Impact of BECLIN1 haploinsufficiency on goblet cell function and susceptibility to colitis Authors: Juliani Juliani, Sharon Tran, Tiffany J Harris, Sarah L Ellis, Aysha H Al-Ani, Yashodha Wijebandara, Komal M Patel, Samuel N Young, Marco Evangelista, David Baloyan, Camilla M Reehorst, Rebecca Nightingale, Laura J Jenkins, Peter De Cruz, Kinga Duszyc, Benjamin T Kile, Alpha S Yap, John M Mariadason, Britt Christensen, André L Samson, James M Murphy, Walter D Fairlie, Erinna F Lee Published: 2026 ABSTRACT AbstractBECLIN1 is a central regulator of autophagy and endocytic trafficking essential for epithelial homoeostasis. While complete intestinal epithelial loss of BECLIN1 causes fatal enteritis originating in the small intestine, the consequences of its partial loss in the gut remain unclear. Given that BECLIN1 expression can vary in human disease, we investigated whether reduced BECLIN1 is sufficient to impair gut barrier function. Heterozygous Becn1 deletion (Becn1IEC+/−) in the mouse intestinal epithelium caused subtle but significant defects. These included shortened small intestines and altered epithelial architecture, despite preservation of basal autophagy, implicating trafficking-related functions. Supporting this conclusion, Becn1IEC+/− small intestinal epithelial cells showed modest increases in RAB5+ve vesicles, redistribution of E-CADHERIN and F-actin along lateral membranes and altered apico-basal cell morphology. Given the absence of overt small intestinal epithelial disruption or inflammation, as seen with complete loss of BECLIN1, we next addressed whether BECLIN1 insufficiency manifests a phenotype under stress or in other gut regions. Indeed, in the colon, Becn1IEC+/− mice exhibited reduced colonic crypt length, baseline goblet cell loss and reduced mucin production, particularly in mature goblet cells, indicating vulnerability of the mucus barrier. When challenged with dextran sulfate sodium (DSS), Becn1IEC+/− mice exhibited greater weight loss, higher disease activity, more severe histological colitis, and disproportionate loss of neutral mucins, with inflammation confined to the mucosa. Together, these findings show that BECLIN1 insufficiency does not trigger spontaneous inflammation but destabilises epithelial organisation and barrier defence, thereby sensitising the gut to inflammatory challenge and further positioning BECLIN1 as a threshold-dependent determinant of intestinal resilience.Subject terms: Mechanisms of disease, Crohn's disease Introduction Genome-wide association studies have identified polymorphisms in autophagy regulators, such as ATG16L1 and IRGM, as strongly associated with inflammatory bowel disease (IBD) [1–7]. These findings highlight the importance of autophagy in intestinal homoeostasis and implicate its disruption in IBD pathogenesis. Functional studies in mouse and cell-based models have revealed both epithelial-intrinsic roles, including maintenance of barrier integrity and cell survival [2, 8–16], and immune-mediated extrinsic roles, such as regulating inflammatory responses and microbial balance [17–21]. Recent studies have shown that the prototypical autophagy regulator BECLIN1 is critical for intestinal homoeostasis. Constitutive activation of BECLIN1-dependent autophagy alleviates goblet cell ER stress, thickening the mucus barrier and protecting against chemical- and infection-driven inflammation [11]. Beyond autophagy, BECLIN1 regulates endocytic trafficking, maintaining barrier integrity via correct junctional protein localisation, cytoskeletal organisation, and epithelial remodelling [22–24]. Complete intestinal epithelial deletion of BECLIN1 is rapidly fatal due to severe enterocolitis with extensive epithelial cell loss, impaired specialised IEC functions, inflammation, and barrier breakdown [23]. BECLIN1 undergoes extensive post-translational modifications that fine-tune its activity and interactions [25–28]. In diseases, however, it is reduced BECLIN1 expression rather than mutation that contributes to pathogenesis [29–31], with changes arising from altered protein interactions, cleavage, epigenetic regulation, or monoallelic deletion [32–35]. Despite this, the consequences of partial BECLIN1 loss in the intestinal epithelium remain poorly understood. Here, we demonstrate that partial loss of BECLIN1 impairs goblet cell function and increases susceptibility to inflammation in the colon. By dissecting its roles in endocytic trafficking, cytoskeletal organisation and mucin secretion, we provide new insights into how BECLIN1 insufficiency destabilises epithelial integrity and compromises barrier defence. Materials and methods Animal experiments All mouse strains used in this study were bred on the C57BL/6 J background. Becn1tm1b(KOMP)Wtsi mice were purchased from the European Conditional Mouse Mutagenesis Program (EUCOMM). Becn1fl/fl mice were generated by breeding Becn1tm1b(KOMP)Wtsi mice onto CAG-FLPe mice. Becn1+/+;, Becn1fl/+;, and Becn1fl/fl;Vil1-CreERT2Cre/+ mice were then subsequently generated by breeding Becn1+/+, Becn1fl/+ and Becn1fl/fl mice to the Vil1-CreERT2 mice [36]. Mice were housed at the La Trobe Animal Research and Teaching Facility (LARTF, La Trobe University, VIC, Australia) under Specific Pathogen Free (SPF) conditions. All