Testicular Lmcd1 regulates phagocytosis by Sertoli cells through modulation of NFAT1/Txlna signaling pathway

Abstract Increased oxidative stress is well known to cause testicular dysfunction in aging males, but the detailed relationships between aging, oxidative stress, and testicular function remain to be elucidated. LIM and cysteine‐rich domains 1 (LMCD1) regulates fundamentally cellular process by interacting with transcription factors. A recent study has identified Lmcd1 as one of the most upregulated nuclear proteins associated with Sertoli cell (SC) differentiation, raising the possibility that testicular actions of LMCD1 are likely to take place. Herein, we reported that LMCD1 was exclusively expressed in the nuclei of SCs. This expression was regulated by TNF‐α signaling produced by apoptotic germ cells (GCs) and was suppressed by oxidative stress in a STAT3‐dependent manner. Ablation of endogenous LMCD1 expression caused lipid accumulation and senescence in GC co‐incubated SCs. Using a previously validated in vivo siRNA approach, we showed that LMCD1 depletion significantly impaired male fertility by inducing oligozoospermia and asthenospermia. Mechanistically, LMCD1 upregulation was associated with the nuclear enrichment of the nuclear factor of activated T cells 1 (NFAT1), a core component of Ca2+/calmodulin‐dependent pathway. LMCD1 facilitated the dephosphorylation and nuclear translocation of NFAT1, which consequently expedited the transactivation of Txlna, a binding partner of the syntaxin family essential for testicular phagocytosis, and thus promoted the removal of apoptotic GCs by phagocytic SCs. Collectively, LMCD1 may operate as a novel pretranscriptional integrator linking SC phagocytosis, lipid homeostasis, and cell senescence.

lipids that are used to produce ATP, thus serving as potent energy sources for SCs (Xiong, Wang, Wu, Chen, & Han, 2009). So, phagocytosis is crucial for both GCs and SCs. Although the physiological responses of testicular phagocytosis have been well-documented, the molecular basis underpinning this unique process is still in its infancy. We previously showed that α-taxilin protein (TXLNA), a binding partner of the syntaxin family that functions as a central player in the intracellular vesicle traffic, functions as a critical regulator of testicular phagocytosis in SCs (Dong et al., 2016). Nevertheless, the transcriptional or post-translational mechanisms responsible for controlling TXLNA expression are poorly understood.
Acting as essential components of the regulatory machinery in cells, LIM domain proteins are key factors modulating cell growth, differentiation, and cell fate determination under pathophysiological circumstances (Bian et al., 2010). LIM and cysteine-rich domains 1 (LMCD1) was originally identified by screening of the expressed sequence tag database (dbEST). Northern blot analysis shows that Lmcd1 is highly expressed in cardiac and skeletal muscles. Functionally, LMCD1 regulates cardiac hypertrophy (Bian et al., 2010) and thrombin-induced smooth muscle cell proliferation (Janjanam et al., 2018) by interacting with different transcription factors. However, the physiological relevance of LMCD1 signaling beyond circulatory and musculoskeletal systems remains largely unknown.
Very recently, by using high-throughput screening in rat SCs, Ryser S et al. have identified Lmcd1 as one of the most upregulated nuclear proteins associated with SC differentiation (Ryser et al., 2011). Given that the expression of LIM domain proteins is emerging as a unique signature of SCs (Yang et al., 1998), it is tempting to propose that additional relevant roles of LMCD1 in the control of protein homeostasis at transcriptional level may also be provided with specific gonadal actions.
On the aforementioned basis, we aimed at assessing the expression pattern and biological relevance of LMCD1 signaling in testis. As our initial evidence demonstrated a steady upregulation of LMCD1 expression in SCs treated with apoptotic GCs, multiple experimental settings were used to define the paracrine regulation of this novel nuclear protein, as well as the molecular basis of its direct control of SC phagocytosis.

