Necrotising enterocolitis triggers a TNF-driven gut-blood-brain cascade that disrupts the blood-brain barrier and induces TNFR1-skewed neuronal apoptosis and pyroptosis

Abstract

Necrotising enterocolitis (NEC) in very-low-birthweight (VLBW) infants is associated with neurodevelopmental difficulties, but the mechanisms linking intestinal disease to brain injury remain unclear. We asked whether NEC propagates along a tumour necrosis factor alpha (TNF-α)-centred gut-blood-brain pathway that compromises the blood-brain barrier (BBB) and activates inflammatory neuronal death programmes. In a retrospective cohort, inflammatory biomarkers at symptom onset (C-reactive protein and procalcitonin) were related to Bayley-III composite scores at 24 months’ corrected age. We then used a neonatal rat NEC model with daily oral etanercept, a soluble TNF receptor fusion protein, or saline from postnatal day 2 to day 14 to assess intestinal histopathology and permeability, circulating microbial products, inflammation, BBB permeability, early neurobehaviour, microglial activation, and neuronal apoptosis and pyroptosis with TNF receptor (TNFR) localisation. An intestinal-epithelial cell–primary cortical neuron transwell co-culture tested whether LPS-injured epithelium was sufficient to release TNF-α, disrupt barrier integrity, and injure neurons, and whether etanercept mitigated these effects. In infants, higher C-reactive protein and procalcitonin at NEC onset were associated with lower Bayley-III motor, cognitive, and language scores. In rats, NEC caused small-intestinal injury with tight-junction loss and permeability, elevated circulating microbial products and TNF-α, increased BBB leak, activated microglia, and induced neuronal apoptosis and pyroptosis with a TNFR1-skewed and TNFR2-reduced profile. NEC also impaired early sensorimotor performance, and worse performance was associated with higher circulating C-reactive protein, procalcitonin, and TNF-α. Etanercept reduced inflammatory burden, improved intestinal and BBB integrity, and improved sensorimotor performance. In co-culture, LPS activated epithelial NF-κB and MAPK signalling, increased epithelial TNF-α, and reduced barrier resistance. Neurons exposed to this inflammatory milieu showed TNFR1 upregulation, TNFR2 downregulation, activation of apoptotic effectors, and NLRP1 inflammasome-linked pyroptotic signalling, accompanied by synaptic injury. Etanercept blunted upstream epithelial signalling, preserved barrier function, and reduced TNFR1-skewed neuronal injury programmes. Acute inflammatory burden at presentation tracked with poorer 24-month neurodevelopment, and experimental data support TNF signalling as a plausible, modifiable relay linking barrier failure to BBB dysfunction, microglial activation, and TNFR1-skewed neuronal injury. These findings support risk stratification using routine biomarkers and motivate mechanism-informed trials targeting inflammatory amplification in NEC. In VLBW infants, higher CRP and procalcitonin at NEC presentation were associated with poorer Bayley-III outcomes at 24 months’ corrected age. Experimental NEC increased intestinal permeability and circulating microbial products, accompanied by BBB leakage and microglial activation. NEC induced TNFR1-skewed neuronal apoptosis and pyroptosis with reduced TNFR2 signalling in vulnerable brain regions. TNF blockade with etanercept attenuated gut injury, systemic inflammation, BBB disruption, microgliosis, and neuronal death signatures. In epithelial-neuron co-culture, epithelial inflammation was sufficient to impair barrier resistance, increase TNF-α exposure, and injure neurons. Symptom-onset CRP and procalcitonin may support risk stratification and inform early interventional study design in NEC.