Primary structure of rat liver 5'-nucleotidase deduced from the cDNA. Presence of the COOH-terminal hydrophobic domain for possible post-translational modification by glycophospholipid

Rat liver 5'-nucleotidase was purified from a crude microsomal fraction, and its molecular mass was estimated to be 73 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The purified protein was subjected to cleavage with CNBr or lysyl endopeptidase, and the resulting 21 peptides as well as the NH2 terminus of the native protein were sequenced by Edman degradation. For further information on the molecular structure, we constructed a lambda gt11 liver cDNA library and isolated two cDNA clones for 5'-nucleotidase, lambda cNTP6 and lambda cNT34. The 3.2-kilobase cDNA insert of lambda cNTP6 contains an open reading frame that encodes a 576-residue polypeptide with a calculated size of 63,965 Da, which is in reasonable agreement with that of 5'-nucleotidase (62 kDa) immunoprecipitated from cell-free translation products. The NH2-terminal 28 residues comprise a signal peptide, which is followed by the NH2-terminal sequence of the purified protein. The predicted structure contains all the other peptide sequences determined by Edman degradation. Five potential N-linked glycosylation sites are found in the molecule, accounting for the difference in mass between the precursor and mature forms. Another characteristic feature is that the primary structure contains a highly hydrophobic amino acid sequence at the COOH terminus, a possible signal for the post-translational modification by glycophospholipid. In fact, labeling experiments of rat hepatocytes demonstrated that 3H-labeled compounds such as ethanolamine, myo-inositol, and palmitic acid, components of the glycolipid anchor, were incorporated into 5'-nucleotidase. Phosphatidylinositol-specific phospholipase C released 5'-nucleotidase from the cell surface, and the released protein no longer contained the radioactivity of [3H]palmitic acid incorporated.     

Functional expression of dipeptidyl peptidase I (Cathepsin C) of the oriental blood fluke Schistosoma japonicum in Trichoplusia ni insect cells

Proenzyme dipeptidyl peptidase I (DPP I) of Schistosoma japonicum was expressed in a baculovirus expression system utilizing Trichoplusia ni BTI-5B1-4 (High Five) strain host insect cells. The recombinant enzyme was purified from cell culture supernatants by affinity chromatography on nickel-nitriloacetic acid resin, exploiting a polyhistidine tag fused to the COOH-terminus of the recombinant protease. The purified recombinant enzyme resolved in reducing SDS-PAGE gels as three forms, of 55, 39, and 38 kDa, all of which were reactive with antiserum raised against bacterially expressed S. japonicum DPP I. NH(2)-terminal sequence analysis of the 55-kDa polypeptide revealed that it corresponded to residues -180 to -175, NH(2)-SRXKXK, of the proregion peptide of S. japonicum DPP I. The 39- and 38-kDa polypeptides shared the NH(2)-terminal sequence, LDXNQLY, corresponding to residues -73 to -67 of the proregion peptide and thus were generated by removal of 126 residues from the NH(2)-terminus of the proenzyme. Following activation for 24 h at pH 7.0, 37 degrees C under reducing conditions, the recombinant enzyme exhibited exopeptidase activity against synthetic peptidyl substrates diagnostic of DPP I. Specificity constants (k(cat)/K(m)) for the recombinant protease for the substrates H-Gly-Arg-NHMec and H-Gly-Phe-NHMec were found to be 14.4 and 10.7 mM(-)1 s(-1), respectively, at pH 7.0. Approximately 1 mg of affinity-purified schistosome DPP I was obtained per liter of insect cell culture supernatant, representing approximately 2 x 10(9) High Five cells.     

