SIRT5–RAC2 Axis Drives Monocyte‐to‐Macrophage Differentiation to Promote Inflammatory Injury in Premature Ovarian Insufficiency
Wenjing TanTai, Yaqi Li, Shengnan Liu, Minjuan Wu, Zhixiao Liu, Junfeng Jiang, Jingjing Chen, Xiaoding Xu, Li Li, Chaoqun Li, Fang Zhao, Ye Liu, Haitao Ni, Tengfei Zhang, Mingjuan Xu, Chaofeng Han
TL;DR
This study shows that the SIRT5–RAC2 pathway influences macrophage behavior in ovarian inflammation, offering a potential treatment target for premature ovarian insufficiency.
Contribution
The study identifies SIRT5 and RAC2 as a novel regulatory axis controlling macrophage differentiation and inflammation in premature ovarian insufficiency.
Findings
SIRT5 deficiency reduces macrophage count and M1 polarization, decreasing ovarian inflammation.
SIRT5 stabilizes RAC2 through desuccinylation, promoting macrophage differentiation and M1 polarization.
Pharmacological inhibition of SIRT5 protects follicular integrity and reduces granulosa cell apoptosis in POI models.
Abstract
Premature ovarian insufficiency (POI) is a major cause of infertility and endocrine dysfunction, in which chronic inflammation plays a critical role. The homeostasis of tissue‐resident macrophages and monocyte‐differentiated macrophages from peripheral blood serves as a key mechanism of inflammation across organs, yet their phenotypic plasticity in ovarian pathologies, including POI, remains poorly understood. Here, we identify that SIRT5 deficiency decreases macrophage count by attenuating monocyte‐macrophage differentiation. SIRT5 deficiency markedly attenuated follicular depletion and granulosa cell apoptosis, coinciding with reduced M1 macrophage infiltration and cytokine expression in the POI model. Mechanistically, we uncovered RAC2 as a novel succinylation substrate of SIRT5. SIRT5 deficiency elevated RAC2 succinylation, promoting its proteasomal degradation and thereby impairing…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
FIGURE 1
FIGURE 2
FIGURE 3
FIGURE 4
FIGURE 5
FIGURE 6
FIGURE 7- —National Natural Science Foundation of China10.13039/501100001809
- —National Key R&D Program of China10.13039/501100012166
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsSirtuins and Resveratrol in Medicine · Immune cells in cancer · Reproductive Biology and Fertility
Introduction
1
Over recent decades, global fertility rates have steadily declined. Premature ovarian insufficiency (POI), affecting at least 1% of women under 40, has emerged as a key factor exacerbating this trend, with significant implications for fertility and long‐term endocrine health [1]. The most established causes of POI include genetic defects, autoimmune processes, and iatrogenic damage. These etiological factors contribute to POI either by directly impairing folliculogenesis and oocyte maturation or by disrupting ovarian immune and/or stromal homeostasis [2, 3]. Ovarian homeostasis is tightly coordinated through the crosstalk between oocytes, granulosa, theca, and stroma cells and a diverse repertoire of immune cells, mediated by their secreted growth factors, cytokines, and chemokines. Disruption of this immune–endocrine interplay has been implicated in a considerable proportion of POI cases [4, 5]. For example, ovarian‐tropic viral infections often elicit robust local immune responses marked by T cells, B cells, and macrophage infiltration into developing follicles. While essential for viral clearance, these immune cells also release pro‐inflammatory cytokines and cytotoxic mediators that cause tissue damage. Similarly, in autoimmune POI, dysregulated immune responses involve CD4⁺ and CD8⁺ T cells, B cells, and macrophages, along with elevated cytokine production, which collectively mediate aberrant immune attacks on ovarian tissue, ultimately increasing POI risk [4, 6]. It is postulated that such immune dysregulation may establish a pro‐inflammatory milieu that exacerbates damage to the ovarian reserve. Preliminary clinical evidence also indicates that POI patients exhibit elevated levels of pro‐inflammatory cytokines and immune cells in peripheral blood [7]. However, many aspects remain unresolved, including the mechanisms governing immune cell recruitment and activation within the ovary, and the detailed pathways by which cytokines impair folliculogenesis.
In the ovary, macrophages constitute the predominant immune cell population and play a central role in regulating local inflammation, while also supporting angiogenesis, steroidogenesis, and tissue remodeling during various physiological stages [8]. Under chronic or excessive inflammatory conditions, dysregulation of macrophage activity has emerged as a critical driver of immune imbalance within the ovarian microenvironment. This is characterized by a phenotypic shift toward pro‐inflammatory M1‐polarized macrophages and suppressed reparative M2‐polarized phenotypes, an imbalance that perpetuates local ovarian inflammation [9, 10]. During ovarian aging, this imbalance contributes to extensive remodeling of the ovarian niche. Cytokines secreted by aberrantly activated M1‐like macrophages activate MAPK and NF‐κB signaling cascades in granulosa cells, triggering their senescence and/or apoptosis, which in turn promotes stromal fibrosis and reproductive decline [10, 11]. However, despite preliminary evidence linking macrophage dysfunction to ovarian inflammation and the progression of POI, critical knowledge gaps remain. Ovarian macrophages, once believed to derive solely from bone marrow‐resident monocytes, have recently been shown to also originate from embryonic yolk sac and fetal liver progenitors [12, 13]. In the context of POI, the ontogeny of infiltrating macrophages, their replenishment dynamics under cytotoxic stress, and the regulatory mechanisms orchestrating their differentiation and polarization remain incompletely understood. These knowledge gaps continue to hinder the development of macrophage‐targeted strategies aimed at preserving ovarian function.
Lysine succinylation alters protein charge by adding a succinyl group to lysine residues, thereby modulating protein stability, localization, structure, and enzymatic activity [14]. SIRT5 is a mitochondrial NAD⁺‐dependent lysine desuccinylase that removes succinyl, malonyl, and glutaryl groups from target proteins, and primarily regulates energy and redox homeostasis, as evidenced by widespread hyper‐succinylation of metabolic enzymes involved in ketogenesis, fatty acid β‐oxidation, the TCA cycle, and ATP synthesis in SIRT5‐deficient mice [15, 16]. Although traditionally viewed as a mitochondrial desuccinylase, SIRT5 also functions in the cytosol and peroxisomes, where it regulates glycolysis, redox balance, and stress responses through newly identified cytosolic substrates [17, 18, 19]. However, the broader physiological relevance of its activity on cytoplasmic, non‐metabolic substrates remains incompletely understood. Another important open question is whether these modifications influence subsequent posttranslational modifications such as ubiquitination or phosphorylation, enabling the integration of diverse signals in a context‐dependent manner. We previously elucidated that cytoplasmic SIRT5 competes with SIRT2 for NF‐κB p65 binding, sustaining p65 acetylation and transcriptional activity during sepsis [20]. This finding positioned SIRT5 as a key regulator of immune signaling beyond its metabolic functions, leading us to hypothesize that it may contribute to diverse pathologies marked by persistent or chronic inflammation. In the course of this work, we also observed that SIRT5 is highly enriched in ovarian macrophages, an essential population involved in maintaining tissue homeostasis and coordinating immune–endocrine interactions. Prompted by this distinct expression pattern, we investigated whether SIRT5 influences ovarian immune dynamics.
Our study identifies SIRT5 as a critical regulator of the ovarian immune microenvironment by modulating macrophage differentiation and subsequent overproduction of inflammatory cytokines. We show that SIRT5 deficiency markedly reduces ovarian macrophage infiltration and granulosa cell apoptosis driven by excessive pro‐inflammatory cytokine production (TNF‐α, IL‐1β). This ultimately mitigates follicular developmental defects and preserves the follicular pool following cyclophosphamide insult. Mechanistically, we identify Ras‐related C3 botulinum toxin substrate 2 (RAC2) as a novel substrate of SIRT5, and demonstrate that SIRT5‐mediated desuccinylation protects RAC2 from proteasomal degradation, thereby promoting monocyte‐to‐macrophage differentiation and enhancing subsequent pro‐inflammatory macrophage activation. We also show that targeting the SIRT5–RAC2 axis represents a promising strategy to preserve ovarian homeostasis through the dynamic regulation of macrophages.
Results
2
SIRT5 Deficiency Decreases Macrophage Infiltration in the Ovary
2.1
To investigate immune perturbations in the ovarian microenvironment during POI pathogenesis, we first established a chemotherapy‐induced POI model using an alkylating agent, one of the most commonly employed experimental systems for mimicking human iatrogenic ovarian insufficiency. Eight‐week‐old female mice were intraperitoneally injected with cyclophosphamide (CTX, 50 mg/kg), a widely accepted POI model (Figure S1A) [21, 22]. CTX exposure induced pronounced ovarian dysfunction, a significant decrease in ovarian weight and follicle reserve, along with widespread follicular atresia (Figure S1B). qRT‐PCR analysis of whole ovarian tissue revealed that CTX administration significantly downregulated key regulators of folliculogenesis, including the granulosa cell‐specific genes anti‐Müllerian hormone (Amh), forkhead box L2 (Foxl2), and follicle‐stimulating hormone receptor (Fshr), as well as the oocyte‐specific gene growth differentiation factor 9 (Gdf9), indicating impaired follicle development (Figure S1C). In parallel, the expression of cytochrome P450 family 19 subfamily A member 1 (Cyp19a1), a granulosa cell‐specific enzyme responsible for estrogen biosynthesis, was markedly reduced, suggesting disrupted steroidogenic function (Figure S1C,D). Conversely, CTX‐treated ovaries showed increased expression of antioxidant genes glutathione peroxidase 1 (Gpx1) and superoxide dismutase 1 (Sod1), indicating a compensatory response to elevated intracellular oxidative stress (Figure S1E). Additionally, CTX increased expression of the pro‐apoptotic gene Bcl‐2‐associated X protein (Bax) and reduced expression of the anti‐apoptotic gene B‐cell lymphoma 2 (Bcl2), suggesting activation of apoptotic signaling pathways in ovarian cells (Figure S1F). These observations together demonstrated that the CTX‐induced model recapitulates key pathological features of clinical POI.
We next examined whether aberrant immune cell infiltration could also be recapitulated in the CTX‐induced POI model by detecting the expression of a subset of marker genes in ovaries with or without CTX administration. These marker gene expression levels served as proxies for the relative abundance of key immune cell populations, including T cells, B cells, dendritic cells, plasma cells, and macrophages, all of which have been implicated in immune dysregulation during POI progression. Among these populations, we observed a pronounced upregulation of the canonical macrophage markers Adgre1 and Cd68 in CTX‐treated ovaries, whereas markers for T cells, plasma cells, and dendritic cells showed only modest increases (Figure 1A). These findings were further validated by flow cytometric analysis, which confirmed that macrophages represent the most prominently expanded immune population in CTX‐treated ovaries compared to controls (Figure 1B and Figure S2A). Consistently, immunofluorescence (IF) staining for F4/80 also revealed a significant accumulation of macrophages in CTX‐treated ovaries (Figure 1C), highlighting macrophage expansion as a prominent immune perturbation associated with POI pathogenesis.
