affinity Ca2+-binding protein (HACBP) of rabbit skeletal muscle sarcoplasmic reticulum was isolated and sequenced. The cDNA encoded a protein of 418 amino acids, but a comparison of the deduced amino acid sequence with the NH2-terminal amino acid sequence of the purified protein indicates that a 17-residue NH2-terminal signal sequence was removed during synthesis. This was confirmed by studies of in vitro translation of mRNA encoding the protein. Structural predictions did not reveal any potential transmembrane segments in the protein. The COOH-terminal sequence of the high affinity Ca2+-binding protein, Lys-Asp-Glu-Leu, is the same as that proposed to be an endoplasmic reticulum retention signal (Munro, S., and Pelham, H. R. B. (1987) Cell 48, 899-907). All of these characteristics suggest that the protein is localized in the lumen of the sarcoplasmic reticulum. The mature protein of Mr 46,567 contains 109 acidic and 52 basic amino acids. Structural predictions suggest that the first half of the molecule forms a globular domain of 8 anti-parallel beta-strands with a helix-turn-helix motif at the extreme NH2 terminus. The next one-third of the sequence is proline-rich. This segment can be subdivided into a charged region which contains a 17-amino acid repeat, followed by a proline, serine, and threonine-rich segment extending from Pro-246 to Thr-316. Thirty-seven acidic residues are clustered within 56 amino acids at the COOH terminus of the protein. Although the protein binds 1 mol of Ca2+/mol with high affinity, no "EF-hand" consensus sequence was observed in the protein. The acidic COOH terminus, however, could account for the low affinity, high capacity Ca2+ binding observed in the protein. In agreement with other involved laboratories, we have chosen the name calreticulin for the protein.     

Amino terminal amino acid sequences and carbohydrate of the two major forms of rabbit plasminogen

Two major forms of plasminogen exist in the plasma of many animal species and are distinguished by their affinities for certain antifibrinolytic amino acids. Quantitative end group analysis demonstrated that each isolated form of rabbit plasminogen possessed a single amino terminal residue of glutamic acid. Amino acid sequence analysis indicated that at least the first twelve amino terminal amino acids were identical in the two forms. The unique amino terminal sequence obtained for each form was NH₂-glu-pro-leu-asp-asp-tyr-val-asn-thr-gln-gly-ala-. Analysis of the carbohydrate content of each major plasminogen form revealed some striking differences. The first major form of rabbit plasminogen isolated from affinity chromatography columns contained 1.5–1.7 percent neutral carbohydrate and 3.0–3.3 moles of sialic acid per mole of protein. The second major form of rabbit plasminogen isolated from affinity chromatography columns contained 0.6–0.8 percent neutral carbohydrate and 1.8–2.2 moles of sialic acid per mole of protein.     

Primary structure of rat liver mannan-binding protein deduced from its cDNA sequence

cDNA clones encoding rat liver mannan-binding protein (MBP), a lectin specific for mannose and N-acetylglucosamine, were isolated from a rat liver cDNA library carried in lambda gt 11, by screening with affinity purified polyclonal rabbit anti-rat liver MBP antibodies. The nucleotide sequence of the cDNA determined by the dideoxy method revealed the complete amino acid sequence of the MBP (226 residues). The NH2-terminal residue of the MBP, glutamic acid, was preceded by a predominantly hydrophobic stretch of 18 amino acids, which was assumed to be a signal peptide. Near the NH2-terminal, there was a collagen-like domain, which consisted of 19 repeats of the sequence Gly-X-Y. Here, X and Y were frequently proline and lysine. Three proline and lysine residues were hydroxylated, and one of the latter appeared to link to galactose. Computer analysis of several lectins for sequence homology suggested that the COOH-terminal quarter of the MBP is associated with the calcium binding as well as carbohydrate recognition.     

Synthesis, purification, and properties of a peptide that enhances the activation of human [Glu1]plasminogen by tissue plasminogen activator and retards fibrin polymerization

Solid phase synthesis of the hexapeptide, GPRVVE, which represents the amino terminal six amino acids of the alpha-chain of human fibrin, yielded a product that contained a modified glutamic acid. The nature of the modification was established as the Friedel-Crafts acylation product of the peptide and anisole, the latter reagent employed in the HF deblocking step. The anisoylated peptide selectively enhanced the activation rate of native [Glu1]plasminogen by recombinant tissue plasminogen activator, accelerated clot lysis, and retarded the polymerization of nascent fibrin.     