Primary structure of human placental 5'-nucleotidase and identification of the glycolipid anchor in the mature form

A cDNA was cloned coding for human placental 5'-nucleotidase. The 3547-bp cDNA contains an open reading frame that encodes a 574-residue polypeptide with calculated size of 63 375 Da. The NH2-terminal 26 residues comprise a signal peptide, which is followed by the NH2-terminal sequence of the purified protein. four potential N-linked glycosylation sites are found in the molecule, accounting for a larger mass of the mature form (71 kDa). The predicted structure contains a hydrophobic amino acid sequence at the COOH terminus, a possible signal for the post-translational modification by glycophospholipid. To confirm this possibility, we tried to isolate and characterize the membrane-anchoring domain of 5'-nucleotidase. BrCN-cleaved fragments of the protein were extracted with hexane and subjected to HPLC, resulting in purification of a single component of 2.3 kDa. Chemical analyses revealed that the purified fragment contains the tetradecapeptide Lys-Val-Ile-Tyr-Pro-Ala-Val-Glu-Gly-Arg-Ile-Lys-Phe-Ser, ethanolamine, glucosamine, mannose, inositol, palmitic acid, and stearic acid. The peptide sequence determined is identified at positions 510-523 in the primary structure deduced from the cDNA sequence, which predicts a further extension to position 548, containing the hydrophobic amino acid sequence. Thus, it is concluded that the mature 5'-nucleotidase lacks the predicted COOH-terminal peptide extension (524-548), which has been replaced by the glycophospholipid functioning as the membrane anchor of 5'-nucleotidase.     

Primary structure of sorghum malate dehydrogenase (NADP) deduced from cDNA sequence. Homology with malate dehydrogenase (NAD)

Malate dehydrogenase (NADP) (NADP-MDH) is an important enzyme of the photosynthetic CO2 fixation pathway of C4 plants. We have isolated two clones from a sorghum lambda gt11 cDNA library (CM3, 932 bp, and CM7, 1441 bp). Nucleotide sequence analysis of the cDNAs CM3 and CM7 showed the existence of two NADP-MDH mRNA species encoding different enzyme subunits. Microsequencing of the N-terminus of the mature protein indicated that a specific cleavage of 13 amino acids occurred during the purification steps of the enzyme. The full-length cDNA CM7 contains a large open reading frame encoding an NH2-terminal transit peptide of 40 amino acids and a mature protein of 389 amino acids (42.207 kDa). Alignment of the NADP-MDH sequence with those of several malate dehydrogenases revealed some similarities with NAD-MDHs.     

Isolation and nucleotide sequence of a cDNA clone encoding rat mitochondrial malate dehydrogenase.

We have determined the complete sequence of the rat mitochondrial malate dehydrogenase (mMDH) precursor derived from nucleotide sequence of the cDNA. A single synthetic oligodeoxynucleotide probe was used to screen a rat atrial cDNA library constructed in lambda gt10. A 1.2 kb full-length cDNA clone provided the first complete amino acid sequence of pre-mMDH. The 1014 nucleotide-long open reading frame encodes the 314 residue long mature mMDH protein and a 24 amino acid NH2-terminal extension which directs mitochondrial import and is cleaved from the precursor after import to generate mature mMDH. The amino acid composition of the transit peptide is polar and basic. The pre-mMDH transit peptide shows marked homology with those of two other enzymes targeted to the rat mitochondrial matrix.     

Correction: Dynamic Changes of Post-Translationally Modified Forms of CXCL10 and Soluble DPP4 in HCV Subjects Receiving Interferon-Free Therapy

Serum levels of the interferon (IFN)-stimulated chemokine CXCL10 are increased during chronic HCV infection and associate with outcome of IFN-based therapy. Elevated levels of NH2-terminal truncated CXCL10 (3-77aa), produced by DPP4 cleavage, negatively associate with spontaneous clearance of acute HCV infection and sustained virological response (SVR) with IFN-based therapy for chronic infection. The association of different CXCL10 forms and DPP4 with outcome during IFN-free HCV therapy has not been examined. Using novel Simoa assays, plasma was analyzed from HCV genotype-1 (GT1) subjects who relapsed (n = 11) or achieved SVR (n = 10) after sofosbuvir and ribavirin (SOF/RBV) treatment, and from SOF/RBV relapsers who achieved SVR with a subsequent SOF/ledipasvir regimen (n = 9). While the NH2-truncated form of CXCL10 was elevated in HCV infection relative to healthy controls, pre-treatment plasma concentrations of CXCL10 forms failed to stratify subjects based on treatment outcome to IFN-free regimens. However, a trend (statistically non-significant) towards elevated higher levels of total and long CXCL10 was observed pre-treatment in subjects who relapsed. All forms of CXCL10 decreased rapidly following treatment initiation and were again elevated in subjects who experienced HCV relapse, indicating that CXCL10 production may be associated with active viral replication. While soluble DPP4 (sDPP4) and NH₂-truncated CXCL10 concentrations were highly correlated, on-treatment sDPP4 levels and activity declined more slowly than CXCL10, suggesting differential regulation. Conclusion: These data suggest post-translationally modified forms of CXCL10 will not support the prediction of treatment outcome in HCV GT1 subjects treated with SOF/RBV.     

