Adenovirus-mediated expression of human prorelaxin promotes the invasive potential of canine mammary cancer cells

This study reports the characterization of a recombinant adenoviral vector containing a tetracycline-regulatable promoter, driving the bicistronic expression of the human H2 preprorelaxin (hH2) cDNA and enhanced green fluorescent protein, via an internal ribosomal entry site. An hH2 ELISA was used to measure the secreted levels of recombinant hH2 in transfected canine (CF33.Mt) and human (MDA-MB-435) mammary cancer cell lines over a 6-d period; secreted peptide peaked on d 2 and 4 for the canine and human cell types, respectively. An unprocessed hH2 immunoreactive form of approximately 18 kDa was identified by Western blotting analysis and confirmed by mass spectrometry, suggesting that prorelaxin remains unprocessed in these cell types. The biological activity of the adenovirally expressed human prorelaxin was measured in the established human monocytic cell line THP-1 cAMP ELISA and in an in vitro Transwell cell migration system. Exogenous recombinant hH2 and adenovirally-mediated delivery of prorelaxin to CF33.Mt cells conferred a significant migratory action in the cells, compared with controls. Cell proliferation assays were performed to discount the possibility that the effect of relaxin was mitogenic. Thus, we have demonstrated that prorelaxin has the ability to facilitate cell migration processes exclusive of its ability to stimulate cell proliferation. In validating this adenovirus-based system, we have created a potential tool for further exploration of the physiology of relaxin in mammalian systems.     

cDNA clones encoding IgE-binding factors from a rat-mouse T-cell hybridoma.

cDNA clones encoding rodent IgE-binding factors (IgE-BF) were isolated from cDNA libraries of a rat-mouse T hybridoma that secretes IgE-suppressive factor (IgE-SF) upon incubation with rat IgE. COS7 cells transfected with two of the cDNAs expressed IgE-BF, which selectively potentiate an in vitro IgE response. IgE-BF expressed in COS7 cells are glycoproteins of approximately equal to 60 and approximately equal to 11 kDa. DNA sequence analysis of an IgE-BF cDNA revealed a 556-amino acid (62 kDa) protein coding region. The results suggest that IgE-potentiating and IgE-suppressive factors share common precursor polypeptides and that the 11-kDa IgE-BF is derived from a 60-kDa precursor.     

Structure of rhesus monkey relaxin predicted by analysis of the single-copy rhesus monkey relaxin gene

The gene encoding rhesus monkey relaxin has been investigated. A cDNA library was prepared using corpus luteal RNA from a pregnant rhesus monkey, cDNA clones encoding relaxin were isolated and the nucleotide sequence was determined. The amino acid sequence of rhesus monkey preprorelaxin, predicted from the cDNA, demonstrates that the sequence has not been strongly conserved when compared with that of man, although features characteristic of the relaxin molecule have been maintained. This structural information will allow production of rhesus monkey relaxin, leading to studies investigating the bioactivity of relaxin in a homologous primate system. Southern blot analysis indicated that there is only one relaxin gene in the rhesus monkey and baboon genomes. In this respect these primate genomes are different from the human genome which contains two relaxin genes.     

Development of a homologous radioimmunoassay for rat relaxin.

Highly purified rat relaxin has been radioiodinated to specific activities of approximately 100 micro Ci/microgram with the Bolton and Hunter reagent [N-succinimidyl 3-(4-hydroxy-5-[125I]iodophenyl) propionate]. A rabbit antirat relaxin serum, applicable in a final dilution of 1:100,000, was developed in a rabbit using unconjugated highly purified rat relaxin. A specific and precise double antibody RIA for rat relaxin sufficiently sensitive to routinely measure from 32--2000 pg rat relaxin was developed. Using this RIA, relaxin immunoactivity levels in extracts of pregnant rat ovaries were found to rise from 0.8 microgram/geq ovarian fresh tissue on day 8 of pregnancy to 723 microgram/geq ovarian fresh tissue on day 20 of pregnancy and then to drop precipitously to 6 microgram/geq ovarian fresh tissue on day 1 of lactation. Consistent with the occurrence and relative levels of relaxin in the ovarian extracts, levels of relaxin in pregnant rat serum were less than 2 ng/ml on day 10 of pregnancy, approximately 150 ng/ml on days 20 and 22 of pregnancy, and 12 ng/ml on day 2 of lactation.     