*SIRT5 deficiency reduces macrophage infiltration in the mouse ovary. (A) qRT‐PCR analysis of the markers of macrophage (Adgre1, Cd68), dendritic cell (Cd209a), T cell (Cd3e), and plasma cell (Cd138) in ovarian tissues from wild‐type (WT) and cyclophosphamide (CTX)‐induced premature ovarian insufficiency (POI) mice (n = 6 per group). (B) Quantification of immune cell populations in ovarian single‐cell suspensions from WT and CTX‐POI mice, analyzed by flow cytometry. The bar graph shows the relative proportions of distinct immune cell subsets, including plasma cells (B220−CD138⁺), dendritic cells (CD11c⁺MHCII⁺), macrophages (CD11b⁺F4/80⁺), B cells (B220⁺), and T cells (CD3⁺). n = 3 per group. (C) Left: Representative immunofluorescence (IF) staining images showing F4/80⁺ macrophages (green) in ovarian tissues from WT and CTX‐induced POI mice. Scale bar: 50 µm. Right: Quantitative analysis of the number of F4/80⁺ macrophages in the two groups (n = 3 per group). (D) Single‐cell RNA sequencing analysis of adult murine ovarian tissues showing Sirt5 mRNA expression (log2[TPM + 1]) across distinct ovarian cell populations. (E) Co‐IF staining for SIRT5 (red), F4/80 (green), and AMH (white) in ovarian sections from Sirt5 +/+ and Sirt5 −/− mice. Scale bar: 10 µm. (F) Immune cell infiltration in the ovarian microenvironment was estimated using ImmuCellAI‐mouse based on transcriptomic data from Sirt5 +/+ and Sirt5 −/− ovarian tissues. (G) qRT‐PCR was performed to analyze the mRNA expression of general macrophage markers (Adgre1, Cd68, Fcgr1, and Cx3cr1), and M1 or M2 polarization markers (Nos2, Cd80, and Cd86 for M1; Arg1 and Mrc1 for M2) in ovarian tissues from Sirt5 +/+ and Sirt5 −/− mice (n = 6 per group). (H–J) IF staining for macrophages (H: F4/80⁺ cells, green; n = 5 per group), M1 macrophages (I: CD86⁺F4/80⁺, red/green; n = 3 per group) and M2 macrophages (J: CD206⁺F4/80⁺, red/green; n = 4 per group) in ovarian sections from Sirt5 +/+ and Sirt5 −/− mice; Quantification of F4/80⁺, CD86⁺F4/80⁺ and CD206⁺F4/80⁺ cell proportions is shown in the right panels. Scale bar: 50 µm. Statistical significance was assessed using an unpaired t‐test (A–C,G–J). Data are presented as Mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; ns, not significant.
Our previous studies demonstrated that the Sirtuin family member SIRT5 enhances macrophage‐mediated innate inflammatory responses in murine models of sepsis by modulating key inflammatory signaling pathways [20]. Intriguingly, analysis of adult mouse ovary scRNA‐seq data from the Mouse Cell Atlas revealed that Sirt5 expression is highly enriched in ovarian macrophages and luteal cells, with negligible expression in oocytes (Figure 1D) [23]. Examination of two independent human ovary scRNA‐seq datasets showed a strikingly similar distribution, with SIRT5 preferentially expressed in monocyte–macrophage lineage clusters rather than other ovarian cell types (Figure S2B) [24, 25]. IF staining further validated robust SIRT5 expression in F4/80⁺ macrophages that were mainly distributed around AMH‐positive developing follicles within the ovarian stroma (Figure 1E), suggesting a potential role for SIRT5 in regulating ovarian immune homeostasis.
To explore the physiological relevance of SIRT5 in ovarian immune homeostasis and reproductive function, we generated Sirt5‐deficient mice and assessed their ovarian immune landscape and fertility outcomes. Although complete ablation of Sirt5 mRNA and protein was confirmed in ovaries of SIRT5 knockout mice (Figure 1E and Figure S3A,B), histological analyses revealed that Sirt5‐deficient ovaries were morphologically normal, containing follicles at multiple developmental stages and intact theca cell layers (Figure S3C,D). Although SIRT5 is also expressed in luteal cells, the ovarian mRNA levels of key steroidogenic enzymes (Star, Cyp11a1, Cyp17a1, Hsd3b1, and Hsd3b2) were comparable between *Sirt5^−/−^
- and control ovaries, indicating that SIRT5 loss has minimal impact on luteal steroidogenic function. Consistently, fertility testing showed no difference in litter size between *Sirt5^−/−^
- and wild‐type females (Figure S3E,F). Together, these data demonstrate that SIRT5 deficiency does not impair folliculogenesis, ovarian hormone biosynthesis, or reproductive capacity under physiological conditions.
Given the selective enrichment of SIRT5 in ovarian macrophages, we first examined its potential role in regulating the ovarian immune microenvironment under physiological conditions. To this end, we performed transcriptomic profiling of ovaries from SIRT5‐deficient and wild‐type untreated mice. Subsequent immune cell infiltration analysis using ImmuCellAI‐mouse [26] revealed a significant reduction in total macrophage infiltration, including both M1‐like and M2‐like subtypes, in SIRT5‐deficient ovaries compared to controls (Figure 1F). This decrease was accompanied by reduced expression of pan‐macrophage markers (Adgre1, Cd68, Fcgr1, and Cx3cr1) as well as M1 (Nos2, Cd80, Cd86) and M2 (Arg1, Mrc1) signature genes (Figure 1G). IF staining and FACS quantification further confirmed a significant loss of total (F4/80⁺), M1 (CD86⁺F4/80⁺), and M2 (CD206⁺F4/80⁺) macrophages in *Sirt5^−/−^
- ovaries (Figure 1H–J and Figure S3G). Together, these findings indicate that SIRT5 deficiency reduces macrophage abundance in the ovary without perturbing reproductive capacity, uncovering a previously unappreciated role for SIRT5 in modulating immune cell infiltration and shaping the ovarian inflammatory microenvironment.
SIRT5 Deficiency Disrupts Monocyte‐to‐Macrophage Differentiation in the Ovary
2.2
To characterize the molecular and functional changes in ovarian macrophages following SIRT5 depletion, we performed Smart‐seq analysis on FACS‐purified F4/80⁺CD11b^+^ macrophages isolated from control and *Sirt5^−/−^
- ovaries (Figure S3G). Transcriptomic profiling identified a distinct set of genes differentially expressed between Sirt5‐deficient and wild‐type ovarian macrophages (Figure S4A). Gene set enrichment analysis (GSEA) revealed that genes downregulated in *Sirt5^−/−^
- ovarian macrophages were significantly enriched (p = 0.0015) for transcripts induced during monocyte‐to‐macrophage differentiation under macrophage colony‐stimulating factor (M‐CSF) stimulation, with 31 out of 160 genes overlapping (referred to as the ‘M‐CSF cluster’) (Figure 2A and Figure S4B). This observation raised the possibility that M‐CSF‐responsive transcriptional programs may be compromised in Sirt5‐deficient monocytes undergoing macrophage differentiation. To further evaluate this hypothesis, we analyzed the expression of two gene clusters in control and *Sirt5^−/−^
- monocytes under basal conditions and following M‐CSF stimulation. Cluster 1 consisted of genes from the M‐CSF cluster with previously reported roles in macrophage function (e.g., Card6, Cd55, Cisd1, Gimap8, and Igbp1). Cluster 2 included canonical downstream targets of CSF1R signaling, a class III receptor tyrosine kinase that serves as a central hub for monocyte differentiation in response to M‐CSF stimulation, including key effectors such as Ccr7, c‐Jun, and Mmp9 [27]. At baseline, Cluster 1 expression was reduced in *Sirt5^−/−^
- monocytes, whereas Cluster 2 genes remained largely unchanged. Upon M‐CSF treatment, both clusters were robustly induced in control monocytes, but this induction was markedly attenuated in the *Sirt5^−/−^
- group (Figure 2B). These findings suggest that SIRT5 deficiency may compromise the transcriptional responsiveness of monocytes to M‐CSF stimulation.
*SIRT5 deficiency impairs monocyte‐to‐macrophage differentiation in the ovary. (A) Gene Set Enrichment Analysis (GSEA) showing that genes downregulated in Sirt5 −/− ovarian macrophages compared with wild‐type controls are enriched for pathways required for macrophage differentiation. (B) qRT–PCR analysis of the indicated genes during M‐CSF‐induced monocyte‐to‐macrophage differentiation (n = 4 per group). (C) IF staining and quantification for macrophages (F4/80+, green) in ovaries from Sirt5 +/+ and Sirt5 −/− mice at day 2 and day 21 following macrophage depletion using CSF1‐neutralizing antibody and clodronate liposomes (n = 4 per group). Scale bar: 20 µm. (D) qRT–PCR analysis of Sirt5 mRNA expression during M‐CSF‐induced monocyte‐to‐macrophage differentiation (n = 3 per group). (E) qRT–PCR showing changes in mRNA levels of Adgre1, Cd68, and Fcgr1 in negatively selected monocytes from Sirt5 +/+ and Sirt5 −/− mice following M‐CSF treatment, normalized to β‐actin (n = 3 per group). (F) IF staining for macrophages (F4/80+, green) in negatively selected monocytes isolated from Sirt5 +/+ and Sirt5 −/− mice. Scale bar: 20 µm. (G) Western blot analysis of phosphorylated PAK1/2/3, JAK2, and AKT in sorted monocytes from Sirt5 +/+ and Sirt5 −/− mice, with or without M‐CSF stimulation. (H,I) IF analysis of p‐AKT and p‐STAT5 in CCR2⁺ monocytes. Ovarian sections from the indicated genotypes and time points after macrophage depletion were stained for CCR2 (green) and (H) p‐AKT or (I) p‐STAT5 (red). Scale bar: 20 µm. Bar graphs show the percentage of p‐AKT⁺CCR2⁺ or p‐STAT5⁺CCR2⁺ cells (n = 4 per group). For all quantification graphs, data are presented as Mean ± SD. Statistical significance was assessed using one‐way ANOVA with Tukey's multiple comparisons test (B,C,E) and unpaired t‐test (D,H,I). *p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; ns, not significant.
During embryogenesis, yolk sac‐derived macrophages are the first to populate the ovary, whereas after birth, circulating Ly6C^+^ bone marrow‐derived monocytes are recruited to the ovarian microenvironment and give rise to phenotypically distinct tissue‐resident macrophage subsets [13, 28]. Given the marked reduction in ovarian macrophage populations observed in adult *Sirt5^−/−^
- mice, along with the suppressed expression of CSF1R signaling targets in SIRT5‐deficient monocytes following M‐CSF stimulation (Figure 2B), we hypothesized that the dynamic replenishment of ovarian macrophages via monocyte‐to‐macrophage differentiation might be impaired in the absence of SIRT5. To investigate this, we employed a depletion strategy using clodronate liposomes in combination with anti‐CSF1 antibody to eliminate the majority of ovarian macrophages in both control and Sirt5‐deficient mice (Figure 2C and Figure S4C) [28]. In control animals, the macrophage pool could be substantially regenerated within 21 days, likely via recruitment and differentiation of bone marrow‐derived monocytes. In sharp contrast, this regenerative process was severely impaired in *Sirt5^−/−^
- ovaries, as evidenced by IF showing a profound reduction of F4/80⁺ monocyte‐derived macrophages (Figure 2C). These findings suggest that SIRT5 is essential for the reconstitution of the ovarian macrophage pool, likely by supporting monocyte‐to‐macrophage differentiation.
