台灣乳癌病理成因之研究-乳癌檢體中含變異之動情激素接受體 與臨床上之關聯性(第三年)

Abstract

為了研究臺灣乳癌檢體中所含變 異型動情素接受體a與臨床上預後的關 聯性，我們從199 個原發性的乳癌檢 體中挑選出98 個檢體，並取其中19 個檢體的正常對照組織，以及8 個良 性的乳腺檢體為對照組，檢查它們的 密碼區(coding region)。我們在38 個其 密碼區被整個檢查過的病例中發現1 個點突變及3 個多形性區域，該點突 變(Pro324Ser)發生於動情素接受體a的 LBD 的第一個a螺旋的末端， 而 proline 一般相信對蛋白質的二級結構 及局部的方向性有很大的影響。而為 了進一步了瞭3 個多形性區域的生理 意義，我們將上述的其它部分檢體進 行多形性區域的分析，發果發現腫瘤 檢體中這3 個多形性區域在統計分析 中顯示出極度的分佈不平衡現象，而 且其中有一個多形性區域顯示在癌症 組織中的偏好表現，然而與癌症的復 發並不具有顯著相關。而比較19 個正 常對照組織與其腫瘤檢體的多形性分 佈情形，我們發現多形性區域的分佈 不平衡現象可能與腫瘤組織失去同時 表現兩個alleles 的能力有關。比較腫 瘤與正常組織中exon 刪除型動情素接 受體a的表現情形時，以往曾經被報告 過的刪除型動情素接受體a亦都存在於 我們的檢體中，我們更發現了二型從 未被報告過的刪除型動情素接受體a-- 單獨刪除exon 1 最後32 個核酸，以 及此32 核酸連同exon 2 一同刪除， 此二型變異型均廣泛的表現於乳癌的 檢體之中。此外，在乳癌檢體中所表 現的各種變異型動情素接受體a之對野 生型動情素接受體a相對量均較正常對 照組織為高。但經由統計分析並無法 測量出變異型動情素接受體a的表現與 腫瘤復發之間的關聯性。

