Chromatin remodelling initiation in spermatids: differences among human males

The inherited methylome landscape is directly altered with paternal aging and associated with offspring neurodevelopmental disorders

Paternal aging and the prevalence of neurodevelopmental disorders in offspring are well documented. Yet, the underlying mechanism and the mode of inheritance have not been conclusively established. Advancing paternal age is a subtle and varying phenotype. As such, it is likely that a threshold for cumulative risk may exist that, if surpassed, culminates in a predisposition to disease and ultimately an observed phenotype in offspring. Epigenetic regulation provides a plausible explanation for the nongenetic paternal transmission of disease susceptibility. With the use of whole‐genome methylation sequencing, the data described herein substantiate an increasingly compromised DNA methylation profile as sperm ages and, for the first time, also demonstrate a generational correlation in sperm and blastocyst of an altered methylome associated with advanced paternal age. Methylation alterations are not randomly distributed across the genome, but appear clustered at certain chromosomal locations, and significantly colocalize with regions of nucleosome retention. Genes associated with autism spectrum disorder, schizophrenia, and bipolar disorder are significantly enriched with causative methylation aberrations in both sperm and embryos from aged fathers. The long‐term health burden and societal economic impact of these conditions are substantial and will continue with increasingly prevalent diagnosis. This work provides a mechanistic link between the paternal age effect and offspring neurodevelopmental disorders leading to a better understanding of causation and investigation into potential future therapy.

Paternal impacts on development: identification of genomic regions vulnerable to oxidative DNA damage in human spermatozoa

STUDY QUESTION

Do all regions of the paternal genome within the gamete display equivalent vulnerability to oxidative DNA damage?

SUMMARY ANSWER

Oxidative DNA damage is not randomly distributed in mature human spermatozoa but is instead targeted, with particular chromosomes being especially vulnerable to oxidative stress.

WHAT IS KNOWN ALREADY

Oxidative DNA damage is frequently encountered in the spermatozoa of male infertility patients. Such lesions can influence the incidence of de novo mutations in children, yet it remains to be established whether all regions of the sperm genome display equivalent susceptibility to attack by reactive oxygen species.

STUDY DESIGN, SIZE, DURATION

Human spermatozoa obtained from normozoospermic males (n = 8) were split into equivalent samples and subjected to either hydrogen peroxide (H2O2) treatment or vehicle controls before extraction of oxidized DNA using a modified DNA immunoprecipitation (MoDIP) protocol. Specific regions of the genome susceptible to oxidative damage were identified by next-generation sequencing and validated in the spermatozoa of normozoospermic males (n = 18) and in patients undergoing infertility evaluation (n = 14).

PARTICIPANTS/MATERIALS, SETTING, METHODS

Human spermatozoa were obtained from normozoospermic males and divided into two identical samples prior to being incubated with either H2O2 (5 mm, 1 h) to elicit oxidative stress or an equal volume of vehicle (untreated controls). Alternatively, spermatozoa were obtained from fertility patients assessed as having high basal levels of oxidative stress within their spermatozoa. All semen samples were subjected to MoDIP to selectively isolate oxidized DNA, prior to sequencing of the resultant DNA fragments using a next-generation whole-genomic sequencing platform. Bioinformatic analysis was then employed to identify genomic regions vulnerable to oxidative damage, several of which were selected for real-time quantitative PCR (qPCR) validation.

MAIN RESULTS AND THE ROLE OF CHANCE

Approximately 9000 genomic regions, 150–1000 bp in size, were identified as highly vulnerable to oxidative damage in human spermatozoa. Specific chromosomes showed differential susceptibility to damage, with chromosome 15 being particularly sensitive to oxidative attack while the sex chromosomes were protected. Susceptible regions generally lay outside protamine- and histone-packaged domains. Furthermore, we confirmed that these susceptible genomic sites experienced a dramatic (2–15-fold) increase in their burden of oxidative DNA damage in patients undergoing infertility evaluation compared to normal healthy donors.