Synthetic peptide substrates for a tyrosine protein kinase

Immunoprecipitates containing the transforming protein of the avian sarcoma virus, Y73, together with its associated tyrosine-specific protein kinase, have an activity which will phosphorylate the synthetic peptide Lys-Leu-Ile-Glu-Asp-Asn-Glu-Tyr-Thr-Ala-Arg at the tyrosine residue. This peptide corresponds to 10 out of 11 amino acids surrounding the phosphorylated tyrosine in both pp60src and P90, the transforming proteins of Rous sarcoma virus and Y73 virus, respectively. The apparent Km for phosphorylation of the peptide was about 5 mM. A second peptide with the sequence Lys-Leu-Ile-Asp-Asn-Glu-Tyr-Thr-ala-Arg differing from the first peptide only by the absence of the glutamic acid 4 residues from the tyrosine was also phosphorylated, but the apparent Km for the reaction was 40 mM. Several sites of tyrosine phosphorylation in viral transforming proteins have been found to have one or more glutamic acids close to the phosphorylated tyrosine on the NH2-terminal side. Taken together with our in vitro phosphorylation studies, this suggests that the primary sequence surrounding target tyrosines may play a role in recognition of substrates by tyrosine protein kinases, and in particular, that glutamic acid residues on the NH2-terminal side may be important.     

Antipeptide antibodies define the NH2-terminal structure of the pan-specific hemopoietin interleukin 3

Antipeptide antibodies were raised in rabbits against a synthetic peptide including the six NH2-terminal amino acids of the pan-specific hemopoietin interleukin 3 (IL 3). Affinity-purified antibody preparations specific for epitopes determined by residues 1 to 6 were immobilized and used as affinity columns. Up to 98% of IL 3 bioactivity in T cell-conditioned medium was depleted by these columns, as was 71 and 74%, respectively, of IL 3 aberrantly produced by the myeloid leukemias WEHI-274.14 and WEHI-3B. IL 3 produced in vivo in WEHI-3B tumor-bearing mice also bound to the anti-1-6 antibody column, up to 70% of the bioactivity being depleted from ascites fluid and up to 84% from the serum. These results suggest that all IL 3 secreted by T cells and the majority of the IL 3 molecules secreted by myeloid leukemias express epitopes determined by residues 1-6 and cannot have the NH2-terminal amino acid sequence initially reported for IL 3. These six NH2-terminal amino acids share similarities with the NH2-terminal amino acids of several other lymphokines, suggesting an important function for this hexapeptide.     

Studies on the nature of the catalytically essential ionizing group of plasmin with pK 8.4.

A fully carbamylated derivative of plasminogen having no free amino groups has been prepared and converted by urokinase to an active enzyme, called carbamyl plasmin A, with a single free NH2-terminal amino group (Val-561). Carbamyl plasmin A was shown to possess a catalytically essential ionizing group having pK 8.6. Carbamylation of the free NH2-terminal amino group of carbamyl plasmin A led to complete loss of catalytic activity. The results of solvent perturbation studies of normal plasmin (EC 3.4.21.7) indicate that the group with pK 8.4 is a neutral acid group. It is suggested that the catalytically essential ionizing group of plasmin having a pK of 8.4 is the alpha-ammonium group of the NH2-terminal Val-561 or the light chain of plasmin, forming an ion pair with a COO- group of an aspartate or glutamate residue.     

