The amino acid sequence of human insulin-like growth factor I and its structural homology with proinsulin

The complete amino acid sequence of human insulin-like growth factor I (IGF-I), a polypeptide isolated from serum, has been determined. IGF-I is a single chain polypeptide of 70 amino acid residues cross-linked by three disulfide bridges. The calculated molecular weight is 7649. IGF-I displays obvious homology to proinsulin: positions 1 to 29 are homologous to insulin B chain and positions 42 to 62 to insulin A chain. A shortened "connecting" peptide with 12 residues (positions 30 to 41) compared to 30 to 35 in proinsulins shows no homology to proinsulin C peptide. An octapeptide sequence at the COOH-terminal end is also a feature not found in proinsulins. The number of differences in amino acid positions between IGF-I and insulins suggests that duplication of the gene of the common ancestor of proinsulin and IGF occurred before the time of appearance of the vertebrates. Of the 19 residues known to be invariant in all insulins so far sequenced, only glutamine A5 and asparagine A21 are replaced in IGF-I by glutamic acid and alanine, respectively. The fact that all half-cystine and glycine residues and most nonpolar core residues of the insulin monomer are conserved is compatible with a three-dimensional structure of IGF-I similar to that of insulin.     

Isolation and structure elucidation of a novel 5-kDa peptide from neurohaemal lobes of the corpora cardiaca of Locusta migratoria (Insecta, Orthoptera)

Two predominant peptides have been isolated from neurohaemal lobes of corpora cardiaca of 8000 adults of Locusta migratoria. Both peptides have been unambiguously characterized by automated peptide microsequencing and liquid secondary-ion mass spectrometry as a 50-residue peptide (5K peptide) and a 48-residue isologue (5K' peptide). Computer search of sequence data banks did not reveal any significant similarity with other identified proteins. The 5K peptides are remarkably rich in alanine residues (25%) and contain a stretch of five consecutive alanines. This structure suggests that these molecules could correspond to spacer peptides. This assumption is corroborated in the accompanying paper [Lagueux et al. (1990) Eur. J. Biochem. 187, 249-254] on the molecular cloning of the precursor protein which attributes to the 5K peptides a role analogous to that of the C peptides of insulins.     

Isolation and structural characterization of an insulin-related molecule, a predominant neuropeptide from Locusta migratoria

Neurohaemal lobes of corpora cardiaca of Locusta migratoria are an established storage site for neurohormones produced by the neurosecretory cells of the brain. As previously reported [Hietter, H., Van Dorsselaer, A., Green, B., Denoroy, L., Hoffmann, J.A. & Luu, B. (1990) Eur. J. Biochem. 187, 241-247], the isolation and characterization of a novel 5-kDa peptide from these lobes served as the basis for oligonucleotide screening of cDNA libraries prepared from poly(A) RNA from neurosecretory cells of the central nervous system. From subsequent cDNA cloning studies [Lagueux, M., Lwoff, L., Meister, M., Goltzené, F. & Hoffmann, J.A. (1990) Eur. J. Biochem. 187, 249-254], the existence of a 145-residue precursor protein was deduced, which contained, in addition to the 5-kDa peptide, amino-acid sequences with homology to the A and B chains of an insulin-related peptide. In the present study we have isolated the native molecule from corpora cardiaca of Locusta and characterized, by Edman degradation and plasma-desorption mass spectrometry, the two chains as follows: A chain, Gly-Val-Phe-Asp-Glu-Cys-Cys-Arg-Lys-Ser-Cys-Ser-Ile-Ser-Glu-Leu-Gln-Thr- Tyr-Cys - Gly (Ile, isoleucine); B chain, Ser-Gly-Ala-Pro-Gln-Pro-Val-Ala-Arg-Tyr-Cys-Gly-Glu-Lys-Leu-Ser-Asn-Ala- Leu-Lys - Leu-Val-Cys-Arg-Gly-Asn-Tyr-Asn-Thr-Met-Phe. Taken in conjunction with the previous cloning studies, our data lead to a clear picture of the processing of Locusta preproinsulin. They indicate that locusta corpora cardiaca contain remarkably large amounts of one single insulin form, in contrast to multiple insulin isoforms of Bombyx mori, the only other insect species from which insulin-related peptides have been isolated and characterized [Nagasawa, H., Kataoka, H., Isogai, A., Tamura, S., Suzuki, A., Mizoguchi, A., Fujiwara, Y., Suzuki, A., Takahashi, S. & Ishizaki, H. (1986) Proc. Natl Acad. Sci. USA 83, 5840-5843].     