The general mitochondrial matrix processing protease from rat liver: structural characterization of the catalytic subunit.

A critical step in the import of nuclear-encoded precursor proteins into mitochondria involves proteolytic cleavage of their amino-terminal leader peptides by processing proteases found in the mitochondrial matrix. We report here the characterization of the general matrix processing protease from rat liver mitochondria. The final enzyme preparation consisted of two polypeptides, a catalytically active 55-kDa subunit and a 52-kDa one. To deduce the complete primary structure of the 55-kDa subunit, we first sequenced its mature amino terminus and several tryptic peptides derived from the pure protein. Next, using mixed oligonucleotide primers that had sequences based on two of these peptides, we synthesized a partial cDNA probe by selective amplification of liver RNA with the polymerase chain reaction. The amplified probe was then used to obtain a nearly full-length clone from a rat liver cDNA library. This cDNA codes for 508 amino acid residues, including 16 residues of an amino-terminal leader peptide, the cleavage site of which is located two polypeptide bonds downstream from an arginine residue. The mature portion has a predicted molecular mass of 55.2 kDa; it shows 36% identity with the mitochondrial processing peptidases of Saccharomyces cerevisiae and Neurospora crassa. A conserved structural feature is a putative, negatively charged alpha-helix, located in the amino-terminal half of the subunit; this element might be important for the recognition of positively charged leader peptides characteristic of mitochondrial precursor proteins.     

Molecular cloning and deduced amino acid sequence of nonspecific lipid transfer protein (sterol carrier protein 2) of rat liver: a higher molecular mass (60 kDa) protein contains the primary sequence of nonspecific lipid transfer protein as its C-terminal part.

Two types of cDNA for nonspecific lipid transfer protein (nsLTP), identical to sterol carrier protein 2, of rat liver were cloned; one was 787 base pairs (bp) long containing a 429-bp open reading frame of 143 amino acids, with a mass of 15,303 Da (15-kDa protein). The cDNA from the other type was 1966 bp long, including a 1641-bp open reading frame of 547 amino acids, giving a mass of 59,002 Da (60-kDa protein). The deduced primary sequence for the 15-kDa protein was exactly the same as the published sequence of purified nsLTP, except for an extra N-terminal sequence of 20 amino acids, consistent with the finding that nsLTP is synthesized as a larger precursor and processed to a mature form. The sequence for the 60-kDa protein contained, at the 3' end, the full sequence of the 15-kDa protein, a larger precursor to nsLTP. The 15- and 60-kDa proteins, synthesized in vitro from the respective cDNAs, were both immunoprecipitated by rabbit anti-rat liver nsLTP antibody and comigrated in SDS/PAGE with the proteins made in vitro from total liver RNA. These results shed new light on the dispute among several groups of investigators about the crossreactivity of anti-nsLTP antibody with a higher molecular mass, 60-kDa protein. In Northern blot analysis, two major RNA bands, 0.85 and 2.2 kilobases (kb) long, were detected together with two minor bands of 1.6 and 2.9 kb. The 0.85- and 2.2-kb RNAs most likely encode the 15-and 60-kDa proteins, respectively.     

The 37/40-kilodalton autoantigen in insulin-dependent diabetes mellitus is the putative tyrosine phosphatase IA-2.

