Glutaminase-1 inhibition alleviates senescence of Wharton's jelly-derived mesenchymal stem cells via senolysis

Abstract

Replicative senescence of mesenchymal stem cells (MSCs) caused by repeated cell culture undermines their potential as a cell therapy because of the reduction in their proliferation and therapeutic potential. Glutaminase-1 (GLS1) is reported to be involved in the survival of senescent cells, and inhibition of GLS1 alleviates age-related dysfunction via senescent cell removal. In the present study, we attempted to elucidate the association between MSC senescence and GLS1. We conducted in vitro and in vivo experiments to analyze the effect of GLS1 inhibition on senolysis and the therapeutic effects of MSCs. Inhibition of GLS1 in Wharton's jelly-derived MSCs (WJ-MSCs) reduced the expression of aging-related markers, such as p16, p21, and senescence-associated secretory phenotype genes, by senolysis. Replicative senescence-alleviated WJ-MSCs, which recovered after short-term treatment with bis-2-(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethyl sulfide 3 (BPTES), showed increased proliferation and therapeutic effects compared to those observed with senescent WJ-MSCs. Moreover, compared to senescent WJ-MSCs, replicative senescence-alleviated WJ-MSCs inhibited apoptosis in serum-starved C2C12 cells, enhanced muscle formation, and hindered apoptosis and fibrosis in mdx mice. These results imply that GLS1 inhibition can ameliorate the therapeutic effects of senescent WJ-MSCs in patients with muscle diseases such as Duchenne muscular dystrophy. In conclusion, GLS1 is a key factor in modulating the senescence mechanism of MSCs, and regulation of GLS1 may enhance the therapeutic effects of senescent MSCs, thereby increasing the success rate of clinical trials involving MSCs.

Keywords: BPTES; GLS1; Wharton’s jelly; mesenchymal stem cells; replicative senescence; senolysis.

Graphical Abstract

Figure 1.

Reduced cell survival after glutaminase-1 (GLS1) inhibition in replicatively senescent Wharton’s jelly-derived mesenchymal stem cells (WJ-MSCs). (A) Senescence-uninduced MSCs (U-MSCs) and replicatively senescent MSCs (rS-MSCs) were stained with senescence-associated (SA)-β-galactosidase. Scale bar = 200 µm. (B) Cell numbers were counted 0, 24, 48, and 72 hours after cell seeding. (C) Doubling time was calculated by counting the numbers of U-MSCs and rS-MSCs. (D) Immunocytochemical analysis of laminB1 in U-MSCs and rS-MSCs. Scale bar = 50 µm. The chart was used to quantify the intensity of lamin B1 divided by the intensity of 4ʹ,6-diamidino-2-phenylindole (DAPI) staining. (E) qRT-PCR and (F) immunoblotting of U-MSCs and rS-MSCs for the confirmation of GLS1 expression. (G) Suppression of GLS1 was examined by qRT-PCR after small interfering RNA (siRNA) transfection of U-MSCs and rS-MSCs. (H) Cell viability was examined 72 hours after siGLS1 transfection of U-MSCs and rS-MSCs. (I) Cell viability was measured after treating U-MSCs and rS-MSCs with dimethyl sulfoxide (DMSO) or bis-2-(5-phenylacetamido-1,2,4-thiadiazol-2-yl)ethyl sulfide 3 (BPTES; 30 µM) for 72 hours. Unpaired 2-tailed Student’s t test (A-F, I), one-way analysis of variance (ANOVA) with Tukey’s multiple comparisons post hoc test (G), and one-way ANOVA with Dunnett’s multiple comparisons post hoc test (H) were performed. ***P < .001, **P < .01, and *P < .05.

Figure 2.

BPTES inhibits senescence of Wharton’s jelly-derived mesenchymal stem cells (WJ-MSCs). (A) Illustration of the experimental procedure. (B) Relative glutamine levels in the culture medium, (C) relative intracellular glutamine levels, (D) relative intracellular glutamate levels, and (E) relative intracellular α-ketoglutarate levels in which rS-MSCs were treated with DMSO or BPTES (30 µM) for 24 hours. (F) Relative cell numbers of the rS-MSCs treated with DMSO or BPTES (30 µM). (G) Levels of p21, p16, and beta-galactosidase-1 (GLB1) proteins in U-MSCs, rS-MSCs, and BPTES-treated rS-MSCs. (H) mRNA expression of p21 and p16 and (I) senescence-associated secretory phenotype (SASP) genes (insulin-like growth factor binding protein 3 [IGFBP3], IGFBP5, and IGFBP7) in U-MSCs, rS-MSCs, and BPTES-treated rS-MSCs. Two-tailed Student’s t test (B-F) and one-way ANOVA with Tukey’s multiple comparisons post hoc test (G-I) were performed. ***P < .001, **P < .01, and *P < .05.

Figure 3.

