Stress granule–mediated ZBP1 activation drives necroptotic cell death in non-obstructive azoospermia and testicular aging

Significance

The causes of male infertility and reproductive aging remain largely elusive. This study identifies a necroptotic pathway, initiated by stress granule formation and mediated by ZBP1 and RIPK3, as a key mechanism driving the loss of testicular spermatogonia and Sertoli cells in both non-obstructive azoospermia and aging. By linking cellular stress responses to regulated cell death in the testis, these findings reveal a central pathway that underlies male reproductive decline across both pathological and physiological contexts.

Abstract

Male infertility remains a major unmet medical challenge, with poorly defined molecular mechanisms and no effective therapies. Here, we identify a stress granule–mediated necroptotic pathway as a key driver of non-obstructive azoospermia, a severe form of male infertility marked by the loss of spermatogenesis. Environmental or physiological stress activates eIF2α kinases, inducing stress granule formation and the recruitment of ZBP1 and RIPK3 into a cytoplasmic complex. This assembly triggers RIPK3 activation, MLKL phosphorylation, and necroptotic death of spermatogonia and Sertoli cells. Genetic ablation of Zbp1 or Ripk3 protects mice from heat-induced testicular degeneration, establishing their essential role in stress-induced testicular damage. Importantly, activation of this pathway is also observed in aged human testes, linking stress-responsive necroptosis to both pathological infertility and the broader process of reproductive aging. These findings reveal an unrecognized mechanism that couples cellular stress responses to regulated cell death in the male reproductive system.

Male infertility affects up to 15% of couples worldwide, with azoospermia—complete absence of sperm in semen—among its most severe forms (1, 2). Non-obstructive azoospermia (NOA), the majority form of azoospermia characterized by testicular failure, remains poorly understood at the molecular level despite its clinical prevalence (3, 4). Testicular aging similarly leads to diminished spermatogenesis (5–7), yet the cellular processes linking infertility and aging remain elusive.

Necroptosis, a regulated form of inflammatory cell death, has emerged as a common feature of neurodegeneration, viral infection, and tissue injury, but its relevance to infertility is unknown (8–10). The activation of necroptosis in spermatogonia and Sertoli cells of aged mice, along with phospho-MLKL enrichment in aged human testes, was observed; however, the mechanism driving such an activation remains unknown (11, 12).

Here, we demonstrate that NOA is driven by stress granule–induced necroptosis mediated through the ZBP1–RIPK3–MLKL pathway. We show that environmental stress triggers eIF2α kinase activation and stress granule assembly, which recruits ZBP1 and RIPK3 to initiate necroptosis. Testicular biopsy samples from NOA patients exhibited robust phospho-MLKL and stress granule markers in all cases examined. Genetic deletion of Zbp1 or Ripk3 prevented heat-induced testicular atrophy in mice, confirming the essential role of this pathway. Notably, we also detect activation of this pathway in aged human testes, revealing a shared mechanism linking infertility and reproductive aging. These findings identify stress granule–mediated necroptosis as a key contributor to male reproductive decline and suggest molecular targets for intervention.

Results

Patient Cohort and Testicular Histopathology.

We recruited 50 unrelated patients diagnosed with NOA for histological and molecular analyses. Diagnosis was based on AUA/ASRM clinical guidelines (13), and the final cohort included 40 cases of idiopathic NOA and 10 with a history of cryptorchidism, a condition in which one or both testes fail to descend into the scrotum, resulting in prolonged exposure of the testes to the higher intra-abdominal temperature (SI Appendix, Fig. S1). Control testicular samples were obtained from 17 patients undergoing orchiectomy for testicular torsion, characterized by ischemia-related loss of blood flow. Histological evaluation revealed that NOA testes lacked sperm and showed a Johnsen Score≤7, indicative of complete spermatogenic failure, while torsion samples retained normal spermatogenesis (Fig. 1 A–C and SI Appendix, Fig. S2 and Dataset S1). The mean age of NOA patients at diagnosis was 32 y.

Fig. 1.

Phospho-MLKL(p-MLKL) were detected in the seminiferous tubules of NOA patients’ testes. (A–C) Hematoxylin and Eosin (H&E) staining of testis sections from human testicular torsion (n = 17) and azoospermia patients (n = 50) in (A). Johnsen Score evaluation of testicular torsion and azoospermia patients based on (A) and shown in (B). The number of seminiferous tubules with sperm was counted based on (A) and quantification in (C), seminiferous tubules with sperm were counted in five fields per testis. (Scale bar, 100 μm.) (D and E) Immunohistochemistry (IHC) analysis of human testicular torsion and NOA testis sections with p-MLKL antibody in (D). The number of seminiferous tubules with positive p-MLKL signal was counted based on IHC staining and quantification in (E). (Scale bar, 100 μm.) (F) Western blotting analysis of RIPK3, p-MLKL, and MLKL in the testis from two NOA patients, GAPDH was used as loading control. HeLa and HT29 cells treated with the combination of T/S/Z as negative and positive control for western blotting analysis, respectively. The asterisk (*) indicates nonspecific bands. (G) Immunofluorescence analysis of NOA testis sections with antibodies against p-MLKL (red) PIWIL4 (spermatogonium specific protein, green) and SOX9 (Sertoli cell–specific protein, green). (Scale bar, 50 μm.) Quantified data in (B, C, and E) represent the mean ± s.e.m. ****P < 0.0001. P values were determined by two-sided unpaired Student’s t tests.

Phospho-MLKL Is Robustly Induced in NOA Testes.

Given prior evidence linking phospho-MLKL–driven necroptosis to aging-related testicular decline in mice and humans (11, 12, 14), we investigated whether this pathway was similarly activated in NOA. IHC using a phospho-MLKL–specific antibody revealed robust signal in the seminiferous tubules of all NOA patients (100%, n = 50), but no staining in torsion controls (0%, n = 17) (Fig. 1 D and E). To confirm antibody specificity, we performed two independent controls: 1) peptide competition with phosphorylated MLKL peptides and 2) lambda protein phosphatase (LPP) pretreatment. Both eliminated the IHC signal, validating the specificity of phospho-MLKL detection (SI Appendix, Fig. S3 A and B).

RIPK3, Not RIPK1, Drives MLKL Activation in NOA.

As RIPK3 is the only known kinase that phosphorylates MLKL, its activity is inferred by phospho-MLKL presence (8–10, 15). We next asked whether RIPK3 was activated via canonical RIPK1-dependent signaling. Phospho-RIPK1 was undetectable in all NOA samples (0%, n = 50), indicating that RIPK3 activation in NOA occurred in a RIPK1-independent mode (SI Appendix, Fig. S4A). Western blotting of testis lysates from two NOA patients confirmed the presence of phospho-MLKL. Negative and positive controls included HeLa and HT29 cells treated with TNF-α, Smac mimetic, and the pan-caspase inhibitor Z-VAD-fmk (T/S/Z), respectively (Fig. 1F). Cleaved caspase-3, a marker of apoptosis, was undetectable in all NOA samples (0%, n = 50), further supporting necroptosis as the dominant cell death pathway in NOA (SI Appendix, Fig. S4B).

