Impact of Weight Loss on the Human Sperm Epitranscriptome (EPI-SPERM)

1. Background and present state of the art in the research field. Sperm non-coding RNA is an important vector for epigenetic inheritance of diet-induced obesity Obesity is a multifactorial pathology associated with a high heritability (50-75%). To date, multiple Genome Wide Association Studies (GWAS) have attempted and failed to identify the genetic factors that contribute to the etiopathogenesis of obesity. In fact, the variance explained by a high Body Mass Index (BMI) at these GWAS mutations is very low, estimated to be approximately 2%[1]. Recent epidemiological and experimental studies strongly suggest that this high heritability of obesity might be associated with the transmission of obesity-induced epigenetic modifications.

   Non-genetic inheritance of a newly acquired phenotype is a concept in biology whereby changes induced by a specific signal and/or environmental stress can be passed to the next generation in the absence of a genetic change [2, 3]. The biological relevance of this process is proven by its diversification in various organisms including plants, insects, nematodes, and mammals [4, 5]. However, evidence is now emerging that a wide range of dietary factors including fats and proteins lead to changes in gene expression that are maintained through mitotic and meiotic divisions[6]. These changes can significantly impact the health of the offspring.

   Very recently, the investigator has gained some insight into the molecular mechanisms involved in this process, in particular those that pertain to the epigenetic inheritance of newly acquired paternal pathologies [7] However, much remains to be uncovered about the functional and molecular characterization underlying the mode of action of the specific factors responsible.

   Changes in DNA methylation [8-10], chromatin modification [11] and expression levels of non-coding small RNA, including miRNA, piRNA and transfer RNA fragments [12-14] occur in germ cells upon changes in environmental cues. However, it is not clear whether these changes would mediate epigenetic inheritance. While several studies have indicated that an altered DNA methylation signature of spermatozoa from HFD-obese individuals could be transmitted to progeny [8], a recent study indicated that the sperm methylome is shaped by genetic and epigenetic variations, but not diet [15], thus disproving the idea that DNA methylation is the vector of epigenetic inheritance in this HFD model.

   Based on this first evidence showing that small non-coding RNAs (sncRNAs) act as trans-generational vectors of epigenetic information in mice [16], the investigative team and others extended this discovery by deep investigation of the role of sperm small RNAs as determinants of the inheritance in an acquired metabolic disorder [12-14]. In this regard, experimental approaches include the micro-injection of one-cell embryos with well-defined sperm RNAs from individuals fed by high-fat (HFD) or control diet (CD). They found that mice derived from the microinjection of HFD sperm RNAs into naive one-cell embryos developed adult onset diet-induced pathologies, such as obesity or signs of type 2 diabetes even though they had been fed a control diet [14]. Deep-sequencing analysis of small RNAs of testis from HFD-fed animals and control revealed that several small-RNAs are indeed deregulated upon HFD. Among those small RNAs, we found miRNAs, piRNAs and tRNA-derived small RNAs (tsRNA). Importantly, microinjection into naive one-cell embryos of one of the deregulated microRNAs - namely microRNA-19b, induced metabolic alterations that were similar to the diet-induced phenotypes. Likewise, other groups found that tsRNA may also contribute to intergenerational inheritance of metabolic disorders. Indeed, Chen et al. showed that micro-injecting sperm tsRNAs from HFD-fed individuals into naive one-cell embryos caused impaired glucose tolerance in the resulting offspring [12]. However, in this model, synthetic microRNAs or synthetic tsRNAs did not induce metabolic disorders in the offspring, suggesting a role of post-transcriptional RNA modifications in the transfer of epigenetic information. This hypothesis is consistent with the elevated levels of m5C and m2G modifications of tsRNAs in sperm from the HFD group [12]. Hence, sperm small RNAs represent a type of paternal epigenetic vector involved in intergenerational inheritance of diet-induced metabolic phenotypes.

   The majority of data on RNA-mediated epigenetic inheritance are from mouse models [12-14]. Whether this mechanism is conserved in human as yet unknown. Noteworthy, a study performed in human by Donkin et al. indicated that obesity might modify the sperm epigenetic landscape as well as sperm epitranscriptome. However, epigenetic and transcriptomic analysis were performed in only a small number of men[17] Taken together these observations from human and rodent studies suggest that obesity-induced small RNA changes are transmitted to the next generation, and then result in heritable metabolic changes. However, although in rodents the obesity-induced changes have been quite well described, the impact of obesity on the sperm epitranscriptome/epigenome in human remains an open question of urgency in the field.