experiments performed were approved by the La Trobe University animal ethics committees (approvals AEC18024, AEC18036) in accordance with the Australian code for the care and use of animals for scientific purposes. All research with these mice has complied with all relevant ethical regulations for animal use. To induce gene deletion, male and female mice aged between 6 and 17 weeks (unless otherwise specified), were selected indiscriminately and intraperitoneally injected with 4 mg tamoxifen (T5648, Sigma-Aldrich, St. Louis, Missouri, USA) in sunflower seed oil (25007, Sigma-Aldrich, St. Louis, Missouri, USA), delivered as one 200 μl injection per day of a 10 mg/ml stock, over two consecutive days. The study was not powered for sex-specific analyses. Mice were humanely euthanised by CO2 asphyxiation. Animals were allocated to experimental groups based on genotype and no formal randomisation was applied. Mice were otherwise selected indiscriminately with respect to sex, weight, and cage origin, and all outcome analyses were performed blinded to genotype to minimise bias. Dextran Sulphate Sodium (DSS) treatment Following Tamoxifen treatment to induce gene deletion, 7–14 weeks old mice were administered DSS (2% (w/v), MP Biomedicals, Irvine, California, USA) in autoclaved drinking water provided ad libitum for 5 days. Control groups received autoclaved drinking water without DSS for the same duration. During the experimental period, mice were monitored daily and weighed. The average daily intake of 2% (w/v) DSS water was calculated by measuring the weight difference of drinking bottles at the start and end of the experiment and dividing by the number of treatment days. The disease activity index (DAI) was assessed using a cumulative scoring system based on parameters outlined in Table 1, including weight loss, stool consistency and the presence of hematochezia (presence of blood in stool). At the experimental endpoint, mice were humanely euthanised using CO2 asphyxiation. The small and large intestine were harvested, and their lengths were measured prior to rolling and fixation in 10% (v/v) neutral buffered formalin for 48 h. Table 1Scoring system for the evaluation of colitis severity in mice based on weight loss, stool consistency and haematochezia parameters.ScoreWeight loss (%)Stool consistencyHaematochezia0NormalNormalNo blood1<5-Visible blood in rectum26–10Pasty, semiformedVisible blood on fur311–20--4>20Liquid, sticky or unable to defecate after 5 min- Scoring system for the evaluation of colitis severity in mice based on weight loss, stool consistency and haematochezia parameters. Intestinal organoid culture Organoids were established by culturing crypt-enriched fractions from the duodenum of 8–12 weeks old untreated mice as described previously [23]. To induce Becn1 deletion, organoids were seeded into media containing 200 nM 4-hydroxytamoxifen (4-HT) (H7904, Sigma-Aldrich, St. Louis, Missouri, USA) for three days and then maintained as per normal. For organoid re-passaging experiments, 4-HT-treated organoids were passaged at day 7 and maintained as above. Intestinal epithelial cell isolation The mouse small intestine was opened longitudinally, rinsed with Dulbecco’s Phosphate Buffered Saline (dPBS), and incubated in dPBS + 15 mM EDTA for 15 min at 37 °C with agitation. Dissociation of IECs was achieved by vortexing. The cell suspension was then washed once with ice-cold dPBS, and the cell pellet snap-frozen on dry ice and stored at −80 °C until use. Histology, immunohistochemistry and organoid whole-mount immunofluorescence Preparation of intestinal “Swiss-rolls”, haematoxylin and eosin (H&E), Periodic acid-Schiff, Alcian blue (PAS-AB) and Ki-67 staining were all performed as described previously [23]. All slides were scanned using the Aperio AT2 (Leica, Wetzlar, Germany) and images captured and analysed using Aperio ImageScope v12.4.2.5010 and Fiji ImageJ. Whole-mount staining of intestinal organoids was performed as previously described [22]. Detailed image acquisition and quantification procedures are described in the Supplementary information. Immunoblotting of mouse tissues Preparation of mouse cell lysates and immunoblotting was performed as described previously [23]. Primary antibodies were: Beclin1, 1:500 (3495, Cell Signalling Technology, Danvers, Massachusetts, USA); p62/SQSTM1, 1:500 (5114, Cell Signalling Technology, Danvers, Massachusetts, USA); LC3B, 1:500 (NB100-2220, Novus Biologicals, Centennial Colorado, USA); β-Actin, 1:5000 (A2228, Sigma-Aldrich, St. Louis, Missouri, USA); GAPDH, 1:5000 (MA5-15738, Invitrogen, Waltham, Massachusetts, USA). Secondary antibodies used were: Donkey anti-Rabbit IgG, 1:10,000 (NA943V, GE Healthcare, Chicago, Illinois, USA); Goat anti-Mouse IgG, 1:10,000 (A0168, Sigma-Aldrich, St. Louis, Missouri, USA). Results Generation of mice with inducible monoallelic deletion of Becn1 in the intestinal epithelium We previously reported that homozygous Becn1 deletion in the intestinal epithelium results in rapid, fatal enterocolitis, with severe epithelial disruption [23]. To model the haploinsufficiency seen in human disease where BECLIN1 is reduced, we generated mice in which monoallelic Becn1 deletion could be induced in the intestinal epithelium