SIRT5 Promotes Monocyte‐to‐Macrophage Differentiation by Potentiating CSF1R Signaling
2.3
CSF1–CSF1R signaling directs monocyte‐to‐macrophage differentiation and is required for maintaining ovarian macrophage populations, as Csf1r deficiency markedly depletes these cells [29]. To determine whether SIRT5 plays a functional role in CSF1‐induced macrophage differentiation, we employed two in vitro models using monocyte‐containing populations. In the first model, total bone marrow cells from wild‐type (WT) and Sirt5 knockout (*Sirt5^−/−^ *) mice were cultured in the presence of CSF1 to assess macrophage differentiation in a mixed hematopoietic context. In the second model, monocytes were purified from bone marrow via a negative selection approach to obtain a more defined population and then subjected to CSF1 stimulation under identical conditions. In both models, Sirt5 mRNA levels were markedly upregulated during the course of monocyte‐to‐macrophage differentiation, supporting a potential regulatory role for SIRT5 in this process (Figure 2D and Figure S4D). Notably, loss of SIRT5 significantly impaired macrophage differentiation in both models, as evidenced by reduced transcription of canonical macrophage markers including Adgre1, Cd68, and Fcgr1 (Figure 2E and Figure S4E). IF further confirmed a decreased proportion of F4/80⁺ cells in *Sirt5^−/−^
- monocyte‐derived populations relative to wild‐type controls following CSF1 treatment (Figure 2F and Figure S4F).
To mechanistically explore how SIRT5 influences this process, we examined the phosphorylation status of JAK2 and AKT, two central effectors downstream of CSF1R that drive monocyte survival and differentiation through the JAK‐STAT and PI3K/AKT pathways, respectively [30, 31, 32]. CSF1‐induced phosphorylation of both JAK2 and AKT was markedly reduced in *Sirt5^−/−^
- monocytes compared to controls (Figure 2G and Figure S4G). Consistent with our in vitro findings, a similar impairment in JAK/STAT and AKT signaling activation was observed during re‐establishment of macrophage populations following SIRT5 depletion. Specifically, upon recruitment to the ovarian microenvironment, SIRT5‐deficient monocytes exhibited significantly diminished levels of p‐AKT and p‐STAT5 (Figure 2H,I), indicating that SIRT5 is required to support CSF1R‐mediated signaling both in vitro and in vivo. Collectively, our data demonstrate that SIRT5 is required for CSF1‐driven macrophage differentiation within the ovarian microenvironment by sustaining JAK/STAT and PI3K/AKT signaling.
SIRT5 Promotes Inflammation in the Ovarian Microenvironment
2.4
Macrophages are key regulators of the ovarian microenvironment, contributing to tissue remodeling, hormone signaling, and immune surveillance. Aberrant macrophage infiltration and polarization, particularly a shift toward the pro‐inflammatory M1 phenotype, have been implicated in chronic inflammation, granulosa cell apoptosis, and compromised ovarian function. To assess how impaired monocyte‐to‐macrophage differentiation resulting from SIRT5 deficiency affects ovarian immune homeostasis, we analyzed RNA‐sequencing data from *Sirt5^+/+^
- and *Sirt5^−/−^
- ovaries (Figure 3A). Gene ontology and pathway enrichment analyses of differentially expressed genes (DEGs) revealed significant enrichment of the cytokine–cytokine receptor interaction pathway, indicating altered inflammatory cytokine signaling in the SIRT5‐deficient ovary (Figure 3B). Both RNA‐sequencing and qRT–PCR validation revealed that SIRT5 deficiency markedly reduced the mRNA expression of several macrophage‐derived cytokines and chemokines, predominantly pro‐inflammatory factors such as Tnf, Il1b, Il1a, Cxcl9, and Cxcl10 (Figure 3C,D). RNA‐sequencing further confirmed that these cytokines/chemokines are predominantly produced by macrophages residing in the ovarian stroma (Figure S4H). This reduction in cytokine expression, together with the observed decrease in macrophage abundance in SIRT5‐deficient ovaries (Figure 1H–J and Figure S3G), suggests that loss of SIRT5 reshapes the ovarian microenvironment by limiting macrophage expansion and consequently reducing pro‐inflammatory cytokine production.
*SIRT5 deficiency reduces M1 macrophage accumulation and ovarian inflammation in a CTX‐induced POI model. (A) Volcano plot showing differentially expressed genes (DEGs) between ovaries from Sirt5 +/+ and Sirt5 −/− mice, with upregulated genes displayed in red and downregulated genes in blue. (B) KEGG pathway enrichment analysis of the DEGs in (A), presented as a bubble plot. (C) Volcano plot highlighting significantly altered cytokine‐ and chemokine‐related genes in ovaries upon SIRT5 depletion. (D) qRT–PCR validation of decreased expression of cytokine genes in ovaries from Sirt5 −/− mice compared to wild‐type controls (n = 3 per group). (E) IF staining for Ly6C (red) and F4/80 (green) in ovaries from Sirt5 +/+, Sirt5 −/−, CTX‐treated Sirt5 +/+, and CTX‐treated Sirt5 −/− mice. Scale bar: 20 µm. (F) IF staining (top) and quantitative analysis (bottom) for CD86 (red) and F4/80 (green) to identify M1 macrophages in ovaries from the indicated groups (n = 3 per group). (G) TNF‐α and IL‐1β levels were quantified by ELISA in ovarian tissue homogenates and serum from the indicated groups (n = 5 per group). (H,I) IF staining in (H) or qRT‐PCR analysis in (I) of TNF‐α and IL‐1β in ovaries from the indicated groups (n = 3 per group). Scale bar: 50 µm. For all quantification graphs, data are presented as Mean ± SD. Statistical significance was assessed using an unpaired t‐test (D) and one‐way ANOVA with Tukey's multiple comparisons test (F,G,I). Significance levels are denoted as *p < 0.05; **p < 0.01; ***p < 0.001; and ***p < 0.0001; ns, not significant.
SIRT5 Deficiency Attenuates CTX‐Induced Ovarian Dysfunction
2.5
In the CTX‐induced POI model, CTX exposure elicited a marked expansion of monocyte‐derived macrophages (Ly6C⁺F4/80⁺) (Figure 3E), indicating that enhanced monocyte‐to‐macrophage differentiation contributes to the macrophage infiltration observed in POI ovaries (Figure 1A–C). CTX treatment also induced a strong increase in M1‐like macrophages within the ovarian stroma (Figure 3F), whereas M2‐like macrophage abundance remained largely unchanged (Figure S5A). ELISA assays of mouse serum and ovarian tissue revealed that CTX treatment markedly elevated the pro‐inflammatory cytokines TNF‐α and IL‐1β, with higher accumulation in ovarian tissue than in circulation (Figure 3G). Consistently, IF staining and qRT–PCR analyses revealed robust increases in TNF‐α and IL‐1β protein and mRNA levels in CTX‐treated ovaries (Figure 3H,I). These findings led us to hypothesize that, during POI pathogenesis, SIRT5 functions as a central immunoregulator by modulating the accelerated differentiation of monocytes into macrophages, thereby promoting M1‐like macrophage accumulation and establishing a pro‐inflammatory ovarian milieu. To assess whether SIRT5 loss reverses this inflammatory shift, we analyzed immune remodeling in the ovaries of SIRT5‐deficient POI mice. Strikingly, SIRT5 deficiency substantially mitigated the CTX‐induced expansion of monocyte‐derived macrophages in the ovary, resulting in a reduced accumulation of M1 macrophages and lower expression of TNF‐α and IL‐1β at both transcript and protein levels, compared to wild‐type controls (Figure 3E–I). Morphologically, SIRT5 depletion also attenuated the CTX‐induced loss of ovarian weight and partially preserved the follicle pool (Figure 4A,B). Immunostaining for mouse Vasa homolog (MVH) further revealed that CTX treatment markedly reduced the abundance of resting follicles (primordial follicles) and significantly lowered the number of growing follicles, including primary, secondary, and antral follicles, in control ovaries. In contrast, this germ cell loss was substantially mitigated in SIRT5‐deficient ovaries (Figure 4C,D). Beyond histological evidence, transcriptional analysis showed that SIRT5 deficiency alleviated CTX‐induced suppression of key genes involved in follicle development (Amh, Foxl2, Gdf9, and Fshr) and estrogen biosynthesis (Cyp19a1) (Figure 4E,F). Moreover, the intracellular response to oxidative stress and the pro‐apoptotic shift triggered by CTX were markedly reduced in SIRT5‐deficient ovaries (Figure 4G,H). Collectively, these findings establish SIRT5 as a key regulator of the ovarian immune environment and a potential target for mitigating CTX‐induced ovarian injury.
*SIRT5 deficiency ameliorates ovarian injury in POI. (A,B) Gross morphology (A) and H&E staining (B) of ovaries from Sirt5 +/+, Sirt5 −/−, CTX‐treated Sirt5 +/+, and CTX‐treated Sirt5 −/− mice. Scale bar: 100 µm. (C,D) IF staining for MVH (green, C) and quantification of follicle numbers (D, n = 3 per group) in ovaries from the indicated groups. Scale bar: 100 µm. Primordial follicle (Pdf); growing follicles include primary follicles (PrF), secondary follicles (SeF), and antral follicles (AnF). (E–H) qRT–PCR analysis of: (E) follicle development‐related genes (Amh, Foxl2, Gdf9, Fshr); (F) hormone synthesis enzyme gene (Cyp19a1); (G) antioxidant‐related genes (Gpx1, Sod1); (H) apoptosis regulators (Bax, Bcl2) in ovaries from Sirt5 +/+, Sirt5 −/−, CTX‐treated Sirt5 +/+, and CTX‐treated Sirt5 −/− mice (n = 3 per group). Statistical significance was assessed using one‐way ANOVA with Tukey's multiple comparisons test (D–H). Data are normalized to β‐actin and presented as Mean ± SD. *p < 0.05; **p < 0.01; ***p < 0.001; ***p < 0.0001; ns, not significant.
SIRT5 Deficiency Mitigates Caspase‐Dependent Granulosa Cell Apoptosis
2.6
Under chemotherapeutic stress, granulosa cells (GCs) in the ovary undergo apoptosis and/or senescence, triggered by genotoxic insults, oxidative stress, and DNA damage [33, 34, 35]. These cellular responses foster a deleterious intra‐ovarian environment characterized by chronic inflammation, stromal remodeling, and disrupted endocrine signaling, thereby accelerating follicular dysfunction and promoting follicular atresia [36]. Consistent with this, immunohistochemical (IHC) analysis revealed a reduction in Ki67‐positive GCs in growing follicles and a marked increase in cleaved Caspase‐3 staining in CTX‐treated ovaries, indicating diminished proliferation and enhanced apoptosis of GCs (Figure 5A,B). TUNEL and senescence‐associated β‐galactosidase (SA‐β‐Gal) assays further corroborated these findings, showing increased DNA fragmentation and senescence activity in GCs upon CTX exposure (Figure 5C,D). Notably, these pathological changes were substantially attenuated in CTX‐treated Sirt5‐deficient ovaries, suggesting that SIRT5 deficiency mitigates CTX‐induced GC apoptosis and senescence, thereby preserving follicular integrity (Figure 5A–D). We hypothesized that the reduced apoptosis in GCs observed in Sirt5‐deficient ovaries is driven, at least in part, by attenuated pro‐inflammatory signaling mediated by macrophage‐derived TNF‐α and IL‐1β.
