Gut microbiota in congenital scoliosis: full-length PacBio 16S rRNA analysis of patient and subtype profiles.

BACKGROUND

Congenital scoliosis (CS) is a pediatric spinal disorder with largely unknown etiology. Emerging evidence suggests that the gut microbiota (GM) may influence bone development and spinal health, as shown in idiopathic scoliosis studies. However, GM in CS has not been comprehensively analyzed, especially using third-generation full-length 16S rRNA sequencing. We therefore compared the fecal microbiota of CS patients with that of healthy controls, and also evaluated differences between CS subtypes - Failure of Vertebral Formation (FVF) and Failure of Vertebral Segmentation (FVS) - -using PacBio full-length 16S sequencing.

MATERIALS AND

METHODS

We collected pre-treatment stool from 29 CS children (13 FVF, 16 FVS) and 11 age- and sex-matched healthy controls. DNA was sequenced on the PacBio SMRT to generate high-quality full-length 16S reads. QIIME2 (DADA2) produced amplicon sequence variants. We computed α-diversity (Shannon, Simpson, etc.) and β-diversity [(Bray-Curtis principal coordinates analysis/non-metric multidimensional scaling (PCoA/NMDS)]. Taxonomy at phylum, family, and genus levels was compared by Wilcoxon tests, and linear discriminant analysis Effect Size (LEfSe) was used to identify differentially abundant taxa. Phylogenetic Investigation of Communities by Reconstruction of Unobserved States (PICRUSt2) predicted functional pathways.

RESULTS

Alpha diversity indices (Shannon, Simpson, etc.) showed no significant differences between CS and healthy controls (all P > 0.05), nor between FVF and FVS subgroups. Beta diversity (PCoA/NMDS) indicated overlapping community structure across groups. Both cohorts were dominated by Firmicutes and Bacteroidota. CS had slightly lower Bacteroidota ( P = 0.0456) and Firmicutes_A ( P = 0.0394). At family level, Ruminococcaceae, Oscillospiraceae, Acutalibacteraceae, Tannerellaceae, and Acidaminococcaceae were modestly reduced in CS ( P < 0.05); Bacteroides_H was lower at genus level ( P = 0.0119). No taxa differed between FVF and FVS. LEfSe and PICRUSt2 detected no robust biomarkers or pathway shifts.

CONCLUSION

This is the first study using PacBio full-length 16S sequencing to profile GM in CS. We found no significant dysbiosis or diversity differences in CS vs healthy controls or between subtypes, consistent with recent reports in idiopathic scoliosis and CS. Some subtle compositional shifts were detectable (e.g., reduced Bacteroides_H in CS). These may reflect the deeper taxonomic resolution achieved by full-length sequencing. Despite largely negative findings, our work fills a knowledge gap and underscores clinical relevance: understanding microbiota adds to the holistic view of scoliosis pathogenesis. Future work will extend this into a longitudinal pre- and post-operative study to assess perioperative microbiome dynamics.