Peptide YY reduces cytotoxicity of Candida albicans in alcohol-associated liver disease

Abstract

Background & Aims

Transitioning from yeast to hyphal morphology enables Candida albicans (C. albicans) to secrete candidalysin, invade the intestinal mucosa and translocate to the blood stream. Patients with alcohol-associated hepatitis show increased intestinal abundance of C. albicans, and the candidalysin-encoding gene is associated with reduced survival. Paneth cell-derived peptide YY (PC-PYY) inhibits hyphal growth of C. albicans. In this study, we evaluated the potential of different C. albicans strains isolated from patients with alcohol-associated hepatitis to cause systemic infections and explored the therapeutic potential of PC-PYY in ethanol-induced liver disease in mice.

Methods

C. albicans strains isolated from fecal samples of patients with alcohol-associated hepatitis (n = 105) were co-cultured with intestinal epithelial Caco-2 cells to assess in vitro cytotoxicity. Caco-2 cells and primary mouse hepatocytes were incubated with C. albicans in the presence or absence of PC-PYY. Mice were subjected to a chronic plus binge ethanol-feeding model.

Results

C. albicans strains isolated from stool of patients with alcohol-associated hepatitis induced significant cytotoxicity in Caco-2 cells, and high cytotoxicity was associated with worse 30-day survival (log-rank p = 0.032). This cytotoxicity was primarily mediated by the hyphal form and largely driven by candidalysin. PC-PYY significantly reduced C. albicans–induced cytotoxicity in Caco-2 cells (Wilcoxon rank-sum test, p = 0.015) and in primary mouse hepatocytes (p = 0.03) compared with a scrambled peptide control, by inhibiting hyphal morphogenesis. The peptide YY–to–chromogranin A ratio in intestinal crypts was significantly increased in ethanol-fed mice compared with both isocaloric (p = 0.005) and antifungal-treated controls (p = 0.009), indicating that fungal overgrowth stimulates PC-PYY release. In ethanol-fed mice, PC-PYY administration attenuated liver injury (p = 0.032) and steatosis (p = 0.0498) and reduced fecal hyphae formation (p = 0.0159).

Conclusion

PYY inhibits filamentous growth of C. albicans in vitro and alleviates ethanol-induced liver disease in mice, highlighting its potential as a therapy for patients with alcohol-associated liver disease.

Impact and implications

Candida albicans (C. albicans) and particularly its toxin candidalysin are associated with poor outcomes in patients with alcohol-associated hepatitis but the extent to which the cytotoxicity of individual C. albicans strains influences patient survival, and the role of Paneth cell-derived PYY (PC-PYY) in this context remains elusive. This study identifies a link between the cytotoxic effect of patient-derived C. albicans strains and survival in patients with alcohol-associated hepatitis and demonstrates that PC-PYY plays a protective role in ethanol-induced liver disease by limiting candidalysin-producing hyphae. Our work provides insight into why some patients with alcohol-associated liver disease have worse outcomes and highlights the potential of PC-PYY as a therapy for patients with alcohol-associated liver disease.

Graphical abstract

Highlights

• •Functional cytotoxicity of fecal C. albicans strains is associated with worse survival in patients with alcohol-associated hepatitis.
• •Paneth cell-derived peptide YY, which inhibits hyphal growth of C. albicans, is elevated in crypts of ethanol-fed mice.
• •Paneth cell-derived peptide YY reduced C. albicans-induced cytotoxicity of hepatocytes.
• •Paneth cell-derived peptide YY administration attenuated ethanol-induced liver disease in mice.

Introduction

Alcohol-associated hepatitis is an acute, cholestatic manifestation within the spectrum of alcohol-associated liver disease and is characterized by high mortality rates.¹ Short-term survival of patients with alcohol-associated hepatitis depends primarily on liver disease severity, assessed by the model for end-stage liver disease (MELD) and Lille scores, while long-term survival is significantly influenced by continued abstinence from alcohol.² Chronic alcohol intake changes the composition of the gut microbiome, reducing fecal fungal diversity and promoting the overgrowth of Candida albicans (C. albicans), which correlates with disease severity.[3], [4], [5], [6]

C. albicans is a polymorphic fungus that can grow in diverse morphologies, including yeast, pseudohyphae and hyphae.[7], [8], [9] The ability to transition from yeast to hyphae is a key virulence trait that facilitates tissue invasion and damage.¹⁰ However, both yeast and hyphae are found during gut colonization in models with a depleted bacterial microbiome, and the dimorphic transition is essential for the colonization of the intestinal environment in the presence of bacteria, as yeast-locked C. albicans cells are unable to colonize the mouse intestine.^11,12 Hyphae of C. albicans are characterized by the secretion of several hyphae-associated virulence factors, including the cytotoxin candidalysin, which is considered a key mediator of epithelial damage and fungal translocation across intestinal epithelial barriers.^8,9,13,14

The hyphal form of the fungus invades human epithelial cells through two distinct mechanisms: induced endocytosis and active penetration. Induced endocytosis is initiated by the host cell and facilitated by the fungal invasion protein Als3.^11,[15], [16], [17], [18] In contrast, active penetration is entirely driven by the fungus.^16,18 Candidalysin is encoded by the ECE1 gene, processed from a complex precursor polyprotein and secreted to an invasion pocket created by invading hyphae.^14,19,20 While hyphal invasion per se does not necessarily cause cell damage, candidalysin secretion into the invasion pocket induces significant epithelial damage, when expressed at high levels.¹⁴ In contrast, synthetic candidalysin alone is less cytotoxic than when produced by live filamenting, invading C. albicans hyphae.^13,19 Hence, effective damage of host cells requires coordinated fungal activities including epithelial adhesion, invasion, hyphal extension, high levels of ECE1 transcription, proper Ece1 processing, and delivery of the processed peptide toxin into the invasion pocket.¹⁹ Our own studies showed that genomic DNA encoding ECE1 is increased in the feces of patients with alcohol-associated hepatitis and its presence is linked to increased mortality.³

Peptide YY (PYY) is produced in enteroendocrine cells and Paneth cells in the intestine of mice and humans.²¹ Paneth cell-derived PYY (PC-PYY) is a 36 amino acid peptide with antimicrobial activity. In contrast enteroendocrine-derived PYY undergoes cleavage by dipeptidyl peptidase IV (DPP-IV), which removes the first two amino acids, generating a truncated form that functions as a satiety hormone.^21,22 Unlike its processed counterpart, PC-PYY is protected from DPP-IV degradation due to its secretion into the intestinal mucus.²² Functionally, PC-PYY shows selective antifungal activity against hyphae, but not yeast cells of C. albicans, thereby limiting the production of hyphal associated virulence factors, including candidalysin, and reducing fungal pathogenicity.^22,23 For the following intervention studies, we used the uncleaved peptide released by Paneth cells (PC-PYY).

