Author's reply to the Letters regarding “Cysteine metabolism dysregulation promotes ferroptosis in osteoarthritis: Insights from multi-species bioinformatics and experimental validation” Authors: Luxi Zheng, Zhengmeng Yang, Sihui Chen, Gongli Cai, Nan Hou, Wenxuan Lin, Carina Hey Pui Cheung, Huisheng Zhou, Junjie Chen, Yican Wang, Huifeng Chen, Yaofeng Wang, Micky D Tortorella, Jinyu Zhu, Lu Feng, Gang Li Published: 2026 Dear Editor, We thank Dr. Ni, Dr. Liu, and colleagues for their thoughtful reading of our article [1] and for their constructive commentary regarding sex as a biological variable (SABV) in osteoarthritis (OA) and ferroptosis research. They raise an important and timely point concerning the demographic composition of our human cartilage multi-omics validation cohort (consisting of three elderly postmenopausal female patients) and correctly highlight the potential influence of estrogenic status and sexual dimorphism on cysteine metabolism and ferroptotic signaling. We welcome this opportunity to address their comments, clarify the scope of our findings, and present additional re-analyses of public datasets to address the generalizability of the cysteine metabolism-ferroptosis axis across sexes. We fully acknowledge the methodological limitation pointed out by Dr. Ni and Dr. Liu that our primary human cartilage proteomic and metabolomic validation was indeed performed on a highly specific clinical subset (three postmenopausal females aged 77-80 years, Kellgren–Lawrence grades III–IV). As the commentators accurately note, core ferroptosis regulators, including GPX4 and SLC7A11, are subject to hormonal modulation, particularly by 17β-estradiol [2]. Therefore, from a strict clinical validation view, the multi-omics profile derived from this small human cohort should be interpreted as direct evidence within this specific subpopulation. However, we wish to emphasize that the overall conclusions of our study do not rely solely on these three human tissue samples. The mechanistic framework of the "cysteine metabolism-ferroptosis axis" in OA was established through a multi-tiered, cross-species investigation. The cross-species transcriptomic integration across human, murine, and cellular models consistently identified dysregulated cysteine metabolism and down-regulated ferroptosis-defense machinery as central features of OA chondrocytes. Also, our primary IL-1β induced OA chondrocyte model demonstrated that pharmacological or molecular targeting of cysteine catabolism directly modulated intracellular GSH levels, reactive oxygen species (ROS) accumulation, lipid peroxidation, and ferroptotic cell death in vitro. Furthermore, surgical DMM-induced OA mouse models validated the therapeutic efficacy of L-cysteine supplementation in mitigating cartilage degeneration and ferroptosis markers in vivo. To directly address the commentators' concern regarding whether the cysteine metabolism-ferroptosis axis operates independently of sex, we conducted a post-hoc sex-stratified re-analysis of publicly available human OA cartilage transcriptomic datasets (GSE114007, [3]) containing both male and female patients with annotated clinical metadata. This dataset comprised 18 normal and 20 OA cartilage samples. After stratification, 13 normal and 8 OA samples were from male patients, and 5 normal and 12 OA samples were from female patients. Raw counts were analyzed with DESeq2 using an OA versus normal comparison (p < 0.05, |fold change|>1.2). Consistent downregulation in OA was observed in both sexes for SLC3A2 (male log2FC = −1.57, p = 3.8 × 10−7; female log2FC = −1.90, p = 1.2 × 10−6), GPX4 (male log2FC = −0.82, p = 1.6 × 10−3; female log2FC = −0.76, p = 4.9 × 10−2), and CTH (male log2FC = −2.07, p = 5.1 × 10−9; female log2FC = −2.47, p = 2.1 × 10−12) (Fig. 1). SLC7A11 and GCLC did not reach significance in either sex, indicating that the conserved response centers on the SLC3A2-GPX4 axis rather than uniform suppression