| SC-specific expression of LMCD1 in mouse testis
Initial RT-qPCR analysis was performed using testicular tissues  2010). Lmcd1 expression began to emerge after SC differentiation (≧23-day) and increased thereafter, with peak expression in adult testis (70-day, Figure 1a). Immunoblotting analysis validated a precisely controlled appearance of LMCD1 in functionally mature SCs (as evidenced by the disappearance of CK18, the marker of immature SCs, at postnatal 23 days (Dong et al., 2016), at the translational level (Figure 1b and Figure S1). Consistently, immunohistochemistry showed the predominant presence of LMCD1 in the SCs (Figure 1c). The exclusive expression of LMCD1 in SCs was further verified in purified testicular cells (Figure 1d). Using immunoblotting, we observed that LMCD1 expression was totally abolished in 540-day-old testes, compared with a strong and a moderate expression in 120-day-old and 23-day-old testes, respectively. Intriguingly, i.p. injection with epigallocatechin-3-gallate (EGCG), a bioactive polyphenol from green tea that exerts a strong antioxidant activity (Ding et al., 2015), ameliorated age-related oxidative stress ( Figure S2) and partially restored the protein levels of LMCD1 in 540-day-old testes ( Figure 1e).

| LMCD1 depletion induces lipid accumulation and senescence in GC co-incubated SCs
The above data point to an association between LMCD1 deregulation and testicular aging. To address this, we generated the 15P-1 cells that were stably deprived of endogenous Lmcd1 expression ( Figure 2a) and then subjected these cells to co-incubation with apoptotic GCs. LMCD1 depletion caused upregulation of the cyclin-dependent kinase inhibitors p16, p19, and damage-related γH2AX, while proliferating cell nuclear antigen (PCNA) was reduced in GC co-incubated 15P-1 Lmcd1sh cells. These phenotypes were undetectable when 15P-1 Lmcd1sh cells were cultured alone (Figure 2b,c). Consistent with these descriptive data, the GC coincubated 15P-1 Lmcd1sh cells showed significantly increased SAß-GAL activity, suggesting stimulated senescence ( Figure 2d).
Because apoptotic GCs are phagocytosed and degraded by SCs to serve as an important energy sources for SCs (Xiong et al., 2009), and because lipid metabolism is tightly linked to cell senescence (Flor, Wolfgeher, Wu, & Kron, 2017), we sought to determine the evaluation of the amount of lipids in GC co-incubated 15P-1 Lmcd1sh cells using lipophilic fluorescent dye BODIPY 493/503. The GC co-incubated 15P-1 Lmcd1sh cells exhibited more fluorescence than their culture-alone counterparts (Figure 2e).

| Oxidative stress directly targets the Lmcd1 transcription
Figure 1e also suggests that testicular oxidative stress may be involved in age-related deregulation of LMCD1 expression. To directly address this, 15P-1 cells were treated in vitro with 0.5 or 1 mM of H 2 O 2 or oxidative stress inducer diamide for 6 h, and LMCD1 protein expression was determined. H 2 O 2 /diamide treatment inhibited LMCD1 expression in a dose-dependent manner, whereas co-treatment with WP1066 (an inhibitor of STAT3 activation) completely neutralized the inhibitory effects of H 2 O 2 / diamide on LMCD1 expression (Figure 3a,b). To ask whether the inhibition of LMCD1 expression by diamide was due to a direct transcriptional effect, 15P-1 cells were transiently transfected with pGL4-Lmcd1-Luc and treated with diamide or diamide plus WP1066. While diamide caused a ~62.6% reduction in luciferase activity relative to control, co-treatment with WP1066 completely restored the luciferase activity in diamide-challenged cells ( Figure 3c). Oxidative stress is known to trigger STAT3 tyrosine phosphorylation (Carballo et al., 1999). Our bioinformatic analysis revealed a putative STAT3-binding site in the Lmcd1 promoter.
We next used EMSA to further examine direct binding of STAT3 onto the Lmcd1 chromatin. STAT3 in the nuclear extracts from GC-stimulated 15P-1/STAT3 cells clearly formed a shifted DIG-

