Search

Two chemoprevention trials have shown that retinoic acid (RA) may be harmful in patients at risk for lung cancer, and RA administration to this high-risk group results in RARB2 reactivation. Although RARB2 is thought to possess tumor suppressive activity, its expression has recently been correlated with poorer prognosis in patients with nonsmall cell lung cancer. We hypothesized that RARB2 expression is necessary for the growth and maintenance of the oncogenic phenotype in lung cancer cells in which RARB2 has not been inactivated. We tested various antisense oligodeoxynucleotides (ASO) against RARB2 in multiple lung cancer cell lines and used microarray technology to compare the patterns of gene expression following ASO treatment versus RA treatment in the A-549 lung cancer cell line. We show that ASO treatment reduces proliferation and causes apoptosis in 3 RARB2-expressing lung cancer cell lines but has no apparent effect in at least two other lung cancer cells lines having lost RARB2 expression or one normal lung RARB2-expressing cell line; we demonstrate a correlation between resulting RARB2 expression levels and cell growth; and identify transcriptional effects related to both RA and RARB2 signaling. In particular, five genes known to contribute to carcinogenesis or chemotherapeutic resistance are down-regulated following ASO treatment: three of these are up-regulated following RA treatment. This work demonstrates a dual role for RARB2 (tumor suppression and tumor promotion) and identifies a challenge with respect to using RARB2 as a target for treatment or prevention strategies., (Copyright © 2011 Wiley-Liss, Inc.)

Retinoic acid receptor B2 (RARB2) is frequently inactivated in cancer. Methylation in the 5'-untranslated region and first exon is known to play a role; however, few studies have analyzed the detailed methylation pattern of the promoter region. We show that hypo- and hypermethylated alleles coexist in 5/11 cell lines in which RARB2 is inactivated. We present evidence supporting the mitotic transmission of these divergent methylation patterns and find a correlation between methylation divergence and heterozygosity at the 3p24 loci, suggesting an allelic methylation bias in these lines. Using a newly devised strategy based on allelic identification via methylation-sensitive restriction enzyme digestion combined with the use of a single nucleotide polymorphism, rs755661, we demonstrate that such a bias exists in three cancer cell specimens heterozygous at rs755661 and therefore amenable to this study. This previously unreported phenomenon of allelic methylation bias suggests that a promoter methylation-independent mechanism may be responsible for inactivation at the hypomethylated allele and this inactivation is reminiscent of an aberrant form of de novo imprinting. Approaches to interpreting methylation data should incorporate the notion of allelic methylation bias.

We report a higher frequency of loss of heterozygosity at loci on the short arm of chromosome 3 in human epidermoid lung tumors than in other non-small cell lung tumors. This observation together with the already known involvement of the retinoids in the development of epidermoid metaplasia and neoplasia, especially in lung tissue, prompted us to investigate by RNAase protection assays the status of expression of a gene that maps on chromosome band 3p24 and codes for the B receptor for retinoic acid (RARB). We show that expression of RARB is detectable in normal lung tissue and in most of the cell lines derived from lung tumors, including the five non-small cell lines that clearly had a non-epidermoid phenotype. Strikingly, however, only one of the five epidermoid-tumor-derived cell lines studied showed detectable expression of RARB. Of two lines derived from adenosquamous tumors, one had a clear epidermoid differentiation, and this line also did not express RARB. Taken together, our results strongly implicate RARB in the evolution of epidermoid lung tumors.

