The effects of deleting the propeptide from human preproapolipoprotein A-I on co-translational translocation and signal peptidase processing

The two principal protein components of human high density lipoprotein particles, apoA-I and apoA-II, are initially synthesized as prepropolypeptides. The function of their NH2-terminal prosegments is not known. We have previously shown that deletion of the pentapeptide prosegment (Ala-Leu-Val-Arg-Arg) from human preproapoA-II redirects signal peptidase cleavage to a site located between the second and third residues of the mature protein (Folz, R. J., and Gordon, J. I. (1986) J. Biol. Chem. 261, 14752-14759). The hexapeptide prosegment of human preproapoA-I differs from other NH2-terminal propeptides in that it terminates with paired glutamine residues (Arg-His-Phe-Trp-Gln-Gln). To examine its role in the early events of protein export, we isolated a full-length human preproapoA-I cDNA and deleted its propeptide coding region by oligonucleotide-directed mutagenesis. The effects of this deletion on co-translational translocation and signal peptidase processing were assessed using an in vitro transcription/translation/microsomal membrane processing system. Propeptide deletion reduces the efficiency of co-translational translocation/processing in both reticulocyte and wheat germ lysates but does not affect the fidelity of signal peptidase cleavage. This reduced efficiency does not appear to reflect differences in signal recognition particle-nascent protein interaction as measured by a translational arrest assay. However, differences in translocation/processing rates were noted in a postinitiation translocation/cleavage assay. This assay employed the initiation inhibitor edeine to generate nascent wild type and mutant proteins of increasing chain length which were subsequently presented to canine pancreatic membranes. Unlike preproapoA-I, pre(delta pro)apoA-I retained its ability to undergo translocation and proteolytic processing even after the entire protein was synthesized. These data suggest that the propeptide in human preproapoA-I may play a role in promoting an optimal structure for co-translational translocation and processing.

The human and rodent intestinal fatty acid binding protein genes. A comparative analysis of their structure, expression, and linkage relationships

Intestinal fatty acid binding protein (I-FABP) is believed to participate in the uptake, intracellular metabolism, and/or transport of long chain fatty acids within enterocytes. The 15.1-kDA rodent proteins is a member of a family of low Mr cytoplasmic proteins that have evolved to bind different ligands. We have now determined the nucleotide sequence of the gene encoding human I-FABP and defined the primary structure of its protein product. The human I-FABP gene spans 3382 nucleotides and contains 4 exons (103 or 128, 173, 108, and 312 base pairs). interrupted by 3 introns (1194, 1023, and 444 base pairs). The 132-residue rat and human I-FABPs have 82% amino acid sequence identity. Blot hybridization studies of RNAs prepared from a variety of adult rhesus monkey tissues as well as human intestine and liver indicate that I-FABP mRNA is confined to the intestine. I-FABP mRNA was not detectable in a number of cultured human enterocyte-like cell lines, suggesting it may be a sensitive marker for differentiated, villus-associated, small intestinal lining cells. Given the similar patterns of tissue-specific expression exhibited by the rat and human genes, we compared their 5' nontranscribed regions. Optimal alignments of the two sequences disclosed 64% identity among the 260 nucleotides immediately 5' to the start site of transcription. Matrix plots revealed a 14-nucleotide long repeated sequence (5'-TGAACTTTGAACTT-3') in the 5' nontranscribed region of both genes as well as in a comparable region of another family member that is expressed in enterocytes, cellular retinol binding protein II. The linkage relationships between I-FABP and the homologous liver FABP (L-FABP) gene were defined in mice and humans. The mouse genes were mapped using restriction fragment length polymorphisms and recombinant inbred strains. The I-FABP gene is located on mouse chromosome 3 between the amylase 1,2 (Amy 1,2) and alcohol dehydrogenase 3 (Adh-3) loci while the L-FABP gene is on mouse chromosome 6 within 3 centimorgans of the lymphocyte antigen-2 (Ly-2) locus. Mouse L-FABP may be identical to the major liver protein-1 (Lvp-1) which is encoded by a gene situated within a centimorgan of Ly-2. Human gene mapping studies were carried out using a panel of mouse-human somatic cell hybrid clones as well as in situ hybridization to metaphase chromosomes.(ABSTRACT TRUNCATED AT 400 WORDS)

Crystallization of rat cellular retinol binding protein II. Preliminary X-ray data obtained from the apoprotein expressed in Escherichia coli

Rat cellular retinol-binding protein II (CRBP II) is a member of a family of cytoplasmic proteins which bind hydrophobic ligands. CRBP II is thought to participate in the intestinal absorption and intracellular metabolism of retinoids. We have previously described the crystallization of a homologous rat intestinal fatty acid-binding protein (I-FABP) isolated from Escherichia coli containing a suitably constructed prokaryotic expression vector (Sacchettini, J. C., Meininger, T. A., Lowe, J. B., Gordon, J. I., and Banaszak, L. J., J. Biol. Chem. 262, 5428-5430). We have now efficiently expressed rat CRBP II in E. coli. The E. coli-derived protein, which does not contain any bound retinoid, has been purified and crystals grown from solutions of polyethylene glycol 4000. Crystals of apo-CRBP II are triclinic, space group P1, a = 36.8 A, b = 64.0 A, c = 30.4 A; alpha = 92.8 degrees, beta = 113.5 degrees, gamma = 90.1 degrees. Each unit cell contains two molecules of the 134-residue apoprotein. X-ray diffraction data suggest that the unit cell parameters of crystalline apo-CRBP II resemble those of I-FABP. Comparison of the tertiary structures of E. coli-derived rat I-FABP and CRBP II should provide insights about how these proteins evolved to bind different hydrophobic ligands.