LIMITATIONS, REASONS FOR CAUTION

The limited number of samples analysed in this study warrants external validation, as do the implications of our findings. Selection of male fertility patients was based on high basal levels of oxidative stress within their spermatozoa as opposed to specific sub-classes of male factor infertility.

WIDER IMPLICATIONS OF THE FINDINGS

The identification of genomic regions susceptible to oxidation in the male germ line will be of value in focusing future analyses into the mutational load carried by children in response to paternal factors such as age, the treatment of male infertility using ART and paternal exposure to environmental toxicants.

STUDY FUNDING/COMPETING INTEREST(S)

Project support was provided by the University of Newcastle’s (UoN) Priority Research Centre for Reproductive Science. M.J.X. was a recipient of a UoN International Postgraduate Research Scholarship. B.N. is the recipient of a National Health and Medical Research Council of Australia Senior Research Fellowship. Authors declare no conflict of interest.

Nuclear envelope remodelling during human spermiogenesis involves somatic B-type lamins and a spermatid-specific B3 lamin isoform

The nuclear lamina (NL) is a filamentous protein meshwork, composed essentially of lamins, situated between the inner nuclear membrane and the chromatin. There is mounting evidence that the NL plays a role in spermatid differentiation during spermiogenesis. The mouse spermatid NL is composed of the ubiquitous lamin B1 and the spermatid-specific lamin B3, an N-terminally truncated isoform of lamin B2. However, nothing is known about the NL in human spermatids. We therefore investigated the expression pattern and localization of A-type lamins (A, C and C2) and B-type lamins (B1, B2 and B3) during human spermiogenesis. Here, we show that a lamin B3 transcript is present in human spermatids and that B-type lamins are the only lamins detectable in human spermatids. We determine that, as shown for their mouse counterparts, human lamin B3, but not lamin B2, induces strong nuclear deformation, when ectopically expressed in HeLa cells. Co-immunofluorescence revealed that, in human spermatids, B-type lamins are present at the nuclear periphery, except in the region covered by the acrosome, and that as the spermatid matures the B-type lamins recede towards the posterior pole. Only lamin B1 remains detectable on 33-47% of ejaculated spermatozoa. On spermatozoa selected for normal head density, however, this fell to <6%, suggesting that loss of the NL signal may be linked to complete sperm nucleus compaction. The similarities revealed between lamin expression during human and rodent spermiogenesis, strengthen evidence that the NL and lamin B3 have conserved functions during the intense remodelling of the mammalian spermatid nucleus.

A mutation study of sperm head shape and motility in the mouse: lessons for the clinic

Mouse mutants that show effects on sperm head shape, the sperm tail (flagellum), and motility were analysed in a systematic way. This was achieved by grouping mutations in the following classes: manchette, acrosome, Sertoli cell contact, chromatin remodelling, and mutations involved in complex regulations such as protein (de)phosphorylation and RNA stability, and flagellum/motility mutations. For all mutant phenotypes, flagellum function (motility) was affected. Head shape, including the nucleus, was also affected in spermatozoa of most mouse models, though with considerable variation. For the mutants that were categorized in the flagellum/motility group, generally normal head shapes were found, even when the flagellum did not develop or only poorly so. Most mutants are sterile, an occasional one semi-sterile. For completeness, the influence of the sex chromosomes on sperm phenotype is included. Functionally, the genes involved can be categorized as regulators of spermiogenesis. When extrapolating these data to human sperm samples, in vivo selection for motility would be the tool for weeding out the products of suboptimal spermiogenesis and epididymal sperm maturation. The striking dependency of motility on proper sperm head development is not easy to understand, but likely is of evolutionary benefit. Also, sperm competition after mating can never act against the long-term multi-generation interest of genetic integrity. Hence, it is plausible to suggest that short-term haplophase fitness i.e., motility, is developmentally integrated with proper nucleus maturation, including genetic integrity to protect multi-generation fitness. We hypothesize that, when the prime defect is in flagellum formation, apparently a feedback loop was not necessary as head morphogenesis in these mutants is mostly normal. Extrapolating to human-assisted reproductive techniques practice, this analysis would supply the arguments for the development of tools to select for motility as a continuous (non-discrete) parameter.