Given the critical role of candidalysin in fungal pathogenesis and its association with disease severity in patients with alcohol-associated hepatitis,³ this study aimed to investigate the functional cytotoxicity of different C. albicans strains by analyzing their ability to induce candidalysin-mediated host cell damage. Furthermore, we explored the role of PC-PYY in ethanol-induced liver disease to better understand its potential as a therapeutic drug.

Patients and methods

Patient cohorts

Cohorts derived from two NIH-funded, independent, multicenter observational studies were studied in this manuscript. For screening of C. albicans-positive stool samples, we included 71 patients with alcohol-associated hepatitis from the Integrated Approaches for Identifying Molecular Targets in Alcoholic Hepatitis (InTeam) Consortium and 34 patients with alcohol-associated hepatitis from Alcoholic Hepatitis Network (AlcHepNet). Baseline characteristics for patients with alcohol-associated hepatitis are summarized in Table 1.

Table 1.

Clinical characteristics of patients with alcohol-associated hepatitis (n = 105).

	Median (IQR) or n (%)
Clinical parameters
Age (years), n = 102	32 (21–43)
Sex (% female)	33 (32%)
Laboratory parameters
Albumin (g/dl), n = 73	2.5 (2.2–3.15)
ALP (IU/L), n = 105	158 (121–161)
ALT (IU/L), n = 105	44 (30–66)
AST (IU/L), n = 105	120 (83–176)
Bilirubin (mg/dl), n = 78	17 (8–22)
Creatinine (mg/dl), n = 78	0.77 (0.58–1.13)
INR, n = 105	1.76 (1.5–2.2)
Cholesterol (mg/dl), n = 38	97 (29–158)
Triglycerides (mg/dl), n = 39	87 (32–154)
WBC (x106), n = 73	11 (7–16)
Hemoglobin (g/dl), n = 97	9.7 (8.5–11.4)
Platelets (x106), n = 73	129 (84–177)
Clinical presentation and treatment at admission
Steroids, n = 79	32 (41%)
Antibiotics, n = 79	24 (30%)
Clinical scores and outcome
MELD, median (range), n = 78	22 (17–28)
Child-Pugh stage (A/B/C), n = 102	1 (1%)/33 (33%)/66 (65%)
30-day-mortality, n = 105	11 (10%)

The number of patients for whom data was available is indicated in the first column.

ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; INR, international normalized ratio; MELD, model of end-stage liver disease; WBC, white blood cell count.

The InTeam patient cohort has been described.^5,6,[24], [25], [26] Fecal samples were used from patients with alcohol-associated hepatitis who were enrolled from June 2014 to May 2017 in the InTeam Consortium in 11 centers from five different regions worldwide (BCN Vall de’ Hebron, Lille, King’s College London, Monterrey Mexico, University of North Carolina Chapel Hill, University of Wisconsin Madison, Veterans Medical Research Foundation San Diego) (ClinicalTrials.gov identifier number: NCT02075918). Patients were eligible for the InTeam Consortium if they reported active alcohol abuse (>50 G/day for men and >40 G/day for women) in the last 3 months. Further inclusion criteria included aspartate aminotransferase (AST) higher than alanine aminotransferase (ALT), total bilirubin higher than 3 mg/dl in the past 3 months and liver biopsy and/or clinical picture consistent with alcohol-associated hepatitis. Liver biopsy was only performed when indicated as part of clinical routine. Exclusion criteria were autoimmune liver disease (anti-nuclear antibody >1/320), chronic viral hepatitis, hepatocellular carcinoma, complete portal vein thrombosis, extrahepatic terminal disease, pregnancy and lack of signed informed consent. Biospecimens were collected at admission to hospital.

The AlcHepNet cohort has been described.^26,27 Fecal samples were collected from patients with alcohol-associated hepatitis enrolled between April 2021 and July 2022 at seven clinical centers in the United States (Cleveland Clinic; University of Louisville; Beth Israel Deaconess Medical Center, Boston; University of Massachusetts; University of Pittsburgh; University of Texas Southwestern; and Virginia Commonwealth University). Patients aged ≥21 years with a clinical diagnosis of alcohol-associated hepatitis fulfilling the following criteria were included: Serum bilirubin >3 mg/dl; subject or guardian ability to understand and willingness to provide written consent; and re-enrollment of a patient with alcohol-associated hepatitis was permissible up to 4 times if the donor presents with a new episode of alcohol-associated hepatitis 24 weeks or longer after the most recent enrollment in the study (no cases of re-enrollment are included in this current study). Patients were excluded if any of the following criteria were present: hemochromatosis, autoimmune liver disease, Wilson disease, metabolic dysfunction–associated steatotic liver disease (MASLD; formerly NAFLD), or acute viral hepatitis; pregnancy or breastfeeding; or, in the investigator’s judgment, inability to understand or comply with the study requirements. Chronic hepatitis B, hepatitis C or HIV infection did not exclude patients in the absence of significant liver disease.

The protocol was approved by the Ethics Committee of each participating center and patients were enrolled after written informed consent was obtained from each patient.

