In a recent study published in Nature Metabolism, Li et al. 1 identified mitochondrial proteotoxicity due to impaired protein folding as a key driver of β-cell failure in type 2 diabetes (T2D), shifting the focus from canonical endoplasmic reticulum (ER) stress paradigms. The chaperone function of the LONP1–mitochondrial heat shock protein 70 (mtHSP70) axis was found to be essential for maintaining mitochondrial proteostasis, offering a novel therapeutic target for preserving β-cell function in diabetic patients. The global prevalence of T2D is projected to exceed 1.3 billion by 2050, underscoring the urgent need to identify novel mechanisms of β-cell failure and druggable targets. Large-scale genome-wide association study (GWAS) and multi-omics technologies have advanced our understanding of T2D's genetic heterogeneity. A trans-ancestry meta-analysis of over 2.5 million individuals identified eight mechanistic clusters, each with distinct clinical manifestations and complication risks, highlighting the necessity for targeted therapies to enable precise treatment. While ER stress has long been implicated in β-cell dysfunction, Li et al. provide compelling evidence that mitochondrial proteotoxicity represents an earlier and more prominent event in human T2D islets. Using unbiased proteomics, they found significant enrichment of insoluble mitochondrial protein aggregates in islets from T2D donors, a signature distinct from ER protein misfolding. The mitochondrial protease LONP1 was identified as a crucial guardian against this proteotoxicity, with its expression notably reduced in T2D β-cells. Using β-cell-specific Lonp1 knockout mice, the authors demonstrated that LONP1 deficiency recapitulates key features of human T2D islets observed in their proteomic analysis, including accumulation of misfolded mitochondrial proteins, bioenergetic deficits, oxidative stress, and β-cell apoptosis. Direct causal evidence in human islets, however, remains to be established. Most importantly, LONP1 protects β-cells via a protease-independent, chaperone-like function by forming a complex with mitochondrial HSP70. The structural basis involves LONP1's chaperone domain interacting with mtHSP70's substrate-binding domain, though the regulation of this complex under diabetic stress requires further investigation. Furthermore, the study linked inadequate adaptive response in T2D to downregulation of ATF5, a known regulator of the mitochondrial unfolded protein response and direct transcriptional activator of LONP1 (Figure 1). However, to fully appreciate the implications of Li et al.'s findings on LONP1, they must be contextualized within the highly integrated mitochondrial quality control (MQC) network. Mitochondria maintain proteostasis through a coordinated system involving compartment-specific proteases, chaperones, and organellar dynamics. Furthermore, the inner membrane protein TMBIM5 directly binds and inhibits the m-AAA protease, thereby coupling mitochondrial energy status to protein turnover 2. This raises the intriguing possibility that LONP1 activity itself may be subject to similar metabolic regulation, extending beyond the transcriptional control mediated by ATF5 as identified by Li et al. Indeed, the mitochondrial unfolded protein response (UPRmt) network is orchestrated in a multi-layered manner: PERK-eIF2α signaling induces ATF4, which cooperates with CHOP to regulate ATF5 expression. Notably, CHOP exhibits a dual nature, transient activation promotes homeostasis, whereas chronic overactivation triggers apoptosis. As an additional layer of complementary quality control, PINK1-mediated mitophagy plays a critical role 3. Single-cell analysis has identified PINK1 as the strongest predictor of β-cell health in type 2 diabetes, suggesting that LONP1-mediated protein folding and PINK1-mediated organellar clearance represent sequential, interdependent steps within a unified MQC process that becomes disrupted in diabetes. Therefore, understanding β-cell failure in T2D requires examining LONP1 function within this dynamic and interconnected MQC network. This study establishes the LONP1-mtHSP70 axis as a critical new node in the pathogenesis of T2D. As a core housekeeping protein essential for maintaining mitochondrial integrity in all cell types, LONP1 likely has functional significance that extends beyond the pancreatic β-cell. Emerging evidence indicates that LONP1 dysfunction in peripheral insulin-sensitive tissues also contributes to systemic metabolic regulation. Downregulation of LONP1 in the liver impairs hepatic insulin sensitivity and promotes gluconeogenesis. In adipose tissue, LONP1 expression correlates with pathways