Primary structure of beta-galactoside alpha 2,6-sialyltransferase. Conversion of membrane-bound enzyme to soluble forms by cleavage of the NH2-terminal signal anchor

This report describes the primary structure of a rat liver beta-galactoside alpha 2,6-sialyltransferase (EC 2.4.99.1), a Golgi apparatus enzyme involved in the terminal sialylation of N-linked carbohydrate groups of glycoproteins. The complete amino acid sequence was deduced from the nucleotide sequence of cDNA clones of the enzyme. The primary structure suggests that the topology of the enzyme in the Golgi apparatus consists of a short NH2-terminal cytoplasmic domain, a 17-residue hydrophobic sequence which serves as the membrane anchor and signal sequence, and a large lumenal, catalytic domain. NH2-terminal sequence analysis of a truncated form of the enzyme, obtained by purification from tissue homogenates, reveals that it is missing a 63-residue NH2-terminal peptide which includes the membrane binding domain. These and supporting results show that soluble forms of the sialyltransferase can be generated by proteolytic cleavage between the NH2-terminal signal-anchor and the catalytic domain.     

Molecular cloning and nucleotide sequence of the cDNA for human liver glutamate dehydrogenase precursor

Two cDNA clones (lambda GDHh1 and lambda GDHn61) for glutamate dehydrogenase (GDH) were isolated from a human liver cDNA library in lambda gt11. The clone, lambda GDHh1, was isolated from the library using a synthetic 45mer oligodeoxy-ribonucleotide, the sequence of which was derived from the known amino acid sequence near the NH2-terminus of human liver GDH. Subsequently, lambda GDHn61 was isolated from the same library using lambda GDHh1 as a probe. The inserts of both clones contained an overlapping cDNA sequence for human liver GDH, consisting of a 5'-untranslated region of 70 bp, an open reading frame of 1677 bp, a 3'-untranslated region of 1262 bp and a 15 base poly(A) tract. The predicted amino acid sequence revealed that the human liver GDH precursor consisted of a total of 558 amino acid residues including the NH2-terminal presequence of 53 amino acids. The sequence deduced for the mature enzyme showed 94% homology to the previously reported amino acid sequence of human liver GDH.     

Isolation and characterization of a full-length cDNA encoding the 55-kDa rabbit zona pellucida protein.

A full-length cDNA (rc55) encoding the major rabbit zona pellucida (ZP) glycoprotein (55 kDa) has been cloned and sequenced. A lambda gt11 expression library was constructed using poly(A)+ mRNA isolated from sexually immature rabbit ovaries which contain large numbers of developing follicles. The rc55 cDNA was identified using affinity purified polyclonal antibodies specific to ZP antigens which are shared among mammalian species. The deduced amino acid sequence of the full-length rc55 clone was matched to the NH2-terminal 25-amino acid sequence obtained for this protein. The predicted amino acid sequence consists of 540 amino acids including a putative signal peptide of 18-24 residues and six potential N-glycosylation sites. The cDNA hybridizes to a 2000-base species of mRNA from rabbit ovary which is not detected in other rabbit tissues. The message is present early in ovarian follicular development and is approximately 600-fold greater in sexually immature as compared with sexually mature rabbit ovaries. This cDNA was expressed as a cro-beta-galactosidase fusion protein using the pEX expression vector. Antibodies against native rabbit ZP, affinity-purified on the recombinant 55-kDa ZP protein, were found to recognize the native rabbit ZP glycoprotein, indicating partial conservation of native epitopes in the expressed recombinant protein.     

