Elucidating variable disease severity of X-linked adrenoleukodystrophy through multiomics and metagenomics of gut microbiome

Abstract

[eng] X-linked adrenoleukodystrophy (X-ALD) is a rare and devastating neurodegenerative disorder characterized by inflammatory demyelination of the brain and/or axonal degeneration in the spinal cord. This arises from a loss-of-function mutation in the ABCD1 gene, a peroxisomal transporter protein. Consequently, very long-chain fatty acids (VLCFAs) accumulate in tissues and plasma, serving as a pathognomonic biomarker for the disease. X-ALD manifests in three main phenotypes: childhood and adult cerebral X-ALD (ccALD and cAMN, respectively) and adrenomyeloneuropathy (AMN). Notably, these phenotypes can occur within a single family, suggesting a complex interplay of genetic, environmental, and stochastic factors influencing disease expressivity. This thesis employed untargeted metabolomic analysis to identify a distinct metabolic signature within the normal-appearing white matter of X-ALD patient brains. Additionally, shotgun metagenomic analysis of fecal samples revealed bacterial adaptations indicative of environmental stress in X-ALD patients. These adaptations included an upregulation of pathways associated with redox metabolism, bacterial biofilm production, and lipopolysaccharide (LPS) modification. Furthermore, we observed perturbations in branched-chain amino acids, likely a consequence of gut microbiota dysbiosis. We quantified neuronal and glial biomarkers like neurofilament light chain (NfL), glial fibrillary acidic protein (GFAP), ubiquitin C-terminal hydrolase L1 (UCHL1), and TAU protein levels in both plasma and cerebrospinal fluid (CSF) across the different X-ALD phenotypes. Notably, NfL and GFAP exhibited significantly greater elevations in cerebral X-ALD compared to AMN patients in both fluids. Interestingly, NfL, but not GFAP, emerged as a potential predictive marker for patients at risk of conversion to the rapidly progressive cerebral form of the disease. Additionally, we observed a decrease in plasma methionine levels specifically in cerebral X-ALD patients, while levels remained unaltered in AMN patients. Importantly, methionine levels exhibited an inverse correlation with NfL levels specifically in the cerebral X-ALD cohort. Furthermore, this work confirmed the presence of two established hallmarks of X-ALD pathogenesis: mitochondrial dysfunction and inflammation. We observed

elevated levels of mitochondrial DNA (mtDNA) and GDF15, supporting these established features. The findings presented in this thesis shed light on the potential significance of altered metabolite profiles in X-ALD pathogenesis, particularly their association with mitochondrial dysfunction, impaired myelinogenesis, and neuroinflammatory processes. We highlight the importance of plasma methionine as a readily accessible and measurable biomarker, especially in the context of cerebral inflammatory X-ALD. Additionally, this work proposes a novel hypothesis suggesting that gut microbiota dysbiosis may contribute to some of the metabolic alterations observed in X-ALD patients. Our human studies identified several disrupted homeostatic pathways, some of which may hold promise as potential therapeutic targets. This suggests the possibility of a multi-targeted therapeutic approach for X-ALD. Such an approach could combine medications to normalize amino acid levels, mitigate gut microbiota alterations, and potentially reduce intestinal inflammation, leading to synergistic and disease-modifying effects. Moreover, the assessment of plasma methionine levels holds significant promise as a valuable metric for monitoring disease activity and progression in clinical trials aimed at addressing demyelination in cerebral X-ALD patients.