Dipeptide metabolite, glutamyl-glutamate mediates microbe-host interaction to boost spermatogenesis

Abstract

The decrease in sperm count and infertility is a global issue that remains unresolved. By screening environmental bacterial isolates, we have found that a novel lactic acid bacterium, Lactiplantibacillus plantarum SNI3, increased testis size, testosterone levels, sperm count, sexual activity and fertility in mice that have consumed the bacteria for four weeks. The abundance of L. plantarum in the colon microbiome was positively associated with sperm count. Fecal microbiota transplantation (FMT) from L. plantarum SNI3-dosed mice improved testicular functions in microbiome-attenuated recipient animals. To identify mediators that confer pro-reproductive effects on the host, untargeted in situ mass spectrometry metabolomics was performed on testis samples of L. plantarum SNI3-treated and control mice. Enrichment pathway analysis revealed several perturbed metabolic pathways in the testis of treated mice. Within the testis, a dipeptide, glutamyl-glutamate (GluGlu) was the most upregulated metabolite following L. plantarum SNI3 administration. To validate the pro-reproductive feature of GluGlu, systemic and local injections of the dipeptide have been performed. γ-GluGlu increased sperm count but had no effect on testosterone. These findings highlight the role of γ-GluGlu in mediating spermatogenetic effects of L. plantarum on the male mouse host and -following relevant human clinical trials- may provide future tools for treating certain forms of male infertility.

Fig. 1

L. plantarum SNI3 affects body composition, increases testis weight, sperm count and testosterone level.  Top left: Schematics of the experimental groups and design. (A) and (B). Body weight (BW) and body composition of mice in each group. Mean ± Standard Error of Mean (SEM). (C) Testis weight and epididymis weight of mice drinking autoclaved tap water (control) or L. plantarum SNI3 for 4 weeks. Mean ± SEM values. (D) Area of seminiferous tubuli and thickness of seminiferous epithelium in control and treated mice. Mean ± SEM values. Below: representative hematoxylin-eosin stained sections from the testes of control and L. plantarum SNI3-treated mice. (E) Sperm count in control and L. plantarum SNI3-treated mice as determined in Makler chamber; Percentage of motile sperms in control and treated groups assessed by CASA (computer-assisted sperm analysis) system. Fluorescent labeling of individual sperms obtained from control and L. plantarum SNI3-treated mice. Peanut lectin-FITC (green , acrosome) , phalloidin-Alexafluor595 (red , actin cytoskeleton) and DAPI (blue , DNA) labeling. (F) Expression of Gonadotropin-releasing hormone (GnRH) Follicle-stimulating hormone (FSH)- and Luteinizing hormone (LH)- β subunits in control and L. plantarum SNI3-treated mice. Mean ± SEM values. (G) Mean ± SEM of serum testosterone (TESTO) levels in control and L. plantarum SNI3-treated mice.  On all graphs: *p < 0.05; ** p < 0.01; ****p < 0.0001 using unpaired , two-tailed Student t-test.

Fig. 2

L. plantarum SNI3 administration impacts sexual behavior. Top: Schematics of the experimental groups and design. (A) Color-coded ethograms of control and L. plantarum SNI3-treated mice. Quantification of sexual behavioral activity of control and L. plantarum SNI3-treated male mice. Six distinct behaviors that belong to sexual interaction sequence were analyzed. Each row represents one mouse.  (B) and (C) Summary graphs showing frequency (B) and duration (C) of selected sexual behavior elements in control and SNI3-treated male mice. Mean ± SEM values. Unpaired t-test between control and SNI3- treated animals. *p < 0.05; **p < 0.01.

