Cilia are complex cellular appendages that are composed of membranes, soluble compartments, axonemes, basal bodies, and ciliary tips (1). As both mechanical and chemical sensors, primary cilia are integral in interpreting the extracellular environment and orchestrating intricate signaling pathways that are essential for the development and maintenance of cellular homeostasis (2, 3). These cilia play crucial roles in modulating key signal transduction mechanisms that govern a wide range of cellular processes. These include the transmission of signals from molecules (4), and platelet-derived growth factor (PDGF) (5). Deficiencies in primary cilia structure and length can disrupt crucial signaling pathways and cellular functions, leading to a group of diseases that are known as ciliopathies (6). In the kidneys, these ciliopathies often result from genetic mutations that impair the ability of the primary cilia to sense fluid flow and signal effectively (6, 7). Such dysfunctions are linked to renal disorders, including polycystic kidney disease and nephronophthisis that are characterized by cyst development and advancing renal failure (8, 9).

Ferroptosis describes a distinct form of cell death, separate from traditional apoptosis and necrosis (10). Ferroptotic cell death process is intricately linked to intracellular oxidative stress, highlighting the critical need for a balanced interplay between oxidative damage and antioxidant defenses (10, 11). Accumulation of iron in cells contributes to the production of reactive oxygen species (ROS) through the Fenton reaction. The ROS attacks polyunsaturated fatty acid-rich phospholipids in cell membranes, leading to lipid peroxides accumulation and cell death (12). One of the key systems in this process is the glutathione peroxidase 4 (GPX4) pathway, which is essential for maintaining cellular redox homeostasis and preventing harmful lipid peroxide accumulation (13). This process converts lipid peroxides in cells into alcohols, which is essential for preventing ferroptosis (14, 15). Therefore, understanding the mechanisms balancing oxidative stress and antioxidant defenses is crucial for comprehending ferroptosis and its implications in pathology.

The ER is an essential organelle critical for cellular homeostasis (16). Cells react to ER stress by triggering processes like the unfolded protein response (UPR) and ER-associated degradation (ERAD) pathways. These pathways aim to correct protein folding issues or remove damaged proteins, restoring ER equilibrium (16, 17). ER stress can be induced by external stimuli, activating processes such as UPR and ERAD to counteract disruptions. ER stress management involves upstream signaling proteins that are pivotal to maintaining ER stability, but when excessively activated, can also trigger apoptosis (18, 19). A key inducer of ER stress is ROS abundance, significant in initiating ER stress responses (20, 21). Recent studies have revealed a connection between ER stress and ferroptosis. However, this relationship is complex and context-dependent, and is influenced by multiple factors (22, 23). The interplay between ER stress and ferroptosis is critical research, as it offers insights into how cellular stress responses contribute to pathological conditions.

Protein disulfide isomerase A member 6 (PDIA6), belonging to the protein disulfide isomerase (PDI) family, is essential for the formation, cleavage, and reorganization of disulfide bonds in proteins (24). Recently, PDIA6 has been identified as a potential regulator cisplatin-resistant non-small cell lung cancer (25). Nevertheless, the exact role of PDIA6 in the process of ferroptosis has yet to be elucidated. In this study, we investigated the role of PDIA6 in both primary ciliogenesis, and in ferroptosis exacerbated by ER stress.