| Requirement of endogenous Lmcd1 for murine spermatogenesis
To understand the biological effects of Lmcd1, we knocked down its expression in vivo using a previously validated protocol (Dong et al., 2016;Gonzalez-Herrera et al., 2006). Treatment with Lmcd1 Stealth siRNA as indicated ( Figure 4a) caused a ~43.2% reduction in Lmcd1 mRNA expression in testis compared with that in testis from Ctrl siRNA-treated mice ( Figure 4b). The effectiveness of siRNA treatment was further confirmed by immunostaining ( Figure 4c). LMCD1 ablation resulted in more frequent and notable vacuoles and lack of mature sperms in the lumen of seminiferous tubules ( Figure 4d). Moreover, the testes treated with Lmcd1 siRNA exhibited a significantly higher level of apoptosis in situ ( Figure 4e). In accordance with these morphological changes, Lmcd1 knockdown significantly impaired fertility potential by inducing oligozoospermia and asthenospermia (Table S1). The integral Lmcd1 expression is thus necessary for normal spermatogenesis.

| Regulation of SC phagocytosis by TNF-α/ Lmcd1 signaling pathway
Because efficient phagocytic clearance of apoptotic GCs by SCs is critical for homeostasis inside the seminiferous tubules (Dong et al., 2016), we were curious whether the abnormal accumulation of testicular apoptosis is due to a potential impairment in phagocytosis. Deregulated metabolism of the lipid droplets is a hallmark of disrupted phagocytic function in SCs (Dong et al., 2016). As shown by Oil Red O staining, lipid droplet accumulation was dramatically increased in Lmcd1 siRNA-treated testis compared with a negligible F I G U R E 1 Testicular expression of LMCD1 is restricted to the nuclei of differentiated Sertoli cells (SCs), and this expression is compromised in aged testis. (a) Lmcd1 mRNA level along the testicular development was assessed using RT-qPCR. Amplification of Gapdh mRNA was used as an internal control. Quantitative values were presented as mean ± SD of three independent experiments. Different superscript letters denote groups that are statistically different (p < 0.05). UD, undetectable. (b) Immunoblotting analysis of LMCD1 expression along the postnatal testicular development. (c) Localization of LMCD1 protein in testicular sections from adult mouse was evaluated using immunohistochemistry. Scale bar, 20 μm. LCs, Leydig cells; Spc, primary spermatocyte; Rsd, round spermatid; Esd, elongated spermatid. (d) Expression profile of LMCD1 was evaluated in different spermatogenic cells and in mouse testis using immunoblotting. Tubulin served as a loading control. (e) 540-day-old mice received i.p. injections with EGCG as described in detail in EXPERIMENTAL PROCEDURES. Following EGCG treatment, testes from different ages of mice were subjected to immunoblotting analysis

| Establishment of TXLNA as the main downstream effector of Lmcd1 signaling during phagocytosis
As further exploration of the molecular basis underpinning Lmcd1 action, we investigated the expression levels of several key factors known to be essential for SC phagocytosis, in Stealth siRNAtreated testis using RT-qPCR. Interestingly, Lmcd1 inhibition was accompanied by decreased Txlna expression by ~62.4% in Stealth siRNA-treated testis ( Figure 6a). This impaired TXLNA expression was further confirmed at the protein level by immunostaining ( Figure 6b). In keeping with this, co-incubation with apoptotic GCs

