In the age of Assisted Reproductive Technology (ART), where defective gametogenesis is no longer a barrier to procreation, it is crucial to understand how the genome is normally processed in germ cells, detect cases where processing is abnormal and determine the consequences of anomalies for genome stability. We will thus characterize the nuclear lamina during human spermatogenesis, and discover proteins required for nuclear remodelling by searching for genetic causes of abnormal sperm head development (teratozoospermia) and the creation of specific knockout mouse models. Our group has a strong clinical component and enjoys close collaborations with La Conception and La Timone Hospital sites: the Reproductive Biology Laboratory, Centre for the study and conservation of human oocytes and sperm (CECOS, directed by C. Metzler-Guillemain), the Molecular Genetics Laboratory in the Department of Medical Genetics and the Germethèque, a labelled Biobank. We also have a collection of normal testicular material, from patients with brain death, deposited in the Germethèque Biobank and authorized by the Biomedicine Agency (ABM).

Over the last four years, we have characterised the nuclear lamina (NL) and its associated proteins during spermiogenesis and in the mature spermatozoon, in human. During spermiogenesis the round spermatids differentiate into highly-specialised motile spermatozoa, involving dramatic remodelling of the chromatin and the nucleus. We have therefore focused our attention on defining the interface between the NL and the chromatin. Our aims were to improve understanding of the roles of the NL during spermatogenesis and to identify biomarkers of gamete quality.

• We have determined the composition of the NL in human spermatids. This had previously only been investigated in rodents 47, 48. We have shown that, in human, the NL is composed exclusively of type-B lamins, including lamin B3, a spermatid-specific isoform of lamin B2. We showed that lamin B3 destabilises the NL and induces deformation of the nucleus when expressed in HeLa cells, indicating that lamin B3 may facilitate nuclear remodelling in spermatids by rendering the NL less rigid 49.
• The NL has been shown to be linked to chromatin in somatic cells by LBR and LEM-domain proteins interacting with the chromatin protein BAF 50, 51. We have explored the NL-chromatin interface during spermiogenesis by studying LBR, seven LEM-domain proteins, and two BAF proteins during spermiogenesis. Only LAP2 had previously been studied during spermiogenesis, in rodents 52. We showed that five of these proteins LAP2, LEMD1, LEMD2 (a short isoform only), BAF and BAF-L, are present in the round spermatid nucleus, but retreat to the posterior pole of the nucleus as the acrosome spreads. We determlined that the full-length isoform of LEMD2, known to contribute to nuclear integrity in somatic cells, was absent from spermatids. Only BAF and BAF-L were retained in elongated spermatids and were detected at the posterior pole of the spermatozoa nucleus by immunofluorescence, and by Western blot. In somatic cells, BAF is known to be required for the re-assembly of the nuclear envelope after mitosis 53, 54. We have therefore proposed that BAF, together with BAF-L, could play roles in the ordered condensation of the spermatid chromatin, its decondensation in the oocyte and the reassembly of the nuclear membrane around the male pronucleus after fertilisation. (Elkhatib et al., 2017, Reproduction – doi: 10.1530/REP-17-0358).
• We have investigated the localisation of the proteins that we find in normal human spermatozoa and in spermatozoa from infertile men with teratozoospermia. We have completed a first study of spermatozoa from men with globozoospermia due to mutations in the DPY19L2 gene. The sperm in these men have a round head, detached acrosome and retain histones 55. We found that lamin B1 is retained throughout the nuclear periphery of globozoospermic spermatozoa, but that neither BAF nor BAF-L is detectable. Furthermore the mRNA for BAF, which we have shown is normally absent from spermatozoa, was detected in RNA extracted from globozoospermic spermatozoa, suggesting that the BAF transcript could be a biomarker of abnormal chromatin remodelling in human spermatozoa. (Paci et al., submitted).
• Finally we have launched whole exome analysis to identify genetic causes of sperm head anomalies found in infertile men. Preliminary results have been encouraging with identification of a homozygous loss-of-function variant in a spermatid-specific nuclear envelope protein in an infertile man who produces spermatozoa with enlarged sperm heads. We have also identified the second reported case of a loss of SUN5 function in men with decapitated spermatozoa syndrome, establishing that SUN5 is required for the formation of the sperm nucleus-flagellum junction. (Elkhatib et al., 2017, Human Molecular Genetics, doi: 10.1093/hmg/ddx200)

We have thus established a solid basis of knowledge concerning the structure and behaviour of the NL during human spermiogenesis, on which we have built our project. An important focus will be the condensation and decondensation of the paternal genome before and after fertilisation, processes of vital importance for male fertility, and for the stable transmission of the human genetic and epigenetic identity to future generations.

Our current scientific objectives are: 

1) Determine the role of the NL, BAF and BAF-L in nuclear remodelling during spermiogenesis and the formation of the male pronucleus in the oocyte after fertilisation.
To gain insights into nuclear remodelling, we will identify protein partners of lamin B1, lamin B3, BAF and BAF-L in spermatids and spermatozoa. Knockout mouse models are underway to identify proteins with key functions in the nuclear remodelling that occurs before and after fertilisation. We will use ChIP seq to define the genomic regions to which BAF and BAF-L are bound in the sperm nucleus.

2) Define new biomarkers for human spermatozoa quality. We have identified specific B-type lamins and lamin partners that are present on human ejaculated spermatozoa. We will evaluate these proteins as biomarkers by measuring their presence or absence on quality-selected spermatozoa and different types of teratozoospermic spermatozoa.

3) Characterize the NL and the NL-chromatin interface during meiosis.
In rats, the spermatocyte NL is composed of lamin B1 and lamin C2, a meiotic specific isoform 56. The lamin C2 appears in the form of discontinuous domains associated with meiotic telomeres 57. In mice expressing lamin A/C but lacking lamin C2, spermatogenesis blocks during meiosis 58. In humans, there is no published data on the nature of the NL and partner proteins in male meiosis. We will identify specific lamin isoforms and NL-associated proteins in human meiotic cells, and follow their localisation in spermatocytes through each step of human meiosis.

4) Use exome analyses to identify causal mutations in men with abnormal sperm heads. Recruitment through the Reproductive Biology Laboratory is around 10-15 cases per year. Based on phenotype, pre-screening will exclude cases with causal mutations in AURKC (macrocephalic sperm), DPY19L2 and SPATA16 (globozoospermia) (estimated 30-50% of cases). Variants in genes encoding components of the nuclear envelope will be validated by modelling in the mouse.
By exome analysis, we have identified a homozygous loss-of-function variant predicted to inactivate a spermatid-specific component of the nuclear envelope, in a man with macrocephalic haploid spermatozoa in which histones have not been replaced by protamines. We will create and characterise a mouse model to gain insights into the role of the nuclear envelope in chromatin remodelling during spermiogenesis.