Putative human sperm Interactome: a networks study

Abstract

BackgroundFor over sixty years, it has been known that mammalian spermatozoa immediately after ejaculation are virtually infertile. They became able to fertilize only after they reside for long time (hours to days) within female genital tract where they complete their functional maturation, the capacitation. This process is finely regulated by the interaction with the female environment and involves, in spermatozoa, a myriad of molecules as messengers and target of signals. Since, to date, a model able to represent the molecular interaction that characterize sperm physiology does not exist, we realized the Human Sperm Interactme Network3.0 (HSIN3.0) and its main component (HSNI3.0_MC), starting from the pathway active in male germ cells.ResultsHSIN3.0 and HSIN3.0_MC are scale free networks, adherent to the Barabasi-Albert model, and are characterised by an ultra-small world topology. We found that they are resistant to random attacks and that are designed to respond quickly and specifically to external inputs. In addition, it has been possible to identify the most connected nodes (the hubs) and the bottlenecks nodes. This result allowed us to explore the control mechanisms active in driving sperm biochemical machinery and to verify the different levels of controls: party vs. date hubs and hubs vs. bottlenecks, thanks the availability of data from KO mice. Finally, we found that several key nodes represent molecules specifically involved in function that are thought to be not present or not active in sperm cells, such as control of cell cycle, proteins synthesis, nuclear trafficking, and immune response, thus potentially open new perspectives on the study of sperm biology.ConclusionsFor the first time we present a network representing putative human sperm interactome. This result gives very intriguing biological information and could contribute to the knowledge of spermatozoa, either in physiological or pathological conditions.Electronic supplementary materialThe online version of this article (10.1186/s12918-018-0578-6) contains supplementary material, which is available to authorized users.

Transcriptomic and computational analysis identified LPA metabolism, KLHL14 and KCNE3 as novel regulators of Epithelial-Mesenchymal Transition

“…According to Barabási-Albert (BA) model 14,15 both networks could be classified as scale-free 34,35 . A total of 149 and 203 hubs, defined as those node highly connected 36 , were found in MC_mAEC and in MC_eAEC, respectively (Supplementary Dataset S3).…”

“…In each cell population network, 200 bottlenecks, defined as those nodes having many "shortest paths" going through them 39 , were selected, according to Ordinelli et al 36 . Analyzing the two bottleneck lists (Supplementary Dataset S5A-B), some of them resulted DEGs, then supporting their topological importance within the AEC networks.…”

“…Kernel density estimation. The KDE analysis was performed using Past3 software, according to Bernabò et al 68 and Ordinelli et al 36 .…”

Cyclin–CDK Complexes are Key Controllers of Capacitation-Dependent Actin Dynamics in Mammalian Spermatozoa

“…In the following step, we listed the nodes based on their values of bottleneck score and selected the first thirty-one nodes (the same number than hubs), in keeping with a previous study [20]. The list of the selected nodes is presented in Table 4.…”

“…Here, we adopted a computational biology approach to gather the available data in order to reconstruct the active pathway responsible for the control of AD in spermatozoa. This strategy has been recently developed and is very promising in giving important data in several branches of biology, from cancer [23] to biochemical machinery [24] to germ cells physiology [20,25,26,27]. More in detail, we used the biological networks to carry out an operation of reverse engineering in four steps: first, we collected the data from the literature (Step 1); we enriched them by using a specific systems biology tool (STRING) (Step 2); we created, analyzed, and visualized the networks (Step 3); and, finally, we took the inference about AD control mechanisms (Step 4).…”