Characterization of rat cellular retinol-binding protein II expressed in Escherichia coli

Rat cellular retinol-binding protein II (CRBP II) is a small (15.6 kDa) intracellular protein that binds all-trans-retinol. In the adult rat, expression of the CRBP II gene is essentially limited to the small intestinal lining cells (enterocytes), suggesting that CRBP II may be uniquely adapted for intestinal metabolism of newly absorbed retinol. Functional and structural analysis of this protein has been hampered by difficulties in freeing rat intestinal CRBP II from its ligand without denaturation. To circumvent this problem, we have obtained efficient expression of rat apoCRBP II in Escherichia coli. The purified E. coli-derived apoprotein, when complexed with all-trans-retinol, demonstrates fluorescence excitation-emission spectra and absorption spectra indistinguishable from that of CRBP II-retinol isolated from rat intestine. Quantitative ligand binding studies were performed by monitoring either the fluorescence of bound retinol or the quenching of protein fluorescence. They revealed that E. coli-derived CRBP II binds retinol tightly (the apparent dissociation constant is estimated to be 10(-7)-10(-8) M), with a stoichiometry of 1:1. Fluorescence quenching studies used acrylamide as a probe for the exposure of the 4 tryptophan residues to solvent. The results indicate that although there is heterogeneity in the exposure of these 4 tryptophan residues to solvent, they are situated in a relatively nonpolar environment. These studies suggest that E. coli-derived apoCRBP II will serve as a useful model for studying retinol-protein interactions.

Biosynthesis of human insulin-like growth factor I (IGF-I). The primary translation product of IGF-I mRNA contains an unusual 48-amino acid signal peptide

The human insulin-like growth factor I (IGF-I) gene can specify two primary translation products, IGF-IA and IGF-IB, which differ at their COOH termini and which must undergo extensive proteolytic processing to produce the mature 70-residue circulating peptide. We have used an in vitro transcription/translation/microsomal membrane processing system to define the early events in IGF-I biosynthesis. Our results demonstrate that protein synthesis is initiated at the first in-frame methionine codon for both human IGF-IA and IGF-IB mRNAs, yielding primary translation products of 153 and 195 residues, respectively. Both are cotranslationally translocated into the lumen of canine pancreatic microsomes. Cotranslational proteolytic processing results in the removal of an unusually large 48-amino acid signal peptide. Computer-assisted sequence analysis revealed that this prepeptide contains two domains. Its COOH-terminal 19 residues possess features which are typically encountered in eukaryotic signal peptides. The rest of the IGF-I signal sequence is represented by an unusual "extended" NH2-terminal domain containing an overall net charge of +5 and 3 clustered cysteines. The size and charge of this NH2-terminal domain distinguish it from comparable regions of other eukaryotic signal peptides including the presegment of human IGF-II. Although no specialized biological role can currently be ascribed to this unique NH2-terminal domain, its remarkable sequence conservation between human and rat suggests an as yet unknown functional significance.

Primary structure and comparative sequence analysis of an insect apolipoprotein. Apolipophorin-III from Manduca sexta

The amino acid sequence of an insect apolipoprotein, apolipophorin-III from Manduca sexta, was determined by a combination of cDNA and protein sequencing. The mature hemolymph protein consists of 166 amino acids. The cDNA also encodes for an amino-terminal extension of 23 amino acids which is not represented in the mature hemolymph protein. The existence of a precursor protein was confirmed by in vitro translation of fat body mRNA. Computer-assisted comparative sequence analysis revealed the following points: 1) the protein is composed of tandemly repeating tetradecapeptide units with a high potential for forming amphiphilic helical structures. Compared to mammalian apolipoproteins the repeat units in the insect apolipoprotein show considerable length variability; 2) the sequence has a striking resemblance to several human apolipoproteins including apoE, AIV, AI, and CI. However, the homology seems to be entirely functional since, although the insect and mammalian apoproteins contain very similar types of amino acid residues, the actual degree of sequence identity is quite low. Whether the mammalian and insect apoproteins are derived from a common ancestral amphiphilic helix forming, lipid-binding protein, or arose by convergent evolution can not be determined at present. This represents the first complete amino acid sequence for an insect apolipoprotein.

Computer-assisted predictions of signal peptidase processing sites

Computer programs are presented which incorporate 2 different algorithms for predicting the site of signal peptide cleavage for eukaryotic preproteins. These programs can be used to identify sites of signal peptidase cleavage of putative preproteins to facilitate the design and interpretation of signal peptide mutagenesis experiments, and to engineer artificial prepolypeptides.

Analysis of the tissue-specific expression, developmental regulation, and linkage relationships of a rodent gene encoding heart fatty acid binding protein

The rat contains at least three homologous cytosolic proteins that bind long chain fatty acids, termed liver (L-), intestinal (I-), and heart (H-) fatty acid binding protein (FABP). I-FABP mRNA is confined to the gastrointestinal tract while L-FABP mRNA is abundantly represented in hepatocytes as well as enterocytes. We have isolated a rat heart FABP cDNA clone and determined the pattern of H-FABP mRNA accumulation in a wide variety of tissues harvested from late fetal, suckling, weaning, and adult rats. RNA blot hybridizations and primer extension analysis disclosed that the distribution of H-FABP mRNA in adult rat tissues is different from that of I- or L-FABP mRNA. H-FABP mRNA is most abundant in adult heart. This mRNA was also present in an adult slow twitch (type I) skeletal muscle (soleus, 63% of the concentration in heart), testes (28%), a fast twitch skeletal muscle (psoas, 17%), brain (10%), kidney (5%), and adrenal gland (5%). H-FABP mRNA was not detected in adult small intestine, colon, spleen, lung, or liver RNA. Distinct patterns of developmental change in H-FABP mRNA accumulation were documented in heart, placenta, brain, kidney, and testes. Myocardial H-FABP mRNA levels rise rapidly during the 48 h prior to and after birth, reaching peak levels by the early weaning period. The postnatal increase in myocardial H-FABP mRNA concentration and its relative distribution in adult fast and slow twitch skeletal muscle are consistent with its previously proposed function in facilitating mitochondrial beta-oxidation of fatty acids. However, the presence of H-FABP mRNA in brain, a tissue which does not normally significantly oxidize fatty acids in late postnatal life, suggests that H-FABP may play a wider role in fatty acid metabolism than previously realized. Mouse-hamster somatic cell hybrids were utilized to map H-FABP. Using stringencies which did not produce cross-hybridization between L-, I-, and H-FABP DNA sequences, we found at least three loci in the mouse genome, each located on different chromosomes, which reacted with our cloned H-FABP cDNA. None of these H-FABP-related loci were linked to the gene which specifies a highly homologous adipocyte-specific protein termed aP2 or to genes encoding two other members of this protein family, cellular retinol binding protein and cellular retinol binding protein II.(ABSTRACT TRUNCATED AT 400 WORDS)