Necroptosis Occurs in Spermatogonia and Sertoli Cells.

To identify the cell types undergoing necroptosis, we performed IHC costaining using cell-type-specific markers. Phospho-MLKL colocalized with PIWIL4 (spermatogonia) and SOX9 (Sertoli cells) and to a lesser extent with DDX4 (spermatocytes/spermatids) (Fig. 1G and SI Appendix, Fig. S5). In contrast, Leydig cells, which reside outside the seminiferous tubules and lack RIPK3 expression, did not exhibit phospho-MLKL signal.

Heat Stress Induces ZBP1–RIPK3–Dependent Necroptosis.

The testes are positioned outside the body to maintain a temperature 2 to 4 °C below core body temperature, a requirement critical for spermatogenesis (16, 17). In our NOA patient cohort, we noted that the 10 individuals with cryptorchidism exhibited stronger phospho-MLKL staining than those with idiopathic NOA, suggesting that elevated testicular temperature may promote necroptosis. Given prior reports implicating ZBP1 in heat stroke–induced necroptosis (18), we examined whether heat shock (HS) alone could induce necroptotic death in mouse testicular cell lines. Exposure of GC-2spd(ts) (spermatocytes), 15P-1 (Sertoli cells), and MA-10 (Leydig cells) to 43 °C did not elicit substantial cell death (Fig. 2 A and B), nor did it in L929 mouse fibroblast cells, even 24 h post–heat shock (SI Appendix, Fig. S6 A and B). Because ZBP1 expression is known to be IFN-β–inducible (19, 20), we pretreated those cells with interferon-β(IFN-β) before heat shock. Strikingly, IFN-β–primed GC-2spd, 15P-1, and L929 cells exhibited significant cell death upon heat stress, whereas vehicle (DMSO)-treated controls did not (Fig. 2B and SI Appendix, Fig. S6 B and C). These results indicate that IFN-β–induced gene expression is required for heat-induced necroptosis in those cell lines. Consistent with this, phospho-MLKL (Ser345)—the murine equivalent of human Ser358—was readily detected in GC-2spd and 15P-1 cells after IFN-β/HS treatment (Fig. 2C). In contrast, MA-10 cells, which lack RIPK3 expression, failed to show either cell death or phospho-MLKL signal following IFN-β/HS exposure (Fig. 2 B and C).

Fig. 2.

Heat shock–induced necroptosis depends on ZBP1 and RIPK3. (A–C) Schematic illustrating experiment design in (A). Cultured GC-2spd, 15P-1, and MA-10 cells were treated with DMSO or IFN-β for 18 h, then the cells were transferred to 1.5 mL Eppendorf Tubes (EP) and put into 37 °C or 43 °C water for 2 h. 6 h after heat shock (HS) cell viability as measured by Cell Titer-Glo in (B). The levels of p-MLKL, MLKL, and RIPK3 were analyzed by immunoblotting in (C), GAPDH was used as loading control. (D) Cultured L929 cells with wild type (WT), Ripk1, Ripk3, or Mlkl gene knocked out were treated with IFN-β and HS for 18 h and 1.5 h as described in SI Appendix, Fig. S6A, 2 h after heat shock cell viability as measured by Cell Titer-Glo. (E and F) Cultured L929 cells with wildtype or Zbp1 gene knocked out were treated with IFN-β and HS for 18 h and 1.5 h as described in SI Appendix, Fig. S6A, 2 h after heat shock cell viability as measured by Cell Titer-Glo in (E). The levels of p-MLKL, MLKL, and ZBP1 were analyzed by immunoblotting in (F), GAPDH was used as loading control. (G) Cultured L929(Ripk3^−/−)-HA-3 × Flag-mRIPK3 cells were treated with IFN-β/HS as indicated. The cell extracts were prepared and subjected to immunoprecipitation with an anti-Flag antibody. The extracts (Input) and the immuno-precipitates (IP: Flag) were then subjected to western blotting analysis using antibodies as indicated. (H–L) Schematic illustrating the experiment design in SI Appendix, Fig. S8A. The lower body of 12-wk-old Ripk3^+/+, Ripk3^−/− littermate, and Zbp1^−/− male mice (n = 7 per each genotype) was put in 43 °C waters for 20 min every other day, for a total of three exposures. Mice were euthanized 7 d after the final treatment. Testes were collected for analysis of weight in (H), histology via H&E staining in (I), seminiferous tubules thickness quantification in (J), p-MLKL IHC staining in (K), and quantification of p-MLKL-positive cells in (L). (Scale bar, 100 μm.) Data in (B, D, and E) are mean ± SD of triplicate wells. Quantified data in (H, J, and L) represent the mean ± s.e.m. ****P<0.0001. P values were determined by two-sided unpaired Student’s t tests.

To dissect the molecular requirements for this process, we employed IFN-β/HS treatment in L929 cells lacking key necroptosis regulators. Genetic deletion of Ripk3 or Mlkl abolished necroptosis, while Ripk1 knockout enhanced IFN-β/HS-induced cell death (Fig. 2D), implicating RIPK1 as a negative regulator in this context. Consistently, phospho-MLKL levels were elevated in Ripk1-deficient L929 cells, while entirely absent in Ripk3-null cells following IFN-β/HS treatment (SI Appendix, Fig. S6D).

These findings were further supported in MEF, GC-2spd, and 15P-1 cells, all of which underwent necroptosis in response to IFN-β/HS (SI Appendix, Fig. S6 E and F). Collectively, these results demonstrate that heat stress–induced necroptosis requires IFN-β–driven ZBP1 expression, and proceeds through a RIPK3–MLKL–dependent mechanism, with RIPK1 acting as an inhibitory regulator in this context.

IFN-β/Heat Stress–Induced Necroptosis Requires Jak/STAT-Mediated ZBP1 Expression.

To directly test the role of ZBP1 in heat-induced necroptosis, we generated Zbp1-knockout L929, GC-2spd, and 15P-1 cells. Upon IFN-β and heat shock (IFN-β/HS) treatment, all Zbp1-deficient cell lines were completely resistant to necroptotic cell death, in contrast to wild-type controls (Fig. 2E and SI Appendix, Fig. S7 A and B). ZBP1 protein was robustly induced in wild-type cells by IFN-β/HS, but absent in Zbp1-deficient counterparts (Fig. 2E and SI Appendix, Fig. S7 A and B). Furthermore, phospho-MLKL was detected in IFN-β/HS-treated wild-type L929 cells, but abolished in Zbp1-knockout cells, confirming ZBP1 as essential for necroptosis induction (Fig. 2F).

Because ZBP1 is an interferon-stimulated gene, we next tested whether the Jak/STAT signaling pathway mediates its expression. Pharmacological inhibition of Jak kinases completely protected L929, GC-2spd, and 15P-1 cells from IFN-β/HS-induced cell death (SI Appendix, Fig. S7 C and D). This protection was accompanied by a loss of ZBP1 expression and phospho-MLKL, indicating that Jak/STAT signaling is required upstream of necroptosis in this context (SI Appendix, Fig. S7E).