Stool cultures and selection of Candida strains

Stool samples were stored at -80 °C. About 50 mg of stool was diluted with 1 ml of autoclaved water until dissolved. 200 μl of this suspension were spread on a yeast extract–peptone–dextrose (YPD) plate (1% yeast, 2% peptone, 2% dextrose, 100 μg/ml chloramphenicol, 100 μg/ml gentamicin) and incubated at 30 °C for 2 days. Colony forming units (CFUs) were counted and 10 colonies were randomly selected. Quantitative PCR (qPCR) for C. albicans was performed on single colonies using specific primers. A maximum of three C. albicans strains were selected per patient and grown overnight in a liquid YPD culture (5 ml) at 30 °C shaking 150 rpm. 660 μl of the liquid culture were mixed with 330 μl of 50% sterile glycerol and stored at -80 °C for further usage. CHROMagar™ Candida plus was prepared according to the manufacturer’s description.

qPCR fecal samples

DNA was extracted from cecum samples of mice using DNeasy PowerSoil kit (Qiagen) according to the manufacturer’s protocol. The abundance of Candida spp. was measured via qPCR using specific primers and normalized by total fungal 18S rRNA (CTAT table).^28,29

Cytotoxicity assay

Frozen C. albicans isolates were streaked onto YPD agar plates and grown for 48 h at 30 °C. A single colony was selected and inoculated into 5 ml liquid YPD medium (1% yeast extract, 2% peptone, 2% dextrose, 100 μg/ml chloramphenicol, 100 μg/ml gentamicin) and cultured at 30 °C with shaking at 150 rpm for 18–24 h. Following incubation, optical density was measured, and cultures were diluted to an OD600 of 0.1 in 10 ml of fresh YPD medium and grown for an additional 16–18 h at 30 °C with shaking at 150 rpm. Subsequently, 500 μl of culture were transferred to a new tube, washed twice with PBS (5,000 g, 5 min), and OD600 was measured to adjust the concentration to 1 × 10⁵ cells/ml in DMEM/F12 medium (1% FBS, 1% penicillin/streptomycin). The suspension was then added to prepared 96-well plates containing Caco-2 cells.

Caco-2 and intestinal Mode-K cells were stored in liquid nitrogen and, upon use, thawed and cultured in 10-cm culture dishes in DMEM or RPMI medium, respectively, supplemented with 10% FBS and 1% penicillin/streptomycin. Medium was changed every 2–3 days, and cells were passaged every 7 days. For experiments, cells were harvested, adjusted to 2 × 10⁵ cells/ml, and 200 μl were seeded per well in 96-well plates. After 48 h, the medium was replaced with DMEM/F12 or RPMI containing 1% FBS and 1% penicillin/streptomycin. Medium alone served as a negative control, and 1% Triton X-100 in medium served as a positive control. C. albicans and Caco-2 or Mode-K cells were co-incubated for 24 h at 37 °C in 5% CO₂. PC-PYY (20 μM; YPIKPEAPGEDASPEELNRYYASLRHYLNLVTRQRY; ChinaPeptides Co., Ltd.) or scrambled peptide (20 μM; HPYAPLLYTLPQADRSRVLSYEYRNGKEYRAIPENE; ChinaPeptides Co., Ltd.) was added at the start of incubation.²² Plates were then centrifuged (250 g, 5 min, 4 °C), and 50 μl of supernatant were transferred to flat-bottom 96-well plates. Three biological and two technical replicates were performed for each sample.

LDH activity in the supernatant was measured using the CyQUANT™ LDH Cytotoxicity Assay (Invitrogen, #C20300) according to the manufacturer’s instructions. Briefly, 50 μl of reaction mixture were added to 50 μl of supernatant and incubated at room temperature for 30 min. Subsequently, 50 μl of stop solution were added, and absorbance was measured at 490 nm and 680 nm. Final LDH activity was calculated by subtracting OD680 from OD490.

To calculate LDH cytotoxicity for individual isolates, the following formula was used:

$$Cytotoxity(\%)=\frac{\left(ODsample-ODnegativecontrol\right)}{\left(ODpositivecontrol-ODnegativecontrol\right)}$$

Isolation of primary mouse hepatocytes

After induction of anesthesia and surgical exposure of the inferior vena cava and portal vein, the inferior vena cava was cannulated with a 22G needle. The liver was perfused sequentially with washing buffer (HBSS, 0.5 mM EDTA, 25 mM HEPES) and digestion medium (DMEM high glucose, 1% penicillin/streptomycin, 1 mg/ml collagenase). Following excision, the liver was placed in a Petri dish containing hepatocyte medium and gently dissociated. Hepatocytes were counted and seeded onto collagen type I–coated 24-well plates at a density of 2 × 10⁵ cells per well. After 3 h, C. albicans and either PC-PYY (20 μM; YPIKPEAPGEDASPEELNRYYASLRHYLNLVTRQRY; ChinaPeptides Co., Ltd.) or scrambled peptide (20 μM; HPYAPLLYTLPQADRSRVLSYEYRNGKEYRAIPENE; ChinaPeptides Co., Ltd.) were added. LDH activity was measured 3 h after the addition of C. albicans and PC-PYY or scrambled peptide.

LDH release of primary mouse hepatocytes

C. albicans was cultured in 5 ml YPD medium (1% yeast extract, 2% peptone, 2% dextrose, 100 μg/ml chloramphenicol, 100 μg/ml gentamicin) for 18–24 h and adjusted to an OD600 of 0.1 in 2 ml fresh YPD medium. To induce hyphal growth, 1% FBS was added and samples were incubated at 37 °C with shaking at 150 rpm; to maintain yeast growth, samples were incubated at room temperature with shaking at 150 rpm. Where indicated, PC-PYY (20 μM; YPIKPEAPGEDASPEELNRYYASLRH; ChinaPeptides Co., Ltd.) was added to the YPD medium. After 16–18 h, 500 μl of culture was transferred to a new tube, washed twice with PBS, and adjusted to 1 × 10⁵ cells/ml in DMEM/F12 medium (1% FBS, 1% penicillin/streptomycin) containing a non-specific protease inhibitor (cOmplete™ Mini Protease Inhibitor Cocktail, Roche). Two hundred microliters of the suspension were added to prepared hepatocytes. Where indicated, 20 μM PC-PYY or scrambled peptide was added to the wells.

After 3 h of incubation at 37 °C in 5% CO₂, plates were centrifuged and LDH activity was measured as described above.

Mouse model of ethanol feeding

C57BL/6 mice were purchased from Charles River. All protocols were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee of the University of California, San Diego (La Jolla, CA). Mice (age, 8–10 weeks, male and female) were fed a chronic plus binge ethanol diet (NIAAA model) or a chronic Lieber DeCarli diet model for 8 weeks. For the NIAAA model, mice received an isocaloric control liquid Lieber DeCarli diet from day 0 to day 5. Afterwards, mice were fed a Lieber DeCarli diet containing 5% (vol/vol) ethanol for 10 days.³⁰ On the last day, mice were gavaged with a single dose of ethanol (31.5% (vol/vol)) at night and sacrificed 9 h later (the next morning). Pair-fed control mice were placed on a control liquid diet during the 16 days and received isocaloric maltodextrin before being euthanized 9 h later.