using mice heterozygous for the Becn1 floxed allele (Becn1fl/+) and carrying the Vil1-CreERT2 transgene (Becn1fl/+;Vil1-CreERT2Cre/+). Littermates lacking (Becn1+/+) or homozygous (Becn1fl/fl) for the floxed allele but positive for Vil1-CreERT2Cre/+ (Becn1+/+;Vil1-CreERT2Cre/+ or Becn1fl/fl;Vil1-CreERT2Cre/+ respectively) served as controls. Intraperitoneal Tamoxifen injections induced deletion in the small intestine (duodenum, jejunum, ileum) and colon of Becn1fl/+;Vil1-CreERT2Cre/+ mice, confirmed genomically and by protein analysis showing approximately 50% reduced BECLIN1 levels (Fig. 1A, Supplementary Fig. 1A, B). Tamoxifen-treated Becn1+/+;Vil1-CreERT2Cre/+, Becn1fl/+;Vil1-CreERT2Cre/+, and Becn1fl/fl;Vil1-CreERT2Cre/+ mice are hereafter respectively referred to as Becn1IEC+/+, Becn1IEC+/- and Becn1IEC-/- mice. Monoallelic Becn1 deletion induces mild intestinal epithelial alterations under basal conditions, distinct from the severe defects observed with homozygous Becn1 deletion. A Representative Western blots (top) showing BECLIN1 levels in the indicated tissues, with densitometry quantification (bottom). To enable comparison across Western blots, BECLIN1 band volumes were quantified and expressed as BECLIN1:GAPDH ratios, then normalised to the mean BECLIN1:GAPDH ratio of the Becn1IEC+/+ group (set to 1.0). Data represent n = 6 for Becn1IEC+/+ and Becn1IEC+/- mice and n = 3 for Becn1IEC-/- mice. All blots showing all samples used in the quantification are in Supplementary Fig. 1B; GAPDH was used as a loading control. B Representative images of intestinal tracts of Becn1IEC+/+, BecnIEC+/- and Becn1IEC+/+ mice (top) with the measurements of intestinal length at Day 7 post Tamoxifen administration (bottom). Data represent n = 12 to 17 animals per genotype from n > 3 independent experiments. C H&E-stained FFPE sections of the intestinal tracts from Becn1IEC+/+, Becn1IEC+/- and Becn1IEC-/- mice at Day 7 post Tamoxifen administration. Scale bars = 100 μm. Data represent n = 9 biologically independent mice of each genotype from n = 3 independent experiments. D Quantification of distal colon crypt length in Becn1IEC+/+ and Becn1IEC+/- mice at days 7, 14 and 35 post-tamoxifen administration. Only visible full-length crypts were included in the analysis, with n > 5 crypts measured per mouse and averaged. Data represent n = 4 to 7 animals per genotype from n = 3 independent experiments. E Representative Western blot (single biological sample per genotype) showing markers of basal autophagy across different sections of the gastrointestinal tract in Becn1IEC+/+ and Becn1IEC+/- mice at Day 7 post 4-HT treatment. The duodenum of Becn1IEC-/- mice was included as a positive control for defective autophagy. β-ACTIN was used as a loading control. Asterisks denote non-specific bands arising from carryover on the previous membrane or from antibody cross-reactivity. F Representative phase contrast microscopy images of Becn1IEC+/+, Becn1IEC+/- and Becn1IEC-/- organoids at Day 7 post-4-HT treatment and G re-passaged surviving Becn1IEC+/+, Becn1IEC+/- and Becn1IEC-/- organoids at Day 5, along with respective quantification of the number of budding crypt-like projections per organoid. Scale bar = 100 µm. H Representative Ki-67 immunostaining of FFPE sections from Becn1IEC+/+ and Becn1IEC+/- mice at Day 7 post-tamoxifen administration across different segments of the small and large intestine. Red arrowheads indicate intact proliferative activity within the intestinal stem cell compartments. Data represent n = 6 mice from n = 3 independent experiments. Scale bar = 100 μm. Accompanying graphs show quantification of Ki-67 staining expressed as the percentage of Ki-67+ve nuclei per crypt across representative regions of the small (ileum) and large (distal colon) intestine. For each mouse, 3–5 crypts were analysed and averaged, with 3 to 6 mice per genotype included. Data in (F, G) were obtained from n = 20 organoids per genotype across n = 3 independent experiments. Graphs show the mean ± S.E.M. Statistical significance was determined by unpaired (Student’s) t test in (A, D, H), and ordinary one-way ANOVA in (B, F, G). SI small intestine, Stom stomach, Duo duodenum, Jej jejunum, Ile ileum, Col Colon, MW molecular weight, 4-HT 4-hydroxytamoxifen, ns not significant (p > 0.05). Mice with reduced BECLIN1 have shortened small intestines and reduced colonic crypt lengths without other overt phenotypes At 7 days post-BECLIN1 deletion, there were no significant body weight differences between Becn1IEC+/+ and Becn1IEC+/- mice (Supplementary Fig. 2A), unlike the marked weight loss previously seen with homozygous Becn1 deletion in the intestinal epithelium [23]. Despite appearing grossly normal, Becn1IEC+/- mice displayed significantly shorter small intestines than Becn1IEC+/+ controls, while colon lengths were unchanged (Fig. 1B). Histological examination by H&E staining revealed preserved epithelial architecture in the small and large intestine of Becn1IEC+/- mice, with intact villi, organised crypt structure, and no evidence of epithelial damage or inflammation at 7 days post deletion (Fig. 1C). At this time point, distal colonic crypt length (Fig. 1D) and width (Supplementary Fig. 2B) did not differ between genotypes. However, by 