Testicular germ cell tumor susceptibility genes from the consomic 129.MOLF-Chr19 mouse strain

Chromosome substitution strains (CSS or consomic strains) are useful for mapping phenotypes to chromosomes. However, huge efforts are needed to identify the gene(s) responsible for the phenotype in the complex context of the chromosome. Here, we report the identification of candidate disease genes from a CSS using a combination of genetic and genomic approaches as well as by using knowledge about the germ cell tumor disease etiology. We utilized the CSS, 129.MOLF-Chr 19 chromosome substitution strain (or M19), in which males develop germ cell tumors of the testes at an extremely high rate. We are able to identify 3 protein-coding genes and 1 microRNA on chromosome 19 that have previously not been implicated to be testicular tumor susceptibility genes. Our findings suggest that changes in gene expression levels in the gonadal tissues of multiple genes from Chr 19 likely contribute to the high TGCT incidence of the M19 strain. Our data advances the use of CSS to identify disease susceptibility genes and demonstrates that the 129.MOLF-Chr 19 strain serves as a useful model to elucidate the genetics and biology of germ cell transformation and tumor development. For gene expression profiling, we used male gonads from different developmental stages, E13.5 and PN1, from the M19 and 129 strains. The idea was to detect the common gene expression changes in the gonads at stages when germ cells are known to transform to embryonal carcinoma (EC) cells. E13.5 stage was chosen because tumor development is reported to start around E13.5 in 129 strains (Stevens 1973b; Stevens and Hummel 1957). PN1 stage gonads were chosen because our studies with M19;Oct4-GFP mice, as described above, showed that both germ cells and EC cells are present at this stage suggesting that germ cell transformation may also be ongoing at this stage in the testes of M19 strain. We also compared gene expression differences in gonads to that of embryos to determine whether gene expression differences are specific to the gonads. Our goal was to identify gene(s) whose expression is consistently changed in the gonads at E13.5 and PN1 in the M19 as these would likely be candidate TGCT susceptibility gene(s). Fig. 3A indicates the 3 different tissue samples collected from the M19 and 129 strains: gonads (genital ridges) dissected from male E13.5 embryos and from PN1 mice and male embryos at E13. Because tumor incidence in M19 is approximately 80% we anticipated that gene expression changes could be masked because ~20% of cells/tissues within a sample may have normal levels of gene expression or 20% of the testes may have normal gene expression. Moreover, the amount of RNA extracted from a single genital ridge or PN1 testis is insufficient for a microarray experiment. To minimize the effect of individual samples and to isolate sufficient RNA for microarray analysis from these small sized gonads, we pooled RNA from multiple samples (Table 1). Six RNA pooled samples were prepared (Fig.3A and Supplementary Table 2): E13.5 gonad RNA from males of 129 (129-GR) and M19 (M1-GR); PN1 testes RNA from 129 (129-NBT) and M19 (M1-NBT); E13.5 embryo RNA from