A comparative study on the NH2-terminal amino acid sequences and some other properties of six isozymic forms of human pepsinogens and pepsins

Six pepsinogen isozymogens, including five forms of pepsinogen A (PGA) and an apparently single form of pepsinogen C (PGC), were isolated simultaneously from the purified total pepsinogen fraction of human gastric mucosa by fast protein liquid chromatography on a Mono Q column, and their NH2-terminal amino acid sequences and some other properties were compared. Upon activation at pH 2.0, all the isozymogens were converted to the corresponding pepsins in a stepwise manner through intermediate forms. The activation rates and the cleavage sites in the activation peptide segment to generate intermediate forms were significantly different among the isozymogens. The NH2-terminal 85-residue amino acid sequences of these isozymogens were determined, including the sequences of the activation peptide segments and the NH2-terminal regions of the corresponding pepsins. Differences in amino acid sequence were found at positions 43 and 77 among the pepsinogen A isozymogens; the residue at position 43 was Lys in PGA-5, PGA-4, and PGA-3a, and Glu in PGA-3 and PGA-2, and the residue at position 77 was Leu in PGA-5 and PGA-4 and Val in PGA-3 and PGA-2. Phosphate was not found in any of the isozymogens. The corresponding pepsins also showed significant variations in properties such as specific activities toward synthetic and protein substrates, pH dependence of activity, susceptibility to various inhibitors, and thermal and alkaline stabilities.     

Mutations in the NH2-terminal domain of the signal peptide of preproparathyroid hormone inhibit translocation without affecting interaction with signal recognition particle

The amino-terminal domain of a eukaryotic signal peptide, from bovine parathyroid hormone, was altered by in vitro mutagenesis of the cDNA. The function of "internalized" signal sequence mutants and of deletion mutants was assayed using an in vitro translation-translocation system. The addition of 11 amino acids to the NH2 terminus of the signal peptide did not prevent normal processing of the precursor protein, whereas a 23-amino acid extension blocked processing. These data suggest that the NH2-terminal sequences of internal signal peptides must be permissive of the signal function. Deletion of 6 NH2-terminal amino acids from the signal peptide had no effect on its cleavage by microsomal membranes, but removal of 10 or 13 amino acids, including all charged residues prior to the hydrophobic core, prevented processing. For both the extension and deletion mutations, processed proteins were protected from proteolytic digestion, whereas unprocessed forms were not, which indicated that the unprocessed mutant proteins were not translocated across the microsomal membrane. Translation of both the extension and deletion translocation-deficient mutants was arrested by signal recognition particle, and salt-washed microsomal membranes reversed the translational arrest. These data demonstrate that the NH2-terminal domain is not required for the interaction of signal recognition particle with the signal peptide or with signal recognition particle receptor, but is required for formation of a maximally translocation-competent complex with the microsomal membrane.     

Isolation and sequencing of a cDNA clone encoding rat liver lysosomal cathepsin D and the structure of three forms of mature enzymes

We isolated and sequenced a cDNA clone corresponding to the entire coding sequence of rat liver lysosomal cathepsin D. The deduced amino acid sequence revealed that cathepsin D consists of 407 amino acid residues (Mr 44,608) and the 20 NH2-terminal residues seem to constitute a cleavable signal peptide after which 44 amino acid residues follow as a propeptide. Two putative N-linked glycosylation sites and aspartic acid in the active site are as well conserved as those of human lysosomal cathepsin D. In the NH2-terminal sequence analysis of two isolated heavy chains of the mature enzyme, the termini were assigned as tryptophan (118th residue) and glycine (165th or 166th residue), respectively, hence demonstrates that the two heavy chains derive from a split of the single chain of cathepsin D at position between 117th and 118th or between 164th and 165th or 165th and 166th amino acids. We conclude that cathepsin D in rat liver lysosomes is a mixture of three forms composed of a single and two two-chain forms. However, the amounts of the two two-chain forms are low compared with that of the single chain form. Densidometric determination after SDS-PAGE revealed that the two two-chain forms account for less than 5% of the single chain form. There is a 82% similarity in amino acid level between rat and human liver lysosomal cathepsin D.     

