Fig. 3

ERE in the sperm fractions sorted based on CMA3− uptake. We compared the observed number of differentially methylated estrogenic and non-estrogenic Alu sequences and other short transposable sequences based on the expected number in the fractions of spermatozoa either stained or not with CMA3 and in the sperm fractions of infertile patients versus healthy semen donors. Differences between the observed and expected numbers were evaluated with the Chi-squared (X²)-test (degrees of freedom, DoF: 1) (A). The observed number of hypomethylated estrogenic Alu sequences in the condensed CMA3− sperm fraction of infertile patients was significantly lower than expected, whereas the observed number of hypomethylated estrogenic Alu sequences was significantly higher in the CMA3+ sperm fraction of infertile patients (p = 0.0137). Conversely, the observed number of hypermethylated estrogenic Alu sequences was significantly higher than the expected number in the CMA3− fraction of fertile semen donors (p = 0.003). Based on the calculated X²-values of the estrogenic sequences, heat maps were constructed to visualize differences between observed and expected number of sequences based on CMA3− uptake (B). In the comparisons, in which the observed numbers exceeded the expected number, the X²-values were colored red (positive), whereas in the comparisons, in which the observed number remained below the expected number, the X²-values were colored blue (negative)