Role of TGF-beta signaling and Ski/SnoN mRNA expression in cervical carcinomas

Jung, Jae Don; Lee, Seon Kyung; Kim, Seung Bo; Chi, Sung Gil

Abstract

OBJECTIVE
TGF-beta signaling is dependent on the heterodimerization of the type II TGF-beta receptor (TbetaR-II) with the type I TGF-beta receptor (TbetaR-I). which mediate intracellular signals through Smad proteins. Whereas physiologic concentrations of SnoN and Ski allow a feedback regulation of TGF-beta signaling, deregulation of SnoN or Ski expression leads to total inhibition of TGF-beta signaling and of the tumor suppressors Smad2 and Smad4, which can explain the role of SnoN and Ski as oncogenes. In order to identify possible molecular mechanisms responsible for TGF-beta resistance, the author investigated the mutation and expression of TGF-beta1, its receptors, Ski/SnoN in cervical carcinomas.
METHODS
From December 1995 to December 1999, 45 carcinomas and 7 normal cervical tissue specimens were obtained by surgical resection in the Kyung Hee University Medical Center. Tissue specimens were snap-forzen in liquid N2 and stored at -70 degrees C until used. Total RNA was extracted from specimens and evaluated the expression levels using densitometric analysis of quantitative RT-PCR products (TGF-beta1, Tbeta1R-I, Tbeta1R-II, Ski/SnoN), and the mutations were investigated by quantitative genomic-PCR followed by nonisotopic RT-PCR-SSCP analysis (Tbeta1R-II, Tbeta1R-I, Ski/SnoN). The abnorally expressed levels of RT-PCR products (TGF-beta1, Tbeta1R-II) were analysed for the clinicopathologic characteristics.
RESULTS
Quantitative RT-PCR analysis demonstrated variable expression of TGF-beta1 mRNA (0.05-0.89) in tumors and significantly increased TGF-beta1 expression level (>0.48) in 15 of 45 samples (33.3%). There is no significant reduction of Tbeta1R-I expression (<0.38) in tumors, but 9 of all tumors (20.0%) show significantly reduced levels of Tbeta1R-II expression (<0.58). Using quantitative DNA-PCR analysis, all of 9 specimens with abnormally low Tbeta1R-II expression show abnormally low levels (<0.47) of the Tbeta1R-II gene at genomic level which suggests allelic deletion of the gene in these specimens. Gene mutations of TGF-beta1 receptors were analysed using specific primers by RT-PCR-SSCP analysis, and the results revealed no mutational alterations of TGF-beta1 receptors and no mutation in poly (A) region of Tbeta1R-II. Quantitative RT-PCR analysis demonstrated variable expression of Ski/SnoN (0.65-1.46/0.75-1.62) in tumors and significantly increased Ski expression level (>1.36) in 2 of 45 samples (4.4%), and there is no amplification of Ski/SnoN gene by quantitative genomic-PCR analysis.
CONCLUSIONS
The overexpression of TGF-beta1 mRNA and the reduced or absent expression of Tbeta1R-II may be an important contributing factors, and the abnormally low genomic levels and no mutational alterations of Tbeta1R-II is caused by monoallelic deletion suggesting that Tbeta1R-II might play as a tumor suppressor of haloinsufficiency in cervical carcinomas. We could not show that high levels of Ski/SnoN expression could produce a disruption of TGF-beta signaling in cervical carcinomas.