Identification of differentially expressed genes in human prostate cancer using subtraction and microarray.

We have identified human prostate cancer- and tissue-specific genes using cDNA library subtraction in conjunction with high throughput microarray screening. Subtracted cDNA libraries of prostate tumors and normal prostate tissue were generated. Characterization of subtracted libraries showed enrichment of both cancer- and tissue-specific genes. Highly redundant clones were eliminated by colony hybridization. The remaining clones were selected for microarray to determine gene expression levels in a variety of tumor and normal tissues. Clones showing overexpression in prostate tumors and/or normal prostate tissues were selected and sequenced. Here we report the identification of two genes, P503S and P504S, from subtracted libraries and a third gene, P510S, by subtraction followed by microarray screening. Their expression profiles were further confirmed by Northern blot, real-time PCR (TaqMan), and immunohistochemistry to be overexpressed in prostate tissues and/or prostate tumors. Full-length cDNA sequences were cloned, and their subcellular locations were predicted by a bioinformatic algorithm, PSORT, to be plasma membrane proteins. The genes identified through these approaches are potential candidates for cancer diagnosis and therapy.

M-Ras, a widely expressed 29-kD homologue of p21 Ras: expression of a constitutively active mutant results in factor-independent growth of an interleukin-3-dependent cell line.

M-Ras, a recently identified homologue of p21 Ras, is widely expressed, with levels of the 29-kD protein in spleen, thymus, and NIH 3T3 fibroblasts equaling or exceeding those of p21 Ras. A G22V mutant of M-Ras was constitutively active and its expression in an interleukin-3 (IL-3)-dependent mast cell/megakaryocyte cell line resulted in increased survival in the absence of IL-3, increased growth in IL-4, and, at high expression levels, in factor-independent growth. Expression of M-Ras G22V, however, had a negative effect on growth in the presence of IL-3, suggesting that M-Ras has both positive and negative effects on growth. Expression of M-Ras G22V in NIH-3T3 fibroblasts resulted in morphological transformation and growth to higher cell densities. M-Ras G22V induced activation of the c-fos promoter, and bound weakly to the Ras-binding domains of Raf-1 and RalGDS. Expression of a mutant of M-Ras G22V that was no longer membrane-bound partially inhibited (40%) activation of the c-fos promoter by N-Ras Q61K, suggesting that M-Ras shared some, but not all, of the effectors of N-Ras. An S27N mutant of M-Ras, like the analogous H-Ras S17N mutant, was a dominant inhibitor of activation of the c-fos promoter by constitutively active Src Y527F, suggesting that M-Ras and p21 Ras shared guanine nucleotide exchange factors and are likely to be activated in parallel. Moreover, M-Ras was recognized by the monoclonal anti-Ras antibody Y13-259, commonly used to study the function and activity of p21 Ras. Mammalian M-Ras and a Caenorhabditis elegans orthologue exhibit conserved structural features, and these are likely to mediate activation of distinctive signaling paths that function in parallel to those downstream of p21 Ras.

Insulin receptor substrate-2 is the major 170-kDa protein phosphorylated on tyrosine in response to cytokines in murine lymphohemopoietic cells.

Insulin receptor substrate 1 (IRS-1), and its structural relative IRS-2, are both phosphorylated on tyrosine following treatment of cells with interleukin-4 (IL-4) and insulin. We have investigated whether both IRS-1 and IRS-2 are expressed in murine lymphohemopoietic cells. T and B lymphocytes and macrophages from primary cultures expressed only IRS-2, which became phosphorylated on tyrosine following stimulation with both IL-4 and insulin. Likewise, the murine myeloid cell line FD-5 expressed only IRS-2, which was tyrosine phosphorylated in response to IL-4 and insulin, as well as interleukin-3 and granulocyte-macrophage colony stimulating factor. Neither IRS-1 nor IRS-2 were expressed at detectable levels in primary bone marrow mast cells although these cells do respond to IL-4. Moreover, a factor-dependent lymphocyte cell line, CT.4S, which grows continuously in IL-4, did not express detectable levels of IRS-1 or IRS-2. IRS-2 from FD-5 cells stimulated with either IL-4 or insulin bound to glutathione S-transferase fusion proteins of the p85 subunit of phosphoinositol 3'-kinase, Grb2, and Syp, paralleling reported associations of IRS-1 with these molecules and indicating phosphorylation of the corresponding residues on IRS-2.

A novel strategy for generating monoclonal antibodies from single, isolated lymphocytes producing antibodies of defined specificities.