Major targets for autoantibodies associated with the development of insulin-dependent diabetes mellitus (IDDM) include tryptic fragments with a molecular mass of 37 kDa and/or 40 kDa of a pancreatic islet cell antigen of unknown identity. The assay identifying autoantibodies against the 37/40-kDa antigen in human sera is based on the immunoprecipitation of 35S-labeled rat insulinoma cell proteins with sera from IDDM patients, followed by limited trypsin digestion of the immunoprecipitated material. To identify cDNA clones coding for the 37/40-kDa antigen, we have screened a cDNA expression library from rat insulinoma cells with a serum from an IDDM patient that precipitated the 37/40-kDa antigen in our assay. Among the cDNA products that reacted with the IDDM serum, we identified one cDNA clone whose open reading frame encodes a protein with a predicted mass of 105 kDa that we termed "ICA105" for 105-kDa islet cell antibody. The deduced amino acid sequence has high homology to a recently cloned putative tyrosine phosphatase IA-2 from human and mouse cDNA libraries. Translation of the cDNA in vitro results in a polypeptide with the expected molecular mass of 105 kDa. The evidence that ICA105 is indeed the precursor of the 37/40-kDa tryptic fragments is based on the following three results: (i) Sera from IDDM patients containing autoantibodies to the 37/40-kDa antigen precipitate the in vitro translated polypeptide, whereas sera from healthy subjects as well as sera from IDDM patients not reactive with the 37/40-kDa antigen do not precipitate the cDNA product. (ii) Immunoprecipitation of the in vitro translated protein with sera containing autoantibodies to the 37/40-kDa antigen followed by limited trypsin digestion of the precipitated proteins results in a 40-kDa polypeptide. (iii) The protein derived from our cDNA but not from an unrelated control cDNA clone can block immunoprecipitation of the 37/40-kDa antigen from a labeled rat insulinoma cell extract. The availability of the cloned 37/40-kDa antigen should facilitate the identification of individuals at risk of IDDM with increased accuracy. Furthermore, the identification of the 37/40-kDa antigen as the putative tyrosine phosphatase IA-2 is of relevance in elucidating the role of this antigen in the development of IDDM.     

Chromogranin-B, a putative precursor of eight novel rat glucagonoma peptides through processing at mono-, di-, or tribasic residues

Chromogranin-B (CgB), a secretory granule protein, is normally synthesized in a variety of neuroendocrine tissues, including the pancreatic islet alpha-cells. We have demonstrated that rat CgB is expressed and extensively processed by limited proteolysis in a transplantable glucagonoma tumor line. Eight peptides (fragments 1-8) purified by HPLC from acidic tumor extracts were partially sequenced and showed homology to CgB amino acid sequences deduced from rat, mouse, and human cDNA. Similar peptides were not found in insulinomas of common origin. The determined amino acid sequence represents approximately 35% of the rat precursor CgB. Ten of a total of 231 sequenced residues deviated from the published rat cDNA sequence. The differences were clustered in 3 fragments, suggesting allelic polymorphism. Five of the 8 peptides could be derived from the precursor by processing at paired basic amino acids, but processing N-terminally at a single basic residue was also seen. One peptide equivalent to the previously reported C-terminal CgB (CCB) is released by processing at a tribasic segment. Fragment 1 containing the N-terminal sequence Ala-Pro-Val-Asp represents the actual N-terminus of CgB after removal of the putative signal peptide sequence. All processing sites used in glucagonoma tissue to derive the 8 isolated fragments were conserved between murine and human CgB. At least 3 dibasic sites present in rat, but not human, CgB sequence were actually not used. A previously reported CgB-derived pituitary peptide, GAWK, was further processed at a conserved internal dibasic site to yield fragment 6, indicating alternative processing in different tissues. The small undecapeptide, fragment 7, is 100% conserved among murine and human CgB and, thus, may have an important biological function. We conclude that CgB is extensively processed in glucagonoma tissue by limited proteolysis as prohormones at conserved basic residues. The proglucagon-converting enzymes present in transformed alpha-cells are likely candidates to be involved in tissue-specific CgB processing. Distinct biological activities of any of the CgB-derived fragments remain to be identified.     

Characterization of the rat neutral and basic amino acid transporter utilizing anti-peptide antibodies.