To investigate the correlation between the presence of ERa variants and prognosis of breast tumors from Taiwan, we detailed examined the coding region of ERa gene. One sense point mutation and 3 polymorphisms have been found. The point mutation, P324S, is located at the end of the beginning helix of ER ligand binding domian. The 3 polymorphisms showed strong disequilibrium in allele distribution and one of the polymorphisms showed allele preference in tumor. However, the allele preference is not significantly associated with tumor recurrence. To elucidate the possible reason of disequilibrium in allele distribution, we compared polymorphisms of paired tumor-normal counterpart tissues, and found the disequlibrium may due to a loss expression of one allele phenomenon. In searching for exon deletion variants, alternative spliced variants were found. Moreover, a deletion of 32 nucleotide 3’ to the exon 1 only (d32nt) or together with the whole exon 2 (d32nt-D2) through aberrant splicing, which has never been reported, was found to exist extensively in tumor tissues in our study. In addition, types and amount of deletion variants were more dominant in tumor tissue than normal tissues. However, the elevated expression of deleted variant forms did not showed significant association with tumor recurrence. Introduction Estrogen exposure is one of the major risk factors of breast cancer development. Many epidemiological surveys support this relationship (reviewed in Pike et al. 1993). The biological mechanism is poorly understood. One hypothesis is the estrogen metabolic derivatives play a genotoxic role in the carcinogenesis of breast tissue. The other possible mechanism is the cell proliferationpromoting ability through estrogen receptor(ER) pathway. There are two known estrogen receptors-- ERa and ERb. ERa protein plays a major role in the development and normal physiology of the breast tissue and is important for the regulation of cell growth and differentiation. To investigate the correlation between the presence of ERa variants and prognosis of breast tumors from Taiwan, we detailed examined the coding region of ERa gene from 98 out of 199 primary breast cancers, 19 normal counterparts and 8 benign masses. Results Detection of Nucleotide Changes by PCR-SSCP Analysis PCR-SSCP analysis was used to detect small insertions/deletions and point mutations present in ERa gene. Nine overlapped amplification regions (about 250 base pairs each in size) were carefully examined and only 1 point mutation and was found from these 38 putative ERa misfunctioning cases. The point mutation, which changed a proline residue (CCC) to serine (TCC) at codon 324, is located in the end of the beginning helix (H1, numbered based on the canonical structure of NR LBD, Wurtz et al., 1996) of the ligand binding domain. Since a proline residue may play a role in determination and orientation of local secondary structure, this mutant is possibly with altered function. However, functional assay of this mutant has not been conducted yet. Detection of Polymorphisms by PCRSSCP Analysis 3 neutral polymorphisms were found in the 38 putative ERa-misfunctioning cases by PCR-SSCP analysis. These polymorphiic sites are located in codon 10 [TCT®TCC (Ser), exon 1], codon 325 [CCC®CCG (Pro), exon 4], and codon 594 [ACA®ACG (Thr), exon 8], respectively. Polymorphism patterns of ERa-PR consistent and normal group were also analyzed in order to compare with those of the ERa-PR inconsistent group. When the alleleic distribution pattern of these polymorphic regions being examined, all of the 3 polymorphisms in the ERa-PR inconsistent group showed linkage disequilibriums but only polymorphism in codon 594 showed a slight allele preference (Table 1). Detection of Exon Deletion Var iants by PCR-Southern Blotting Analysis In the PCR-Southern blotting based detection of truncated transcripts, truncated transcripts (such as clone 4) or longer exon deletion variants (such as D2-3-4) can not be detected by our strategy. We could not also discriminate transcripts with single exon deletion or combine exon deletion ( for instance, exon 4 and exon 7 deletions were found in the same transcript). But these transcripts present in relatively low level and the majority of variant transcripts bearing shorter deletions. In our study, the alternative spliced transcripts, D2, D3, D4, D5, D7, D2-3, D3-4, D4-5, reported by previous studies of other groups were also detected. We also found a truncated transcript comprising 32 nucleotides 3’ to the exon 1 (denoted as d32nt) which has never been reported in previous publishes. It might because the shorter PCR spanning region in our study is favored finding out small deletions. The 32 nt bears a sequence character with a beginning of GT and an ending of AG, which is seemly the feather of most introns. Thus the d32nt may owe to be mistaken as an intron and spliced out. A deletion of this 32 nt together with the whole exon 2 (denoted as d32nt-D2) was also found in our studies. Protein translated from d32nt and d32nt-D2 variants will terminated prematurely just after Thr140 in addition to 5 or 35 more extra amino acids and bearing only an incomplete A/B domain possibly without any biological activity. Compared with the basal level by quantifying the relative ratio of variant to wildtype bands, tumor groups beard significant increase in D4, D5, D7, and D2-3 expression (Table 2). Discussion In the present study for the search of ERa variants, we used a preliminary IHC analysis to raise the possibility of finding out ERa variants with altered function. However, we only found 1 sense point mutation out of 38 primary breast cancer cases, much fewer from the expected number, 2~10 mutations based on multiply 1~5% reported mutation rate and 199 cases. It is possible that some mutations with unaltered function may exist in the 161 cases we did not detailed screened. 2 variants comprising deletion of 32 nucleotides 3_ to exon 1 only (d32nt) or together with the whole exon 2 (d32nt-D2) were first reported in this study. They were probably generated from aberrant splicing mechanism since the 2 sequences are flanked with genuine or cryptic splice sites. Aberrant splicing mechanism is reported important in altering normal function of some cellular proteins but aberrant splicing variant of ERa was never identified except a 69- nucleotide insertion between exon 5 and 6. The variant was resulted from a new splice donor site generated by a point mutation in the intron 5 (Wang et al., 1997). However, it is the first time that aberrant splicing is found to occur within the ERa coding sequences. It is also the first time to suggest that aberrant splicing may play some role in altering the signaling pathway of ERa gene. Since these variants were only reported in Taiwan population, more extensive studies must be undertaken in different populations to determine whether they are ethnic-specific variants or they were found in the advantage of shorter PCR amplified region comparing with those used by other groups. By using polymorphisms as genetic markers, we also found significantly strong disequilibrium in allele distribution comparing with the result by an U.S. group (Roodi et al., 1995) Only 2 out of 6 polymorphisms showed relatively slighter significance. Furthermore, the polymorphisms they analyzed did not reveal the loss expression of one allele phenomenon we observed in our cases. It is thus strongly suggested that changes in ERa signaling may play a different role in breast cancer tumoriogenesis in Taiwan.