To investigate the effect of PDIA6 on primary cilia, HK-2 cells were transfected with specific siRNA against PDIA6. We performed immunostaining for ARL13B, a common ciliary marker. Decreases in the number and length of primary cilia were observed in PDIA6-knockdown cells (Fig. 1A, B, Supplementary Fig. 1). Ferritin heavy chain 1 (FTH1) is a component of ferritin, which is the main iron storage protein. FTH1 is elevated in the mechanism of anti-ferroptotic effects, thereby contributing to iron metabolism. GPX4 is a major enzyme that eliminates lipid peroxidation within cells. The collapse of the GPX4 antioxidant system leads to the buildup of lipid peroxidation, which then causes ferroptotic cell death. Transferrin receptor (TFR) recognizes transferrin in the transferrin-mediated iron uptake pathway. The upregulation of TFR leads to overload of Fe²⁺. Subsequently, excessive Fe²⁺ promotes lipid peroxidation, ultimately resulting in cell ferroptosis (26, 27). Depletion of PDIA6 indicates changes in proteins related to ferroptosis. As shown in Fig. 1C, in PDIA6-knockdown cells, the expression of FTH1 and GPX4 decreased, whereas TFR expression increased. PDIA6-knockdown cells exhibited elevated levels of ROS, as indicated by the upregulation of HO-1. DCFH-DA analysis showed consistent results to those obtained with HO-1 (Fig. 1D). Lipid peroxidation triggers ferroptosis execution. PDIA6-depleted cells were treated with lipophilic antioxidants, such as ferrostatin-1 (Fer-1) and liproxstatin-1 (Lip-1) (10, 28). C11-BODIPY is used as a sensor for lipid peroxidation in cell membranes; it binds to membranes, emitting green fluorescence that is indicative of its oxidized form. We observed an increase in the production of oxidized C11-BODIPY in PDIA6-depleted cells. However, Fer-1 significantly blocked the production of oxidized C11-BODIPY production induced by PDIA6-knockdown (Fig. 1E). 4-Hydroxynonenal (4-HNE) is a product generated during the peroxidation cascade, with increased levels in PDIA6 depletion, which were alleviated by Fer-1 treatment (Fig. 1F). Additionally, Fe²⁺ levels, indicated by FerroOrange, increased in PDIA6-knockdown cells (Fig. 1G). These results suggest that silencing PDIA6 sensitizes cells to ferroptosis. To investigate the specificity for ferroptosis, we used other forms of cell death inhibitors (Supplementary Fig. 2). To investigate whether the reduced primary cilia in PDIA6-depleted cells are associated with ferroptosis, we employed another ferroptosis inhibitor, liproxstatin-1 (Lip-1). Following treatment of ferroptosis inhibitors, we observed restored ciliary length and number in the PDIA6-knockdown group, along with the recovery of ciliary signaling proteins like Gli2 and CTNNB1 (Fig. 2, Supplementary Fig. 3). These results suggest that the dysregulation of primary ciliogenesis is linked to ferroptosis in PDIA6-depleted cells.

Recent evidence suggests that ER stress can trigger ferroptosis (23, 29), and PDIA6 depletion is known to induce ER stress (24, 30). Treatment of PDIA6-reduced cells with TUDCA, an ER stress alleviator, results in decreased HSPA5 and DDIT3 (Fig. 3A), along with restoration of the number and length of primary cilia (Fig. 3B, C, Supplementary Fig. 4). This elucidates the link between ER stress and PDIA6-mediated reduction in ciliogenesis.

We generated Pdia6 knockout mice using CRISPR/Cas9 to target exons 3-4 of the murine Pdia6 gene, resulting in a 58 bp deletion in exon 3, and a frameshift mutation in exon 4 of Pdia6 ^＋/− (Het) mice, as confirmed by PCR genotyping (Fig. 4A). The intercrossing of Pdia6 heterozygous mice resulted in no homozygous offspring, suggesting the homozygous depletion of Pdia6 is embryonic lethal on the C56BL/6J background. Pdia6^+/− mice kidneys revealed increased TFR protein levels, but decreased GPX4 levels compared to WT mice (Fig. 4B). Consistently, primary ciliogenesis and ciliary elongation in proximal tubules were decreased in Pdia6^+/− mice, while restoration was observed in the Fer-1-treated Pdia6^+/− mice (Fig. 4C, D). In addition, the kidney pathological changes in Pdia6^+/− mice were evaluated to resemble glomerulosclerosis. Following Fer-1 treatment, renal injury decreased, compared to with saline-injected Pdia6^+/− mice (Fig. 4E). Perls Prussian blue staining revealed a slight increase in iron accumulation in Pdia6^+/− mice compared to WT mice, while the Fer-1 injection group did not show observable iron accumulation (Fig. 4F).