Structure and expression of the human apolipoprotein A-IV gene

We have isolated the human apolipoprotein (apo) A-IV gene from a cosmid library and determined its complete nucleotide sequence. The gene contains three exons of 162, 127, and 1180 nucleotides separated by two introns of 357 and 777 nucleotides. A sequence polymorphism has been identified in the 3' noncoding portion of the third exon. The human apoA-IV gene lacks an intron in the area encoding the 5' nontranslated region of its mRNA, which distinguishes it from all the other human apolipoprotein genes whose sequences are known. Comparison matrix analysis of the human apoA-IV gene sequence revealed evidence for an ancestral 11-nucleotide repeat unit that spans the third exon. These repeated sequences are much more highly conserved than those present in either rat apoA-IV or in any other human apolipoprotein. Optimal alignments of the 5' flanking regions of the rat and human apoA-IV genes disclosed multiple deletions in the rat sequence as well as a highly conserved region of 90 nucleotides (90% sequence identity) located within 170 nucleotides of the start site of transcription. The 5' flanking regions of the human and rat apoA-IV genes were ligated to the bacterial chloramphenicol acetyltransferase gene, then transfected into different cultured cells. The apoA-IV gene sequences elicited preferential expression of chloramphenicol acetyltransferase activity when introduced into intestinally derived Caco-2 cells and liver-derived Hep-G2 cells, consistent with the tissue specificity of the native gene. Analysis of deletion mutants of the human apoA-IV 5' flanking region indicated that regions from -293 to -233 and from -127 to -60 upstream of the transcription start site contain sequences required for maximum gene expression. These findings on the structure and expression of rat and human apoA-IV should prove useful in studying the control of the apoA-IV gene.

Comparison of the tissue-specific expression and developmental regulation of two closely linked rodent genes encoding cytosolic retinol-binding proteins

Cellular retinol-binding protein (CRBP) and cellular retinol-binding protein II (CRBP II) are two highly homologous cytoplasmic proteins that bind all-trans-retinol. We have recently demonstrated that the mouse genes encoding CRBP and CRBP II are closely linked on chromosome 9 and that both human genes are located on chromosome 3 (Demmer, L.A., Birkenmeier, E.H., Sweetser, D.A., Levin, M.S., Zollman, S., Sparkes, R.S., Mohandas, T., Lusis, A.J., and Gordon, J.I. (1987) J. Biol. Chem. 262, 2458-2467). We have now used RNA blot hybridization analysis to assess the degree to which these genes are coordinately expressed in fetal, suckling, weaning, and adult rat tissues. Both genes exhibit different developmental patterns of expression in liver, intestine, lung, kidney, testes, and placenta. In the intestine, CRBP mRNA was detected during the 16th day of gestation--prior to the development of a well-differentiated absorptive epithelium--and remained essentially unchanged throughout the peri- and postpartum periods. By contrast, the pattern of intestinal CRBP II mRNA accumulation closely parallels the times of first appearance, and subsequent proliferation, of the intestinal absorptive columnar epithelium, supporting the hypothesis that CRBP II is involved in the intestinal uptake or intracellular trafficking of this hydrophobic vitamin. In the fetal liver, both genes were expressed by gestational day 16. Whereas the concentration of hepatic CRBP mRNA increased markedly during the suckling and early weaning periods, CRBP II mRNA levels fell abruptly immediately after birth. These peripartum changes were not paralleled by remarkable alterations in the steady state levels of hepatic retinol. Marked changes in the expression of CRBP in the liver and of CRBP II in the intestine were also documented in pregnant and lactating female rats. These differences in CRBP/CRBP II gene expression strongly suggest that their proteins serve different physiological functions. The peripartum liver may provide a useful model for dissecting the relative roles played by these homologous proteins in retinoid metabolism as well as the factors which modulate activation and repression their genes.

Purification and characterization of yeast myristoyl CoA:protein N-myristoyltransferase

Myristoyl CoA:protein N-myristoyltransferase (NMT) catalyzes the addition of myristic acid to the amino-terminal glycine residues of a number of eukaryotic proteins. Recently, we developed a cell-free system for analyzing NMT activity and have begun to characterize the substrate specificity of this enzyme by using a series of synthetic peptides. We have now purified NMT from Saccharomyces cerevisiae to apparent homogeneity. The native enzyme is a 55-kDa protein, exhibits no requirement for divalent cation, and appears to contain a histidine residue critical for enzyme activity. A total of 42 synthetic peptides have been used to define structure/activity relationships in NMT substrates. An amino-terminal glycine is required for acylation; substitution with glycine analogues produces peptides that are inactive as substrates or inhibitors of NMT. A broad spectrum of amino acids is permitted at positions 3 and 4, while strict amino acid requirements are exhibited at position 5. Replacement of Ala5 in the peptide Gly-Asn-Ala-Ala-Ala-Ala-Arg-Arg with Asp ablates the peptide's myristoyl-accepting activity. A serine at this position results in a decrease by a factor of approximately equal to 500 in the apparent Km in the context of three different sequences. Penta- and hexa-peptides are substrates, but with decreased affinity. These studies establish that structural information important for NMT-ligand interaction exists beyond the first two amino acids in peptide substrates and that the side chains of residue 5 play a critical role in the binding of substrates to this enzyme.

Expression of rat intestinal fatty acid-binding protein in Escherichia coli. Purification and comparison of ligand binding characteristics with that of Escherichia coli-derived rat liver fatty acid-binding protein