To confirm the sufficiency of ZBP1, we used HeLa-RIPK3/TetOn-ZBP1 cells, which stably express RIPK3 and allow doxycycline (DOX)-inducible ZBP1 expression. Upon DOX/HS treatment, these cells underwent robust RIPK3-dependent necroptosis, consistent with a direct role for ZBP1 in initiating the pathway (SI Appendix, Fig. S7F). We further confirmed that ZBP1 physically interacts with RIPK3 following IFN-β/HS treatment using coimmunoprecipitation, demonstrating ZBP1–RIPK3 complex formation under necroptotic conditions (Fig. 2G). To dissect the structural requirements for ZBP1 function (21), we introduced various ZBP1 domain mutants into Zbp1^−/−L929 cells: full-length ZBP1, ZBP1(∆Zα1/2) (lacking both Zα domains), ZBP1(∆C) (lacking the C-terminal domain), and ZBP1mut (bearing point mutations in both RHIM domains) (SI Appendix, Fig. S7G). Reexpression of wild-type ZBP1 or ZBP1(∆C) restored necroptosis, while ZBP1(∆Zα1/2) and ZBP1mut failed to do so—even though all constructs were expressed at similar levels (SI Appendix, Fig. S7 G and H). These results indicate that both the Zα domains and RHIM motifs are required for ZBP1-mediated activation of RIPK3 and necroptosis in response to heat stress. Finally, unlike the cell lines requiring IFN-β priming, ZBP1 is constitutively expressed at similar levels in human and mouse testicular tissue (SI Appendix, Fig. S7 I and J). This basal ZBP1 expression allows for direct activation of ZBP1–RIPK3–MLKL–mediated necroptosis in vivo, helping to explain the susceptibility of testicular cells to heat-induced cell death and azoospermia.

Zbp1 or Ripk3 Deficiency Protects Against Heat-Induced Testicular Damage In Vivo.

To assess whether heat stress triggers ZBP1–RIPK3–dependent necroptosis in vivo, we established a mouse model in which the scrotum was immersed in 43 °C water for 20 min every other day over 5 d (22–24). Mice were killed on day 7 following the final exposure, and their body weights, testicular weights, and histological features were examined (SI Appendix, Fig. S8 A and B). In wild-type mice, heat exposure led to significant testicular atrophy, as indicated by reduced testis weight. In contrast, both Ripk3-knockout and Zbp1-knockout mice were largely protected from testis weight loss (Fig. 2H and SI Appendix, Fig. S8C). Histological analysis revealed severe cell depletion and near-empty seminiferous tubules in heat-treated wild-type testes, whereas tubules in Ripk3-knockout and Zbp1-knockout remained intact, with lumens containing abundant germ cells and spermatozoa—resembling untreated controls (Fig. 2I and SI Appendix, Fig. S8D). Heat-induced thinning of seminiferous tubule walls was also significantly attenuated in Ripk3- and Zbp1-deficient mice (Fig. 2J). Furthermore, phospho-MLKL was robustly detected in the seminiferous epithelium of wild-type testes post–heat stress but was undetectable in the same regions of Ripk3- and Zbp1-deficient testes (Fig. 2 K and L). These findings demonstrate that heat stress activates necroptosis in testicular tissue via a ZBP1–RIPK3–MLKL axis and that genetic ablation of Zbp1 or Ripk3 is sufficient to prevent heat-induced testicular damage and germ cell loss, supporting a central role for this pathway in heat-related azoospermia.

Stress Granule Formation Is Required for Heat-Induced Necroptosis.

Stress granules (SGs) are membrane-less, phase-separated cytoplasmic assemblies of ribonucleoproteins that form in response to acute stressors such as heat shock, viral infection, tumorigenesis, and neurodegeneration (25, 26). Key RNA-binding proteins, including TIA-1 and G3BP1, are essential for SG nucleation (27, 28). Although SGs have been proposed to promote cell survival (29), recent work has shown that certain cytotoxic insults, such as sodium arsenite, can trigger ZBP1–RIPK3–dependent necroptosis via recruitment of RIPK3 to SGs (30).

To investigate whether SGs are similarly required for heat-induced necroptosis, we treated parental or G3bp1, G3bp2, or G3bp1/2 double-knockout (DKO) L929 cells with IFN-β and heat shock (IFN-β/HS). Western blot analysis showed that G3bp1/2-DKO cells failed to activate phospho-MLKL and those cells were completely resistant to IFN-β/HS-induced necroptosis (Fig. 3 A and B and SI Appendix, Fig. S9 A and B).

Fig. 3.

Stress granule is required for heat shock–induced necroptosis. (A and B) Cultured L929 cells with wild type or G3bp1 and G3bp2 double gene knocked out were treated with IFN-β and HS for 18 h and 1.5 h as described in SI Appendix, Fig. S6A, 2 h after heat shock cell viability as measured by Cell Titer-Glo in (A). The levels of p-MLKL, MLKL, ZBP1, G3BP1, and G3BP2 were analyzed by immunoblotting in (B), GAPDH was used as loading control. Data in (A) are mean ± SD of triplicate wells. (C) Cultured L929(Ripk3^−/−)-HA-3 × Flag-mRIPK3 cells were treated with IFN-β/HS as indicated. The cell extracts were prepared and subjected to immunoprecipitation with an anti-Flag antibody. The extracts (Input) and the immuno-precipitates (IP: Flag) were then subjected to western blotting analysis using antibodies as indicated. The asterisk (*) indicates nonspecific bands. (D) Cultured L929(Zbp1^+/+) and L929(Zbp1^−/−) cells were treated with the indicated stimuli for 18 h (IFN-β) and 0.5 h (HS). The RIPK3 and G3BP1 were detected by immunofluorescence. (Scale bar, 10 μm.) (E and F) IHC analysis of human testicular torsion and NOA testis sections with G3BP1 antibody in (E). The number of seminiferous tubules with positive G3BP1 dot signal was counted based on IHC staining and quantification in (F). (Scale bar, 100 μm.) Data represent the mean ± s.e.m. ****P < 0.0001. P values were determined by two-sided unpaired Student’s t tests. (G and H) Immunofluorescence analysis of NOA testis sections (unknown cause of NOA, n = 5; cryptorchidism, n = 5) with antibodies against G3BP1(green) and RIPK3(red) in (G). (Scale bar, 50 μm.) Profiling of representative white dotted line traces the intensities of RIPK3 and G3BP1 signal based on (G) and analyzed in (H). (I and J) Immunofluorescence analysis of NOA testis sections (unknown cause of NOA, n = 5; cryptorchidism, n = 5) with antibodies against G3BP1(green) and ZBP1(red) in (I). (Scale bar, 50 μm.) Profiling of representative white dotted line traces the intensities of ZBP1 and G3BP1 signal based on (I) and analyzed in (J).

