Poster Presentation 50th International Society for the Study of the Lumbar Spine Annual Meeting 2024

INTRODUCTION:

Inflammation is a key hallmark of painful vertebral bone marrow (BM) lesions called Modic type 1 changes (MC1). From histological and bulk RNA sequencing analyses, there are indications for monocyte, neutrophil, T- and B-cell involvement in inflammatory MC1 processes. However, a comprehensive understanding of the MC1 cellular composition and their interactions is still lacking. Here we used single-cell RNA sequencing to resolve the cellular composition of MC1 BM.  

METHODS:

From low back pain patients with MC1 undergoing lumbar spondylodesis (n=4), one MC1 and one intra-patient control BM biopsy was collected. Single cell suspensions were obtained by digestion and flushing, CD45⁺CD66b⁺ neutrophils were depleted by cell sorting, and 10'000 cells were sequenced (10x Genomics). Reads were aligned and counted, dimensionality was reduced, highly variable features were identified, clustering was performed, data was integrated, and clusters were annotated. Fractions between MC1 and intra-patient controls were compared with paired t-tests. Cellular interactions between clusters were explored using the “CellChat” tool. Differentially expressed genes (p<0.05) between MC1 and intra-patient controls per cell clusters were identified by pseudobulk differential expression analysis. Dysregulated processes/pathways were identified with overrepresentation analysis (ORA).

RESULTS:

From a total of 69’415 cells (MC1: 37’365; controls: 32’050) included in the analysis, 74 clusters were identified. Clusters were assigned to 19 major cell types (Figure 1A). Changes in cell type fractions were very heterogeneous among patients. CD16⁺monocytes (MC1 vs. control: +66.7%, p=0.08) and plasmacytoid dendritic cells (pDCs) (+55.0%, p=0.12) showed the largest fraction increase in MC1 and were the only two cell types consistently higher in MC1 among all patients (Figure 1B). CD16⁺monocytes are pro-inflammatory cells shown to expand in and contribute to cartilage destructive joint diseases. pDCs, like conventional dendritic cells, link the innate and adaptive immune system and are involved in T-cell activation. Fractions of CD4 (+35.9%, p=0.29) T-cells were further found to be increased in MC1. Besides the fractional increase, MC1 CD16⁺monocytes (+55.0%) and pDCs (+43.0%) also made substantial more cellular interactions than control. The interaction increase in MC1 pDCs was mostly attributed to an interaction increase with CD4 T-cells (+26.2%). ORA revealed a pro-inflammatory transcriptome in both MC1 CD16⁺monocytes and pDCs (Figure 1B). Pathway analysis of MC1 pDCs further showed enriched T-cell activation (FDR<1.0E-3) and response to purine-containing compounds (FDR<1.0E-3).

DISCUSSION:

Fractions and cellular interactions of CD16⁺monocytes and pDCs are increased in MC1 BM. This could indicate a pathomechanistic relevant role for CD16⁺monocytes and pDCs in MC1. CD16⁺monocytes become activated by antigen-antiobody-immune complexes. This could be an activation mechanism in MC1, since MC1 BM contains increased antibody-producing plasma cell infiltrates. pDCs can become activated by neutrophil extracellular traps (NETs) and they are able to induce T-cell responses. MC1 BM neutrophils are activated and pDCs have an upregulated response to purine-containing compounds (which are parts of NETs), suggesting an expansion/activation mechanism. Additionally, the observed T-cell fraction increase might also result from induction by pDCs. In summary, this study unraveled CD16⁺monocytes and pDCs as potential cellular drivers of inflammatory MC1 processes and lays the foundation for future pathomechanistic studies.