Rat intestinal fatty acid-binding protein (I-FABP) is an abundant, 15,124-Da polypeptide found in the cytosol of small intestinal epithelial cells (enterocytes). It is homologous to rat liver fatty acid-binding protein (L-FABP), a 14,273-Da cytosolic protein which is found in enterocytes as well as hepatocytes. It is unclear why the small intestinal epithelium contains two abundant fatty acid-binding proteins. A systematic comparative analysis of the ligand binding characteristics of the two FABPs has not been reported. To undertake such a study we expressed the coding region of a full length I-FABP cDNA in Escherichia coli and purified large quantities of the protein. We also purified rat L-FABP from a similar, previously described expression system (Lowe, J. B., Strauss, A. W., and Gordon, J. I. (1984) J. Biol. Chem. 259, 12696-12704). Analysis of fatty acids associated with each of the homogeneous E. coli-derived FABPs suggested that the two proteins differed in their ligand binding specificity and capacity. All of the fatty acids associated with I-FABP were saturated while 30% of the E. coli fatty acids bound to L-FABP were unsaturated (16:1, 18:1, 18:2). We directly analyzed the ability of I- and L-FABP to bind fatty acids of different chain length and degree of saturation using a hydroxyalkoxypropyl dextran-based assay. Scatchard analysis revealed that each mole of L-FABP can bind up to 2 mol of long chain fatty acid while each mole of I-FABP can bind only 1 mole of fatty acid. L-FABP exhibited a relatively higher affinity for unsaturated fatty acids (oleate, arachidonate) than for saturated fatty acid (palmitate). By contrast, we were not able to detect a significant difference in the affinity of I-FABP for palmitate, oleate, and arachidonate. Neither protein exhibited any appreciable affinity for fatty acids whose chain length was less than C16. The observed differences in ligand affinities and capacities suggest that these proteins may have distinct roles in metabolism and/or compartmentalization of fatty acids within enterocytes.

Crystallization of rat intestinal fatty acid binding protein. Preliminary X-ray data obtained from protein expressed in Escherichia coli

Rat intestinal fatty acid binding protein has been expressed in Escherichia coli, purified with bound long chain fatty acids and crystals grown from solutions of polyethylene glycol 4000. The crystals are monoclinic, space group P2(1), a = 3638 A, b = 57.2 A, c = 31.9 A, and beta = 113.9 degrees. Each unit cell contains two monomers of this 132-residue, 15.1-kDa polypeptide. The crystals are remarkably resistant to x-ray damage. X-ray diffraction data have been observed to 2.0 A resolution. Platinum chloride was used to generate a potential isomorphous heavy atom derivative.

The cellular retinol binding protein II gene. Sequence analysis of the rat gene, chromosomal localization in mice and humans, and documentation of its close linkage to the cellular retinol binding protein gene

Cellular retinol binding protein II (CRBP II) is an abundant, 134-residue protein present in the small intestinal epithelium. It is thought to participate in the uptake and/or intracellular metabolism of vitamin A. We have isolated and sequenced the rat CRBP II gene. Its four exons span 0.65 kilobases and are interrupted by three introns with an aggregate length of 19.5 kilobases. Southern blot hybridization analysis indicated that this gene is highly conserved in rats, mice, and humans. CRBP II belongs to a protein family that contains eight known members. Computer-assisted comparative sequence analyses indicated that a region of internal homology spans its first two exons and that oligopeptide domains specified by these first two exons exhibit significant homology to all other family members as well as to a portion of the all-trans-retinol binding domain that has previously been defined in serum retinol binding protein. The CRBP II gene was mapped in mice using recombinant inbred strains and restriction fragment length polymorphisms. It is located on chromosome 9 within 5.3 centimorgans of the phosphoglucomutase-3 locus and is closely linked (within 3.0 centimorgans) to the gene specifying a highly homologous intracellular retinol binding protein known as CRBP. Mouse-human somatic cell hybrids were used to determine that both the CRBP and CRBP II genes are located on human chromosome 3.

Characterization of sites of tyrosine sulfation in proteins and criteria for predicting their occurrence

A wide variety of secretory proteins have recently been found to undergo post-translational sulfation of specific tyrosine residues. Here, amino acid sequences surrounding known sulfation sites in proteins are analyzed in order to identify factors which determine the specificity of sulfation. Several distinctive features of sulfation sites are identified, including: abundance of acidic amino acid residues, lack of basic residues, low hydropathy, absence of neighboring cysteine residues, lack of extended secondary structure. Rules are proposed for predicting likely sites of sulfation based on the amino acid sequence of a protein.

Deletion of the propeptide from human preproapolipoprotein A-II redirects cotranslational processing by signal peptidase

The functions of NH2-terminal propeptides are not known. We have used apoA-II as a model to study prosegment structure/function relationships. The primary translation product of human apolipoprotein A-II mRNA contains an 18-amino acid signal peptide, a 5-amino acid propeptide, and the mature 77-amino acid plasma protein sequence. Its propeptide was deleted by site-directed mutagenesis of a cloned cDNA. The effects of this mutation on cotranslational translocation and proteolytic processing were assessed using an in vitro transcription/translation/microsomal membrane processing system. Deletion of the propeptide did not affect cotranslational translocation. However, without its propeptide, signal peptidase cleavage was redirected to a different site located between the 2nd and 3rd residues of the mature protein. Since the primary structure of the signal peptide was not altered in the mutant, these results suggest that sequences located downstream from the signal peptidase cleavage site (e.g. in propeptides) may modulate, or participate in defining, the correct site of cotranslational proteolytic processing.

Tissue-specific expression and developmental regulation of the rat apolipoprotein B gene

Expression of the apolipoprotein B (apoB) gene was examined in a variety of fetal, neonatal, and adult rat tissues by probing RNA blots with a cloned rat apoB cDNA. Among 10 adult male tissues surveyed, small intestine had the highest concentration of apoB mRNA. Its abundance in liver and adrenal gland was 40% and 0.5%, respectively, of that in small bowel, while none was detected in colon, kidney, testes, spleen, lung, heart, or brain. ApoB mRNA is as abundant in 18-day fetal liver as at any subsequent period of hepatic development. In contrast, the concentration of apoB mRNA remains low in fetal intestine until the last (21st) day of gestation, when it increases sharply to levels that are several-fold higher than in the liver. ApoB mRNA levels in fetal membranes harvested during this late gestational period were 10 times greater than in fetal liver. Since the major lipoprotein species in 19-day fetal plasma is low density lipoprotein, these observations suggest that fetal liver, and particularly its functional homologue, the yolk sac, are the principal sites of fetal lipoprotein synthesis at this stage of development. A 20-fold increase in placental apoB mRNA concentrations during the last 48 hr of pregnancy (to a level that is 50% of that encountered in fetal membrane RNA) suggests a specific role for this organ in maternal-fetal lipid transport immediately prior to parturition. Pulse-labeling experiments using 21-day fetal tissue slices showed that the liver synthesizes both apoB-100 (B-PI) and apoB-48 (B-PIII) albeit in somewhat different ratios than the adult organ. Fetal intestine produces almost exclusively the smaller apoB species, while fetal membranes and placenta synthesize only the larger peptide. The postnatal pattern of apoB mRNA accumulation is similar in liver and intestine. Profound decreases were observed during the late suckling and weaning periods, followed by an increase to adult levels. These final concentrations were similar to those encountered at birth. Analysis of these developmental changes offers an opportunity to generate testable hypotheses about the factors that modulate apoB synthesis.