Autocatalytic processing of procathepsin E to cathepsin E and their structural differences

The processing of human gastric procathepsin E to its mature form, cathepsin E, was studied at pH 3.5. The results revealed the autocatalytic and apparently one-step conversion of procathepsin E to cathepsin E within 10 min of incubation at 14 degrees C under the conditions used. Analyses of the amino acid sequences of both procathepsin E and cathepsin E showed that cleavage occurred at the Met36-Ile37 bond to produce the mature form, cathepsin E. The NH2-terminal amino acid sequence of procathepsin E thus determined was identical with that predicted from the cDNA sequence by Azuma et al. except that the NH2-terminal glutamine residue in the latter was converted into a pyroglutamic acid residue in the former and that the glycine residue at position 2 in the latter sequence was deleted in the former. On the other hand, the NH2-terminal amino acid sequence of cathepsin E was identical with that reported previously by us.     

Characterization and cDNA cloning of dipeptidyl peptidase IV from the venom of Gloydius blomhoffi brevicaudus

Dipeptidyl peptidase activity was investigated in snake venoms from Gloydius blomhoffi brevicaudus, Gloydius halys blomhoffii, Trimeresurus flavoviridis and Crotalus atrox. The strongest dipeptidyl peptidase IV (DPP IV) activity was found in venom from G. blomhoffi brevicaudus. The substrate specificity, susceptibility to inhibitors, and pH optimum of the partially purified enzyme were similar to those of known DPP IVs from bacteria and eukaryotes. The G. blomhoffi brevicaudus venom gland cDNA library was screened to isolate cDNA clones using probes based on amino acid sequences highly conserved in known DPP IVs. Two cDNA species encoding DPP IV were obtained, and designated as DPP IVa and DPP IVb. This is the first study to report the primary structure of DPP IV from a reptile. The deduced amino acid sequences for DPP IVa and DPP IVb both consist of 751amino acid residues and are highly homologous to each other. A putative catalytic triad for serine proteases, Ser-616, Asp-694, and His-726, is present. It is of particular interest that the deduced NH(2)-terminal sequence associated with the characteristic signal peptide is identical to that determined from the purified DPP IV. This indicates that the signal peptide of snake venom DPP IV is not cleaved off during biosynthesis, unlike those of other snake venom proteins.     

High and low molecular weight rabbit secretory components. Evidence for the deletion of the second and third domains in the smaller polypeptide

Rabbit secretory components exist in two forms which differ in apparent mass by about 25 kDa. Each of these two forms were reduced, carboxymethylated, and extensively digested with trypsin. The resulting peptides were purified by reverse-phase high performance liquid chromatography and characterized by NH2- and COOH-terminal sequence determination and/or amino acid analysis. They were aligned with the protein sequence predicted from the cDNA nucleotide sequence encoding the rabbit poly(Ig) receptor (Mostov, K. E., Friedlander, M., and Blobel, G. (1984) Nature 308, 37-43). All peptides belonging to the fourth and fifth domains except one (positions 488-496) were accounted for in both forms. In addition, limited tryptic proteolysis of the native low Mr secretory components produced the intact 18-kDa NH2-terminal domain (positions 1-117) and the 30-kDa fragment encompassing the fourth and fifth domains. These results suggest that the smaller polypeptide derives from the larger secretory component form by the deletion of the second and third domains.     

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.     

Molecular cloning and sequencing of the cDNA of rat alpha 1-protease inhibitor and its expression in COS-1 cells

A cDNA clone for alpha 1-protease inhibitor (pc alpha 1P1212) was isolated from a lambda ZAP rat liver cDNA library. The 1.4 kb cDNA insert of pc alpha 1P1212 contained an open reading frame that encodes a 411-residue polypeptide (46,125 Da), in which a signal peptide of 24 residues was identified by comparison with the NH2-terminal sequence of the purified protein. Three potential sites for N-linked glycosylation were found in the molecule, accounting for the difference in molecular mass between the predicted form and the purified protein (56 kDa). The deduced primary structure of rat alpha 1-protease inhibitor showed 68.5% homology to that of the human inhibitor. We then constructed the expression plasmid pSV2 alpha 1PI from pSV2-gpt and pc alpha 1P1212, and transfected it into COS-1 cells. The transfected cells synthesized a molecule which was precipitated with anti-(rat alpha 1-protease inhibitor)-IgG and had the same molecular size as that of the inhibitor produced by rat hepatocytes.     

Cloning and expression of the Gal beta 1, 3GalNAc alpha 2,3-sialyltransferase.