| Requirement of dephosphorylation of NFAT1 by LMCD1 for TXLNA activation in SCs
Because LMCD1 functions at the transcriptional level (Du et al., 2016), we initially hypothesized that LMCD1 may directly regulate the Txlna transcription, but apparently this was not the case since transfection with pCMV3-Lmcd1 failed to potentiate the apoptotic GC-induced transactivation of Txlna in 15P-1 cells (Figure 7a). The calcium-dependent pathway plays an essential regulatory role along germ cell development (Yuan et al., 2006). A recent study has shown that LMCD1 regulates cardiac hypertrophy by targeting calciumrelated NFAT signaling (Bian et al., 2010). These findings made us curious about a potential involvement of NFATs in SC function. In resting cells, NFATs are normally localized in the cytoplasm due to hyperphosphorylation. Instead, upon different stimulations, NFAT proteins are dephosphorylated, thus translocating to the nucleus to activate target gene expression (Sharma et al., 2011). Interestingly, co-incubation of primary SCs with apoptotic GCs for 6 h only promoted the translocation of NFAT1 to the nuclei but had no effects F I G U R E 3 Oxidative stress directly regulates the Lmcd1 transcription. (a-b) 15P-1 cells were treated with different doses of H 2 O 2 or diamide, in the presence or absence of co-treatment with WP1066 (5 μM), for 6 h, followed by immunoblotting analysis. (c) 15P-1 cells were transiently transfected with pGL4-Lmcd1-Luc. 48 h later, cells were treated with diamide (1 mM) or diamide plus WP1066 (5 μM) for 6 h, followed by the measurement of relative luciferase activity. Different superscript letters denote groups that are statistically different (p < 0.05). (d) 15P-1 cells were transfected with Stat3 Flag pRc/CMV, which renders the STAT3 molecule constitutively active. 48 h later, 15P-1 cells were treated with 1 mM of diamide for another 6 h, followed by nonradioactive electrophoretic mobility shift assay (EMSA). (e) ChIP analysis of pSTAT3 after diamide stimulation (1 mM) of the Lmcd1 promoter. (f) 15P-1 cells were transfected with WT STAT3 (Stat3 Flag pRc/CMV) or DN STAT3 (Stat3 Y705F Flag pRc/CMV) using Lipofectamine™ 3000. 48 h after transfection, cells were starved for 10 h and were then treated for 6 h with 1 mM of diamide, followed by immunoblotting analysis on the cytoplasmic localization of NFAT2 (Figure 7a Figure 8d). Thus, the disturbance in the dephosphorylation and F I G U R E 6 Lmcd1 depletion compromises TXLNA-mediated testicular phagocytosis. (a) Expression levels of different key factors essential for testicular phagocytosis in Lmcd1 siRNA-treated testis were determined using RT-qPCR. **p < 0.01 compared with the values in Ctrl siRNA-treated testis. (b) Expression levels of LMCD1 and TXLNA in Lmcd1 siRNA-treated testis were evaluated using immunohistochemistry at translational level. Scale bar, 20 μm. (c) Wild-type 15P-1 and 15P-1 Lmcd1−/− cells were incubated with apoptotic GCs for 6 h, followed by immunoblotting analysis. (d) Wild-type 15P-1 and 15P-1/TXLNA cells were transduced with Lmcd1 shRNA for 48 h, followed by immunoblotting analysis. (e) Wild-type 15P-1 and 15P-1 Lmcd1−/− cells were transiently transfected with pCMV6-Txlna or empty vectors. 48 h later, cells were subjected to the RB binding and phagocytosis assay, as described above. **p < 0.01 comparing Lmcd1 shRNA+vector to Lmcd1 shRNA+pCMV6-Txlna F I G U R E 7 LMCD1 regulates the dephosphorylation and nuclear translocation of NFAT1, which consequently expedites the transactivation of Txlna, a binding partner of the syntaxin family that functions as a central regulator of testicular phagocytosis. (a) 0.5 μg reporter plasmid and pRL-TK Renilla reporter plasmid were co-transfected into 15P-1 cells with different plasmids as indicated. 48 h later, cells were incubated with apoptotic GCs for 6 h, followed by the measurement of luciferase activities using a dual-luciferase reporter system from Promega. Different superscript letters denote groups that are statistically different (p < 0.05). (b) Primary SCs were incubated with apoptotic GCs for 6 h. Cells were then subjected to nuclear and cytoplasmic fractionation and immunoblotting analysis. (c) Primary SCs were incubated with apoptotic GCs for 6 h, followed by immunofluorescence. (d) Wild-type 15P-1 and 15P-1 Lmcd1−/− cells were incubated with apoptotic GCs for 6 h, followed by immunoblotting analysis. (e) Wild-type 15P-1 and 15P-1 Lmcd1−/− cells were transiently transfected with pCMV6-DDK-Nfat1. 48 h later, cells were incubated with apoptotic GCs for 6 h, followed by Co-IP analysis. (f) 0.5 μg reporter plasmid and pRL-TK Renilla reporter plasmid were co-transfected into wild-type 15P-1 and 15P-1 Lmcd1−/− cells with different plasmids as indicated. 48 h later, cells were incubated with apoptotic GCs for 6 h, followed by the measurement of luciferase activities using a dual-luciferase reporter system from Promega. Different superscript letters denote groups that are statistically different (p < 0.05). (g) The 15P-1/DDK-Nfat1 cells were incubated with apoptotic GCs for 6 h, followed by ChIP assay using anti-DDK antibody. (h) Wild-type 15P-1 and 15P-1 Lmcd1−/− cells were transfected with pCMV6-DDK-Nfat1. 48 h later, cells were incubated with apoptotic GCs for 6 h, followed by ChIP-qPCR analysis using anti-DDK antibody. Different superscript letters denote groups that are statistically different (p < 0.05) | 11 of 14 JIN et al.
nuclear translocation of NFAT1 mediates, at least in part, LMCD1 deficiency-impaired phagocytosis in SCs.

