A preclinical candidate of cyclophilin D inhibition improves alcohol-associated liver injury

Summary

Currently, no therapies are approved for alcohol-associated liver disease (ALD). Here, we identify cyclophilin D (CypD) as a critical mediator in the progression of ALD. We observe elevated expression of CypD in ALD patients and a corresponding mouse model. Hepatocyte-specific knockout of CypD mitigates hepatic mitochondrial dysfunction, steatosis, inflammation, and oxidative stress. Conversely, overexpression of CypD exacerbates hepatic mitochondrial stress. In vivo and in vitro experiments demonstrate that a CypD inhibitor, RN-0001, effectively and safely alleviates hepatic damage induced by ethanol exposure; these protective effects are absent in CypD-deficient mice. Biophysical assays indicate that RN-0001 directly binds to CypD. Additionally, absorption, distribution, metabolism, excretion, and toxicity (ADMET) tests and first-in-human phase I clinical trial identify RN-0001 as a promising translational candidate for ALD therapy. Collectively, our study highlights the pathological role of CypD in ALD and introduces a preclinical candidate for its management. This study was registered at chictr.org.cn (ChiCTR2500106709).

Graphical abstract

Highlights

• •Cyclophilin D is elevated in ALD and drives hepatocyte mitochondrial stress
• •Hepatocyte CypD loss reduces ethanol-induced steatosis, inflammation, and oxidative stress
• •RN-0001 binds CypD to improve mitochondrial function and reduce liver injury
• •ADMET and phase I data support RN-0001 as a preclinical candidate for ALD

Che et al. show that cyclophilin D promotes mitochondrial dysfunction and liver injury in alcohol-associated liver disease. In ethanol-fed mice, hepatocyte CypD deletion or RN-0001 treatment reduces steatosis, inflammation, and oxidative stress. ADMET profiling and a first-in-human phase I clinical trial support the safety and translational potential of RN-0001.

Introduction

Alcohol-associated liver disease (ALD) is a significant global health concern that stands as the predominant cause of cirrhosis in Western countries, accounting for 60% of cases across Europe and North America.¹ Its escalating prevalence and associated morbidity and mortality impose a considerable burden on healthcare systems and governance.^1,2 ALD manifests a spectrum of clinical presentations ranging from simple steatosis to alcohol-associated hepatitis and fibrosis. Notably, about 10%–20% of those affected progress to cirrhosis, driven by expedited fibrogenesis.³ The pathophysiological mechanisms underlying ALD are intricate and not fully understood. Key features include a reduction in hepatic β-oxidation of fatty acids accompanied by enhanced de novo lipid synthesis, leading to alcohol-associated steatosis.^4,5 Currently, the principal treatments for ALD consist of sustained abstinence in its early stages and liver transplantation in advanced disease stages.⁶ No therapies are approved by Food and Drug Administration (FDA) for ALD, though agents like N-acetylcysteine, granulocyte-colony stimulating factor, and metadoxine show potential benefits.⁷ Emerging insight into the mechanisms driving ALD progression is crucial for identifying preclinical therapeutic candidates.⁸

Mitochondrial dysfunction plays a foundational role in ALD pathogenesis.^9,10 Alcohol consumption prompts early hepatocyte injury, visible through morphological and functional changes in mitochondria.¹¹ Post-alcohol ingestion, mitochondria exhibit increased oxygen consumption as an adaptive measure to more rapidly oxidize the toxic metabolite acetaldehyde and augment the supply of NAD⁺ required for alcohol metabolism.^11,12 An elevated NADH/NAD⁺ ratio suppresses mitochondrial β-oxidation, culminating in steatosis. Additionally, liver tissues from patients with alcohol-associated steatohepatitis show significantly increased levels of cytochrome P450 2E1,^13,14 an alcohol-inducible isoform located in mitochondria that generates reactive oxygen species (ROS) and forms the 1-hydroxyethyl radical.¹⁵ ROS-mediated mitochondrial DNA damage, lipid peroxidation, and the oxidation of mitochondrial proteins and lipids further contribute to mitochondrial dysfunction, perpetuating the cycle of ROS production in both alcohol-associated patients and ethanol-fed mice.^16,17,18 Thus, targeting mitochondrial dysfunction may hold the key to slowing or halting the progression of ALD.

The mitochondrial permeability transition pore (mPTP) is a critical regulator of mitochondrial function, orchestrating key aspects of energy metabolism and cellular homeostasis.¹⁹ Calcium-induced mitochondrial permeability transition (mPT) is characterized by a precipitous increase in inner mitochondrial membrane permeability, resulting in dissipation of the mitochondrial membrane potential (ΔΨm), mitochondrial swelling, and outer membrane rupture.²⁰ Cyclophilin D (CypD), encoded by the peptidyl-prolyl cis-trans isomerase F gene (PPIF), is a key regulator of the mPTP complex. CypD modulates mPTP opening, consequently influencing ATP synthesis and ROS production.²¹ Exaggerated CypD activity has been associated with enhanced mPTP opening, causing endoplasmic reticulum stress and hepatic steatosis.²² Cyclosporine A (CsA) and related non-immunosuppressive analogs like alisporivir effectively inhibit mPTP by binding to CypD.²³ Supporting this, alisporivir has shown promise in mitigating mitochondria-dependent apoptosis/necrosis in muscular pathology models and has enhanced recovery while reducing mortality post-acute myocardial infarction in mice.^24,25,26 Despite its established role in mitochondrial homeostasis, the contribution of the mPTP to alcohol-induced hepatic mitochondrial calcium overload, dysfunction, and steatosis remains elusive. Our findings revealed a significant upregulation of CypD expression in both human ALD patients and corresponding mouse models. Hepatocyte-specific CypD knockout in ALD mice conferred a protective effect against hepatic mitochondrial dysfunction, steatosis, inflammation, and oxidative stress, whereas hepatocyte-specific overexpression of CypD intensified hepatic mitochondrial dysfunction. Crucially, we synthesized a CypD inhibitor (RN-0001) that effectively mitigated alcohol-associated liver damage in mice. Chemical absorption, distribution, metabolism, excretion, and toxicity (ADMET) profiling and first-in-human phase I clinical trial further support RN-0001 as a promising preclinical candidate for ALD therapy.

Results

CypD expression is upregulated in the liver of ALD patients and mouse NIAAA model

Mitochondrial dysfunction is a central pathogenic mechanism underlying ALD. Using the well-established NIAAA chronic-binge ALD mouse model, which recapitulates human ALD progression through chronic-plus-binge ethanol feeding and mimics the clinical trajectory of heavy drinkers who develop acute-on-chronic liver injury (Figure 1A),²⁷ we observed pronounced mitochondrial swelling in hepatocytes of ALD mice compared to healthy controls via transmission electron microscopy (Figure 1B). To systematically reveal the molecular basis of mitochondrial dysfunction, we conducted a relative quantitative analysis of key mitochondrial proteins in the liver and found that the expression profile of mitochondrial-related genes in the ALD group underwent significant reprogramming: the expression of the key mitochondrial permeability protein Ppif (encoding CypD) was upregulated; the expression of Vdac1 was slightly suppressed; the expression levels of the division regulatory factors Dnm1l and Mief2 were downregulated; the expressions of mitochondrial autophagy markers Mst1 and Bnip3 decreased; the mitochondrial apoptotic factors Ripk1 and Pgam5 were downregulated; and the expressions of biogenic regulatory proteins Ppargc1a, Tfam, and Ppara were repressed. The expression of mitochondrial-fusion-related protein Mfn1 decreased, while the expression of Miag1 increased. The expressions of transportation-related proteins Trak1 and Rhot2 were downregulated (Figure 1C). Given the core position of Ppif in the regulation of mPTP and its most significant expression difference, we further detected the CypD protein level and found that the expression level of CypD protein in the liver of ALD mice significantly increased compared with the control group (Figure 1D). To further verify the relationship between CypD and the disease progression of ALD, we found that CypD (mRNA and protein levels) was positively correlated with the pathological score of liver steatosis (Figures 1E and 1F), suggesting that CypD served as a key target for the disease progression of ALD. To verify the clinical relevance of this phenomenon, we conducted an analysis through published clinical cohorts.²⁸ The level of CypD protein in the liver of ALD patients was significantly upregulated with disease progression (Figure 1G). Immunohistochemical analysis further confirmed that the signal intensity of CypD protein in the liver tissues of ALD patients was upregulated compared with that in normal liver tissues (Figure 1H). In addition, through the analysis of human (GSE115469) and mouse (SCP1404) liver single-cell sequencing data, Ppif expression is almost exclusively localized to hepatocytes in the liver (Figures 1I, 1J, and S1).

Figure 1.

Cyclophilin D (CypD) expression is upregulated in alcohol-associated liver disease (ALD)

(A) Schematic representation of the chronic-binge ALD model (the NIAAA model).

(B) Transmission electron microscopy analysis of liver mitochondria in healthy mice and ALD mice (left, mitochondrial swelling marked with a red arrow; scale bars, 2 μm [left panel]/500 nm [magnified part]); quantitative analysis of mitochondrial swelling area in the liver (right, n = 30 per group).

(C) Quantitative PCR results of key mitochondrial functional proteins in the livers of healthy (control) mice and ALD mice (n = 4 mice per group).

(D) Representative western blot results of CypD protein in the livers of healthy mice and ALD mice; corresponding quantitative analyses were conducted (n = 3 mice per group).

(E) Representative western blot results of CypD protein in mouse liver tissues at different stages of steatosis; corresponding quantitative analysis was conducted (n = 3 mice per group).

(F) The expression level of Ppif mRNA in mouse liver tissues at different stages of steatosis (n = 4 mice per group).

