Skeletal site-specific variations in myeloid cells: insights from single-cell RNA sequencing of the mandible and femur

Abstract

To understand differences that exist between the cell populations in different skeletal sites, we performed an unbiased genetic survey via single-cell RNA sequencing of CD11b+ myeloid cells from mandibular- and femur-derived bone marrow of 2-mo-old C57BL/6 mice. Our results reveal transcriptomic evidence that suggests a uniquely inflammatory genetic profile of the mandibular-derived CD11b+ cells. The monocyte cell population found within the CD11b+ cells analyzed by scRNA-seq expressed Csf1r and Tnfrsf11a suggesting that this population contained osteoclast precursors. Osteoclasts of the craniofacial region facilitate processes such as tooth eruption and jawbone development. Evidence from multiple researchers suggests that craniofacial osteoclasts exhibit differences in size, gene expression, and activity compared to their counterparts within the appendicular skeleton. A biological mechanism to explain the observable difference between craniofacial osteoclasts and osteoclasts in the long bones has not been previously explored. This monocyte population had enhanced inflammatory gene expression by qRT-PCR which correlated with an increase in select areas of open chromatin by assay for transposase-accessible chromatin using sequencing. Further exploration into a specific upregulated gene determined KLF4 was both necessary and important for proper differentiation in mandibular- but not femur-derived cells.

Keywords: craniofacial; inflammation; osteoclasts; sequencing; transcription.

Figure 1

Single-cell RNA sequencing of mouse femur- and mandibular-derived CD11b+ cells. (A) Schematic of bone marrow isolation, CD11b+ cell magnetic selection, and single-cell RNA sequencing of CD11b+ cells from the femur and mandible of 2-mo-old male and female C57BL/6J mice. Six samples total were submitted for sequencing (3 femur and 3 mandible). (B) Uniform manifold approximation of sequenced cells and cluster identification by cell type. (C) Cell percentages found in each cluster. (D) expression of marker genes in each cluster.

Figure 2

Gene set enrichment analysis (GSEA) analysis of pathways differentially regulated in mandibular-derived cell populations compared to femur-derived cell populations in each cell cluster. (A) Hallmark gene set and (B) KEGG gene set.

Figure 3

Heat maps of differentially regulated genes in clusters 0-4. (A) Cluster 0, (B) cluster 1, (C) cluster 2, (D) cluster 3, and (E) cluster 4. Abbreviations: F, femur; M, mandible.

Figure 4

AMELX-positive monocytes and osteoclasts detected in mandibular-derived cells. (A) Heat map of differentially regulated genes in cluster 7. Red arrow indicates presence of Amelx. Abbreviations: F, femur; M, mandible. (B) Amelx expression verification in mandibular-derived monocytes by RT-qPCR. Data shown are from at least 3 independent experiments. Data is graphed relative to Hprt. Samples compared using a Student’s t-test. (C) Amelx expression in mandibular and femur-derived osteoclast precursors stimulated in M-CSF (day 0) or M-CSF and RANKL (days 1 and 3). (D, E) Immunofluorescence staining of AMELX+ osteoclasts at differentiation time point day 2 after RANKL addition. Data shown are from 2 independent experiments. Data is graphed relative to Hprt.

Figure 5

Verification of mandibular-derived differentially expressed genes in monocytes identified in cluster 5. (A) Heat-map of cluster 5 differentially regulated genes in cluster 5 in both femur and mandible. Abbreviations: F, femur; M, mandible. (B-H) Verification of upregulation of monocytic genes, including Egr1 (B), Nr4a1 (C), JunB (D), Klf4 (E), and Ccl3 (F), and downregulation of S100a8 (G) and S100a9 (H) as seen in the single-cell RNA sequencing data. Data shown are from at least 3 independent experiments. Data is graphed relative to Hprt. Samples were compared using Student’s t-test. (I-M) Genes upregulated in mouse mandibular monocytes are also upregulated in human PBMCs. Expression of human genes, including CCL3 (I), JUNB (J), KLF4 (K), EGR1 (L), and NR4A1 (M) was verified by RT-qPCR human CD11b+ cells isolated from PBMCs collected by IV arm line or third molar socket aspiration during third molar extraction surgery. Data shown are from 10 independent patients. Data is graphed relative to GAPDH. Samples were compared using Student’s t-test.

Figure 6

Monocytes in the mandible exhibit increased chromatin accessibility in genes associated with inflammatory processes. (A) Distribution of genomic region annotations. (B) Gene set enrichment analysis (GSEA) of mandible and femur-derived monocytes showing enrichment of genes involved in inflammatory pathways. (C-F) Assay for transposase-accessible chromatin with sequencing peak signal plots for genes Egr1, Nr4a1, Junb, and Klf4. Sequencing peak signal plot derived from femur cells is on the top and sequencing peak signal plot derived from mandible cells is on the bottom. (G-L) Monocytes were transfected with either control or Klf4-targeting siRNA and treated with CMG14-12 conditioned media and receptor activator of NF-kappa B ligand for 4 d. qRT-PCR of control and Klf4 siRNA-treated (G) femur-derived monocytes (H) mandibular-derived monocytes. Representative tartrate-resistant acid phosphatase images of control and Klf4 siRNA-treated (I) femur-derived monocytes and (J) mandibular-derived monocytes. Average size of TRAP⁺ (K) femur-derived (L) mandibular-derived cells. qRT-PCR data is graphed relative to Hprt. Samples were compared using Student’s t-test.