| DISCUSS ION
By using old testis and 15P-1 Lmcd1−/− cells, we firstly found that LMCD1 deficiency promoted cellular senescence in SCs, likely as a result of lipid accumulation. In support of our hypothesis, deregulation of lipid metabolism has been shown to potentiate senescence in many cell lines including retinal epithelial cells (Chen et al., 2019), endothelial cells (Nannelli et al., 2018), and cancer cells (Flor et al., 2017). Of note, the promotion of cellular senescence was only detectable in the GCincubated 15P-1 Lmcd1−/− cells but not in 15P-1 Lmcd1−/− cells alone, suggesting that the LMCD1 action in SCs may require interaction with adjacent GCs. In this regard, the promotion effect of LMCD1 deficiency on SC senescence may be attributed to the disruption in phagocytic function of SCs, but unlikely due to metabolic disturbance in SCs themselves. In favor of our ratiocination, previous studies using neutrophils and macrophages have shown that a failure to mount an efficient autophagy, a similar phagocytic process inside cells, is a deviation on the cell's part from normal cellular function into cell senescence and cessation of the cell cycle (Gomez-Cambronero & Kantonen, 2014). Therefore, such a LMCD1 deficiency-based cellular senescence mechanism may possibly involve two sequential inputs: impaired phagocytosis and ineffectiveness in clearing away decomposed GC fragments that ultimately lead to lipid accumulation (Figure 8e).