(G) Expression changes of CypD protein in liver tissues at different stages of steatosis in a published clinical cohort (left panel, n = 12–36 participants per group). The expression of CypD protein is shown in liver tissue stratified by the absence of steatosis or inflammation (S/I 0), varying degrees of steatosis (S 1–3), and varying degrees of inflammatory activity (I 1–3) (right panel, n = 43–63 participants per group). S: grade of steatosis, I: grade of inflammation. S/I 0: no steatosis or inflammation; S 1–3: steatosis grade 1–3; I 1–3: inflammatory activity grade 1–3.

(H) Representative result maps of CypD immunohistochemistry in liver tissues from healthy individuals and ALD patients (scale bars, 100 μm; n = 4 participants per group). Arrows indicate regions/cells with enhanced CypD immunoreactivity.

(I and J) Single-cell sequencing data of Ppif expression in different liver cell types of humans. This set of data was downloaded from the Human Protein Atlas (GSE115469). All data are presented as the means ± SDs. i.g., intragastric administration.

Hepatocyte-specific ablation of CypD prevents ethanol-induced hepatic injury

To further elucidate CypD’s role in ALD progression, we engineered a hepatocyte-specific CypD knockout mouse model (CypD^ΔHep) due to the predominant Ppif expression in hepatocytes (Figures 1I, S2A, and S2B). Post-ethanol treatment, no significant differences in weight loss or dietary intake were observed between control CypD^flox/flox (CypD^f/f) and CypD^ΔHep mice (Figures 2A and 2B), suggesting that CypD deficiency does not affect the basal energy metabolism. Remarkably, the levels of serum alanine aminotransferase (ALT) and aspartate aminotransferase (AST) in the CypD^ΔHep-ALD group of mice were significantly lower than those in the CypD^f/f-ALD group (Figure 2C), indicating that alcohol-induced hepatocyte necrosis was improved. Lipid metabolism analysis revealed that the levels of circulating total cholesterol (TC) and triglycerides (TG) decreased simultaneously (Figure 2C), while the release of pro-inflammatory cytokines tumor necrosis factor alpha (TNF-α) and interleukin-6 (IL-6) was also significantly inhibited (Figure 2D), suggesting that CypD deficiency systematically alleviated alcohol-induced metabolic disorders and inflammatory reactions. Further analysis of the liver phenotype revealed that compared with the mice in the CypD^f/f-ALD group, the liver weight/body weight ratio of the mice in the CypD^ΔHep-ALD group was significantly decreased (Figure 2E), and the accumulation of hepatic free fatty acid (FFA) and TG contents in the liver was reduced (Figure 2F). Histological evidence supports the above findings. Compared with mice in the CypD^f/f-ALD group, the lipid droplet deposition area in hepatocytes of mice in the CypD^ΔHep-ALD group was reduced, and inflammatory cell infiltration was decreased (Figure 2G). Changes in hepatic protein levels of markers related to mitochondrial function (dynamin-related protein 1 [DRP1], mitochondrial fission 1 protein [FIS1], mitofusin 1 [MFN1], mitofusin 2 [MFN2], and optic atrophy 1 [OPA1]), oxidative stress (cytochrome P450 2E1 [CYP2E1] and inducible nitric oxide synthase [iNOS]), antioxidation (phosphorylated nuclear factor erythroid-2-related factor 2 [p-NRF2], superoxide dismutase 1 [SOD1], and catalase [CAT]), and apoptosis (cleaved caspase-3 [c-casp3], cytochrome c [Cyto C], B cell lymphoma 2 [Bcl2], and BCL2-associated X protein [Bax]) corroborated this phenotype (Figure 2H). These findings confirm that CypD is a key regulatory factor for mitochondrial-dependent damage during the process of ALD.

Figure 2.

Hepatocyte-specific knockout of cyclophilin D (CypD) improves alcohol-induced hepatic steatosis

(A) Schematic diagram of the chronic-binge ALD model (the NIAAA model) with hepatocyte-specific knockout of CypD.

(B) Body weight changes of CypD^flox/flox (CypD^f/f) or CypD^ΔHep mice stimulated or not stimulated by ethanol (left, n = 7 mice per group); the changes in the dietary intake of mice (right, n = 15 per group).

(C) The levels of serum biochemical parameters (ALT, AST, TG, and TC) of mice in each group (n = 4 mice per group).

(D) Levels of pro-inflammatory cytokines (TNF-α and IL-6) in the serum of mice in each group (n = 5 mice per group).

(E) Liver weight and liver/body weight ratios were assessed in mice from indicated groups (n = 5 mice per group).

(F) Changes in mouse liver FFA and TG levels from indicated groups (n = 5 mice per group).

(G) Representative H&E and oil red O stained images of mouse liver sections from indicated groups are shown. Scale bars, 50 μm. NAFLD activity score (NAS) and the measurement of oil red O staining (n = 4–5 mice per group).

(H) Protein level changes in the liver from indicated groups (left) and their quantification (right; n = 3 per group). All data are presented as the means ± SDs. i.g., intragastric administration.

CypD-dependent mitochondrial permeability transition regulates ethanol-induced liver damage

To determine if enhanced mitochondrial function underlies the improved steatosis and inflammation in CypD^ΔHep mice, we assessed hepatic mitochondrial stress in ALD models. Ethanol treatment induced marked mitochondrial swelling (indicating mPTP opening)²⁹ in CypD^f/f-ALD mice, whereas CypD^ΔHep mice showed significantly attenuated swelling under identical alcohol stimulation, emphasizing the key role of CypD in ethanol (Figure 3A). The results of transmission electron microscopy showed that ethanol stimulation led to mitochondrial swelling and crest disappearance (indicated by red arrowheads) in CypD^f/f mice. However, after specific CypD knockout in mouse hepatocytes, the mitochondrial swelling caused by ethanol weakened and the cristae were well preserved (Figure 3B). Consistent with prior studies,¹⁷ basal mitochondrial Ca²⁺ levels elevated by ethanol were reduced in CypD^ΔHep-ALD mice (Figure 3C). Evaluations of Ca²⁺ retention capacity (CRC) revealed that ethanol diminished CRC in mitochondria from CypD^f/f mice, an effect significantly reversed in CypD^ΔHep-ALD mice (Figure 3D). Assessment of mitochondrial respiratory function showed that the maximal mitochondrial respiratory capacity (elicited by FCCP) of CypD^ΔHep-ALD mice was elevated compared with that of the CypD^f/f group after ethanol exposure (Figures 3E and 3F), suggesting that CypD knockdown reversed ethanol-induced oxidative phosphorylation damage. Under resting conditions, there was no difference in the polarization of mitochondria between mouse groups as assessed by high-resolution, Ca²⁺-sensitive electrode measurements. The repetitive Ca²⁺ pulses produced fast, transient depolarizations in CypD^f/f-ALD mitochondria, whereas in CypD^ΔHep-ALD mitochondria, Ca²⁺ pulses produced slower depolarizations with much smaller amplitude (Figure 3G). At the end of these experiments, mitochondria were treated with the uncoupler 2,4-dinitrophenol (2,4-DNP) to produce maximal depolarization. After alcohol stimulation, the mitochondrial respiratory function and ATP production of CypD^f/f mice were inhibited. CypD^ΔHep could effectively reverse the mitochondrial respiratory capacity and ATP production disorders caused by ethanol (Figures 3H and 3I). Excessive mPTP opening, indicated by mitochondrial swelling, promotes ROS generation, cytochrome c release, and subsequent apoptosis.^30,31 MitoSOX Red, a mitochondrial-targeted fluorogenic dye, selectively detects superoxide (O₂^·−) within the mitochondrial matrix through oxidation to a red fluorescent product, enabling spatial resolution of ROS generation in ethanol-stressed hepatocytes. CypD^ΔHep-ALD mice exhibited significantly reduced mitochondrial ROS production compared to CypD^f/f-ALD (Figure 3J). Collectively, these findings demonstrate that hepatic CypD deficiency attenuates ethanol-induced mitochondrial swelling, mPTP opening, and subsequent mitochondrial Ca²⁺ accumulation.

Figure 3.

Hepatocyte-specific knockout of cyclophilin D (CypD) inhibits mPTP opening and improves hepatic mitochondrial dysfunction

(A) Mitochondrial swelling, induced by calcium, was assessed in isolated liver mitochondria. The percentage decrease in initial optical density at 520 nm was used as a measure of swelling severity, determined by the slope of the resulting curve (n = 5 mice per group).

(B) The swelling degree of mitochondria in each group was analyzed by transmission electron microscopy (the mitochondrial swelling was marked by red arrows). Scale bars, 2 μm (upper panels) and 500 nm (zoom-out part). The swollen area of mitochondria in each group was quantitatively analyzed (right part; n = 30 per group).

(C) Ca²⁺ concentration in the liver tissues of indicated groups (n = 5 mice per group).

(D) Mitochondrial calcium retention capacity (CRC) was assessed in purified liver mitochondria from CypD^f/f and CypD^ΔHep mice using a Fluo-4 AM assay. Mitochondria were subjected to sequential 40 μM CaCl₂ pulses until reaching mPTP opening or calcium uptake saturation (n = 5 mice per group).

(E) Oxygen consumption rates (OCR) were measured in CypD^f/f and CypD^ΔHep mouse liver mitochondria. Mitochondrial respiration was challenged sequentially with oligomycin (2 μM), FCCP (2 μM), and rotenone (0.5 μM) (n = 5 mice per group).

(F) Quantification of state 3 respiration from (E) (n = 5 mice per group).

(G) The impact of calcium on mitochondrial membrane potential was examined in CypD^f/f and CypD^ΔHep mitochondria.

(H) The maximal respiratory control ratio (RCR; state 3/state 4 respiration) was determined for isolated liver mitochondria from indicated groups (n = 5 mice per group).

(I) ADP/ATP ratio the in liver from of indicated groups (n = 5 mice per group).

(J) Mitochondrial ROS generation was detected in liver sections using MitoSox Red fluorescence. Scale bars, 100 μM (left). The ratio of MitoSox-red-stained area to DAPI-blue-stained area was quantified (n = 5 mice per group) (right). All data are presented as the means ± SDs.

