Genomic organization of human neuropilin-1 and neuropilin-2 genes: identification and distribution of splice variants and soluble isoforms.

Neuropilin-1 (NRP1) and neuropilin-2 (NRP2) are both receptors for semaphorins, which regulate neuronal guidance, and vascular endothelial growth factor (VEGF), an angiogenic factor. The two human NRP1 and NRP2 genes were cloned, and the exon-intron boundaries were determined. The NRP1 and NRP2 genes span over 120 and 112 kb, respectively, and are composed of 17 exons. Five of the exons are identical in size in the two genes, suggesting that they arose by gene duplication. Both NRP genes are characterized by multiple alternatively spliced variants. Two NRP2 isoforms, NRP2a and NRP2b, were cloned. A striking feature of these two isoforms is that they have identical extracellular domains but have divergent transmembrane and cytoplasmic domains. In these domains, NRP2a is closer in sequence identity to NRP1 than to NRP2b. As determined by Northern blot analysis, both NRP2a and NRP2b are expressed in a variety of tissues, mostly in a nonoverlapping manner. Within NRP2a and NRP2b, there are several alternatively spliced species: NRP2a(17), NRP2a(22), NRP2b(0), and NRP2b(5). In addition to full-length NRPs, there are truncated NRPs as well, which contain only the extracellular a/CUB and b/coagulation factor domains. These genes encode proteins that are soluble (sNRP) and released by cells. In addition to s12NRP1, which was previously cloned, s11NRP1 and s9NRP2 have now been cloned. These sNRP molecules are characterized by having intron-derived sequences at their C-termini. Altogether, eight NRP isoforms are described in this report. It was concluded that there are multiple NRP1 and NRP2 isoforms including intact and soluble forms. Characterization of these isoforms should help to elucidate the function of NRPs in neuronal guidance and angiogenesis.

Identification of a natural soluble neuropilin-1 that binds vascular endothelial growth factor: In vivo expression and antitumor activity.

Neuropilin-1 (NRP1) is a 130-kDa transmembrane receptor for semaphorins, mediators of neuronal guidance, and for vascular endothelial growth factor 165 (VEGF(165)), an angiogenesis factor. A 2.2-kb truncated NRP1 cDNA was cloned that encodes a 644-aa soluble NRP1 (sNRP1) isoform containing just the a/CUB and b/coagulation factor homology extracellular domains of NRP1. sNRP1 is secreted by cells as a 90-kDa protein that binds VEGF(165), but not VEGF(121). It inhibits (125)I-VEGF(165) binding to endothelial and tumor cells and VEGF(165)-induced tyrosine phosphorylation of KDR in endothelial cells. The 3' end of sNRP1 cDNA contains a unique, 28-bp intron-derived sequence that is absent in full-length NRP1 cDNA. Using a probe corresponding to this unique sequence, sNRP1 mRNA could be detected by in situ hybridization differentially from full-length NRP1 mRNA, for example, in cells of liver, kidney, skin, and breast. Analysis of blood vessels in situ showed that NRP1, but not sNRP1, was expressed. sNRP1 was functional in vivo. Unlike control tumors, tumors of rat prostate carcinoma cells expressing recombinant sNRP1 were characterized by extensive hemorrhage, damaged vessels, and apoptotic tumor cells. These results demonstrate the existence of a naturally occurring, soluble NRP1 that is expressed differently from intact NRP1 and that appears to be a VEGF(165) antagonist.

Characterization of the promoter for the human antisense fibroblast growth factor-2 gene; regulation by Ets in Jurkat T cells.

Human lymphoid cells were found to synthesize predominantly antisense, and not sense, fibroblast growth factor-2 (FGF-2) mRNA. Two cDNAs corresponding to human 1069- and 1173-nucleotide antisense FGF-2 mRNAs were cloned from Jurkat T cells. The two cDNAs each possess a unique exon 1 and common exon 2, 3, 4, and 5 sequences. Exon 4 and 5 sequences overlap in the 3' untranslated region of FGF-2 cDNA, but not in the FGF-2 open reading frame. This is unlike the Xenopus antisense FGF-2 homologue, which overlaps with parts of both the FGF-2 3' untranslated region and its open reading frame. To investigate the regulation of human antisense FGF-2 gene expression, a 2.5-kilobase (kb) promoter region was isolated and characterized. Transient transfection of promoter-luciferase constructs demonstrated the antisense FGF-2 promoter to be active in Jurkat cells. Using transient transfection and in vitro binding assays, specific mutations within the promoter sequence have implicated that Ets-like transcription factors are significant in regulating the human antisense FGF-2 gene in Jurkat cells.

Characterization of a SpAN promoter sufficient to mediate correct spatial regulation along the animal-vegetal axis of the sea urchin embryo.