males of 129 (129-E) and M19 (M1-E). The six RNA samples were hybridized to six Affymetrix mouse genome genechips individually. Hierarchical clustering analysis revealed a dendrogram in which gene expression profiles of the same tissue type clustered together (Fig. 3B). Profiles of the gonads, E13.5 and PN1, were related more closely than to those of embryos. Comparison of the expression profiles of E13.5 gonads from M19 and 129 revealed 10 genes with more than a 2-fold difference in expression levels (Fig. 3A and Supplementary Table 3A). Six of the 10 genes mapped to Chr 19. For the PN1 testes, 266 genes showed significant fold changes between the two strains and of these 13 genes mapped to Chr 19 (Fig. 3A and Supplementary Table 3C). For the E13.5 embryos, 35 genes showed greater than a 2-fold change in expression between the M19 and 129 and 5 out of 35 genes mapped to Chr 19 (Fig. 3A and Supplementary Table 3B). By analyzing the data of differentially expressed genes present in the 3 samples, as well as selecting those that map to Chr 19, we were able to exclude a majority of the genes but found 3 genes in common. The 3 genes map to Chr 19 and were found to be downregulated in the M19 strain E13.5 and PN1 gonads as well as E13.5 embryos. These are Zfp162, D19Bwg1357e and Cox15. These 3 genes have not been previously implicated in testicular tumorigenesis and are novel TGCT candidate susceptibility genes.染色体替换品系（Chromosome substitution strains，CSS；又称同系染色体替换品系consomic strains）是将表型定位至染色体的有效工具。然而，在染色体的复杂遗传背景中鉴定出与表型相关的责任基因，仍需投入大量研究工作。本研究结合遗传学与基因组学方法，并借助生殖细胞肿瘤的疾病病因学知识，从某一染色体替换品系中筛选出候选致病基因。我们使用的染色体替换品系为129.MOLF-Chr 19染色体替换品系（简称M19），该品系的雄性个体罹患睾丸生殖细胞肿瘤的概率极高。我们在19号染色体上鉴定出3个蛋白编码基因与1个微小RNA（microRNA），此前这些基因均未被报道与睾丸肿瘤易感性相关。本研究结果表明，19号染色体上多个基因在性腺组织中的表达水平改变，可能是M19品系睾丸生殖细胞肿瘤（testicular germ cell tumor, TGCT）发病率极高的原因。本研究数据拓展了CSS在鉴定疾病易感基因中的应用，同时证明129.MOLF-Chr 19品系是阐明生殖细胞转化与肿瘤发生的遗传学与分子生物学机制的理想模型。在基因表达谱分析中，我们采集了M19与129品系雄性个体在不同发育阶段的性腺样本：胚胎第13.5天（embryonic day 13.5, E13.5）与出生后第1天（postnatal day 1, PN1）。实验设计旨在检测生殖细胞已知转化为胚胎性癌（embryonal carcinoma, EC）细胞的发育阶段中，性腺内共有的基因表达变化。选择E13.5阶段的原因是，已有研究报道129品系的肿瘤发生始于E13.5左右（Stevens 1973b; Stevens and Hummel 1957）。选择PN1阶段性腺样本的原因是，我们此前针对M19;Oct4-GFP小鼠的研究表明，该阶段同时存在生殖细胞与EC细胞，提示M19品系睾丸内的生殖细胞转化可能在该阶段仍在进行。我们还将性腺的基因表达差异与胚胎进行对比，以明确基因表达差异是否为性腺所特有。本研究的目标是筛选出在M19品系E13.5与PN1阶段性腺中表达持续改变的基因，此类基因极有可能为TGCT候选易感基因。图3A展示了从M19与129品系中采集的3类不同组织样本：雄性E13.5胚胎的性腺（生殖嵴）、PN1阶段雄性小鼠的睾丸，以及E13阶段雄性胚胎。由于M19品系的肿瘤发生率约为80%，我们推测基因表达变化可能被掩盖：样本中约20%的细胞/组织可能呈现正常的基因表达水平，或20%的睾丸组织基因表达正常。此外，从单个生殖嵴或PN1阶段睾丸中提取的RNA总量不足以进行微阵列实验。为降低单个样本的影响，并从这类小型性腺中获取足够的RNA用于微阵列分析，我们将多个样本的RNA进行了混合（表1）。我们共制备了6个混合RNA样本（图3A与补充表2）：129品系与M19品系雄性E13.5性腺的RNA（分别记为129-GR与M1-GR）；129品系与M19品系雄性PN1阶段睾丸的RNA（分别记为129-NBT与M1-NBT）；以及129品系与M19品系雄性E13.5胚胎的RNA（分别记为129-E与M1-E）。将这6个RNA样本分别与6张Affymetrix小鼠基因组基因芯片进行杂交。层级聚类分析得到的树状图显示，同一组织类型的基因表达谱聚为一类（图3B）。E13.5与PN1阶段性腺的表达谱之间的相关性，高于其与胚胎表达谱的相关性。对比M19与129品系雄性E13.5性腺的表达谱，共发现10个表达水平差异超过2倍的基因（图3A与补充表3A）。这10个基因中有6个定位于19号染色体。针对PN1阶段睾丸的对比分析显示，两品系间共有266个基因存在显著表达差异，其中13个定位于19号染色体（图3A与补充表3C）。针对E13.5胚胎的对比分析显示，两品系间共有35个基因表达差异超过2倍，其中5个定位于19号染色体（图3A与补充表3B）。通过分析3类样本中的差异表达基因数据，并筛选定位于19号染色体的基因，我们排除了绝大多数基因，最终得到3个共有基因。这3个基因均定位于19号染色体，且在M19品系的E13.5与PN1阶段性腺以及E13.5胚胎中均呈现表达下调，分别为Zfp162、D19Bwg1357e与Cox15。这3个基因此前均未被报道与睾丸肿瘤发生相关，属于全新的TGCT候选易感基因。