Identification and structural analysis of human RBM8A and RBM8B: two highly conserved RNA-binding motif proteins that interact with OVCA1, a candidate tumor suppressor.

The OVCA1 gene is a candidate for the breast and ovarian tumor suppressor gene at chromosome 17p13.3. To help determine the function(s) of OVCA1, we used a yeast two-hybrid screening approach to identify OVCA1-associating proteins. One such protein, which we initially referred to as BOV-1 (binder of OVCA1-1) is 173 or 174 amino acids in length and appears to be a new member of a highly conserved RNA-binding motif (RBM) protein family that is highly conserved evolutionarily. Northern blot analysis revealed that BOV-1 is ubiquitously expressed and that three distinct messenger RNA species are expressed, 1-, 3.2-, and 5.8-kb transcripts. The 1-kb transcript is the most abundant and is expressed at high levels in the testis, heart, placenta, spleen, thymus, and lymphocytes. Using fluorescence in situ hybridization and the 5.8-kb complementary DNA probe, we determined that BOV-1 maps to both chromosome 5q13-q14 and chromosome 14q22-q23. Further sequence analysis determined that the gene coding the 1- and the 3.2-kb transcripts (HGMW-approved gene symbol RBM8A) maps to 14q22-q23, whereas a second highly related gene coding for the 5.8-kb transcript resides at chromosome 5q13-q14 (HGMW-approved gene symbol RBM8B). The predicted proteins encoded by RBM8A and RBM8B are identical except that RBM8B is 16 amino acids shorter at its N-terminus. Molecular modeling of the RNA-binding domain of RBM8A and RBM8B, based on homology to the sex-lethal protein of Drosophila, identifies conserved residues in the RBM8 protein family that are likely to contact RNA in a protein-RNA complex. The conservation of sequence and structure through such an evolutionarily divergent group of organisms suggests an important function for the RBM8 family of proteins.

Spontaneous transformation of rat ovarian surface epithelial cells: association with cytogenetic changes and implications of repeated ovulation in the etiology of ovarian cancer.

BACKGROUND: Ovarian surface epithelial cells undergo several rounds of division to repair the wound created by follicular rupture at the time of ovulation. This cyclical requirement for cell division, when not interrupted by the long anovulatory rest periods that occur during pregnancy and lactation, may contribute to the development of ovarian cancer. PURPOSE AND METHODS: To test this hypothesis, we isolated rat ovarian surface epithelial cells from 10 adult female Fisher rats, initiated two mixed-population and seven clonal cell lines, and repeatedly subcultured these cells in vitro for more than 20 passages. We then tested them for the acquisition of the following four features associated with transformation: 1) the loss of contact inhibition, 2) the capacity for substrate-independent growth, 3) the ability to form tumors when injected subcutaneously and/or intraperitoneally into athymic mice, and 4) cytogenetic abnormalities. RESULTS: Loss of contact inhibition was observed in all nine late-passage cell lines. Six of the nine late-passage, but none of the early-passage, cell lines tested exhibited a capacity for substrate-independent growth that was augmented in a dose-dependent manner by epidermal growth factor. Two late-passage cell lines (clone 2 and mixed-population 2) generated tumors in athymic BALB/c mice within 3 weeks following subcutaneous injection of 5 x 10(6) cells, whereas similar numbers of early-passage cells from the same cell lines failed to generate palpable tumors. Late-passage clone 7 cells were tumorigenic when 5 x 10(7) cells were injected intraperitoneally. Two of the cell lines analyzed exhibited alterations involving losses of part or all of one member of the chromosome 5 pair. Clone 2 possessed an interstitial deletion, del(5)(q21.3q24), consistent with the loss of an uncloned putative tumor suppressor gene at 5q22q23 previously reported to reside near the loci for the interferon alpha, interferon beta, and c-jun genes. Early-passage clone 7 cells exhibited chromosome 5 monosomy, while late-passage cells contained one normal chromosome 5 and a derivative (5q12q). Southern analysis of the three cell lines revealed no consistent loss of loci for the interferon and c-jun genes, although early-passage clone 7 cells had one half the gene copy number for the interferon beta and c-jun genes and both early- and late-passage clone 7 cells lacked DNA sequences hybridizing with the probe for interferon alpha. CONCLUSION: This pattern of passage-dependent spontaneous transformation of rat ovarian surface epithelial cells in vitro supports the hypothesis that repetitious ovulation contributes to the etiology of human ovarian cancer.

Increased levels of glutathione S-transferase pi transcript as a mechanism of resistance to ethacrynic acid.

Subpopulations of HT 29 human colon carcinoma cells (HT/M and HT/S) were selected for resistance to the glutathione S-transferase (GST) inhibitor ethacrynic acid (EA). Both clones displayed a 2-fold resistance to the selection agent and required its constant presence for the maintenance of the resistant phenotype. Purification and characterization of GST isoforms showed similar profiles in the wild-type (WT) and EA-resistant clones, with microheterogeneous forms of the pi isoenzyme detected in each case. Metabolism of EA in vitro in the presence of GSH and the isolated GST from each cell line was characterized by a biphasic disappearance of the parent drug; the initial rate at which each of these enzymes metabolized EA was similar. These enzymes also displayed similar Km values for 1-chloro-2,4-dinitrobenzene. However, the amount of GST isolated per total cellular protein was 3.0-fold in HT/M and 1.6-fold in HT/S relative to WT in the continuous presence of EA. Under these conditions GST activity was increased by 2.3-fold in HT/M and 3.2-fold in HT/S as were GSH levels (2.7- and 4.1-fold for HT/M and HT/S respectively). When EA was removed, enzyme activity and GSH concentrations decreased to values similar to those of the WT. Slot-blot and Southern analyses of the DNA gave no evidence of GST-pi-gene amplification or rearrangement. However, RNA analyses by both slot-blot and Northern studies indicate a 2.5-3.5-fold elevation in the GST pi transcript in the EA-resistant population. Results from these studies indicate that: (1) maintenance of the EA-resistant phenotype requires constant presence of the agent; (2) the 2-fold resistance to EA can be quantitatively related to a 2-3-fold increase in GST activity and amount which appears to be the result of a 2.5-3.5-fold elevation in GST transcript; (3) EA, a Michael-reaction acceptor, can induce GST at the transcriptional level.

Evidence for a glycoconjugate form of glutathione S-transferase pI.

The anionic form of glutathione S-transferase from human (GST pi) and rat (GST Yp) sources has been shown to exist in multiple forms which have similar molecular weights but different isoelectric points (pIs). Treatment with endoglycosidase H caused the acidic forms of GST Yp to be converted to proteins with more basic pIs as compared to the untreated control mixtures, suggesting that an N-linked mannose moiety containing acidic residues had been removed. Inability to detect these carbohydrates by techniques requiring unsubstituted vicinal hydroxyls further suggested acidic substitutions on the sugar moiety. GST pi/Yp carbohydrate modifications were also identified by differential staining procedures. These data represent the first indication that glycosylation of GST can occur. Additionally, this may offer an explanation for the often seen microheterogeneity within a class of GST isozymes.