A truncated laminin chain homologous to the B2 chain: structure, spatial expression, and chromosomal assignment

We describe the identification of a novel laminin chain. Overlapping clones were isolated from a human fibrosarcoma HT1080 cell cDNA library spanning a total of 5,200 bp. A second set of clones contained an alternative 3' end sequence giving a total of 4,316 bp. The longer sequence contained an open reading frame for a 1,193-residue-long polypeptide. The alternative sequence was shortened at the carboxyl-terminal end coding for a 1,111-residue-long polypeptide. The amino acid sequence contained 21 amino acids of a putative signal peptide and 1,172 residues or alternatively 1,090 residues of a sequence with five distinct domains homologous to domains I-V in laminin chains. Comparison of the amino acid sequences showed that the novel laminin chain is homologous to the laminin B2 chain. However, the structure of the novel laminin chain isolated here differs significantly from that of the B2 chain in that it has no domain VI and domains V, IV, and III are shorter, resulting in a truncated laminin chain. The alternative sequence had a shortened domain I/II. In accordance with the current nomenclature, the chain characterized here is termed B2t. Calculation of possible chain interactions of laminin chains with the B2t chain domain I/II indicated that the B2t chain can replace the B2 chain in some laminin molecules. The gene for the laminin B2t chain (LAMB2T) was localized to chromosome 1q25-q31 in close proximity to the laminin B2 chain gene. Northern analysis showed that the B2t chain is expressed in several human fetal tissues but differently from the laminin B1 and B2 chains. By in situ hybridization expression of the B2t chain was localized to specific epithelial cells in skin, lung, and kidney as opposed to a general epithelial and endothelial cell expression of the laminin B2 chain in the same tissues.     

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.     

The c-sis gene encodes a precursor of the B chain of platelet-derived growth factor.

The relationship between platelet-derived growth factor (PDGF) and the proto-oncogene c-sis has been determined by amino acid sequence analysis of PDGF and nucleotide sequence analysis of c-sis genomic clones. The nucleotide sequences of five regions of the human c-sis gene which are homologous to sequences of the transforming region (v-sis) of simian sarcoma virus (SSV) were determined. By alignment of the c-sis and v-sis nucleotide sequences the predicted amino acid sequence of a polypeptide homologous to the putative transforming protein p28sis of SSV was deduced. Both predicted sequences use the same termination codon and additional coding sequences may lie 5' to the homologous regions. Amino acid sequence analysis of the PDGF B chain shows identity to the amino acid sequence predicted from the c-sis sequences over 109 amino acid residues. Polymorphism may exist at two amino acid residues. These results suggest that c-sis encodes a polypeptide precursor of the B chain. A partial amino acid sequence of the PDGF A chain is also described. This chain is 60% homologous to the B chain and cannot be encoded by that part of c-sis which has been sequenced but could be encoded by sequences which lie 5' to the five regions of v-sis homology in c-sis, or at a separate locus.     

Isolation and identification of the AKH III precursor-related peptide from Locusta migratoria

We have isolated an 8770Da peptide from extracts of corpora cardiaca of adult male and female Locusta migratoria. The N-terminal amino-acid sequence as partially established by Edman degradation is Ala-Leu-Gly-Ala-Pro-Ala-Ala-Gly-Asp. These nine amino acids correspond to the first nine N-terminal amino acids of the adipokinetic hormone precursor-related peptide gamma-chain (APRP-gamma), a peptide that is predicted from the gene encoding the adipokinetic hormone III precursor. The APRP-gamma chain has a monoisotopic mass of 4387Da and contains two cysteine residues. It is known that both AKH I and AKH II precursors occur as dimers. After processing they give rise to the active hormones and three dimeric (two homodimers and one heterodimer) adipokinetic hormone precursor related peptides (APRPs). Based on the mass of 8770Da and the established N-terminal sequence tag, we conclude that the isolated peptide is a homodimer consisting of two APRP-gamma units, covalently linked to each other by two disulphide bounds. In analogy with the previous identified APRPs (APRP-1, APRP-2, and APRP-3), this APRP will be designated as APRP-4.     