We report a novel approach to the generation of monoclonal antibodies based on the molecular cloning and expression of immunoglobulin variable region cDNAs generated from single rabbit or murine lymphocytes that were selected for the production of specific antibodies. Single cells secreting antibodies for a specific peptide either from gp116 of the human cytomegalovirus or from gp120 of HIV-1 or for sheep red blood cells were selected using antigen-specific hemolytic plaque assays. Sheep red blood cells were coated with specific peptides in a procedure applicable to any antigen that can be biotinylated. Heavy- and light-chain variable region cDNAs were rescued from single cells by reverse transcription-PCR and expressed in the context of human immunoglobulin constant regions. These chimeric murine and rabbit monoclonal antibodies replicated the target specificities of the original antibody-forming cells. The selected lymphocyte antibody method exploits the in vivo mechanisms that generate high-affinity antibodies. This method can use lymphocytes from peripheral blood, can exploit a variety of procedures that identify individual lymphocytes producing a particular antibody, and is applicable to the generation of monoclonal antibodies from many species, including humans.

Genetic basis of hypo-responsiveness of A/J mice to interleukin-3.

Hematopoietic progenitor cells of the A/J strain of mice show a pronounced defect in the ability to form colonies or proliferate in response to interleukin-3 (IL-3). Comparison of immunoblots of A/J mast cells and of mast cells from the C57BL/6 strain that respond normally to IL-3 showed that, in both strains, a 125-kD band of the expected size was recognized by an antibody against the beta chain of the IL-3 receptor, the AIC2A molecule. However, in the C57BL/6 cells, there was an additional 110-kD species not seen in cells of the A/J strain. Analyses using bone marrow-derived mast cells from a panel of A/J x C57BL/6 and A/J x C57BL/6 recombinant inbred (RI) mice showed that the hypo-responsiveness to IL-3 is governed by a single gene. However, the absence of this 110-kD species in the A/J strain did not co-map with IL-3 hypo-responsiveness but did indeed map to the AIC2A genetic locus. These data show that this trait in the A/J strain was due to a polymorphism of the AIC2A gene unrelated to IL-3 hypo-responsiveness. Typing of the RI strains for the markers D14Mit98, D14Mitl4, and D14Mit133 mapped the locus determining hypo-responsiveness to IL-3 to the subtelomeric region of chromosome 14, the region that also bears the gene encoding the alpha chain of the IL-3 receptor (lL-3Ralpha). Immunofluorescence analyses indicated that IL-3Ralpha protein was undetectable on fresh bone marrow cells from A/J mice, although clearly detectable on cells from the responder C57BL/6 strain. However, IL-3Ralpha was readily detectable at normal levels on A/J mast cells generated by culture of A/J bone marrow cells in a combination of IL-3 and steel factor. Moreover, IL-3Ralpha on these A/J mast cells appears to be functional in that IL-3 stimulation of these cells results in tyrosine phosphorylation events characteristic of IL-3 signaling, including tyrosine phosphorylation of the beta chain of the IL-3 receptor, Jak-2 kinase, and SHPTP2. Collectively, these data indicate that the hypo-responsiveness of A/J mice to IL-3 is due to a defect in the gene encoding IL-3Ralpha and that, although this defect gives rise to reduced expression of alpha chain on primary bone marrow cells, this defect is not absolute and that, under certain circumstances, A/J cells can express functional receptors.

Interleukin (IL)-3 and granulocyte/macrophage colony-stimulating factor, but not IL-4, induce tyrosine phosphorylation, activation, and association of SHPTP2 with Grb2 and phosphatidylinositol 3'-kinase.

Binding of interleukin (IL)-3 and granulocyte/macrophage colony-stimulating factor (GM-CSF) to their high affinity cell surface receptors induces tyrosine phosphorylation of a similar set of protein substrates. We have identified one of these common substrates (p70) as the protein-tyrosine phosphatase SHPTP2. The Src homology 2 (SH2) domain of the adaptor protein Grb2 bound with high affinity to tyrosine-phosphorylated SHPTP2 following treatment of cells with IL-3 or GM-CSF, but not IL-4. This interaction was inhibited by two phosphotyrosine peptides, based on sequences within SHPTP2, which conform to the postulated consensus sequence for Grb2 SH2 recognition. Following treatment with IL-3 or GM-CSF, but not IL-4, SHPTP2 co-immunoprecipitated with antibodies directed against the p85 subunit of PI 3'-kinase. This was partially blocked by the same phosphopeptides that blocked Grb2-SH2 binding to SHPTP2. Importantly, treatment with IL-3 resulted in a 2-3-fold increase in SHPTP2 phosphatase activity. These results suggest that SHPTP2 may play an important role in integrating signals from the IL-3 and GM-CSF receptors to both Ras and PI 3'-kinase.