Stimulating electrodes were placed on the terminal branches of the pelvic nerves to the urinary bladder and the pudendal nerve to the sphincters in seven Rhesus monkeys and two chimpanzees. The proximity of the electrodes to these structures assured organ specificity. Evoked responses produced by stimulation of these terminal nerve branches were recorded in the fascicles and rootlets of the lower thoracic, lumbar, and sacral nerve roots. During identical stimulating and recording conditions, the amplitude as well as presence or absence of the evoked responses recorded was variable within the various roots. The amplitude of the evoked responses or their absence depended on the number of fibers within a particular fascicle which conducted impulses to the urinary bladder or the urethral and anal sphincters. By this method, it was determined that there was segregation or compartmentalization of the nervous innervation to the urinary bladder and sphincters within the spinal roots. In addition, the segmental spinal cord origin of the innervation of the urinary bladder was determined for the Rhesus monkey and chimpanzee. In the Rhesus monkey, the pelvic nerves to the urinary bladder arose from the first and second sacral segments and to a much lesser extent from the seventh lumbar segment. In the chimpanzee, the sacral segments one to four gave rise to innervation of the urinary bladder.

Stimulating electrodes were placed on the terminal branches of the pudendal nerve to the external urethral and external anal sphincters. The proximity of the electrodes to the sphincters assured organ specificity. Evoked responses produced by stimulation of these terminal nerve branches were recorded in the fascicles and rootlets of the lower thoracic, lumbar, and sacral nerve roots. By this method, the segmental spinal cord origin of the innervation of the external urethral and anal sphincters was determined for the Rhesus monkey and chimpanzee. The data indicated that the pudendal nerves to the urethral and anal sphincters in the Rhesus monkey arose from the sixth and seventh lumbar spinal segments and the first and second sacral spinal segments which are homologous to the S-1 and S-4 segments found to give innervation to these structures in the chimpanzee. The primate experiments thus indicate that the spinal origin of the pudendal nerve was more rostrally located by one segment or more than was the origin of the pelvic nerves to the urinary bladder.

Evidence is presented which suggests that a CHO-derived thymidine kinase (Tk) heterozygote, ts201, can reversibly inactivate the wild type Tk allele, and that this event may result in simultaneous inactivation of a linked allele, galactokinase (Glk). The clone ts201 was isolated as a revertant of a stable Tk- 5-bromodeoxyuridine (BrdU)-resistant mutant of CHO. The reacquired Tk activity differed from that of the wild type with respect to Km and heat resistance, supporting the contention that ts201 was genetically heterozygous (Tk+/-). At frequencies varying from 0.02-0.5, segregants of ts201 could be isolated by cloning in BrdU at 39 degrees. These derivatives were phenotypically Tk- (by enzymology and by autoradiography of 3H-dT labeled cells), but after removal of BrdU, reacquired the Tk+ phenotype at frequencies varying randomly from clone to clone. From mutagenized populations of ts201 two variants were isolated, 71t and 72c, by selection at 39 degrees in 2-deoxygalactose (2-dgal). Resistance to 2-dgal has been correlated with a mutation in the gene for Glk, which is syntenic with that for Tk and evidence is presented suggesting that 71t and 72c are Glk+/-. Cloning efficiencies in doubly selective media and autoradiographic data showed that at 39 degrees a coordination existed between the levels of Tk and Glk, which in 71t was inverse, and in 72c, direct. These data led to the hypothesis that pairs of linked alleles can be inactivated or re-expressed simultaneously. The inactivating effect was temperature-sensitive, since at 33 degrees, the cloning efficiency in BrdU and in 2-dgal was very low.

Mutants of SV40-transformed mouse fibroblasts have been isolated that have greatly increased cell-substratum adherence. The adherent phenotype (COL-) is recessive, and all mutants analyzed belong to one complementation group. No consistent qualitative differences between wild-type and mutant cells were found with respect to the protein content of the substrate-attached material (SAM), a cell surface fraction left after removal of cells from the substrate with a gentle Ca2+-chelating agent. However, the mutants yielded 2.5-10-fold more SAM than the parental cell line, and the SAM deposited by mutants was able to mediate attachment of transformed cells to a much greater degree than was the SAM from the parental cell line. The mutation, which appears to control the generation of footpads, was shown to cosegregate with resistance to the drug 6-thioguanine, which suggests X-linkage.