Coimmunoprecipitation experiments revealed that RIPK3 was recruited to SGs and formed a complex with ZBP1 and G3BP1 following IFN-β/HS treatment (Fig. 3C), implicating a functional interaction. To investigate how RIPK3 is targeted to SGs, we examined the subcellular localization of RIPK3 and G3BP1 in L929, GC-2spd, and 15P-1 cells, with or without Zbp1, before and after IFN-β/HS exposure. In untreated cells, RIPK3 was diffusely cytoplasmic, but after treatment, it became punctate and filamentous, colocalizing with G3BP1 (Fig. 3D and SI Appendix, Fig. S9C). This redistribution was abolished in Zbp1-deficient cells, indicating that ZBP1 is required for RIPK3 recruitment to stress granules.

To assess whether SG formation occurs in vivo, we performed IHC staining of NOA patient testes using a G3BP1 antibody. Punctate G3BP1 signals were prominently detected in the seminiferous tubules of most NOA samples but were mostly absent in testicular torsion controls (Fig. 3 E and F). Importantly, these G3BP1 puncta colocalized with ZBP1 and RIPK3 within the same testicular cells, including those from patients without cryptorchidism (Fig. 3 G–J). Similarly, heat-treated mouse testes exhibited robust G3BP1 puncta, which were nearly absent in untreated controls (SI Appendix, Fig. S9 D–F). Collectively, these results demonstrate that stress granules are essential platforms for ZBP1–RIPK3 complex formation and necroptosis activation, both in vitro and in vivo, and that this mechanism underlies testicular degeneration in NOA.

eIF2α Kinases Are Required for Stress-Induced Necroptosis.

The integrated stress response (ISR) is a cytoprotective signaling cascade triggered by phosphorylation of eukaryotic initiation factor 2 alpha (eIF2α) at Ser51, a key regulatory event that inhibits translation during stress (31, 32). Four eIF2α kinases mediate ISR activation in response to distinct stimuli: HRI (heme depletion), PKR (viral infection), PERK (ER stress), and GCN2 (amino acid deprivation) (32). Deletion of all four kinases abrogates ISR signaling entirely (33).

To examine whether eIF2α kinases are required for necroptosis, we generated quadruple-knockout L929 cells (lacking HRI, PKR, PERK, and GCN2; SI Appendix, Fig. S10A). Upon IFN-β/heat shock (HS) treatment, quadruple-knockout cells were resistant to necroptosis, and phospho-MLKL as well as G3BP1 puncta were not detected (Fig. 4 A–C). Stepwise deletion revealed that PKR and HRI deficiency partially reduced necroptosis, while additional GCN2 knockout, almost completely abolished cell death. In contrast, PERK knockout had no additional protective effect to the PKR/HRI knockout cells (SI Appendix, Fig. S10B). Consistently, phospho-MLKL and phospho-eIF2α were totally abolished in cells in which GCN2 was knocked out in addition to PKR/HRI/PERK knockout while PKR/HRI/PERK triple knockout cells showed similar levels of signals as in PKR/HRI double knockout cells (SI Appendix, Fig. S10C), indicating that out of these four stress kinases, PERK is the only one not participating in IFN-β/HS-mediated necroptosis induction. We next tested whether other ISR-activating stressors could trigger ZBP1–RIPK3–dependent necroptosis. Exposure to arsenite, halofuginone (GCN2-dependent), and 8-azaadenosine (PKR-dependent) all induced necroptosis in L929 cells, while tunicamycin (PERK-dependent) did not (SI Appendix, Fig. S11 A–D). These results further suggested that GCN2, PKR, and possibly HRI—but not PERK—mediate necroptosis induction under stress. To explore whether ISR is similarly activated in vivo, we performed IHC staining of testis samples from NOA patients. Phosphorylated GCN2 (p-GCN2) was detected in 90% of NOA samples (45/50) and the signal colocalized with phosphor-MLKL in the same seminiferous tubule cells (Fig.4 D–F and J), while phospho-PKR was detected in 46% (23/50) of the NOA testis samples and the signal also colocalized with phosphor-MLKL (Fig. 4 G–I and J). Both signals were nearly absent in testicular torsion controls (Fig. 4 D, E, G, and H). Detection of phospho-PERK was minimal in both NOA and control samples (SI Appendix, Fig. S11 E and F) and HRI phosphorylation could not be evaluated due to a lack of specific antibodies. Of the 50 NOA samples, 45 exhibited phospho-GCN2, 23 exhibited phospho-PKR, and 18 were positive for both (Fig. 4 J and K). These findings demonstrate that activation of the ISR, particularly through GCN2/PKR, is essential for stress granule formation and ZBP1–RIPK3-mediated necroptosis. The presence of activated eIF2α kinases in NOA testis supports a model in which chronic or acute stress triggers ISR-driven necroptosis, contributing to testicular degeneration and infertility.

Fig. 4.

Stress kinases induce ZBP1 and RIPK3-dependent necroptosis in NOA. (A and B) Cultured L929 cells with wildtype or Pkr, Hri, Perk, and Gcn2 four gene knocked out (quadruple-KO) were treated with IFN-β and HS for 18 h and 1.5 h as described in SI Appendix, Fig. S6A, 2 h after heat shock cell viability as measured by Cell Titer-Glo (A). The levels of p-MLKL and MLKL were analyzed by immunoblotting in (B), GAPDH was used as loading control. Data in (A) are mean ± SD of triplicate wells. (C) Cultured L929(wildtype) and L929(quadruple-KO) cells were treated with the indicated stimuli for 18 h (IFN-β) and 0.5 h (HS). The RIPK3 and G3BP1 were detected by immunofluorescence. (Scale bar, 10 μm.) (D and E) IHC analysis of human testicular torsion and NOA testis sections with p-GCN2 antibody in (D). The number of seminiferous tubules with positive p-GCN2 signal was counted based on IHC staining and quantification in (E). (Scale bar, 100 μm.) Data represent the mean ± s.e.m. ****P < 0.0001. P values were determined by two-sided unpaired Student’s t tests. (F) Immunofluorescence analysis of NOA testis sections (n = 5) with antibodies against p-MLKL (red) and p-GCN2 (green). (Scale bar, 50 μm.) (G and H) IHC analysis of human testicular torsion and NOA testis sections with p-PKR antibody in (G). The number of seminiferous tubules with positive p-PKR signals was counted based on IHC staining and quantification in (H). (Scale bar, 100 μm.) Data represent the mean ± s.e.m. **P < 0.01. P values were determined by two-sided unpaired Student’s t tests. (I) Immunofluorescence analysis of NOA testis sections (n = 5) with antibodies against p-MLKL (red) and p-PKR (green). (Scale bar, 50 μm.) (J and K) Analysis of p-GCN2 and p-PKR-positive testis sections from the testicular torsion (n = 17) and NOA (n = 50) testes in (J). Analysis of NOA testis sections with single-positive p-GCN2, single-positive p-PKR, and double-positive p-GCN2 and p-PKR in (K).

Stress Markers and Phospho-MLKL Are Elevated in Aging Human Testes.