On computer-assisted analysis of biological sequences: proline punctuation, consensus sequences, and apolipoprotein repeats

During the past several years, the use of computer programs in the analysis of protein and DNA sequences has become commonplace. In all but the simplest procedures, the ability to critically review the results obtained with computer methods requires a basic knowledge of the algorithms employed (and the assumptions upon which they are based), an awareness of the capabilities and limitations of the particular program that implements an algorithm, and some familiarity with probability and statistics. We describe a number of computer methods that have been applied to the analysis of apolipoprotein sequences. We discuss the suitability of these methods for particular problems, how the choice of initial "parameters" can affect the results, and what the results can tell us about protein or gene sequences. We also identify some outstanding problems of apolipoprotein sequence analysis where further work is needed.

Rat cellular retinol-binding protein II: use of a cloned cDNA to define its primary structure, tissue-specific expression, and developmental regulation

The primary structure of rat cellular retinol-binding protein (CRBP) II has been determined from a cloned cDNA. Alignment of this 134-amino acid, 15,580-Da polypeptide with rat CRBP revealed that 75 of 133 comparable residues are identical. Both proteins contain four tryptophan residues, which occupy identical relative positions in the two primary structures, providing a structural explanation for their similar fluorescence spectra when complexed to retinol. Two of the three cysteines in each single-chain protein are comparably positioned. Both polypeptides contain reactive thiol groups, but the rate of disruption of CRBP II-retinol complexes by p-chloromercuribenzoate is greater than that of CRBP-retinol. The small intestine contains the highest concentrations of CRBP II mRNA in adult rats. CRBP II mRNA is first detectable in intestinal RNA during the 19th day of gestation, a time that corresponds to the appearance of an absorptive columnar epithelium. Unlike in intestine, a dramatic fall in liver CRBP II mRNA concentration occurs immediately after birth. The CRBP II gene remains quiescent in the liver during subsequent postnatal development. These data suggest that ligand-protein interactions may be somewhat different for the two rat CRBPs. They also support the concept that CRBP II plays a role in the intestinal absorption or esterification of retinol and suggest that changes in hepatic metabolism of vitamin A occur during development.

Rat heart fatty acid-binding protein is highly homologous to the murine adipocyte 422 protein and the P2 protein of peripheral nerve myelin

The rat heart contains an abundant cytosolic protein which binds long chain fatty acids. We have determined its primary structure by Edman degradation of peptides generated from chymotryptic, tryptic, and elastase digestions. This polypeptide (Mr = 14,992) contains 134 amino acids and has a blocked (acetylated) NH2 terminus. The sequence of rat heart fatty acid-binding protein (FABP) is remarkably similar to the murine adipocyte 422 protein and the P2 protein of peripheral nerve myelin. Computer-assisted alignment of heart FABP and 422 revealed that 82 of 132 comparable residues are identical (62%). There are 77 identities out of 131 possible matches between this protein and the human myelin P2 protein (59%). Similar comparisons demonstrate that heart FABP has significant homology to several other proteins which bind hydrophobic ligands. The rank of order of similarity to heart FABP is: 422 greater than myelin P2 greater than cellular retinoic acid-binding protein greater than cellular retinol-binding protein II greater than cellular retinol-binding protein greater than intestinal FABP greater than liver FABP. These eight sequences form a family of paralogous homologues. Heart FABP has a region of internal homology involving tandemly arrayed oligopeptides spanning residues 71-100 and 101-131. This feature is not found in the 422 and P2 sequences. The endogenous ligands bound by the 422, P2, and heart FABP sequences have not been defined. Interpretation of the biological significance of their structural similarities and differences will require information about their ligand specificities and affinities.

Evolution of the apolipoproteins. Structure of the rat apo-A-IV gene and its relationship to the human genes for apo-A-I, C-III, and E

We have determined the nucleotide sequence of the rat apolipoprotein (apo-) A-IV gene and analyzed its structural and evolutionary relationships to the human apolipoprotein A-I, E, and C-III genes. The rat A-IV gene is 2.4 kilobases in size and consists of three exons (142, 126, and 1157 base pairs) interrupted by two introns (277 and 673 base pairs). The 5'-nontranslated region and most of the signal peptide are encoded by the first exon. Thus, the apo-A-IV gene lacks an intron in the 5'-nontranslated region of its mRNA in contrast to all other known apolipoprotein genes. Sequences coding for amphipathic docosapeptides span both the second and third exons of the rat A-IV gene. We demonstrate that this is also true for the human apolipoprotein genes. This gene family seems to have evolved by the duplication of an ancestral minigene that resulted in the formation of two exons. Thereafter, evolution of these sequences was dominated by intraexonic amplification of repeating units coding for amphipathic peptides. Sequence divergence of these repeats resulted in the functional differentiation of the apolipoproteins. However, conservation of the fundamental amphipathic pattern allowed members of this protein family to retain their lipid-binding properties.

The nucleotide sequence of the rat liver fatty acid-binding protein gene. Evidence that exon 1 encodes an oligopeptide domain shared by a family of proteins which bind hydrophobic ligands

We have determined the nucleotide sequence of the gene encoding rat liver fatty acid-binding protein (L-FABP). Previous structural studies have shown that L-FABP belongs to a family of low molecular weight cytosolic proteins which bind hydrophobic ligands. Rat L-FABP is the first member of this family whose genomic organization has been defined. The L-FABP transcription unit spans 3790 nucleotides and contains four exons (115, 173, 93, and 121 base pairs) interrupted by three introns (1454, 1224, and 610 base pairs). No other abundant mRNAs appear to be transcribed from sequences located within 4 kilobases 5' or 6.5 kilobases 3' of this gene. Sequence analyses have detected the presence of several related amino acid sequence blocks within L-FABP which may serve similar structural roles. A variety of computational techniques were used to compare the oligopeptides specified by each exon with other known members of the protein family. The results indicate that only the amino acid sequence present in the first exon is conserved among the homologous proteins. This suggests that the first exon may encode a shared structural and functional domain important in the interaction of these proteins with their ligands.