14- and 35-days post deletion, Becn1IEC+/- mice exhibited significantly shorter colonic crypts and reduced crypt volume fraction [37] compared to controls (Fig. 1D). Observation over one month post deletion also revealed a persistently shorter small intestine (Supplementary Fig. 2C) without weight loss (Supplementary Fig. 2A), and otherwise normal intestinal morphology (Supplementary Fig. 2D). These phenotypes contrasted sharply with Becn1IEC-/- mice, where small intestines were shortened, swollen, and lytic, with villus stunting along the entire length within 7 days of gene deletion (Fig. 1B, C) [23]. In summary, apart from reduced small intestinal and colonic crypt lengths, heterozygous Becn1 deletion does not cause overt abnormalities under basal conditions up to one month post deletion, in contrast to its homozygous loss [23]. Intestinal epithelial cells with reduced BECLIN1 levels display no disruption to basal autophagy or proliferation Monoallelic Becn1 deletion caused no dramatic changes in basal autophagy in a representative mouse where BECLIN1 protein levels were reduced (Fig. 1A), with P62 levels and LC3B (total and LC3B-I:LC3BII) levels appearing comparable to wild-type IECs at 7 days post deletion (Fig. 1E, Supplementary Fig. 1C). In contrast, Becn1IEC-/- IECs showed pronounced increases in P62, total LC3B and LC3B-I:LC3BII (Fig. 1E, Supplementary Fig. 1C) [23]. We previously identified an essential role for BECLIN1 in endocytic trafficking in small intestinal-derived organoids [22, 23]. To test whether monoallelic Becn1 loss induced similar defects, we generated small intestinal organoids from Becn1+/+; Becn1fl/+; and Becn1fl/fl;Vil1-CreERT2Cre/+ mice. Treatment with 4-hydroxytamoxifen (4-HT) led to reduced BECLIN1 expression in Becn1IEC+/- organoids (Supplementary Fig. 3A) which did not impact basal autophagy (as with intestinal epithelial cells from mice, Supplementary Fig. 3B), unlike homozygous Becn1 deletion. At 7 days post 4-HT treatment, Becn1IEC-/- organoids showed significantly reduced budding crypt formation (Fig. 1F) [23], whereas Becn1IEC+/- organoids formed crypt-like projections comparable to Becn1IEC+/+ organoids (Fig. 1F). While budding morphology alone cannot establish proliferation or differentiation status, this suggests that gross organoid growth capacity is preserved in the heterozygous setting. Re-passaging of Becn1IEC+/+ and Becn1IEC+/- cultures yielded mature organoids by Day 5, with similar numbers of budding crypts (Fig. 1G). Importantly, this is consistent with Ki-67 staining of intestinal tissues from Becn1IEC+/- mice, which showed no significant change in the percentage of Ki-67+ve nuclei per crypt compared to controls (Fig. 1H). In contrast, surviving Becn1IEC-/- organoids failed to generate viable organoid structures upon re-passaging, producing significantly fewer crypt-like projections by Day 5 (Fig. 1G). Together, these findings indicate that reduced BECLIN1 expression does not impair basal autophagy or epithelial cell survival, and proliferation remains largely intact in the heterozygous state. Intestinal epithelial cells with reduced BECLIN1 levels alters endocytic trafficking, E-CADHERIN distribution, and actin-dependent epithelial architecture We previously reported that homozygous BECLIN1 deletion in small intestinal epithelial cells (IECs) caused pronounced trafficking defects that compromise barrier integrity via mislocalisation of junctional proteins [22, 23, 38–42]. To determine whether this also occurs under heterozygous conditions, we examined Becn1IEC+/− small intestinal-derived organoids. Compared with Becn1IEC+/+, Becn1IEC+/− organoids displayed a modest but significant increase in cytoplasmic RAB5+ve early endosomes (Fig. 2A, B, Supplementary Fig. 4A) though without the aberrant enlargement seen in Becn1IEC-/- organoids (Fig. 2A, C). In the homozygous setting, the junctional protein E-CADHERIN is sequestered within the enlarged cytoplasmic RAB5+ve early endosomes (Fig. 2A, D, E) [23]. However, in Becn1IEC+/− IECs, E-CADHERIN levels were instead increased at apical and lateral membranes (Fig. 2F, G) and in the cytoplasm (Fig. 2D), but notably, with no change in RAB5 co-localisation (Fig. 2E, H, I). Thus, reduced BECLIN1 alters early endosomal dynamics, but E-CADHERIN localisation changes appear independent of RAB5-mediated trafficking. BECLIN1 reduction alters E-CADHERIN trafficking in a RAB5-independent manner. A Representative whole-mount immunofluorescence staining for RAB5 (red), E-CADHERIN (green) and DAPI (blue) in Becn1IEC+/+, Becn1IEC+/−and Becn1IEC−/− organoids. Increased cytoplasmic RAB5 is indicated by white arrows. Aberrant E-CADHERIN localisation is denoted by white (apical), red (lateral), and purple (cytoplasmic) arrowheads. Scale bar = 5 μm. B Quantification of cytoplasmic RAB5 pixels per cell. C Measurement of the average size of RAB5+ve vesicles. D Quantification of cytoplasmic E-CADHERIN pixels per cell, along with its (E) co-localisation with cytoplasmic RAB5, measured using Pearson’s correlation coefficient. Quantification of (F) apical and (G) lateral E-CADHERIN pixels per intestinal epithelial cell. Measurement of the degree of colocalisation between RAB5 and E-CADHERIN on the (H) apical and (I) lateral membrane