Sequence information obtained by NH2-terminal sequence analysis of two molecular weight forms (45 and 48 kDa) of the porcine Gal beta 1,3GalNAc alpha 2,3-sialyltransferase was used to clone a full-length cDNA of the enzyme. The cDNA sequence revealed an open reading frame coding for 343 amino acids and a putative domain structure consisting of a short NH2-terminal cytoplasmic domain, a signal-anchor sequence, and a large COOH-terminal catalytic domain. This domain structure was confirmed by construction of a recombinant sialyltransferase in which the cytoplasmic domain and signal-anchor sequence of the enzyme was replaced with the cDNA of insulin signal sequence. Expression of the resulting construct in COS-1 cells produced an active sialyltransferase which was secreted into the medium in soluble form. Comparison of the cDNA sequence of the sialyltransferase with GenBank produced no significant homologies except with the previously described Gal beta 1,4GlcNAc alpha 2,6-sialyltransferase. Although the cDNA sequences of these two enzymes were largely nonhomologous, there was a 45-amino acid sequence which exhibited 65% identity. This observation suggests that the two sialyltransferases were derived, in part, from a common gene.     

家蝇防御素基因的cDNA克隆及序列分析

Defensin is a kind of cationic.inducible antimicrobial peptide found in a large range of living organisms that contributes to host defense by disrupting the cytoplasmic membrane of microorganisms.with their broad antimicrobial spectrum and strong pharmaceutical effects.antimicrobial peptides,including defensins,represent a source of novel antibiotic agents.A novel full-length 430 base pairs cDNA of an insect defensin was cloned using polymerase chain reaction (PCR) from the cDnA library of houseflies(Musca domestica) that had been challenged by E.coli and staphylococcus taincd an NH2-terminal signal sequence(1-22)followed by a propeptide and the mature peptide(53-92),The sequence identity with other insect defensin is between 51% and 73%.The mature peptide,with a predicted molecular weight of 4.0kDa,and pI of 8.69,has 1 negative charged amino acid and 4 positice ones,the putative housefly defensin is characterized by 6 invariant cysteine residues forming 3 disulfide bonds,Cys1-Cys4,Cys2-Cys5 and Cys3-Cys6,These results suggest that the novel full-length cDNA of the defensin gene.Denominated Mdde,has been successfully cloned from houseflies.     

Two soluble forms of bovine carboxypeptidase H have different NH2-terminal sequences

Carboxypeptidase H is an important enzyme in the biosynthesis of many peptide hormones. Development of a rapid isolation procedure led to the purification of two soluble forms from acidic extracts of bovine pituitary glands. These two forms differed in apparent molecular size (56 and 53 kDa). Both forms were found in the anterior lobe while only the 53-kDa form was found in posterior lobe. Digestion with N-glycosidase F demonstrated that these two forms are not due to alternative glycosylation of a common polypeptide core. Both forms bind antibodies raised against a COOH-terminal peptide of the full-length protein showing that the difference between them is not due to proteolysis at the COOH terminus. These results also argue against the idea that proteolysis of COOH-terminal domains converts the membrane-associated form of this protein into a soluble form. NH2-terminal sequence analysis demonstrated different NH2 termini. The NH2-terminal sequence of the 56-kDa form begins at the site predicted for signal peptide cleavage. Ion-exchange chromatography resolved the 56-kDa form from the 53-kDa form. The two forms were catalytically active with very similar properties. These results show that bovine carboxypeptidase H can be posttranslationally processed at alternative sites and provide evidence against the idea of a prosequence that must be removed before enzyme activity can be expressed.     

Cloning and sequence analysis of a rat liver cDNA encoding acyl-peptide hydrolase

