Mitigating chemotherapy-induced granulosa cell damage: role of hUCMSC-EVs in regulating the lncRNA HCP5-miR-20a-5p-YAP1 network.

The positive outcomes of hUCMSC-EVs on promoting angiogenesis within ovarian tissue and reinstating ovarian function in POF-afflicted mice have been evidenced (Yang et al. 2019 ). To investigate whether hUCMSC-EVs have a repairing effect on CP-induced GC damage, we isolated and purified hUCMSC-EVs and characterized hUCMSCs. Optical observation revealed that hUCMSCs displayed elongated shuttle-like or spindle-like shapes while growing in colonies. The cells exhibited a swirling arrangement when closely packed, as demonstrated in Figure S1 A. Simultaneously, differentiation experiments were performed on hUCMSCs for adipogenic, osteogenic, and chondrogenic lineages, which exhibited robust differentiation capabilities (Figure S1 B). Flow cytometry revealed positive expression of CD44, CD73, CD105 and negative expression of CD45, CD31 and CD34 on the surface of hUCMSCs (Figure S1 C). The above results indicated that we obtained hUCMSCs with high purity. EVs were additionally separated from hUCMSCs. Evaluation under TEM demonstrated that EVs isolated from the supernatant of hUCMSCs had a cup-like morphology (Fig.  1 A). NTA indicated that the characteristics of EVs ranged predominantly from 50 to 150 nm (Figure S1 D). Based on WB assay results (Figure S1 E), EVs specific markers CD81, Alix and Hsp70 were detected in hUCMSCs-derived EVs, with the absence of endoplasmic reticulum marker protein Calnexin, indicating successfully isolated hUCMSCs-EVs. Fig. 1 Effects of hUCMSC-EVs on CP-induced GC damage. A , TEM observation of EVs morphology. B , Immunofluorescence microscopy observation of the entry of hUCMSC-EVs labeled with fluorescent PKH67 into GCs. PKH67-labeled EVs are in green, and DAPI-stained nuclei are in blue. C , Left panel: Assessment of GCs proliferation viability using CCK8, right panel: Doubling time, determined through cell counting. D , RT-qPCR to detect expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs. E, Immunofluorescence to detect expression of γH2 AX in CP-induced GCs. F, WB to detect expression of γH2 AX and cle-PARP protein expression in CP-induced GCs. * p  < 0.05. All cell experiments were repeated three times Effects of hUCMSC-EVs on CP-induced GC damage. A , TEM observation of EVs morphology. B , Immunofluorescence microscopy observation of the entry of hUCMSC-EVs labeled with fluorescent PKH67 into GCs. PKH67-labeled EVs are in green, and DAPI-stained nuclei are in blue. C , Left panel: Assessment of GCs proliferation viability using CCK8, right panel: Doubling time, determined through cell counting. D , RT-qPCR to detect expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs. E, Immunofluorescence to detect expression of γH2 AX in CP-induced GCs. F, WB to detect expression of γH2 AX and cle-PARP protein expression in CP-induced GCs. * p  < 0.05. All cell experiments were repeated three times Investigating the impact of hUCMSC-EVs on GCs required the co-culturing of PKH67-labeled hUCMSC-EVs with GCs, and the cells were stained and photographed after 12 h. Observations revealed the visibility of green fluorescence-tagged hUCMSC-EVs within GCs, contrasted with the lack of green fluorescence in GCs subjected solely to PBS treatment (Fig.  1 B). Further, CP was used to induce an in vitro GC damage model, which significantly reduced the proliferation viability of GCs, accompanied by a significantly prolonged doubling time (Fig.  1 C). Quantification by RT-qPCR manifested a pronounced rise in the pro-apoptotic gene Bax and a marked decrease in the anti-apoptotic gene Bcl-2 expression in GCs after CP exposure (Fig.  1 D). The γH2 AX can be used for DNA damage signaling and DNA repair (Taneja et al. 2004 ; Banáth et al. 2004 ). Immunofluorescence and WB results illustrated that γH2 AX fluorescence intensity and γH2 AX and cle-PARP protein expression were significantly increased in CP-induced GCs (Fig.  1 E-F). Further treatment of the CP-induced GCs with EVs increased the viability of GCs, accompanied by decreased Bax expression, γH2 AX fluorescence intensity, γH2 AX and cle-PARP protein expression but increased Bcl-2 expression (Fig.  1 C-F). The aforementioned results demonstrated that hUCMSC-EVs could reverse CP-induced GC damage. To investigate the molecular mechanism of hUCMSC-EVs in mitigating GC damage, differential comparison was conducted utilizing 4 transcriptome sequencing datasets between control and CP-induced GCs to identify variances in mRNA and lncRNA expression levels. A total of 9 differentially expressed lncRNAs were obtained, with 7 lncRNAs upregulated and 2 lncRNAs downregulated (Fig.  2 A). We imported these 9 lncRNAs into the exoRBase 2.0 database and found that 7 lncRNAs were enriched in EVs from multiple samples, with HCP5 being the most enriched one (Fig.  2 B). Further, HCP5 expression was substantially elevated in hUCMSC- EVs relative to that in the hUCMSCs (Fig.  2 C). Therefore, we speculated that hUCMSC-EVs might affect DNA damage repair in GCs through delivery of HCP5. Fig. 2 RNA transcriptome sequencing and bioinformatics analysis to screen for differentially expressed lncRNAs in GCs after CP induction. A , Heatmap of differentially expressed lncRNAs in control GCs and CP-induced GCs (n = 4). B , exoRBase 2.0 database to detect the expression of lncRNAs in EVs. C , RT-qPCR to detect the expression of HCP5 in EVs. D , RT-qPCR to detect the expression of HCP5 in EVs treated with RNase and detergent. E , Immunofluorescence microscopy to observe the expression of labeled fluorescent Cy3-HCP5 in GCs. Cy3-HCP5-labeled EVs are in red, DAPI-stained nuclei in blue and Phalloidin-stained cytoskeleton in green. F , RT-qPCR to detect HCP5 expression in GCs after co-culture with EVs. * p  < 0.05. All cell experiments were repeated three times RNA transcriptome sequencing and bioinformatics analysis to screen for differentially expressed lncRNAs in GCs after CP induction. A , Heatmap of differentially expressed lncRNAs in control GCs and CP-induced GCs (n = 4). B , exoRBase 2.0 database to detect the expression of lncRNAs in EVs. C , RT-qPCR to detect the expression of HCP5 in EVs. D , RT-qPCR to detect the expression of HCP5 in EVs treated with RNase and detergent. E , Immunofluorescence microscopy to observe the expression of labeled fluorescent Cy3-HCP5 in GCs. Cy3-HCP5-labeled EVs are in red, DAPI-stained nuclei in blue and Phalloidin-stained cytoskeleton in green. F , RT-qPCR to detect HCP5 expression in GCs after co-culture with EVs. * p  < 0.05. All cell experiments were repeated three times We started by verifying the presence of HCP5 within EVs or its display on the EVs'surface utilizing RNase protective assays. After RNase intervention, the HCP5 concentration in the solution remained unchanged, yet concurrent application of RNase A and Triton X100 led to a notable decline in HCP5 levels, suggesting that HCP5 was enclosed within the membrane rather than being released directly (Fig.  2 D). To deepen the understanding of HCP5 