Testicular dysfunction caused by increased oxidative stress is a
hallmark of aging males . Nevertheless, whether oxidative stress damages male fertility through affecting testicular phagocytosis remains unknown. We have shown that oxidative stress negatively regulated Lmcd1 transcription through a STAT3dependent pathway in SCs. Because ROS-mediated activation of STAT3 signaling is involved in stress-induced cellular senescence (Liu et al., 2017), and because ROS/STAT3 signaling pathway also regulates autophagy of cancer cells (Yoon et al., 2010), we reason that LMCD1 may function as an important converging point linking oxidative stress, testicular phagocytosis, and SC senescence. Moreover, since disruption of lipid homeostasis is well known for induction of the increased oxidant species in senescent cells (Chen et al., 2019) and apoptotic GCs are phagocytosed and degraded into lipids by SCs to fulfill energy requirement (Xiong et al., 2009), a reciprocal feedback loop is very likely to exist between ROS/LMCD1 signaling and lipid-elicited oxidative stress in phagocytic SCs.
TNF-α, a pleiotropic cytokine produced by GCs, binds to TNFR1 and mediates multiple biologic responses in SCs including androgen receptor expression (Delfino, Boustead, Fix, & Walker, 2003), production of lactate (Riera et al., 2001), immune homeostasis (Wu et al., 2016), and SC apoptosis (Wang & Su, 2018). Accumulating data evidence that protein kinase C (PKC) is one of the core intracellular mediators of TNF-α actions (Chang & Tepperman, 2001). Intriguingly, our results showed that co-incubation with either TNF-α inhibitor SPD304 or PKC inhibitor Ro318220 could both effectively abolish the GC-elicited upregulation of Lmcd1 transcription. The ability of TNF-α to acutely stimulate Lmcd1 mRNA was further confirmed using the 15P-1 Tnfr1−/− cells, which demonstrated also that TNFR1 depletion totally abrogated dead GCinduced Lmcd1 mRNA expression. Overall, our current results strongly suggest that SC-specific expression of LMCD1 is subjected to precise regulation, involving a paracrine control by GC-released TNF-α.
Further evidence for an essential regulation of male fertility by LMCD1 through control of SC phagocytosis is provided by the promoting effects of LMCD1 depletion on the accumulation of lipid droplets in SCs and by the deleterious effects of LMCD1 depletion on phagocytic activities in vitro. To note, unlike other key factors regulating phagocytosis in SCs (Elliott et al., 2010), mice treated with Lmcd1 Stealth siRNAs in vivo only displayed impaired male infertility but not absolute sterility. One tempting possibility is that our in vivo siRNA approach only partially inhibited but not totally depleted the LMCD1 expression in SCs. Further experimental work, including employment of a Sertoli-specific Lmcd1 KO mouse, is needed to fully elucidate the functions of LMCD1 in SCs.
Ca 2+ /calmodulin-dependent pathways are essentially involved in the regulation of testicular functions under both physiological and pathological conditions. The nuclear factor of activated T cells (NFAT) family is such a striking example. NFAT2 is implicated in corticosterone-induced rat Leydig cell apoptosis (Chai, Wang, & Gao, 2007). NFAT1-regulated anti-inflammatory response exerts a potent protective effect on testicular macrophages upon infection with F I G U R E 8 Overexpression of a constitutively active NFAT1 ameliorates the LMCD1 deficiency-impaired Txlna expression and phagocytic activities in 15P-1 cells. (a) 15P-1 Lmcd1−/− cells were transfected with different mutant plasmids as described in the EXPERIMENTAL PROCEDURES section, followed by immunoblotting. (b) 15P-1 Lmcd1−/− cells were transfected with different mutant plasmids as described in the EXPERIMENTAL PROCEDURES section, followed by immunofluorescence. (c) 15P-1 Lmcd1−/− cells with different transfections were incubated with apoptotic GCs for 6 h, followed by the measurement of Txlna expression using RT-qPCR. Different superscript letters denote groups that are statistically different (p < 0.05). (d) 15P-1 Lmcd1−/− cells with different transfections were subjected to the RB binding and phagocytosis assay, as described above. Different superscript letters denote groups that are statistically different (p < 0.05). (e) Proposed working model by the current study: SC-specific expression of LMCD1 is under the precise control of apoptotic GC-produced TNF-α signaling. LMCD1 facilitates the dephosphorylation and nuclear translocation of NFAT1, which consequently potentiates the Txlna transactivation and thus promotes the removal of apoptotic GCs by phagocytic functions of SCs. Disruption in phagocytosis by LMCD1 deficiency causes lipid accumulation, increase in oxidative stress, and cellular senescence in phagocytic SCs uropathogenic E. coli (Bhushan et al., 2011). Nevertheless, the functional details of distinct NFAT members in different testicular cell types and corresponding molecular basis remain poorly understood.
Our findings expand this knowledge by identifying NFAT1 as a downstream effector of LMCD1 pathway in SCs. Facilitation of the nuclear translocation of NFAT1 via potentiating its dephosphorylation, thereafter promoting the NFAT1-mediated transactivation of Txlna, may represent a novel mechanism contributing to the maintenance of SC  (Bian et al., 2010), the available data thus strongly suggest that LMCD1 functions as an essential transcriptional cofactor that helps maintain the function of NFAT1/Txlna cascade during phagocytosis. In favor of this, it has been shown that LMCD1 inhibits DNA binding of transcriptional factor GATA6 and thus restricts its function during lung and heart development (Rath, Wang, Lu, & Morrisey, 2005). Likewise, by acting as a co-activator with E2F transcription factor 1, LMCD1 stimulates human aortic smooth muscle cell (HASMC) proliferation and thereby promotes human atherogenesis (Janjanam et al., 2018). In this context, a precise balance between interactions with LMCD1 and NFAT1 is indispensable for the regulation of spatial and temporal activity of NFAT1 as well as other key transcriptional factors in phagocytic SCs.
In summary, our current results unambiguously demonstrate that SC-specific expression of LMCD1 is essential for phagocytic activity and this expression is under the precise control of both apoptotic GC-produced TNF-α and lipid-elicited ROS signaling pathways. One major mechanism underlying the LMCD1 action originates from its ability to facilitate the dephosphorylation and nuclear translocation of NFAT1, which consequently expedites the Txlna transactivation and thus promotes the removal of apoptotic GCs by phagocytic function of SCs (Figure 8e). Overall, LMCD1 may operate as a novel pretranscriptional integrator linking SC phagocytosis, lipid homeostasis, and cell senescence.