Multiple hemopoietins, with the exception of interleukin-4, induce modification of Shc and mSos1, but not their translocation.

Shc, grb2, and Son-of-sevenless (mSos1) proteins are potential upstream regulators of p21ras. We show that p52Shc and p46Shc comprise a significant portion of two of the major protein substrates phosphorylated on tyrosine in response to interleukin-2 (IL-2), IL-3, granulocyte-macrophage colony stimulating factor (GM-CSF), Steel factor (SLF), and colony-stimulating factor-1 (CSF-1). Once tyrosine phosphorylated, p52Shc and p46Shc associated with grb2. However, in contrast to published results with epidermal growth factor, treatment with GM-CSF, IL-3, and SLF failed to induce significant biochemically detectable translocation of Shc, grb2, or mSos1 from the cytosol to the plasma membrane. In addition, we did not observe significant epidermal growth factor-induced translocation of Sos1 to the membrane in Rat-1 cells. Treatment with SLF or IL-3 did increase tyrosine phosphorylation of membrane-localized p52Shc, which could then associate with grb2, although the majority of tyrosine-phosphorylated Shc was located in the cytosol. SLF, IL-3, and phorbol ester induced a decrease in the electrophoretic mobility of mSos1. This occurred with slower kinetics than p21ras activation and unlike hemopoietin-induced activation of p21ras was partially inhibited by a specific protein kinase C inhibitor. Thus, growth factor-induced modification of mSos1 may represent a downstream event, subsequent to p21ras activation. Significantly, IL-4, a cytokine that fails to activate p21ras, also failed to induce significant tyrosine phosphorylation of Shc or a shift in mSos1 mobility for the first time correlating these events with the ability of a growth factor to activate p21ras. Together, these data suggest that the current model for regulation of p21ras, which proposes a stable association of Shc-grb2-Sos1 complexes at the plasma membrane, may be an oversimplification.

Intracisternal A-type particle-mediated activations of cytokine genes in a murine myelomonocytic leukemia: generation of functional cytokine mRNAs by retroviral splicing events.

Previously we have described the derivation of three distinct classes of leukemic cell clones from a single in vivo-passaged myelomonocytic leukemia, WEHI-274, that arose in a mouse infected with the Abelson leukemia virus/Moloney leukemia virus complex (K. B. Leslie and J. W. Schrader, Mol. Cell. Biol. 9:2414-2423, 1989). The three classes of cell clones were characterized by distinct patterns of growth in vitro, the production of cytokines, and the presence of cytokine gene rearrangements. However, all three classes of WEHI-274 clones bore a common rearrangement of the c-myb gene, suggesting that all were derived from the one ancestral cell and that at least three distinct and independent autostimulatory events were involved in the progression of a single myeloid leukemic disease. In this article, we demonstrate that the autocrine growth factor production by the WEHI-274 leukemic clones resulted from cytokine gene activations mediated by the insertion of an intracisternal A-type particle (IAP) sequence 5' to the interleukin-3 (IL-3) gene, in the case of the class I clone, or 5' to the gene for granulocyte-macrophage colony-stimulating factor (GM-CSF), in the case of the class II clones. IAPs are defective murine retroviruses encoded by endogenous genetic elements which may undergo transpositions and act as endogenous mutagens. The functional IL-3 and GM-CSF mRNAs were generated by mechanisms in which the splice donor apparatus of the IAP sequence has been used in IAP gag-to-IL-3 or -GM-CSF splicing events.

The role of interleukin-1 and granulocyte-macrophage colony-stimulating factor in the paracrine stimulation of an in vivo-derived murine myeloid leukemia.