transfer, we transfected Cy3-HCP5 (red) into hUCMSCs, followed by extraction of EVs. Following incubation with GCs, red fluorescence was detected in the recipient GCs, suggesting that the cells took up EVs containing Cy3-labeled HCP5 (Fig.  2 E). Meanwhile, we examined the expression of HCP5 in GCs after co-culture using RT-qPCR and found that the expression of HCP5 in GCs was markedly lower after CP induction, whereas it was significantly elevated after co-culture of EVs (Fig.  2 F). Collectively, HCP5 could be delivered by hUCMSC-EVs into GCs. To investigate whether hUCMSC-EVs affect the damage repair of GCs by delivering HCP5, we further knocked down HCP5 in hUCMSCs and selected sh-HCP5, which had the best interference effect, for subsequent experiments. Proven by the RT-qPCR analysis indicated in Fig.  3 A, a significant decrease in HCP5 expression occurred in hUCMSCs and hUCMSC-EVs after treatment with sh-HCP5. As illustrated in Fig.  3 B, the expression of HCP5 experienced a marked growth in the CP-induced GCs co-cultured with hUCMSC-EVs, which could be decreased after the application of sh-HCP5 treatment. Fig. 3 hUCMSC-EVs deliver HCP5 to affect CP-induced GC damage. A , RT-qPCR to detect the expression of HCP5 in hUCMSCs and hUCMSC-EVs. B , RT-qPCR to detect the expression of HCP5 in CP-induced GCs after co-culture with hUCMSC-EVs. C , CCK-8 to detect the viability of CP-induced GCs after co-culture with hUCMSC-EVs. D , RT-qPCR to detect the expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs after co-culture with hUCMSC-EVs. E , Immunofluorescence to detect the expression of γH2 AX in CP-induced GCs after co-culture with hUCMSC-EVs. F , WB to detect the protein expression of γH2 AX and cle-PARP. in CP-induced GCs after co-culture with hUCMSC-EVs. * p  < 0.05. All cell experiments were repeated three times hUCMSC-EVs deliver HCP5 to affect CP-induced GC damage. A , RT-qPCR to detect the expression of HCP5 in hUCMSCs and hUCMSC-EVs. B , RT-qPCR to detect the expression of HCP5 in CP-induced GCs after co-culture with hUCMSC-EVs. C , CCK-8 to detect the viability of CP-induced GCs after co-culture with hUCMSC-EVs. D , RT-qPCR to detect the expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs after co-culture with hUCMSC-EVs. E , Immunofluorescence to detect the expression of γH2 AX in CP-induced GCs after co-culture with hUCMSC-EVs. F , WB to detect the protein expression of γH2 AX and cle-PARP. in CP-induced GCs after co-culture with hUCMSC-EVs. * p  < 0.05. All cell experiments were repeated three times Further functional experiments illustrated that, in relation to the CP + PBS group, the CP + EVs-shNC group displayed significantly increased GC proliferation, decreased doubling time, and reduced apoptosis and DNA damage. On the contrary, as opposed to the CP + EVs-shNC group, the CP + EVs-shHCP5 group showed significantly reduced GC proliferation, increased doubling time, and increased apoptosis and DNA damage (Fig.  3 C-F). Taken together, hUCMSC-EVs could deliver HCP5, attenuating CP-induced GC damage. To explore the downstream targets of HCP5 affecting damage repair in GCs, we first searched the top 1000 genes associated with"DNA damage"through GeneCards and intersected them with the 549 DEGs obtained from sequencing, which yielded 12 intersecting genes (Figure S2 A). We then uploaded the proteins encoded by these 12 genes into the STRING database and sorted them according to the number of nodes. The top 10 genes are shown in Figure S2 B, of which only YAP1 was significantly down-regulated after CP induction (Figure S2 C). Hence, our theory suggests that YAP1 might be a fundamental gene influencing the repair mechanism in GCs. We further investigated whether HCP5 could regulate YAP1 expression. RT-qPCR and WB assays illustrated that the expression of HCP5 was markedly enhanced in the presence of oe-HCP5. YAP expression was markedly suppressed with CP induction and HCP5 overexpression reversed the CP-induced down-regulation of YAP1 in GCs (Fig.  4 A, B). Fig. 4 HCP5 regulates YAP1 expression to affect CP-induced GC damage. A , RT-qPCR to detect the mRNA level of YAP1 in CP-induced GCs after overexpression or knockdown of HCP5. B , WB to detect the protein expression of YAP1 in CP-induced GCs after overexpression or knockdown of HCP5. C , RT-qPCR to detect the mRNA level of YAP1 in CP-induced GCs treated with hUCMSC-EVs. D , WB to detect protein expression of YAP1 in CP-induced GCs treated with hUCMSC-EVs. E , Left panel: Detection of the proliferation ability of co-cultured GCs using CCK-8 assay, right panel: Determination of doubling time by cell counting. F , RT-qPCR to detect expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs treated with hUCMSC-EVs. G , Immunofluorescence to detect expression of γH2 AX in CP-induced GCs treated with hUCMSC-EVs. H , WB to detect protein expression of γH2 AX and cle-PARP in CP-induced GCs treated with hUCMSC-EVs. * p  < 0.05. All cell experiments were repeated three times HCP5 regulates YAP1 expression to affect CP-induced GC damage. A , RT-qPCR to detect the mRNA level of YAP1 in CP-induced GCs after overexpression or knockdown of HCP5. B , WB to detect the protein expression of YAP1 in CP-induced GCs after overexpression or knockdown of HCP5. C , RT-qPCR to detect the mRNA level of YAP1 in CP-induced GCs treated with hUCMSC-EVs. D , WB to detect protein expression of YAP1 in CP-induced GCs treated with hUCMSC-EVs. E , Left panel: Detection of the proliferation ability of co-cultured GCs using CCK-8 assay, right panel: Determination of doubling time by cell counting. F , RT-qPCR to detect expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs treated with hUCMSC-EVs. G , Immunofluorescence to detect expression of γH2 AX in CP-induced GCs treated with hUCMSC-EVs. H , WB to detect protein expression of γH2 AX and cle-PARP in CP-induced GCs treated with hUCMSC-EVs. * p  < 0.05. All cell experiments were repeated three times In addition, to explore whether HCP5 delivered by hUCMSC-EVs could regulate YAP1, we treated GCs with hUCMSC-EVs after knocking down YAP1. RT-qPCR and WB assays displayed that hUCMSC-EVs promoted the mRNA and protein expression of YAP1 in CP-induced GCs, while the knockdown of YAP1 inhibited the up-regulation of YAP1 induced by hUCMSC-EVs (Fig.  4 C, D). Opted for further experiments was the sh-YAP1 variant with the most significant reverse function. In the investigation of GCs proliferation and apoptosis, it was found that in contrast with the CP + PBS + sh-NC group, the CP + EVs + sh-NC group displayed a marked growth in cell proliferation, a notable drop in doubling time, and a significant reduction in cellular apoptosis (Fig.  4 E, F). DNA damage repair assay showed that γH2 AX fluorescence intensity, as well as γ H2 AX and cle-PARP protein expression in the CP-induced GCs were significantly decreased by hUCMSC-EVs (Fig.  4 G, H). In contrast, interference with YAP1 reversed the ameliorative effect of hUCMSC-EVs on CP-induced GC damage (Fig.  4 E-G). The aforementioned results demonstrated that hUCMSC-EVs up-regulated YAP1 