The apical amino-acid transporter b0,+ (SLC3A1/SLC7A9) imports cystine and selenocystine into enterocytes and defines a selenium-utilization program in the healthy human intestine. Prior work established this program in normal tissue; the present study aimed to determine how disease states—specifically intestinal inflammation and cancer—remodel this epithelial redox axis.
The authors applied a multi-cohort single-cell framework with donor-level statistics to compare epithelial states across cohorts. Complementary experimental systems included small-intestinal enteroids treated with inflammatory cytokines (IFNγ, TNF), a network knockout approach to probe dependency relationships, analysis of an independent patient proteome, and a Caco-2 polarization model. These orthogonal datasets were used to test whether changes reflected transcriptional disruption, altered cell differentiation, or substrate-level supply effects on the selenoproteome.
In adult Crohn’s disease, co-expression of b0,+ in small-intestinal enterocytes was substantially reduced: median 4.9% of enterocytes positive versus 34% in the comparator group (P = 0.008). By contrast, paediatric inflammatory bowel disease did not show this suppression and preserved b0,+ expression. The observed loss in adults did not reflect a primary transcriptional uncoupling of the b0,+ program; instead, it corresponded to replacement of b0,+-high mature enterocytes by a dedifferentiated epithelial state. Importantly, coupling between SLC7A9 and SELENOP remained intact despite the shift in cell state.
Downstream within the selenium-utilization pathway, expression of the selenoenzyme GPX4 was selectively suppressed while the machinery for selenocysteine incorporation was preserved. This pattern establishes a severity-graded priming of the epithelium toward ferroptosis under inflammatory conditions, because substrate uptake is compromised at the level of b0,+ and key antioxidant selenoprotein GPX4 is reduced.
Functional experiments supported a model in which b0,+ sustains the selenoproteome primarily by supplying substrate rather than by driving transcription of selenoprotein genes. Treatment of small-intestinal enteroids with inflammatory cytokines IFNγ and TNF suppressed the selenium pole, consistent with observations in patient-derived single-cell data. A network knockout analysis further indicated that b0,+ dependency on downstream selenoproteins operates through availability of imported substrates (cystine/selenocystine) rather than transcriptional regulation of the selenoprotein program. These mechanistic inferences were corroborated by an independent patient proteome dataset and by a polarization model using Caco-2 cells.
In colorectal cancer colonocytes the pattern opposed that seen in inflammation. Tumour colonocytes showed near-absent b0,+ expression while inducing the xCT/thiol antioxidant program. This shift was statistically significant (P = 5.6 × 10−5) and configured the malignant epithelium toward ferroptosis resistance rather than susceptibility. In essence, cancer appears to bypass the selenium-dependent antioxidant supply node and instead mobilizes a thiol-based compensatory axis.
The findings define a dual-pole redox axis in the intestinal epithelium with b0,+ at the supply node: inflammatory states collapse the selenium pole, producing a dedifferentiated, ferroptosis-prone epithelial configuration, whereas cancer shifts the balance away from selenium toward an xCT-driven thiol antioxidant pole and ferroptosis resistance. The baseline state of this axis was reported as an exploratory correlate of anti-TNF treatment response. Consequently, the selenium-thiol axis represents a potential disease-associated vulnerability that could inform stratification or therapeutic targeting, pending further validation.
This report synthesizes multi-cohort single-cell analyses and several experimental systems; however, it is a preprint and has not been peer-reviewed. Specific methodological details, cohort compositions beyond the reported adult versus paediatric contrasts, and longitudinal treatment-response data were not fully detailed in the abstract. Causal translation into clinical interventions remains to be established, and additional work will be needed to define how modulation of b0,+ or the xCT axis affects disease course and therapeutic outcomes.