*SIRT5 deficiency mitigates CTX‐induced granulosa cell apoptosis by attenuating inflammatory cytokine signaling. (A) Immunohistochemical (IHC) staining of Ki67 in ovarian sections from Sirt5 +/+, Sirt5 −/−, CTX‐treated Sirt5 +/+, and CTX‐treated Sirt5 −/− mice. Scale bar: 100 µm. (B) IF staining for Cleaved Caspase‐3 in ovarian sections from the indicated groups. Scale bar: 50 µm. (C) TUNEL assay showing apoptotic cells (green) in ovarian sections from the indicated groups. Scale bar: 50 µm. (D) Senescence‐associated β‐galactosidase (SA‐β‐Gal) staining in ovarian tissues from the indicated groups. Scale bar: 20 µm. (E–G) Control or SIRT5‐knockdown KGN cells were treated with vehicle, 4‐HC, IL‐1β + TNF‐α, or a combination of 4‐HC and IL‐1β + TNF‐α, and subjected to the following assays: (E) IF staining and quantification of Ki67⁺ cells (n = 5 per group). Scale bar: 50 µm. (F) Flow cytometric analysis of apoptotic cells (n = 3 per group). (G) Cell viability assay (n = 3 per group). (H) Heatmap of differentially expressed genes specifically altered between the 4‐HC and 4‐HC + IL‐1β + TNF‐α groups in KGN cells, while remaining unchanged between the vehicle and 4‐HC groups. These genes collectively define the Cyto‐POI gene set. (I) KEGG pathway enrichment analysis of upregulated or downregulated genes within the Cyto‐POI signature. (J) Left: Heatmap showing upregulated Cyto‐POI genes associated with cell death‐related pathways; right: Bubble plot indicating fold changes of individual genes between 4‐HC and 4‐HC + IL‐1β + TNF‐α or between vehicle and 4‐HC. (K) IF staining for Caspase‐7 (green) in granulosa cells from ovarian sections of Sirt5 +/+, Sirt5 −/−, CTX‐treated Sirt5 +/+, and CTX‐treated Sirt5 −/− mice. Scale bar: 50 µm. In panels (A,C,D) square dashed boxes indicate regions shown at higher magnification. In panels (A–C,K), follicular boundaries are outlined with dashed lines. Data in panels (E–G) are presented as the Mean ± SD. Statistical analyses were conducted using two‐way ANOVA with Tukey's multiple comparisons test for panels (E,F) and one‐way ANOVA with Tukey's multiple comparisons test for panel (G). Significance levels are indicated as *p < 0.05, ***p < 0.001, and ***p < 0.0001; ns, not significant.
To clarify the role of inflammatory cytokines in sensitizing granulosa cells to chemotherapeutic stress, we employed the human granulosa cell line KGN. Cells were treated with 4‐Hydroperoxycyclophosphamide (4‐HC), cytokines (IL‐1β and TNF‐α) alone, or their combination. Either 4‐HC or cytokines alone induced only modest apoptosis and exerted limited effects on proliferation. In sharp contrast, combined 4‐HC and cytokine treatment markedly exacerbated cellular injury, resulting in substantially increased apoptosis and pronounced suppression of cell proliferation compared with either single treatment (Figure 5E–G). To determine whether SIRT5 loss confers a granulosa cell–intrinsic survival advantage, we generated a stable SIRT5 knockdown (SIRT5 KD) KGN cell line and subjected it to the same apoptotic stimuli. We found that SIRT5 knockdown did not alter apoptosis or proliferation under any treatment condition (Figure 5E–G and Figure S6A). These results suggest that TNF‐α and IL‐1β synergize with chemotoxic stress to intensify granulosa cell injury.
Although TNF‐α and IL‐1β are well‐established mediators of stress‐induced apoptosis, the downstream molecular mechanisms by which they promote granulosa cell death within the ovarian microenvironment remain incompletely defined. To further dissect the underlying mechanisms, we performed RNA‐sequencing on KGN cells treated with 4‐HC alone or in combination with TNF‐α and IL‐1β. Treatment with 4‐HC alone altered the expression of genes involved in PI3K/AKT and MAPK signaling as well as cytokine–receptor interactions, consistent with its pro‐apoptotic effects (Figure S6B,C). Notably, co‐treatment with TNF‐α and IL‐1β induced a substantially broader and more robust transcriptional changes, suggesting an amplified apoptotic signaling (Figure 5H). Importantly, a distinct subset of genes was uniquely altered only in the cytokine and 4‐HC co‐treatment group, but not in cells treated with 4‐HC alone, relative to control. We defined this subset as the Cyto‐POI gene signature, which was significantly enriched for pathways associated with inflammation‐driven cell death, including p53 signaling, NOD‐like receptor signaling, apoptosis, and necroptosis (Figure 5I and Figure S6D). Protein–protein interaction analysis of upregulated genes from the Cyto‐POI signature identified caspase family members, including Caspase‐1, Caspase‐7, and Caspase‐10, appeared as central hub nodes within the apoptosis‐related cluster (Figure S6E), consistent with their established roles in programmed cell death and their selective induction under cytokine plus 4‐HC co‐treatment (Figure 5J). Caspase‐1 is involved in inflammasome activation and bridges inflammatory signaling with cell deaths [37]. Caspase‐10 serves as an initiator caspase in apoptotic cascades, while Caspase‐7 acts as an executioner caspase downstream of both intrinsic and extrinsic apoptotic cues [38]. We next validated that the upregulation of these caspases functionally links macrophage‐derived inflammatory signals to follicular somatic cell death under chemotherapeutic stress in vivo. IF staining revealed that CTX treatment triggered elevated expression of Caspase‐1 and ‐7 in somatic cells within the follicle. In contrast, their expression was markedly attenuated in CTX‐treated Sirt5‐deficient ovaries with reduced pro‐inflammatory cytokines (Figure 5K and Figure S6F). Together, these findings suggest that SIRT5 contributes to CTX‐induced follicular somatic cell apoptosis by amplifying the inflammatory microenvironment and activating caspase‐mediated cell death pathways, with granulosa cells likely representing the predominant apoptotic population.
SIRT5 Promotes Monocyte‐to‐Macrophage Differentiation by Desuccinylating and Stabilizing RAC2
2.7
Previous studies have established that SIRT5, through its desuccinylase activity, functions as a pivotal regulator of cellular metabolism by modulating the enzymatic activity of diverse substrates [14, 15, 39, 40]. Consistent with this, we observed a marked increase in global lysine succinylation levels in ovarian tissue lysates from *Sirt5^−/−^
- mice compared to wild‐type controls (Figure S7A). Although multiple SIRT5 substrates have been identified in various tissues, its specific targets within the monocyte/macrophage population, particularly in the ovarian microenvironment, remain largely uncharacterized. To address this, we performed co‐immunoprecipitation followed by mass spectrometry (IP‐MS) in CSF1‐stimulated monocytes to identify SIRT5‐interacting proteins involved in monocyte‐to‐macrophage differentiation (Figure 6A). From 116 candidate interactors enriched in SIRT5 immunoprecipitates (MS intensity > 50 000), we prioritized RAC2, a hematopoietic‐specific Rho GTPase, as a biologically relevant SIRT5 target based on its ovarian macrophage‐enriched expression and the established involvement of Rho GTPases in modulating CSF1R‐mediated signaling including the PI3K/AKT, JAK/STAT, and NF‐κB pathways (Figure S7B) [41, 42]. We first validated the interaction between SIRT5 and RAC2 through reciprocal co‐immunoprecipitation (co‐IP) in RAW264.7 cells overexpressing HA‐tagged RAC2 and Myc‐tagged SIRT5 (Figure 6B), and further confirmed endogenous SIRT5–RAC2 binding in mouse ovarian tissue (Figure 6C). Notably, SIRT5 deficiency led to marked hyper‐succinylation of RAC2 in both cultured RAW264.7 cells and negatively selected primary monocytes (Figure 6D). These findings identify RAC2 as a previously unrecognized succinylation substrate of SIRT5. To determine whether RAC2 succinylation also occurs in the ovary, IP was performed using an antibody against endogenous RAC2 in mouse ovarian lysates, revealing succinylation signals at the expected molecular weight of RAC2 (Figure 6E).
*Ubiquitination‐mediated degradation of RAC2 impairs macrophage differentiation in the Sirt5‐deficient ovarian environment. (A) Identification of SIRT5‐interacting proteins in M‐CSF‐stimulated monocytes by immunoprecipitation (IP) followed by mass spectrometry (MS), showing the number of peptides detected for interacting proteins. (B) Reciprocal co‐IP in RAW264.7 cells co‐expressing HA‐tagged RAC2 and Myc‐tagged SIRT5 showing their direct interaction. (C) Endogenous interaction between SIRT5 and RAC2 confirmed by reciprocal co‐IP in mouse ovarian tissues. (D) SIRT5 deficiency increases RAC2 succinylation, as shown by elevated succinylation levels in SIRT5‐knockdown RAW264.7 cells (left) and Sirt5 −/− monocytes (right). (E) RAC2 was immunoprecipitated from mouse ovarian lysates, followed by a western blot to assess lysine succinylation. (F) Western blot analysis showing decreased RAC2 protein levels and increased global succinylation in RAW264.7 cells (left) with pharmacological SIRT5 inhibition (MC3482) or in Sirt5 −/− monocytes (right) compared to controls. (G) Co‐IF staining for RAC2 (red) and F4/80 (green) in ovarian sections reveals diminished RAC2 expression in F4/80⁺ macrophages from Sirt5 −/− mice compared to wild‐type controls. (H) RAC2 ubiquitination is elevated under SIRT5‐deficient conditions, as shown by increased ubiquitination in SIRT5‐knockdown or MC3482‐treated RAW264.7 cells. (I) Both ubiquitination and succinylation of RAC2 are elevated in monocytes from Sirt5 −/− mice compared with those from wild‐type controls. (J,K) Impaired monocyte‐to‐macrophage differentiation upon RAC2 inhibition. (J) qRT–PCR analysis shows reduced expression of macrophage markers (Adgre1, Fcgr1, Cd68) in negatively selected monocytes following NSC23766 (RAC2 inhibitor) treatment (n = 3 per group; one‐way ANOVA with Tukey's multiple comparisons test). Data are normalized to β‐actin and presented as Mean ± SD. **p < 0.01; ***p < 0.0001. (K) IF staining for F4/80 (green) confirms impaired macrophage differentiation in the NSC23766‐treated group. Scale bar: 20 µm.
Intriguingly, depletion of SIRT5 or inhibition of its desuccinylase activity resulted in a pronounced reduction of RAC2 protein levels, without a corresponding decrease in Rac2 mRNA expression, in either *Sirt5^−/−^
- monocytes or RAW264.7 cells treated with MC3482, a selective small‐molecule inhibitor of SIRT5 (Figure 6F and Figure S7C) [43]. Consistently, Sirt5^−/−^ ovarian macrophages also exhibited diminished RAC2 protein levels, whereas RNA sequencing revealed comparable Rac2 transcript levels between *Sirt5^−/−^
- and wild‐type ovarian macrophages (Figure 6G and Figure S7D). This disconnect between transcript and protein abundance suggests that SIRT5 may regulate RAC2 stability through a post‐translational mechanism. Supporting this notion, increased ubiquitination of RAC2 was detected in both *Sirt5^−/−^
- monocytes and RAW264.7 cells treated with either MC3482 or Sirt5‐targeting siRNA. In those cells, RAC2 was hyper‐succinylated, suggesting that succinylation promotes its proteasomal degradation in the absence of SIRT5, thereby impairing RAC2 stability (Figure 6H,I).
