A Game Network Model Of Gene-driven Cell Interactions In Primordial Follicles In Primordial Follicles

Primordial follicles are of utmost importance to maintain female fertility reserve,determining a woman's fertility and reproductive lifespan.A primordial follicle is regarded as a functional complex in which germ cells and granulosa cells determine folliculogenesis through their intricate but wellorchestrated interaction networks.Single-cell sequencing technique has made it possible to collect a torrent of gene expression data at the individual cell level,providing valuable resources for characterizing the molecular mechanisms of germ-soma interactions.However,not in parallel to the increasing data from various reproductive laboratories,there is little methodology to analyze these data and further extract general biological principles from them.This scarcity has significantly limited the best use of these data and the further development of reproductive medicine.The purpose of this dissertation is to develop a computational model for unveiling the genomic signatures that guide the interactions pattern between germ cells and granulosa cells during follicle development.The model cross-pollinates different disciplines ideas into a unified conceptual framework under which cell interactions at different scales can be classified,curated,and quantitated.A primordial follicle can be viewed as an ecosystem in which germlines and somas,as the two underlying compartments,interact with each other through lens of game theory.The expression of a gene on one compartment is determined jointly by the strategy of this compartment and the strategy of its counterpart.To better unravel interaction variation and trajectories,such a game theory should be extended to evolutionary game theory that models how player strategies change over time.However,it is very difficult to monitor the temporal expression profiles of individual genes within a single cell,making the direct use of evolutionary game theory impossible.To overcome this issue,we introduce allometric scaling law by creating a new concept of niche index,to derive a system of quasi-dynamic ordinary differential equations(qd ODEs)for characterizing how germlines interact with somas across different primordial follicles.The qd ODE model was used to analyze single-cell RNA-seq data collected from human female embryos,obtaining the following results:(1)Cell interactions across germlines and somas obey different patterns.The model classifies these interactions into five types,synergism,antagonism,directional synergism,directional antagonism,and altruism/selfness.A majority of genes(70%)mediate synergistic relationships,whereas only 4% incurs antagonism,which complies with the general rule that positive promotion dominates gene interactions in eukaryotes.(2)Cell development is controlled by particular gene networks.Gene networks of germlines and somas are similar in structural mainstay but different in organizational details.Both networks are dominated by directional synergism and directional antagonism,a phenomenon also observed in natural macro-communities.Hub genes change reciprocally between the two networks;i.e.,a hub as a regulator in the germline network becomes a target in the soma network and vice versa and bridge a functional link between them.Germlines play a leadership role in follicle development.(3)Genes determine how meiotic germlines(late)sense the signals transduced from miotic germlines(early).NODAL is a signaling pathway that serializes germlines from their early stage to late stage.The qd ODE model has successfully identified specific genes interaction pattern during this process,providing an important clue for dissecting the developmental trajectories of primordial follicles.(4)Gene-driven competition and cooperation are drivers of cell heterogeneity.There is strong and widespread cell heterogeneity within a developing embryo,which is thought to facilitate natural selection for superior oocytes.Results by the qd ODE model suggest that germ cells in the same stage tend to compete with each other while those in different stages tend to cooperate.The genomic machineries underlying cell-cell competition and cooperation help interpret how germline cells undertake “the survival of the fittest”.In summary,this dissertation presents one of the first models that can characterize and quantitate the molecular mechanisms underlying primordial follicles.This model has identified the previously unknown inner workings that govern cellular heterogeneity,interactions and evolution within primordial follicle.