Complete primary structure of a scallop striated muscle myosin heavy chain. Sequence comparison with other heavy chains reveals regions that might be critical for regulation

We have determined the primary structure of the myosin heavy chain (MHC) of the striated adductor muscle of the scallop Aequipecten irradians by cloning and sequencing its cDNA. It is the first heavy chain sequence obtained in a directly Ca(2+)-regulated myosin. The 1938-amino acid sequence has an overall structure similar to other MHCs. The subfragment-1 region of the scallop MHC has a 59-62% sequence identity with sarcomeric and a 52-53% identity with nonsarcomeric (smooth and metazoan nonmuscle) MHCs. The heavy chain component of the regulatory domain (Kwon, H., Goodwin, E. B., Nyitray, L., Berliner, E., O'Neall-Hennessey, E., Melandri, F. D., and Szent-Gy?rgyi, A. G. (1990) Proc. Natl. Acad. Sci. U.S.A. 87, 4771-4775) starts at either Leu-755 or Val-760. Ca(2+)-sensitive Trp residues (Wells, C., Warriner, K. E., and Bagshaw, C. R. (1985) Biochem. J. 231, 31-38) are located near the C-terminal end of this segment (residues 818-827). More detailed sequence comparison with other MHCs reveals that the 50-kDa domain and the N-terminal two-thirds of the 20-kDa domain differ substantially between sarcomeric and nonsarcomeric myosins. In contrast, in the light chain binding region of the regulatory domain (residues 784-844) the scallop sequence shows greater homology with regulated myosins (smooth muscle, nonmuscle, and invertebrate striated muscles) than with unregulated ones (vertebrate skeletal and heart muscles). The N-terminal 25-kDa domain also contains several residues which are preserved only in regulated myosins. These results indicate that certain heavy chain sites might be critical for regulation. The rod has features typical of sarcomeric myosins. It is 52-60% and 30-33% homologous with sarcomeric and nonsarcomeric MHCs, respectively. A Ser-rich tailpiece (residues 1918-1938) is apparently nonhelical.     

Human B lymphocytes define an alternative mechanism of adhesion to fibronectin. The interaction of the alpha 4 beta 1 integrin with the LHGPEILDVPST sequence of the type III connecting segment is sufficient to promote cell attachment

In this report we have studied the mechanism of human B lymphocyte adhesion to fibronectin and to proteolytic fragments of this protein. B cells adhered to fibronectin and to a 38-kDa fragment, derived from the A chain, containing the Hep II domain and most of the type III connecting segment, IIICS, of fibronectin. Cells did not bind to an 80-kDa fragment containing the RGD adhesive sequence of fibronectin. Attachment to fibronectin or to the 38-kDa fragment was not affected by the 80-kDa fragment, the GRGDSPC synthetic peptide, or by a mAb specific for the alpha chain of the RGD-dependent fibronectin receptor, alpha 5 beta 1. However, B cell adhesion to fibronectin was inhibited by the synthetic peptides CS-1, comprising the first 25 amino acids of IIICS and B12, containing the sequence LHGPEILDVPST of CS-1 (residues 14-25). Moreover, this sequence was shown to be sufficient to induce stable cell adhesion when coated on plastic surfaces. A mAb specific for the alpha-subunit of the alpha 4 beta 1 integrin, completely inhibited B cell attachment to B12, CS-1, 38 kDa, and fibronectin coated substrata. These data clearly indicate that adhesion of B lymphocytes to fibronectin is exclusively mediated by the interaction of alpha 4 beta 1 with residues 14-25 of the IIICS region in fibronectin. Therefore this interaction constitutes an alternative pathway of adhesion to fibronectin, independent of RGD and alpha 5 beta 1.     