Acyl-peptide hydrolase catalyzes the removal of an N alpha-acetylated amino acid residue from an N alpha-acetylated peptide. Two overlapping degenerate oligonucleotide probes based on the sequence of a CNBr tryptic peptide, derived from purified rat acyl-peptide hydrolase, were synthesized and used to screen a rat liver lambda gt11 cDNA library. A 2.5-kilobase cDNA was cloned and sequenced. This clone contained 2364 base pairs of rat acyl-peptide hydrolase sequence but lacked a translational initiation codon. Using a 220-base pair probe derived from near the 5'-end of this almost full-length cDNA to rescreen the library, full-length clones were isolated, which contained an in-frame ATG codon at nucleotides 6-8 and encoded the NH2-terminal sequence, Met-Glu-Arg-Gln.... The DNA sequence encoded a protein of 732 amino acid residues, 40% of which were confirmed by protein sequence data from 19 CNBr or CNBr tryptic peptides. The isolated enzyme is NH2-terminally blocked (Kobayashi, K., and Smith, J. A. (1987) J. Biol. Chem. 262, 11435-11445), and based on the NH2-terminal protein sequence deduced from the DNA sequence and the sequence of the most NH2-terminal CNBr peptide, it is likely that the NH2-terminal residue is an acetylated methionine residue, since such residues are frequently juxtaposed to glutamyl residues (Persson, B., Flinta, C., von Heijne, G., and Jornvall, H. (1985) Eur. J. Biochem. 152, 523-527). The RNA blot analysis revealed a single message of 2.7 kilobases in various rat tissues examined. Although this enzyme is known to be inhibited by diisopropyl fluorophosphate and acetylalanine chloromethyl ketone (Kobayashi, K., and Smith, J. A. (1987) J. Biol. Chem. 262, 11435-11445), no strong similarity in protein sequence has been found with other serine proteases. This result suggests that acyl-peptide hydrolase may be a unique serine protease.     

Molecular cloning and nucleotide sequence of cDNA encoding the entire precursor of rat liver medium chain acyl coenzyme A dehydrogenase

cDNA encoding the precursor of rat liver medium chain acyl-CoA dehydrogenase (EC 1.3.99.3) was cloned and sequenced. The longest cDNA insert isolated was 1866 bases in length. This cDNA encodes the entire protein of 421-amino acids including a 25-amino acid leader peptide and a 396-amino acid mature polypeptide. The identity of the medium chain acyl-CoA dehydrogenase clone was confirmed by matching the amino acid sequence predicted from the cDNA to the NH2-terminal and nine internal tryptic peptide sequences derived from pure rat liver medium chain acyl-CoA dehydrogenase. The calculated molecular masses of the precursor medium chain acyl-CoA dehydrogenase, the mature medium chain acyl-CoA dehydrogenase, and the leader peptide are 46,600, 43,700, and 2,900 daltons, respectively. The leader peptide contains five basic amino acids and only one acidic amino acid; thus, it is positively charged, overall. Cysteine residues are unevenly distributed in the mature portion of the protein; five of six are found within the NH2-terminal half of the polypeptide. Comparison of medium chain acyl-CoA dehydrogenase sequence to other flavoproteins and enzymes which act on coenzyme A ester substrates did not lead to unambiguous identification of a possible FAD-binding site nor a coenzyme A-binding domain. The sequencing of other homologous acyl-CoA dehydrogenases will be informative in this regard.     

A collagen-binding protein in the endoplasmic reticulum of myoblasts exhibits relationship with serine protease inhibitors.

Several cDNA clones encoding a 46-kDa collagen-binding glycoprotein (gp46) from rat skeletal myoblasts were isolated and sequenced. The cDNA encoded a 17-amino acid signal peptide and a 400-amino acid mature protein, containing three potential N-linked oligosaccharide attachment sites. The cDNA sequence of gp46 shows 93% identity in the coding region with J6, a retinoic acid-inducible gene coding for a protein of unknown function described from embryonal carcinoma F9 cells. The first 41 NH2-terminal amino acids of the predicted J6 sequence are, however, different from the gp46 sequence as a result of a 7-base pair insertion in the gp46 cDNA. In addition, the NH2-terminal amino acid sequence of hsp47, a collagen-binding protein found in chick embryo fibroblasts, shows 64% identity to gp46 over 36 residues. Interestingly, this alignment begins 10 residues inward from the first amino acid in the mature form of gp46. A significant sequence similarity was observed between gp46 and members of the serine protease inhibitor (serpin) family. Unlike other serpins, however, gp46 is both a heat shock and a collagen-binding protein and is localized to the lumen of the endoplasmic reticulum, as suggested by the presence of the RDEL sequence at the COOH terminus. This sequence is similar to other proposed endoplasmic reticulum retention signals.