In this study we revealed epigenetic mechanisms that are likely to be involved in the expression of key genes in spermatogonia, namely Pou5f1 and Gfra1. These experiments implicate histone H3 methylation and acetylation in the regulation of their expression and suggest that transcription is influenced by HDACs and KDM1. The underlying molecular mechanisms revealed by pharmacological treatments of the GC-1 cell line suggest that in spermatogonia expression of Pou5f1 and Gfra1 is maintained by an enrichment of histone H3 acetylation and dimethylation of histone H3 at lysine 4. Based on our results we suggest that as spermatogonia differentiate the loss of expression of Pou5f1 and Gfra1 may be concomitant with gene-specific activity of HDACs and KDM1 and the consequent removal of gene activating histone H3 acetylation and H3K4 dimethylation. These epigenetic changes are summarized in a model shown in .
Epigenetic remodeling in Pouf5f1 and Gfra1 promoter regions - a proposed model to explain changes in gene expression in GC-1 cells after Tranylcypromine and TSA treatment.
Histone acetylation is intimately involved in the epigenetic regulation of transcription and at a single gene locus a fine balance of histone acetyltransferase and deacetylase activity controls gene expression 
. Histone hyperacetylation is associated with an open chromatin state and general transcriptional activity 
. Inhibition of histone deacetylase activity by TSA has many faces: it can for instance result in de-differentiation of primordial germ cells into embryonic germ cells 
, accelerate murine stem cell differentiation 
, force cancer cells into a differentiated state 
, or in concert with a DNA methyltransferase inhibitor convert neurosphere cells into neurosphere-derived haematopoietic cells 
. While we did not observe changes in cell morphology or behavior to indicate the GC-1 cell line altered its differentiation state, we did observe specific changes in gene expression that were associated to gene-specific changes in the epigenetic profile. In the brief time-scale of these experiments changes in differentiation would not be expected. For example in the case of induced pluripotent stem cells, virally transduced pluripotency factors (e.g. Pou5f1
) need to be expressed for approximately twelve to eighteen days to initiate de-differentiation and reprogramming of mouse embryonic fibroblast cells 
Interestingly TSA treatment of the GC-1 cell line turned on specific genes expressed in spermatogonia as we saw an activation of spermatogonia cell markers Gfra1
. The lack of change in expression in the panel of genes examined suggests that expression of Pou5f1
and Gfra1 i
s a consequence of gene-specific epigenetic effects of the HDAC and KDM1 inhibitors and not due to global changes to the epigenome. In the present study, induced Pou5f1
expression was associated with increased H3K4 dimethylation, H3K9 acetylation and pan-H3 acetylation at regulatory regions of the mouse Pou5f1
promoter. Although we demonstrated that Pou5f1
expression was maximal under the co-suppression of KDM1 and HDAC1, a closer look at the epigenetic marks on the promoter revealed that co-inhibition of HDAC and KDM1 did not further enhance gene activating H3K4 methylation marks. Therefore the enhanced gene activation under KDM1/HDAC1 inhibition may be a consequence of indirect KDM1/HDAC1 signaling. We postulate that the inhibition by TSA on HDACs and the consequent increased H3K4 dimethylation was due to an indirect effect on KDM1 since TSA does not target demethylases, but specifically inhibits histone deacetylases by binding to the catalytic core 
. Evidence supporting this potential molecular mechanism comes from experiments showing that hyperacetylation of lysine residues on histone H3 abolished or attenuated KDM1 function, and that acetyl groups had to be removed before H3K4 demethylation by KDM1 occurred 
. Moreover we detected KDM1 and HDAC at our target genes promoters suggesting that effects are a consequence of drug action on the KDM1/HDAC complex.
The gain in histone H3K4 dimethylation at the Pou5f1
promoter and associated gene activation we report here following HDAC inhibition and under the combined suppression of KDM1/HDAC is similar to what has been observed in embryonal carcinoma cells 
. However, unlike the GC-1 cells, embryonal carcinoma cells increased Pou5f1
expression when KDM1 was inhibited by T alone 
. The differences in cellular response to T between the embryonal carcinoma cells and GC-1 cell used here may be due to differences in sensitivity to inhibition of KDM1 by T. Unlike GC-1 cells, embryonal carcinoma cells express high levels of KDM1 and have detectable levels of Pou5f1
under control conditions 
, the expression of Gfra1
in GC-1 cells seemed to be directed principally by HDACs since treatment with TSA caused a strong increase in activating H3K9- and total H3 acetylation in the Gfra1
promoter region and consequent gene expression. In contrast to the major changes in acetylation only a slight increase in gene activating H3K4 dimethylation under the combined inhibition of KDM1/HDAC1 by TSA and T occurred, and this was not associated with enhanced expression in comparison to TSA alone. In this study there was no effect of inhibition of KDM1 by T alone on global histone modifications or changes in gene expression. Rather, treatment with T was effective only when used in combination with TSA. These effects are in line with previous studies where it has been reported that there was a lack of globally increased histone H3 methylation after genetic ablation of KDM1 in murine ES cells 
, and following the suppression of KDM1 by T, or its knockdown by siRNA in human cervical cancer cell lines 
. Taken together these findings suggest that KDM1 activities are highly gene-specific and likely to involve its function in a protein complex with HDAC and other proteins.
At some gene promoters there is crosstalk between histone modifications and DNA methylation. In a previous study Ou and coworkers associated a TSA induced increase of histone acetylation, with a significant decrease in global DNA methylation 
. It was thereby pertinent to assess the DNA methylation status at the promoter regions of Pou5f1
. Overall no significant changes in DNA methylation occurred in either the Pou5f1
promoter indicating that histone methylation and acetylation are the principles means of regulating gene expression at these promoters in GC-1 cells. However, depending on the treatment and the position of the CpG dinucleotides we observed a slight but significant decrease in DNA methylation in the analyzed Pou5f1
promoter region, but none in the Gfra1
promoter. We cannot completely rule out that the changes in DNA methylation observed in the Pou5f1
promoter had an impact on the induction of its expression in treated GC-1 cells. But this seems to be unlikely as >60 to 70% methylation remained following treatment at the Pou5f1
promoter and low methylation was found on the Gfra1
As the mechanisms and interpretation discussed here are based on a transformed spermatocyte-like cell line, this underlies the extrapolation to events in vivo
. The use of an immortalized cell line such as the GC-1 holds certain limitations but the positive factors outweighed the negatives when choosing a model system and examining the alternatives. The GC-1 cells represent a relatively homogenous cell line that can be readily and robustly expanded without being co-cultured with somatic cells, while sustaining germ cell characteristics 
. Moreover they are easy to manipulate and deliver an unlimited source of material, which is critical for ChIP experiments, which require millions of cells per treatment for analysis.
Importantly this study has revealed epigenetic mechanisms likely to be involved in the complex gene regulation in spermatogenesis. Worth considering is that environmental induced epigenetic mutations in developing germ cells may have consequences for lifetime fertility as there is likely to be plasticity in histone modifications.