of glucose and lipid metabolism and appears to be protective against metabolic dysfunction. Age-related decline in LONP1 in the kidney contributes to tubular cell senescence and fibrosis. Even in skeletal muscle, where LONP1 deficiency produces a paradoxical phenotype characterized by mitochondrial proteostasis impairment accompanied by resistance to diet-induced obesity and improved systemic insulin sensitivity, this apparent systemic benefit comes at the expense of local muscle health 4. Together, these findings suggest that LONP1 dysfunction may amplify the systemic metabolic defects of T2D by creating a “double hit” that simultaneously impairs insulin secretion from β-cells and insulin sensitivity in peripheral tissues. Therefore, extending the investigation of the LONP1-mtHSP70 axis to peripheral insulin-sensitive tissues will not only help test whether this mechanism represents a common T2D pathway across different tissues but also provide a more systematic perspective for understanding the synergistic relationship between β-cell dysfunction and peripheral insulin resistance. While Li et al. demonstrate distinct ER stress and mitochondrial proteotoxicity pathways in β-cells, these organelles are functionally coupled via mitochondria-associated ER membranes (MAMs). In β-cells, MAMs are essential for glucose-stimulated insulin secretion: acute glucose enhances MAM formation and ER-mitochondria Ca²⁺ transfer to boost ATP production, while chronic glucotoxicity paradoxically increases MAM tethering but impairs functional Ca²⁺ exchange, resulting in insulin secretion defects. Recent evidence reveals potential LONP1-MAMs crosstalk. PERK at MAMs induces ATF4/ATF5 to activate LONP1 transcription, while in cardiomyocytes 5, LONP1 localizes to MAMs under ER stress to regulate inter-organelle protein homeostasis. Whether this axis operates in β-cells, and whether stress-induced LONP1 localization to MAMs modulates its chaperone activity, remains to be established. MAMs dysfunction extends beyond β-cells. In adipocytes, disrupted MAMs integrity impairs differentiation and induces insulin resistance. In diabetic kidney disease, MAMs dysregulation promotes podocyte injury and tubular fibrosis. These findings support combination therapies targeting inter-organellar crosstalk. For example, tauroursodeoxycholic acid (TUDCA), an ER chemical chaperone, improves β-cell survival in diabetic models and enhances insulin sensitivity in obese humans. 4-phenylbutyrate (4-PBA), another ER chaperone, prevents amyloid-induced β-cell dysfunction in hIAPP transgenic mice. Pairing these ER chaperones with mitochondrial proteostasis enhancers, such as NAD⁺ precursors to activate UPRmt or LONP1-mtHSP70 stabilizers, may synergistically protect β-cells under metabolic stress. In summary, the study by Li et al. represents a seminal advancement in the field, effectively reframing a core aspect of type 2 diabetes etiology as a disorder of mitochondrial proteostasis. By identifying the chaperone function of the LONP1-mtHSP70 complex, the authors provide a robust scientific foundation for exploring novel therapeutic strategies aimed at preserving β-cell function. Nevertheless, the translational path forward is laden with challenges, including issues of target specificity, system-level integration, and the inherent complexity of mitochondrial quality control networks. Future efforts should focus on rigorously addressing the pivotal questions raised by this work. Can cell-type selective modulation of this pathway be achieved? How can the integrated UPRmt network be safely engaged in the context of a multi-organ disease? Might combination therapies targeting interorganellar crosstalk, such as those involving ER chaperones and mitochondrial proteostasis enhancers, yield transformative clinical outcomes? By opening this crucial new front in the fight against diabetes, the study by Li et al. calls for a balanced approach that combines innovative biology with therapeutic prudence. Junlin Wei: conceptualization, writing – original draft, visualization. Fang Wang: supervision, writing – review and editing, funding acquisition. Both authors have read and approved the final manuscript. The authors thank BioRender (www.biorender.com) for the creation of scientific illustrations used in the preparation of this manuscript. This work was supported by the Natural Science Foundation of Hunan province (2024JJ5520); the Changsha Municipal Natural Science Foundation (kq2403054). The authors have nothing to report. The authors have nothing to report. The authors declare no conflicts of interest. The authors have nothing to report.
Wei et al. (Sun,) studied this question.