To study the effects of an antifungal drug on PYY production and ethanol-induced liver disease, mice received nystatin (40,000 unit/kg per day) which was dissolved in the liquid Lieber DeCarli ethanol-containing diet for the last 10 days.^3,31 Xanthan gum was added to liquid diet to increase viscosity of the liquid diet solutions and prevent gravitational separation of nystatin.

For the 8-week feeding model, mice were fed an isocaloric control liquid Lieber–DeCarli diet from day 0 to day 5. Thereafter, mice received a Lieber–DeCarli diet containing 4% (v/v) ethanol for 6 weeks, followed by 5% (v/v) ethanol for the final 10 days. On the last day, mice were gavaged at night with a single dose of ethanol (31.5% v/v) and sacrificed 9 h later (the following morning). Pair-fed control mice were maintained on the control liquid diet for the full 8 weeks, received isocaloric maltodextrin, and were euthanized 9 h after the final gavage. During the last 14 days of the study, mice were gavaged daily with either 100 μl of scrambled peptide (2 μg/μl; HPYAPLLYTLPQADRSRVLSYEYRNGKEYRAIPENE; ChinaPeptides Co., Ltd.) or PC-PYY (2 μg/μl; YPIKPEAPGEDASPEELNRYYASLRHYLNLVTRQRY; ChinaPeptides Co., Ltd.).

Isolation of small intestinal crypts and villi

The proximal and distal small intestine was removed from mice, and luminal contents were flushed with PBS. The tissue was then shaken in RPMI medium containing 1.5 mM dithiothreitol (DTT). The intestine was cut into small pieces, mixed with RPMI medium, and incubated for 20 min at 37 °C with shaking at 250 rpm. Cells from the villi were collected by passing the suspension through a 70 μm cell strainer and centrifuged at 280 g for 10 min. Tissue remaining in the strainer was transferred back into RPMI medium, incubated for 20 min at 37 °C with shaking at 250 rpm, and filtered again through a cell strainer. After centrifugation, the resulting pellet contained the intestinal crypts.

Liquid culture of feces

Feces was collected from mice at harvesting and stored at -80 °C. After thawing it was diluted 1:10 with PBS and 100 μl of the dilution was added to 5 ml YPD media +1% FBS. Cultures were incubated for 48 h at 37 °C with shaking at 150 rpm. After 48 h, samples were vortexed and 10 μl of each sample was transferred to a microscopy slide.

Biochemical analysis and histology

Serum alanine aminotransferase (ALT) and hepatic triglyceride levels were measured using the ALT (SGPT) Kinetic assay (Teco Diagnostics) and the Triglyceride Liquid Reagents kit (Pointe Scientific), respectively. Lipid accumulation in liver tissue was assessed by Oil red O staining (Sigma–Aldrich). Six to ten images per mouse were quantified using ImageJ, and representative images are shown for each experimental group.

qRT-PCR

Mouse liver RNA and intestinal RNA were extracted using TRIzol reagent (Life Technologies) and converted to cDNA using High-Capacity cDNA reverse Transcription Kit (Invitrogen). Gene expression was measured by ABI StepOnePlus real-time PCR system using SYBR Green (Bio-Rad Laboratories). Expression was normalized to 18S ribosomal RNA (rRNA) (CTAT table).

Immunofluorescence staining

After sectioning and deparaffinization, tissue sections were incubated in 3% H₂O₂ for 10 min. Antigen retrieval was performed using DAKO Antigen Retrieval Solution in a household steamer. Cells were permeabilized with 0.1% Triton X-100 and blocked with DAKO Protein Block before incubation with the first primary antibody, lysozyme recombinant rabbit monoclonal (ST50-02, Invitrogen), overnight at 4 °C. Slides were washed thoroughly and incubated with the first secondary antibody (Alexa Fluor 488 goat anti-rabbit) for 1 h at room temperature. After washing, slides were re-incubated with the first secondary antibody for 20 min, washed again, and incubated with the second primary antibody, PYY rabbit polyclonal (1:200; abcam ab22663), overnight at 4 °C.

The following day, slides were incubated with the second secondary antibody (Alexa Fluor 568 donkey anti-rabbit) for 15–20 min and mounted with immunofluorescence mounting medium containing DAPI. Two representative images were acquired using an Olympus IX71 microscope and analyzed with QuPath 0.5.1. A pixel classifier was first trained to detect lysozyme-positive cells and annotate the area, followed by a second classifier to detect Pyy-positive cells within the annotated region.

Immunoblot

Proteins from primary cultures of hepatocytes exposed to C. albicans and PYY or control peptide were extracted and homogenized in 150 μl RIPA buffer (89901, ThermoFisher) supplemented with protease inhibitors (11836153001, Roche) and phosphatase inhibitors (4906845001, Roche) as previously described.³² For the extraction of total liver proteins from mice fed an isocaloric or ethanol diet and treated with either PYY or control peptide, 600 μl of RIPA buffer was used.³³ The proteins were resolved on polyacrylamide gels by SDS-PAGE (Bio-rad), and then transferred to polyvinylidene difluoride membranes (1620175, Bio-rad). Immunoblot analysis was performed using anti-gasdermin D (cleaved and full-length peptides) (1:1,000) (MA5-44666, ThermoFisher), anti-IL1b (cleaved form) (1:1,000) (ab9722, Abcam) and anti-GAPDH (1:5000) (GTX627408, GeneTex). For protein detection, horseradish peroxidase-conjugated secondary antibodies and chemiluminescence (34577, ThermoFisher) were used. Protein levels were normalized to GAPDH for each sample and expressed as fold change compared with the control group. Image acquisition was performed using Image Lab 2.0 software (Biorad) and protein expression was analyzed by densitometric quantification using Image J 1.52p software. Uncropped images of immunoblots accompanied by the location of molecular weight markers are shown in Fig. S3.

Statistical analysis

Results are expressed as median and range unless stated otherwise. Two groups were compared using Mann-Whitney-Wilcoxon rank sum test or T-test as appropriate. Three or more groups were compared using Kruskal Wallis test or ANOVA as appropiate followed by the tukey method as a post-hoc test. Statistical tests were two sided. For survival analysis, Kaplan–Meier curves were generated and compared using the log-rank test. Patients were censored at the time of last follow-up. The date of liver transplantation was considered as a time point of death. Patients were stratified into high- and low–C. albicans cytotoxicity groups using maximally selected rank statistics. Statistical analyses were performed using R version 4.4.2 (2025).