Liver-specific overexpression of CypD exacerbates ethanol-induced hepatic mitochondrial stress

To further investigate the involving role of CypD in ALD development, we established a liver-specific CypD overexpression mice model by tail vein injection of AAV8-CypD viral vector targeting the liver (Figures S3A and S3B). Compared with the control vector (AAV-Vector-Ctrl), AAV-CypD-Ctrl did not significantly alter body weight or dietary intake in mice (Figure S3C). Following chronic-binge ethanol administration, liver-specific CypD overexpression did not substantially modulate most ALD-associated parameters, including serum ALT, AST, TC, TG, TNF-α, and IL-6 levels, as well as hepatic FFA and TG contents (Figures S3D, S3E, and S3F). Hepatic histology (H&E and oil red O staining) further confirmed this observation (Figure S3G). However, CypD overexpression significantly increased mitochondrial ROS production (Figure S3H) and exacerbated mitochondrial swelling compared to AAV-Vector mice post-ethanol consumption (Figure S3I), indicating heightened mitochondrial stress. These results suggested that while CypD overexpression might not influence liver toxicity and steatosis directly but aggravated ethanol-induced mitochondrial damage and oxidative stress.

Finally, complementary in vitro examinations employing both knockdown and overexpression of CypD in a well-recognized cellular ALD-like model³² confirmed that CypD modulation significantly influences hepatocyte viability, apoptosis, and lipidomic responses to ethanol exposure (Figure S4).

CypD is associated with hepatic lipid metabolism and oxidation pathways in ALD mice

To systematically clarify the possible pathways involved in the improvement of ALD by CypD deletion, in this study, whole transcriptome sequencing (RNA sequencing) was performed on the liver tissues of CypD^ΔHep-ALD mice, CypD^f/f-Ctrl, and CypD^f/f-ALD mice. Differential gene analysis (|log2FC| > 1, FDR <0.05) revealed that ethanol exposure disturbed metabolism-related genes, exacerbating metabolic disorders in hepatocytes. CypD knockout ameliorated alcohol-induced liver injury, potentially by enhancing fatty acid oxidation, promoting lipid droplet dynamics, and strengthening antioxidant defenses (Figure S5A). These findings further suggest that the CypD/mPTP pathway serves as a central hub linking mitochondrial dysfunction to metabolic dysregulation. Reactome pathway enrichment analysis further revealed that CypD knockout specifically regulates core pathways such as “biological oxidation processes” and “fatty acid metabolism” (Figure S5B). KEGG analysis verified the above findings (Figure S5C). Meanwhile, GSEA revealing significant regulation by CypD in fatty acid metabolism and biological oxidations pathways (Figure S5D). These data indicate that CypD deficiency reshaped liver energy metabolism homeostasis by synergistically activating mitochondrial oxidative metabolism and lipid breakdown pathways.

RN-0001 ameliorates alcohol-induced hepatocytes injury via targeting CypD in cell models

Given the promising therapeutic potential of inhibiting CypD in the management of ALD, we sought to modify cyclosporin A (CsA), a potent cyclophilin inhibitor that failed to advance through late-stage clinical trials due to multiple organ-toxicity and immunosuppressive properties.³³ Through structural modification of CsA, we developed a CypD inhibitor named RN-0001, a macrocyclic peptide (Figure 4A). The chemical synthetic route of RN-0001 was depicted in Figure 4B. In vitro assays demonstrated that RN-0001 effectively inhibited human CypA and CypD, with Ki values of 4.1 nM and 12.0 nM, respectively, comparable to the inhibitory capacity of CsA against CypA and CypD. CsA exhibited minimal inhibitory effects on calcineurin activity in the absence of CypA, with an IC₅₀ exceeding 1,000 nM. In contrast, RN-0001 demonstrated modest inhibition of calcineurin activity (IC₅₀ = 2,276 nM) under the same conditions. The formation of complexes between CsA or RN-0001 with CypA significantly enhanced calcineurin inhibition, resulting in IC₅₀ values of 107 nM and 1,551 nM, respectively. Notably, RN-0001 exhibited calcium ion retention capacity (EC₅₀ = 916 nM) similar to CsA (EC₅₀ = 369 nM). Furthermore, while CsA potently suppressed IL-2 release (EC₅₀ = 3.78 nM), RN-0001 showed markedly reduced immunosuppressive activity, as evidenced by higher IL-2 release (EC₅₀ = 347.90 nM) (Figure 4C). To support the selection of RN-0001 as a lead compound and to guide the dosing strategy for subsequent in vivo studies, we then characterized its pharmacokinetic (PK) properties (Figure 4D; Table S1). Following a single intravenous administration of RN-0001 (1, 3, or 10 mg/kg) to healthy Sprague-Dawley rats, peak plasma concentrations (Cmax) were 3,670; 7,470; and 24,800 ng/mL, respectively (Figure 4D), with corresponding elimination half-lives of 3.17, 3.53, and 5.06 h (Table S1). In the ALD-like cell model, concentration-response experiments determined the IC₅₀ of RN-0001 to be 102.7 μM for AML12 and 60.73 μM for RAW264.7 (Figure 4E). These results demonstrate that RN-0001 exhibits strong activity in both cell types. Subsequently, we selected concentrations of 0.01, 0.1, 1, and 10 μM for further bioactivity testing of RN-0001. Compared to control group, ethanol exposure (ALD-no drug group) significantly reduced hepatocyte viability, increased apoptosis, elevated LDH release, and exacerbated lipid accumulation and inflammation (Figures 4F, 4G, 4H, and 4J). RN-0001 effectively suppressed the ethanol-induced upregulation of lipogenic markers (SREBP1), pro-inflammatory cytokines (TNF-α and IL-6), oxidative stress marker (iNOS), and apoptotic marker (cleaved caspase-3), which were all induced by ethanol exposure (Figure 4I). Across the tested range (0.01–10 μM), RN-0001 produced a concentration-dependent protection that approached a plateau at 0.1 μM. Therefore, 0.1 μM was selected as the lowest concentration providing consistent near-plateau protection for subsequent mechanistic experiments. Notably, RN-0001 outshone Debio-025 (alisporivir, DEB025), a well-recognized cyclophilin inhibitor in liver diseases,^34,35 in alleviating ethanol-induced hepatocyte injury at equivalent doses.

Figure 4.

RN-0001 is superior to Debio-025 in mitigating alcohol-induced hepatocyte injury

(A) Chemical structure of RN-0001, a selective inhibitor of CypD.

(B) Chemical synthesis route of RN-0001 from cyclosporine A (CsA).

(C) Biological functional tests of RN-0001 and CsA in vitro.

(D) Pharmacokinetics of RN-0001 were determined in Sprague-Dawley rat. RN-0001 was administered via intravenous (i.v.) injection at doses of 1, 3, and 10 mg/kg. Blood concentrations of RN-0001 were measured between 0 and 24 h (n = 6 per group).

(E) Dose-response curves and calculated half maximal inhibitory concentration (IC₅₀) of AML12 and Raw264.7 treated with RN-0001 (n = 3 per group).

(F–H) Results of cell viability (F), lactate dehydrogenase (LDH) activity (G), and apoptosis rate (H) with or without various concentrations of RN-0001 (0.01, 0.1, 1, and 10 μM) or Debio-025 (DEB025, 1 and 10 μM) in ALM12 cell (n = 3–5 per group).

(I) Protein level changes in AML12 from indicated groups and their quantification (n = 3 per group, lower panels).

(J) Results of Nile Red staining (left) and quantification of Nile Red staining (right) in AML12 from indicated groups. Scale bars, 20 μm (n = 5 per group). All data are presented as the means ± SDs.

After knocking down CypD expression in AML12, an increase in cell viability post-ethanol treatment was noted; however, such anti-ALD effects were not enhanced further with RN-0001 treatment. In contrast, with CypD overexpression, ethanol treatment markedly suppressed cell activity, which was significantly mitigated upon co-treatment with RN-0001 (Figures S6A, S6B, and S6C). Additionally, the protein levels associated with lipid metabolism, oxidative stress, inflammatory responses, and apoptosis were inversely regulated by co-administration with RN-0001 (Figure S6D). Nile red staining illustrated that CypD knockdown effectively reduced ethanol-induced lipid accumulation in hepatocytes. Conversely, ethanol-induced lipid accumulation was exacerbated by CypD overexpression but was reversed upon co-treatment with RN-0001 (Figure S6E).

In addition, we examined the activity changes of mitochondrial complexes I, II, III, and IV. Knockdown of CypD or administration of RN-0001 alleviated the inhibition of alcohol on the activity of mitochondrial complexes I–IV (Figure S7A). At the same time, ethanol inhibited intracellular ATP production and mtDNA content; however, knockdown of CypD or administration of RN-0001 reversed the ethanol-induced reduction in ATP production and mtDNA content (Figures S7B and S7C). To further validate the effects of CypD knockdown and RN-0001 administration on mitochondrial membrane potential, we performed Rhodamine 123 staining in ALD cell model. The results showed that Rhodamine 123 fluorescence intensity was significantly reduced in the ALD group (Figure S7D), indicating significant mitochondrial depolarization. This finding is consistent with our earlier observations of electron transport chain (ETC) dysfunction (e.g., impaired ATP production) and oxidative stress (e.g., elevated iNOS and reduced SOD1/CAT), suggesting that ΔΨm collapse is a downstream consequence of disruption of ETC efficiency and ROS overproduction. However, knockdown of CypD or administration of RN-0001 alleviated the effects of alcohol on mitochondrial membrane potential. We further investigated the nuclear factor κB (NF-κB) p65 nuclear translocation. It could be seen that alcohol induced the translocation of NF-κB p65 from the cytoplasm into the nucleus; however, knockdown of CypD or administration of RN-0001 alleviated the alcohol-induced NF-κB p65 nuclear translocation situation (Figures S7E and S7F). These findings underscored that RN-0001 exerted its therapeutic effects against ALD through targeting CypD in cellular models.