Identification of the actin and plasminogen binding regions of group B streptococcal phosphoglycerate kinase

Phosphoglycerate kinase (PGK), present on the surface of group B streptococcus (GBS), has previously been demonstrated to bind the host proteins actin and plasminogen. The actin and plasminogen binding sites of GBS-PGK were identified using truncated GBS-PGK molecules, followed by peptide mapping. These experiments identified two actin and plasminogen binding sites located between amino acids 126-134 and 204-208 of the 398-amino acid-long GBS-PGK molecule. Substitution of the lysine residues within these regions with alanine resulted in significantly reduced binding to both actin and plasminogen. In addition, conversion of the glutamic acid residue at amino acid 133 to proline, the amino acid found at this position for the PGK protein of Streptococcus pneumoniae, also resulted in sign ificantly reduced binding to actin and plasminogen. These results demonstrate that the lysine residues at amino acid positions 126, 127, 130, 204, and 208 along with the glutamic acid residue at amino acid position 133 are necessary for actin and plasminogen binding by GBS-PGK.     

Opossum (Didelphis virginiana) "little" and "big" gastrins

1. "Little" gastrins from most mammalian species are 17 amino acid peptides and the precursor "big" gastrins are 34 amino acid peptides. 2. "Little" gastrins of the New World hystricomorphs, guinea-pig and chinchilla, are 16 amino acid peptides due to deletion of a glutamic acid in the region 6-9 from their NH2-terminus and the corresponding "big" gastrins are 33 amino acid peptides. 3. Antral gastrins from the opossum, a New World marsupial, have a glutamic acid deletion in the same region as the hystricomorph gastrins. 4. Opossum "big" gastrin is a 33 amino acid peptide with the following sequence: less than ELGPQDLPYLTADLSKKQGPWLEEEEAYGWMDF#.     

NH2-terminal amino acid distribution and amino acid composition of Streptococcus faecalis R soluble and ribosomal proteins

The NH(2)-terminal amino acid distribution of Streptococcus faecalis R soluble and ribosomal proteins isolated from cells at different stages of growth on either folate-sufficient or folate-deficient medium was determined by the dinitrophenyl method. The NH(2)-terminal residues do not follow the random distribution observed for the total amino acid composition of S. faecalis soluble and ribosomal proteins. Methionine and alanine occur most frequently; serine, threonine, aspartic and glutamic acids, and glycine are also present at the NH(2)-terminal position of S. faecalis R proteins. The absence of folic acid yields cells that are incapable of formylating methionyl-transfer ribonucelic acid tRNA(f) (Met), but does not affect either the qualitative or quantitative NH(2)-terminal distribution of total soluble or total ribosomal proteins compared to cells grown with folate. A small quantitative difference was observed in the frequency of distribution of certain amino acids at the NH(2)-termini between log and stationary phase soluble proteins. The amino acid residues found at the NH(2)-terminal position of S. faecalis proteins are qualitatively similar to those reported for several other organisms.     

Amino acid residues 24-31 but not palmitoylation of cysteines 30 and 45 are required for membrane anchoring of glutamic acid decarboxylase, GAD65

The smaller isoform of the GABA synthesizing enzyme glutamic acid decarboxylase, GAD65, is synthesized as a soluble protein that undergoes post-translational modification(s) in the NH2-terminal region to become anchored to the membrane of small synaptic-like microvesicles in pancreatic beta cells, and synaptic vesicles in GABA-ergic neurons. A soluble hydrophilic form, a soluble hydrophobic form, and a hydrophobic firmly membrane-anchored form have been detected in beta cells. A reversible and hydroxylamine sensitive palmitoylation has been shown to distinguish the firmly membrane-anchored form from the soluble yet hydrophobic form, suggesting that palmitoylation of cysteines in the NH2-terminal region is involved in membrane anchoring. In this study we use site-directed mutagenesis to identify the first two cysteines in the NH2-terminal region, Cys 30 and Cys 45, as the sites of palmitoylation of the GAD65 molecule. Mutation of Cys 30 and Cys 45 to Ala results in a loss of palmitoylation but does not significantly alter membrane association of GAD65 in COS-7 cells. Deletion of the first 23 amino acids at the NH2 terminus of the GAD65 30/45A mutant also does not affect the hydrophobicity and membrane anchoring of the GAD65 protein. However, deletion of an additional eight amino acids at the NH2 terminus results in a protein which is hydrophilic and cytosolic. The results suggest that amino acids 24-31 are required for hydrophobic modification and/or targeting of GAD65 to membrane compartments, whereas palmitoylation of Cys 30 and Cys 45 may rather serve to orient or fold the protein at synaptic vesicle membranes.