Results

C. albicans strains isolated from patients with alcohol-associated hepatitis show different cytotoxicity which predicts survival

To examine the functional relationship between C. albicans and disease outcome in patients with alcohol-associated hepatitis, we plated stool samples of 105 patients with alcohol-associated hepatitis (Table 1), obtained from two multicenter national and international cohorts, on YPD plates. A maximum of ten colonies per patient were randomly selected for single colony qPCR to confirm C. albicans identity and exclude other Candida species. CFUs per mg of feces and positivity rates were not significantly different between patients who received antibiotics and those who did not (mean: 7.05 CFU/mg feces vs. 7.6 CFU/mg feces, p = 0.304; 70% vs. 82.3%, p = 0.315).

Of the 105 patients with alcohol-associated hepatitis, 56% (n = 59) were positive for fungal species, and in 29.5% (n = 31) C. albicans was detected in fecal samples, consistent with our previous findings.³ Fungi-positive patients who were negative for C. albicans showed growth of C. glabrata, C. krusei, C. tropicalis, and C. auris (Fig. S1).

From each C. albicans-positive patient, we selected up to three randomly selected strains for further experiments. To screen for cytotoxicity of each strain, we co-incubated intestinal epithelial Caco-2 cells with individual C. albicans strains for 24 h and measured LDH release as a marker of epithelial cell death (Fig. 1A).

Fig. 1.

Candida albicans cytotoxicity predicts survival in patients with alcohol-associated hepatitis.

(A) Workflow, created in BioRender. (B) Feces of 105 patients was tested for C. albicans. A maximum of three C. albicans strains were selected and tested for LDH release on Caco-2 cells. They were compared with a WT C. albicans lab strain (a Ura3-positive derivate of strain BWP17) and an ece1Δ/Δ C. albicans strain. Results are presented as mean and standard deviation. (C) To determine the cytotoxic effect of C. albicans per patient we calculated the mean of all three tested strains. Then we performed qPCR to quantify the amount of C. albicans in the feces and normalized the cytotoxicity for the amount of C. albicans present. An optimal threshold was calculated to classify patients into high- and low-cytotoxicity groups using max-rank statistics. Patients with high cytotoxicity exhibited significantly lower 30-day survival compared with patients with low cytotoxicity (log-rank, p = 0.032). (D) MELD scores did not differ between high- and low-cytotoxicity groups. p values were calculated using the Wilcoxon test. C. albicans, Candida albicans; MELD, model for end-stage liver disease; qPCR, quantitative PCR; WT, wild-type.

The wild-type C. albicans strain (a Ura3-positive derivate of strain BWP17)¹⁴ (Fig. 1B, first column), used as a positive control, induced 30% LDH release, whereas a candidalysin-knockout (ece1Δ/Δ) mutant (derived from BWP17) (Fig. 1B, second column), serving as a negative control, showed no LDH release. While C. albicans strains from individual patients exhibited consistent cytotoxicity within a patient, which might indicate clonal colonization, notable variations were observed between patients (Fig. 1B).

We sought to determine if different cytotoxicity levels were associated with differences in clinical outcome. Fecal C. albicans abundance was quantified by qPCR, and total cytotoxicity per patient was calculated by multiplying mean cytotoxicity with C. albicans load. We classified patients into high and low cytotoxic effect groups based on an optimal cut-off value. Patients with high C. albicans cytotoxicity showed significantly lower 30-day survival compared to patients with low normalized cytotoxicity (log-rank p = 0.032) (Fig. 1C). No significant difference was observed in the MELD score between the two groups (Fig. 1D), nor in markers of liver injury and chronic liver disease, including AST, ALT, international normalized ratio and bilirubin (not shown). There was no significant difference in the cytotoxic effect of C. albicans between patients who received antibiotics and those who did not (p = 0.12). There were also no differences in steroid use or the incidence of infections.

In summary, our data indicate that patient outcomes are influenced by the potential of C. albicans strains to cause host cell damage. Hyphal morphogenesis in C. albicans is central to the production of virulence factors,¹⁴ including candidalysin and thus seems to be crucial for functional toxicity.

Peptide YY reduces cytotoxicity of C. albicans in vitro

Previous work by Pierre et al. demonstrated that PC-PYY can suppress hyphal formation in C. albicans.²² Therefore, in the following experiments, we aimed to investigate the effects of PC-PYY on C. albicans in the context of alcohol-associated liver disease.

First, we investigated the potential protective effect of PC-PYY against intestinal epithelial cell cytotoxicity induced by C. albicans. For this purpose, C. albicans was co-incubated with Caco-2 cells and either the PC-PYY or a scrambled peptide of PYY for 24 h. Cytotoxicity was assessed by measuring LDH release. The addition of PC-PYY significantly reduced the cytotoxicity of C. albicans compared with the scrambled peptide or the strain alone (Fig. 2A). Increasing the concentration of PC-PYY resulted in a dose-dependent reduction in cytotoxicity (Fig. 2B), which is consistent with previous findings.²² We found the same effect in Mode-K intestinal cells (Fig. S2).

Fig. 2.

Paneth cell–derived PYY reduces C. albicans cytotoxicity.

(A) Caco-2 cells were incubated with WT C. albicans (Ura3-positive derivative of strain BWP17) with or without 20 μM SP or PC-PYY for 24 h. LDH release was measured in the supernatant and expressed as percent reduction relative to C. albicans WT alone. p values were calculated using the Wilcoxon test. 6-9 independent experiments in triplicates each were performed; each dot represents one independent experiment. (B) Caco-2 cells were incubated with WT C. albicans and increasing concentrations of SP or PC-PYY for 24 h. LDH release is shown as percent reduction relative to 0 μM SP or PC-PYY. Twp to three independent experiments in triplicates each were performed; each dot represents one independent experiment. (C) Primary mouse hepatocytes were incubated for 3 h with C. albicans WT, an ECE1-deleted strain (ece1Δ/Δ), an ECE1 mutant lacking the candidalysin-encoding region (ece1Δ/Δ + ECE1Δ184–279), or an ECE1 reintegrant strain (ece1Δ/Δ + ECE1). LDH release was measured and normalized to WT-induced LDH release. p values were calculated using the Wilcoxon test. Six to seven independent experiments in duplicates each were performed; each dot represents one independent experiment. (D) Primary mouse hepatocytes were incubated for 3 h with WT C. albicans in the presence or absence of 20 μM SP or PC-PYY and a protease inhibitor. LDH release is expressed as percent reduction relative to C. albicans WT alone. 7-8 independent experiments in duplicates each were performed. (E) C. albicans was incubated overnight in YPD medium with 20 μM PC-PYY or SP, and 20 μl of culture was plated onto microscope slides for morphological analysis. C. albicans, Candida albicans; PC-PYY, Paneth cell derived peptide YY; SP, scrambled peptide; WT, wild type; YPD, yeast extract–peptone–dextrose.