Cloning and characterization of a novel cDNA sequence encoding the precursor of a novel venom peptide (BmKbpp) related to a bradykinin-potentiating peptide from Chinese scorpion Buthus martensii Karsch

Based on the amino acid sequence of a bradykinin-potentiating peptide (Bpp) (peptide K-12) from scorpion Buthus occitanus, a full-length cDNA sequence encoding the precursor of a novel venom peptide (named BmKbpp) related to this Bpp, has been isolated and analyzed. The cDNA encodes a precursor of 72 amino acid residues, including a signal peptide of 22 residues and an extra Arg-Arg-Arg tail at the C-terminal end of the precursor, which have to be removed in the processing step. The C-terminal region (21 residues) of the precursor is homologous (57% identical) with the sequence of peptide K-12. Thus, according to the primary structure of the BmKbpp precursor, there may be a propeptide between the signal peptide and the putative mature BmKbpp at the C-terminal region of the precursor.     

Characterization of a cDNA clone encoding molluscan insulin-related peptide II of Lymnaea stagnalis

A cDNA clone encoding molluscan insulin-related peptide (MIP) II was isolated from a cDNA library of the central nervous system (CNS) of the freshwater snail, Lymnaea stagnalis, using a heterologous screening with a previously identified MIP cDNA (renamed MIP-I cDNA). The MIP-II cDNA encodes a preprohormone resembling the organization of preproinsulin, with a putative signal sequence, and A and B chains; however, in this case connected by two distinct C peptides, C alpha and C beta, instead of a single C peptide, a phenomenon which represents a new development in the prohormone organization of peptides belonging to the insulin superfamily. The A and B chains of MIP II and I differ remarkably in primary structure; in contrast, the C alpha peptide domains are fully identical. MIP II has only limited sequence similarity with insulins and related peptides. Both MIP II and I exhibit structural features, which make them a unique class of the insulin superfamily. The MIP I and II genes are expressed in a single type of neuron: the growth-controlling neuroendocrine light green cells of the Lymnaea CNS.     

Isolation, cDNA cloning, and characterization of an 18-kDa hemagglutinin and amebocyte aggregation factor from Limulus polyphemus.

An 18-kDa hemagglutinin which possesses the property of inducing both aggregation of amebocytes and agglutination of erythrocytes has been isolated from Limulus polyphemus amebocytes and purified by ion exchange chromatography. This nonglycosylated, single chain polypeptide with an M(r) of 18,506 and isoelectric point of 8.3 is stored exclusively in the large secretory granules of amebocytes. Based on the partial N-terminal amino acid sequence of 63 residues, DNA probes have been synthesized for screening a pBR322 cDNA library constructed from Limulus amebocytes. The cDNA coding for this protein reveals the presence of a 19-residue signal peptide preceding the 153-residue open reading frame. Northern blot analysis indicates the presence of a single mRNA species. The primary structure derived from the corresponding cDNA sequence reveals an internal homology consisting of two consensus sequences, Val-Asn-Asp/Ser-Trp-Asp and Glu-Asp-Arg-Arg-Trp. The formation of 5 disulfide bonds between 10 half-cysteines divides the molecule into three looped domains each containing the Glu-Asp-Arg-Arg-Trp repeating unit. One of the novel features of this protein is that it shares 37% identity with a 22-kDa mammalian extracellular matrix protein isolated from fetal bovine skin (Neame, P.J., Choi, H.U., and Rosenberg, L.C. (1989) J. Biol. Chem. 264, 5474-5479). The two proteins exhibit a similar pattern of looped domains, each domain containing a homologous consensus sequence (i.e. Glu-Asp-Arg-Arg-Trp). The overall structure of both proteins seems to be highly related, with the exception of an N-terminal tyrosine-rich region present only in the mammalian extracellular matrix protein. The functional properties of the two proteins are similar in that the Limulus 18-kDa protein agglutinates horse erythrocytes and aggregates Limulus amebocytes, and the mammalian 22-kDa protein is an effective adhesion promoter for dermal fibroblasts. On the basis of these unique properties, the newly characterized hemagglutinin has been termed Limulus 18K agglutination-aggregation factor (18K-LAF).     