of every cysteine-handling gene. Across the KEGG ferroptosis pathway gene set, six genes were significantly changed in both sexes, and all six changed in the same direction. No ferroptosis-related gene showed opposite significant directions between sexes (Spearman r = 0.81 for male versus female log2FC). Sex-specific differences were still observed in individual genes, including ACSL4 down-regulation in males only and NCOA4/STEAP3/CYBB up-regulation in females only, consistent with the estrogen-regulated baseline susceptibility cited by Dr. Ni and Dr. Liu. Overall, the directionality of cysteine metabolism dysregulation and ferroptosis activation remained highly consistent between male and female OA patients. These data are summarized in Table 1, Table 2, Table 3. These additional bioinformatic findings suggest that while sexual dimorphism and postmenopausal estrogen deficiency may exacerbate or modulate the severity of ferroptosis in cartilage, cysteine metabolism dysregulation represents a fundamental, conserved pathophysiological hallmark of OA that transcends single-sex boundaries. Sex-stratified ferroptosis-related gene expression changes in human OA cartilage (GSE114007) A. RNA-seq raw counts from 18 normal and 20 OA human knee cartilage samples were analyzed separately by sex (male: 13 normal and 8 OA; female: 5 normal and 12 OA) with DESeq2 using an OA versus normal contrast. Differentially expressed genes were defined by p < 0.05 and |fold change| > 1.2 (equivalent to |log2FC| > 0.263). Key cysteine metabolism and ferroptosis genes with log2 fold changes shown for male (teal) and female (red) samples; positive values indicate higher expression in OA, negative values indicate lower expression in OA, and dashed lines mark the |fold change| = 1.2 threshold. Corresponding p-values and normalized mean counts are provided in Table 1. B. Significant ferroptosis DEGs in female samples, ranked by log2 fold change. C. Significant ferroptosis DEGs in male samples, ranked by log2 fold change. In panels B and C, red bars indicate up-regulation in OA and teal bars indicate down-regulation in OA; dashed lines mark the |fold change| = 1.2 threshold. Table 1Sex-stratified differential expression summary of human knee cartilage in GSE114007.Table 1SexNormal samplesOA samplesGenes testedDEGsUp in OADown in OATotal samplesMale1382262134281484194421Female5122242350042946205817 Sex-stratified differential expression summary of human knee cartilage in GSE114007. Table 2Sex-stratified DESeq2 results for key cysteine metabolism and ferroptosis genes in GSE114007.Table 2SYMBOLlog2FC_Malepvalue_Malesignificant_Maledirection_Malelog2FC_Femalepvalue_Femalesignificant_Femaledirection_Femalesame_signboth_sigconsistencyFC_MaleFC_FemaleSLC7A11−0.11790.8322FALSENot significant0.51380.4799FALSENot significantFALSEFALSENeither significant0.92151.4278GPX4−0.82420.0016TRUEDown in OA−0.75590.0494TRUEDown in OATRUETRUEBoth significant, same direction0.56480.5922CTH−2.07120.0000TRUEDown in OA−2.46710.0000TRUEDown in OATRUETRUEBoth significant, same direction0.23800.1809ACSL4−1.54280.0011TRUEDown in OA−0.13670.6331FALSENot significantTRUEFALSEOnly one sex significant0.34320.9096LPCAT3−0.51470.0645FALSENot significant−0.76440.0412TRUEDown in OATRUEFALSEOnly one sex significant0.69990.5887TFRC−0.96800.0137TRUEDown in OA−0.48020.2481FALSENot significantTRUEFALSEOnly one sex significant0.51120.7169NCOA40.25020.1615FALSENot significant0.57240.0002TRUEUp in OATRUEFALSEOnly one sex significant1.18941.4869HMOX1−0.63650.1958FALSENot significant−0.55650.2190FALSENot significantTRUEFALSENeither significant0.64330.6800FTH1−0.12490.7761FALSENot significant−0.81950.1849FALSENot significantTRUEFALSENeither significant0.91710.5666FTL0.25360.6657FALSENot significant−0.36320.6553FALSENot significantFALSEFALSENeither significant1.19220.7774GCLC0.01660.9411FALSENot significant0.17710.5156FALSENot significantTRUEFALSENeither significant1.01161.1306GCLM−0.38940.4218FALSENot