In order to investigate how the maternally specified animal-vegetal axis of the sea urchin embryo is established, we have examined the molecular basis of regulation of several genes transcribed differentially in nonvegetal and vegetal domains of the very early blastula. Here we present an initial characterization of the regulatory region of one of these, SpAN, which encodes a protease in the astacin family related to Drosophila tolloid and vertebrate BMP-1 (Reynolds et al., Development 114, 769-786). Tests of SpAN promoter function in vivo show that high-level activity and correct not-vegetal expression are mediated by sequences within 300 bp upstream of the basal promoter. In vitro studies have identified six protein binding sites serviced by at least five different proteins. Comparison of the structure of the SpAN promoter to that of SpHE, whose expression pattern is identical, shows that both promoters contain multiple positively acting upstream elements close to the basal promoter. We show that two elements are critical for high-level transcription of SpAN, since exact replacement of either results in 10- to 20-fold reduction in promoter strength. These shared elements are, however, not essential for spatially correct SpHE gene transcription. We conclude that the coordinate strong activities of the SpAN and SpHE promoters in the nonvegetal domain of the embryo rely primarily on different transcription factor activities.

Characterization of the SpHE promoter that is spatially regulated along the animal-vegetal axis of the sea urchin embryo.

To understand how the maternally determined animal-vegetal polarity of the sea urchin embryo is established, we have begun to examine the regulatory apparatus of the gene encoding the Strongylocentrotus purpuratus hatching enzyme (SpHE). Previous studies have shown that the pattern of SpHE mRNA accumulation reflects the animal-vegetal developmental axis in that transcription is strongly upregulated during early cleavage in more animal blastomeres, but not in those around the maternally specified vegetal pole of the 16-cell embryo [Reynolds et al., Development 114, 769-786 (1992)]. Tests of SpHE promoter function in vivo using chloramphenicol acetyltransferase and beta-galactosidase enzymatic reporters define a regulatory region within several hundred nucleotides of the transcription initiation site. This region is sufficient to mediate both strong expression in the early blastula and spatially correct transcription. However, neither this region nor longer upstream sequences are sufficient to reproduce the transcriptional downregulation after very early blastula stage that is observed for endogenous genes. Biochemical assays of protein-DNA interactions within the regulatory region identify at least nine sites binding at least six different factors. These cis elements include Otx (an orthodenticle homologue), CCAAT, ets-related, and three unidentified motifs. Deletions and/or replacements of these cis-elements, alone and in combination, indicate that no single factor is essential for SpHE promoter activity, but instead that various combinations of subsets of these elements are capable of eliciting levels of transcription similar to those of the unaltered regulatory region. This density of regulatory elements is consistent with the intense transcription of endogenous SpHE genes during cleavage.

Peripheral blood T lymphocytes and lymphocytes infiltrating human cancers express vascular endothelial growth factor: a potential role for T cells in angiogenesis.

CD3+ peripheral blood T lymphocytes were evaluated for expression of vascular endothelial growth factor (VEGF), an endothelial cell mitogen and potent angiogenic factor. VEGF mRNA expression was confirmed in CD3+ cells and Jurkat cells, a human T-cell line, by reverse transcription-PCR and in CD4+ and CD8+ T cell subtypes by Northern blot hybridization. Steady-state levels of VEGF mRNA were inducible in CD3+ T cells by hypoxia, a known inducer of VEGF mRNA accumulation. Secreted VEGF was detected in CD4+ and CD8+ T cell- and Jurkat cell-conditioned medium, indicating that T lymphocytes are capable of exporting bioactive concentrations of VEGF into the extracellular space. Human prostate and bladder cancers (prostatic adenocarcinoma and transitional cell carcinomas) were evaluated for VEGF mRNA expression by in situ hybridization. Tumor-infiltrating lymphocytes (TIL), identifiable immunocytochemically as T cells, along with tumor cells in these cancers, expressed VEGF mRNA. TIL in bladder cancers could be labeled with a specific anti-VEGF mAb, indicating that TIL are likely to be able to secrete VEGF protein in situ at bioactive concentrations. The finding that peripheral T cells and TIL in human tumors synthesize a factor known to be a specific mediator of neovascularization suggests a role for T lymphocytes as cellular effectors of angiogenesis.

Posttranscriptional regulation of ectoderm-specific gene expression in early sea urchin embryos.

During development of the sea urchin Strongylocentrotus purpuratus embryo, transcription of the Spec1 and actin CyIIIa genes is activated and the corresponding mRNAs accumulate specifically in ectoderm cells. We show that in gastrulae this tissue specificity of mRNA accumulation is regulated largely if not entirely at a posttranscriptional level. We used RNAase protection assays with intron and exon probes to measure the levels of nuclear precursors and mature message, respectively, in total RNA from embryo fractions enriched for ectoderm (Ect) or endoderm+mesenchyme (E/M) cells. These measurements demonstrate that E/M cells, which do not accumulate Spec1 and actin CyIIIa mRNAs, contain high levels of intron transcripts, indicating that cells of the E/M tissues transcribe these genes. At later stages, transcripts containing intron sequences are restricted to ectoderm cells. These results indicate that there is a transition from posttranscriptional to transcriptional regulation of tissue-specific mRNA accumulation during the gastrula stage. Measurements of transcription rate by nuclear run-on assays substantiate this conclusion for Spec1 and extend it to two other genes, SpEGFI and Spec2c, which also encode ectoderm-specific mRNAs. Posttranscriptional regulation was not observed for the SM50 gene whose mRNA accumulates only in primary mesenchyme cells, or for actin CyI which is expressed predominantly in E/M cells of gastrulae.