Fig. 3

L. plantarum SNI3 affects colon microbiome.  (A) Abundance of main phyla and top 20 taxa at genus level in the colon microbiome. Mean ± SEM values **** p  < 0.0001 vs. control group. (B) Correlation between the abundance of different microbial populations in the colon and sperm count. Scatter plots showing the relationship between microbial population abundance and sperm number. Each point represents one individual sample (n = 14). (C) Impact of fecal material transplantation (FMT) from L.plantarum SNI3 treated mice on total bacterial DNA concentration in colon content, sperm count and serum testosterone level. No FMT: mice received antibiotic treatment and no FMT (n = 5), Control FMT: mice received antibiotic treatment and fecal microbiome from control, untreated animals (n = 6), SNI3 FMT: mice received antibiotic treatment and fecal microbiome from L. plantarum SNI3 treated mice (n = 13). All microbiome data from two independent cohorts, # FMT data were analyzed by one-way ANOVA followed by Dunnett’s multiple comparison test (n = 6–13 per group). Mean ± SEM values *p < 0.05, **p < 0.01.

Fig. 4

Hierarchical clustering analysis, sPLSDA analysis and ROC analysis for discriminant metabolic comparison. (A) Hematoxylin-eosin staining of testis samples from control and L. plantarum SNI3-treated mice. Each section represents an individual mouse. (B) Hierarchical clustering analysis demonstrating different metabolic profiles of L. plantarumSNI3-treated and control groups on testis samples. Colored cells correspond to intensity values of metabolites. Euclidean distance and Ward method were applied for cluster analysis. (C) Partial least squares discriminant analysis (sPLSDA) for discriminant metabolic comparison ofL. plantarumSNI3- treated and control groups.  (D) An overview of ROC curves from different biomarker models using different numbers of features. ROC nalysis was performed to evaluate the discerning performances based on metabolic profiles. “ROC view”  a provides an overview comparing the ROC curves for all models created by MetaboAnalyst (Model 1–6  with 5, 10, 15, 25, 50 and 100 features respectively). Testis samples give the performance with AUC values  > 0.965.

Fig. 5

Metabolic correlation network of the testis in L. plantarum SNI3 treatment compared with control samples.  Nodes represent metabolites , with size reflecting the log2 fold change between control and L. plantarum SNI3 treatment in testis samples. Node sizes increase with the log2 fold change , while the criteria for labelling nodes are a log2 fold change cutoff value greater than 0 for upregulation (green color) and less than 0 for downregulation (red color) , respectively. Edges represent shared KEGG pathways. Selected pathways are colored as follows: arachidonic acid metabolism (yellow), linoleic acid metabolism (pink), glycerophospholipid metabolism (blue), fatty acid metabolism (green), protein digestion and amino acid metabolism (orange). Metabolites in green and red squares highlight the most upregulated (Glutamyl glutamate) and the most downregulated metabolite (Phosphatidyl choline , PC(22:6(4Z , 7Z , 10Z , 13Z , 16Z , 19Z)/16:0 , respectively.

Fig. 6

Identification of differential metabolites and pathways in control vs.L. plantarum SNI3-treated testes.  (A) The volcano plot is a combination of fold change (FC) and t-tests. The x-axis is log (FC). Y-axis is log10 (p value) based on FDR adjusted p values. The discriminative metabolites were selected by a volcano plot with fold change > 2 or < 0.5 and p-value < 0.05. (B) Pathway enrichment analyses were performed by KEGG pathway analysis. Discriminative pathways are summarized here. (C)  In situ mass spectrometry (MSI) images of select discriminative metabolites. On the left, MSI image of phosphatidylcholine , which is reduced upon L. plantarum SNI3 treatment; on the right, MSI image of glutamyl-glutamate, GluGlu, which is enhanced upon L. plantarum SNI3 treatment in the testis. Unpaired t-test was used for statistical analysis (*: p ≤ 0.05; ***: p ≤ 0.001).

Fig. 7

Systemic and intratesticular γ-GluGlu increases spermatogenesis.  C57Bl6 male mice received γ-GluGlu orally , intraperitoneally and locally (intratesticularly). Spermatozoa from the cauda epididymis were isolated and counted. Bar graphs showing the sperm number of mice injected with PBS or different doses of γ-GluGlu. Mean ± SEM values. One way ANOVA , *p < 0.05; **p < 0.01.