expression through delivery of HCP5, thereby attenuating CP-induced GCs damage. More studies have been conducted to reveal the function of non-coding RNAs by studying the regulatory correlation between miRNAs and lncRNAs. It has been shown that HCP5 could play a ceRNA mechanism to mediate the process of miRNA-influenced diseases (Yang et al. 2022 ; Li et al. 2021 ; Zhang et al. 2022 ). To investigate whether HCP5 could affect YAP1 expression via miRNAs, we used the Starbase database to search for HCP5-binding miRNAs, as well as YAP1 upstream regulatory miRNAs, and the mirPath v.3 database KEGG Reverse Search for DNA damage response, detection of DNA damage|GO: 0042769 related miRNAs; the intersection between them was take to obtain a total of 4 intersection miRNAs (Fig.  5 A). Next, we examined the expression of the miRNAs in CP-induced GCs, and the results showed that miR-20a-5p was significantly up-regulated in the CP-induced GCs than in the control GCs, and that miR-20a-5p showed the most markedly up-regulated expression in the CP-induced GCs (Fig.  5 B). Hence, our speculation is that HCP5 could potentially enhance the YAP1 expression via miR-20a-5p. Fig. 5 HCP5 regulates the expression of YAP1 through a miRNA-dependent mechanism. A , Venn plots of the intersection of miRNAs targeting HCP5 or YAP1 predicted through multiple databases. B , RT-qPCR to detect miRNA expression in GCs after CP induction. C , Binding sites and mutation sites between miR-20a-5p and HCP5 or YAP1. D , Dual luciferase reporter gene assay to measure HCP5/YAP1-WT/MUT luciferase activity. E , RIP assay to detect enrichment of HCP5, miR-20a-5p and YAP1. F , Enrichment of HCP5 and YAP1 in GCs by biotin-miR-20a-5p detected by RNA pull-down assay. G , RT-qPCR to detect expression of miR-20a-5p and YAP1 in GCs after transfection with miR-20a-5p mimic. * p  < 0.05. All cell experiments were repeated three times HCP5 regulates the expression of YAP1 through a miRNA-dependent mechanism. A , Venn plots of the intersection of miRNAs targeting HCP5 or YAP1 predicted through multiple databases. B , RT-qPCR to detect miRNA expression in GCs after CP induction. C , Binding sites and mutation sites between miR-20a-5p and HCP5 or YAP1. D , Dual luciferase reporter gene assay to measure HCP5/YAP1-WT/MUT luciferase activity. E , RIP assay to detect enrichment of HCP5, miR-20a-5p and YAP1. F , Enrichment of HCP5 and YAP1 in GCs by biotin-miR-20a-5p detected by RNA pull-down assay. G , RT-qPCR to detect expression of miR-20a-5p and YAP1 in GCs after transfection with miR-20a-5p mimic. * p  < 0.05. All cell experiments were repeated three times We further predicted the binding sites through the StarBase website and the results demonstrated intersecting binding domains between HCP5 and miR-20a-5p, and also between miR-20a-5p and YAP1 (Fig.  5 C). We mutated the binding sites between them and the dual luciferase reporter gene assay indicated that the miR-20a-5p mimic significantly inhibited the luciferase activity of HCP5-WT as well as YAP1-WT, without impacting the functionality of HCP5-MUT and YAP1-MUT (Fig.  5 D). Interaction between AGO2 and miR-20a-5p, HCP5, or YAP1 was demonstrated through RIP assays (Fig.  5 E). More enriched YAP1 and HCP5 were detected with Biotin-miR-20a-5p in the RNA pull-down assay results as opposed to Biotin-cel-miR67 (Fig.  5 F), indicating the precise recognition of HCP5 and YAP1 by miR-20a-5p. Further in vitro cell validation displayed that miR-20a-5p expression was notably amplified in the CP-induced GCs than in the control GCs. Co-culture of EVs decreased the miR-20a-5p expression in the CP-induced GCs, while knockdown of HCP5 could reverse the effect. The expression of miR-20a-5p was considerably escalated and that of YAP lessened in the presence of miR-20a-5p mimic (Fig.  5 G). Collectively, HCP5 could sponge miR-20a-5p to up-regulate YAP1 expression. To investigate the impact of HCP5 on CP-induced GC damage via the miR-20a-5p/YAP1 axis, we evaluated the effects of overexpressing HCP5 or miR-20a-5p on GC functions. Analysis of RT-qPCR data indicated that in the CP + oe-HCP5 + mimic-NC group, miR-20a-5p expression decreased significantly while YAP1 expression increased significantly contrasted against the CP + oe-NC + mimic-NC group. Conversely, the miR-20a-5p expression significantly rose, while the YAP1 expression notably lessened in the CP + oe-HCP5 + miR-20a-5p mimic group in contrast to the CP + oe-HCP5 + mimic-NC group. HCP5 expression was notably higher in the CP + oe-HCP5 + mimic-NC group in relation to the CP + oe-NC + mimic-NC group, with no significant difference observed in HCP5 expression between the CP + oe-HCP5 + miR-20a-5p mimic and CP + oe-HCP5 + mimic-NC groups (Fig.  6 A). Biological function assays revealed that, as opposed to the CP + oe-NC + mimic-NC group, the CP + oe-HCP5 + mimic-NC group displayed significantly increased proliferation, decreased doubling time, and reduced apoptosis (Fig.  6 B, C). DNA damage repair assays demonstrated that overexpression of HCP5 lowered γH2 AX fluorescence intensity, as well as γH2 AX and cle-PARP protein expression (Fig.  6 D, E), which could be counteracted by overexpression of miR-20a-5p. Fig. 6 HCP5 regulates miR-20a-5p/YAP1 axis to affect CP-induced GC damage. A , RT-qPCR to detect the expression of miR-20a-5p and YAP1 in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. B , Left panel: CCK-8 detection of the proliferative capacity of transfected GCs, right panel: Doubling time, determined by cell counting. C , RT-qPCR to detect the expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. D , Immunofluorescence to detect the expression of γH2 AX in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. E , WB to detect the protein expression of γH2 AX and cle-PARP in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. * p  < 0.05. All cell experiments were repeated three times HCP5 regulates miR-20a-5p/YAP1 axis to affect CP-induced GC damage. A , RT-qPCR to detect the expression of miR-20a-5p and YAP1 in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. B , Left panel: CCK-8 detection of the proliferative capacity of transfected GCs, right panel: Doubling time, determined by cell counting. C , RT-qPCR to detect the expression of apoptosis-related genes Bax and Bcl-2 in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. D , Immunofluorescence to detect the expression of γH2 AX in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. E , WB to detect the protein expression of γH2 AX and cle-PARP in CP-induced GCs after transfection with oe-HCP5 alone or combined with miR-20a-5p mimic. * p  < 0.05. All cell experiments were repeated three times Collectively, HCP5 sponged miR-20a-5p to up-regulate YAP1 expression, which attenuated CP-induced GC damage. To delve deeper into the impact of hUCMSCs-EVs delivering HCP5 on CP-induced ovarian damage, we used CP-induced rats to establish a POF model after tail vein injection of hUCMSC-EVs, as shown in Fig.  7 A. We used PKH26 to label hUCMSC-EVs and found that PKH26-labeled hUCMSC-EVs could be distributed