Taken together, these findings suggest that PDIA6 plays a role in renal ciliogenesis and the regulation of ferroptosis, indicating that Fer-1 could significantly alleviate renal damage. We summarize the progression resulting from PDIA6-knockdown in a schematic (Fig. 4G).

Primary cilia are key to detecting extracellular signals and transducing these into cellular responses, regulating cell proliferation, migration, and differentiation. The role of primary cilia as signaling hubs for growth factors and morphogens is increasingly recognized. Consequently, mutations in cilia-associated genes lead to ciliopathies (31, 32).

In this study, we have shown that the depletion of PDIA6 disrupts primary cilia regulation via ER stress. PDIA6, which can be induced under ischemic and hypoxic conditions, plays a pivotal role in protein folding by catalyzing the formation of disulfide bonds between cysteine residues. Its levels increase in response to elevated ER stress (31). Our findings suggest that ferroptosis may be triggered by ER stress following PDIA6-knockdown, with ER stress leading to iron accumulation and lipid peroxidation. These phenomena are hypothesized to stem from increased ROS levels driven by ER stress, which changes in transcription factors such as XBP1. Piezo1, a novel Ca²⁺-permeable channel, has been shown to play an important role in Ca²⁺-dependent ferroptosis. Ferroptosis can also occur through the regulation of Ca²⁺ via ER stress (33). We postulate that defects in ciliary structure arise as a consequence of ferroptosis induced under these stressful conditions. Conversely, it is also conceivable that ciliary defects might precipitate ferroptosis. Recent evidence suggests that ciliary signal transduction plays a role in regulating mRNA expression related to iron homeostasis (32). Therefore, it is possible that PDIA6-knockdown results in ciliary abnormalities and subsequent iron accumulation, culminating in ferroptosis. However, further comprehensive studies are necessary to elucidate the role of ferroptosis in primary ciliogenesis.

Mice with Pdia6 mutation exhibit growth retardation, hypoinsulinemia, and hyperglycemia, due to the loss of pancreatic β-cell function (34, 35). A patient with PDIA6 loss-of-function gene displayed a phenotype similar to the mouse model, characteristic of ciliopathies (36). Our results suggest that PDIA6 may regulate the protein folding of ciliopathy-related proteins, either directly or indirectly.

Ferroptosis is characterized by lipid peroxidation catalysis due to ferrous iron or lipoxygenase action, leading to decreased GPX4 and cell death (37). The PERK pathway in ER stress can reduce GSH synthesis by impeding the expression of cystine/glutamate transporter via the p53 gene, thus promoting ferroptosis (38). Our findings indicate that PDIA6 suppression increases ER stress and ferroptosis. Following PDIA6 suppression, we observed increased ER stress and ferroptosis, along with reduced primary cilia. Treatment with ferrostatin-1 and liproxstatin-1, which reduce lipid peroxidation and ferroptosis, led to cilia restoration (Fig. 2). Furthermore, the use of TUDCA, a known ER stress alleviator, demonstrated the link between ER stress and cilia regulation (Fig. 3). However, more research is needed to fully understand how PDIA6 regulates primary cilia and their role in ferroptosis. Our study confirms that PDIA6 inhibition induces ferroptosis and reduces primary cilia.

Recent studies have implicated ferroptosis in kidney abnormalities (39). Previous research on Pdia6^+/− mice identified syndromic disorders, including growth retardation and diabetes, but did not specifically address kidney abnormalities (34). Given the known association of primary cilia with kidney function, our study establishes a link between cilia, ferroptosis, and abnormal kidney development in mice. Intraventricular hemorrhage has observed defects in primary cilia alongside iron accumulation (40). Pdia6^+/− mice displayed glomerular abnormalities that are suggestive of glomerulosclerosis. As glomerulosclerosis progresses, the abnormalities in cilia become more apparent (41). While the exact mechanisms linking ferroptosis and cilia are still unclear, these findings could provide insights into future renal disease treatment.