We have previously demonstrated that synthetic candidalysin induces hepatocyte death.³ However, efficient delivery of candidalysin to the host cells is crucial to fully enhance its cytotoxic effects.¹⁹ In this study, we aimed to evaluate functional cytotoxicity of candidalysin secreted by C. albicans, as well as the potential protective role of PC-PYY. To achieve this, we isolated primary mouse hepatocytes and incubated them with different C. albicans strains. We normalized the cytotoxic effect on hepatocytes to the cytotoxic effect of wild-type C. albicans. Incubation with the wild-type strain resulted in cytotoxicity, while incubation with the mutant lacking the ECE1 gene (ece1Δ/Δ) reduced the cytotoxic effect. Reintroducing a full-length allele into the ece1Δ/Δ strain (ece1Δ/Δ + ECE1) restored cytotoxicity, whereas reintroducing the mutant allele lacking only the candidalysin encoding region of ECE1 (ece1Δ/Δ + ECE1_Δ184–279) had no cytotoxic effect on hepatocytes (Fig. 2C), indicating the crucial role of candidalysin for mediating the cytotoxic effect.

Next, we exposed hepatocytes to the C. albicans wild-type strain in the presence of either PC-PYY or a scrambled peptide. A protease inhibitor was included to prevent cleavage of PC-PYY by DPP-IV, a serine protease expressed by hepatocytes. The addition of PC-PYY, along with the protease inhibitor, resulted in a significant reduction in LDH release from hepatocytes compared with the scrambled peptide (Fig. 2D). To confirm that the reduction in hyphal formation was the primary mechanism underlying the decrease in cytotoxicity, we microscopically examined representative C. albicans samples incubated with either PC-PYY or the scrambled peptide. C. albicans treated with PC-PYY exhibited reduced levels of hyphal morphology (Fig. 2E).

In summary, our data highlight the cytotoxic effect of C. albicans on hepatocytes and the importance of efficient delivery mechanisms to maximize cytotoxicity. Moreover, we demonstrate the protective role of PC-PYY in mitigating the cytotoxic effects of C. albicans in hepatocytes.

Peptide YY is upregulated in mice fed with ethanol diet

In humans and mice, different antimicrobial peptides are present in the mucus layer of the intestine to defend against bacterial translocation. Some of these peptides are decreased in patients with alcohol-associated liver disease and mice fed an ethanol diet.^34,35 To investigate the impact of ethanol on the expression of Pyy, we subjected mice to the chronic plus binge feeding model (NIAAA model) with an ethanol or an isocaloric control diet. Ethanol consumption leads to fungal overgrowth in mice, so we included a group of ethanol-fed mice treated with the non-absorbable antifungal nystatin to determine whether observed effects were driven by ethanol or by fungal overgrowth.³¹ Nystatin reduced fecal fungi and ethanol-induced liver disease including steatosis in mice fed an ethanol diet.³¹ After harvesting, intestinal crypts containing Paneth cells and intestinal villi containing enteroendocrine cells were isolated to assess the expression of PC-Pyy with antimicrobial properties and Pyy from enteroendocrine cells acting as a satiety hormone, respectively. Expression of Pyy and chromogranin A (Chga), a marker for enteroendocrine cells, was measured using qPCR.

Pyy levels were significantly increased in intestinal crypts of mice fed an ethanol diet compared with isocaloric controls and were significantly reduced when fungal growth was inhibited with nystatin (Fig. 3A). To distinguish between the two forms, Pyy expression was normalized to Chga (Fig. 3B). The Pyy/Chga ratio was significantly higher in ethanol-fed mice compared with both isocaloric controls and the ethanol plus nystatin group (Fig. 3C). No differences in Chga expression, Pyy levels, or Pyy/Chga ratio were observed in the villi among the three groups (Fig. 3D–F). Nystatin treatment also reduced liver injury, as measured by ALT levels, and steatosis, as determined by hepatic triglyceride content, in ethanol-fed mice compared with mice fed ethanol alone (Fig. 3G,H). Body weight did not differ between groups (Fig. 3I). These results indicate that Paneth cell-derived Pyy is upregulated during chronic ethanol feeding, potentially as a response to fungal overgrowth.

Fig. 3.

Gene expression of peptide YY is increased in crypts of ethanol-fed mice.

Mice were fed an ethanol diet, ethanol diet and nystatin, or isocaloric diet following the chronic plus binge feeding model (NIAAA) for 16 days. Nystatin was added for the last 10 days. Experiments were conducted in three independent experiments with 11 mice in the isocaloric group, 9 in the ethanol group and 10 in the ethanol + nystatin group. p values were calculated using the Kruskal-Wallis or Wilcoxon tests. Each dot represents one mouse. (A-F) intestinal level of Pyy and Chga, in crypts (A-C) and villi (D-F). (G) Serum levels of ALT. (H) Hepatic triglyceride content. (I) Body weight. ALT, alanine aminotransferase.

To assess protein levels, we stained intestinal sections from mice fed either ethanol or control diet for lysozyme to detect Paneth cells, along with PC-Pyy protein expression. PC-Pyy expression was significantly increased in Paneth cells of ethanol-fed mice (Fig. 4A,B), further indicating that antimicrobial PC-Pyy is increased in crypts of ethanol-fed mice.

Fig. 4.

Protein expression of peptide YY is increased in Paneth cells of ethanol-fed mice.

Mice were fed an ethanol diet or isocaloric diet following the chronic plus binge feeding model (NIAAA) for 16 days. Experiments were conducted in two independent experiments with 8 mice in the isocaloric and 7 in the ethanol group. p values were calculated using the Wilcoxon test. (A) The number of red pixels in Lyz-positive cells was quantified using Qupath. (B) Representative staining of intestinal sections. Red = Pyy, green = Lyz, blue = DAPI. Lyz, lysozyme.

