Cureus. 2026 Jul 28;18(7):e113549. doi: 10.7759/cureus.113549. eCollection 2026 Jul.
ABSTRACT
BACKGROUND: Major depressive disorder (MDD) and amyotrophic lateral sclerosis (ALS) are usually treated as unrelated, yet depressive symptoms occur in a substantial minority of people with ALS and may appear early. These symptoms are heterogeneous and may reflect syndromal MDD, psychological and functional burden, fatigue, apathy, pseudobulbar affect, frontotemporal involvement, sleep or respiratory disturbance, medication effects, or shared affective vulnerability. A proposed pruning-continuum model suggests both disorders may share vulnerability in microglia-mediated synaptic pruning, with ALS amplified by autophagy and protein-quality-control failure and MDD by RNA-processing, stress, and immune dysregulation. We performed an exploratory secondary transcriptome-wide association study (TWAS)/pathway-integration analysis to test whether predefined nicotinamide mononucleotide (NMN)-nominated pathways map onto this vulnerability.
METHODS: We integrated precomputed S-PrediXcan outputs for MDD and ALS across available brain-relevant tissues. Ten Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were predefined from a prior re-analysis of NMN-associated transcriptional programs in aged mouse metabolic tissues. Mouse-derived candidates were represented by human ortholog symbols before the human TWAS screen. The analysis tested nominated pathways rather than the 35-gene NMN-robust list as a standalone set. Cross-tissue screening used Stouffer Z aggregation, tissue-level Wilcoxon testing, competitive permutation testing, percentile bootstrap intervals, pairwise disease statistics, Levene variance tests, concordance measures, and leave-one-out sensitivity analysis. No analysis was treated as confirmatory or evidence of causal mediation.
RESULTS: MDD showed the strongest Stouffer-based exploratory signal in the synaptic vesicle cycle pathway, with a meta-across-tissue Stouffer Z of 3.41 and a wide bootstrap 95% confidence interval of -0.46 to 7.40. This signal did not survive competitive permutation testing (p = 0.1222) or Wilcoxon testing (p = 0.1926). The strongest tissue-level result occurred in the amygdala (Z = 4.057; nominal Wilcoxon p = 0.0093), although tissue-level permutation testing was not performed in the multi-gene-set run. ALS showed no significant meta-across-tissue enrichment among the 10 nominated pathways but displayed candidate gene-level signals in autophagy, endosomal, and vesicle-related genes, including TBK1 and C9orf72. Exploratory Levene tests indicated variance heterogeneity in the regulation of the actin cytoskeleton, endocytosis, and neuroactive ligand-receptor interaction; the actin cytoskeleton and endocytosis remained significant in pooled global false discovery rate (FDR) analysis. Fourteen genes were influential in at least two focus pathways, including EGF, KNG1, FGF8, RAC1, PAK1, PAK2, RAF1, MAPK1, and FGFR1.
CONCLUSIONS: These findings are hypothesis-generating. MDD and ALS may stress overlapping cellular logistics processes while engaging largely different genes. MDD showed the strongest exploratory pathway-level signal in synaptic vesicle biology, whereas ALS showed candidate gene-level coherence in autophagy and endosomal processes without significant meta-pathway enrichment. NMN/NAD+ repletion is not established as a treatment for MDD, ALS, or their comorbidity. These findings generate hypotheses about NAD+-linked cellular stress pathways for future preclinical and clinical studies.
PMID:42529685 | PMC:PMC13419321 | DOI:10.7759/cureus.113549