Developmental regulation of a gene that encodes a cysteine-rich intestinal protein and maps near the murine immunoglobulin heavy chain locus

Mouse and rat small intestinal cDNA libraries were screened for recombinants derived from mRNAs whose concentration changed during the transition from suckling to weaning. cDNAs transcribed from a 570-nucleotide-long mRNA were isolated. Dot blot hybridization analyses of RNA recovered at various stages of rat gastrointestinal ontogeny indicated that the concentration of this mRNA begins to increase during the mid-suckling period, reaching a peak during weaning. There is considerable variation in the relative amount of this mRNA in adult tissues, with highest levels encountered in the rat small intestine and colon. Its concentration in duodenum, jejunum, and ileum is approximately the same. It is more concentrated in villi than in crypts. The rat mRNA encodes a 77 amino acid, 8.55-kDa polypeptide that has seven cysteine residues. This cysteine-rich intestinal protein (named CRIP) has two internal repeated sequence blocks. Computer-assisted comparisons of CRIP to proteins of known function disclosed that it is homologous to certain ferredoxins. Southern blot analyses revealed that sequences homologous to the rat gene are present in sea squirt, fish, bird, and human DNA, indicating that this gene is highly conserved and that related proteins may be present in many if not all vertebrates. Recombinant inbred mouse strains were utilized to show that the CRIP gene is closely linked to the immunoglobulin heavy chain constant region locus, Igh-c, on chromosome 12. CRIP mRNA is a molecular marker for the suckling-to-weaning transition of rodent intestinal development. The cloned cDNA may be a useful probe for identifying factors that regulate intestinal development during this period.

The nucleotide and derived amino acid sequence of human apolipoprotein A-IV mRNA and the close linkage of its gene to the genes of apolipoproteins A-I and C-III

Both cDNA and genomic clones encoding human apolipoprotein (apo-) A-IV have been isolated and characterized. Southern blot analyses of apo-A-IV gene-containing cosmids revealed that the apo-A-IV gene is linked to the apo-A-I and apo-C-III genes within a 20-kilobase span of chromosome 11 DNA. The apo-A-IV gene is located about 14 kilobases downstream from the apo-A-I gene in the same orientation, with the apo-C-III gene located between them in the opposite orientation. The nucleotide sequence of the corresponding human apo-A-IV mRNA was determined, and the derived amino acid sequence showed that mature plasma apo-A-IV contained 376 residues. Throughout most of its length, human apo-A-IV was found to contain multiple tandem 22-residue repeated segments having amphipathic, alpha-helical potential. Amino acid substitutions within these homologous segments were generally conservative in nature. A comparison of the sequences of human and rat apo-A-IV revealed a 79% identity of amino acid positions in the amino-terminal 60 residues and a 58% identity in the remainder of the sequences, with the human protein containing 5 extra residues near the carboxyl terminus. An examination of the distribution of apo-A-IV mRNA in different tissues of the rat, marmoset, and man showed that apo-A-IV mRNA was abundant in both the liver and small intestine of the rat, but abundant in both the liver and small intestine of the marmoset and man. It was expressed in only trace amounts in all other tissues that were examined. These findings on the structure and expression of apo-A-IV and the close linkage of its gene to those of apo-A-I and apo-C-III suggest a regulatory relationship between the three genes.

Proteolytic processing and compartmentalization of the primary translation products of mammalian apolipoprotein mRNAs

The steps involved in the initial assembly of apolipoproteins and lipids into supramolecular arrays (nascent lipoprotein particles) are largely unknown. Examination of the proteolytic processing and compartmentalization of the primary translation products of apolipoprotein mRNAs represents one approach to deciphering the molecular details of lipoprotein assembly. The structures of the primary translation products of seven mammalian apolipoprotein mRNAs has been determined in the past several years. The organization of apolipoprotein signal peptides is typical of eukaryotic prepeptides, although an unusual degree of sequence conservation is present among the signal segments of apo AI, AIV, and E. For those apolipoprotein sequences studied in detail, SRP-dependent cotranslational translocation and proteolytic processing appears to be highly efficient and results in sequestration of the processed protein within the lumen of the endoplasmic reticulum (ER). However the mechanism by which these lipid-binding proteins avoid arrest during their translocation through the lipid bilayer of the ER membrane remains obscure. The two principal human HDL apolipoproteins undergo novel extracellular post-translational proteolytic processing, which results in removal of nonhomologous propeptides. The proteases responsible for proapo AI and AII processing appear to be different. The processing of these proapolipoproteins provides a potential series of steps for regulating the ordered assembly of HDL constituents.

Rat apolipoprotein A-IV: application of computational methods for studying the structure, function, and evolution of a protein

There are a great variety of computational methods available to study protein and nucleic acid sequences. The choice of a computer program appropriate to a particular problem and the critical interpretation of the results can lead to specific, and experimentally testable, predictions of a protein's structure and function and may yield insights into its evolution and the location of its gene. We have shown that rat apoA-IV bears a striking structural similarity to human apoA-I (summarized in Fig. 7). Statistical analyses of homologies between apolipoproteins A-I, A-IV, and E demonstrate conclusively that all three sequences diverged from a common ancestral gene. That apoA-IV largely composed of 22 amino acid amphipathic segments with alpha-helical potential suggests that it possesses the structural requirements for LCAT activation. Analysis of the apoB, E receptor-binding domain of human apoE3 has demonstrated that it evolved from an ancestral repeated sequence. Assuming that the genes for these proteins evolved as a result of a series of intra- and intergenic unequal crossovers, it is likely that their genetic loci were at one time linked. The repeated sequences of which these genes are composed have propagated themselves in an expansionary manner. Given this fact, the existence of other genes or pseudogenes based upon this repeated sequence motif is a distinct possibility.