| Animal
Male C57BL/6J mice were obtained from the Animal Research Center of our university. Mice were fed ad libitum and were allowed to acclimatize for at least 1 week before the experiment. Mice were finally euthanized by CO 2 inhalation followed by cervical dislocation. To study oxidative stress regulation of LMCD1 expression, 540-day-old mice received i.p. injections with EGCG (50 mg/kg, MedChemExpress, Shanghai, China) for consecutive 3 days, followed by a 4-day break. Mice received a total of 4 cycles of EGCG treatment.
In vivo ablation of endogenous LMCD1 expression in testis was achieved using a previously established protocol (Dong et al., 2016;Gonzalez-Herrera et al., 2006). Briefly, mice were injected intraperitoneally (i.p.) with 100 μl solution of Lmcd1 Stealth siRNAs or scramble Ctrl siRNAs daily for 10 days in isotonic saline solution (3 μg/ mouse) followed by 2 days of break. Mice received a total of three cycles of siRNA injections (~35 days), and this duration was selected because one cycle of murine spermatogenesis consists of 35 days (Timmons, Rigby, & Poirier, 2002). At the end of 35 days after the first siRNA injection, mice were subjected to the evaluation of reproductive capacity and characterization of epididymal sperms status, as described in detail in our previous work (Dong et al., 2016).