of intestinal epithelial cells. Data are representative of at least n = 3 different slices per organoid and of at least n = 3 biologically independent organoids. For each z-section, at least n = 3 individual IECs with clear apical to basal delineation were used. Graphs indicate the mean ± S.E.M. Statistical significance was determined using ordinary one-way ANOVA. ns not significant (p > 0.05). Cytoskeletal remodelling can influence cadherin localisation [43]. Consistent with this, we previously showed that homozygous BECLIN1 loss disrupts the cytoskeleton and compromises the intestinal epithelial barrier [22]. Therefore, we next examined F-actin organisation, which is critical for adhesion and endocytosis [44, 45]. Becn1IEC+/− organoids exhibited increased F-actin, and co-localised E-CADHERIN, along the lateral membrane but not apical membrane (Fig. 3A–E), consistent with an F-actin-driven “cadherin flow” mechanism, whereby cadherins are transported along remodelling actin networks [46]. Notably, E-CADHERIN localisation was redistributed within the lateral membrane toward apicolateral junctions (Fig. 3E), potentially reinforcing cell-cell adhesion and the preservation of epithelial integrity. Given this increased lateral membrane localisation of both E-CADHERIN and F-actin in Becn1IEC+/- IECs (Figs. 2G, 3C, E), Becn1IEC+/− organoids had a longer apico-basal axis and shorter IEC width (Fig. 3A, F, G, Supplementary Fig. 4B), likely contributing to the shorter small intestinal length (Fig. 1B). Reduced BECLIN1 therefore alters early endosomal dynamics and F-actin-guided adhesion remodelling, leading to compensatory redistribution of junctional proteins and changes in epithelial architecture. Monoallelic Becn1 deletion alters F-actin organisation in intestinal epithelial cells. A Representative whole-mount immunofluorescence staining for F-actin (red), E-CADHERIN (green) and DAPI (blue) in Becn1IEC+/+, Becn1IEC+/−and Becn1IEC−/− organoids. Increased lateral F-actin in Becn1IEC+/− organoids is indicated by white arrowheads. Scale bar = 5 μm. Quantification of (B) apical and (C) lateral F-actin signals per cell. Colocalisation between F-actin and E-CADHERIN on the (D) apical and (E) lateral membranes of intestinal epithelial cells, assessed using Pearson’s correlation coefficient. F Measurement of cell length (apical to basal membrane) and (G) cell width (lateral to lateral membrane). Data are representative of at least n = 3 different z-slices per organoid and of at least n = 3 biologically independent organoids. For each z-slice, at least n = 3 individual cells with clear apical to basal delineation were used for quantification. Graphs indicate the mean ± S.E.M. Statistical significance was determined using ordinary one-way ANOVA. ns not significant (p > 0.05). Monoallelic Becn1 deletion in the intestinal epithelium causes goblet cell defects Complete loss of BECLIN1 in the intestinal epithelium causes severe, spontaneous enteritis originating in the small intestine [23]. Accordingly, the fatal nature of this phenotype, and the need for early euthanasia limited assessment of whether the colon is similarly affected. In the heterozygous setting, however, we observed no spontaneous small intestinal inflammation. Instead, from 14 days post gene deletion onwards (Fig. 1D), Becn1IEC+/- mice exhibited shortened colonic crypts, suggesting that the colon may represent a site of vulnerability to partial BECLIN1 loss over time. Guided by this observation, we next examined the colonic epithelium more closely. Periodic-Acid-Schiff (PAS) and Alcian Blue (AB) staining revealed a significant reduction in PAS-AB staining in the distal colon (Fig. 4A - top panels, B, Supplementary 5A), but not the proximal colon (Fig. 4B), of Becn1IEC+/- mice compared to Becn1IEC+/+ mice (Fig. 4A, B, Supplementary Fig. 5A), suggesting reduced goblet cell-associated mucin content and/or goblet cell abundance [13]. As goblet cell maturation involves migration to the crypt surface, we also quantified PAS-AB+ve mucin area in the lower (immature) versus upper (mature) colonic crypt compartments (Fig. 4C, Supplementary Fig. 5B) [13, 47]. PAS-AB+ve mucin staining was reduced in both lower and upper crypt compartments, but more so in the latter (50% versus 15%) (Fig. 4C). This maturation defect was supported by the enlarged cytoplasmic mucin (theca) areas in the upper crypt compartments of remaining mature goblet cells of Becn1IEC+/- mice compared with Becn1IEC+/+ mice (Fig. 4A, D), consistent with impaired mucin secretion. These results show that reduced BECLIN1 results in reduced goblet cell-associated mucin content and defective mucin production which can potentially compromise mucosal barrier function. Monoallelic Becn1 deletion leads to goblet cell defects. A Representative PAS-AB-stained distal colon sections from untreated and DSS-treated Becn1IEC+/+ and Becn1IEC+/− mice. Scale bar = 100 μm. Black arrowheads depict enlarged goblet cells in Becn1IEC+/− mice compared with Becn1IEC+/+ goblet cells indicated with orange arrowheads. Data represent n = 9 biologically independent mice of each genotype from n = 3 independent experiments. B Quantification of PAS-AB+ve area (measured per µm of muscularis mucosae) normalised to gross colon length