Expression of rat apolipoprotein A-IV and A-I genes: mRNA induction during development and in response to glucocorticoids and insulin

Rat apolipoproteins (apo-) A-IV and A-I share many structural similarities, the most notable of which is a domain of repeated docosapeptides with amphipathic helical potential. Although the genes for apo-A-IV and apo-A-I probably diverged from a common ancestor, these proteins seem to have developed different functions in their evolution. In the present study, cloned cDNAs were used to characterize the expression of apo-A-IV and apo-A-I mRNAs in a wide variety of adult rat tissues, as well as in small intestine and liver obtained from fetal, suckling, and weanling animals; comparisons were made to the expression of apo-E mRNA. The apo-A-IV and apo-A-I mRNAs were most abundant in adult small intestine and liver, with trace amounts detected in other tissues. Substantial amounts of these mRNAs were detected in the yolk sac, suggesting that this fetal tissue plays an important role in lipid metabolism during gestation. Noncoordinate accumulation of apo-A-IV and apo-A-I mRNAs was observed within and between the liver and small intestine during neonatal development. The apo-A-IV mRNA levels in the developing small intestine and liver appeared to correlate with their triglyceride secretion rates, suggesting that this protein plays an important role in the metabolism of triglyceride-rich lipoproteins. When dexamethasone (0.1 microM), insulin (0.01 microM), or insulin and dexamethasone together were incubated with primary cultures of nonproliferating adult rat hepatocytes, apo-A-IV mRNA levels were 4-, 7-, and 11-fold higher, respectively, than in non-hormone-treated control hepatocytes. Hormone administration resulted in a 2-fold greater amount of apo-A-I mRNA in each case, with no significant change in the level of apo-E mRNA. The overall results suggest that these structurally related apolipoproteins are regulated in substantially different ways.

Extracellular processing of proapolipoprotein A-II in Hep G2 cell cultures is mediated by a 54-kDa protease immunologically related to cathepsin B

Apolipoprotein A-II is the second most abundant polypeptide found in human plasma high density lipoprotein particles. The primary translation product of human apo-A-II mRNA is a prepropolypeptide. We have previously reported (Gordon, J. I., Sims, H. F., Edelstein, C., Scanu, A. M., and Strauss, A. W. (1984) J. Biol. Chem. 259, 15556-15563) that the prosegment of apo-A-II was removed following export from a human hepatoma cell line (Hep G2). This represented a novel processing compartment for prosegments terminating with paired basic residues and differed from the processing of proalbumin which occurred with high efficiency prior to export from these cells. We have now characterized the enzyme responsible for this extracellular cleavage. The proapo-A-II converting activity is blocked by the thiol protease inhibitors antipain, E-64, leupeptin, and Ala-Lys-Arg chloromethyl ketone. Incubation of 125I-iodotyrosylated Ala-Lys-Arg chloromethyl ketone with serum-free media harvested from cell cultures over a 12-h period revealed a time-dependent accumulation of a 54-kDa protease. Although small quantities of the 54-kDa protease were detected in cell lysates, the major intracellular sequences labeled by the affinity probe had masses of 31.5 and 6 kDa. The 54-kDa extracellular, as well as 31.5- and 6-kDa intracellular, species were all immunoprecipitated by monospecific anti-human liver cathepsin B IgG. Addition of this antibody to media inhibited extracellular conversion of proapo-A-II to the mature protein. Based on these observations, we conclude that a "pro" cathepsin B-like protease exported by Hep G2 cells is responsible for proapo-A-II prosegment removal. It appears that cathepsin B-like proteases exhibit a complex pattern of segregation within the secretory pathway and that larger molecular weight forms of cathepsin B-like proteases are capable of accurately processing propolypeptides.

Analyzing the structures, functions and evolution of two abundant gastrointestinal fatty acid binding proteins with recombinant DNA and computational techniques

The structures of intestinal and liver fatty acid binding proteins (FABPs) have been determined from an analysis of the nucleotide sequences of cloned cDNAs. The primary translation product of intestinal FABP mRNA contains 132 residues (Mr = 15 124). Liver FABP mRNA encodes a 127 amino acid polypeptide (Mr = 14 273). In vitro co-translational cleavage and translocation assays showed that neither sequence has a cleavable signal peptide or signal peptide equivalent - suggesting that the FABPs do not enter the secretory apparatus but rather are targeted to the cytoplasm. A variety of computational techniques were used to compare the two FABP sequences. The results indicate that liver and intestinal FABP are paralogous homologues. A superfamily of proteins was defined which includes the FABPs, the cellular retinol and retinoic acid binding proteins, the P2 protein of peripheral nerve myelin, and a polypeptide known as 422 whose synthesis is induced during differentiation of 3T3-L1 cells to adipocytes. No sequence homologies were noted between any of these small molecular weight cytosolic proteins and nonspecific lipid transfer protein (sterol carrier protein 2), phosphatidylcholine transfer protein, serum albumin or apolipoprotein AI. The FABPs may have structural features responsible for lipid-protein interactions that are not present in these non-homologous sequences. The distribution of intestinal and liver FABP mRNAs in adult rat tissues and the changes in FABP gene expression which occur during gastrointestinal development support the notion that these proteins are involved in fatty acid uptake, transport and/or compartmentalization. However, differences in tissue distribution and periods of non-coordinate expression during gastrointestinal ontogeny suggest that the two FABPs have distinct functions. The relationship between intestinal and liver FABPs and similar sized cytosolic FABPs isolated from brain, skeletal and cardiac muscle remains unclear. Recombinant DNA techniques combined with comparative sequence analyses offer a useful approach for defining unique as well as general structure-function relationships in this group of fatty acid binding proteins.

Human liver fatty acid binding protein. Isolation of a full length cDNA and comparative sequence analyses of orthologous and paralogous proteins

We have determined the primary structure of human liver fatty acid binding protein from an analysis of a full length cDNA. This 127-residue 14,178-Da protein exhibits a high degree of sequence conservation when compared to its orthologous homologue, rat liver fatty acid binding protein. It appears likely that this polypeptide arose from two intragenic duplication events. Using a variety of computational techniques, we were unable to find any evidence of amphipathic alpha helical domains in this protein nor any sequence similarities to apolipoproteins and serum albumins. A family of paralogous proteins was defined, whose members share a remarkable degree of sequence homology with share a remarkable degree of sequence homology with human liver fatty acid binding protein. These include rat intestinal fatty acid binding protein, the cellular the P2 protein of myelin. It appears that the small cytosolic fatty acid binding proteins have evolved structural features necessary for lipid-protein interaction which are different from those present in some familiar and better studied extracellular sequences.