We next investigated whether the reduction in RAC2 directly compromises monocyte‐to‐macrophage differentiation. Pharmacological inhibition of RAC2 using NSC23766 phenocopied the effects of SIRT5 loss by impairing CSF1‐induced macrophage differentiation (Figure 2E,F and Figure S4E,F) [44]. This was evidenced by reduced expression of canonical macrophage markers (Adgre1, Fcgr1, and Cd68) and a lower proportion of F4/80⁺ cells in an in vitro differentiation assay (Figure 6J,K). NSC23766 treatment suppressed the phosphorylation of key CSF1R downstream effectors, including JAK2 and AKT, linking RAC2 activity to the activation of the CSF1R signaling cascade (Figure S7E,F). Together, these findings identify RAC2 as a previously unrecognized substrate of SIRT5 and reveal that SIRT5‐mediated desuccinylation stabilizes RAC2 protein to ensure effective transduction of CSF1 signaling during monocyte‐to‐macrophage differentiation.
In addition to the suppression of canonical JAK/STAT signaling, phosphorylation of p21‐activated kinase (PAK), another key downstream effector of RAC2, was also markedly reduced in monocytes lacking SIRT5 or treated with NSC23766 during CSF1‐induced macrophage differentiation (Figure 2G and Figure S7E–G) [45]. Given the established role of PAK signaling in driving M1 macrophage polarization, we next investigated whether the SIRT5‐RAC2 axis contributes to the reduced infiltration of M1‐like macrophages observed in SIRT5‐deficient ovaries (Figure 3F) [46]. To functionally test this, we employed an LPS‐induced activation model in RAW264.7 cells. Disruption of RAC2 function, either through siRNA‐mediated knockdown or pharmacological inhibition with NSC23766, significantly blunted LPS‐induced upregulation of inducible nitric oxide synthase (iNOS) at both the transcript and protein levels (Figure S8A,C–E,G). Notably, RAC2 inhibition also impaired activation of the NF‐κB signaling pathway, a central regulator of M1 polarization, as indicated by reduced phosphorylation of p65 and decreased expression of the downstream cytokines TNF‐α and IL‐1β (Figure S8A–H) [47, 48]. These in vitro findings are consistent with our in vivo observations that SIRT5 deficiency suppresses TNF‐α and IL‐1β production in the CTX‐induced POI model (Figure 3G–I). Collectively, these findings demonstrate a dual role for RAC2 in driving both monocyte‐to‐macrophage differentiation and M1 polarization, positioning RAC2 as a critical molecular integrator downstream of SIRT5 in shaping the ovarian microenvironment.
Pharmacological Inhibition of SIRT5 by MC3482 Alleviates Ovarian Damage in the CTX‐Induced POI Model
2.8
Building upon our earlier findings that dysregulated succinylation gives rise to a pro‐inflammatory ovarian microenvironment that aggravates CTX‐induced tissue injury, we next evaluated whether pharmacological inhibition of SIRT5 could confer therapeutic benefit. For the treatment paradigm, female mice received MC3482 (5 mg/kg, i.p.) every three days during the first week. In weeks two and three, MC3482 administration continued and POI was induced by concurrent CTX administration, while the control group received vehicle dosing and underwent the same CTX‐induced POI protocol (Figure 7A). Morphological assessment and MVH immunostaining revealed markedly improved ovarian architecture and preservation of the follicle pool in MC3482‐treated mice compared to vehicle controls, as demonstrated by increased numbers of resting and growing follicles, along with reduced follicular atresia (Figure 7B–E). Gene expression analysis revealed restoration of key regulators of folliculogenesis and steroid hormone biosynthesis, accompanied by reduced expression of genes involved in apoptosis and oxidative stress in MC3482‐treated ovaries (Figure S9). Consistent with these molecular changes, MC3482 treatment led to a significant increase in Ki67⁺ granulosa cells within growing follicles, indicative of enhanced proliferative activity (Figure 7F). This was accompanied by a marked reduction in cleaved Caspase‐3–positive granulosa cells (Figure 7G) and further corroborated by TUNEL assay, both of which confirmed decreased granulosa cell apoptosis in MC3482‐treated ovaries (Figure 7H). We next examined the impact of MC3482 treatment on ovarian immune cell infiltration and inflammatory cytokine expression. MC3482 administration markedly suppressed CTX‐induced expansion of the ovarian macrophage population (Figure 7I). Consistently, IF staining, ELISA, and qRT–PCR analyses showed marked reductions in TNF‐α and IL‐1β at both protein and mRNA levels in MC3482‐treated ovaries compared with vehicle controls (Figure 7J–L), supporting an anti‐inflammatory effect of MC3482 treatment within the ovary. These findings demonstrate that pharmacological inhibition of SIRT5 with MC3482 mitigates CTX‐induced ovarian injury by restraining macrophage‐driven inflammation and limiting granulosa cell apoptosis, providing proof‐of‐concept that targeting SIRT5 offers a viable immunomodulatory strategy for protecting the ovary from CTX‐associated damage.
*Pharmacological inhibition of SIRT5 with MC3482 mitigates CTX‐induced ovarian injury. (A) Schematic of experimental timeline illustrating administration of MC3482 (5 mg/kg) or vehicle in a CTX (50 mg/kg)‐induced POI model. (B,C) Representative ovarian morphology (B) and H&E staining (C) of ovarian tissues from control, POI + vehicle, and POI + MC3482 groups. (D,E) IF staining for MVH (green, D) and quantification of follicle numbers (E, n = 4 per group) in ovaries from the specified groups. Scale bar: 100 µm. Primordial follicle (Pdf); growing follicles include primary follicles (PrF), secondary follicles (SeF), and antral follicles (AnF). (F,G) Representative IHC staining for Ki67 (F, scale bar: 100 µm) and IF staining for cleaved Caspase‐3 (G, scale bar: 50 µm) in ovarian sections from the indicated groups. (H) TUNEL assay evaluating granulosa cell apoptosis in ovarian tissues from the indicated groups. Scale bars: 50 µm. (I) IF staining for macrophages (F4/80⁺ cells) in ovarian tissues from control, POI + vehicle, and POI + MC3482 groups. Scale bar: 50 µm. (J,L) IF staining (J) and qRT‐PCR (L, n = 3 per group) showing expression and transcription of the inflammatory cytokines TNF‐α and IL‐1β in ovarian tissues from the indicated groups. Scale bar: 50 µm. (K) ELISA quantification of TNF‐α and IL‐1β levels in ovarian tissue homogenates and serum from the indicated groups (n = 5 per group). Data are presented as Mean ± SD. Statistical significance was assessed using one‐way ANOVA with Tukey's multiple comparisons test (E,K,L). **p < 0.01; ***p < 0.001; ***p < 0.0001.
Discussion
3
In this study, we investigated how macrophages contribute to inflammatory remodeling of the ovarian immune microenvironment in an alkylating agent‐induced POI mouse model. Consistent with prior reports that alkylating agents induce ovarian infiltration by macrophages, T lymphocytes, and neutrophils in patients, our CTX‐induced POI model similarly exhibited prominent immune cell infiltration, with macrophage expansion as the dominant feature [49]. The mechanisms driving macrophage infiltration into the ovarian cortex following tissue injury remain poorly understood. Our findings indicate that the increased macrophage presence in CTX‐treated ovaries may be driven by accelerated differentiation of Ly6C⁺ monocytes, as demonstrated by a pronounced expansion of Ly6C⁺F4/80⁺ monocyte‐derived macrophages within the ovarian microenvironment. This pattern resembles observations in models of kidney injury, where Ly6C⁺ monocytes are actively recruited to inflamed tissue and differentiate into F4/80⁺ macrophages in response to injury signals [50]. Previous studies have shown that infiltrating macrophages display substantial plasticity, adopting either pro‐inflammatory or tissue‐reparative states depending on local cues [51]. Under physiological conditions, M1‐like macrophages promote early follicular activation via PI3K/AKT/mTOR signaling, and their depletion impairs oocyte maturation and follicle development [52, 53]. In contrast, our POI model revealed a pathological shift: increased ovarian M0 macrophage accumulation is accompanied by enhanced M0‐to‐M1 polarization, and the functional role of newly differentiated M1 macrophages shifts toward amplifying ovarian injury. Limiting M1 macrophage infiltration preserved follicle number and integrity, coinciding with restored expression of folliculogenesis‐ and steroidogenesis‐related genes. Mechanistically, we identified that excessive production of TNF‐α and IL‐1β is functionally required for this pathological shift, directly driving granulosa cell apoptosis. While the epigenomic basis for this macrophage reprogramming under pathological stress remains to be elucidated, our findings suggest that restraining excessive monocyte‐to‐macrophage transition may offer a strategy to restrain innate immune‐driven ovarian damage.
After infiltrating injured tissues, classical monocytes (e.g., Ly6C⁺ in mice) encounter local signals such as GM‐CSF, IL‐4, and IL‐10 that initiate transcriptional programs driving their differentiation into tissue resident monocyte‐derived macrophages (moMACs). This process is regulated by lineage‐specific signaling pathways and varies substantially across species and tissue contexts [54]. For example, TGF‐β receptor signaling promotes tolerogenic macrophages in the colon, while ASGR1–NF‐κB/ATF5 signaling supports macrophage differentiation during hepatic sepsis [55, 56]. Despite these insights, the regulatory mechanisms orchestrating macrophage differentiation within the ovarian microenvironment remain poorly defined under both homeostatic and stress conditions. While most studies have focused on macrophage dynamics in ovarian cancer, little is known about their roles in non‐malignant ovarian physiology [57]. Here, we identify SIRT5 as a previously unrecognized regulator of ovarian moMAC differentiation, promoting this process by stabilizing RAC2 through its desuccinylase activity and sustaining CSF1R signaling. This axis is essential for the differentiation, maintenance, and activation of ovarian macrophages under both homeostatic and chemotherapy‐induced stress. Pharmacological depletion of macrophages and blockade of CSF1R, as well as genetic disruption of the SIRT5–RAC2 pathway, impaired macrophage differentiation and M1‐like polarization. These findings align with prior studies showing that CSF1 deficiency leads to ovarian dysfunction marked by prolonged estrous cycles and reduced ovulation [58]. We propose that under CTX‐induced stress, SIRT5 is essential for CSF1–CSF1R–driven expansion of the M0 macrophage pool, which primes subsequent M1 polarization. This underscores the pivotal role of the SIRT5–RAC2 axis in moMAC differentiation and ovarian immune regulation. Whether this mechanism extends to other tissues where macrophage identity is critically shaped by distinct local cues remains to be determined.
Sirtuins (SIRTs) are increasingly recognized as critical modulators of both innate and adaptive immune responses. For example, SIRT1 and SIRT2 suppress inflammatory signaling by deacetylating NF‐κB, leading to reduced transcription of pro‐inflammatory cytokines such as IL‐6, IL‐8, and TNF‐α [59, 60]. Findings on the role of SIRT5 in cellular homeostasis and immune responses have been conflicting and appear to vary across different pathological models [18, 20]. SIRT5‐mediated regulation of macrophage differentiation and activation is context dependent, shaped by tissue‐specific environments and pathological conditions. In murine oncogene‐driven HCC, SIRT5 loss increased hepatocyte‐derived taurocholic acid (TCA), promoting M2‐like macrophage infiltration in the liver [61]. Interestingly, we observed that under homeostatic conditions, SIRT5 deficiency reduced ovarian macrophage infiltration, affecting both M1‐ and M2‐like subsets. Upon CTX‐induced stress, this reduction was more pronounced for M1 macrophages, suggesting that SIRT5 not only supports moMAC differentiation but also promotes M1 polarization in inflammatory settings. These findings highlight the context‐dependent role of SIRT5 in immune regulation and suggest that ovarian immune networks may be reprogrammed under chemotoxic stress.