WEHI-274.3 is a cell line isolated from an in vivo-derived, murine myelomonocytic leukemia. Although the survival and growth of WEHI-274.3 cells in vitro is absolutely dependent on the addition of exogenous growth factors such as interleukin-3 (IL-3), granulocyte-macrophage colony-stimulating factor (GM-CSF), or colony-stimulating factor-1, when injected into syngeneic mice the cell line is tumorigenic. Sera from normal mice contain low levels of an activity that sustains survival of WEHI-274.3 but does not stimulate growth. In contrast, sera from mice bearing the WEHI-274.3 leukemia contained levels of CSF-1 and GM-CSF that stimulated the growth of WEHI-274.3 cells. Supernatants of cultures of WEHI-274.3 cells contained an activity that stimulated 3T3 fibroblasts to release an activity that stimulated the growth of the WEHI-274.3 cells. The 3T3-stimulatory activity released by the WEHI-274.3 cells was neutralized completely with an antiserum specific for murine IL-1 alpha, but not with antiserum specific for IL-1 beta. Moreover, WEHI-274.3 cells both in vitro and in vivo contained high levels of IL-1 alpha and IL-1 beta mRNAs. The leukemia-stimulatory activity released by the 3T3 cells was neutralized by an antiserum specific for GM-CSF. We postulate that the IL-1 alpha constitutively released by the WEHI-274.3 cells stimulates the production of GM-CSF from host cells such as fibroblasts or endothelial cells. A similar paracrine mechanism of growth stimulation may occur in acute myeloid leukemias in humans.

Connective tissue mast cells in contact with fibroblasts express IL-3 mRNA. Analysis of single cells by polymerase chain reaction.

Mast cell-fibroblast interactions have been extensively investigated in the last few years. Fibroblasts support the in vitro survival but not proliferation of mouse connective-tissue type mast cells. However, the factor(s) that allow their survival on fibroblast monolayers has not been identified. We have investigated the presence of mRNA for IL-3 and granulocyte-macrophage-CSF in single mouse mast cells, before and after co-culture with 3T3 fibroblasts, using the polymerase chain reaction technique. The system was calibrated first by using in vitro generated population of mouse bone-marrow derived mast cells (BMMC). Significant differences in the amplification of IL-3 cDNA were observed in each of the BMMC cells examined, whereas the amplification of cDNA for the alpha-subunit of the Fc epsilon RI were similar. Inasmuch as murine cultured IL-3-dependent mast cells differentiate into connective tissue-like mast cells when co-cultured with 3T3 fibroblasts without any exogenous supply of growth factors, it was of interest to determine whether these connective tissue-like mast cells produce IL-3 message. Separation of the differentiated BMMC from the fibroblast monolayer, by either trypsinization or by single cell manipulation revealed the synthesis of a detectable amount of IL-3 mRNA in these mast cells. Whether this IL-3 mRNA was induced by fibroblasts was further investigated using connective tissue mast cells freshly purified from the mouse peritoneal cavity. Only about 20% of these connective tissue mast cells produced detectable amount of granulocyte-macrophage-CSF mRNA whereas in less than 10% of the cells IL-3 mRNA was detected. However, when these connective tissue mast cells were co-cultured with 3T3 fibroblasts for 18 hours and then separated, IL-3 mRNA were detected in most of the cells whereas no such mRNA was detected in tissue mast cells incubated for 18 h with medium derived from 3T3 fibroblasts. Therefore we conclude that fibroblasts induce the accumulation of IL-3 mRNA in connective tissue mast cells. The production of IL-3 may play a role in the survival of this type of mast cells on the fibroblast monolayer.

Tumor treatment and cytokines.

In summary, molecular biology and advances in biochemical and biological techniques have led to the rapid characterization of a large number of cytokines. These are released in response to injury or invasion of the body and regulate the growth and function of a broad variety of cell types. Many cytokines are now in clinical trials and show promise in modulating the defence and repair responses of the body. The direct application of these cytokines to the killing of tumor cells remains problematic. However, it does seem likely that the use of these substances in a para-physiological mode i.e. in activating defence mechanisms and in particular stimulating hemopoetins may be a very important adjunct to more conventional means of tumor therapy such as chemotherapy or radiotherapy. The hemopoietic growth factors show particular promise here, although other substances such as IL-1 or IL-6 may find similar applications. In the future more information on the direct role of cytokines in the development of particular tumors may lead to the development and use of cytokine antagonists with effects directed at tumor cells or their environment.

Growth factor gene activation and clonal heterogeneity in an autostimulatory myeloid leukemia.