Tissue specific expression and developmental regulation of two genes coding for rat fatty acid binding proteins

We have examined the tissue distribution and developmental regulation of two low molecular weight cytosolic fatty acid binding proteins. Based on their initial site of isolation, they have been referred to as liver and intestinal fatty acid binding proteins (FABP). Cloned cDNAs were used to probe blots of RNAs extracted from a wide variety of adult rat tissues as well as small intestine and liver RNA obtained from fetal, suckling, and weaning animals. The highest concentrations of "liver" FABP mRNA were found in small intestine and liver. "Intestinal" FABP mRNA is most abundant in small bowel RNA while only trace amounts were encountered in liver. Both mRNAs were detectable in stomach, colon, pancreas, spleen, lung, heart, testes, adrenal, and brain RNA at 1-8% the concentrations observed in small intestine. Accumulation of both mRNAs in the small intestinal epithelium increases during development. The mRNAs are first detectable between the 19th and 21st day of gestation. They undergo a coordinated 3-4-fold increase in concentration within the first 24 h after birth. Thereafter, gut levels of intestinal FABP mRNA remain constant during the suckling period while liver FABP mRNA increases an additional 2-fold. Liver FABP mRNA levels are also induced in hepatocytes during the first postnatal day but subsequently do not change during the suckling and weaning phase, despite marked alterations in hepatic fatty acid metabolism. These observations support the concept that the major role of these proteins is to facilitate the entry of lipids into cells and/or their subsequent intracellular transport and compartmentalization. The data also raise questions about the identity of extragastrointestinal FABPs.

Regulation of the biosynthesis of two distinct fatty acid-binding proteins in rat liver and intestine. Influences of sex difference and of clofibrate

Two distinct fatty acid-binding proteins (FABPs) have been identified in rat intestine, gFABP (15,063 Da) which is confined to intestinal epithelium and hFABP (14,184 Da) which is found in both liver and intestine. We have examined the influence of sex difference and the effect of clofibrate, both of which affect cellular fatty acid metabolism and hFABP levels, on the concentration, and mRNA levels of both hepatic and intestinal FABPs. In the liver, hFABP concentration was approximately 2-fold greater in females and in clofibrate-treated males than in untreated male rats. These differences were not accompanied by changes in the fractional turnover of the polypeptide but rather by parallel increases in hFABP mRNA. In the intestine, the two FABPs exhibited different regulatory responses. Intestinal hFABP turnover was 33% greater in females than in males, whereas mRNA concentration was 50% greater. Thus, unlike hFABP in liver, there was no sex-related difference in the steady-state level of hFABP in intestine. However, clofibrate treatment, similar to its effects in the liver, doubled intestinal hFABP protein and mRNA concentration. In contrast to hFABP, neither gFABP protein nor mRNA concentration were sex dependent, whereas clofibrate produced only a modest increase in gFABP concentration without significantly changing gFABP mRNA levels. The results indicate that the influence of sex difference and the effect of clofibrate on hepatic fatty acid metabolism are both associated with changes in hFABP synthesis mediated pretranslationally. The differential response of hFABP and gFABP in intestine suggests that these proteins play distinct roles in the cellular metabolism of fatty acids.

Comparative analysis of repeated sequences in rat apolipoproteins A-I, A-IV, and E

To understand the structural, functional, and evolutionary relationships among the principal protein components of rat high density lipoprotein particles, we undertook a systematic comparative analysis of the primary structures of apolipoproteins (apo)-A-I, -A-IV, and -E. Human apo-A-I and rat apo-A-IV have been shown previously to contain repeated sequences that presumably arose by intragenic duplication of 11- or 22-amino acid amphipathic segments. For apo-A-I, these segments are thought to be the structures responsible for lipid binding and activation of lecithin:cholesterol acyltransferase. From an analysis of the sequence of a full-length cDNA clone, rat apo-A-I is shown to contain eight tandem repetitions of a 22-amino acid segment. However, compared with human apo-A-I, the rat protein has undergone three deletions, two of which involve multiple amino acids in the repeated sequence domain. This disruption of the periodic structure of the protein raises the possibility of species-specific variation in the ability of rat apo-A-I to interact with high density lipoproteins and activate lecithin:cholesterol acyltransferase. Statistical analysis of the structure and organization of repeated sequences in apo-A-I, -A-IV, and -E demonstrates that all three proteins are paralogous members of a dispersed gene family. Despite overall similarity in sequence organization, different portions of these sequences have evolved at different rates. Diversification of a duplicated ancestral sequence has resulted in three lipid-binding proteins with distinct and shared functions.

Human proapolipoprotein A-II is cleaved following secretion from Hep G2 cells by a thiol protease

The two principal high-density lipoprotein apolipoproteins A-I and A-II are both initially synthesized as preproproteins. The prosegment of apo-A-I is unusual: it ends with paired glutamine residues and is removed extracellularly. The apo-A-II prosegment resembles the propeptides of prohormones and proalbumin: it ends with paired basic amino acids. We have studied the processing of proapo-A-II in a human hepatoma cell line (Hep G2) which is known to accurately and efficiently remove the prosegment from proalbumin prior to secretion. Pulse-chase experiments were performed in order to determine if the apo-A-II prosegment is removed prior to or after secretion. Apo-A-II was purified from cell lysates and media at various times during the chase and subjected to automated sequential Edman degradation. The results indicate that proteolytic processing of proapo-A-II is largely an extracellular event. These cells secrete the protease responsible for prosegment removal. The converting activity present in media is not blocked by serine protease inhibitors (phenylmethanesulfonyl fluoride, aprotinin, and furoyl saccharin) or by a metalloprotease inhibitor (o-phenanthroline). It is inhibited by the thiol protease reagents p-chloromercuribenezene-sulfonic acid and leupeptin. Prosegment removal changes the pI of the dominant apo-A-II isoform from 6.61 to 4.95. The presence of the propeptide does not prevent specific in vitro recombination of apo-A-II with high-density lipoprotein3 particles present in normolipemic serum. Extracellular processing after a single basic amino acid has been described for a variety of precursor proteins. Extracellular cleavage of the apo-A-II propeptide after paired COOH-terminal basic residues represents a novel processing pathway.