RN-0001 engages several cyclophilins, but its ALD protection is predominantly CypD dependent

To define the interaction mode of RN-0001 with CypD and to anchor subsequent mechanistic and in vivo studies, we combined cross-species molecular docking/molecular dynamics (MD) simulations with orthogonal biophysical assays (SPR and MST). Docking and 200-ns MD simulations supported stable occupancy of RN-0001 within the canonical PPIase pocket of CypD, with conserved key contacts across species (mouse His95/Arg96; human Arg97/Gln105/Gly114) (Figures 5A, 5B, 5C, and 5D). Consistent with this binding mode, RN-0001 exhibited high-affinity binding to CypD in both species (MST Kd: mouse 51.8 nM; human 25.2 nM; SPR Kd: mouse 48.9 nM; human 23.7 nM) (Figures 5E and 5F).

Figure 5.

RN-0001 has a direct interaction force with the CypD protein of humans and mice

(A and B) Molecular docking of RN-0001 with mouse/human CypD (left), and calculation of the binding free energy of the mouse/human CYPD-RN-0001 complex (MM-GBSA) (right).

(C) Root-mean-square variance (RMSD) analysis of mouse CypD (unbound) and mouse CYPD-RN-0001 complex (right) and root-mean-square fluctuation (RMSF) analysis (left).

(D) RMSD analysis of human CypD (unbound) and human CYPD-RN-0001 complex (right) and RMSF analysis (left).

(E) MST affinity test of RN-0001 with human and mouse CypD (n = 3 per group).

(F) SPR affinity test of RN-0001 with human and mouse CypD. All data are presented as the means ± SDs.

Because the PPIase pocket is highly conserved among cyclophilins, we additionally profiled RN-0001 against PPIA (CypA), PPIB (CypB), and PPIG (CypG) using SPR and MST (Figure S8), showing measurable engagement of CypA and CypB in the nano-molar range and weaker binding to CypG in the micro-molar range.

To evaluate the functional relevance of Ppia and Ppib in ALD, we integrated human transcriptomic evidence with loss-of-function studies. In the GSE142530 cohort, PPIA and PPIB did not differ substantially across healthy controls, alcohol-associated hepatitis (AH), and alcohol-associated cirrhosis (AC) (Figure S9A). Concordantly, siRNA-mediated knockdown of Ppia or Ppib in AML12 conferred minimal protection against ethanol-induced loss of viability and lipid accumulation (Figures S9B, S9C, and S9D). In vivo, hepatocyte-targeted knockdown of Ppia or Ppib using AAV8-HAAT-shRNA produced only limited improvement in alcohol-induced liver injury and did not significantly rescue mitochondrial dysfunction (Figures S9E, S9F, S9G, S9H, and S9I). Together, these data indicate that although RN-0001 can bind CypA/CypB, Ppia/Ppib contribute minimally to the mitochondrial injury phenotypes emphasized in our ALD models. We therefore focus the mechanistic framework of this study on Ppif (CypD) and the mPTP-mitochondrial vulnerability axis as the principal link between RN-0001 target engagement, mitochondrial protection, and attenuation of ALD progression.

RN-0001 ameliorated ALD-associated hepatic injury in mice models via CypD targeting

In mouse models of ALD (Figure 6A), treatment with RN-0001 (10 mg/kg) substantially ameliorated signs of liver injury. Compared with the mice in the ALD group, the mice in the ALD-RN-0001 group significantly alleviated alcohol-induced liver injury after administration of RN-0001, including partial recovery of alcohol-induced weight loss (Figure 6B), improvements of serum biomarkers (ALT, AST, TG, and TC) caused by alcohol (Figure 6C), and reduction of the elevated levels of pro-inflammatory cytokines, TNF-α, and IL-6 (Figure 6D). Furthermore, it restored the liver weight/body weight ratio of mice to normal levels (Figure 6E), while alleviating the increase of liver lipid indicators (FFA and TG) caused by alcohol (Figure 6F). Importantly, these improvements occurred without affecting daily dietary intake. Co-treatment with RN-0001 significantly reduced hepatic fat droplet accumulation and macrophage infiltration, reflected by a lower NAFLD activity score (NAS; Figure 6G). Additionally, RN-0001 reversed aberrations in key proteins involved in mitochondrial function (DRP1, FIS1, MFN1, MFN2, and OPA1), oxidative stress (CYP2E1 and iNOS), antioxidation (p-NRF2, NRF2, SOD1, and CAT), and apoptosis (cleaved caspase-3, Cyto C, Bcl2, and BAX) (Figure 6H). Notably, RN-0001 treatment significantly reduced mitochondrial swelling post-ethanol exposure (Figure 6I).

Figure 6.

RN-0001 effectively alleviates alcohol-induced liver damage through CypD

(A) Schematic diagram of the RN-0001 administration in the NIAAA chronic-binge ALD mouse model.

(B) Changes in body weight and diet intake of mice from indicated groups (n = 15 per group).

(C) Changes in the levels of mouse serum biochemical parameters (ALT, AST, TG, and TC) in mice from indicated groups (n = 3-5 per group).

(D) Changes in mouse serum pro-inflammatory mediators (TNF-α and IL-6) levels from indicated groups (n = 5 mice per group).

(E) Liver weight and liver/body weight ratios were assessed in mice from indicated groups (n = 5 mice per group).

(F) Changes in mouse liver FFA and TG levels from indicated groups (n = 5 mice per group).

(G) Representative H&E and oil red O staining images of mouse liver sections from indicated groups are shown. Scale bars, 50 μm (left), NAFLD activity score (NAS score) and the measurement of oil red O staining (right, n = 4–5 mice per group).

(H) Protein level changes in the liver from indicated groups (left) and their quantification (n = 3 per group, right).

(I) Degree of mitochondria swelling in each group analyzed by transmission electron microscopy (mitochondrial swelling marked by red arrows). Scale bars, 2 μm (upper panels) and 500 nm (zoom-out part). The swollen area of mitochondria in each group was quantitatively analyzed (right; n = 30 per group).

(J) Changes of serum biochemical indexes (ALT, AST, TG, and TC) between CypD^f/f and CypD^ΔHep mice after ethanol and RN-0001 treatments (n = 5 mice per group).

(K) Representative H&E and oil red O staining images of liver sections of CypD^f/f and CypD^ΔHep mice after ethanol and RN-0001 treatments. Scale bars, 50 μm. (upper panels); NAFLD activity score (NAS score) and measurement of oil red O staining (below; n = 5 mice per group).

(L) Degree of mitochondria swelling in CypD^f/f and CypD^ΔHep mice after ethanol and RN-0001 treatments analyzed by transmission electron microscopy. Scale bars, 2 μm (upper panels) and 500 nm (zoom-out part). The swollen area of mitochondria in each group was quantitatively analyzed (below; n = 30 per group). All data are presented as the means ± SDs. i.g., intragastric administration; i.p., intraperitoneal injection.

To further benchmark RN-0001 against clinically relevant cyclophilin-directed strategies, we performed parallel efficacy studies with the pan-cyclophilin inhibitor rencofilstat (CRV431). Under our experimental conditions, CRV431 provided weaker protection than RN-0001: in ethanol-injured hepatocytes, CRV431 was less effective at matched concentrations in suppressing alcohol-induced cell death and lipid accumulation (Figures S10A and S10B), and in the NIAAA chronic-plus-binge ALD mouse model, CRV431 at 50 mg/kg conferred only modest benefit and remained inferior to RN-0001 at 10 mg/kg, particularly in steatosis-related endpoints (Figures S10C, S10D, and S10E). These results indicate that CRV431 exerts limited activity in reversing ALD-associated hepatic lipid accumulation in our settings. Furthermore, we compared RN-0001 with the clinically investigated CsA analogues rencofilstat (CRV431) and SCY-635 with respect to residue-level structural modifications and functional profiles (Table S2).

To specifically assess whether RN-0001 hepatoprotective effects stemmed from its action on CypD, we administered RN-0001 along with an alcohol-associated liquid diet to both CypD^f/f and CypD^ΔHep mice. In CypD^ΔHep mice, RN-0001 did not manifest additional improvements in serum biomarkers (ALT, AST, TG, and TC; Figure 6J) or further reductions in histopathological damage, lipid accumulation, and mitochondrial swelling (Figures 6K and 6L). These observations confirm that RN-0001’s therapeutic efficacy in ALD primarily hinges on its interaction with hepatic CypD.

Long-term administration with RN-0001 was safe in healthy mice

We evaluated the safety of RN-0001 in healthy mice through a long-term administration assay. Mice received once every 2 days intraperitoneal injections of RN-0001 (10 mg/kg) for 90 consecutive days. No death of mice was observed during the administration period. Visceral indices and histological analysis of the liver, skeletal muscle, lung, spleen, kidneys, and heart revealed no apparent signs of injury in mice treated with or without RN-0001 (Figures S11A and S11B). Moreover, comparison of the control and control+RN-0001 groups revealed no significant changes in serum biochemical markers, including ALT, AST, TG, and TC (Figure S11C). These results suggest that long-term administration of RN-0001 did not adversely affect the overall health of the mice.

RN-0001 demonstrated favorable pharmacokinetics and safety in rodents/non-rodents and healthy volunteers

Investigational new drug (IND)-enabling Good Laboratory Practice (GLP) studies with RN-0001 showed no adverse effects on cardiovascular, neurological, or respiratory systems. Additionally, 4-week toxicity studies in rodents and non-rodents demonstrated no drug-related toxicity. The no-observed-adverse-effect levels (NOAELs) were determined to be 20 mg/kg/day for rats and 18 mg/kg/day for dogs (Table S3). To evaluate the safety, tolerability, and pharmacokinetics of RN-0001, a placebo-controlled, exploratory phase I, first-in-human study (China’s Center for Drug Evaluation Clinical Trials, CXHL2000510; Chinese Clinical Trial Register, ChiCTR2500106709) was conducted with 80 healthy adult volunteers, who were allocated to 10 treatment groups (Figure S12). No severe or serious adverse events or dose-limiting toxicities were reported. Mild to moderate adverse effects occurred in 22.6% of RN-0001-treated subjects (14/62) compared to 11.1% of placebo-treated subjects (2/18) (Table S4). RN-0001 exhibited rapid absorption, with peak plasma concentrations achieved within 2.50–2.98 h post-dosing and a terminal elimination half-life of 15.51–26.19 h. Multiple-ascending dose cohorts demonstrated a degree of fluctuation (DF) of 133.76%–208.40% and an accumulation index (Rac) of 1.44–1.50 (Table S5). Overall, in above ADMET tests and first-in-human phase I clinical trial, RN-0001 demonstrated favorable pharmacokinetics and safety, supporting its progression to phase II efficacy trials.