The primary structure of coagulation factor IX/factor X-binding protein isolated from the venom of Trimeresurus flavoviridis. Homology with asialoglycoprotein receptors, proteoglycan core protein, tetranectin, and lymphocyte Fc epsilon receptor for immunoglobulin E

An anticoagulant protein, factor IX/factor X-binding protein (IX/X-bp), isolated from the venom of Trimeresurus flavoviridis, binds with factor IX and factor X in the presence of Ca2+ with a 1 to 1 stoichiometry (Atoda, H., and Morita, T. (1989) J. Biochem. (Tokyo) 106, 808-813). Analysis of S-pyridylethylated IX/X-bp by sodium dodecyl sulfate-polyacrylamide gel electrophoresis revealed a 16.0-kDa band (designated the A chain) and a 15.5-kDa band (designated the B chain). These two chains were separated by reversed-phase high performance liquid chromatography, and their complete amino acid sequences were determined by sequencing of the peptides obtained after digestion with lysyl endopeptidase, chymotrypsin, and V8 protease from Staphylococcus aureus and after chemical cleavage with cyanogen bromide. The A chain had an amino-terminal sequence of Asp-Cys-Leu-Ser-Gly- and consisted of 129 residues with Mr 14,830. The B chain has an amino-terminal sequence of Asp-Cys-Pro-Ser-Asp- and consists of 123 residues of Mr 14,440. There was 47% identity between the A and the B chain. The sequence of IX/X-bp showed 25-37% identity with that of the C-type carbohydrate recognition domain-like structure of acorn barnacle lectin, human and rat asialoglycoprotein receptors, the human lymphocyte Fc epsilon receptor for immunoglobulin E, proteoglycan core protein, pancreatic stone protein, and tetranectin. The sequences of the first 18 amino acid residues of both the A and B chains were also, to a certain extent, homologous to the partial amino acid sequence of the b subunit of factor XIII, a member of the beta 2-glycoprotein I-like family. In this region, some similarity with the amino-terminal amino acid sequence of botrocetin was also observed.     

Evidence for conservation of the vasopressin/oxytocin superfamily in Annelida

Annetocin is a structurally and functionally oxytocin-related peptide isolated from the earthworm Eisenia foetida. We present the characterization of the annetocin cDNA. Sequence analyses of the deduced precursor polypeptide revealed that the annetocin precursor is composed of three segments: a signal peptide, an annetocin sequence flanked by a Gly C-terminal amidation signal and a Lys-Arg dibasic processing site, and a neurophysin domain, similar to other oxytocin family precursors. The proannetocin showed 37.4-45.8% amino acid homology to other prohormones. In the neurophysin domain, 14 cysteines and amino acid residues essential for association of a neurophysin with a vasopressin/oxytocin superfamily peptide were conserved, suggesting that the Eisenia neurophysin can bind to annetocin. Furthermore, in situ hybridization experiments demonstrated that the annetocin gene is expressed exclusively in neurons of the central nervous system predicted to be involved in regulation of reproductive behavior. These findings confirm that annetocin is a member of the vasopressin/oxytocin superfamily. This is the first identification of the cDNA encoding the precursor of an invertebrate oxytocin-related peptide and also the first report of the identification of an annelid vasopressin/oxytocin-related precursor.     

Further characterization of a conserved actin-binding 27-kDa fragment of actinogelin and alpha-actinins and mapping of their binding sites on the actin molecule by chemical cross-linking

A conserved actin-binding domain (Mr = 27,000) of rat hepatic actinogelin, rat skeletal muscle, and chicken gizzard alpha-actinins (Mimura, N., and Asano, A. (1986) J. Biol. Chem. 261, 10680-10687) was separated into two components having different isoelectric points (peptides A and B) by chromatofocusing. Thermolysin digestion of peptide A generated peptide B with concomitant loss of peptide A. Amino acid compositions and tryptic maps of peptides A and B also demonstrated that peptide A is a precursor of peptide B upon thermolysin digestion. All of peptides A and B retained the activity to bind with F-actin competitively to each other. By the gel-filtration method, it was also shown that the native actin-binding 27-kDa fragments are monomeric and globular. The non-actin-binding 50- or 53.5-kDa fragment of actinogelin/alpha-actinins was, however, found to be asymmetric and dimeric in the native state. Chemical cross-linking of the 27-kDa fragment with F-actin with a water-soluble carbodiimide produced at least four different complexes (I-IV). Chemical cleaving analysis of the cross-linked products (complexes I and II) indicated that the 27-kDa fragment possesses two possible binding sites on actin at the NH2-terminal residues 1-12 (for complex I) and at residues spanning 86-119 or 123 (for complex II).     