in ovarian tissues (Fig.  7 B). Moreover, compared with those in the normal control rats, miR-20a-5p expression was significantly elevated and YAP1 expression was decreased in the CP-induced POF rats. hUCMSC-EVs resulted in lowered miR-20a-5p expression and increased YAP1 expression in the CP-induced POF rats, while HCP5 knockdown could reverse the effect (Fig.  7 C). Fig. 7 hUCMSC-EVs deliver HCP5 to affect CP-induced ovarian damage in in vivo animal experiments. A , Flowchart of POF rat model construction. B , Distribution of PKH26-labeled hUCMSC-EVs in ovarian tissues after in vivo injection. C , RT-qPCR to detect expression of YAP1 and miR-20a-5p in ovarian tissues of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. D , HE staining to observe the pathological features of ovarian tissues and follicle statistics. E , ELISA to detect E2 and FSH in serum of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. F , TUNEL to detect apoptosis index in ovarian tissue of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. G , Immunofluorescence to detect fluorescence quantification of γH2 AX in CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. ns p  > 0.05. * p  < 0.05. n = 10 hUCMSC-EVs deliver HCP5 to affect CP-induced ovarian damage in in vivo animal experiments. A , Flowchart of POF rat model construction. B , Distribution of PKH26-labeled hUCMSC-EVs in ovarian tissues after in vivo injection. C , RT-qPCR to detect expression of YAP1 and miR-20a-5p in ovarian tissues of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. D , HE staining to observe the pathological features of ovarian tissues and follicle statistics. E , ELISA to detect E2 and FSH in serum of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. F , TUNEL to detect apoptosis index in ovarian tissue of CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. G , Immunofluorescence to detect fluorescence quantification of γH2 AX in CP-induced POF rats in response to hUCMSC-EVs alone or combined with HCP knockdown. ns p  > 0.05. * p  < 0.05. n = 10 The effect of hUCMSCs-EVs delivery of HCP5 on CP-induced ovarian damage was further explored. Morphological analysis showed different stages of follicles in the normal control rats; while in the CP-induced POF rats, primordial follicles were reduced and atretic follicles were significantly increased. Relative to the CP-induced POF rats, the CP-induced POF rats treated with hUCMSCs-EVs had increased primordial follicles and reduced atretic follicles, which could be reversed by HCP5 knockdown (Fig.  7 D). ELISA assay outcomes displayed that relative to those in the normal control rats, the serum level of FSH increased and that of E2 decreased in the CP-induced POF rats. Treatment with hUCMSCs-EVs decreased the serum level of FSH and elevated that of E2 in the CP-induced POF rats, while HCP5 knockdown could negate the effect (Fig.  7 E). TUNEL and immunofluorescence results displayed that compared to the normal control rats, the CP-induced POF rats had increased apoptotic index and γH2 AX fluorescence quantification. Treatment with hUCMSCs-EVs resulted in marked increase in the apoptotic index and γH2 AX fluorescence quantification in the CP-induced POF rats, the effects of which could be negated by HCP5 knockdown (Fig.  7 F, G). Taken together, hUCMSC-EVs delivering HCP5 up-regulated YAP1 expression, which could attenuate CP-induced ovarian damage in vivo.

Execution of the study was sanctioned by the Ethics Committee of Guangzhou Red Cross Hospital. Adherence to the Guide for the Care and Use of Laboratory Animals was ensured throughout the animal investigation process. Procured from Procell Life Science & Technology Co., Ltd. in Wuhan, China, the hUCMSCs (CP-CL11) and hGCs (CP-H192) were cultured in DMEM/F12 medium (11320033, Gibco, Thermo Fisher Scientific, Rockford, IL) augmented with 10% FBS (12483020, Gibco, Thermo Fisher Scientific). Cultivation of cells occurred in an incubation chamber set at 37 °C, 5% CO 2 , and maintained humidity. Passage of cells was conducted every 72 h, with hUCMSCs from passages 3 to 7 being designated for subsequent examinations (Zhu et al. 2021 ). The creation of an in vitro POF model involved exposing hGCs to 100 µg/mL of cyclophosphamide (CP) (6055–19-2, Sigma-Aldrich, St. Louis, MO) for a duration of 24 h to induce cellular damage (Zhao et al. 2022 ). We sequenced and analyzed control GC cells and CP-induced GC cells respectively, and CapitalBio Technology in Beijing, China, conducted the generation and sequencing of the sequencing libraries. In short, TRIzol reagent (15,596,026, Invitrogen, Carlsbad, CA) facilitated the extraction of total RNA from the samples. Preparation of sequencing libraries involved the utilization of premium RNA samples, followed by the examination of differentially expressed genes (DEGs) and the identification of enriched differentially expressed lncRNAs. Gene counts for each sample were normalized to the base mean, and p values and fold change (FC) in each comparison were calculated. Next, the screening criteria for differentially expressed lncRNAs were established as |log 2 FC|> 1 and p   2 and p  < 0.05. A volcano map and a heatmap were plotted using R software"ggplot2"and"ggpubr"packages. The lncRNA expression in Exos was analyzed via the exoRBase 2.0 database. The acquisition of candidate targets'gene and protein–protein interaction (PPI) networks was facilitated by the STRING database. Subsequent to the acquisition of findings related to PPI, Cytoscape 3.6.0 application was utilized to visualize and identify pivotal gene. The online platform Starbase was employed to study the miRNAs regulated by lncRNA HCP5 and the upstream regulatory miRNAs of YAP1. Next, mirPath v.3 KEGG Reverse Search was used to retrieve DNA damage response, followed by detection of DNA damage|GO: 0042769 related miRNAs and delineation of the Venn diagram to obtain intersecting miRNAs. According to the instruction of hUCMSC induction differentiation (osteogenesis, lipogenesis and chondrogenesis) kits (PD-017/018/019, Procell Life Science & Technology Co., Ltd.), Alizarin Red S, Oil Red O, and Alcian Blue stains were applied to hUCMSCs to observe their capability in differentiating into osteogenic, lipogenic, and chondrogenic lineages (Li et al. 2018 ). Selected for surface marker analysis were the hUCMSCs in their third passage, showing an approximate 80% coverage. Following removal of medium, detachment and centrifugation, adjustments were made to reach a concentration of 1 × 10 6 cells/mL. The cells were moved to 15 mL centrifuge tubes encompassing 100 μL of PBS buffer with 2% FBS, followed by supplementation of specific fluorescent flow antibodies, including PE-labeled CD73, CD105, CD44, CD34 and CD31 and APC-labeled CD45 (FAB5795P/FAB10971P/FAB3660P/FAB7227P/FAB3567P/FAB1430 A, R&D Systems, Minneapolis, MN). After incubation in darkness for 30 min at 4℃, the cells were reinvigorated in 3 mL of PBS saline, spun down, and combined with 300 μL of PBS buffer. The reconstitution of cells in 3 mL of PBS buffer was followed by centrifugation, after