Discussion

ALD represents a formidable global health challenge, affecting approximately 75 million people worldwide who suffer from alcohol use disorders out of some 2 billion alcohol consumers.³⁶ It ranks among the top causes of alcohol-related mortality, despite significant strides in understanding the pathophysiology of alcohol-related liver conditions. While FDA-approved pharmacotherapies exist for alcohol use disorder (e.g., disulfiram, acamprosate, and naltrexone),³⁷ no FDA-approved treatments are currently approved to directly treat the hepatic manifestations of ALD. Existing alcohol use disorder medications, though valuable for reducing alcohol intake, do not mitigate alcohol-induced hepatocyte injury, mitochondrial dysfunction, or fibrogenesis-pathological hallmarks requiring targeted intervention. Clinical trials targeting various aspects of ALD pathogenesis—including inflammation, oxidative stress, and metabolic disturbances—have largely failed to advance, partially due to the paucity of effective therapeutic targets.^38,39

The liver, like other metabolically active organs, is rich in mitochondria and heavily dependent on their function for energy production and cellular metabolism.⁴⁰ Alcohol consumption induces significant structural and functional alterations in hepatic mitochondria, contributing substantially to ALD progression.¹⁰ These alterations manifest as induction of mPTP, leading to increased permeability, depolarization, swelling, and damage to mitochondrial membranes.^41,42 Chronic ethanol exposure has been demonstrated to induce mPTP formation, ultimately leading to mitochondrial dysfunction and cell death.⁴³ However, the underlying mechanisms are not well elucidated. CypD, a critical regulator of the mPTP complex, is implicated in the pathophysiology of various liver conditions, including hepatic ischemia/reperfusion injury and metabolic-dysfunction-associated fatty liver disease.^22,23 Its translocation to the inner mitochondrial membrane triggers mPTP opening, manifesting in loss of mitochondrial membrane potential, decreased cristae density, mitochondrial swelling, and reduced ATP production.^44,45 These deleterious effects are reversible with pharmacological or genetic inhibition of CypD,^46,47,48 highlighting CypD as a promising target for therapeutic intervention in ALD.

Our findings positioned mitochondrial CypD as a central effector in ALD, echoing its established role in metabolic-dysfunction-associated steatotic liver disease (MASLD) where CypD mediated lipid-peroxidation-induced mPTP activation and hepatocyte death.^35,49 This mechanistic overlap underscored mitochondrial permeability transition as a convergent pathway in etiologically distinct liver diseases. However, our data revealed critical distinctions rooted in ethanol-specific pathogenesis. While MASLD engaged CypD through systemic metabolic stress (e.g., insulin resistance and lipotoxicity),⁴⁹ ALD directly targeted mitochondria via ethanol-metabolism-derived insults: (1) ethanol induced an increase in CYP2E1 expression, generating excessive ROS and leading to oxidative stress; (2) acetaldehyde adducts impaired mitochondrial complex I–IV function, causing NAD⁺ depletion and sensitizing hepatocytes to CypD-dependent apoptosis. These direct metabolite-mitochondria interactions distinguished ALD from the secondary mitochondrial dysfunction characteristic of MASLD.

Cyclophilins A and B (CypA/CypB) are implicated in MASLD and general hepatic injury pathways. CypA promotes inflammatory signaling via NF-κB activation,⁵⁰ while CypB modulates endoplasmic reticulum (ER) stress responses.⁵¹ However, our data reveal that hepatocyte knockdown of CypA (Ppia) or CypB (Ppib) partially mitigated ALD; it did not meaningfully restore mitochondrial dysfunction. This suggests that while these cyclophilins may contribute to inflammation or ER stress, they are not central to the mitochondrial dysfunction observed in ALD. Unlike cytosolic CypA and ER-localized CypB, CypD is the sole mitochondrial cyclophilin isoform and a direct regulator of the mPTP. Alcohol metabolism induces oxidative stress specifically within mitochondria, triggering CypD-dependent mPTP opening. This mechanistic divergence explains why CypD knockout mice exhibited profound protection in our ALD models, whereas studies in MASLD highlight CypA/B roles. RN-0001’s therapeutic efficacy in wild-type ALD models mirrored CypD ablation and was entirely abolished in CypD^ΔHep mice. This genetic epistasis confirmed that RN-0001’s benefits depend specifically on CypD targeting, despite its broader in vitro profile. Thus, in ALD, CypD inhibition alone was sufficient to mitigate pathology, and RN-0001’s effects were mechanistically attributable to CypD blockade. Our work demonstrated that selective CypD inhibition was both necessary and sufficient for efficacy in experimental ALD. Future studies exploring isoform-specific inhibitors could further refine therapeutic strategies for distinct liver diseases.

Pan-cyclophilin inhibitors like CRV431 have shown promise in ALD through broad antifibrotic effects.⁵² However, our data demonstrate that CRV431’s efficacy in ALD is limited compared to RN-0001. This outcome is also consistent with how CRV431 has been positioned and evaluated in the literature. Preclinical and translational studies of CRV431/rencofilstat have predominantly emphasized anti-fibrotic biology, namely reductions in liver fibrosis across chronic liver disease models and modulation of extracellular matrix (ECM) remodeling, rather than robust anti-steatotic effects as a primary endpoint. Early clinical development has similarly focused on fibrosis-linked readouts (e.g., collagen/fibrogenesis-associated biomarkers) in patients with stage F2/F3 NASH.⁵³ Together, these considerations support a model in which pan-cyclophilin inhibition may preferentially modulate fibrosis/ECM-driven pathology, whereas RN-0001, by prioritizing the mitochondrial cyclophilin axis central to early ethanol-triggered injury, achieves a more pronounced reduction in ALD-associated steatosis under our experimental conditions.

Most importantly, the preclinical and phase I clinical data collectively establish RN-0001 as a viable candidate for phase II trials in early-stage ALD. Demonstrated safety margins across species (NOAELs in rodents and non-rodents) and acceptable tolerability in healthy volunteers (mild-to-moderate AE incidence comparable to placebo) provide translational confidence for human studies. These findings warrant phase II evaluation of RN-0001’s efficacy in ALD patients, prioritizing mitochondrial function biomarkers alongside conventional clinical endpoints to validate its mechanism-driven therapeutic potential.

Our findings reveal that ethanol aggravates liver injury primarily through its impact on hepatocyte mitochondria: the oxidative metabolism of ethanol induces mPTP activation, leading to mitochondrial depolarization, ATP depletion, and cell death, all of which can be effectively counteracted by CypD inhibition. RN-0001, a targeted CypD inhibitor, emerges as a promising and safe therapeutic option for ALD, providing an approach to ameliorate ethanol-induced hepatocyte injury and potentially reversing mitochondrial dysfunction in ALD models. Given the current lack of effective pharmacotherapies for ALD, RN-0001 offers a significant opportunity for forthcoming clinical trials.

Limitations of the study

There are several limitations to this study that should be addressed. First, while intravenous (i.v.) dosing was used in the phase I trial to ensure precise control over systemic exposure and to facilitate clear safety and pharmacokinetic evaluations, the murine efficacy studies employed intraperitoneal (i.p.) administration. This was because repeated i.v. injections in small rodents can cause procedural stress and local tissue injury, which may confound liver injury and inflammation readouts. In contrast, i.p. dosing provides more consistent systemic exposure with fewer technical and welfare-related issues. Additionally, dosing frequency varied across in vivo studies: ALD efficacy experiments used once-daily i.p. dosing to align with the ethanol challenge schedule, whereas the 90-day mouse safety study used an every-other-day (q2d) regimen to minimize long-term injection burden and enhance tolerability. Lastly, the oral formulation and absolute bioavailability of RN-0001 have not yet been established. Developing an oral formulation and conducting bioavailability studies will be key priorities for future research.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Jia Xiao (edwinsiu@connect.hku.hk).

Materials availability

All unique reagents generated in this study are available from the lead contact with a completed Materials Transfer Agreement.

Data and code availability

• •The RNA-seq data obtained from mouse liver were deposited at GEO repository under GSE316128. The chemical structure of RN-0001 has been deposited in PubChem under CID: 146639799. Publicly available datasets analyzed in this study include a human liver scRNA-seq data (Human Protein Atlas; GEO: GSE115469), a mouse liver scRNA-seq dataset (Single Cell Portal: SCP1404), and a human liver alcohol-associated liver disease cohort transcriptome dataset (GEO: GSE142530).
• •This study did not generate code.
• •Study data are available within the article and its supplemental information. Any additional information required to reanalyze the data reported in this work paper is available from the lead contact upon request.

Acknowledgments

This study was supported by National Natural Science Foundation of China U23A20401 and 82122009 to J.X.; the Science and Technology Projects in Guangzhou 202201020066 to J.X.; and the China Postdoctoral Science Foundation 2025M782024 to Z.C..