A teratocarcinoma glycoprotein carrying a developmentally regulated carbohydrate marker is a member of the immunoglobulin gene superfamily

Using the lambda gt11 expression library of murine teratocarcinoma cells, we isolated cDNA clones encoding a core protein carrying a developmentally regulated carbohydrate marker, namely the binding site for Dolichos biflorus agglutinin. The deduced amino acid sequence of the polypeptide (molecular weight, 37,102) revealed a leader sequence, nine potential asparagine glycosylation sites, and a transmembrane region. Sequence homology to variable domain of immunoglobulin kappa chain has been detected in a domain; homologous amino acids near cysteine residues are those conserved in many members of the immunoglobulin gene superfamily. In contrast to other members of the superfamily, the core protein gene was significantly expressed in embryonal carcinoma cells, which are similar to undifferentiated cells of early embryos.     

The primary structure of calf chymosin.

The complete amino acid sequence of calf chymosin (rennin) (EC 3.4.23.4) has been determined. The sequence consists of a single peptide chain of 323 amino acid residues. The primary structure of the precursor part of calf prochymosin was published previously (Pedersen, V.B., and Foltmann, B. (1975) Eur. J. Biochem. 55, 95-103), thus we are now able to account for the total 365 amino acid residues of calf prochymosin. Comparison of the sequence of calf prochymosin with that of pig pepsinogen A (EC 3.4.23.1) shows extensive homology. In the precursor part of the sequence, 15 residues are located at identical positions, as compared to 189 identical residues in the respective enzymes. Furthermore comparison to Penicillium janthinellum acid proteinase (penicillopepsin) (EC 3.4.23.7) shows that 76 residues are common to this enzyme and to the two gastric proteinases. These homologies in sequence further suggest that the folding of the peptide chain in chymosin is very similar to that of other acid proteinases.     

Complete primary structure of the transacylase (E2b) subunit of the human branched chain alpha-keto acid dehydrogenase complex

We isolated from a placental cDNA library by immunoscreening a cDNA clone encoding the transacylase (E2b) precursor of the human branched chain alpha-keto acid dehydrogenase (BCKDH) complex. The cDNA insert consists of 2,649 base pairs with an open reading frame of 1,431 base pairs which can be translated into 477 amino acids and a 3'-untranslated region of 1,205 base pairs. The deduced amino acid sequence includes a leader peptide of 56 amino acid residues, a lipoyl-bearing domain, a E3-binding domain and an inner core domain. A mature human E2b subunit is likely to contain 421 amino acid residues with a calculated Mr 46,322. The nucleotide sequence of the open reading frame and the deduced amino acid sequence of the human E2b shows 91.6% and 92.0% homology with those of the bovine E2b subunit, respectively.     

Molecular cloning and characterization of the Endo B cytokeratin expressed in preimplantation mouse embryos

A cDNA clone of a keratin-related, intermediate filament protein, designated Endo B, was constructed from size-fractionated parietal endodermal mRNA and characterized. The 1466-nucleotide cDNA insert contains an open reading frame of 1272 nucleotides that would result in 5' and 3' noncoding sequences of 54 and 60 nucleotides, respectively. The predicted amino acid composition, molecular weight (47,400), and peptide pattern correlate well with data obtained on the isolated protein. The predicted amino acid sequence fits easily into the general domain structure suggested for all intermediate filament proteins with a unique amino-terminal head domain, a large conserved central domain of predominantly alpha-helical structure, and a relatively unique carboxyl-terminal or tail domain. Over the entire molecule, Endo B is 43% identical with human 52-kDa epidermal type I keratin. However, over two of the three regions contained in the central domain that are predicted to form coiled-coil structures, the Endo B is 54-68% identical with other type I keratin sequences. This homology, along with the presence of the completely conserved sequence DNARLAADDFR-KYE, which is found in all type I keratins, permits the unambiguous identification of Endo B as a type I keratin. Comparison of the Endo B sequence to other intermediate filament proteins reveals 22 residues which are identical in all intermediate filament proteins regardless of whether filament formation requires only one type of protein subunit (vimentin, desmin, glial fibrillar acidic protein, or a neurofilament protein) or two dissimilar types (type I and type II keratins). Endo B mRNA was detectable in RNA isolated from F9 cells treated with retinoic acid for 48 h. Approximately three to five genes homologous to Endo B were detected in the mouse genome.