significant0.01570.9701FALSENot significantFALSEFALSENeither significant0.76341.0110GSR0.21050.6441FALSENot significant0.51210.2742FALSENot significantTRUEFALSENeither significant1.15711.4261NQO10.70940.0140TRUEUp in OA1.17820.0003TRUEUp in OATRUETRUEBoth significant, same direction1.63512.2630NFE2L2−0.62010.1871FALSENot significant−0.61900.2156FALSENot significantTRUEFALSENeither significant0.65060.6511SLC3A2−1.57030.0000TRUEDown in OA−1.90010.0000TRUEDown in OATRUETRUEBoth significant, same direction0.33670.2679STEAP30.59420.0917FALSENot significant1.45600.0000TRUEUp in OATRUEFALSEOnly one sex significant1.50962.7435SAT1−1.60260.0020TRUEDown in OA−0.40820.4230FALSENot significantTRUEFALSEOnly one sex significant0.32930.7535SAT2−0.20750.5266FALSENot significant−1.08770.0079TRUEDown in OATRUEFALSEOnly one sex significant0.86600.4705MAP1LC3B−1.07390.0000TRUEDown in OA−0.80960.0177TRUEDown in OATRUETRUEBoth significant, same direction0.47500.5705CYBB1.14020.2172FALSENot significant7.96060.0000TRUEUp in OATRUEFALSEOnly one sex significant2.2041249.1105SLC39A14−1.53150.0017TRUEDown in OA−0.70630.0714FALSENot significantTRUEFALSEOnly one sex significant0.34590.6129SLC39A80.45190.1023FALSENot significant1.06770.0061TRUEUp in OATRUEFALSEOnly one sex significant1.36782.0962 Sex-stratified DESeq2 results for key cysteine metabolism and ferroptosis genes in GSE114007. Table 3Directional consistency of ferroptosis-related gene expression changes between male and female OA patients in GSE114007.Table 3ConsistencyGene_countBoth significant, same direction6Neither significant26Only one sex significant13male_only_sig5female_only_sig8both_sexes_sig6pearson_all0.6782439774spearman_all0.8055335968pearson_significant0.7165165607 Directional consistency of ferroptosis-related gene expression changes between male and female OA patients in GSE114007. We sincerely appreciate the constructive suggestions provided by the commentators. Moving forward, we fully agree that incorporating sex-matched male cartilage specimens and conducting sex-stratified translational cohorts are mandatory steps for validating metabolism-targeted therapies in OA. In our ongoing and future research projects, we are prospectively collecting gender-balanced, age-matched human articular cartilage specimens (spanning both male and female patients across mild to severe KL grades) to perform comprehensive sex-stratified transcriptomic, proteomic, and metabolomic profiling. In conclusion, we thank Dr. Ni and Dr. Liu once again for their valuable insights. Their commentary highlights sex as a crucial biological dimension, which enriches the interpretation of our work. While our original human multi-omics validation cohort was demographically restricted to elderly female patients, our broader cross-species experimental evidence and newly performed sex-stratified transcriptomic re-analyses support the broader relevance of the cysteine metabolism–ferroptosis axis in OA. We believe that incorporating SABV into metabolic OA research will ultimately enhance translational rigor and paved the way toward personalized, sex-informed therapeutic strategies. Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Contributor Information Luxi Zheng, Email: ZHENG-luxi@link.cuhk.edu.hk. Zhengmeng Yang, Email: zhengmengyang@link.cuhk.edu.hk. Sihui Chen, Email: chensihuihuihui@163.com. Gongli Cai, Email: gonglicai@link.cuhk.edu.hk. Nan Hou, Email: nan.hou@link.cuhk.edu.hk. Wenxuan Lin, Email: linwx@alum.ubc.ca. Carina Hey Pui Cheung, Email: Carinaa@connect.hku.hk. Huisheng Zhou, Email: huishengzhou@126.com. Junjie Chen, Email: cjj9760@163.com. Yican Wang, Email: yicanwang3-c@my.cityu.edu.hk. Huifeng Chen, Email: huifechen4-c@my.cityu.edu.hk. Yaofeng Wang, Email: yaofeng.wang@hkisi-cas.org.hk. Micky D. Tortorella, Email: m.tortorella@crmh-cas.org.hk. Jinyu Zhu, Email: zhujinyu@hotmail.com. Lu Feng, Email: lufeng@link.cuhk.edu.hk. Gang Li, Email: gang.li6@siat.ac.cn. References