High-titer, site-specific antibodies have been produced against the rat kidney broad-spectrum, sodium-independent neutral and basic amino acid transporter (NBAA-Tr) whose cDNA we cloned earlier. These antibodies have allowed us to characterize the transporter protein in normal rat tissues and in various cellular and in vitro expression systems. Western analysis detected 84- to 87-kDa glycosylated species enriched in rat renal and jejunal epithelial cell brush border membranes. In vitro translation of NBAA-Tr complementary RNA in the rabbit reticulocyte lysate system yielded a 78-kDa protein, a molecular mass that was predicted by the amino acid sequence deduced from the cloned cDNA. Translation in the presence of rough microsomal membranes yielded a glycosylated 89-kDa species. Glycosylated 87- to 89-kDa species were also expressed in Xenopus oocytes microinjected with NBAA-Tr complementary RNA and in COS-7 cells transfected with NBAA-Tr cDNA. Localization of NBAA-Tr in renal and intestinal brush border membranes is consistent with its proposed role in transepithelial transport of amino acids.     

Cloning, characterization, and autoimmune recognition of rat islet glutamic acid decarboxylase in insulin-dependent diabetes mellitus.

A 64-kDa islet protein is a major autoantigen in insulin-dependent diabetes mellitus (IDDM). Autoantibodies against the 64-kDa protein were recently shown to immunoprecipitate glutamic acid decarboxylase (GAD; L-glutamate 1-carboxy-lyase, EC 4.1.1.15) from brain and from islets. We present evidence that the autoantisera also recognize a hydrophilic islet protein of approximately 67 kDa in addition to the amphiphilic 64-kDa form. We have isolated a full-length rat islet GAD cDNA encoding a hydrophilic 67-kDa protein, which appears to be identical to rat brain 67-kDa GAD. A partial sequence of human insulinoma 67-kDa GAD was identical to human brain 67-kDa GAD. Allelic variations were observed in rat as well as in human 67-kDa GAD sequences. The expressed rat islet 67-kDa GAD protein is functional and is immunoprecipitated by IDDM sera; it comigrates electrophoretically with the 67-kDa islet autoantigen. The hydrophilic 67-kDa form of GAD in islets is an additional autoantigen in IDDM and is recognized by a different subset of autoantibodies than the 64-kDa autoantigen. Thus, mammalian cell lines expressing functionally active, recombinant GAD may become important tools to study the nature and the role of GAD autoreactivity in IDDM.     

Evidence for a dual source of relaxin in the pregnant rat: immunolocalization in the corpora lutea and endometrium.

Uterine tissue was collected from intact pregnant rats on days 4, 10-15, 18, and 22 of pregnancy and day 2 postpartum and from rats on day 22 of pregnancy after ovariectomy (day 9) and progesterone-estrogen treatment. Placental, cervical, mammary, and pituitary tissues were collected from intact pregnant rats on day 22. Ovarian tissue was collected from intact pregnant rats on days 15, 20, and 22. These tissues were evaluated for relaxin immunostaining at the light microscope level using an antiserum to purified rat relaxin and the avidin biotin immunostaining technique. Since the corpus luteum is the major source of relaxin in the rat, this tissue was used as a positive control. Relaxin immunostaining in the corpus luteum was observed in a discrete region of the cytoplasm. This pattern of staining corresponded with the discrete clustering of relaxin-containing secretory granules found in the rat luteal cells. Relaxin was not observed in the day 22 placenta, cervix, mammary gland, pituitary gland, or day 4-13 pregnant uterus. Relaxin immunostaining was detected in endometrial epithelial cells on days 14-22 of pregnancy and day 2 postpartum. Relaxin immunostaining was more evident at implantation sites than between implantation sites. Antimesometrial epithelial cells at implantation sites contained relaxin immunostaining. The mesometrial surface, which consists of the placenta and decidualized endometrium, did not contain relaxin immunostaining. Relaxin immunostaining was also present in the endometrial epithelial cells of the day 22 pregnant rats that had been ovariectomized on day 9 and given subsequent replacement therapy with ovarian steroids. The presence of relaxin immunostaining in the ovariectomized rat indicates the endometrium is not sequestering relaxin secreted from the corpus luteum. These data provide evidence that the pregnant rat endometrium is a source of relaxin in addition to the corpus luteum. The appearance of relaxin in endometrial epithelial cells after implantation and its prominence at implantation sites may indicate that locally secreted conceptus factors stimulate endometrial relaxin synthesis.     

Human carbonic anhydrase IV: cDNA cloning, sequence comparison, and expression in COS cell membranes.