Mechanistically, we identify RAC2 as a novel substrate of SIRT5. Desuccinylation of RAC2 by SIRT5 enhances its stability, thereby promoting both monocyte‐to‐macrophage differentiation and M1 polarization. Our findings establish the SIRT5–RAC2 axis as a key driver of inflammation that amplifies CTX‐induced ovarian injury. Disruption of this axis, either through SIRT5 deficiency or inhibition of RAC2, effectively mitigates immune dysregulation in the ovary. Rac1 and Rac2 often operate through overlapping downstream signaling pathways, engaging common effectors such as PAK kinases and NF‐κB to regulate diverse cellular processes including migration, transcriptional activity, and immune cell activation [62, 63]. In necrotizing enterocolitis, RAC1 has been implicated in the early differentiation of neonatal monocyte‐derived macrophages and the production of pro‐inflammatory cytokines, contributing to a highly inflammatory intestinal milieu [64]. Consistent with these findings, our study reveals that RAC2 stability and activation are essential for promoting moMAC differentiation and M1 polarization in ovarian monocytes. We identify lysine succinylation as a previously unrecognized modification that destabilizes RAC2, promoting its ubiquitin‐mediated degradation. This highlights succinylation as a regulatory signal marking RAC2 for proteasomal turnover. This finding aligns with a recent report showing that succinylation of SERCA2a promotes its K48‐linked ubiquitination and degradation during sepsis‐induced cardiac dysfunction [65]. Although additional SIRT5‐associated proteins identified through IP‐MS may also influence macrophage differentiation via lysine succinylation‐dependent or ‐independent mechanisms, our current findings suggest that targeting protein succinylation represents a promising strategy for modulating inflammation in the context of ovarian inflammation and macrophage‐driven tissue remodeling. Together, these results highlight lysine succinylation as a novel regulatory PTM controlling Rho GTPase stability, and suggest that targeting succinylation of key immune modulators may represent a promising therapeutic strategy for inflammatory diseases.
Previous studies have shown that granulosa cell apoptosis in CTX‐induced POI involves multiple pathways: alkylating agents trigger p53‐mediated apoptosis via DNA damage and oxidative stress‐induced mitochondrial dysfunction, while also elevating pro‐inflammatory cytokines (e.g., TNF‐α, IL‐1β, IL‐8, and IL‐6), further amplifying cell death [35, 38, 66]. However, whether these cytokines directly trigger granulosa cell apoptosis, the specific intracellular signaling events involved, and how these intersect with canonical apoptotic pathways to amplify stress‐induced cytotoxicity remain poorly defined. Herein, we were able to delineate the specific contribution of TNF‐α and IL‐1β to amplifying CTX‐induced granulosa cell death by defining a unique inflammatory gene set, termed the Cyto‐POI gene set, that enriched for signaling pathways associated with inflammation‐driven apoptosis, including p53 signaling, NOD‐like receptor activation, and necroptosis [38]. Within this gene signature, Caspase‐1, ‐7, and ‐10 emerged as candidate mediators of granulosa cell apoptosis under inflammatory conditions, based on their enrichment within the Cyto‐POI profile and their concomitant upregulation alongside elevated pro‐inflammatory cytokines in the POI model. Consistent with our findings, prior studies have implicated the TRAF2–NF‐κB–Caspase‐1 signaling axis in Benzo(a)pyrene‐induced granulosa cell apoptosis under oxidative stress [67]. We propose that multiple caspases act synergistically to prime granulosa cells for apoptosis when exposed to simultaneous inflammatory and genotoxic stress.
Therapeutic strategies for POI are diverse, spanning hormone replacement therapy, fertility preservation techniques, pharmacological agents, and stem cell‐based approaches [36]. Pharmacologic interventions, including antioxidants and senolytics, show promise in mitigating follicular damage but remain largely experimental [68]. Human amniotic mesenchymal stem cells (hAD‐MSCs) have also been shown to attenuate CTX‐induced ovarian inflammation in rats by downregulating key pro‐inflammatory cytokines such as IL‐1β, IL‐6, and TNF‐α, preserving ovarian tissue integrity and mitigating POI [69]. Our findings provide initial evidence that inhibiting the SIRT5–RAC2 axis may offer a promising strategy to modulate the ovarian microenvironment, particularly its immune and stromal components, to reduce inflammation‐related tissue damage and preserve granulosa cell integrity and follicular reserves. Notably, this protective effect is achieved through limiting macrophage accumulation and activation, thereby reducing the primary cellular source of pro‐inflammatory cytokines within the ovarian microenvironment. Given that cross‐species scRNA‐seq analyses demonstrate conserved high expression of SIRT5 in myeloid cells and macrophages from both human and mouse ovaries, modulation of the SIRT5–RAC2 axis presents a clinically relevant proof‐of‐concept strategy to preserve ovarian function in women undergoing gonadotoxic therapies or in POI patients with elevated inflammatory cytokine levels. However, this approach will necessitate both the development of novel SIRT5 inhibitors and rigorous preclinical evaluation of their target selectivity, pharmacokinetic profiles, and long‐term safety to assess their therapeutic viability.
This study has several limitations. First, our findings are based on a CTX‐induced model that primarily recapitulates chemotherapy‐associated POI. Other major etiologies of human POI, such as genetic or autoimmune causes, may involve distinct immune mechanisms. Defining the immune landscape across diverse POI models will be necessary to determine whether SIRT5–RAC2 signaling constitutes a shared pathogenic axis. Second, although SIRT5 is highly enriched in ovarian macrophages and minimally expressed in other ovarian cell types, including granulosa cells, this low‐level expression does not entirely preclude potential cell‐intrinsic roles. Future studies employing myeloid‐specific Sirt5 conditional knockout models (e.g., Ccr2‐Cre; Sirt5 ^flox/flox^) will be critical to definitively determine the macrophage‐dependent function of SIRT5 in mediating follicular damage under inflammatory stress. Third, while our data suggest that impaired monocyte‐to‐macrophage differentiation underlies reduced macrophage accumulation in SIRT5‐deficient ovaries, we cannot exclude effects on earlier recruitment steps such as monocyte extravasation or context‐dependent lineage commitment. Fourth, ovarian macrophages include both embryonically derived resident and monocyte‐derived subsets. The distinct contributions of SIRT5–RAC2 signaling in these distinct lineages remain to be delineated, ideally using lineage‐tracing approaches.
Methods
4
Animal Models and Induction of Premature Ovarian Insufficiency (POI)
4.1
Female C57BL/6J mice (6–8 weeks old) were obtained from Shanghai Jiesijie Laboratory Animal Co., Ltd. (Shanghai, China; Approval No. SCXK (Hu) 2023‐0004) and housed under standard conditions. Littermate controls were used to reduce genetic background variability and enhance experimental reproducibility. POI was induced in 8‐week‐old mice by daily intraperitoneal injection of cyclophosphamide (CTX; Sigma, USA) at a dose of 50 mg/kg for 15 consecutive days. Control animals received equal volumes of sterile saline following the same injection schedule. At the experimental endpoint, mice were euthanized via CO_2_ asphyxiation, and ovaries were excised for downstream histological and molecular analyses.
MC3482 Pretreatment for Attenuation of CTX‐Induced Ovarian Toxicity
4.2
To investigate potential strategies for mitigating CTX‐induced ovarian damage, the SIRT5 inhibitor MC3482 (MedChemExpress, China) was used to suppress lysine succinylation in ovarian tissue. Female C57BL/6J mice (8 weeks old) were randomly assigned to three groups (n = 6 per group): MC3482 treatment, vehicle control, and a mock (non‐CTX) control group. MC3482 was dissolved in a vehicle containing 10% dimethyl sulfoxide (DMSO), 40% polyethylene glycol 300 (PEG300), 5% Tween‐80, and 45% saline. Mice in the treatment group received intraperitoneal (i.p.) injections of MC3482 at 5 mg/kg (200 µL per dose), three times per week for one week. Vehicle control and mock groups received the same volume of vehicle on an identical schedule. Following the one‐week pretreatment, mice in the MC3482 and vehicle control groups underwent the POI induction protocol, consisting of daily i.p. injections of CTX (50 mg/kg) for 15 consecutive days, as previously described. MC3482 or vehicle administration continued throughout the CTX treatment period. Mice in the mock (non‐CTX) group received sterile saline (i.p.) on the same schedule.
Hematoxylin and Eosin (H&E) Staining
4.3
Ovarian tissues were fixed in 4% paraformaldehyde (PFA) at 4°C for 24 h, dehydrated through a graded ethanol series (30%, 50%, 75%, 95%, and 100%), and cleared in xylene. Samples were embedded in paraffin and sectioned at a thickness of 5 µm using a Leica RM2126 microtome (Leica, Germany). Tissue sections were stained using a hematoxylin and eosin staining kit (Solarbio, China) according to the manufacturer's instructions. Following staining, sections were dehydrated, cleared in xylene, and mounted with neutral resin for imaging using a Zeiss Axio Scan 7 slide scanner (Carl Zeiss AG, Germany).
TUNEL Assay
4.4
Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining was performed using a commercial apoptosis detection kit (Promega, USA) according to the manufacturer's protocol. Ovarian cryosections were fixed with 4% PFA at 4°C for 10 min and washed twice with PBS. After permeabilization with 0.2% Triton X‐100 for 20 min at room temperature and two additional PBS washes, 100 µL of equilibration buffer was applied to each section and incubated for 10 min at room temperature. The TdT reaction mixture was then added to the sections, which were covered with plastic coverslips and incubated at 37°C for 60 min in the dark. The reaction was stopped by washing with 2× SSC buffer for 15 min, followed by three washes with PBS. Nuclei were counterstained with DAPI, and sections were mounted. Fluorescent images were acquired using a Zeiss LSM 880 confocal microscope (Carl Zeiss AG, Germany).
Cell Culture
4.5
Raw264.7 cells (RRID: CVCL_0493) were obtained from the American Type Culture Collection (ATCC; USA) with catalog number ATCC TIB‐71 and purchased in October 2023. KGN cells (RRID: CVCL_0375) were acquired from the National Collection of Authenticated Cell Cultures (NCACC; China) under catalog number SCSP‐5495 and purchased in December 2023. RAW264.7 cells were cultured in Dulbecco's modified Eagle medium (DMEM), while KGN cells were maintained in DMEM/F12 (1:1) medium, both supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA). All cells were maintained at 37°C in a humidified incubator with 5% CO_2_. All cell lines used in this study were verified for identity and purity. Additionally, routine mycoplasma contamination testing was performed using the MycoAlert Mycoplasma Detection Kit (Lonza, Switzerland) prior to experimentation, with all cells testing negative for mycoplasma.