in the distal and proximal colon of untreated and DSS-treated Becn1IEC+/+ and Becn1IEC+/− mice. Data represent n = 5 to 7 biologically independent mice per genotype from n = 3 independent experiments. C Quantification of PAS-AB+ve area in the lower and upper colonic crypt compartments, as delineated in Supplementary Fig. 5B, in Becn1IEC+/+ and Becn1IEC+/- mice. Data represent n = 8 animals per genotype from n = 3 independent experiments. For each mouse, 3–6 crypts were analysed and averaged. D Quantification of goblet cell theca size in the upper colonic crypt compartments of Becn1IEC+/+ and Becn1IEC+/- mice. Data represent n = 7 to 8 biologically independent mice per genotype from n = 2 independent experiments. Graphs show the ± S.E.M. Statistical significance was determined by ordinary one-way ANOVA in (B) and unpaired (Student’s) t-test in (C, D). For crypt-specific quantifications in (C), full-length intact crypts with at least three contiguous intact crypts were analysed, with n > 3 crypts analysed per animal. The average goblet cell theca area in (D) was obtained by measuring >100 goblet cells per animal. Representative images used for quantification shown in Fig. 4B–D are shown in Supplementary Fig. 5A. DSS dextran sulphate sodium, PAS periodic acid Schiff, AB alcian blue, ns not significant (p > 0.05). Reduced BECLIN1 levels in intestinal epithelial cells increases susceptibility to DSS-induced colitis Goblet cells are secretory IECs specialised for mucus production, forming a critical barrier against pathogens and mechanical damage [48]. Disruption of this barrier is a hallmark of chronic inflammatory conditions, including IBD, often arising from goblet cell loss or defects in mucin formation, storage, or secretion [48–51]. To test whether the goblet cell defects seen in Becn1IEC+/− mice increase susceptibility to inflammation, 7- to 14-week old mice were treated with dextran sulphate sodium (DSS) to induce colitis, a widely used model of IBD pathogenesis [52]. Following Tamoxifen-induced gene deletion, mice received either 2% DSS in drinking water for 5 days or normal water (Fig. 5A). There were no genotype-dependent differences in starting body weight or DSS-water consumption (Supplementary Fig. 6A, B). Reduced levels of BECLIN1 in the intestinal epithelium predisposes to Dextran Sulphate Sodium-induced colitis. A Schematic diagram depicting the timeline of Tamoxifen-induced Becn1 gene deletion followed by 5 days of 2% Dextran Sulphate Sodium (2% DSS) treatment. Image created in BioRender.com. B Representative images of abdominal necropsy and C intestinal tracts of Becn1IEC+/+ and Becn1IEC+/− mice, who received normal drinking water (untreated) or 2% DSS drinking water, at endpoint. D Colon lengths of mice from each treatment group at endpoint. E Changes in body weight of Becn1IEC+/+ and Becn1IEC+/− mice during the 2% DSS treatment period, normalised to body weight at the start of DSS treatment (day 9). F Diarrhoeal and (G) haematochezia scores of Becn1IEC+/+ and Becn1IEC+/− mice who received either normal drinking water or 2% DSS drinking water. Scores were determined based on parameters outlined in Table 1. H Disease Activity Index (DAI) score of Becn1IEC+/+ and Becn1IEC+/− mice receiving either normal drinking water or 2% DSS drinking water, calculated by summing the scores of the parameters outlined in Table 1. I H&E-stained FFPE distal colon sections from Becn1IEC+/+ and Becn1IEC+/− mice who received normal drinking water (untreated) or 2% DSS drinking water. The black double-headed arrow denotes the muscularis mucosae, and goblet cells are indicated by *. Immune cell infiltration in the submucosa and mucosa is marked by black and red open arrowheads, respectively. A magnified area (indicated by the box) provides a closer view of the mucosa, illustrating the increased immune infiltration in DSS-treated Becn1IEC+/− mice. J Percentage of injured or inflamed epithelium, normalised to the length of the muscularis mucosae, in the colon of Becn1IEC+/+ and Becn1IEC+/− mice receiving either normal drinking water or 2% DSS drinking water. K The histological colitis scores (HCS) of Becn1IEC+/+ and Becn1IEC+/− mice with and without DSS treatment was calculated based on described methods [55]. L Percentage of eroded epithelium, normalised to the length of the muscularis mucosae, in the colon of Becn1IEC+/+ and Becn1IEC+/− mice receiving either normal drinking water or 2% DSS drinking water. Data are representative of n = 7 to 9 biologically independent mice per genotype and treatment group, from n = 3 independent experiments in (D, F, G, H, J, K, L). Graphs indicate the ± S.E.M. Statistical significance was determined using ordinary one-way ANOVA except in (E) where changes in body weight were determined using two-way ANOVA with Tukey’s post-hoc test. Images in (B, C, I) are representative of at least n = 7 to 9 biologically independent mice from n = 3 independent experiments. At endpoint, both DSS-treated genotypes displayed expected colitic changes, including darker, reddish areas in the colon consistent with hyperaemia/inflammation (Fig. 5B, C) and colon shortening with swelling and constriction (Fig. 5C, D), but gross colonic changes were similar (Fig. 5B–D). The small intestines of Becn1IEC+/− mice were significantly shorter than Becn1IEC+/+ controls regardless of DSS treatment (Supplementary Fig. 6C). Both genotypes lost weight after DSS exposure, but this was greater in Becn1IEC+/− mice (Fig. 5E). They also developed more severe diarrhoea, with a higher proportion producing pasty to liquid stools (Fig. 5F), though hematochezia did not differ (Fig. 5G). Importantly, the composite disease activity index (DAI), incorporating weight loss, stool consistency, and rectal bleeding (Table 1) [53, 54], was significantly higher in Becn1IEC+/− mice, indicating greater colitis severity (Fig. 5H). Histological assessment of DSS-treated colons was performed using a blinded scoring system evaluating crypt architecture, inflammatory cell infiltration, and epithelial integrity, with the composite score reported as the histological colitis score (HCS) [55]. Untreated mice of both genotypes had minimal mucosal damage and low histological colitis scores (HCS) (Fig. 5I, J, K). DSS treatment increased the HCS scores in both genotypes, but Becn1IEC+/− mice scored approximately 2.2-fold higher (Fig. 5K), with more severe epithelial injury and crypt attenuation, epithelial loss, and immune infiltration (Fig. 5I, J). These mice also had more regions of epithelial erosion (Fig. 5L), characterised by complete crypt loss and dense localised immune cell infiltration. Lymphoid follicle counts and sizes did not differ between groups (Supplementary Fig. 6D, E), indicating that inflammation was largely confined to the epithelial layer. Hence, monoallelic Becn1 loss worsens DSS-induced colitis, indicating increased sensitivity to intestinal inflammation. Reduced BECLIN1 levels disrupt neutral mucins at baseline and exacerbates mucin loss following DSS treatment We next investigated goblet cell responses in DSS-treated Becn1IEC+/+ and Becn1IEC+/− mice. The epithelial damage in both DSS-treated Becn1IEC+/+ and Becn1IEC+/− mice (Fig. 5I–L) was accompanied by a significant reduction in PAS-AB+ve mucin area in the distal colon compared with their respective untreated controls (Fig. 4A, B), consistent with the distal colon being the primary site of DSS-induced injury [56]. Notably, the epithelial damage in the distal colon was exacerbated in DSS-treated Becn1IEC+/− mice, where the PAS-AB+ve mucin area was significantly reduced when compared with DSS-treated Becn1IEC+/+ mice (Fig. 4A, B, Supplementary Fig. 5A). We also assessed PAS-AB staining across the entire colon, which revealed a loss of PAS+ve neutral mucins in Becn1IEC+/− mice compared to Becn1IEC+/+ in the untreated groups, with no significant change in AB+ve acidic mucins (Fig. 6). Notably, following DSS treatment, unlike Becn1IEC+/+ mice, Becn1IEC+/− mice exhibited a loss of both neutral and acidic mucins, consistent with the exacerbated epithelial damage observed (Figs. 5I–L, 6). This reveals a vulnerability of the mucosal barrier when BECLIN1 is reduced, reinforcing its role in epithelial homoeostasis. The selective baseline loss of neutral mucins, and subsequent loss of both mucin types after DSS treatment, highlights the critical importance of neutral mucins in protecting against DSS-induced colitis. Reduced BECLIN1 levels disrupt neutral mucins at baseline and exacerbates mucin loss following DSS treatment. Representation of acidic (AB+), neutral (PAS+), and total (Merge) mucins in untreated and DSS-treated Becn1IEC+/+ and Becn1IEC+/− mice. Scale bars = 1 mm. The areas of acidic (AB+ve) and neutral (PAS+ve) mucus were quantified per µm of muscularis mucosae and normalised to gross total colon length. Data are representative of n = 5 to 7 biologically independent mice per genotype and treatment group, from n = 2 independent experiments. Graphs show the ±S.E.M. Statistical significance was determined by ordinary one-way ANOVA. Images are representative of at least n = 5 to 7 biologically independent mice from n = 2 independent experiments. Only animals that received DSS treatment from the same lot number were included for analysis. DSS dextran sulphate sodium, PAS periodic acid Schiff, AB alcian blue, ns not significant (p > 0.05). Discussion We previously identified an essential role for BECLIN1 in maintaining intestinal homoeostasis and barrier integrity mediated through its autophagy and endocytic trafficking functions [22, 23]. Given that homozygous Becn1 deletion causes fatal small intestinal enteritis in adult mice [23], we investigated whether partial reduction, mimicking clinically relevant haploinsufficiency, also impairs intestinal function. Accordingly, this study is specifically focused on the consequences of haploinsufficient BECLIN1 expression, as the rapid and fatal small intestinal pathology associated with homozygous deletion precludes detailed analysis of colonic phenotypes in that setting [23]. Unlike homozygous BECLIN1 deletion, monoallelic loss did not trigger overt spontaneous disease. However, Becn1IEC+/− mice did display subclinical abnormalities, including shortened small intestines and alterations in epithelial cell morphology and organisation. Organoid studies demonstrated modest increases in cytoplasmic RAB5+ve early endosomes but no enlargement, indicating that trafficking disruptions in the heterozygous state fall below the threshold seen in homozygous deletion [22, 23]. Instead changes in E-CADHERIN and F-actin distribution suggested cytoskeletal remodelling and altered cell shape, likely representing adaptive mechanisms to preserve junctional integrity, rather than explicit barrier failure. These data reinforce the idea that the small intestine is the primary site of overt pathology in homozygous deletion, whereas heterozygous loss is tolerated in this compartment under baseline conditions. By contrast, the novel finding of this study is the identification of colonic goblet cell defects in Becn1IEC+/− mice, a phenotype that was not apparent in the homozygous setting due to the rapid onset of fatal small intestinal disease precluding assessment of colonic changes [23]. Goblet cells were disproportionately affected by partial BECLIN1 loss, with reduced numbers, altered maturation (larger thecae), and diminished mucin staining, especially in the upper crypt. This aligns with the known role of autophagy in promoting goblet cell function via ER stress alleviation and mucin handling [11]. Given the central role of goblet cells and mucins in maintaining the mucus barrier, this baseline defect highlights the colon as a site of vulnerability to partial BECLIN1 loss. Following DSS challenge, Becn1IEC+/− mice failed to mount the same level of compensatory mucin response as wild-type controls, showing depletion of both neutral and acidic mucins. The baseline vulnerability due to selective mucin loss, compounded by the broad depletion after DSS treatment, points to a barrier-specific weakness rather than uncontrolled inflammation, consistent with goblet cell dysfunction as a driver. These findings underscore the importance of neutral mucins as a critical buffer against colitis, in line with prior findings from MUC2-deficient mice [51] and mucin abnormalities in IBD patients [57–62]. They also extend current knowledge by positioning BECLIN1 insufficiency as a mechanism that selectively undermines goblet cell function and mucosal defence. Although IBD GWAS have not identified BECLIN1 single nucleotide polymorphisms (SNPs), emerging evidence implicates both IEC-intrinsic and -extrinsic mechanisms that lower BECLIN1 levels, thereby exacerbating colitis. For instance, cyclic GMP-AMP synthase (cGAS) promotes autophagy in IECs through interaction with BECLIN1, and its loss worsens colitis [63]. Similarly, the m6A nuclear reader, YTHDC1, stabilises Becn1 mRNA and is downregulated in DSS colitis and in IBD patient microbiota [35, 64]. Moreover, while no IBD-associated SNPs have been identified within Becn1 itself, variants in upstream regulators such as STAT3, a transcriptional regulator of Becn1, have been implicated in IBD pathogenesis [65–67]. Caspase-mediated cleavage of BECLIN1 during apoptosis may also contribute to reduced protein levels in inflamed tissue, amplifying epithelial injury [68, 69]. Looking forward, resources such as atlases of IBD patient samples [70, 71] provide opportunities to examine BECLIN1 expression at endoscopically defined disease margins. Expanding clinical analyses will be critical to determine whether there are changes in BECLIN1 levels across broader clinical cohorts and to assess its potential as a biomarker of IBD pathogenesis and severity. In summary, we show that while heterozygous BECLIN1 loss is tolerated in the small intestine under baseline conditions, it creates a colonic vulnerability characterised by goblet cell dysfunction and mucin depletion, sensitising the mucosa to DSS-induced colitis. These findings establish BECLIN1 as a critical epithelial defence regulator, with implications for stratifying patients by barrier vulnerability and guiding precision therapeutic interventions in IBD. Supplementary information Supplemental Information Original Data Author contributions JJ and ST designed, performed, and analysed experiments and wrote the paper; TJH, YWNR, SLE, AHA-A, KP, SNY, ME, DB, CMR, RN, and LJJ designed, performed experiments, analysed and interpreted data; PDC, KD, BTK, ASY, JMM, BC, ALS, and JM. Murphy analysed and interpreted data; WDF and EFL designed the project, analysed, interpreted data, and wrote the paper. All authors commented on the manuscript. Materials availability Correspondence and requests for materials should be addressed to Erinna F. Lee or Walter D. Fairlie. Competing interests JMM, ALS, KMP, and SNY contribute to, or have contributed to, the development of necroptosis pathway inhibitors in collaboration with Anaxis Pharma Pty. Ltd. JMM has also received research funding from Anaxis Pharma Pty. Ltd. All other authors declare no competing interests. Footnotes Edited by Professor Mauro Piacentini Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. These authors contributed equally: Walter D. Fairlie, Erinna F. Lee. Contributor Information Walter D. Fairlie, Email: Doug.Fairlie@onjcri.org.au Erinna F. Lee, Email: Erinna.Lee@latrobe.edu.au Supplementary information The online version contains supplementary material available at 10.1038/s41419-026-08984-8. References Associated Data Supplementary Materials Supplemental Information Original Data Data Availability Statement Correspondence and requests for materials should be addressed to Erinna F. Lee or Walter D. Fairlie.