which 300 μL of PBS buffer were added. The control cells were identified using isotype monoclonal antibodies for background labeling, and the BD FACSVerse flow cytometer was utilized for the analysis of fluorescent cells. FlowJo (FlowJo LLC, Ashland, OR) was employed for the computation of surface antigen positivity. Removal of EVs from the serum was achieved by subjecting FBS to centrifugation at 100,000 g for a duration of 24 h. Upon reaching 80% confluence of hUCMSCs, the supernatant from the culture was discarded, and the existing medium was substituted with fresh culture medium containing 10% FBS depleted of EVs. This was succeeded by a 48-h incubation period at 37 °C in a CO 2 -regulated incubation chamber. Sequential centrifugation was applied to the harvested culture supernatant; at 500 g for 15 min, at 2000 g for 15 min and at 10,000 g for 20 min. Post sieving using 0.22 μm filters, the cultural liquid underwent centrifugation at 110,000 g for 70 min, succeeding in suspension within PBS. This was further followed by ultracentrifugation and another suspension in 100 uL of sterile PBS for subsequent analyses. The ultracentrifugation steps were executed at a temperature of 4 °C utilizing a Beckman ultracentrifuge (Optima L-90 K, Beckman Coulter Inc., Chaska, MN) that was fitted with a SW-32 Ti rotor. Conducting the remaining low-speed centrifugation involved the Beckman Allegra X-15R benchtop centrifuge. Implementation of Nanosight nanoparticle size analysis allowed for the measurement of EVs'size distribution, utilizing the NanoSight nanoparticle tracking analysis (NTA) system supplied by Malvern Instruments in the UK (Yan et al. 2018 ). Observation was executed through the use of transmission electron microscopy (TEM). The ultra-centrifuged EVs, measuring 20 μL, were placed on a carbon-coated copper electron microscope grid and subjected to staining with phosphotungstic acid solution (12,501–23-4, Sigma-Aldrich) for 5 min for counterstaining. This was followed by observation using TEM (H7650, Hitachi, Tokyo, Japan) (Fang et al. 2018 ). The identification of EVs'surface markers was accomplished using Western blot (WB) technique. The assessment of protein expression in concentrated EV suspensions was completed by utilizing a BCA kit (BCA1, Sigma-Aldrich). EVs were loaded on SDS-PAGE gels and marker proteins specific to EVs, specifically Hsp70, CD81, Alix, and the non-EV marker Calnexin, were successfully identified (Ramanathan et al. 2019 ). Purified hUCMSC-EVs were tagged using a PKH67 fluorescent kit (PKH67GL, Sigma-Aldrich). The EVs were immersed in 1 mL of Diluent C solution and combined with 4 μL of PKH67 ethanol dye solution to create the 4 × 10 –6  M dye solution. Next, combining a volume of 1 mL EVs suspension with the dye solution for a period of 5 min, the staining procedure was ceased by incubating the solution in 2 mL of 1% FBS devoid of EVs for 1 min. The identified EVs were obtained through ultracentrifugation at a speed of 100,000 × gravity for a duration of 2 h, with EVs concentrated within the sucrose density gradient of 1.13–1.19 g/mL and subsequently acquired. GCs were treated with PKH67-tagged EVs for a duration of 12 h while being maintained at 37 °C. The fixation of cellular components was completed through the application of 4% paraformaldehyde, along with the utilization of DAPI (MBD0020, Sigma-Aldrich) for nucleus visualization. To explore the transfer of HCP5, hUCMSCs were transfected with cy3-labelled HCP5 (serum-free medium) (Genechem, Shanghai, China) applying the Lipo3000 kit (L3000015, Thermo Fisher Scientific) for 6 h. Subsequently, a 48-h incubation took place in a 10% EVs serum-exclusion medium. Cell supernatants were collected and resuspended in PBS. EVs were added to GCs, followed by fixation with 4% paraformaldehyde. Phalloidin-iFluor 488 Reagent (1:1000, ab176753, Abcam, Cambridge, UK, emitting green fluorescence) was applied to mark the framework of GCs for 30 min at room temperature (RT). Nuclei were stained with DAPI (MBD0020, Fluorescent microscopy (ECLIPSE E800, Nikon, Tokyo, Japan) was utilized to visualize the internalization of EVs and EVs-HCP5 by GCs after nuclei were stained with DAPI (MBD0020, Sigma-Aldrich). The utilization of RNase A was implemented to determine the presence of HCP5 on the exterior or within EVs. Suspension of EVs in PBS was followed by treatment with RNase A (11119915001, Sigma-Aldrich) at a concentration of 20 μg/μL for a period of 20 min at 37 °C. In addition, the vesicle's membrane structural integrity suffered damage after exposure to Triton X100 detergent (93443, Sigma-Aldrich), followed by supplementation of RIPA buffer for 20 min and the RNase A application as described above. Next, RNA isolation was facilitated by the addition of a buffer to halt the reaction. Finally, the relative expression of HCP5 was calculated. Shanghai-based Genechem designed and manufactured the lentiviral vector encoding short hairpin RNA (shRNA) specific to HCP5, along with the control vector (sh-NC), used in this research. The 293 T cell line (CRL-3216, ATCC, Manassas, VA) received the lentiviral vector through Lipofectamine 3000 reagent (L3000015, Invitrogen). A change occurred after twenty hours, with the medium being swapped for 12 mL of medium with 5% fetal bovine serum added. The virus-laden supernatant was assembled after an approximate duration of 48 h, strained with a 0.45 μm cellulose acetate filter (HAWG04700, Millipore Corp., Billerica, MA), and placed in storage at −80 °C. At an appropriate multiplicity of infection (MOI = 20), hUCMSCs were infected with lentiviral vectors and cells were screened for 3 d using 1 μg/mL of puromycin (A1113803, Thermo Fisher Scientific) to ascertain screening stable cells. GCs were subjected to transient transfection with overexpression (oe)-HCP5, sh-HCP5, miR-20a-5p mimic or the corresponding NCs purchased from Genechem. Cultivation of GCs took place in cell culture dishes situated in 6-well trays, with a cell density of 4 × 10 5 cells per milliliter. Upon achieving 80% cell confluence, transfection of the cells was conducted utilizing lipofectamine 2000. After transfection, the cells in the distinct subgroups underwent a 48-h incubation at 37 °C in the presence of 5% CO 2 before undergoing additional procedures. The information on the shRNA sequences is listed in Table S1 . Cultured hUCMSCs were treated with PBS as control or subjected to extraction of hUCMSCs-EVs or EVs isolated from hUCMSCs infected with sh-NC or shHCP5. After treatment of hGCs with 100 µg/mL CP for 24 h to initiate cell damage, EVs were co-cultured with GCs and incubated for 24 h with equal amounts of PBS and EVs (5 µg/mL), respectively. GCs were untreated as control; induced with CP; induced with CP and co-cultured with PBS, hUCMSC-EVs or hUCMSC-EVs isolated from lentiviral vector expressing sh-NC or sh-HCP5; transfected with oe-NC or oe-HCP5; transfected with oe-NC or oe-HCP5 and then induced with CP; transfected with sh-NC, co-cultured with PBS and induced with CP; transfected with sh-NC, co-cultured with hUCMSC-EVs and induced with