Author contributions

J.X., F.M., and C.-P.M. conceived and designed the study. J.X., Z.C., Z.L., D.X., F.M., H.Z., Y.S., S.G., Y.Y., H.W., C.-P.M., and K.-F.S. contributed to study implementation through technical support, resource provision, and methodological expertise. J.X., Z.C., Z.L., D.X., F.M., H.Z., Y.S., S.G., and Y.Y. contributed to the experimental and clinical data acquisition. J.X., Z.C., D.X., F.M., S.G., and C.-P.M. contributed to data analysis and interpretation. J.X., Z.C., Z.L., D.X., F.M., H.Z., Y.S., S.G., and Y.Y. provided formal data analysis and curation. J.X. F.M., C.-P.M., and K.-F.S. secured funding for the study. J.X. drafted the initial manuscript. J.X., D.X., F.M., H.W., C.-P.M., and K.-F.S. made revisions and edited the manuscript. D.X., F.M., S.G., and C.-P.M. performed all statistical analyses for the phase I study and had complete access to the study data. All authors contributed to manuscript development, review, and approval and take full responsibility for the accuracy of the data, adherence to the study protocol, and the validity of the statistical analysis.

Declaration of interests

C.-P.M. is a founder of Farsight Medical Technology (Shanghai) Co., Ltd. and a member of its scientific advisory board. F.M. is co-founder, and D.X. and S.G. are employees of Farsight Medical Technology (Shanghai) Co., Ltd.. We have a patent application related to this work (Application no.: PCT/CN2024/079653).

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Anti-Mitofusin 1 antibody	Abcam	Cat#ab221661; RRID:AB_2941083
Anti-Mitofusin 2 antibody	Abcam	Cat#ab124773; RRID:AB_10999860
Anti-Cytochrome C antibody	Abcam	Cat#ab133504; RRID:AB_2802115
Anti-Nrf2 antibody	Abcam	Cat#ab92946; RRID:AB_10561604
Anti-SOD1 antibody	Abcam	Cat#ab16831; RRID:AB_302535
Anti-Catalase antibody	Abcam	Cat#ab52477; RRID:AB_868694
Anti-CYP2E1 antibody	Abcam	Cat#ab28146; RRID:AB_2089985
Anti-DRP1 antibody	Cell Signaling Technology	Cat#8570; RRID:AB_10950498
Anti-FIS1 antibody	Cell Signaling Technology	Cat#32525; RRID:AB_2927422
Anti-OPA1 antibody	Cell Signaling Technology	Cat#80471; RRID:AB_2734117
Anti-Bcl-2 antibody	Cell Signaling Technology	Cat#3498; RRID:AB_1903907
Anti-Bax antibody	Cell Signaling Technology	Cat#2772; RRID:AB_10695870
Anti-PCNA antibody	Cell Signaling Technology	Cat#13110; RRID:AB_2636979
Anti- iNOS antibody	Cell Signaling Technology	Cat#13120; RRID:AB_2687529
Anti-Cleaved Caspase-3 antibody	Cell Signaling Technology	Cat#9661; RRID:AB_2341188
Anti-NF-kappaB p65 antibody	Cell Signaling Technology	Cat#8242; RRID:AB_10859369
Anti-Phospho-NF-kappaB p65 antibody	Cell Signaling Technology	Cat#3033; RRID:AB_331284
Anti-GAPDH antibody	Cell Signaling Technology	Cat#2118; RRID:AB_561053
Anti-Phospho-Nrf2 antibody	Affinity Biosciences	Cat#DF7519
Bacterial and virus strains
AAV8-ApoE/hAATp-EGFP-FT2A-NM_134084-3Flag-SV40 PolyA	GeneChem	N/A
AAV8-hAAT-shPpia-EGFP	GeneChem	N/A
AAV8-hAAT-shPpib-EGFP	GeneChem	N/A
Biological samples
Healthy/ALD adult liver tissue	This paper	N/A
Chemicals, peptides, and recombinant proteins
RN-0001	Farsight Medical Technology	N/A
Debio-025	MedChemExpress	Cat#HY-12559
CRV431	MedChemExpress	Cat#HY-135644
Critical commercial assays
Cytotoxicity LDH Assay Kit	MedChemExpress	Cat#HY-K1090
Apoptosis detection kit	Vazyme	Cat#A211-01
Calcineurin phosphatase assay kit	Enzo Life Sciences	Cat#BML-AK804-0001
Calcium colorimetric assay kit	Beyotime Biotechnology	Cat#S1063S
Mitochondria isolation kit	MedChemExpress	Cat#HY-K1061
ATP/ADP ratio assay kit	Abcam	Cat#ab65313
Triglyceride content detection kit	Solarbio	Cat#BC0625
Non-esterified FFA assay kit	Elabscience	Cat#E-BC-K013
Deposited data
RN-0001	This paper	PubChem CID: 146639799
RNA-seq data of mouse liver	This paper	GEO: GSE316128
Single-cell RNA-seq data of human liver	MacParland et al.54	GEO: GSE115469
Single-cell RNA-seq data of mouse liver	Chen et al.55	SCP1404
RNA-seq data of human liver ALD cohort	Massey et al.56	GEO: GSE142530
Experimental models: Cell lines
RAW264.7	ATCC	ATCC Cat#TIB-71
AML12	ATCC	ATCC Cat#CRL-2254
Experimental models: Organisms/strains
Mouse: C57BL6/J	GemPharmatech	Strain#N000295
Oligonucleotides
Primers, see Table S6	This paper	N/A
Software
GraphPad Prism 10	https://www.graphpad.com/scientific-software/prism	N/A
ImageJ	https://imagej.nih.gov/ij/	N/A
Other
Trial registration number: ChiCTR2500106709	chictr.org.cn	https://www.chictr.org.cn/showproj.html?proj=278532

Experimental model and study participant details

Human subjects

Alcohol-associated hepatitis liver tissue cohort. Liver tissue specimens were obtained from four healthy controls and four patients with alcohol-associated hepatitis at the Third Affiliated Hospital of Sun Yat-sen University. Ethical approval was granted by the ethics committee of the Third Affiliated Hospital of Sun Yat-sen University (IRB: RG2024-049-01). Written informed consent was obtained from all participants before tissue collection.

First-in-human phase I clinical trial cohort. Healthy adult volunteers (18–45 years; BMI 18.5–29.9 kg/m²) were enrolled to evaluate the safety RN-0001 in a first-in-human phase I study (total n = 80). The trial was approved by China’s drug evaluation authority (CXHL2000510) and registered in the Chinese Clinical Trial Register (ChiCTR2500106709, https://www.chictr.org.cn/showproj.html?proj=278532). Written informed consent was obtained from all participants prior to enrollment. The study included both single-ascending-dose and multiple-dose parts. Further details of trial design, dosing, sampling, and safety assessments are described in METHOD DETAILS.

Animals

Male C57BL/6J mice (7 weeks old) were maintained in a specific pathogen-free facility with controlled temperature and a 12-h light/12-h dark cycle. All animal procedures were approved by the Institutional Animal Care and Use Committee (IACUC) of Jinan University (Approval No. IACUC-20220708-06).

Cell lines

Mouse RAW264.7 macrophages were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) at 37°C in a humidified incubator with 5% CO₂. AML12 hepatocytes were cultured in DMEM/F12 supplemented with 10% FBS, insulin (5 μg/mL), transferrin (5 μg/mL), selenium (5 ng/mL), and dexamethasone (40 ng/mL) at 37°C with 5% CO₂.

Method details

NIAAA model of alcohol-associated liver disease and drug administration

The National Institute on Alcohol Abuse and Alcoholism (NIAAA) chronic-plus-binge ethanol feeding model was used.²⁷ Mice were acclimatized to a liquid diet by feeding a control Lieber-DeCarli liquid diet for 5 days. Mice were then fed a Lieber-DeCarli ethanol diet containing 5% (v/v) ethanol for 10 days, while control groups received an isocaloric control diet. Dietary intake and body weight were monitored daily. On day 16, mice received a single intragastric gavage of ethanol (5 g/kg body weight) or isocaloric dextrin-maltose and were euthanized 9 h later. Based on pilot studies of acute and chronic ethanol-induced liver injury (data not shown), RN-0001 was administered intraperitoneally at 10 mg/kg/day. CRV431 was administered intragastrically at 50 mg/kg/day.⁵⁷ All experimental groups, including CypD^F/F and CypD^ΔHep cohorts, underwent identical feeding regimens according to the NIAAA protocol (Figure 1A).

Generation of hepatocyte-specific CypD-deficient mice (CypD^ΔHep)

CypD^ΔHep mice (Alb-Cre⁺ Ppif^F/F) were generated using CRISPR/Cas9-based genome editing targeting intronic regions flanking Ppif exons 4–5 on mouse chromosome 14, followed by germline transmission and breeding to obtain genetically stable lines. Cas9 protein and two gRNAs (gRNA-B1: GCATAGAAGAGGATTCGGGC-GGG; gRNA-B2: CGATGGGATGACAGCCACAC-AGG) were co-injected into fertilized mouse zygotes. Injected embryos were implanted into recipient females to generate F0 founders. F0 founders were screened by PCR and sequencing. Genotyping/sequence validation used primers (F1/R1 and F2/R2 as listed): F2: 5′-GTGTCTACAAAGCTGTTGGGCTT-3’; R2: 5′-AATTCCTAAGAGGACCACAAGCAG-3’. Sequencing-confirmed founders were crossed with wild-type mice to obtain F1 heterozygotes, which were genotyped and intercrossed to generate the experimental cohorts. Seven-week-old male Alb-Cre⁺ Ppif^F/F mice and Alb-Cre⁻ Ppif^F/F age-matched littermates were subjected to the NIAAA model.

AAV8-mediated hepatic CypD overexpression

AAV8-CypD (Ppif) was engineered and packaged by GeneChem. For hepatic overexpression, 7-week-old male C57BL/6J mice received a tail-vein injection of AAV8-CypD (2 × 10ˆ11 viral genomes in 0.2 mL saline). The AAV8 vector was driven by the HAAT promoter for hepatocyte-preferential expression. Overexpression was confirmed 14 days post-injection by western blot (Figure S3B). Mice were then subjected to the NIAAA model two weeks after AAV injection.

