Fold-change correction values for testicular somatic transcripts in gene expression studies of human spermatogenesis

What are the reference values for delineating altered somatic cell gene expression from transcript enrichment/dilution in gene expression studies of human spermatogenesis?

We have designed a crosstable and rule-of-thumb values for different stages of spermatogenic impairment that define the reference cut-off values for altered gene expression in Sertoli and Leydig cells in the context of impaired spermatogenesis.

Morphometrical studies have shown that on the cellular level, impaired spermatogenesis results in a relative enrichment of somatic cell types. However, until now it is not known how this affects transcript levels in gene expression studies.

In this study, 31 testis biopsies from men with different stages of spermatogenic impairment (full spermatogenesis, hypospermatogenesis, meiotic arrest, spermatogonial presence, Sertoli cell-only syndrome, complete tubular atrophy) were used to define reference ratios of somatic transcript enrichment/dilution. The reference ratios were validated on an independent test set of 28 samples and on gene expression data from men with Y-chromosomal microdeletions.

High-quality microarray data were filtered with respect to Sertoli- and Leydig-cell-specific genes. General reference enrichment/dilution factors for these two cell types for all combinations of spermatogenic impairment were calculated using robust permutation statistics. To validate the specificity of the filtered transcripts, we calculated ratios for an independent test set of spermatogenic impairment and for transcriptional data from men with Y-chromosomal microdeletions, and checked the functional enrichment (gene ontology) and cellular localization of the corresponding proteins in a histological database and assessed their correlation with testicular size.

Filtering of Sertoli- and Leydig-cell-specific genes resulted in a set of 54 and 332 transcripts, respectively. These were used in defining robust reference dilution/enrichment factors of somatic transcripts for all spermatogenic levels and were compiled in a reference crosstable. Validation on an independent test set showed ratios within 0.5 units of our reference crosstable. Analysis of the resulting transcripts with respect to functional enrichment for Sertoli- and Leydig-cell-specific functions and protein expression, as obtained from an immunohistochemical database, indicated filtering of data sets highly enriched for Sertoli and Leydig cell function. The dilution/enrichment ratios differed significantly when transcripts were interrogated in samples with Y-chromosomal microdeletions, pointing to an overall decreased expression of somatic markers in a genetically altered background.

The defined reference ratios might apply with some restrictions in samples that display very heterogeneous histology (e.g. Sertoli cell only with a significant proportion of spermatogenic foci) or when spermatogenic impairment is a consequence of an altered genetic background.

The reference dilution/enrichment values for somatic testicular transcripts as defined in this study are to be seen as cut-off values for discriminating between simple transcript dilution/enrichment as a consequence of an altered germ cell composition and actual transcriptional regulation. Future studies dealing with transcriptional changes in testicular somatic cells in a background of altered germ cell quantities should consider these correction factors in order to avoid the description of transcriptional changes that are based simply on shifts in somatic cellular quantities.

Financial support was from grant Sp721/1-3 of the German Research Foundation. There are no competing interests to be declared.