CP; transfected with sh-YAP1, co-cultured with hUCMSC-EVs and induced with CP; transfected with oe-NC + mimic-NC and induced with CP; transfected with oe-HCP5 + mimic-NC and induced with CP; transfected with oe-HCP5 + miR-20a-5p mimic and induced with CP. Prediction of relationships was accomplished by accessing the Starbase database, and the validation of the interactions between HCP5, miR-20a-5p, and YAP1 was established through luciferase reporter gene assays (Song et al. 2021 ). HCP5/YAP1 dual luciferase reporter gene vectors and mutant vectors with mutations in the binding sites of HCP5/YAP1 and miR-20a-5p, namely, pGLO-HCP5/YAP1 3'-UTR WT and pGLO-HCP5/YAP1 3'-UTR MUT, were established (pGLO was purchased from Bio-Rad Laboratories, Hercules, CA). The sequencing results of HCP5-WT, HCP5-MUT, YAP-WT, and YAP-MUT can be found in Table S2 . The two correctly sequenced reporter plasmids were co-transfected into 293 T cells with miR-20a-5p mimic and mimic-NC (Genechem), respectively. Upon completion of 24 h post-transfection, cellular lysis took place, and subsequent centrifugation at 12,000 g for 60 s was conducted to isolate the supernatant. The Dual-Luciferase® Reporter Assay System (E1910, Promega from Madison, WI) was employed to determine the luciferase activity. Relative luciferase activity was assessed based on the ratio of firefly luciferase to Renilla luciferase. The detection process was executed as per the guidelines applying the Magna-RNA binding protein immunoprecipitation kit (17–700, Millipore Corp.). Cell lysis was induced by RIPA lysis buffer, followed by centrifugation at 12,000 g for a duration of 10 min at 4 °C to eliminate the supernatant. Input was derived from a section of the cell extract, while the remaining portion was designated for antibody incubation aimed at co-precipitation. In each co-precipitation reaction setting, 50 μL of magnetic beads were reactivated in 100 μL of RIP cleansing fluid and bound with 5 μg of antibody for attachment. The suspension of the bead-antibody amalgamation in 900 μL of RIP washing solution was followed by an overnight incubation at 4 °C with 100 μL of cell extract, leading to the aggregation of the bead-protein mixtures. Specimens were treated with Proteinase K to detach samples, allowing for the extraction of RNA used in the following RT-qPCR procedure. Antibodies applied in the RIP assay included AGO2 (1:50, ab186733, Abcam) and IgG (1:100, ab205718, Abcam, NC). The hGCs were transfected with 50 nM of RNA biotin-labeled miR-20a-5p (bio-miR-20a-5p) and 50 nM of bio-cel-miR67 (used as a NC). Upon completion of the 48-h transfection period, cell specimens were harvested and subsequently treated with a designated cellular solution for a span of 10 min. Next, dispensing 50 mL of the sample cell lysate, the excess lysate was subjected to incubation with streptavidin magnetic beads (LSKMAGT, Sigma-Aldrich) that were precoated with RNase-free and yeast tRNA (11119915001/AM7119, Thermo Fisher Scientific) for 3 h at a temperature of 4 °C. Extraction of RNA and subsequent RT-qPCR analysis were carried out to assess the expression of HCP5 or YAP1 in the bio-miR-20a-5p or bio-cel-miR-67 samples obtained through pull-down. Cellular survival of GCs was examined by applying CCK-8 kits (40203ES60, Yeasen Co., Shanghai, China). Harvested cells in the exponential growth phase were adjusted to a concentration of 5 × 10 4 cells/mL in the presence of complete culture medium. 96-well culture panels were utilized for the cell distribution, with the addition of 100 μL of cellular incubation medium into each compartment. The following step included placement in the incubator for durations of 0 h, 12 h, 24 h, and 48 h. Afterwards, each well received 10 μL of the CCK-8 solution. The absorbance readings at 450 nm were recorded by a Multiskan FC microplate reader (51119180ET, Thermo Fisher Scientific) after incubating for 2 h at 37 °C. Cell sections were obstructed in 10% BSA for 1 h at 25 °C, and immersed overnight in darkness with specific anti-γH2 AX (1: 100, # ab2893, Abcam) at 4 °C and then with fluorescent secondary antibody Alexa Fluor 488 goat anti-rabbit IgG (1: 2000, #4412S, Cell Signaling Technology, Beverly, MA) for 2 h at 37 °C, pursued by co-incubation with DAPI for 15 min at RT. Fluorescent microscope (ECLIPSE E800, Nikon) was used to obtain immunofluorescence images. The TRIzol reagent (15596026, Invitrogen) was employed to extract total RNA from cells and tissues, following the specified instructions provided by the manufacturer. For detection of miRNA, miRNA was reversely transcribed using miRNA-specific stem-loop primers, followed by miRNA quantification applying the MiRcute miRNA RT-qPCR kit (FP411, TIANGEN Biotechnology Co. Ltd, Beijing, China), with U6 as the internal reference. For detection of lncRNA and mRNA, the PrimeScript RT reagent Kit (RR047 A, Takara, Shiga, Japan) was employed to convert RNA into cDNA through reverse transcription. SYBR Premix Ex Taq II (Takara) was employed for the gene expression analysis. Gene expression was evaluated through the 2 −ΔΔCt method with GAPDH as the internal normalization control. Table S3 contains the primer list information. The breakdown of EVs, cells, or tissues was facilitated by employing the lysis buffer (89901, Thermo Fisher Scientific) and then extracted applying the total ExoRNA and protein isolation set (4478545, Thermo Fisher Scientific), followed by the assessment of overall protein concentrations through the BCA Protein Assay (BCA1, Sigma-Aldrich). After the lysates were placed on SDS-PAGE gels for transfer onto a PVDF membrane, they were treated with the specified primary antibodies from Table S4, followed by incubation with HRP-labeled secondary antibodies goat anti-rabbit IgG (1: 5000, ab205718, Abcam) and goat anti-mouse IgG (1: 5000, ab205719, Abcam). Ultimately, the enhancement chemiluminescent substrate kit (36222ES60, Yeasen) was employed to identify the protein bands, followed by quantifying the protein bands with Image J software. Female Sprague–Dawley (SD) rats, aged 7 weeks, were received from the laboratory animal center affiliated with Guangzhou Red Cross Hospital. The rats were bred in enclosures at a humidity level of 45–60% and a temperature range of 22–24 degrees Celsius, while being provided unrestricted availability to both hydration and nourishment. After one week of acclimatization, the rats were exposed to intraperitoneal infusion of CP (starting dose 50 mg/kg, then 8 mg/kg daily for 14 days) or the same amount of 0.9% saline as a comparison group. Rats were randomly used as normal control rats, or induced by PC to establish a POF model without other treatment or further treated with hUCMSC-EVs or hUCMSC-EVs with sh-HCP5 (n = 10). EVs-treated rats were treated with hUCMSC-EVs (125 μg diluted in 100 μL PBS) in the tail vein at 1, 5 and 10 d after CP injection, respectively. Rats were euthanized on day 30 after CP injection, and blood samples were obtained and ovaries were dissected. To detect the distribution of hUCMSC-EVs in ovarian tissues, hUCMSC-EVs were labeled with PKH-26 (PKH26GL, Sigma-Aldrich) per the guidance provided by the maker. The specified hUCMSC-EVs were injected into SD rats (n = 3) via the tail vein. Ovarian specimens obtained were cryogenically preserved, sectioned finely (at a width of 5 µm), and subsequently scrutinized and pictured using a fluorescence microscope (Nikon ECLIPSE E800). Quantification of follicle stimulating hormone (FSH) and estradiol (E2) in the rat serum was carried out using ELISA testing, and the effect of ovarian endocrine function was analyzed. Collection of blood was performed by accessing the abdominal aorta post administration of intraperitoneal injection with 1% pentobarbital for anesthesia. After freezing, the serum samples were kept at −80℃ for analysis in the future. Rat serum underwent analysis for concentrations of FSH and E2 as per the instructions outlined in the ELISA kit (H206-1–2/H101-1–2/H102-1–1, NanJing JianCheng Bioengineering Institute, Nanjing, China). Measuring the absorbance levels took place at 492 nm with the utilization of a microplate reader (51119180ET, Thermo Fisher Scientific), and the levels of the tested indicators were obtained using the standard curve according to the absorbance value of each sample. The extent of damage to ovarian and uterine tissue structures in rats was assessed by HE staining-based follicle counting. Following overnight fixation in 4% paraformaldehyde at RT, the ovarian tissue was encased in paraffin, cut into sections with a thickness of 5 μm, and stained for 5 min with hematoxylin and 3 min with eosin. The numbers of hair follicles and corpus luteum were calculated according to their shape and diameter. Standard dewaxing of paraffin sections involved immersion in water, subsequent treatment with 20 μg/mL proteinase K for 20 min at 37℃, incubation with 0.1% Triton X-100 for 4 min at 4℃, and then exposure to 3% H 2 O 2 for 10 min at RT. The solution required for the TUNEL assay was readied and incubated at a temperature of 37 °C for a duration of 70 min, subsequently subjected to DAB coloring and counterstained with hematoxylin for a brief 3-min period. The apoptotic cell nuclei are in brownish yellow under light microscopy. A computation was conducted on the average value after counting the TUNEL-positive cells in five fields of high magnification selected randomly. Sections of ovarian tissue, with a thickness of 5 μm, were obtained after the freezing and slicing process. Subsequent to overnight incubation at 4 °C in darkness, the designated primary antibody anti-γH2 AX (ab81299, Abcam) was applied to the compartments, followed by staining with the fluorescent secondary antibody Alexa Fluor 488 goat anti-rabbit IgG (#4412S, Cell Signaling Technology) for 1 h at 37 °C. DAPI (MBD0020, Sigma-Aldrich) was used for staining. Utilization of fluorescence microscopy (Nikon ECLIPSE E800) facilitated the observation and photography of the samples, followed by analysis through Image J software. The application of SPSS 26.0 statistical software (IBM, Armonk, New York) was pivotal in processing all data. Information gathered was portrayed as mean ± standard deviation. Firstly, evaluation of normality and homogeneity in variance was undertaken. An unpaired t test was utilized to perform comparisons between the groups. Statistical analysis involved the comparison of data across different groups using one-way ANOVA or repeated ANOVA, and subsequently conducting the Tukey post hoc evaluation. Significance at the p  < 0.05 level indicated a notable variance.

DNA damage and repair in ovarian GCs shares strong correlation with POI (Takahashi et al. 2021 ; Sha et al. 2021 ). In our study, our research delved into the molecular pathways implicated in hUCMSC-derived EVs in CP-induced POF and demonstrated that hUCMSC-derived EVs could deliver HCP5 to GCs to prevent CP-induced POF through interaction with miR-20a-5p and YAP1. Initially, our study found that hUCMSC-EVs could reverse CP-induced GC damage. Transplantation of hUCMSCs has been highlighted to be therapeutic for POF (Shareghi-oskoue et al. 2021 ). hUCMSCs-derived exosomes could ameliorate ovarian GC apoptosis in a rat model of POF (Qu et al. 2022 ). Of note, it has also been shown that EVs from hUCMSCs (Sun et al. 2017 ) and bone marrow MSCs (Sun et al. 2019 ) significantly up-regulated Bcl-2 levels and down-regulated cleaved Caspase-3 levels in a cisplatin-induced POF model, suggesting that EVs protect against chemotherapeutic drug-induced apoptosis of ovarian GCs. The conclusions drawn from earlier studies are in line with what we have discovered. In our study, we further demonstrated that hUCMSC-EVs could be implicated in the occurrence of CP-induced DNA harm in GCs and in ovarian functional impairment in a POF rat model via delivery of HCP5. Based on the transcriptome sequencing data of GCs after CP induction, we found that HCP5 was significantly down-regulated after CP induction and that HCP5 was significantly enriched in hUCMSC-EVs. HCP5 could exert transcriptional regulation on MSH5 and on DNA damage repair through YB1 and thus prevent the development of POF (Wang et al. 2020 ). Besides, overexpressed HCP5 in hUCMSC-derived exosomes was reported to enhance ovarian GC proliferation through the MSI2/ESR1 axis (Sun et al. 2022 ). Different from our study, these studies fail to explore the delivery of HCP5 by hUCMSC-EVs in the setting of POF. Furthermore, the current investigation uncovered that HCP5 was involved in up-regulating the expression of YAP1 by interacting with miR-20a-5p and that YAP1 was significantly down-regulated after CP induction. Overexpression of HCP5 reversed the down-regulation of YAP1 in CP-induced GCs, and interfering with YAP1 counteracted the mitigating influence of hUCMSC-EVs on CP-induced damage in GCs. Intriguingly, miR-20a-5p was implicated in the pathophysiology of polycystic ovary syndrome (Chen et al. 2022 ). Additionally, miR-20a could regulate NTN4 expression to result in endometriosis (Zhao et al. 2014 ). YAP1 protein is a key gene that regulates proliferation and apoptosis of GCs (Ji et al. 2017 ; Li et al. 2022 ). Active form of YAP1 was predominantly expressed in proliferative GCs and timely YAP1 activation of in GCs is required for ovarian follicle development (Lv et al. 2019 ). Prior research unfolded that YAP1 overexpression conferred protection against TOCP-induced ovarian dysfunction (Hu et al. 2019 ). YAP1 has been shown to be regulated by a miRNA (Sun et al. 2020 ), and HCP5 could exert ceRNA mechanisms to mediate miRNA-influenced disease processes (Yang et al. 2022 ; Li et al. 2021 ; Zhang et al. 2022 ). Interestingly, HCP5 was found to interact with has-miR-20b-5p in breast cancer (Wu et al. 2018 ). Nevertheless, literature has provided minimal discussion on the ceRNA regulatory mechanism connecting HCP5, miR-20a-5p, and YAP1. By acting as a miR-20a-5p sponge, HCP5 augmented YAP1 expression was substantiated in our current research. It was through regulation of the miR-20a-5p/YAP1 axis that HCP5 delivered by.hUCMSC-EVs could contribute to protection against POF. Taken together, the current data