To investigate whether ZBP1–RIPK3–dependent necroptosis occurs during physiological testis aging, we analyzed testicular tissue from 30 prostate cancer patients (age range: 59 to 89 y; mean: 73; and Dataset S1). Testicular torsion samples served as controls. Histological analysis revealed that aging testes had significantly lower Johnsen Scores (mean score = 7.4), indicating impaired spermatogenesis compared to controls (Fig. 5 A and B). IHC using a human phospho-MLKL antibody revealed robust staining in the seminiferous tubules of 29/30 (96.7%) aging testis samples, whereas no signal was detected in torsion controls (0%, n = 17) (Fig. 5 C and D). We next asked whether the ISR–stress granule–necroptosis pathway was also activated in aging testes. IHC for G3BP1, phospho-GCN2, phospho-PKR, and phospho-PERK was performed. Punctate G3BP1 staining, indicative of stress granule formation, was detected in most aging samples but was absent or minimal in torsion controls (Fig. 5 E and F). Similarly, phospho-GCN2 was present in 96.7% (29/30) of aging samples (Fig. 5 G, H, and J), and phospho-PKR in 23% (7/30) (Fig. 5J and SI Appendix, Fig. S11 G and H). In contrast, phospho-PERK was not detected in either group (SI Appendix, Fig.S11 I and J). Colocalization analysis revealed that phospho-GCN2 and phospho-PKR overlapped with phospho-MLKL in the same seminiferous tubule cells (Fig. 5I and SI Appendix, Fig. S11K), suggesting functional integration of these stress pathways. Of the 30 aging samples, 29 exhibited phospho-GCN2, 7 exhibited phospho-PKR, and 6 were positive for both (Fig. 5 J and K). Together, these findings demonstrated that eIF2α kinase activation and stress granule formation occurred in the aging human testis, leading to ZBP1–RIPK3–MLKL–mediated necroptosis. These results suggest that testis aging shares a common molecular pathway with NOA, driven by chronic stress signaling and regulated cell death (SI Appendix, Fig. S12).

Fig. 5.

p-MLKL signals were associated with stress biomarkers in aging testes. (A and B) H&E staining of testis sections from human testicular torsion (n = 17) and prostate cancer patients (aging testis, n = 30) in (A). Johnsen Score evaluation of testicular torsion and prostate cancer patients based on (A) and shown in (B). (Scale bar, 100 μm.) (C and D) Immunofluorescence analysis of human testicular torsion and aging testis sections with p-MLKL antibody in (C). The number of seminiferous tubules with positive p-MLKL signals was counted based on immunofluorescence staining and quantification in (D). (Scale bar, 100 μm.) (E and F) IHC analysis of human testicular torsion and aging testis sections with G3BP1 antibody in (E). The number of seminiferous tubules with positive G3BP1 dot signal was counted based on IHC staining and quantification in (F). (Scale bar, 100 μm.) Data represent the mean ± s.e.m. ****P < 0.0001. P values were determined by two-sided unpaired Student’s t tests. (G and H) IHC analysis of human testicular torsion and aging testis sections with p-GCN2 antibody in (G). The number of seminiferous tubules with positive p-GCN2 signal was counted based on IHC staining and quantification in (H). (Scale bar, 100 μm.) Data represent the mean ± s.e.m. ****P < 0.0001. P values were determined by two-sided unpaired Student’s t tests. (I) Immunofluorescence analysis of aging testis sections (n = 5) with antibodies against p-MLKL (red) and p-GCN2 (green). (Scale bar, 50 μm.) (J and K) Analysis of p-GCN2 and p-PKR-positive testis sections from the testicular torsion (n = 17) and aging testes (n = 30) in (J). Analysis of aging testis sections with single-positive p-GCN2, single-positive p-PKR, and double-positive p-GCN2 and p-PKR in (K).

Discussion

Our findings identify ZBP1–RIPK3–dependent necroptosis as a central mechanism underlying NOA, one of the most severe forms of male infertility. The necroptosis marker phospho-MLKL was detected in 100% of NOA patient testis samples but not in any controls, indicating that this pathway is consistently activated across a clinically and etiologically heterogeneous patient population. Phospho-MLKL localized specifically to spermatogonia and Sertoli cells—two cell types essential for spermatogenesis—strongly implicating necroptotic cell death as a direct cause of germ cell loss in NOA.

Mechanistic studies in cell lines and mouse models revealed that necrotic cell death in the testis is triggered by stress granules, which act as scaffolds to recruit and activate ZBP1, presumably by the Z-form RNA coalesced by the stress granules, leading to downstream activation of RIPK3 and MLKL. SG formation itself is initiated by eIF2α kinase signaling, particularly through GCN2 and PKR, which respond to diverse stressors including heat shock, oxidative stress, and amino acid deprivation. Markers of this stress axis—including phospho-GCN2, phospho-PKR, punctate G3BP1, RIPK3, and phospho-MLKL—were consistently detected and colocalized within the same testicular cell types in NOA patients.

Considering our previous finding that a RIPK1 inhibitor effectively alleviated testicular aging in mice (11), we performed additional phospho-RIPK1 IHC staining in azoospermia testicular tissues. However, we did not detect any phospho-RIPK1 signal. Moreover, knockout RIPK1 in multiple cell lines did not impact their necroptotic response to IFN-β/HS, a cellular model for testis necroptosis during aging. We therefore concluded that the antiaging effect of the RIPK1 inhibitor used in the previous study on mouse testes is due to an unknown off-target mechanism.

Despite the clinical heterogeneity of NOA, the uniform activation of necroptosis across all patients suggests the presence of a common molecular effector pathway. Our results support a model in which diverse environmental and endogenous stress signals converge on eIF2α kinase–driven stress granule assembly, which in turn activates the ZBP1–RIPK3–MLKL cascade. ZBP1 is likely recruited to SGs via its Zα domain binding to the aggregated Z-form RNA, forming necrosomes with RIPK3 and initiating cell death. Notably, multiple eIF2α kinases were activated in NOA patient samples, with GCN2 being the most consistently observed. This raises the possibility that NOA may reflect accumulated responses to a variety of environmental or intrinsic insults, such as hyperthermia, nutrient stress, viral infections, oxidative stress, or endogenous retroelement activation. Interestingly, PERK, which mediates ER stress, does not seem to be involved in the process. We could not assess HRI activity in patient samples due to technical limitations, and its role in activating necroptosis in the testis remains to be determined.

Unlike in most cell lines, where ZBP1 expression requires induction by IFN-β, ZBP1 is constitutively expressed in the testes of both mice and humans, regardless of disease status. This constitutive expression may function as a safeguard mechanism, eliminating germ cells that accumulate endogenous Z-form RNA due to retroelement activation or other cellular stresses. However, under conditions of excessive or persistent stress, this pathway may become pathologically engaged, resulting in widespread cell loss and NOA.

Importantly, the necroptosis pathway characterized here in NOA mirrors that observed in physiological testicular aging. Phospho-MLKL, stress granules, and upstream stress markers were also detected in aging testes from older men, suggesting that chronic, low-level activation of the SG–ZBP1–RIPK3 pathway may contribute to age-associated testicular decline. These findings position NOA as a model of premature testicular aging and offer a mechanistic framework for future studies of male reproductive aging.

Methods

Mice.

The Ripk3^−/− and Zbp1^−/− mice (C57BL/6NCrl strain) were kept in our lab. The primers used for genotyping are listed below.