Cell culture and cilia staining

HK-2 cells were maintained with RPMI (LM011-01) medium, which was supplemented with 10% FBS (S001-07) and 1% penicillin-streptomycin (LS202-02). To stain the primary cilia of HK-2 cells, they were fixed with 4% (w/v) paraformaldehyde (PFA) and 0.1% (v/v) Triton X-100. Primary cilia were immunostained with specific antibodies. Cilia images were captured using a fluorescence microscope. For each experimental condition, cilia counts were obtained from approximately 200 cells (n = 3). Cilia lengths were measured using the free-hand line selection tool in the Olympus CellSens imaging software (Olympus Europa, Hamburg, Germany).

Measurement of lipid peroxidation, ferrous iron and ROS

Intracellular lipid peroxidation was assessed using the BODIPY^TM 581/591 C11 probe. The levels of cellular ferrous iron were measured using the FerroOrange probe. Intracellular ROS were measured using the dichloro-dihydro-fluorescein diacetate (DCFH-DA). All reagents were used according to the protocols provided by the manufacturers. Information about reagents used in this study is provided in the supplementary information. Fluorescence imaging was conducted using a fluorescence microscope (IX71; Olympus, Tokyo, Japan). Intensity was quantified using ImageJ software (NIH).

Animals

Pdia6 knockout mice were generated as the CRISPR/Cas9 system on the C57BL/6J genetic background. Mice were maintained under standard environmental conditions; a temperature of 20-22°C, humidity of 50-60%, light/dark cycles of 12 hours each, and free access to food and water. All animal experiments and handling procedures were performed in agreement with the guidelines of the Institutional Animal Care and Use Committee of the Kyungpook National University (Daegu, Republic of Korea). The approval license number is KNU2023-0450. The mice were divided randomly into three groups: wild-type (WT) group, heterozygous (Het) group, and Het+Ferrostatin-1 (Fer-1) group. Mice were injected intraperitoneally (i.p.) with 5 mg/kg Fer-1 in 0.9% saline once daily for consequently five days. Twenty-four hours post-injection, the animals were sacrificed for subsequent analysis.

Histological analysis

Animals were anesthetized with avertin (200 mg/kg) and transcardially perfused with PBS (pH 7.4). For immunofluorescence staining, the sectioned kidneys were incubated with the following specific antibodies at 4°C overnight, followed by incubation with secondary antibodies at room temperature for 2 hours. All images were captured on a confocal microscope. For assessment of injury, kidney tissues were stained with hematoxylin and eosin (H&E) and Perls Prussian blue to detect tissue iron. The sections were immersed in an acid solution of potassium ferrocyanide, which reacts with ferric iron, and then counterstained with nuclear fast red. Images were acquired using a Leica microscope (Leica DMI3000B, Leica Microsystems Ltd.).

Western blotting

Cell lysates were prepared using 2×Laemmli sample buffer. The total protein content was quantified with the Bradford dye method, following the manufacturer’s instructions. Kidney tissue lysates were extracted using PRO-PREP Protein Extraction supplemented with protease inhibitors. The samples were separated by SDS polyacrylamide gel electrophoresis and transferred onto PVDF membranes. The membranes were then incubated with specific primary antibodies, as detailed in the Supplementary Information.

Statistical analysis

Data were collected from a minimum of three independent experiments and are presented as means ± SD. Statistical analyses were conducted using Student’s t-test and one-way ANOVA, followed by the Bonferroni post hoc test for multiple comparisons. GraphPad Prism 10 software (GraphPad Software, San Diego, CA, USA) was used for all statistical analysis, and P < 0.05 was considered statistically significant for all tests.

This research was supported by the National Research Foundation of Korea, funded by the Ministry of Science & ICT (2020R1A2C2003523), and by Korea Institute for Advancement of Technology funded by the Ministry of Trade, Industry and Energy (P0025489). This research was supported by LAMP Program of the NRF grant funded by the Ministry of Education (RS-2023-00301914).

The authors have no conflicting interests.