Apoptosis Detection and Cell Viability in KGN Cells
4.6
To model the cytokine‐sensitized apoptotic response observed in vivo, a Transwell co‐culture system was established with monocyte‐derived M1 macrophages seeded in the upper chamber and KGN granulosa cells in the lower chamber. Two macrophage‐to‐KGN cell ratios were used to mimic distinct inflammatory environments: the “macrophage‐high” condition (1 × 10⁵ macrophages + 2 × 10⁵ KGN cells) representing CTX‐treated wild‐type ovaries, and the “macrophage‐low” condition (1 × 10⁴ macrophages + 2 × 10⁵ KGN cells) representing Sirt5‐deficient ovaries. Dose–response assays with 0–25 µm 4‐HC revealed that 6.25 µm selectively induced apoptosis in the macrophage‐high group, with minimal effect in the macrophage‐low group, closely mirroring the in vivo differential response. Based on these findings, KGN cells were divided into four groups: (1) control (DMSO); (2) CTX‐only (6.25 µm 4‐HC); (3) cytokines‐only (10 ng/mL IL‐1β and 40 ng/mL TNF‐α) [70, 71, 72, 73]; and (4) CTX + cytokines (6.25 µm 4‐HC combined with 10 ng/mL IL‐1β and 40 ng/mL TNF‐α). After 36 h of treatment, cells were harvested using Accutase Enzyme, washed twice with ice‐cold PBS, and resuspended in FACS binding buffer (HBSS supplemented with 2% FBS and 2 mm EDTA). Apoptosis was assessed using the Annexin V‐APC/7‐AAD Apoptosis Detection Kit (Multi Sciences, China) following the manufacturer's protocol. 5 × 10^5^ Cells were stained with 5 µL Annexin V‐APC and 10 µL 7‐AAD, then analyzed on a CytoFLEX flow cytometer (Beckman Coulter, USA). Data were processed using FlowJo software. Cell viability was measured using the MTS‐based CellTiter AQueous One Solution assay (Promega, USA). MTS reagent (20 µL) was added to each well containing 100 µL culture medium and incubated at 37°C. Absorbance at 490 nm was recorded using a microplate reader (BioTek, USA). Wells without cells served as blanks.
Plasmid Construction and Transfection
4.7
The pcDNA3.1(+)‐Sirt5‐Myc and pcDNA3.1(+)‐Rac2‐HA expression plasmids were constructed by inserting the full‐length coding sequences (CDS) of human SIRT5 and RAC2 into the pcDNA3.1(+) vector with C‐terminal Myc or HA tags, respectively. Corresponding empty vectors, pcDNA3.1(+)‐Myc and pcDNA3.1(+)‐HA, were used as controls. For Co‐IP assays, Raw264.7 cells were transfected with 2 µg of expression plasmids using Lipofectamine 3000 (Thermo Fisher, USA). Cells were harvested 48 h post‐transfection for validation of the SIRT5–RAC2 interaction. To assess the effect of SIRT5 knockdown on RAC2 succinylation, Raw264.7 cells were transfected with Sirt5‐targeting siRNA or Ctrl RNA using RNAiMAX (Thermo Fisher, USA). Twelve hours later, cells were transfected with 1 µg of pcDNA3.1(+)‐Rac2‐HA. Lysates were collected 48 h after siRNA transfection for succinylation analysis. For inhibition of RAC2 activity, cells were either treated with DMSO or NSC23766 (100 µm; MedChemExpress, China), or transfected with RAC2 siRNA using RNAiMAX. After 48 h, cells were stimulated with LPS (10 µm; Sigma, USA) for 6 h to induce M1 differentiation, and samples were collected for western blot (WB) and qRT‐PCR analysis.
Generation of SIRT5 Knockdown (SIRT5 KD) KGN Cells
4.8
Three stable SIRT5 KD KGN cell lines were generated via lentiviral transduction. Short hairpin RNA (shRNA) sequences targeting human SIRT5 (listed in Table S2) were cloned into the pLKO.1‐puro vector. High‐titer lentiviral stocks were produced by co‐transfecting HEK293T cells with the shRNA construct and packaging plasmids (psPAX2 and pMD2.G), followed by ultracentrifugation‐based concentration of the harvested viral supernatants. KGN cells were infected with the concentrated virus in the presence of polybrene (8 µg/mL) and selected with puromycin (1.5 µg/mL) for 5 days. Knockdown efficiency was confirmed by Western blot.
Immunohistochemistry (IHC) Staining
4.9
IHC Staining was performed as previously described [74]. Paraffin‐embedded ovarian tissue sections (5 µm) were deparaffinized in xylene (twice for 10 min each) and rehydrated through a graded ethanol series (100% to 30%) followed by distilled water. Antigen retrieval was performed by heating sections in citrate buffer (pH 6.0) using a microwave oven for 10 min. After cooling, sections were blocked with 5% goat serum/0.3% Triton X‐100 in PBS for 1 h, then incubated with rabbit anti‐Ki67 primary antibody (1:800; Abcam, ab16667) overnight at 4°C. Following five washes with PBS, slides were probed with HRP‐conjugated goat anti‐rabbit secondary antibody (1:1000; Jackson ImmunoResearch) for 1 h. Signal was visualized using a DAB chromogenic kit (Maixin Biotech), counterstained with hematoxylin, and processed for dehydration/clearing before mounting. Images were acquired using a Zeiss Axio Scan 7 slide scanner (Carl Zeiss AG, Germany).
Immunofluorescence (IF) Staining
4.10
Ovarian tissues were fixed in 4% PFA for 1 h at 4°C, followed by immersion in 30% sucrose until fully equilibrated, embedded in OCT compound (Sakura Finetek, Japan), and cryosectioned at 5 µm thickness using a cryostat (Leica CM1950). For in vitro staining, cells were seeded onto sterile 12‐mm glass coverslips in 24‐well plates and fixed with 4% PFA for 15 min at room temperature, followed by three washes with PBS. Both cryosections and coverslips were permeabilized with 0.3% Triton X‐100 for 20 min, blocked with 5% goat serum for 1 h, and incubated with primary antibodies (Table S1) overnight at 4°C. After five washes with PBS, Alexa Fluor‐conjugated secondary antibodies (1:1000; Invitrogen, USA) were applied for 2 h at 4°C. Nuclei were counterstained with DAPI, and slides were mounted with antifade medium (Invitrogen, USA). Images were captured using a Zeiss LSM 880 Confocal laser scanning microscope.
SA‐β‐Gal Staining
4.11
To assess cellular senescence in the ovarian microenvironment, cryosections of frozen ovaries were processed using an in situ β‐galactosidase staining kit (Beyotime Biotechnology, China) according to the manufacturer's instructions. Briefly, sections were fixed with kit‐provided fixative for 10 min at room temperature, washed three times with PBS, and incubated with working solution at 37°C for 4 h. Following final PBS washes, sections were mounted, and bright‐field images were captured using a Zeiss Axio Scan.Z1 scanner.
Flow Cytometry Analysis and Macrophage Sorting
4.12
Ovarian tissues from 8‐week‐old female mice were harvested, minced with sterile surgical blades, and enzymatically digested at 37°C for 30 min in HBSS containing 1 mg/mL Collagenase Type IV (Sigma, USA) and 1 U/mL DNase I (Thermo, USA). The digestion was quenched by the addition of 5 mm EDTA, and the resulting suspension was filtered through a 70 µm cell strainer. Red blood cells were lysed using RBC lysis buffer (BioLegend, China), and the remaining cells were centrifuged and resuspended in FACS buffer (HBSS supplemented with 2% FBS and 2 mm EDTA). To block Fc receptor‐mediated non‐specific binding, single‐cell suspensions were incubated with TruStain FcX (anti‐mouse CD16/32; BioLegend, China) for 10 min at 4°C. For immune cell infiltration analysis, single‐cell suspensions were prepared from ovaries of control, or CTX‐induced POI mice and stained with a panel of fluorophore‐conjugated antibodies against CD45.2, F4/80, CD11b, CD11c, CD138, B220, CD3, and MHC class II (MHC‐II), as detailed in Table S1. Viability Dye eFluor 780 (Thermo Fisher) was used to exclude dead cells. For the isolation of total macrophages and quantification of M1 and M2 subsets, single‐cell suspensions from *Sirt5^+/+^
- and *Sirt5^−/−^
- mice were stained with antibodies against CD45.2, F4/80, CD11b, CD86, and CD206. Dead cells were excluded using DAPI (1 µg/mL; Thermo Fisher). Single‐color controls were included for compensation. Live CD11b⁺F4/80⁺ macrophages were sorted using a BD FACSAria Fusion flow cytometer (BD Biosciences, USA) and collected for downstream molecular and functional assays.
RNA Isolation and qRT–PCR Analysis
4.13
Total RNA was extracted from cultured cells, ovarian tissues or sorted cells using the FastPure Complex Tissue/Cell Total RNA Isolation Kit (Vazyme, RC113) following the manufacturer's protocol. Fresh tissue or cells were homogenized in 600 µL Buffer SRL using a mechanical homogenizer (Servicebio, China) until fully lysed. RNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo, USA). Complementary DNA (cDNA) was synthesized from 1 µg of total RNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara, Japan). Quantitative real‐time PCR (qRT–PCR) was carried out using TB Green Premix Ex Taq II (Takara, Japan) on a QuantStudio 6 Flex Real‐Time PCR System (Thermo, USA) in triplicate. PCR cycling conditions were as follows: 95°C for 30 s, 40 cycles of 95°C for 5 s and 60°C for 34 s, followed by melting curve analysis. Relative gene expression levels in Figures 1A, 1G, 2B, 2E, 3, 4, 4, 6, S1C, S1E, S3F, S4E, S9A, and S9D were calculated using the 2^−^ ^ΔCt^ method and normalized to β‐actin. Primer sequences are listed in Table S2.
Isolation and Differentiation of Bone Marrow–Derived Macrophages (BMDMs)
4.14
Bone marrow cells were isolated from femurs of Sirt5 ^+/+^ and Sirt5 ^−/−^ mice by flushing with Dulbecco's Modified Eagle Medium (DMEM; Gibco) supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin–streptomycin (Gibco, USA) using a 26G needle. The cell suspension was filtered through a 70‐µm cell strainer (WHB, China) to remove debris, and red blood cells were lysed with ACK lysis buffer (Thermo, USA) for 2 min at room temperature. Following two washes with PBS by centrifugation, cells were plated into T75 flasks and incubated at 37°C for 5 h to allow adherence of stromal cells. Non‐adherent cells were collected, counted, and seeded into 12‐well plates in DMEM containing 25 ng/mL recombinant mouse M‐CSF (PeproTech, USA). Half of the medium was replaced every 48 h. After 7 days of culture, adherent differentiated BMDMs were collected for downstream analyses.
Monocyte Isolation and Differentiation
4.15
Monocytes were enriched via negative selection from mouse bone marrow cell suspensions using the MojoSort Mouse Monocyte Isolation Kit (BioLegend, China). Briefly, single‐cell suspensions from freshly harvested murine femurs were adjusted to 1 × 10⁷ cells in 100 µL FACS buffer and incubated with the Biotin–Antibody Cocktail at 4°C for 15 min with gentle agitation. Subsequently, 10 µL of Streptavidin Nanobeads was added, and the samples were incubated for an additional 15 min at 4°C. The reaction volume was then adjusted to 1 mL with FACS buffer before magnetic separation using a MojoSort magnet. The unlabeled cells, representing the enriched monocyte population, were carefully collected for downstream applications. To induce differentiation, isolated monocytes were seeded in 24‐well plates and cultured in DMEM supplemented with 25 ng/mL M‐CSF. For RAC2 inhibition assays, monocytes were treated from day 0 of differentiation with either DMSO (vehicle control) or 100 µm NSC23766 (MedChemExpress, China) dissolved in DMSO. Cells were harvested on day 3 for RNA extraction or Western blot analysis.
Co‐immunoprecipitation (Co‐IP), Immunoprecipitation (IP), and Western Blot (WB)
4.16
For co‐IP assays, mouse ovarian tissues or RAW264.7 cells were lysed in co‐IP lysis buffer (Vazyme, China) supplemented with 1 × protease inhibitor cocktail (Roche, USA) for 30 min on ice according to the manufacturer's protocol. Lysates were clarified by centrifugation at 8000 × g for 10 min at 4°C, and supernatants were incubated overnight at 4°C with primary antibodies against HA, Myc, RAC2, or SIRT5 (Table S1). Protein A/G magnetic beads (Vazyme, China) were then added, and samples were rotated at room temperature for 1 h. Beads were washed three times with co‐IP lysis buffer, and bound proteins were eluted in SDS sample buffer and then subjected to western blot.