Cell lines were isolated from an in vivo-passaged myelomonocytic leukemia, WEHI-274, that arose in a mouse infected with the Abelson leukemia virus-Moloney leukemia virus complex. Clones were isolated in vitro in the presence or absence of a source of a hemopoietic growth factor, interleukin-3 (IL-3), and were divisible into three distinct classes. All three classes were leukemogenic in vivo. In vitro, the class I clone grew slowly at low cell density but responded with an increased growth rate to IL-3, granulocyte-macrophage colony-stimulating factor (GM-CSF), and autoconditioned medium. Supernatants of these cultures contained a factor with the biological, biochemical, and antigenic properties of IL-3. Class II clones grew better in vitro at low cell densities than did the class I clone and also responded with an increased growth rate to IL-3, GM-CSF, and autoconditional medium but produced GM-CSF rather than IL-3. In contrast, class III clones died in vitro at all cell densities unless exogenous IL-3 or GM-CSF was added. Moreover, they produced no autostimulatory factors. In the class I and class II clones, one allele of the respective IL-3 or GM-CSF gene is rearranged, and in each case, grossly abnormal RNA transcripts of the rearranged gene are present. Neither rearrangements nor abnormal RNA transcripts of the IL-3 or GM-CSF gene were detected in the class III clones. All three classes exhibited a common rearrangement of the c-myb gene, which suggested that all were derived from the one ancestral cell. These experiments demonstrate that two distinct and independent autostimulatory events were involved in the progression of a single disease.

Multiple glycosylated forms of T cell-derived interleukin 3 (IL-3). Heterogeneity of IL-3 from physiological and nonphysiological sources.

Interleukin 3 (IL-3) derived from mouse T cells was biosynthetically labeled with either [35S]methionine or [3H]mannose, affinity-purified using various anti-IL-3 antibodies, and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Autoradiography revealed the same three major bands with Mr values of 21,500-22,500, 27,000-31,000, and 32,000-36,000, irrespective of whether the anti-IL-3 antibody had been directed to the N or C termini of the IL-3 polypeptide. Bioassay of eluates from the gels confirmed that all three bands exhibited IL-3 bioactivity. IL-3 produced from two nonphysiological sources, the myelomonocytic leukemia WEHI-3B or Cos 7 cells that had been transfected with an IL-3 cDNA clone, had in each case a different pattern of microheterogeneity. Treatment with either tunicamycin or N-glycanase resulted in IL-3 running as one band with Mr 16,000, corresponding to its 140-amino acid polypeptide chain. No evidence for proteolytic processing was detected. These results show that the Mr heterogeneity of IL-3 was highly dependent on the cellular source and is due to N-linked glycosylation.

Autostimulatory mechanisms in myeloid leukemogenesis.

WEHI-274 is a monocytic leukemia that arose in a BALB/c mouse infected with Abelson murine leukemia virus. A series of subclones were derived from early passages of this tumor. Three subsets of these leukemogenic subclones were identified. Two subsets demonstrated autostimulatory patterns of growth. This was due to the ectopic production of the T-cell lymphokine the panspecific hemopoietin IL-3 in one case and of the T-cell lymphokine granulocyte-macrophage colony-stimulating factor (GM-CSF) in the other. The third type of subclone did not secrete any autostimulatory growth factor. In the subclone producing IL-3, one copy of IL-3 gene was rearranged and abnormal IL-3 RNA transcripts were present in the nucleus. Subclones producing GM-CSF also contained abnormal GM-CSF RNA transcripts, although no rearrangement of the GM-CSF gene was detected. All three sets of subclones shared a common rearrangement of one c-myb oncogene, suggesting that they share a common ancestor. These results suggest that initiation or progression of leukemogenic behavior in this abnormal clone occurred in three different ways, two of which involved autostimulation by the ectopic activation of T-cell lymphokine genes.

Structural homologies among the hemopoietins.

A group of cytokines characterized by a common set of target cells--namely, the pluripotential hemopoietic stem cells or their cellular derivatives--share similarities in the amino acid sequence at their N terminus or in the putative signal peptide immediately prior to the published N terminus. Murine P-cell-stimulating factor (PSF), murine and human interleukin 2 (IL-2), murine and human granulocyte-macrophage colony-stimulating factor (GM-CSF), human erythropoietin, and human interleukin 1 beta all share alanine as the N-terminal amino acid and have some similarities in the succeeding three or four amino acids. In the case of murine PSF and GM-CSF, the six N-terminal amino acids are readily cleaved from mature molecules and are lacking from the N-terminal amino acid sequences reported initially. A sixth cytokine, colony-stimulating factor 1, has an alanine followed by a similar pattern of five amino acids at the end of the putative signal peptide. GM-CSF and IL-2 have more extensive homology, about 25% of residues being identical in three regions that comprise about 70% of the molecules. Only minor similarities of uncertain significance were found among the complete amino acid sequences of the other cytokines. Although its evolutionary origin is uncertain, the homology around the N terminus may provide a structural marker for a group of cytokines active on the pluripotential hemopoietic stem cell and its derivatives.