Replicative senescence-alleviated MSCs (rSA-MSCs) were more effective at inhibiting apoptosis in myoblasts than rS-MSC. (A) The proportions of β-gal-positive cells in U-MSCs, rS-MSCs, and rSA-MSCs. Scale bar = 200 µm. (B) Quantification of cell area, (C) doubling time, and (D) mRNA expression levels of aurora kinase A (AURKA) in U-MSCs, rS-MSCs, and rSA-MSCs. (E) Immunoblotting using anti-cleaved PARP and anti-cleaved Caspase 3 antibodies and (F) immunocytochemical analysis of annexin V were conducted after coculture of serum-starved C2C12 cells with U-MSCs, rS-MSCs, and rSA-MSCs. Scale bar = 100 µm. The chart was used to quantify the proportion of annexin V-positive cells among the total number of cells. One-way ANOVA with Dunnett’s multiple comparisons post hoc test (A-E) and one-way ANOVA with Tukey’s multiple comparisons post hoc test (C, F) were performed. ***P < .001, **P < .01, and *P < .05. Abbreviation: SF, serum-free.

Figure 4.

Compared with rS-MSCs, rSA-MSCs were more effective in the recovery of skeletal muscles in mdx mice. (A) Grip strengths of forelimbs and hindlimbs, and (B) creatine kinase (CK) activity were analyzed by comparing preinjection and postinjection in normal, mdx, U-MSC-injected mdx, rS-MSC-injected mdx, and rSA-MSC-injected mdx mice. (C) Immunoblotting was conducted in normal, mdx, U-MSC-injected mdx, rS-MSC-injected mdx, and rSA-MSC-injected mdx mice using the indicated antibodies. Relative protein levels of (D) myosin heavy-chain (MHC), (E) annexin V, and (F) fibronectin indicate the converted value based on that of the mdx group. (G) Immunohistochemical analysis shows immunostaining of MHC and annexin V and Sirius Red staining. Scale bar = 100 µm. The charts were used to quantify the intensities of (H) MHC and (I) annexin V divided by the intensity of DAPI staining. (J) The chart compares the results of the fibrotic area. One-way ANOVA with Tukey’s multiple comparisons post hoc test (A, D) and one-way ANOVA with Dunnett’s multiple comparisons post hoc test (B, E-J) were performed. ***P < .001, **P < .01, and *P < .05.

Figure 5.

In rS-MSCs, mRNA expression of cathepsin C (CTSC), E2F transcription factor 7 (E2F7), and sorbin and SH3 domain containing 2 (SORBS2) was restored by GLS1 inhibition. (A) Venn diagram displaying the differentially expressed gene (DEG) distribution and (B) heatmap showing DEGs in U-MSCs, rS-MSCs, and BPTES-treated rS-MSCs. (C) Gene categories corresponding to DEGs in rS-MSCs and BPTES-treated rS-MSCs. (D) Levels of CTSC, E2F7, and SORBS2 mRNA in U-MSCs and rS-MSCs. (E) Normalized log2 values of CTSC, E2F7, and SORBS2 in U-MSCs, rS-MSCs, and BPTES-treated rS-MSCs obtained from the mRNA sequencing analysis. (F) Levels of CTSC, E2F7, and SORBS2 mRNA in U-MSCs, rS-MSCs, and BPTES-treated rS-MSCs. (G) Levels of CTSC, E2F7, and SORBS2 mRNA in rS-MSCs after siGLS1 transfection. Unpaired 2-tailed Student’s t test (D, G), one-way ANOVA with Dunnett’s multiple comparisons post hoc test (F; CTSC and E2F7), and one-way ANOVA with Tukey’s multiple comparisons post hoc test (F; SORBS2) were performed. ***P < .001, **P < .01, and *P < .05.

Figure 6.

GLS1 participates in hindering senescence by activating Wnt signaling. (A) Heatmap showing the DEGs in U-MSCs and rS-MSCs groups. The functional classification of 545 DEGs based on (B) biological process (BP) of gene ontology (GO) term analysis and (C) Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. The bars represent the number of DEGs in U-MSCs compared to that in rS-MSCs. (D) Enrichment plots based on gene set enrichment analysis (GSEA) showed that replicative senescence in WJ-MSCs is associated with cell cycle DNA replication and cell-cell signaling by Wnt. (E) Immunoblotting of U-MSCs, rS-MSCs, and rSA-MSCs, and (F) immunoblotting of rS-MSCs after siGLS1 transfection using the indicated antibodies. (G) An increase of GLS1 due to senescence inactivates the Wnt signaling pathway in rS-MSCs, but inhibition of GLS1 by BPTES inhibits replicative senescence by activating the Wnt signaling pathway in rSA-MSCs. One-way ANOVA with Dunnett’s multiple comparisons post hoc test (E) and unpaired 2-tailed Student’s t test (F) were performed. ***P < .001, **P < .01, and *P < .05. Abbreviations: NES, normalized enrichment score; FDR, false discovery rate.