Fig 1.
A 2-step diet intervention results in systemic metabolic responses and increased sperm motility.
(A) Participants (n = 15) were given a highly controlled standard diet for 1 week (100% of RDI based on their TEE), followed by a week with additional sugar (+50% of RDI). (B) Changes in fat and fat free mass were estimated by BodPod measurements. (C) Serum triglycerides during the test period. (D) Semen was collected from each participant at the beginning of the study and at the end of each week. (E) Sperm motility during the test period. Data are available in S2 Table. RDI, recommended daily intake; TEE, total energy expenditure.
Fig 2.
Human sperm tRNA fragments are acutely sensitive to high-sugar diet.
(A) Small RNA-seq profiles from motile sperm of each participant (S1-S16; n = 15) were analyzed at the experimental start point (“Start”), after the first week of healthy diet (“Healthy”), and after the second week of high-sugar diet (“Sugar”). (B) Mean proportion of small RNA across the experiment. (C) The intervention was primarily designed for investigating the effect of Sugar compared to Healthy (baseline). (D) Fold changes in biotypes after 1 week of high-sugar diet (Sugar/Healthy). (E) Types of tsRNA analyzed. (F) Mean number of reads for the different types of tsRNAs. (G) Fold changes of tsRNA types after 1 week of high-sugar diet. (H) Top expressed nuclear and mitochondrial tRNA isodecoders, their mean expression, fold change after 1 week of high-sugar diet, and their composition of tsRNA types. Only tRNA isodecoders with at least 100 RPM are presented. Error bars indicate ± SEM. “*” Indicates at least p < 0.05. Graphs can be reproduced using the script in S1 Text with input from S1 and S2 Data. FC, fold change; lincRNA, long non-coding RNA; miRNA, microRNA; Mt, Mitochondrial; piRNA, piwi interacting RNA; RPM, reads per million; rsRNA, rRNA-derived small RNA; tsRNA, tRNA-derived small RNA.
Fig 3.
Mitochondrial and nuclear tsRNA form distinct clusters.
Dendrogram shows the relatedness of the regulatory responses of individual tsRNAs from the significantly changed tRNA isodecoders in (Fig 2H). Each leaf represents the differences between Sugar and Healthy in a single tsRNA across the 15 participants. Mitochondrial (red) and nuclear i-tsRNAs (blue) shows clear division into 2 clusters. Hierarchical cluster analysis was based on Euclidean distances across center scaled RPM differences between the Sugar and the Healthy sample of each participant. For more information about these tsRNAs, see S5 Table. The graph can be reproduced using the script in S1 Text with input from S2 Data. RPM, reads per million; tsRNA, tRNA-derived small RNA.
Fig 4.
Sperm from obese men display changes in tsRNAs.
(A) Previously published data by Donkin and colleagues on sperm small RNA from obese and lean men (SRA project: SRP065418) [25] was reanalyzed using our analytical workflow. (B) Boxplot show mean, and individual expression, of the different tsRNA types (grey dots lean individuals, black dots obese individuals). Error bars indicate ± SEM. (C) Shows a significant relationship between average tsRNA expression across the current (Sugar/Healthy) and the obesity (Obesity/Lean) studies. Each colored large dot represents individual tsRNAs from significantly changed, sugar-sensitive tRNA isodecoders in Fig 2H. Grey small dots represent other tsRNA. (D) Same tsRNAs as in panel C but plotted as the differences within each of the two studies, comparing Sugar versus Healthy diets, and Obese versus Lean men, respectively. Percentages represent the proportion of tsRNA found in each quadrant of the plot; not bracketed = diet sensitive; in brackets = all analyzed tsRNA. RPM, reads per million; SRA, Sequence Read Archive; tsRNA, tRNA-derived small RNA.
Fig 5.
T-loop cleavage generates sugar-sensitive nitRNAs.
(A–D) Nucleotide coverage and cleavage site analysis. (E–H) Graphical representation of tsRNAs generated in response to a high-sugar diet. (I) Shared T-loop sequence (TΨCGA) across sugar-sensitive nuclear tsRNAs. (J) Positional analysis of 3′ cleavage sites including all tsRNAs in the study. Top panel: Sugar (dark grey) versus Healthy (light grey) diet. Middle and bottom panels: significant tsRNAs (filled colored dots; p < 0.05), nuclear tsRNAs (blue), and mitochondrial tsRNAs (red). Data are available in S2 Data and S4 Table. nitRNA, nuclear internal T-loop tsRNA; RPM, reads per million; tsRNA, tRNA-derived small RNA.
Fig 6.
Sperm motility is positively associated with parallel changes in tsRNA.
Nuclear and MT tsRNA—either sugar-sensitive (= significantly up-regulated in Sugar versus Healthy) or other (= not significantly up-regulated tsRNA)—were reevaluated in relation to Start point values. (A) Shows the changes in nuclear sugar-sensitive nitRNA (dark points) and other nuclear tsRNA (light points) over the whole diet intervention. Note that sugar-sensitive nitRNA specifically responded to Sugar. (B) Changes in sperm motility between Sugar versus Start diets were associated with sugar-sensitive nitRNA (dark points), but not other nuclear tsRNA (light points). (C) No association between changes in nitRNA and other nuclear tsRNA. (D) Changes in sugar-sensitive MT tsRNA (dark points) and other MT tsRNA (light points) in relation to Start point values. Note that all MT tsRNA progressively increased over the diet intervention, but a boost was seen in sugar-sensitive after Sugar. (E) Changes in both sugar-sensitive MT tsRNA (dark points) and other MT tsRNA (light points) were associated with changes in sperm motility. (F) Changes in sugar-sensitive MT tsRNA and other MT tsRNA were strongly correlated. ****p < 0.0001, **p < 0.01, *p < 0.05, # p < 0.1. Data are available in S2 Table (sperm motility) and S2 Data (tsRNA). Healthy, samples taken after healthy diet; MT, mitochondrial; nitRNA, nuclear internal T-loop tsRNA; n.s., not significant; Start, start point samples; Sugar, samples taken after high-sugar diet; tsRNA, tRNA-derived small RNA.
Fig 7.
Alternative hypotheses for rapid responses to diet in human sperm.
Since endogenous nuclear gene transcription in late-stage spermatozoa is strongly repressed, rapid responses are likely dependent on the activation of latent factors already available within the sperm or on the transfer of critical factors from surrounding somatic cells. Possible mechanisms for latent factor activation involve direct sperm sensing of dietary nutrients present in seminal fluid or intercellular signaling by receptor-ligand binding. Exosomes, which are small extracellular vesicles, known to transfer sncRNA between cells, are candidates for soma-to-sperm transfer of sncRNA such as tsRNA, but also other factors affecting sperm motility. In humans, epididymal principal cells and prostate acinar cells in the male reproductive tract are known to release such exosomes into seminal fluid. Thus, exosome transfer is the only known mechanism that may increase nuclear sncRNA in late-stage sperm by de novo transcription. Transcription of mitochondrial sncRNA in mature sperm is less well understood. sncRNA, small non-coding RNA; tsRNA, tRNA-derived small RNA.