Testicular germ cell cancers in young adult men derive from a precursor lesion called carcinoma in situ (CIS) of the testis. CIS cells were suggested to arise from primordial germ cells or gonocytes. However, direct studies on purified samples of CIS cells are lacking. To overcome this problem, we performed laser microdissection of CIS cells. Highly enriched cell populations were obtained and subjected to gene expression analysis. The expression profile of CIS cells was compared with microdissected gonocytes, oogonia, and cultured embryonic stem cells with and without genomic aberrations. Three samples of each tissue type were used for the analyses. Unique expression patterns for these developmentally very related cell types revealed that CIS cells were very similar to gonocytes because only five genes distinguished these two cell types. We did not find indications that CIS was derived from a meiotic cell, and the similarity to embryonic stem cells was modest compared with gonocytes. Thus, we provide new evidence that the molecular phenotype of CIS cells is similar to that of gonocytes. Our data are in line with the idea that CIS cells may be gonocytes that survived in the postnatal testis. We speculate that disturbed development of somatic cells in the fetal testis may play a role in allowing undifferentiated cells to survive in the postnatal testes. The further development of CIS into invasive germ cell tumors may depend on signals from their postpubertal niche of somatic cells, including hormones and growth factors from Leydig and Sertoli cells. [Cancer Res 2009;69(12):5241–50]
Microarray and RT-PCR based methods are important tools for analysis of gene expression; however, in tissues containing many different cells types, such as the testis, characterization of gene expression in specific cell types can be severely hampered by noise from other cells. The laser microdissection technology allows for enrichment of specific cell types. However, when the cells are not morphologically distinguishable, it is necessary to use a specific staining method for the target cells. In this study we have tested different fixatives, storage conditions for frozen sections and staining protocols, and present two staining protocols for frozen sections, one for fast and specific staining of fetal germ cells, testicular carcinoma in situ cells, and other cells with embryonic stem cell-like properties that express the alkaline phosphatase, and one for specific staining of lipid droplet-containing cells, which is useful for isolation of the androgen-producing Leydig cells. Both protocols retain a morphology that is compatible with laser microdissection and yield RNA of a quality suitable for PCR and microarray analysis.