suggested that HCP5 delivered by hUCMSC-EVs may play a ceRNA mechanism in GCs, where HCP5 sponged miR-20a-5p and up-regulated YAP1 expression, thereby attenuating DNA damage in GCs and ultimately protecting against chemotherapy-induced POF (Fig.  8 ). Our study provides a new therapeutic direction and molecular rationale for chemotherapy-induced POF. Nonetheless, there exist various constraints in this research. Primarily, our transcriptome sequencing samples are mainly CP-induced samples; secondly, we did not perform sequencing analysis for differentially expressed lncRNAs in hUCMSC; in addition, there is a lack of clinical samples to validate this paper. Fig. 8 Molecular mechanism of HCP5 delivery by hUCMSC-EVs to mitigate DNA damage in GCs for the treatment of chemotherapy-induced POF. hUCMSC-EVs may deliver HCP5 into GCs to exert the ceRNA mechanism, by which HCP5 sponges miR-20a-5p, and up-regulate the expression of YAP1, thus reducing the DNA damage in GCs and protecting against POF caused by chemotherapy Molecular mechanism of HCP5 delivery by hUCMSC-EVs to mitigate DNA damage in GCs for the treatment of chemotherapy-induced POF. hUCMSC-EVs may deliver HCP5 into GCs to exert the ceRNA mechanism, by which HCP5 sponges miR-20a-5p, and up-regulate the expression of YAP1, thus reducing the DNA damage in GCs and protecting against POF caused by chemotherapy

In the context of individuals under 40, premature ovarian failure (POF), alternatively termed untimely ovarian inadequacy or untimely menopause, signifies the cessation of ovarian function prematurely, leading to clinical manifestations attributed to deficient estrogen levels (Szeliga et al. 2021 ). Significant occurrences of POF can arise in young cancer patients as a major adverse effect of chemotherapy, and cyclophosphamide (CP) can induce follicle activation in this process (Kalich-Philosoph et al. 2013 ). Granulosa cells (GCs) are pivotal in follicular growth and regulation of ovarian function and may be subjected to DNA damage (Abdelnour et al. 2020 ). Therefore, there is a need to find potential therapeutic candidates to mitigate POF based on protection of GCs. Extracellular vesicles (EVs) are important regulators of intercellular communication (Whitford and Guterstam 2019 ). EVs can interact with target cells via their intrinsic surface expression ligands to transport their cargoes (proteins, microRNAs (miR or microRNA) etc.) into the cytoplasm of recipient cells, thereby affecting the phenotype and function of the target cells (Elsharkasy et al. 2020 ). Multiple disease models, such as kidney, liver, cardiovascular diseases, and neurological ailments, have featured in discussions on the potential therapeutic impacts of EVs from stem cells (Urabe et al. 2020 ). Human umbilical cord mesenchymal stem cells (hUCMSC)-derived EVs could be localized in post-transplant ovarian tissue and exert a profound impact on stimulating ovarian angiogenesis and restoring ovarian function in rats with POF rats (Yang et al. 2019 ). However, the specific pathways and mechanisms underlying the protective effects of hUCMSCs-derived EVs on ovarian function are still lacking in experimental studies. The regulation of target mRNAs is influenced by the competition for post-transcriptional control through shared miRNAs binding sites, which is a characteristic of long non-coding RNAs (lncRNAs) (Panni et al. 2020 ). Analysis using bioinformatics tools in the present investigation revealed varying expression levels of lncRNA HLA complex P5 (HCP5) and miRNA-20a-5p in cases of POF. HCP5 participated in POF through transcriptional regulation of MSH5 as well as DNA damage repair through YB1 (Wang et al. 2020 ). Overexpression of HCP5 in hUCMSC-derived exosomes could augment the proliferation of ovarian GCs via the MSI2/ESR1 axis (Sun et al. 2022 ). Intriguingly, participation of miRNA-20a-5p in the pathological mechanism of polycystic ovarian syndrome has been documented (Chen et al. 2022 ). A previous study found that Yes1 associated transcriptional regulator (YAP1) protein is required for the proliferation of GCs, linking to the pathogenesis of polycystic ovary syndrome (Ji et al. 2017 ). Moreover, restoration of YAP1 could regulate mitochondrial function and apoptosis, which aided in GC function regulation to reduce ovarian reserve (Li et al. 2022 ). Our investigation focused on elucidating the plausible molecular mechanisms underlying the conveyance of HCP5 by hUCMSC-EVs for the purpose of overseeing DNA damage repair in GCs, involving the miR-20a-5p/YAP1 signaling pathway, affecting chemotherapy-induced POF, thereby contributing to potential therapeutic strategies for such cases.

Below is the link to the electronic supplementary material. Supplementary file1 Identification of hUCMSCs and hUCMSC-EVs. A, Observation of the morphology of hUCMSCs at passage 3 under a light microscope. B, Oil red O staining, alizarin red S staining or Alcian blue solution staining to detect lipogenic, osteogenic and chondrogenic differentiation abilities of hUCMSCs respectively. C, Flow cytometry to detect hUCMSCs surface markers. D, NTA to detect diameter distribution and concentration of EVs. E, WB to detect expression of the EVs marker proteins CD81, Alix, Hsp70, and the endoplasmic reticulum protein Calnexin. All cell experiments were repeated three times (JPG 2553 KB) Supplementary file2 Bioinformatics screening of HCP5 downstream regulatory genes. A, Venn plot showing the intersecting genes between DEGs in untreated and CP-induced GCs analyzed through transcriptome sequencing and genes related to"DNA damage"in the GeneCard database. B, Top 10 genes in terms of number of nodes of proteins encoded by 12 genes. C, Volcano map of DEGs in untreated and CP-induced GCs. Inverted triangles in green represent down-regulated mRNAs, and red triangles represent up-regulated mRNAs (n = 4) (JPG 1588 KB) Supplementary file3 (DOCX 17 KB) Supplementary file1 Identification of hUCMSCs and hUCMSC-EVs. A, Observation of the morphology of hUCMSCs at passage 3 under a light microscope. B, Oil red O staining, alizarin red S staining or Alcian blue solution staining to detect lipogenic, osteogenic and chondrogenic differentiation abilities of hUCMSCs respectively. C, Flow cytometry to detect hUCMSCs surface markers. D, NTA to detect diameter distribution and concentration of EVs. E, WB to detect expression of the EVs marker proteins CD81, Alix, Hsp70, and the endoplasmic reticulum protein Calnexin. All cell experiments were repeated three times (JPG 2553 KB) Supplementary file2 Bioinformatics screening of HCP5 downstream regulatory genes. A, Venn plot showing the intersecting genes between DEGs in untreated and CP-induced GCs analyzed through transcriptome sequencing and genes related to"DNA damage"in the GeneCard database. B, Top 10 genes in terms of number of nodes of proteins encoded by 12 genes. C, Volcano map of DEGs in untreated and CP-induced GCs. Inverted triangles in green represent down-regulated mRNAs, and red triangles represent up-regulated mRNAs (n = 4) (JPG 1588 KB) Supplementary file3 (DOCX 17 KB)