We have isolated a full-length cDNA for human carbonic anhydrase IV (CA IV) from a lambda gt10 human kidney cDNA library. The 1105-base-pair (bp) cDNA contains a 47-bp 5' untranslated region, a 936-bp open reading frame, and a 122-bp 3' untranslated region. The deduced amino acid sequence is colinear with the N-terminal sequence and the sequence of several tryptic peptides of human lung CA IV. It includes an 18-amino acid signal sequence, a 260-amino acid region that shows 30-36% similarity with the 29-kDa cytoplasmic CAs (CA I, CA II, and CA III), and an additional 27-amino acid C-terminal sequence that ends in a 21-amino acid hydrophobic domain. Of the 17 "active site" residues that are highly conserved in other human CAs, 16 are also present in CA IV. Expression of the cDNA in COS cells produced a 35-kDa enzyme that was membrane associated, resistant to inactivation by SDS, contained no carbohydrate, and reacted on Western blots with antiserum to the 35-kDa CA IV from human lung. Treatment of membranes from transfected COS cells with phosphatidylinositol-specific phospholipase C released 20-30% of the expressed enzyme from membranes, indicating that at least 20-30% of the expressed enzyme was anchored to membranes by a glycosyl-phosphatidylinositol linkage.     

Isolation and characterization of a cDNA clone encoding an alternative oxidase protein of Sauromatum guttatum (Schott).

Polyclonal and monoclonal antibodies that recognize the 35-, 36-, and 37-kDa alternative oxidase proteins of Sauromatum guttatum (Schott) were used to isolate a cDNA clone, pAOSG81, from an S. guttatum cDNA expression library. A fusion protein with an apparent molecular mass of 48 kDa was expressed from a pUC119 derivative of pAOSG81 (pAOSG81-119) in Escherichia coli cells and was recognized by the monoclonal antibodies. When the in vitro translated and immunoprecipitated products made from mRNA hybrid-selected by pAOSG81 were analyzed, a single band corresponding to a protein with an apparent molecular mass of 42 kDa was observed. DNA sequence characterization showed that pAOSG81 contains the entire coding region of a protein with a calculated molecular mass of 38.9 kDa, a putative 63-amino acid transit peptide, and a 9-amino acid match to the authentic N-terminal sequence of the 36-kDa alternative oxidase protein. Analyses of the deduced amino acid sequence indicate: (i) that the transit peptide is predicted to form amphiphilic helices, and (ii) that three regions of the processed protein are likely to form transmembrane alpha-helices. We conclude from these data that pAOSG81 represents a nuclear gene, aox1, encoding a precursor protein of one or more of the alternative oxidase proteins of S. guttatum.     

Human mitochondrial carbonic anhydrase: cDNA cloning, expression, subcellular localization, and mapping to chromosome 16.

A full-length cDNA clone encoding human mitochondrial carbonic anhydrase (CA), CA V, was isolated from a human liver cDNA library. The 1123-bp cDNA includes a 55-bp 5' untranslated region, a 915-bp open reading frame, and a 153-bp 3' untranslated region. Expression of the cDNA in COS cells produced active enzyme. The 34-kDa precursor and 30-kDa mature form of CA V were identified on Western blots of COS-cell homogenates by a CA V-specific antibody raised to a synthetic peptide corresponding to the C-terminal 17 aa of CA V. Both 34-kDa and 30-kDa bands were also present in mitochondria isolated from transfected COS cells, whereas only the 30-kDa band was present in mitochondria isolated from normal human liver. The N-terminal sequence determined directly on the 30-kDa soluble CA purified from transfected COS cells indicated that processing of the precursor to mature human CA V involves removal of a 38-aa mitochondrial leader sequence. The 267-aa sequence deduced for mature human CA V shows 30-49% similarity to amino acid sequences of previously characterized human CAs (CA I-CA VII) and 76% similarity to the corresponding amino acid sequence deduced from the mouse cDNA. PCR analysis of DNAs from human-rodent somatic cell hybrids localized the gene for CA V to human chromosome 16, the same chromosome to which CA VII has previously been mapped.