Administration of peptide YY reduces ethanol-induced liver disease in mice fed chronic plus binge ethanol diet

Mice fed a chronic ethanol diet are more likely to develop fungal overgrowth in the intestine. To further investigate the effect of PC-PYY on fungal overgrowth in ethanol-induced liver disease, we subjected mice to a chronic plus binge ethanol diet for 8 weeks and administered either synthetic PC-PYY or scrambled peptide via oral gavage during the last 10 days.

Mice treated with PC-PYY developed significantly less severe liver injury and reduced hepatic steatosis compared with mice receiving the scrambled peptide (Fig. 5A-C). Although not statistically significant, PC-PYY–treated animals showed a trend toward reduced liver inflammation, as indicated by lower expression of chemokine (C-X-C motif) ligand 1 and 2 (Fig. 5D,E). Body weight and food intake remained comparable between groups (Fig. 5, Fig. 6A). Ethanol is metabolized in the liver primarily by alcohol dehydrogenase 1 and cytochrome P450 family 2 subfamily E polypeptide 1; the expression of both enzymes was similar between groups (Fig. 6B,C). Both ethanol-fed groups exhibited fungal growth (Fig. 6D). In ethanol-fed mice, administration of PC-PYY significantly reduced hyphal formation compared with mice receiving the scrambled peptide (Fig. 6E,F). Together, these findings suggest that PC-PYY mitigates ethanol-induced liver injury by inhibiting C. albicans hyphae formation.

Fig. 5.

Administration of peptide YY ameliorates ethanol-induced liver disease in mice.

Mice were fed an ethanol diet or isocaloric diet following the chronic plus binge feeding model for 8 weeks and gavaged daily with 100 μl of 2 μg/μl PC-PYY or SP during the last 10 days. We conducted three independent experiments with 20 mice in the Ethanol + SP group, 19 mice in the Ethanol + PC-PYY group and 5 mice in each of the isocaloric groups. Each dot represents one mouse. p values were calculated using the Kruskal Wallis and Wilcoxon tests. (A) Serum levels of ALT. (B) Quantification of red O staining with ImageJ. (C) Representative slides of H&E staining and Oil red O staining from mouse liver. (D,E) Hepatic level of Cxcl1 and Cxcl2 mRNAs. (F) Food intake. ALT, alanine aminotransferase; PC-PYY, Paneth cell derived peptide YY; SP, scrambled peptide.

Fig. 6.

Administration of peptide YY does not change body weight, hepatic ethanol metabolism or fungal overgrowth in ethanol-fed mice.

Mice were fed an ethanol diet or isocaloric diet following the chronic plus binge model for 8 weeks and gavaged daily with 100 μl of 2 μg/μl PC-PYY or SP during the last 10 days. We conducted three independent experiments with 20 mice in the Ethanol + SP group, 19 mice in the Ethanol + PC-PYY group and 5 mice in each of the isocaloric groups. Each dot represents one mouse. p values were calculated using Kruskal Wallis and Wilcoxon tests. (A) Body weight. (B,C) Hepatic level of Adh1 and Cyp2e1 mRNAs. (D) Number of CFUs per g feces. (E,F) Hyphae and yeast form in feces from mice gavaged with PC-PYY or SP. Feces was collected at the time of harvesting and stored at -80 °C. After thawing it was cultured in YPD medium with 1% FBS at 37 °C, with shaking overnight. (E) Pictures were taken with brightfield microscopy and analyzed with Qupath. (F) Representative images. CFUs, colony-forming units; PC-PYY, Paneth cell–derived peptide YY; SP, scrambled peptide; YPD, yeast extract–peptone–dextrose.

PC-PYY reduces C. albicans-induced pyroptosis

Pyroptosis has been identified as an important driver of alcohol-associated liver disease.³⁶ C. albicans hyphae are potent inducers of pyroptosis. Candidalysin and cell wall components of C. albicans activate NLRP3 which leads to cleavage of caspase-1 and subsequently induction of pyroptosis via activation of gasdermin D (GSDMD) and IL1b.[37], [38], [39], [40], [41] We cultured primary mouse hepatocytes with C. albicans and PC-PYY or C. albicans and scrambled peptide, and then assessed pyroptosis by quantifying GSDMD by immunoblot. C. albicans increased cleaved GSDMD and the addition of PC-PYY significantly reduced GSDMD cleavage compared with SP (Figs 7A-D and S4A). We confirmed these results in liver tissue of ethanol-fed mice gavaged with PC-PYY or scrambled peptide (Fig. S3A-D and S4C). Additionally, liver tissue from ethanol-fed mice gavaged with PC-PYY showed lower levels of cleaved IL1b compared with mice fed ethanol diet and SP (Figs 7E-F and S4B). These results indicate that PC-PYY reduces pyroptosis induced by C. albicans.

Fig. 7.

Addition of PC-PYY leads to reduction of C. albicans-induced pyroptosis in primary mouse hepatocytes.

(A-D) Primary mouse hepatocytes were incubated with C. albicans wild-type (a Ura3-positive derivate of strain BWP17) and eventually 20 μM SP or PC-PYY and protease inhibitor for 3 h. Proteins were quantified from hepatocytes by immunoblot. p values were calculated using ANOVA followed by post-hoc analysis with the tukey-test for pairwise comparison. We performed three biological replicates. (D) Representative immunoblot. (E,F) Mice were fed an ethanol diet or isocaloric diet following the chronic plus binge model for 8 weeks and gavaged daily with 100 μl of 2 μg/μl PC-PYY or SP during the last 10 days. Each dot represents one mouse. p values were calculated using Kruskal Wallis and Wilcoxon tests. (E) Hepatic protein level of cleaved IL1b. (F) Representative immunoblot. C. albicans, Candida albicans; PC-PYY, Paneth cell–derived peptide YY; PYY, peptide YY; SP, scrambled peptide.

Discussion

In this study we show that C. albicans strains isolated from the feces of patients with alcohol-associated hepatitis have varying degrees of cytotoxicity towards intestinal epithelial cells, and higher cytotoxicity correlates with poorer survival outcomes. Recent studies have demonstrated the antifungal effect of PC-PYY, which specifically inhibits the hyphal, but not yeast morphology of C. albicans.²² We found that PC-PYY reduces C. albicans–induced cytotoxicity in hepatocytes. Moreover, supplementation of ethanol-fed mice with PC-PYY significantly attenuated liver injury.