| Cell treatment
15P-1 cells were obtained from the American Type Culture Collection (Manassas, VA, USA) and maintained in DMEM/F12 containing 10% fetal calf serum (Thermo Fisher Scientific, Shanghai, China) in a humidified atmosphere of 5% CO 2 at 37°C. Primary germ cells (GCs), SCs, and caudal sperms were isolated and purified from 8-week-old male C57BL/6J mice, as described by our previous work Ma et al., 2010;Zhang et al., 2013). GCs were cultured alone for 2 days to induce spontaneous apoptosis (Wang et al., 2006). To answer whether moderate oxidative stress regulated LMCD1 expression, 15P-1 cells were treated with dif- for 48 h, prior to other assays. Retroviral plasmids that expressed the wild-type mouse Nfat1 (WT-Nfat1), a constitutively active mouse Nfat1, mutated to interfere selectively with the NFAT:AP-1 interaction (CA-RIT-Nfat1), and a constitutively active mouse Nfat1 that is unable to interact with AP-1 with four-point mutations in the DNA-binding loop that abolish DNA binding (DBDmut-Nfat1) were all gifts from Anjana Rao (Addgene plasmid #11100, Addgene plasmid #63214, and Addgene plasmid #63669; Addgene, Watertown, MA, USA). The 15P-1 cells that stably expressed different exogenous mouse Nfat1 were generated by transfected with the aforementioned plasmids followed by selection with 50 ng/ ml of puromycin (Sigma-Aldrich) or by fluorescence-activated cell sorting, respectively. To study the potential involvement of different pathways in apoptotic GC-induced Lmcd1 expression, 15P-1 cells were incubated with apoptotic GCs, in the presence of different pathway inhibitors including 5 μM of BLT-1 (SR-BI), 0.5 μM of SCH772984 (ERK1/2), 5 μM of SB203580 (p38 MAPK), 1 μM of SPD304 (TNF-α), and 5 μM of Ro318220 (PKC), for 6 h, followed by RT-qPCR analysis. BLT-1 was from Merck (Shanghai, China), and the other inhibitors were from MedChemExpress (Shanghai, China).

| In vitro binding and phagocytosis assay
RBs were isolated and purified from testes of 8-week-old mice, as described in previous publications (Dong et al., 2016;Yefimova et al., 2008). In brief, seminiferous tubules were digested by in 25 ml of 0.01 M PBS (pH 7.2) containing 0.1% glucose, 3 mM lactate, and 0.25 mg/ml trypsin for 20 min at 33°C. The resultant cell suspension was filtered through sterile surgical gauze, followed by three consecutive centrifugations at 800 g for 10 min. The resultant supernatants that contained RBs were stored at 4°C and were then subjected to biotin labeling using an EZ-Link NHS-PEG4 Biotinylation Kit (Thermo Fisher Scientific). To measure the phagocytotic activity in vitro, 15P-1 cells were seeded in eight-well Labtek chambers and were then incubated with 1.2 × 10 7 biotinylated RBs at 33°C in a humidified atmosphere of 5% CO2 for different durations as indicated.
Upon completion of incubation, unbound RBs were washed away with culture medium and fixed for 5 min with 4% paraformaldehyde/ PBS (pH 7.4), followed by labeling with avidin-FITC at 25°C for another 1 h. The ingested and membrane-bound RBs were observed using an inverted microscope (Axio Imager M1 Microscope, Zeiss), and the binding and phagocytic index was calculated as the ratio of positive green signals/cell number in each well.
Detailed information on histological examination, measurement of testicular oxidative stress, lipid accumulation and cellular senescence determination, RT-qPCR, immunoblotting, luciferase reporter assay, co-immunoprecipitation (Co-IP) and chromatin immunoprecipitation (ChIP), nonradioactive electrophoretic mobility shift assay (EMSA), and statistical analysis pertinent to this paper can be found in Supporting Information.

ACK N OWLED G M ENTS
This work was supported by the National Natural Science Foundation of China (NSFC) 31671198 and 31971070.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest.

AUTH O R CO NTR I B UTI O N S
WL and YZ designed and supervised the experiments, contributed reagents, and analyzed the data. WL wrote and revised the paper and supervised the submission. XJ and SZ completed the majority of experiments and analyzed the data, and XJ wrote the manuscript. TD, PZ, and CZ performed the experiments and helped to analyze the data.