AAV8-mediated knockdown of Ppia and Ppib in hepatocytes

For hepatocyte-specific knockdown of Ppia and Ppib, AAV8-shPpia and AAV8-shPpib were administered via tail-vein injection (2 × 10ˆ11 viral genomes in 0.2 mL saline). Knockdown efficiency was assessed 14 days post-injection by RT-qPCR.

Long-term safety assay

For long-term safety evaluation, male mice were acclimatized for 1 week and randomized to control or RN-0001 groups. The RN-0001 group received intraperitoneal RN-0001 (10 mg/kg) once every two days for 90 days; controls received a corresponding volume of normal saline. Mice were monitored daily for general appearance and health status. On day 91, mice were euthanized.

Synthesis of RN-0001

RN-0001 was synthesized by modification of cyclosporin A (CsA) as follows: Step 1: Synthesis of [2'-(2-Thiopyridyl)-Sar]³-cyclosporine A (Compound 2). Cyclosporine A (20 g, 16.6 mmol), anhydrous lithium chloride (21.1 g, 499 mmol), and dry THF (500 mL) were combined in a dry 1 L flask, purged with argon, and cooled to −45°C. Diisopropylamine (13.5 g, 133 mmol) in THF (120 mL) was cooled to −78°C, and n-butyllithium (53.2 mL of 2.5 M solution, 133 mmol) was added. After 20 min, the diisopropylamide solution was added to cyclosporine, and the mixture was stirred at −45°C for 90 min. A solution of 2-pyridyldisulfide (11 g, 49.9 mmol) in THF (20 mL) was added dropwise, and the mixture was allowed to warm to room temperature overnight. The reaction was quenched with saturated NaCl (200 mL), and the organic phase was separated. After washing and drying, the compound was isolated by silica gel chromatography (7.18 g). Step 2: Synthesis of RN-0001. Copper triflate (0.291 g, 0.8 mmol) and 3 Å molecular sieves were added to a flask containing dry THF (3 mL), purged with argon. A mixture of Compound 2 (0.293 g, 0.223 mmol), dimethylaminoethanol (0.086 g, 0.96 mmol), and 3 Å sieves in THF (2 mL) was stirred for 30 min and added to the copper triflate solution. The mixture was stirred at room temperature overnight, then treated with NaHCO₃ (10 mL) and filtered through Celite. The organic layer was separated, extracted with ethyl acetate, and dried. The crude product was purified by silica gel chromatography to yield RN-0001 (86.4 mg). For full synthetic procedures, please refer to WO2019016572 (https://patentscope2.wipo.int/search/zh/detail.jsf?docId=WO2019016572) and WO2024183651 (https://patentscope.wipo.int/search/zh/detail.jsf?docId=WO2024183651&_cid=P20-MK9FCO-99174-1).

Determination of calcineurin activity

Calcineurin activity was measured using a calcineurin phosphatase assay kit. CsA or RN-0001 (complexed with cyclophilin A) was incubated with calcineurin and calmodulin in 96-well plates at ambient temperature for 30 min. The dephosphorylation reaction was initiated by adding RII phosphopeptide substrate and incubating for 15 min at 30°C. A colorimetric detection reagent was added, incubated for 16 min at room temperature, and absorbance at 620 nm was measured.

Cyclophilin inhibition assay

Inhibitor Ki values were determined using a previously described protease-free peptidyl-prolyl isomerase (PPIase) assay.⁵⁸ The assay was performed in 35 mM HEPES buffer (pH 7.8) containing 4 nM bovine serum albumin at 9.5°C using Suc-Ala-Ala-Pro-Phe-pNA as substrate. Ki values were calculated from assays performed with 8–11 inhibitor concentrations.

Ethanol-induced hepatocyte injury model

To model macrophage-driven paracrine amplification of ethanol-stressed hepatocyte injury, RAW264.7-conditioned medium (CM) was generated and applied to AML12 hepatocytes with minor modifications as previously described.³² RAW264.7 were plated and allowed to adhere for 24 h, serum-starved, and then pretreated with RN-0001, Debio-025, or CRV431 at indicated concentrations for 2 h, followed by LPS stimulation (10 ng/mL) for 6 h. Supernatants were collected and clarified by centrifugation to remove cells/debris to generate CM. AML12 hepatocytes at ∼60–70% confluence were treated with RN-0001, Debio-025, or CRV431 in RAW264.7 CM supplemented with ethanol to a final concentration of 50 mM. Controls were performed in parallel as specified for each experiment.

siRNA knockdown and plasmid overexpression in AML12 hepatocytes

For loss-of-function experiments, AML12 were transfected with siRNAs targeting mouse Ppif (RefSeq NM_134084), Ppia (RefSeq NM_008907), Ppib (RefSeq NM_011149), or a non-targeting control. For gain-of-function experiments, the mouse Ppif coding sequence (RefSeq NM_134084) was PCR-amplified and cloned into pcDNA3.1. Transfections were performed with Lipofectamine 3000 following the manufacturer’s instructions. Cells were transfected for 6 h, then medium was replaced with complete growth medium for an additional 18 h recovery before use in the in vitro ethanol injury model. Knockdown/overexpression efficiency was verified by measuring Ppif, Ppia or Ppib mRNA and/or protein levels.

PCR genotyping

Ppif genotyping was performed using PCR with primers: forward 5′-GTGTCTACAAAGCTGTTGGGCTT-3′ and reverse 5′-AATTCCTAAGAGGACCACAAGCAG-3’. Cycling conditions were 35 cycles of 94°C for 30 s, 60°C for 35 s, and 72°C for 35 s, distinguishing the mutant allele (205 bp) from the wild-type allele (138 bp).

Real-time quantitative PCR (RT-qPCR)

Total RNA was extracted from liver tissue or cells using TRIzol. cDNA was synthesized using PrimeScript RT Master Mix. qPCR was performed using SYBR Green qPCR Master Mix. Relative expression was calculated using the 2^−ΔΔCt method.⁵⁹ Primer sequences are provided in Table S6.

Cell lipid accumulation and viability assays

For lipid droplet staining, AML12 were plated in 24-well plates (5 × 10ˆ4 cells/well) and stained with Nile Red (0.1 mg/mL) for 30 min at 37°C, washed with PBS, fixed in 4% formaldehyde for 10 min, counterstained with DAPI (5 mg/mL) for 10 min, washed, and imaged. For cell viability, AML12 were seeded in 96-well plates (8 × 10ˆ3 cells/well) and incubated with 10% CCK-8 solution for 2 h at 37°C, followed by absorbance measurement at 450 nm. LDH release was measured using the LDH cytotoxicity assay kit following the manufacturer’s protocol.

Flow cytometric analysis of apoptosis

Cells were treated with RN-0001 or Debio-025 at indicated concentrations for 24 h, collected, washed with PBS, and resuspended in 100 μL binding buffer. Annexin V-FITC and PI (5 μL each) were added and incubated for 15 min at room temperature in the dark. Samples were analyzed by flow cytometry (Becton Dickinson Biosciences, Franklin Lakes, NJ).

Biochemical assays

Serum ALT, AST, TG, TC, and blood urea nitrogen (BUN) were measured using a Chemray 800 analyzer (Rayto Life and Analytical Sciences, Shenzhen, China). Serum TNF-α and IL-6 were quantified by ELISA kit. Hepatic TG and free fatty acids were measured using commercial kits.

Histological assays and NAS scoring

Liver tissues were fixed in 10% formalin and embedded in paraffin. Sections (5 μm) were stained with hematoxylin and eosin (H&E). For Oil Red O staining, tissues were processed as appropriate for lipid staining. The NAFLD Activity Score (NAS) was calculated for each group according to the following criteria: Steatosis grade: 0 (none), 1 (≤33%), 2 (34–66%), 3 (≥67%); Lobular inflammation grade: 0 (none), 1 (1–2 foci/200× field), 2 (3–4 foci/200× field), 3 (≥5 foci/200× field); Hepatocellular ballooning: 0 (none), 1 (few), 2 (many). The final NAS was determined by summing the scores from the three components (steatosis, inflammation, and ballooning).⁶⁰

Immunofluorescence

Frozen sections on coverslips were fixed with 4% paraformaldehyde for 10 min at room temperature. Sections were blocked in TBS containing 1% BSA, 0.2% Triton X-100, and 5% normal goat serum, then incubated with primary antibodies overnight at 4°C. Sections were incubated with secondary antibodies for 4 h at room temperature, washed with TBS, counterstained with DAPI, and mounted with ProLong Gold. Images were acquired using a Leica fluorescence microscope. Negative controls used non-specific IgG instead of primary antibodies.

In situ detection of mitochondrial ROS

Mitochondrial ROS in liver was detected using MitoSOX Red (1:2000) as previously described.²⁹ Images were acquired in a blinded manner using consistent acquisition parameters. MitoSOX positive area was quantified using ZEN software.

Protein extraction and western blotting

Cells or liver tissues were washed three times with PBS and lysed in RIPA lysis and extraction buffer supplemented with protease and phosphatase inhibitors. Protein concentration was determined using BCA protein concentration assay kit. Proteins were subjected to standard western blotting procedures, with GAPDH used as a loading control. Signals were detected using ECL reagents and imaged on an Amersham Imager 600 (GE HealthCare, Chicago, IL).

RNA sequencing and analysis

Total RNA was extracted using TRIzol. RNA purity was measured using a KaiaoK5500 spectrophotometer (Kaiao, Beijing, China), and RNA integrity was assessed using an Agilent Bioanalyzer 2100 with the RNA Nano 6000 kit (Agilent, Santa Clara, CA). Libraries were prepared from 2 μg RNA per sample using the NEBNext Ultra RNA Library Prep Kit for Illumina (#E7530L; New England Biolabs, Ipswich, MA) with indexed adapters, clustered using HiSeq PE Cluster Kit v4-cBot-HS, and sequenced as 150 bp paired-end reads on an Illumina platform. Adapters, poly-N, and low-quality reads were removed to generate clean reads. Clean reads were aligned to the mouse reference genome using HISAT2, and gene counts were obtained using HTSeq. Differential expression analysis was performed using DESeq, defining differentially expressed genes as q ≤ 0.05 (Benjamini–Hochberg correction) and |log2 fold change| ≥ 1.