• Mouse Ripk3-KO-F: CAGTGGGACTTCGTGTCCG

• Mouse Ripk3-KO-R: CAAGCTGTGTAGGTAGCACATC

• Mouse Zbp1-WT-F: AGAGTTGGGGGTTCCTACCT

• Mouse Zbp1-WT-R: TGAGGGTTTTCTTGGGCACT

• Mouse Zbp1-KO-F: GTGGCTGAAGCAGGAGGATT

• Mouse Zbp1-KO-R: ATTGGTAGCCCTTGTGAGGC

Animal Husbandry.

Mice were group-housed in a 12 h light/dark (light between 08:00 and 20:00) in a temperature-controlled room (21.1 ± 1 °C) at Sironax with free access to water. The ages of mice are indicated in the figure, figure legends, or methods. All animal experiments were conducted following the Ministry of Health national guidelines for the housing and care of laboratory animals and were performed in accordance with institutional regulations after review and approval by the Institutional Animal Care and Use Committee at Sironax, Beijing.

Heatstroke Model.

Adult male C57BL/6 mice (12 wk) weighing 25 to 30 g were anesthetized by injecting 1.25% Avertin (2,2,2-tribromoethanol, M2920, Aibei Biotechnology, Nanjing, China) intraperitoneally at concentrations of 0.12 ml/10 g. Then, the lower parts of the body (hind legs, tail, and scrotum) were submerged in a thermostatically controlled water bath at 43 or 37 °C for 20 min, every other day for a total of three times (SI Appendix, Fig. S8A). Immediately after the heat stress exposure, mice were subsequently moved back to their original cages with an environmental temperature at 21.1 ± 1 °C and free access to food and water. 7 d after the final heatstroke, mice were killed and perfused with PBS. Testes were weighed and fixed in Bouin’s solution for histological and immunohistochemical assays.

Human Tissues.

Human tissue samples were dissected from human testicular torsion, azoospermia, and prostate cancer patients (n = 17, 13 to 30 y, testicular torsion patients; n = 50, 18 to 57 y, azoospermia patients; n = 30, 59 to 89 y, prostate cancer patients) and snap-frozen in liquid nitrogen and stored at –80 °C. Tissues were cut into appropriately sized pieces and placed in Bouin’s solution for preservation. After several days of Bouin’s solution fixation at room temperature, tissue fragments were transferred to 70% ethanol and stored at 4 °C.

Informed consent was obtained from all participants prior to the study, and all human samples were deidentified to the researcher of this study. The medical ethics committee of the Beijing Chao-Yang Hospital and National Institute of Biological Sciences, Beijing, China, approved the study (2022-KE-491).

Antibodies and Reagents.

Antibodies used in this study were anti-GAPDH-HRP (M171-1, MBL; WB, 1:5000), anti-Flag (F1804, Sigma-Aldrich; WB, 1:5000), anti-Human-p-RIPK1 (#44590S, Cell Signaling; IHC, 1:100), Cleaved Caspae3 (#9661L, Cell Signaling; IHC, 1:100), anti-RIPK3 (#2283, ProSci; WB, 1:1000; IHC, 1:100), anti-Mouse-RIPK3 (#95702, Cell Signaling; WB, 1:1000), anti-MLKL (ab255747; Abcam; WB, 1:1000), anti-Mouse-p-MLKL (ab196436; WB, 1:1000; IHC, 1:100), anti-Human-p-MLKL (ab187091; WB, 1:1000; IHC, 1:100), anti-Human/Mouse-ZBP1 (AG-20B-0010-C100, AdipoGen; WB, 1:1000; IHC, 1:100), anti-G3BP1 (13057-2-AP, Proteintech; WB, 1:1000; IHC, 1:100), anti-G3BP2 (PA5-53797, Invitrogen; WB, 1:1000), anti-PKR (ab184257, Abcam; WB, 1:1000), anti-PKR (Abways, CY5665; WB, 1:1000), anti-PERK (377400, Santa Cruz Biotechnology; WB, 1:1000), anti-GCN2 (3302, Cell Signaling; WB, 1:1000), anti-PIWIL4 (PA5-31448, Thermo; IHC, 1:500), anti-Sox9 (ab185966, Abcam; IHC, 1:100), DDX4 (ab13840, Abcam; IHC, 1:100), anti-p-GCN2(T899) (ThermoFisher, PA5-105886; WB, 1:1000; IHC, 1:500), anti-p-PKR (Abways, CY5271; WB, 1:1000; IHC, 1:100), anti-p-PERK(T981) (CUSABIO, CSB-PA072558; IHC, 1:100), Donkey anti-Mouse, Alexa Fluor 488 (Thermo Fisher, A-21202), Donkey anti-Mouse, Alexa Fluor 555 (Thermo Fisher, A-31570), Donkey anti-Rabbit, Alexa Fluor 488 (Thermo Fisher, A-21206), and Donkey anti-Rabbit, Alexa Fluor 555 (Thermo Fisher, A-31572).

Reagents used in this study were INF-β (Sino Biological, 50708-MCCH), Baricitinib (JAK inhibitor, MedChemExpress, HY-15315), DOX (Sigma, WXBC9363V), NaAsO₂ (Sodium arsenite, Innochem, A25410), Halofuginone (MedChemExpress, S8144), 8-Azaadenosine (MedChemExpress, HY-115686), Tunicamycin (MedChemExpress, S7894), and Lambda Protein Phosphatase (Cat: P0753S, 400,000 U/mL). Phosphorylated MLKL peptides (GYQVKLAGFELRKTQpTpSMSLGTTREKTDRVKS) were synthesized by GenScript.

Constructs.

psPAX2 and pMD2.G constructs were kept in our lab. Full-length mouse ZBP1 and truncated ZBP1[lacked two Zα domain or C-terminal domain (315 to 411)] were subcloned into the pWPI (GFP-tagged) vector to generate pWPI-mZBP1 (WT, ∆Zα, and ∆C) construct. The Quickchange Site-Directed Mutagenesis Kit was used to generate pWPI-mZBP1^mut (two points mutation within two RHIM domain) construct.

The gRNAs for targeting mouse Zbp1, G3bp1/2, Pkr, Perk, Gcn2, and Hri were designed and were cloned into the gRNA-Cas9 expression plasmid pX458-GFP to generate pX458-GFP-ZBP1/G3BP1/G3BP2/PKR/PERK/GCN2/HRI construct. The sequences used for gRNAs targeting are listed below.

• Mouse ZBP1-gRNA: GAAGATCTACCACTCACGTC

• Mouse G3BP1-gRNA: ATGTTCACAACGACATCTTC

• Mouse G3BP2-gRNA: AAGCTCCCGAGTATTTGCAC

• Mouse HRI-gRNA: TCGAAGCACAAACGTCACGC

• Mouse PKR-gRNA: TTGTTCGTTGGTAACTACAT

• Mouse PERK-gRNA: CTCGAATCTTCCTACAAGTT

• Mouse GCN2-gRNA: AAAGCCCGGACATACTCCTC

Cell Culture and Stable Cell Lines.

All cells were cultured at 37 °C with 5% CO₂. All cell lines were cultured as follows: HEK293 T (293T) were obtained from ATCC and cultured in DMEM (HyClone). Mouse embryonic fibroblasts (MEF), MEF(Ripk3^−/−), and HeLa-RIPK3/TetOn-ZBP1(15) cells were cultured in DMEM. L929, L929(Ripk1^−/−), L929(Ripk3^−/−), L929(Mlkl^−/−), L929(Zbp1^−/−), L929(Pkr^−/−), L929(Pkr^−/−Hri^−/−), L929(Pkr^−/−Hri^−/−Perk^−/−), and L929(Pkr^−/−Hri^−/−Perk^−/−Gcn2^−/−) were cultured in DMEM. GC-2spd(ts) and 15P-1 cells were obtained from ATCC and cultured in DMEM. GC-2spd(Zbp1^−/−) and 15P-1(Zbp1^−/−) were cultured in DMEM. L929(Zbp1^−/−) cells were infected with virus encoding ZBP1 (WT, RHIM^mut, ∆Zα, and ∆C) to establish the L929(Zbp1^−/−)-ZBP1, L929(Zbp1^−/−)-ZBP1(RHIM^mut), L929(Zbp1^−/−)-ZBP1(∆Zα) and L929(Zbp1^−/−)-ZBP1(∆C) cell lines. L929(Ripk3^−/−) cells were infected with virus encoding HA-3 × Flag-mRIPK3, and GFP-positive live cells were sorted to establish the L929(Ripk3^−/−)-HA-3 × Flag-mRIPK3 cell lines. All media were supplemented with 10% FBS (Thermo Fisher) and 100 units/ml penicillin/ streptomycin (Thermo Fisher). MA-10 were obtained from ATCC and cultured in DMEM:F12 (HyClone, additional 20 mM HEPES, horse serum to a final concentration of 15%).

Cell Survival Assay.

The cell survival assay was performed using the Cell Titer-Glo Luminescent Cell Viability Assay kit. A Cell Titer-Glo assay (Promega, G7570) was performed according to the manufacturer’s instructions. Luminescence was recorded with a Tecan GENios Pro plate reader.

CRISPR/Cas9 Knockout Cells.

10 μg of pX458-GFP-ZBP1/G3BP1/G3BP2/PKR/PERK/GCN2/HRI plasmid was transfected into 1 × 10⁵ L929 cells using the Neon™ transfection system (Invitrogen™, MPK5000) by following the manufacturer’s instructions. 3 d after the transfection, GFP-positive live cells were sorted into single clones by using a BD FACSArial cell sorter. The single clones were cultured into 96-well plates for another 10 to 14 d or longer, depending upon the cell growth rate. The anti-ZBP1/G3BP1/G3BP2/PKR /PERK/GCN2 immunoblotting was used to screen for the L929(Zbp1^−/−), L929(Pkr^−/−), L929(Pkr^−/−Hri^−/−), L929(Pkr^−/−Hri^−/−Perk^−/−), and L929(Pkr^−/−Hri^−/−Perk^−/−Gcn2^−/−) clones. Genome type of the knockout cells was determined by DNA sequencing.

Cell Stress Exposure.

Cultured L929, L929(Pkr^−/−), L929(Pkr^−/−Hri^−/−), L929(Pkr^−/−Hri^−/−Perk^−/−), L929(Pkr^−/−Hri^−/−Perk^−/−Gcn2^−/−), L929(G3bp1^−/− G3bp2^−/−), L929(Zbp1^−/−), L929(Ripk3^−/−), L929(Mlkl^−/−), MEF, MEF(Ripk3^−/−), HeLa-RIPK3/TetOn-ZBP1, GC-2spd, 15P-1, GC-2spd(Zbp1^−/−), 15P-1(Zbp1^−/−), and MA-10 cells were pretreated with DMSO, Dox, or recombinant mouse INF-β (50708-MCCH, Sino Biological, Beijing, China) at 10 ng/mL for 18 h before were trypsinized and resuspended with the complete medium. Cells were placed in a water bath with a temperature at 43 or 37 °C for the indicated times, and then incubated at 37 °C and humidified 5% CO₂ for the indicated time periods. Finally, cell lysates and supernatants were collected at the indicated time points after heat stress for ATP activity, western blot, immunoprecipitation, or immunofluorescence assay.

Cultured L929, L929(Pkr^−/−Hri^−/−Perk^−/−Gcn2^−/−), L929(G3bp1^−/−G3bp2^−/−), L929(Zbp1^−/−), L929(Ripk3^−/−), and L929(Mlkl^−/−) cells were treated with NaAsO₂(30 μM) and NaAsO₂(30 μM)+IFN-β(10 ng/mL) for 18 h; Halofuginone(500 nM) and Halofuginone(500 nM)+IFN-β(10 ng/ml) for 42 h; 8-Azaadenosine (20 μM) and 8-Azaadenosine(20 μM)+IFN-β(10 ng/ml) for 46 h; and Tunicamycin (2.5 μM) and Tunicamycin(2.5 μM)+IFN-β(10 ng/mL) for 30 h. The intracellular ATP levels were measured by Cell Titer-Glo.

Western Blotting.

Cell pellet samples were collected and resuspended in lysis buffer (100 mM Tris-HCl, pH 7.4, 100 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA, Roche complete protease inhibitor set, and Sigma phosphatase inhibitor set), incubated on ice for 30 min, and centrifuged at 20,000×g for 30 min. The supernatants were collected for western blotting.

Testis tissues were ground and resuspended in lysis buffer, homogenized for 30 s with a Paddle Blender (Prima, PB100), incubated on ice for 30 min, and centrifuged at 20,000×g for 30 min. The supernatants were collected for western blotting.

Immunoprecipitation.

The cells were cultured on 15-cm dishes and grown to confluence. At 70% confluence, cells underwent treatment for indicated time periods based on experimental design. Then, cells were washed once with PBS and harvested by scraping and centrifugation at 800×g for 5 min. The harvested cells were washed with PBS and lysed for 30 min on ice in the lysis buffer (100 mM Tris-HCl, pH 7.4, 100 mM NaCl, 10% glycerol, 1% Triton X-100, 2 mM EDTA, Roche complete protease inhibitor set, and Sigma phosphatase inhibitor set). Cell lysates were then spun down at 12,000×g for 20 min. The soluble fraction was collected, and the protein concentration was determined by Bradford assay. Cell extract was mixed with anti-Flag affinity gel (Sigma-Aldrich, A2220) in a ratio of 1 mg of extract per 30 μL of agarose. After overnight rocking at 4 °C, the beads were pelleted at 2,500×g for 3 min and washed with lysis buffer three times. The beads were then eluted with 0.5 mg/mL of the corresponding antigenic peptide for 6 h or directly boiled in 1 × SDS loading buffer (125 mM Tris, pH 6.8, 2% 2-mercaptoethanol, 3% SDS, 10% glycerol, and 0.01% bromophenol blue).

IHC and Immunofluorescence.

