Bacteria in human lumbar discs – subclinical infection or contamination? Metabolomic evidence for colonization, multiplication, and cell-cell cross-talk of bacteria

Abstract

The accumulating evidence associating sub-clinical infection with disc degeneration (DD) and the controversy of contamination versus infection mandates a further understanding of the microbial activity in the disc and host-microbiome interaction. To utilize a novel approach of metabolomics to probe the presence of bacterial metabolites involved in colonization, survival, and replication in human lumbar intervertebral discs (LIVD). An observational case-control study. Nucleus pulposus from the LIVD of three brain-dead voluntary organ donors (MRI normal and classified as controls) and of three patients undergoing surgery for disc degeneration (DD) (cases) were utilized. Untargeted metabolite profiling was carried out in six discs (3-controls and 3-cases) after extraction using methanol: acetonitrile: water (2:2:1) solvent system and acquired through HPLC-MS/MS platform using C18 reversed-phase column. From the total IVD metabolome, microbial metabolites were filtered by mapping against HMDB, ChEBI, SigMol, Siderophore database, ecdmb database, and PaMet databases. The biological functions of the metabolites were then studied by MSEA pipeline from Metaboanalyst, and the enrichment ratio, p-value, and Variably Importance Projection scores of the metabolites were calculated. Degeneration responsive changes in the abundance of the microbial metabolites were calculated based on the peak intensities between the control and cases. Mass spectrometry identified a total of 17601 and 15003 metabolites, respectively, in the control and degenerated discs. Preliminary mapping of the above metabolites against HMDB indicated the multiple sources, and of these, 64 metabolites were of microbial origin, accounting for 1.6% of the total IVD metabolome. Principle Component Analysis and Orthogonal Partial Least Square-Discriminant Analysis (OPLS-DA) showed distinct clustered patterns between control and disc degene`ration, indicating a strong variation in concentration, peak, and spectral values of the 64 metabolites between controls and cases. After the exclusion of metabolites that were also associated with humans, drugs, and food, 39 metabolites specific to bacteria were isolated. Nine were primary metabolites related to bacterial growth and survival, and the remaining 30 were secondary metabolites related to different environmental stress response activities. The three significant pathways (p<.001) which were predominant in the bacterial metabolites were autoinducer-2 biosynthesis, peptidoglycan biosynthesis, and chorismate pathway. In addition, a significant fold change of >1.0 was found for nine metabolites which included (S)-14-Methyilhexadecanoic acid related to P. acnes, 9-OxoODE, and 13-OxoODE related to gut flora, vibriobactin - a siderophore, tuberculosinol and iso-tuberculosinol, virulence factors of M. tuberculosis. There was also upregulation of Autoinducer- 2, an important "Quorum sensing molecule" involved in bacterial cross-talk. We identified several bacterial-specific metabolites participating in bacterial growth, survival, and cross-talk pathways. These were found in both groups but up-regulated in degenerated discs. The presence of Quorum sensing molecules and cell-cell interactions provides firm proof of colonization and growth. These findings indicate that the bacterial presence may not be mere contamination but could be colonization with a possible role in infection-mediated inflammation in DD. Proof of subclinical infection as an initiator of DD and documentation of exact germ and drug sensitivity will change the way millions of patients with non-specific low back pain (NSLBP) are treated across the world.