To assess RAC2 post‐translational modifications, cultured cells or negatively sorted mouse monocytes were lysed directly in RIPA buffer (150 mm NaCl, 50 mm Tris‐HCl, pH 7.4–7.5, 1% NP‐40, 0.5% sodium deoxycholate, 0.1% SDS) supplemented with 1 × protease inhibitor cocktail (Roche, USA). Cell lysis was incubated on ice for 30 min with occasional mixing to ensure complete disruption. Lysates were clarified by centrifugation at 8000 ×* g* for 10 min at 4°C, and the supernatants were incubated overnight at 4°C with 10 µg of anti‐RAC2 antibody (Proteintech, China) or normal mouse IgG (Proteintech, China) as a negative control. Immune complexes were captured using Protein A/G magnetic beads (Vazyme, China) for 2 h at 4°C, followed by three washes with RIPA buffer. Proteins were eluted by boiling in SDS sample buffer and analyzed by Western blot. For the detection of RAC2 ubiquitination, cells were pre‐treated with 10 µm MG132 (Sigma‐Aldrich, USA) for 4 h to inhibit proteasomal degradation. MG132 (10 µm) was also included in the RIPA lysis buffer during sample preparation.
For WB analysis, protein concentrations were determined using the BCA Protein Assay Kit (Beyotime Biotechnology, China). Equal amounts of protein were subjected to SDS–PAGE and transferred onto PVDF membranes using a rapid transfer system (Bio‐Rad, USA). Membranes were blocked with 2% BSA in TBST for 1 h at room temperature and incubated overnight at 4°C with primary antibodies (details and dilutions listed in Table S1). After washing with TBST and incubating with HRP‐conjugated secondary antibodies, protein bands were detected using a chemiluminescence detection system (Shanghai Tanon Technology, China).
Macrophage Depletion in Mouse Ovaries
4.17
Macrophage depletion was performed as previously described (schematic illustration provided in Figure S4C) [75]. To deplete ovarian macrophages, both Sirt5‐knockout and *Sirt5^+/+^
- mice were intraperitoneally injected with 0.5, 0.25, and 0.25 mg of a CSF1‐neutralizing antibody (clone 5A1; Bio X Cell, USA) on postnatal day 14 (P14), P18, and P22, respectively. Control mice received injections of an isotype control IgG antibody (clone HRPN; Bio X Cell, USA) on the same schedule. In parallel, 50 µL of clodronate liposomes (LIPOSOMA, Netherlands) were administered intravenously via the tail vein on P15, P19, and P23. Control groups were injected with control liposomes according to the same schedule. To assess the efficiency of macrophage depletion, ovaries were surgically collected on P25 for IF staining. To evaluate monocyte‐to‐macrophage differentiation, ovaries were collected on P44, fixed in 4% PFA, and processed for immunofluorescence analysis.
ELISA Detection
4.18
Mouse IL‐1β and TNF‐α levels in serum and ovarian tissues were quantified using Elabscience High‐Sensitivity ELISA Kits (Elabscience Biotechnology, China) according to the manufacturer's instructions. Serum was collected by allowing whole blood to clot at room temperature for 1 h, followed by centrifugation at 1000 × g for 20 min at 4°C. Ovarian tissues were homogenized in ice‐cold PBS containing protease inhibitors, sonicated, and centrifuged at 5000 × g for 10 min at 4°C. Protein concentrations in tissue lysates were determined using the BCA assay. For the ELISA, 100 µL of standards or diluted samples were added to pre‐coated microplate wells and incubated at 37°C for 90 min. After washing, 100 µL of biotinylated detection antibody was added and incubated for 60 min, followed by three washes. Next, 100 µL of HRP‐conjugated secondary antibody was added and incubated for 30 min, followed by five washes. Color development was initiated by adding 90 µL of TMB substrate and incubating for 15 min at 37°C in the dark. The reaction was terminated with 50 µL of stop solution. Absorbance was measured at 450 nm using a BioTek microplate reader (USA), and cytokine concentrations were interpolated from a four‐parameter logistic standard curve.
RNA Sequencing and Data Analysis
4.19
RNA from ovarian tissues, KGN cells, or sorted CD11b⁺F4/80⁺ macrophages was subjected to library construction and paired‐end sequencing by LC Sciences using an Illumina NovaSeq platform. Raw reads were evaluated for base quality, adapter contamination, and sequence duplication with FastQC v0.11.9. Adapter sequences and low‐quality bases (Phred score < 20) were removed using Trim Galore v0.6.7. Clean reads from ovarian tissue and sorted macrophage RNA were aligned to the mouse genome (GRCm39/mm39) with Hisat2 v2.2.1, whereas clean reads from KGN cells were mapped to the human genome (GRCh38/hg38) using the same tool. Gene counts were generated via featureCounts v2.0.1, and transcripts per million (TPM) values were calculated using Salmon v1.9.0. Differential expression analysis was performed in R v4.3.1 with DESeq2 v1.38.3, identifying significantly altered genes at an adjusted p‐value < 0.05 and |log_2_ fold change| ≥ 1. Functional enrichment of GO terms and KEGG pathways was conducted using clusterProfiler v4.8.0 with Benjamini–Hochberg correction for multiple testing (FDR < 0.05). To assess immune cell infiltration within ovarian tissues, transcriptome profiling data from control and SIRT5‐deficient ovaries were analyzed using ImmuCellAI‐mouse [26], a computational tool that estimates the relative abundance of immune cell subsets based on signature gene expression profiles.
Single‐Cell RNA Sequencing Data Analysis
4.20
For mouse ovarian single‐cell transcriptomic analysis, digital gene expression (DGE) matrices and cell‐type annotations were obtained from the Gallery section of the Mouse Cell Atlas 2.0 (MCA 2.0; https://bis.zju.edu.cn/MCA/). Gene expression was normalized to TPM by the database. Mean TPM values for Sirt5 were calculated across annotated cell clusters to assess its cell‐type‐specific expression.
For human ovarian data, annotated scRNA‐seq datasets were accessed via the CELLxGENE Discover platform (https://cellxgene.cziscience.com), specifically from the Human Cell Landscape and Tabula Sapiens ovary datasets. Gene expression was pre‐normalized as ln(CPTT + 1), where CPTT denotes counts per ten thousand. Average ln(CPTT + 1) values for SIRT5 were computed across cell clusters to evaluate its expression distribution in human ovaries.
Statistical Analysis
4.21
All statistical tests were performed using GraphPad Prism v10.5. Data were presented as mean ± SD from at least three biological replicates (n ≥ 3 per group). A two‐tailed unpaired Student's t‐test was used to compare differences between two groups. For comparisons involving more than two groups, a one‐way ANOVA followed by Tukey's multiple comparisons test was used. Statistical significance was defined as P < 0.05. The specific statistical tests used for each experiment are detailed in the corresponding figure legends. Pearson correlation coefficients were calculated pairwise across RNA‐sequencing samples using R, with all expressed genes retained for comparison. Correlation matrices were visualized using heatmaps to assess inter‐sample similarity.
Author Contributions
W.J.T.T., Y.Q.L., and S.N.L. contributed equally to this work. Supervision and funding acquisition were done by C.F.H. and M.J.X. Resources were provided by C.F.H., T.F.Z., and M.J.X. Conceptualization and study design were done by W.J.T.T., C.F.H., T.F.Z., and M.J.X. Experiments and data analysis were done by W.J.T.T., Y.Q.L., and S.N.L., with technical assistance and material support from M.J.W., Z.X.L., J.F.J., J.J.C., X.D.X., L.L., C.Q.L., F.Z., Y.L., and H.T.N. The original draft was written by W.J.T.T. with contributions from Y.Q.L. and S.N.L. Review and editing were done by C.F.H. and M.J.X. with manuscript discussions involving all authors.
Funding
This work was supported by the National Natural Science Foundation of China (grants 82471792 and 31770945 to C.F.H.), the National Key R&D Program of China (grant 2024YFA0918400 to C.F.H.).
Conflicts of Interest
The authors declare no conflicts of interest.
Ethics Statement
All animal experiments were approved by the Medical Research Ethics Committee of Naval Medical University (Approval No. SYXK (Hu) 2022‐0011) and conducted in strict accordance with institutional guidelines and regulations for the ethical use of animals in research.
Supporting information
Supporting File 1: advs73559‐sup‐0001‐SuppMat.docx.
Supporting File 2: advs73559‐sup‐0002‐Data.rar;.
Supporting File 3: advs73559‐sup‐0003‐TableS3.xlsx.
Supporting File 4: advs73559‐sup‐0004‐TableS4.xlsx.
Supporting File 5: advs73559‐sup‐0005‐TableS5.xlsx.
Supporting File 6: advs73559‐sup‐0006‐TableS6.xlsx.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1P. Touraine , N. Chabbert‐Buffet , G. Plu‐Bureau , L. Duranteau , A. H. Sinclair , and E. J. Tucker , “Premature Ovarian Insufficiency,” Nature Reviews Disease Primers 10 (2024): 63.10.1038/s 41572-024-00547-539266563 · doi ↗ · pubmed ↗
- 2X. Jiao , H. Ke , Y. Qin , and Z.‐J. Chen , “Molecular Genetics of Premature Ovarian Insufficiency,” Trends in Endocrinology and Metabolism 29 (2018): 795–807.30078697 10.1016/j.tem.2018.07.002 · doi ↗ · pubmed ↗
- 3Z. Nash and M. Davies , “Premature Ovarian Insufficiency,” BMJ 384 (2024): 077469.10.1136/bmj-2023-07746938508679 · doi ↗ · pubmed ↗
- 4L. Cacciottola , A. Camboni , and M. M. Dolmans , “Immune System Regulation of Physiological and Pathological Aspects of the Ovarian Follicle Pool throughout the Female Reproductive Lifespan,” Human Reproduction 40 (2025): 12–22.39607771 10.1093/humrep/deae 254 · doi ↗ · pubmed ↗
- 5M. Dai , Y. Xu , G. Gong , and Y. Zhang , “Roles of Immune Microenvironment in the Female Reproductive Maintenance and Regulation: Novel Insights Into the Crosstalk of Immune Cells,” Frontiers in Immunology 14 (2023): 1109122.38223507 10.3389/fimmu.2023.1109122 PMC 10786641 · doi ↗ · pubmed ↗
- 6S. Dragojević‐Dikić , D. Marisavljević , A. Mitrović , S. Dikić , T. Jovanović , and S. Janković‐Ražnatović , “An Immunological Insight Into Premature Ovarian Failure (POF),” Autoimmunity Reviews 9 (2010): 771–774.20601203 10.1016/j.autrev.2010.06.008 · doi ↗ · pubmed ↗
- 7J. Liu , X. Huang , X. Cao , X. Feng , and X. Wang , “Serum Biomarker Analysis in Patients With Premature Ovarian Insufficiency,” Cytokine 126 (2020): 154876.31629109 10.1016/j.cyto.2019.154876 · doi ↗ · pubmed ↗
- 8R. Wu , K. H. Van der Hoek , N. K. Ryan , R. J. Norman , and R. L. Robker , “Macrophage Contributions to Ovarian Function,” Human Reproduction Update 10 (2004): 119–133.15073142 10.1093/humupd/dmh 011 · doi ↗ · pubmed ↗