Minor DNA methylation changes are observed in spermatozoa prepared using different protocols

Background DNA methylation patterns can show transgenerational inheritance and are influenced by lifestyle and environmental factors. It is suggested that these patterns can be changed by assisted reproductive technology. Objectives To evaluate the impact of two different sperm preparation methods, conventional density gradient centrifugation (DGC) vs. density gradient centrifugation followed by magnetic-activated cell sorting (MACS) of non-apoptotic spermatozoa, on sperm DNA methylation profile. Materials and methods We analyzed semen of patients included in our IVF treatment program. Half of the semen from each included patient was prepared for ICSI using the DGC method and the other half with DGC followed by MACS. The remaining samples were processed for DNA methylation analysis with reduced representation bisulfite sequencing (RRBS). In addition to the DNA methylation profile, we assessed the morphology and DNA fragmentation of spermatozoa. Results RRBS analysis revealed that the average genome-wide methylation level was similar between both groups (DGC vs. MACS group) and ranged from 0.53 to 0.56. Furthermore, RRBS analysis identified 99 differentially methylated regions (DMRs) and 800 differentially methylated positions (DMPs). In the DGC group, 43 DMRs and 392 DMPs were hypermethylated whereas 56 DMRs and 408 DMPs were hypomethylated compared with those in the MACS group. When DMRs and DMPs were annotated to genes, 3 genes associated with imprinting were found: IGF2, PRDM16, and CLF4/BRUNOL4. The percentage of morphologically normal spermatozoa (MACS vs. DGC; 14.0 ± 10.8 vs. 13.2 ± 10.0; P = .335) and of spermatozoa with fragmented DNA of patients with RRBS analysis (22.9 ± 21.1% vs. 34.4 ± 21.2; P = .529) were also similar between groups. Discussion and conclusion Although the average genome-wide level of sperm DNA methylation was similar in both sample groups, a distinctive number of methylation changes were observed in DMR and DMP levels. A larger number of samples should be analyzed and additional sperm preparation methods should be tested to confirm our findings.