Transmission electron microscopy

Liver ultrastructure was examined as previously described.⁶¹ Liver tissues were cut into 1 mm³ cubes, washed, and post-fixed in 1% OsO₄ plus 1.5% KFeCN₆ for 1 h. Samples were dehydrated in graded ethanol, embedded in TAAB Epon resin, and polymerized at 60°C for 48 h. Ultrathin sections (60 nm) were cut using a Leica EM UC7 ultramicrotome and imaged using a Hitachi HT7700 TEM at 120 kV.

Molecular docking

Docking was performed to model RN-0001 binding to CypD. The human CypD structure was obtained from the Protein DataBank (PDB: 3QYU). A mouse CypD homology model was generated using the human template and the mouse Ppif sequence. Proteins were prepared by removing crystallographic ligands/water, adding missing atoms where needed, assigning protonation states near physiological pH, and minimizing to relieve clashes. RN-0001 was built in 3D, assigned relevant protonation/tautomeric states near physiological pH, and optimized. Docking centered the grid on the canonical PPIase active pocket (cyclosporin-class inhibitor-binding groove). Top-ranked poses with reasonable pocket geometry were selected for subsequent MD simulations.

Molecular dynamics (MD) simulations

Protein–ligand complexes were parameterized using standard biomolecular force fields. Systems were solvated in explicit water under periodic boundary conditions with counterions and physiological ionic strength. After staged minimization, heating, and equilibration, 200 ns production simulations were performed. Trajectories were processed to remove periodicity artifacts and aligned to the protein backbone. Binding free energy was estimated by MM/GBSA using snapshots from the equilibrated portion of the trajectories.

MST and SPR binding assays

For MST, labeled cyclophilin proteins (100 nM) were incubated with RN-0001 (50 μM–1.53 nM) for 30 min at ambient temperature. Samples were loaded into premium capillaries and measured on a Monolith NT.115 Blue/Green instrument (NanoTemper) at 60% MST power and 25°C. Kd values were calculated assuming 1:1 binding. For SPR, experiments were performed on a Biacore T200 (GE Healthcare). NTA chips (#BR100407) were activated with 0.5 mM NiCl₂ at 10 μL/min for 60 s. Running buffer was PBST (1× PBS, 0.005% Tween 20, pH 7.4) with 1% DMSO as indicated. Protein (5 μg/mL) was captured on Fc2 at 10 μL/min for 120 s. RN-0001 was injected at 250, 125, 62.5, 31.25, 15.625, 7.8, and 0 nM at 30 μL/min with 120 s association and 300 s dissociation. Seven cycles were run in ascending concentration. Surfaces were regenerated with 350 mM EDTA for 30 s. Kinetic parameters were analyzed using Biacore T200 Evaluation Software.

Tissue calcium measurement

Liver tissues were homogenized in lysis solution (150–250 μL per 20 mg tissue) and centrifuged at 14,000 × g for 5 min at 4°C. Supernatants were kept on ice and assayed using a calcium colorimetric kit.

ATP/ADP ratio assay

ATP/ADP ratio was determined using the Abcam ATP/ADP kit with minor modifications. Liver tissue was washed in cold PBS, homogenized, centrifuged, and luminescence was measured on a Luminoskan reader to quantify ATP and ADP according to the kit protocol. Ratios were normalized to protein concentration by BCA assay.

Mitochondria isolation

Liver mitochondria were isolated as previously described.²⁹ Liver tissue was homogenized in ice-cold buffer (225 mM mannitol, etc.) using a Dounce homogenizer (DWK Life Sciences, Millville, NJ). After centrifugation (1,300 × g, 5 min) to remove debris, supernatants were layered on 15% Percoll (GE Healthcare) and centrifuged (34,000 × g, 10 min). Pellets were washed twice with BSA-free buffer (8,000 × g, 10 min each). Digitonin treatment (0.02%, 5 min) followed by centrifugation (8,000 × g, 10 min) was used to reduce cytosolic contamination. Protein concentration was determined by Bradford assay.

Measurement of mitochondria bioenergetic function

Mitochondrial respiration was assessed on an XF24 analyzer (Seahorse Bioscience). Mitochondria pooled from three mice per group were resuspended in mitochondrial assay solution (e.g., 1× MAS with 70 mM sucrose; pH 7.2) containing 10 mM succinate and 2 μM rotenone to 0.2 mg/mL. After centrifugation (2000 × g, 20 min, 4°C), 50 μL was loaded per well of an XF24 plate and 450 μL assay buffer was added. Ports contained ADP (40 mM), oligomycin (25 μg/mL), FCCP (40 μM), and antimycin A (40 μM). OCR was measured sequentially under substrate-only, ADP-stimulated, oligomycin-inhibited, FCCP-stimulated, and antimycin A–inhibited conditions.

Mitochondrial swelling assay

Isolated mitochondria (20 μg protein/mL) were resuspended in swelling buffer (150 mM KCl, pH 7.2). Ca²⁺ (1 μmol/mg protein) was added to induce mPTP opening. Swelling was monitored for 12 min by optical density using a microplate reader (Molecular Devices, San Jose, CA).⁶²

Assessment of mitochondrial respiration

Mitochondrial respiration was measured using a Clark-type oxygen electrode (Hansatech, Norfolk, UK). Isolated mitochondria (0.05 mg) were incubated in pre-warmed respiration buffer (0.5 mL, 10 mM MOPS-Tris, etc., pH 7.4) with glutamate (10 mM) and malate (5 mM) at 32°C. State 3 respiration was initiated with 100 nmol ADP. Respiratory control ratio (State 3/State 4) was calculated. DNP (35 nM) was used to initiate uncoupled respiration. ADP/O ratios were calculated per experiment.

Measurement of mitochondrial transmembrane electric current

Membrane potential was assessed using a TPP⁺ electrode in a continuously stirred chamber (0.3 mL, 37°C). The incubation medium (pH 7.4) contained mannitol (e.g., 215 mM) and other solutes as previously described.⁶³ Changes in TPP⁺ distribution were used to infer membrane polarization/depolarization.

Mitochondrial Ca²⁺ uptake/calcium retention capacity (CRC)

CRC was measured in isolated mitochondria (40 μg) at room temperature using Fluo-4 AM⁶⁴ Mitochondria were suspended in buffer (150 mM KCl, pH 7.2) and challenged with sequential CaCl₂ pulses (40 μM). Fluorescence was monitored continuously on a Synergy Neo 2 reader (Bio-Tek) at Ex/Em 506/531 nm. Loss of calcium retention was used as a readout of mPTP opening.

Pharmacokinetics

Sprague–Dawley rats (8 weeks old) received a single intraperitoneal dose of RN-0001 (1, 3, or 10 mg/kg). Blood was collected at indicated time points from 3.3 min to 24 h and stored at −20°C before LC-MS analysis. Non-compartmental pharmacokinetic parameters were calculated using Kinetica 5.0 with trapezoidal integration.

GLP toxicity studies (2-week of repeated dosing, 4-week of recovery) in rodents and non-rodents to support investigational new drug (IND) application

IND-enabling GLP studies included cardiovascular assessments in beagle dogs (n = 8; 4/M, 4/F; 21–46 months) receiving intraperitoneal RN-0001 (4.5, 9, or 18 mg/kg) with monitoring of mortality, clinical signs, body weight, temperature, blood pressure, heart rate, and ECG. Respiratory function was assessed in SD rats (n = 40; 20/M, 20/F; ∼9 weeks) receiving 10, 20, or 40 mg/kg i.p. CNS function was assessed in SD rats (n = 50; 25/M, 25/F; ∼10 weeks) receiving 10, 20, or 40 mg/kg i.p. General toxicity studies included SD rats (n = 120; 60/M, 60/F; ∼8 weeks) receiving 10, 20, or 40 mg/kg/day i.p., and beagle dogs (n = 40; 20/M, 20/F; 7–8 months) receiving 4.5, 9, or 18 mg/kg/day i.p. Animals were monitored daily, and necropsy was performed at study termination.

First-in-human phase I clinical trial

A first-in-human phase I study evaluated safety, tolerability, and exploratory pharmacokinetics of RN-0001 in healthy adults aged 18–45 years (BMI 18.5–29.9 kg/m²). The study was approved by China’s drug evaluation authority (CXHL2000510) and registered (ChiCTR2500106709). The design was randomized, double-blind, placebo-controlled with escalating single and multiple doses (total n = 80). Single ascending-dose cohorts generally included eight subjects (six RN-0001, two placebo), except the 0.1 mg/kg cohort (two RN-0001). Multiple dosing randomized ten male subjects to RN-0001 daily (n = 8) or placebo (n = 2) for eight days (Figure S12). Serial blood samples were collected for pharmacokinetic analyses. Safety assessments included adverse events, vital signs, and routine laboratory testing (hematology, coagulation, clinical chemistry, urinalysis) (Details are provided in Data S1).

Quantification and statistical analysis

Data are presented as mean ± SD. Statistical details (including n and exact tests) are reported in the figure legends. Two-group comparisons used unpaired t-tests (parametric) or Mann-Whitney U tests (non-parametric). Comparisons involving one control and multiple treatment groups used one-way ANOVA with Dunnett’s post hoc test (parametric) or Kruskal-Wallis with Dunn’s post hoc test (non-parametric). two-way ANOVA with Tukey’s multiple comparisons test and false discovery rate (FDR) correction was used for analyses involving two independent variables. Statistical significance was defined as p ≤ 0.05.

Published: March 17, 2026

Supplemental information

Data S1. Trial protocol and CONSORT checklist of the RN-0001-Ph-I

Related to STAR Methods.