Paraffin-embedded specimens were sectioned to a 5 μm thickness and were then deparaffinized, rehydrated, and stained with hematoxylin and eosin (H&E) using standard protocols. For the preparation of the IHC samples, sections were dewaxed, incubated in boiling citrate buffer solution for 15 min in plastic dishes, and subsequently allowed to cool down to room temperature over 3 h. Endogenous peroxidase activity was blocked by immersing the slides in hydrogen peroxide buffer (10%, Sinopharm Chemical Reagent) for 15 min at room temperature and was then washed with PBS. Blocking buffer (1% bovine serum albumin in PBS) was added, and the slides were incubated for 2 h at room temperature. Primary antibody against human IgG, p-mouse-MLKL, Cleaved-caspase3, p-RIPK1, G3BP1, p-GCN2, p-PKR, p-PERK, ZBP1, or p-Human-MLKL(p-MLKL) was incubated overnight at 4 °C in PBS. After three washes with PBS, slides were incubated with secondary antibody (polymer-horseradish-peroxidase-labeled anti-rabbit, Sigma) in PBS. After a further three washes, slides were analyzed using a diaminobutyric acid substrate kit (Thermo Fisher). Slides were counterstained with hematoxylin and mounted in neutral balsam medium (Sinopharm Chemical).

IHC analysis for SOX9/PIWIL4/DDX4/G3BP1/p-GCN2/p-PKR or p-MLKL/RIPK3 was performed using an antibody against SOX9/ PIWIL4/DDX4/G3BP1/p-GCN2/p-PKR and p-MLKL/RIPK3. Primary antibody against SOX9/PIWIL4/DDX4/G3BP1/p-GCN2/p-PKR was incubated overnight at 4 °C in PBS. After three washes with PBS, slides were incubated with DyLight-488/555 conjugated donkey anti-rabbit/mouse secondary antibodies (Life) in PBS for 8 h at 4 °C. After a further three washes, slides were incubated with p-MLKL/RIPK3 antibody overnight at 4 °C in PBS. After a further three washes, slides were incubated with DyLight-488/555 conjugated donkey anti-mouse/rabbit secondary antibodies (Life) for 2 h at room temperature in PBS. After a further three washes in PBS, the cell nuclei were then counterstained with DAPI (Invitrogen) in PBS. Fluorescence microscopy was performed using a Nikon A1-R confocal microscope.

Immunofluorescence analysis for G3BP1 or RIPK3 was performed using an antibody against G3BP1 and RIPK3. Primary antibody against G3BP1 was incubated overnight at 4 °C in PBS. After three washes with PBS, slides were incubated with DyLight-555 conjugated donkey anti-rabbit/mouse secondary antibodies (Life) in PBS for 8 h at 4 °C. After a further three washes, slides were incubated with RIPK3 antibody overnight at 4 °C in PBS. After a further three washes, slides were incubated with DyLight-488 conjugated donkey anti-mouse/rabbit secondary antibodies (Life) for 2 h at room temperature in PBS. After a further three washes in PBS, the cell nuclei were then counterstained with DAPI (Invitrogen) in PBS. Fluorescence microscopy was performed using a Nikon A1-R confocal microscope.

Phosphorylated MLKL Peptides Competition Assay.

Paraffin-embedded specimens were sectioned to a 5 μm thickness. Then sections were dewaxed, incubated in boiling citrate buffer solution for 15 min in plastic dishes, and subsequently allowed to cool down to room temperature over 3 h. Endogenous peroxidase activity was blocked by immersing the slides in Hydrogen peroxide buffer (10%, Sinopharm Chemical Reagent) for 15 min at room temperature and were then washed with PBS. Blocking buffer (1% bovine serum albumin in PBS) was added, and the slides were incubated for 2 h at room temperature. Primary antibody against p-Human-MLKL(p-MLKL) was incubated alone or coincubated with phosphorylated MLKL peptides overnight at 4 °C in PBS. After three washes with PBS, slides were incubated with secondary antibody (polymer-horseradish-peroxidase-labeled anti-rabbit, Sigma) in PBS. After a further three washes, slides were analyzed using a diaminobutyric acid substrate kit (Thermo Fisher). Slides were counterstained with hematoxylin and mounted in neutral balsam medium (Sinopharm Chemical).

Alkaline Phosphatase Dephosphorylation Assay.

The 5 μm paraffin sections were deparaffinized using a robotic autostainer (Leica Microsystems) and pretreated with a high pH (pH 9) buffer. After a 10-min incubation with Peroxidase-Blocking Reagent, add 100 μL of 10X NEBuffer for Protein Metallo Phosphatases (PMP) (Cat: B0761S), 100 μL of 10 mM MnCl2 (Cat: B1761S), and 1 μL of Lambda Protein Phosphatase (Cat: P0753S, 400,000 U/mL) to make a total reaction volume of 500 μL. Incubate the samples at 37 °C for 60 min.

Next, add Calf Intestinal Alkaline Phosphatase (CIAP) (Cat: M2825, 1,000 U/mL). Dilute CIAP in CIAP 1X Reaction Buffer to a final concentration of 500 U/mL for immediate use. Incubate the tissue sections at 37 °C for 24 h.

Following this, the samples were blocked with SuperBlock™ (TBS) Blocking Buffer for 60 min. The tissues were then incubated overnight in a humid chamber at room temperature (RT) with the primary p-MLKL antibody overnight at 4 °C in PBS. After three washes with PBS, slides were incubated with secondary antibody (polymer-horseradish-peroxidase-labeled anti-rabbit, Sigma) in PBS. After a further three washes, slides were analyzed using a diaminobutyric acid substrate kit (Thermo Fisher). Slides were counterstained with hematoxylin and mounted in neutral balsam medium (Sinopharm Chemical).

qRT-PCR.

Cell/Tissue total RNA was extracted with the FastPure® Cell/Tissue Total RNA Isolation Kit (RC112, Vazyme Biotech, Nanjing, China), and cDNA was prepared with HiScript® III RT SuperMix for qPCR Kit (R323, Vazyme Biotech, Nanjing, China) according to the manufacturer’s protocol. Quantitative RT-PCR of ZBP1 was performed with Taq Pro Universal SYBR qPCR Master Mix (Q712, Vazyme Biotech, Nanjing, China) and the primers as follows:

• ZBP1-forward: CAAGTCCTTTACCGCCTGAAG

• ZBP1-reverse: TCGTCATTCCCAGAGCCTTG

• GAPDH-forward: GTGCTGAGTATGTCGTGGAGTC

• GAPDH-reverse: GTGGTTCACACCCATCACAAAC

Data were normalized by GAPDH expression, and relative expression changes were analyzed according to the 2^-ΔΔCT method.

Statistical Analysis.

Statistical tests were used for every type of analysis. The data meet the assumptions of the statistical tests described for each figure. Results are expressed as the mean ±s.e.m or SD. Each experiment was repeated at least two times independently to ensure reproducibility. Differences between experimental groups were assessed for significance using a two-tailed unpaired Student’s t test using GraphPad prism 10. The *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001 levels were considered significant. P > 0.05 was considered not significant (NS).

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information. Correspondence and requests for materials should be addressed to D.L, L.T., and X.W.