We have previously shown that the abundance of C. albicans is increased in patients with alcohol use disorder and that abundance of fecal genomic DNA encoding ECE1 is associated with more severe liver disease in patients with alcohol-associated hepatitis.^3,42 In our current study, we advanced this knowledge and functionally assessed host cell damage induced by C. albicans.

Cytotoxicity is largely due to secreted candidalysin, but functional delivery of candidalysin to the invasion pocket is essential for the observed cytotoxic effect. In inflammatory bowel disease, C. albicans strains isolated from one patient showed different cytotoxicity towards macrophages, and the cell-damaging ability of each C. albicans strain is linked to its filamentous ability.⁴³ In line with the study in inflammatory bowel disease, we show that functional cytotoxicity is different between strains and not solely dependent on the presence of the genomic ECE1 gene encoding candidalysin. Varying levels of cytotoxicity are likely due to differences in the efficiency of candidalysin delivery to target cells, which requires a combination of virulence attributes.¹⁹

Our results show that PC-PYY reduces cytotoxicity of C. albicans in vitro and in vivo. Since PC-PYY inhibits hyphal growth, it not only suppresses candidalysin expression and secretion but also limits hyphal-associated virulence attributes such as adhesion and invasion via the hyphal-associated protein Als3, which is essential for the effective delivery of candidalysin to host cells within an invasion pocket.^17,19,44,45 We propose that antimicrobial PC-PYY functions to maintain C. albicans in its yeast form, thereby preventing hyphal growth, adhesion to host cells, host cell invasion, candidalysin secretion and thus host cell damage.

PC-PYY is released in response to the presence of fungal hyphae.²² Initially, we hypothesized that PC-PYY levels would be reduced in ethanol-induced liver disease, like other antimicrobial peptides.^34,46,47 However, our findings revealed the opposite: PC-PYY levels were increased. Ethanol administration in mice promotes fungal overgrowth, which is linked to disease severity, independent of ethanol consumption, as this effect is reversed by nystatin administration.^3,31 We propose that this increase in disease severity is associated with increased growth of fungi in their hyphal form and thus enhanced secretion of candidalysin. Our study demonstrates that the host responds to this fungal transition by upregulating PC-PYY. In the early stages of infection, sublytic concentrations of candidalysin initiate a danger response by activating the p-MKP1/cFos pathway in host cells,¹⁴ thus signaling an increased level of toxin-producing hyphae.¹⁴ This mechanism may represent a coevolutionary adaptation, in which the host has developed a sensitive toxin detection system to recognize a critical threshold of hyphae (in contact with epithelial cells) and counteract fungal invasion.⁴⁸ We propose that PC-PYY secretion may be an integral component of the host’s danger response to pathogenic C. albicans. Importantly, enteroendocrine PYY levels are not affected by alcohol consumption in humans,⁴⁹ nor were they altered in our mouse model.

In summary, our study establishes a link between the cytotoxic effects of individual C. albicans strains against epithelial cells and survival in patients with alcohol-associated hepatitis, which is in line with data that high-toxicity C. albicans strains are associated with more severe inflammatory bowel disease.⁴³ Additionally, we suggest that PC-PYY plays a protective role in ethanol-induced liver disease by limiting candidalysin-producing hyphae.

Abbreviations

ALT, alanine aminotransferase; AST, aspartate aminotransferase; CFU, colony forming unit; Chga, chromogranin A; C. albicans, Candida albicans; DPP-IV, dipeptidyl peptidase IV; GSDMD, gasdermin D; MELD, model for end-stage liver disease; PC-PYY, Paneth cell–derived peptide YY; PYY, peptide YY; qPCR, quantitative PCR; YPD, yeast extract–peptone–dextrose.

Authors’ contributions

H.K. was responsible for performing experiments, data analysis, and writing the manuscript; M.F.F., A.H., F.R.T, E.M., D.S., P.H. and C.L. provided assistance performing experiments. B.H. and S.L. provided experimental resources and edited the manuscript; P.S., and AlcHepNet investigators were responsible for collection of fecal samples; B.S. was responsible for study concept and design, interpretation of data, editing the manuscript, and study supervision.

Data availability

Data will be made available on request.

Declaration of generative AI and AI-assisted technologies in the writing process

During the preparation of this work the authors used ChatGPT in order to edit the manuscript. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Financial support

H.K. is supported by the Walter-Benjamin-Fellowship of the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) (KR 5843 1-1). D.S. is supported by the Walter-Benjamin-Fellowship of the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) (SCHO 1910/4-1). B.H. is supported by the DFG within the Cluster of Excellence ‘Balance of the Microverse’, under Germany’s Excellence Strategy, EXC 2051, project ID 390713860 and within the Collaborative Research Centre/Transregio 124 “FungiNet” project C1 and C2 (DFG project number 210879364). P.H. is supported by National Institutes of Health (NIH) grant K12 HD105271, University of California San Diego Altman Clinical and Translational Research Institute (ACTRI)/NIH grant KL2TR001444, and Pinnacle Research Award in Liver Diseases Grant #PNC22-159963 from the American Association for the Study of Liver Diseases Foundation. C.L. is supported by NIH grants R01 AA029106, R21 AA030654, P30 AR073761 the D34 HP31027 UC San Diego’s Hispanic Center of Excellence. This study was supported in part by NIH grants R01 AA24726, R37 AA020703, by Award Number BX004594 from the Biomedical Laboratory Research & Development Service of the VA Office of Research and Development, and by an ALF Pilot Research Award from the American Liver Foundation (to B.S.) and services provided by NIH centers P30 DK120515 and P50 AA011999.

Conflicts of interest

P.H.’s institution UCSD has received grant support from Nterica Bio. B.H. is consulting for CureVac, Tübingen, Germany. B.S. has been consulting for Ambys Medicine, Boehringer Ingelheim Pharma, Surrozen and Takeda (prior 24 months). B.S.’s institution UC San Diego has received research support from Axial Biotherapeutics, ChromoLogic, CymaBay Therapeutics, Intercept Pharmaceuticals and Prodigy Biotech (prior 24 months). B.S. is founder of Nterica Bio. UC San Diego has filed several patents with B.S. as inventor related to this work.

Please refer to the accompanying ICMJE disclosure forms for further details.

Supplementary data

The following are the Supplementary data to this article: