Pediatric intestinal enteroids derived from very early onset inflammatory bowel disease reveal differences in growth, morphology, and barrier function

Abstract

Pediatric inflammatory bowel disease (IBD) presents more aggressively than adult-onset disease; however, epithelial pathophysiology remains poorly understood due to limited access to patient tissue. Human intestinal enteroids (HIEs) are a powerful model for investigating disease phenotypes and heterogeneity. We characterised duodenal HIEs from two cases with Crohn’s disease (ED-71 and ED-81) and one control with intestinal obstruction (ED-19), all aged <2years. Analyses included growth curves, epithelial cell heights, immunofluorescence, histology, electron microscopy, gene expression and intestinal permeability. All HIEs maintained consistent growth in vitro, with ED-81 displaying the steepest trajectory. Morphological examination revealed variations in epithelial cell height, with ED-71 and ED-81[median(95%CI): 10.2(7.76–11.4) and 12.5(11.5–15)µm] displaying a more flattened appearance than ED-19[18.7(16.8–21.0)µm, p-value:<0.001]. HIE differentiation into intestinal cell types was confirmed by gene expression and microscopy. Intestinal permeability assays indicated compromised barrier integrity in IBD-derived monolayers, with ED-81 exhibiting the highest baseline permeability (5.27%) and EGTA-induced disruption (50.03%) compared to ED-19 (3.88% and 36.82%). Pediatric intestinal enteroids in IBD demonstrate differences in epithelial growth, morphology, and barrier function. HIEs serve as a potential translational model for pediatric IBD, facilitating the study of epithelial pathophysiology and guiding precision therapy. Further studies with more pediatric HIEs are needed to confirm these findings.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1038/s41598-026-62289-6.

Introduction

Inflammatory bowel disease (IBD), encompassing Crohn’s disease (CD) and Ulcerative colitis (UC), presents significant challenges, affecting patients and burdening the healthcare system globally. Despite historically being reported more often in Europe and North America, the geographic distribution of IBD has since expanded, affecting populations in countries previously considered low-risk areas^1–4. This rising trend in IBD incidence has been attributed to multiple factors, such as increased awareness, improved diagnostics, better access to specialised healthcare, and societal changes such as urbanisation, adoption of a westernised diet, and improved hygiene⁵. This underscores the need for a deeper understanding of the pathogenesis and progression of IBD in these populations. This is even more important in pediatric IBD, where disease characteristics and treatment responses vary from those in adults^3,6. Pediatric IBD is a distinct and often more aggressive subtype that is frequently associated with extensive intestinal involvement, growth failure, and complex therapeutic challenges⁷. The increasing prevalence of pediatric IBD is particularly significant among younger children, under the age of ten⁸. However, the study of pediatric IBD pathophysiology remains limited by the difficulty in obtaining sufficient patient-specific tissue, the absence of suitable animal models, and the inadequacy of conventional cell culture techniques to replicate the complexity of the intestinal epithelium.

Recent advances in three-dimensional intestinal organoid culture systems from patient-derived intestinal crypts or pluripotent stem cells have provided unprecedented opportunities to model human intestinal physiology and pathology ex vivo^9,10. Tissue-derived human intestinal enteroids (HIEs) closely mimic the cellular diversity, polarity, and functional characteristics of the native intestinal epithelium and can be expanded over time while maintaining donor-specific traits^11–13. These have proven especially useful in the study of epithelial barrier function, host–microbe interactions, and responses to pharmacologic agents in IBD^14–17. Recent studies have begun to characterise the pediatric IBD epithelium using organoid models, providing important insights into epithelial barrier dysfunction and inflammatory responses^18,19. However, these studies have largely focused on western pediatric cohorts and primarily focus on molecular characterization using transcriptomics and epigenetic studies. Representation of pediatric-derived HIEs from low-and middle-income countries (LMIC), including south Asia, also remains limited. The south Asian pediatric IBD population may present a distinct epidemiological and genetic context, and data from this setting would complement the existing literature. Expanding IBD organoid studies in LMICs will require more fundamental characterisation of HIEs, including assessment of morphology and barrier function across patient lines and evaluation of reproducibility and consistency.

In this pilot or proof-of-concept study, we utilised an established HIE biobank to characterise duodenal enteroids from pediatric IBD cases and a non-IBD age-matched control in southern India. We employed morphological and intestinal permeability assays to compare epithelial cell properties across IBD and control HIE lines. In this context, this study focuses on the reproducibility, consistency, and characterisation of pediatric IBD-derived organoids as a platform, complementing existing molecular approaches and providing preliminary data to support future mechanistic and therapeutic investigations. The present study was designed as an initial feasibility and characterisation investigation, findings are reported as preliminary observations and trends rather than definitive conclusions. By integrating these methods, we aim to advance the use of pediatric enteroids as a translational platform for intestinal disease biology in IBD, evaluate therapeutic responses and guide precision therapy.

Results

Clinical history, tissue histopathology and other laboratory parameters of IBD cases and control

Three pediatric subjects who were enrolled for duodenal tissue-derived HIE generation were included and comprised two IBD cases (derived HIEs ED-71 and ED-81) and one non-IBD control (ED-19).

ED-71 was a 1-year-old female who underwent endoscopy for complaints of chronic blood mixed diarrhea and failure to thrive. Colonoscopy showed multiple ulcers in the entire colon except the descending colon. Histopathological evaluation of duodenal biopsy showed minimal chronic duodenitis and increased intraepithelial lymphocytes, and colonoscopic biopsies demonstrated chronic active inflammation, features consistent with Crohn’s disease. There was no stricture or fistula, and she improved with immunomodulator therapy (azathioprine and mesalazine). ED-81 was a 2-year-old male who presented with a history of fever, diarrhea with occasional blood in the stool and failure to thrive. Colonoscopy showed multiple colonic ulcers from caecum to sigmoid colon. Histopathological evaluation of the duodenal biopsy showed moderate chronic duodenitis, and colonoscopic biopsy showed patchy chronic active colitis with chronic ulceration and fibrosis. He also had no strictures or fistula and was treated with immunomodulators (methotrexate and mesalazine). Both ED-71 and ED-81 cases are currently 9 and 10 years old, and at the last follow-up (2025 and 2024, respectively) were continuing treatment for Crohn’s disease. The control, ED-19, was an 11-month-old male who underwent intestinal surgery for neonatal intestinal obstruction and was diagnosed postoperatively with duodenal atresia requiring duodenoplasty. Histopathological evaluation of the duodenal mucosal biopsy showed normal mucosa and submucosa with no active inflammation. At the last follow-up at age 9 in 2025, he was healthy, and laboratory parameters were normal. Laboratory parameters for the two cases and the control are shown in Table 1.

Table 1.

Clinical and biomarker profile of study participants.

	ED-71	ED-81	ED-19
Age	1 yr	2 yr	11 months
Sex	Female	Male	Male
Diagnosis	Crohn’s disease	Crohn’s disease	Intestinal obstruction due to duodenal web
Biopsy reports	Duodenum: Minimal chronic duodenitis with focal increase in intraepithelial lymphocytes with minimal activity (~ 30/100 enterocytes). Ileum: Mild chronic active ileitis with mild villous atrophy and increased intraepithelial lymphocytes Cecum to rectum: Mild to patchy moderate colitis with mild activity	Duodenum: Mild to moderate chronic active duodenitis with patchy regenerative change. Ileum: No significant lesion Cecum to rectum: Patchy chronic active colitis with chronic ulceration and fibrosis.	Consistent with duodenal web with normal mucosa and submucosa (excision biopsy)
Fecal calprotectin	> 2000 µg/g	> 2000 µg/g	60.62 µg/g
Myeloperoxidase (MPO)	8962.06ng/mL	101381.33ng/mL	Below detection limit

All three HIE lines used in this study were established from intestinal biopsies obtained during diagnostic endoscopy performed at Christian Medical College, Vellore. Biopsies were collected approximately 8 years prior to the experiments reported here. HIEs were generated from fresh biopsy tissue at the time of collection and subsequently cryopreserved for long-term storage. Frozen mucosal biopsies were not used directly to establish organoids.

All experiments reported in this study were conducted within the last two years using these biobanked organoid stocks. HIE lines were revived from cryopreserved vials at passage 3–4 and expanded under standard conditions as detailed in the Supplementary Methods. All experiments were conducted at passage 6.

Growth curve analysis

To evaluate the proliferative capacity of the pediatric duodenal tissue-derived intact HIEs, we monitored growth over a 7-day culture period. As shown in Fig. 1, all HIE lines exhibited time-dependent increases in the diameter of the enteroids, indicating sustained growth. ED-81 HIEs showed the steepest growth trajectory, particularly between Days 4 and 6. In contrast, ED-71 HIEs showed moderate growth, while ED-19 HIEs displayed the slowest growth trajectory. These differences suggest inherent variability in growth potential among donor-derived lines.

Fig. 1.

Growth curves of intact HIEs derived from three ED patients (ED-19, ED-71, and ED-81) over 7 days. Shaded regions denote 95% confidence intervals. Growth curves were generated by tracking the same 10 organoids per dome longitudinally over 7 days per replicate per line. Representative phase contrast images of ED71 shown in Fig S1.

Differentiation of HIE

To evaluate epithelial lineage differentiation in the pediatric HIEs, we performed immunofluorescence staining and confocal imaging of lineage-specific markers on differentiated HIEs (Fig. 2, S2). Enterocytes of all three HIE lines identified by sucrase-isomaltase (SI) staining localised prominently to the apical brush border of the epithelium, confirming mature absorptive cell differentiation. Goblet cells detected using mucin-2 (MUC2) staining showed cytoplasmic staining and localisation towards the lumen of the HIEs, indicative of mucin production. Paneth cells visualised by lysozyme (LYZ) staining, showed a discrete punctate red signal. Enteroendocrine cells, marked by chromogranin A (CHGA) staining, were sparsely distributed but positive within the epithelial layer.

Fig. 2.

Confocal imaging of immunofluorescence staining of differentiated pediatric HIEs. Representative confocal microscopy images showing expression of intestinal epithelial lineage markers in differentiated HIEs. (a) Sucrase-isomaltase (SI, green) indicates brush border enterocytes (ED81). (b) MUC2 (red) indicates goblet cells (ED81) (c) LYZ (lysozyme, red) indicates Paneth cells, with punctate expression near crypt regions (ED19) (d) CHGA (ED81) (chromogranin A, red) identifies sparse enteroendocrine cells. Nuclei are counterstained with DAPI (blue). Scale bars = 20 μm for 60X magnification and 60 μm for 20X magnification as indicated.

To investigate epithelial alterations in pediatric IBD, we assessed gene expression markers representing epithelial lineages and stemness compared to controls. SI expression, an enterocyte differentiation marker, was decreased in both ED71 and ED81 IBD cases (p < 0.0001) compared to the ED-19 control, suggesting impaired epithelial maturation in IBD HIEs. The Paneth cell marker LYZ and the goblet cell marker MUC2 were elevated in ED81 (p < 0.0001) while CHGA, an enteroendocrine marker, was downregulated in both ED71 and ED81 (p < 0.0001). LGR5, which is a stem cell marker, showed increased expression in ED81 and decreased expression in ED71 compared to the ED-19 control. Ki-67, a marker of proliferation, was elevated in ED71 and 81 compared to ED19 (p < 0.001), indicating hyperproliferative responses in IBD HIE (Fig. 3), as seen in the growth curves. TLR4, TLR3, and IFN-β gene expression was upregulated in ED81 (p < 0.01 to p < 0.0001), whereas they were downregulated in ED71. These variations could be attributed to donor-specific differences in innate immune gene expression or polymorphisms, or potentially to the disease stage of IBD.

Fig. 3.

Altered gene expression in undifferentiated pediatric IBD HIEs compared to healthy control.qPCR analysis of epithelial and immune markers in human HIEs derived from healthy control (ED19, red) and IBD cases (ED71, green; ED81, blue). Shown are fold change relative to the non-IBD control line ED-19, calculated using the 2^-ΔΔCt method following normalization to GAPDH. Markers assessed include enterocyte (SI), Paneth cell (LYZ), enteroendocrine (CHGA), stemness (LGR5), proliferation (Ki-67), and innate immune response genes (TLR4, TLR3, and IFN-β). Data represent mean ± SD of triplicate qPCR measurements from three independent experiments using separate cryopreserved vials for each line. Statistical significance was determined using an unpaired t-test comparing ED-71 and ED-81 to ED-19. Significance markers indicate trends as: *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns = not significant.

Barrier function and structural integrity

Cell height measurements revealed significant variability among the pediatric tissue-derived HIE monolayers (Fig. 4a, S4). Quantitative analysis demonstrated that ED19 monolayers had the highest mean cell height, followed by ED81, while ED71 exhibited the lowest mean height. Statistical analysis indicated that the cell height of ED71 was reduced compared to both ED19 (p < 0.05) and ED81 (p < 0.05). These findings are supported by hematoxylin and eosin (H&E) staining in intact HIEs (Figure S4), which showed taller and more columnar epithelial morphology in ED19. The variation in cell height may reflect differences in epithelial differentiation and structural integrity among the ED lines, potentially correlating with disease severity or barrier function.

Fig. 4.

(a), Cell Height Measurements of HIE Monolayers. Box plot showing epithelial cell height (in µm). Statistical significance indicated by asterisks (p < 0.05). Outliers are shown as dots. (b,c) TEER Measurement and FITC-Dextran Permeability Assay in HIE Monolayers: Mean trans-epithelial electrical resistance (TEER) values of HIE monolayers derived from ED19, ED71, and ED81 measured at 2 h post-treatment. Error bars represent standard deviation. Mean FITC-dextran flux (µg/mL) across ED19, ED71, and ED81 HIE monolayers in untreated and treated conditions at 2 h post-treatment. (d), qPCR analysis of tight junction markers in human HIEs derived from healthy control (ED19, red) and IBD cases (ED71, green; ED81, blue). Shown are fold change relative to the non-IBD control line ED-19 for tight junction (ZO-1, Claudin-1). Data represent mean ± SD of triplicate qPCR measurements from three independent experiments.

Trans-epithelial electrical resistance (TEER) and FITC-dextran flux assays were performed on differentiated HIE monolayers (grown in transwells) derived from ED19, ED71, and ED81 to evaluate epithelial barrier integrity. At baseline (0 h) all three HIE lines exhibited high TEER values, indicating intact monolayers. After 2 h, TEER values decreased significantly on EGTA treatment in all three HIE lines, suggesting a potential compromise in barrier function (Fig. 4b). EGTA, a calcium chelator, was used as a positive control to impair tight junction integrity and increase paracellular permeability. This is consistent with its previously reported use as a barrier-disrupting agent in infant HIEs²⁰. FITC-dextran flux revealed significant differences in permeability upon treatment with EGTA. Under untreated conditions, all three lines maintained low permeability; however, baseline FITC-dextran flux was slightly higher in ED81 than ED19 and ED71, suggesting inherently weaker barrier properties in the former. Upon exposure to treatment with EGTA, all HIEs exhibited a marked increase in FITC-dextran flux (Fig. 4c). Notably, ED81 showed the highest permeability, followed by ED71 and ED19. These results indicate that the ED81-derived monolayer is the most susceptible to barrier disruption, with elevated baseline permeability and heightened sensitivity to EGTA treatment. Paradoxically, an upregulation in the expression of genes associated with tight junctions, specifically Claudin-1 and ZO-1, was observed in ED81 (Fig. 4d). These observations regarding baseline permeability and sensitivity to EGTA were reproducible across three independent experiments.

Ultrastructure of HIE

Electron microscopy of all three HIEs showed cells arranged in rings and nests. Some of the rings were lined by flattened to cuboidal epithelium with rudimentary microvilli on the inner luminal surface. Others were lined by tall columnar epithelium with more prominent microvilli. Some of the columnar cells had mucin vacuoles in the apical cytoplasm, suggesting differentiation towards a goblet cell phenotype. (Fig. 5a, c, d). Apart from rings, ‘nests’ of epithelial cells were also seen. Some of the nests had small lumina between the tall columnar cells, the luminal surface of which showed prominent microvilli (Fig. 5b). All epithelial cells in both rings and nests showed apical intercellular tight junctions and desmosomes between lateral cell margins. ED71 HIEs showed rings and nests with focal prominent clumping of nuclear chromatin, consistent with pyknosis, an early change of apoptosis. These cells also showed increased size and number of lysosomes. Many shrunken, dark cells with nuclear fragments containing condensed chromatin and degenerate organelles suggestive of apoptosis were also seen in ED71, some being extruded into the lumen of rings or smaller lumina of nests (Fig. 5e, f). These features were entirely absent in the control line ED19 and in the remaining IBD-derived line ED81, distinguishing ED71 as a line with a distinct apoptotic phenotype.

Fig. 5.

Transmission electron microscopy (TEM) images of ultra-thin sections: Representative images from enteroids demonstrate rings and nests. (a) Part of a ring lined by flattened to cuboidal epithelium, with rudimentary apical microvilli (ED81). (b) Part of an epithelial cell nest with small lumina lined by short to tall microvilli (green arrow) (ED19). (c, d) Tall columnar cells, lining rings, with prominent microvilli (green arrows) suggestive of enterocytes. Other cells show apical mucin vacuoles (orange arrows), suggestive of goblet cell differentiation (ED19). (e) Epithelial cells with nuclear condensation or pyknosis indicative of early apoptosis (yellow arrow) (ED71). (f) Shrunken apoptotic cells being extruded into the lumen of a ring (red arrow) (ED 71). Scale bar: 1–5 μm (red).

Discussion

In this study, we characterised pediatric patient-derived HIEs from two cases with IBD and one non-IBD control, generating a comparative platform to assess characteristics of the intestinal epithelium pertinent to pediatric IBD. By combining morphological, molecular, and permeability assays, we offer a thorough baseline examination of HIE characteristics in both IBD and non-IBD conditions, highlighting the value of pediatric HIEs for translational research in IBD. Reports from LMICs remain limited, and pediatric IBD HIEs from these regions are largely underrepresented. Recent emerging studies in pediatric and very early-onset IBD (VEO-IBD) have highlighted significant epithelial-intrinsic alterations, including barrier dysfunction and dysregulated inflammatory signalling, through the use of organoid-based models^18,19. Notably, these investigations have primarily focused on western pediatric cohorts and have emphasised deep molecular and multi-omic profiling approaches. In contrast, our study complements this evolving body of work by focusing on a south Asian pediatric cohort. By prioritising the reproducibility, consistency, and functional characterisation of the organoid platform itself, this pilot study addresses a critical geographic gap and establishes a robust foundation for future mechanistic and precision-medicine investigation in this underrepresented population.

The HIE growth curve analysis revealed significant variability in proliferative capacity among patient-derived lines, with the Crohn’s-derived line (ED-81) exhibiting the steepest growth trajectory, while the healthy control (ED-19) demonstrated the slowest growth. These differences may reflect disease-specific epithelial adaptations or altered crypt biology associated with chronic inflammation, as previously observed in adult IBD studies^9,14 or could be individual-specific²¹. Similar variability has been observed in adult IBD organoids, where studies involving single-cell analysis and biobanks have identified growth differences linked to disease activity and variations between patients^22–24. Additionally, epigenetic and transcriptional profiling have demonstrated that epithelial programs can categorise disease severity and phenotype^14,17, reinforcing the notion that the variability seen in pediatric HIEs represents authentic disease-associated adaptations.

All three pediatric HIEs exhibited the anticipated epithelial lineages; however, HIEs derived from IBD demonstrated altered lineage and stemness markers, indicative of impaired maturation and increased proliferative capacity⁹. Similar variability in innate immune gene expression has been reported, largely attributable to donor heterogeneity and inflammatory state at biopsy²¹. Several studies have noted enrichment of Paneth cell markers (e.g., LYZ) and upregulation of antigen-presentation pathways in IBD colonoids, suggesting heightened epithelial immune activation^25,26. Interestingly, Kelsen et al. observed reduced growth efficiency in pediatric IBD colonoids, measured as lower crypt-to-spheroid conversion, despite concurrent Paneth and immune marker upregulation²⁵. This contrast may reflect differences between conversion efficiency and long-term expansion potential^27,28, as well as variations in intestinal segment (our HIEs are duodenal-derived, whereas Kelsen et al. used colonoids), disease activity at the time of biopsy, inflammatory stress, passage number, and culture conditions²¹. We speculate that in our cases, chronic low-grade inflammatory signalling may have upregulated epithelial proliferative programmes, consistent with reported alterations in Wnt and EGFR signalling in Crohn’s disease. This interpretation is supported by the gene expression data presented in Fig. 3, which demonstrate that ED-81 exhibits higher relative expression of LGR5 and Ki-67 compared with ED-19, consistent with an enhanced stem cell and proliferative compartment in this IBD-derived line. Notably, ED-81 also shows elevated IFN-β-expression, suggesting a low-grade innate inflammatory tone that may further potentiate epithelial renewal and growth kinetics, in keeping with the known interplay between interferon signalling and intestinal epithelial proliferation. Furthermore, as these organoids were derived from cases diagnosed at a very young age (VEO-IBD), these distinct growth phenotypes may reflect intrinsic genetic underpinnings that fundamentally alter epithelial kinetics, a phenomenon increasingly reported in other VEO-IBD organoid models where epithelial-intrinsic alterations persist ex-vivo¹⁹. Collectively, these findings suggest that epithelial remodeling in inflammatory bowel disease (IBD) encompasses both the augmentation of immune lineage signatures and changes in growth dynamics, highlighting the complex nature of responses observed in patient-derived organoids. These observations are presented as speculative trends, given the pilot scale of this study.

Histological analyses revealed that ED-71 monolayers show reduced epithelial height and flattened morphology compared to ED-19 (columnar) and ED-81 (intermediate), reflecting impaired differentiation (Shown as Fig.S3). Similar morphological alterations in IBD-derived colonoids show reduced size, altered polarity, and loss of budding structures under inflammatory contexts²⁹. Epithelial flattening in organoid and monolayer systems correlates with compromised barrier integrity and heightened susceptibility to injury, with such phenotypes reversed by treatment^29,30. These changes mirror in vivo crypt atrophy and epithelial distortion, established markers of disease activity in IBD^31–33. Our patient-derived monolayer phenotypes recapitulate the structural features seen in vivo, suggesting epithelial height metrics may indicate differentiation capacity and barrier functionality.

The ultrastructural features observed across all three HIEs, including apical tight junctions, desmosomes, microvilli, and goblet cell differentiation, are consistent with native intestinal epithelium, confirming that these organoids faithfully recapitulate key structural and functional properties of the intestinal epithelial barrier. Strikingly, apoptotic features, including pyknosis and extrusion of degenerate cells into luminal spaces, were observed exclusively in ED71 and were absent in the control line ED19 and the other IBD-derived line ED81. This complete contrast suggests that the apoptotic phenotype in ED71 reflects a line-specific biological characteristic rather than a universal feature of IBD-derived enteroids. Increased epithelial apoptosis is well documented in IBD mucosa³⁴, and its preservation specifically in ED71 cells may reflect a particularly pronounced disease state in the donor, potentially linked to compromised barrier integrity. Formal quantification across a larger panel of lines would be required to establish this definitively.

Our intestinal permeability assays revealed donor-specific differences: ED-81 showed higher baseline permeability and greatest EGTA-induced FITC flux, suggesting tight-junction fragility. Similar barrier defects occur in IBD organoid models, where reduced TEER and increased flux correlate with impaired differentiation^20,35. EGTA chelation probes tight-junction stability, with increased sensitivity reflecting weakened junctional integrity³⁶. Elevated fecal calprotectin and MPO correlate with mucosal permeability³⁷, consistent with ED-81’s barrier vulnerability. These findings support an epithelial contribution to IBD pathology beyond immune-mediated injury^35,38.

Although ED81 HIE showed elevated expression of tight junction genes like Claudin-1 and ZO-1, increased permeability in FITC-dextran assays suggests a disconnect between gene expression and barrier integrity. The increased expression of innate immune genes like TLR3, TLR4, and IFN-β indicates inflammatory pathway activation, which disrupts epithelial junctions through cytokine-mediated modulation and increased cell turnover. These inflammatory signals may override repair mechanisms, causing barrier dysfunction despite upregulated junctional transcripts. Animal studies using DSS-treated organoids confirmed that inflammatory insults degrade tight-junction integrity and increase paracellular permeability, shown by increased FITC-dextran flux 1³⁶. In adult IBD colonoids, reduced TEER and disrupted tight junction integrity persist in remission samples^35,39.

The major limitation of this work is the small sample size, including the availability of only one non-IBD control, which restricts the generalizability of our findings. However, our results align with prior reports showing that patient-derived intestinal organoids retain donor- and disease-specific molecular and functional signatures. While our sample size is limited, the differences between the HIEs were robust. Additionally, we demonstrated reproducibility of our results across three independent experiments using matched passage numbers for the cultures. Together, these studies establish proof-of-principle for IBD HIE studies in LMICs. Additional studies with larger numbers of subjects are needed to make meaningful biological conclusions. Further, these HIEs were derived from duodenal biopsies; HIEs derived from colonic biopsies could have provided more information.

This exploratory study therefore supplements this growing body of literature by extending these observations to pediatric HIEs and highlighting the potential of this model to interrogate patient-level variation in epithelial physiology. Co-culture with immune cells or microbiota could further enhance the physiological relevance of this model. Future studies expanding patient numbers and incorporating transcriptomic and proteomic analyses will help identify molecular drivers of observed phenotypes. This platform could evaluate reactions to cytokines, microbial challenges, and therapeutic agents to reveal age and disease-specific vulnerabilities, aiding in precision therapy for pediatric IBD.

Methods

Patient recruitment and establishment of HIE biobank

A pediatric HIE biobank was established at Christian Medical College (CMC), Vellore, India between 2016 and 2018 as part of a study to establish HIE from children with environmental enteric dysfunction (EED). Briefly, children under the age of five years undergoing abdominal surgery (n = 26) or endoscopic biopsies (n = 25) for a range of clinical indications were recruited after written informed consent from an adult caregiver (parent or legal guardian) was obtained by a trained research nurse. Along with intestinal biopsy tissue, a stool sample and a blood sample were also collected, processed and banked. Derivation of enteroids from intestinal tissue were conducted in accordance with established protocols^40,41, as detailed in supplementary methods with 19 duodenal, 10 jejunal and 12 ileal HIE for a total of 41/51 successfully grown HIEs and banked in liquid nitrogen for further characterisation and future studies.

IBD and control HIE for characterisation

Three duodenal HIEs were selected for in-depth characterisation: two from patients diagnosed with very early onset IBD (Crohn’s disease) and one without any abnormalities in tissue histopathology, serving as a non-IBD age-matched control. The diagnosis of Crohn’s disease was ascertained by gastroenterologists in accordance with the Asia Pacific criteria⁴² with thorough assessment of clinical manifestations, endoscopic observations, histopathological evaluation, and radiological findings. Fecal calprotectin (Epitope Diagnostics, California) and myeloperoxidase (MPO, Immundiagnostik, IDK, Germany) were also tested. While the biobank contains 41 successfully grown HIE lines, the remaining 38 lines were excluded from this characterisation as they were derived from pediatric cases with varying clinical etiologies.

HIEs were revived from cryopreserved stocks at passage 3 or 4 and subsequently expanded as mentioned in the supplementary methods. For differentiation, expansion media was replaced with differentiation media at Day 4 and all assays were performed with differentiated HIE at Day 7. All experiments were performed using HIE lines at passage 6 with biological and technical replicates. To ensure reproducibility, gene expression and all functional assays were independently validated using three separate cryopreserved vials that were thawed and expanded separately. HIEs were grown as monolayers in transwells as described in the supplementary methods for cell height and permeability assays.

Growth curves

Growth curve measurements were performed using Image J/Fiji software version 2.14.0/1.54f by calibrating images to a known scale bar (10 μm) and measuring intact 3D HIE diameter using the line tool. For the HIE growth curve analysis, ten organoids were selected per dome at Day 0 and were tracked at subsequent time points across the 7-day experiment. All measurements obtained from 10 selected HIEs per replicate at each time point (days 1 to 7) were recorded in micrometres, and output data were saved as TIFF images and corresponding measurement files. LOESS smoothing was applied to visualise growth trends, and 95% confidence intervals were plotted to indicate variability.

Transmission electron microscopy (TEM) of HIE

Intact 3D HIEs were fixed in 3% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) with 2 mM CaCl₂ for 3 h at 4 °C. After washing in sodium cacodylate-sucrose buffer, samples were post-fixed in 1% OsO₄ in 0.2 M sodium cacodylate buffer for 2 h at 4 °C. Samples were dehydrated through graded ethanol (50–100%), treated with propylene oxide, and infiltrated with epoxy resin over three changes. HIEs were embedded and polymerized at 60 °C for 48 h. Semi-thin section (0.5–1 μm) were cut using Leica UCT ultramicrotome and stained with toluidine blue in borax. Ultrathin sections (~ 50–70 nm) were cut, mounted on copper grids, and stained with uranyl acetate and Reynold’s lead citrate. Sections were examined using a FEI Tecnai T12 microscope at 120 keV.

Immunofluorescence staining and confocal imaging of HIE

Intact 3D HIEs grown in Matrigel domes were fixed with 4% formaldehyde for 1 hour at room temperature (RT), followed by three PBS (1X) washes. Samples were blocked and permeabilised with 5% fetal bovine serum (FBS, MP Biomedicals, Ohio) and 2% Triton X-100 for 2 hours at RT. Primary antibodies (1:50 in 5% FBS, 0.25% Triton) were added and incubated at RT in the dark with gentle shaking. After washing, secondary antibodies (1:1000) were applied for 3 hours at RT. Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI, Invitrogen, California) (1:1000) for 30 min, followed by three PBS washes and clearing for 2 h. Slides were mounted with Prolong gold antifade reagent (Invitrogen, California) and sealed. Imaging was performed using confocal microscope with appropriate fluorescent filters. Antibodies are listed in Supplementary Table 5. Samples were imaged using an Olympus Laser Scanning Confocal Microscope system (spectral version), model Olympus FV1000. Images were captured at 20X and 60X magnifications using Olympus Fluoview Ver.3.16 software.

Gene expression of HIE

Intact 3D HIEs were incubated in cold water for 10 min, followed by adding 800 µL cold PBS and centrifugation at 800 × g for 3 min at 4 °C. The pellet was lysed in 350 µL RLT (RNA Lysis Tissue) buffer with 1% β-mercaptoethanol, incubated for 10 min at room temperature, and mixed with 350 µL 70% ethanol. The lysate was processed using spin columns (Qiagen, Germany) for RNA isolation. RNA was quantified using a Nanodrop, and 500 ng–2 µg was reverse-transcribed into cDNA (cDNA REV TRANS KIT, ThermoScientific, USA) in a 20 µL reaction using manufacturer’s protocol. Gene expression was assessed by qPCR using 5 ng/µL diluted cDNA, TaqMan Fast Advanced Master Mix, and gene-specific primer-probe pairs. All qPCR experiments were performed in triplicate and validated across three independent experiments using HIEs revived from a fresh cryopreserved vial in each instance to ensure biological and technical reproducibility. This included Sucrose Isomaltase (SI), Lysozyme (LYZ), Chromogranin A (CHGA), LGR5, Ki-67, ZO-1, Claudin-1, Toll-like receptor4 (TLR4), TLR3, and Interferon β (IFN-β) (listed in Supplementary Table) Flex (Applied Biosystems^™) as per manufacturer’s protocols. Negative controls included no-template and no-reverse transcriptase reactions to confirm the absence of contamination. GAPDH was used for normalisation.

Cell height measurement

Cell height was assessed in Hematoxylin and Eosin (H&E)-stained sections of differentiated HIE monolayers grown in transwells and then mounted on slides (Figure S4). Only typical enterocytes were included; cells with inclusions, vacuoles, or multilayered stacking were excluded. Heights were measured from the base of the cell to the apical surface, bisecting the nucleus. Three cells were measured per image for each sample across three separate images. Measurements were conducted using Image J/Fiji version 2.14.0/1.54f software after setting the image scale using the embedded scale bar. A blinded analyst performed all analyses to eliminate bias.

Intestinal permeability assays

Transepithelial electrical resistance (TEER) of differentiated HIEs grown on transwells (Corning, New York) was recorded using Millicell ERS-2 before and after a 2-hour treatment with 10mM EGTA (Millipore, Merck, USA). FITC-Dextran (Sigma-Aldrich, USA) serial dilutions (5 mg/ml) for the standard curve and a 1:100 working dilution were prepared during incubation. Following treatment, the media was replaced with 100 µL of diluted FITC-Dextran and incubated for two additional hours. Final TEER was measured, and 50 µL each of apical, basolateral, and standard curve samples were transferred to a 96-well plate. To assess permeability, fluorescence was read at 490 ± 9 nm excitation and 520 ± 9 nm emission using TECAN Infinite M200 PRO reader.

Statistical analysis

All analyses were conducted using R version 4.4, and all images were analysed using ImageJ/Fiji version 2.14.0/1.54 f. Data from 3 biological and technical replicates were included, and for imaging-based analyses, at least 10 measurements per replicate were obtained to ensure statistical robustness. For experiments involving continuous variables such as HIE diameter, cell height, gene expression, and permeability assays, data were analysed using descriptive statistics (mean ± standard deviation). For comparison between control and IBD HIE lines, unpaired two-tailed Student’s t-test was used. P-values < 0.05 were considered statistically significant. The study was approved by the Institutional Review Board, Christian Medical College, Vellore, India (IRB No: 8956). Written informed consent was obtained from the parents or legal guardians of all pediatric participants prior to biopsy collection. All procedures were conducted in accordance with the Declaration of Helsinki.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

B.V. contributed to study design, performed laboratory work and data analysis, and drafted the manuscript. B.S., S.B., and A.P. conducted laboratory experiments and A.J.J., A.K.D., and E.S. were involved in clinical data acquisition. M.E., H.W., and S.R. contributed to study design and provided training and resources. S.S.R.A. supervised the study, contributed to study design, provided resources and finalised the manuscript. All authors meet the criteria for authorship, reviewed the manuscript, and approved the final version for submission.

Funding

The EED study that led to the development of the HIE biobank was funded by the Gates Foundation OPP1108228 to Tufts Medical Center, Boston, MA, USA; Baylor College of Medicine, TX, USA and Christian Medical College, Vellore, India. Characterisation of HIE from IBD cases was supported by institutional funds at the Wellcome Trust Research Laboratory, CMC, Vellore.

Data availability

Data available with the corresponding author on request.

Declarations

Competing interests

The authors declare no competing interests.