   Reversal of the obesity-induced epigenetic changes: a hope for the next generation? Although epigenetic modifications are mitotically heritable, they are erased and re-established twice during development [18]. This reprogramming takes place early in embryogenesis and during gametogenesis in primordial germ cells that will give rise to eggs and sperm. [18]. Furthermore, phenotypes conferred by environmental compromise can also be reversed via active intervention, such as change to healthy diet or physic exercise during the preconception period [19, 20]. It is not understood whether there is a relationship between developmental reprogramming and reversibility of environmentally induced phenotypic states. Although the erasure of these newly established modifications has been clearly described in rodents [20, 21] only limited studies have been addressed in human [17, 22, 23]. Those studies only focused on DNA methylation pattern modulated by obesity in somatic and germ cells. To date, nothing is known about the reversibility of trans-generational epigenetic RNA in human sperm.

   Preliminary results Obesity negatively impacts on sperm epigenome and early embryonic development in human

   After having demonstrated the role of sperm RNA (small and long RNA) in paternal heredity of obesity and its associated diseases, the investigative team is currently investigating the relevance of sperm RNA-mediated epigenetic inheritance of metabolic disorders in human. To achieve this goal, they have carried out two complementary sets of experiments. First, they recently found that the development capacity of human embryos derived from obese men was impaired and noted that pre-implantation embryo morphokinetic parameters were altered when sperm derived from obese fathers was used to fertilize oocytes (Raad et al. 2019), suggesting obesity induced epigenetic changes in sperm. Then, to determine whether obesity induced sperm small-non coding RNA dysregulation in human, they performed a small RNA-seq differential expression analysis between obese and non-obese subjects. Briefly, upon multivariant analysis (Principal Component Analysis, PCA), they identified distinct small RNA profiles between obese and lean men. However, while sperm small RNA profiles among lean men (N1, N2, N3 and N4 samples) were homogeneous, small RNA profiles among obese men (O1, O2, O3 and O4 samples) were very heterogeneous, suggesting stochastic dysregulation of sperm small RNA expression in obesity.

   Together, our results highlight the negative impacts of excessive weight mass on the sperm epi-transcriptome and human early embryo development and strongly suggest that paternal obesity negatively affects the sperm epigenome and subsequently influences the health of the descendants.

2. Descriptive of the objectives Based on the published and unpublished data described above, the investigative team hypothesize that epi-transcriptomes are responsive to transient environmental changes such as obesity, leading to modulation of the progeny's epigenomes in human and subsequently to the development of the father's phenotypes.

To validate this hypothesis, the investigator will use sperm collected from 2 groups of men: the first group will be composed of non-obese men and the second one will be composed of obese men taken before and after weight loss induced by surgery. Thanks to these cohorts, we propose to:

• Identify the sperm epitranscriptomic signature of obese men Team 1 has recently demonstrated that sperm RNAs are vectors of epigenetic inheritance of obesity and its associated pathologies in mice. In parallel, while confirming the role of these RNAs in epigenetic inheritance, two other independent groups have shown the involvement of RNA posttranscriptional modifications in this process in mice. However, it is important to determine whether such process exists for human. By combining the complementary expertise of three groups in either epigenetic inheritance in mice (Team1) and in the handling and collection of human sperm samples (Teams 2 and 3), the investigative team will have the opportunity to answer this unresolved question.
• Study the reversibility of the sperm epitranscriptomic signature of obese men The question of reversibility of sperm epigenetic signature induced by obesity in human is an important question in term of public health. Barrey's group has recently shown a possible reversion of obesity-induced epigenetic upon surgically incurred-weight loss. However, this study relies on sperm DNA methylation studies [17]. The role of DNA methylation as vector of epigenetic inheritance of obesity remains to be established, however, we and others have recently demonstrated the involvement of sperm epitranscriptome in this process in mice, [12, 14]. The possible reversibility of these obese-induced epitranscriptomic changes remains to be demonstrated both in mice and in human.

The impact of bariatric surgery with weight loss on metabolic health in progeny of obese fathers could be, in principle, studied directly on human cohorts. However, the few longitudinal studies currently available from offspring born after parental weight loss are based on maternal weight loss studies [22, 24, 25]. These results are not only inconsistent but also biased by the fact that the surgery procedure leads to malnutrition risks in the mother. Given the importance of nutrition during pregnancy, the potential impact of surgical weight loss on gametic epigenomic reversibility and on metabolic health of offspring could not be evaluated in this context. By performing comparison of epitranscriptomic profiles between obese men before and after surgery, the investigative team will bring new insights into this very important biological question.

The investigative team expect that this proposal will provide clues about the molecular mechanisms involved in this process, and will allow the identification of obese-susceptibility loci as well, which their expression may be modulated by environmental factors and consequently affects the inheritance of the disease.