Expression of two mRNAs encoding EGF-related proteins identifies subregions of sea urchin embryonic ectoderm.

Many proteins containing domains related to epidermal growth factor (EGF) function in intercellular interactions that mediate specification of cell fate. We have used in situ hybridization to show that the expression of two EGF-related genes (SpEGF I and SpEGF II) is restricted to the same subset of ectodermal cells in sea urchin pluteus larvae. However, the concentration of EGF I mRNA in different epithelial cells of aboral ectoderm and postoral facial epithelium is constant while that of EGF II mRNA is highly modulated. RNase protection assays show that both genes are activated during the period when ectoderm funder cells are established, i.e., between fourth and fifth and between fifth and sixth cleavages for EGF I and EGF II, respectively. By mesenchyme blastula stage EGF I mRNA reaches maximum abundance (800-1000 copies/expressing cell) as a result of a high transcription rate, while EGF II mRNA peaks at about half that concentration by gastrula stage. EGF I expression begins at early stages of oogenesis while EGF II expression appears to be confined to embryogenesis.

Studies of the cyclic adenosine monophosphate chemoreceptor of Paramecium.

A doublet of proteins (approximately 48,000 Mr) from the Paramecium cell body membrane fits several criteria for the external cAMP chemoreceptor. These criteria include: (i) selective elution from a cAMP affinity column, matching a specificity that could be predicted from the behavioral response and whole-cell binding; (ii) binding to wheat germ agglutinin indicating the presence of carbohydrate moieties indicating surface exposure; and (iii) selective inhibition of the intact cells' chemoresponse to cAMP by antibodies against the doublet. Additional evidence for the existence of a receptor, in general, comes from selective elimination of the cAMP chemoresponse by photoaffinity labeling of while cells with 8-N3-cAMP. The doublet proteins are not identical to the regulatory subunit of a cAMP-dependent protein kinase from Paramecium, the Dictyostelium cAMP chemoreceptor, or the 42-45 kDa range proteins related to the large surface glycoprotein in Paramecium. The doublet proteins are not readily separable and, as in Dictyostelium, may represent two different covalent modification states of the same protein. Amino acid analysis indicates that the proteins are similar, but does not distinguish between the possibilities of proteolysis and covalent modification. Once cloned, this doublet may prove to be only the fifth external, eukaryotic chemoreceptor to be identified.

Progressively restricted expression of a homeo box gene within the aboral ectoderm of developing sea urchin embryos.

A homeo box-containing gene, Hbox1 is expressed in an unusual and highly conserved spatial pattern in embryos of two different species of sea urchin, Tripneustes gratilla and Strongylocentrotus purpuratus. Hybridization in situ shows that this mRNA accumulates initially throughout the aboral ectoderm; however, between blastula and pluteus stages, the region containing Hbox1 mRNA retracts gradually until only a small area around the vertex is labeled in pluteus larvae. Aboral ectoderm appears cytologically uniform and also accumulates uniform levels of other tissue-specific mRNAs. Therefore, the Hbox1 pattern reveals a previously unsuspected heterogeneity of aboral ectoderm cells and a polarity within this tissue. In S. purpuratus, the Hbox1 gene product probably is not involved in initial specification of cell fate, as this message does not achieve a significant fraction of its peak abundance until almost hatching blastula stage, well after the time aboral ectoderm cells have initiated a tissue-specific program of gene expression. RNA blot and RNase protection analyses revealed low levels of Hbox1 mRNA in all adult tissues examined. However, this message was not detectable in mature eggs, suggesting that the Hbox1 gene does not have a maternal function. In addition to highly conserved spatial and temporal patterns of expression, the homeo box genes of these two urchin species also are conserved highly in sequences outside the homeo domain, despite the divergence of these two species (30-45 my). Two notable features of the protein shared with several vertebrate homeo proteins are a short conserved sequence encoded by an exon upstream of that encoding the homeo domain and a large region of high serine and proline content.

Correlations between cyclic AMP binding and chemoreception in Paramecium.

Paramecium tetraurelia is attracted to cyclic AMP, which probably, as other attractants, signifies the presence of food. Attraction to cyclic AMP was specific, saturable, and, therefore, likely to be receptor-mediated. In these studies, we measured the binding of cyclic [3H]AMP to whole cells and found it to be saturable, reversible, and displaying specificity similar to that of attraction. An HPLC method of separating nucleotides was devised and used to determine that external cyclic AMP was degraded in the absence of IBMX, a phosphodiesterase inhibitor, and that cyclic AMP was taken into the cells in small amounts. Since binding and attraction were subsequently measured in the presence of IMBX, it was cyclic AMP and not a degradation product that served as the attractant stimulus for Paramecium.