Comparison of inbred mouse strains shows diverse phenotypic outcomes of intervertebral disc aging

Abstract

Intervertebral disc degeneration presents a wide spectrum of clinically degenerative disc phenotypes; however, the contribution of genetic background to the degenerative outcomes has not been established. We characterized the spinal phenotype of 3 mouse strains with varying cartilage-regenerative potential at 6 and 23 months: C57BL/6, LG/J and SM/J. All strains showed different aging phenotypes. Importantly, LG/J mice showed an increased prevalence of dystrophic disc calcification in caudal discs with aging. Quantitative-histological analyses of LG/J and SM/J caudal discs evidenced accelerated degeneration compared to BL6, with cellular disorganization and cell loss together with fibrosis of the NP, respectively. Along with the higher grades of disc degeneration, SM/J, at 6M, also differed the most in terms of NP gene expression compared to other strains. Moreover, although we found common DEGs between BL6 and LG/J aging, most of them were divergent between the strains. Noteworthy, the common DEGs altered in both LG/J and BL6 aging were associated with inflammatory processes, response to stress, cell differentiation, cell metabolism and cell division. Results suggested that disc calcification in LG/J resulted from a dystrophic calcification process likely aggravated by cell death, matrix remodelling, changes in calcium/phosphate homeostasis and cell transformation. Lastly, we report 7 distinct phenotypes of human disc degeneration based on transcriptomic profiles, that presented similar pathways and DEGs found in aging mouse strains. Together, our results suggest that disc aging and degeneration depends on the genetic background and involves changes in various molecular pathways, which might help to explain the diverse phenotypes seen during disc disease.

Keywords: Aging; LG/J; SM/J; calcification; intervertebral disc; transcriptome.

FIGURE 1

LG/J mice present high prevalence of age‐dependent disc calcification. (a‐c) μCT showed higher prevalence and widespread distribution of disc calcification in caudal spine of 23M LG/J mice. Yellow, orange and red arrows (a) and colour scale (c) represent small, medium and large size calcifications, respectively. p ≤ .001***, χ² test, BL6 (n = 10), LG/J (n = 14) and SM/J (n = 8). (d‐d’) Disc calcifications in LG/J mice were present in both NP and AF. (e‐e’) Alizarin Red staining showed staining of the disc, not limited to the calcified nodules. (f) FTIR analysis showed comparable spectral profile between disc mineralization and vertebrae. Green arrow—amide peak – 1,665/cm; red arrow—phosphate peak – 960/cm; purple arrow—carbonate peak – 870/cm; *phosphate peak—1,030/cm; ** tape artefact. (g‐h’) Phosphate and carbonate levels in the disc were comparable to the vertebrae. (i‐i’) Phosphate to protein ratio was higher in the disc. (j) Plasma measurements of calcium, phosphorous, TNAP, fetuin, glucose, creatine, sodium and albumin from 6M and 23M mice. Mann–Whitney test was used for comparing differences between the groups. ns = not significant; p ≤ .05*; p ≤ .01**; p ≤ .01**; BL6 (n = 10), LG/J (n = 14) and SM/J (n = 8). Scale bar d‐e’ = 200 µm

FIGURE 2

Inbred strains show distinct disc phenotypes during aging. (a‐f’’) Histology of 6M and 23M mice showed differences in tissue architecture and cell morphology in the NP, AF and EP. Panels d', e' and f' show the area enclosed in dotted boxes in panels d, e and f, respectively. (g‐i) Modified Thompson Grading distribution and averages showed small degenerative changes in LG/J mice and BL6 but high degenerative grades in SM/J mice at 6M. At 23M, both LG/J and SM/J mice showed higher degenerative grades than BL6 mice in the NP. (j‐m) Picro‐Sirius Red staining and polarized microscopy of discs (n‐o) Analysis of percentage of thin (green), intermediate (yellow) and thick fibres (red). Mann–Whitney test was used for comparing 6M to 23M in each strain. Kruskal–Wallis and chi‐squared test were used to perform comparison between the 3 strains. BL6 (n = 10), LG/J (n = 14) and SM/J (n = 8), 4 levels per strain were analysed. Scale bar A‐F and J‐L’= 200 µm; Scale bar d’‐f = 50 µm. AF, annulus fibrosus; EP, endplate; GP, growth plate; NP, nucleus pulposus; VB, vertebra

FIGURE 3

Aged inbred mouse strains evidenced changes in matrix composition. Staining and abundance in NP and AF compartments of important extracellular matrix proteins in 23M BL6, LG/J and SM/J caudal discs (a‐a’’’’) collagen I, (b‐b’’’’) collagen II, (c‐c’’’’) collagen X, (d‐d’’’’) aggrecan, (e‐e’’’’) chondroitin sulphate and (f‐f’’’’) ARGXX neoepitope. ANOVA or Kruskal–Wallis test was used as appropriate; n = 6 mice/strain, 2–3 levels per mouse were analysed. Scale bar A‐F’’=50 µm

FIGURE 4

SM/J and LG/J showed divergent transcriptomic profiles at 6M. (a) Schematic summarizing microarray analysis. (b) Transcriptomic profiles of 6M BL6 (n = 7), LG/J (n = 5) and SM/J (n = 6) mice clustered distinctly along principal components. (C) Venn diagram from DEGs, Fold Change ≥ 2, FDR ≤ 0.05, (d) volcano plot, showing up‐ and downregulated DEGs from LG/J versus BL6 comparison, used for GO process enrichment analysis, (e) representative GO processes of upregulated genes in LG/J versus BL6, (f) representative GO processes of downregulated genes in LG/J versus BL6. (g) Volcano plot, showing up‐ and downregulated DEGs from SM/J versus BL6 comparison. (h) Representative GO processes of upregulated genes in SM/J versus BL6. (i) Representative GO processes of downregulated genes in SM/J versus BL6. GO analysis was performed using PANTHER overrepresentation test, GO database annotation, binomial statistical test with FDR ≤ 0.05. Representative genes from select GO processes different between (j) LG/J versus BL6 and (k) SM/J versus BL/6

FIGURE 5

BL6 and LG/J mice comparison uncovers biological pathways that are independent and dependent on the strain background. (a) Transcriptomic profiles of 6M BL6 (n = 7), 23M BL6 (n = 7), 6M LG/J (n = 5) and 23M LG/J (n = 5) clustered distinctly along principal components. (b) Venn diagram of upregulated DEGs from BL6 23M versus 6M and LG/J 23M versus 6M, FDR ≤ 0.05. (c) Representative GO processes of the common upregulated genes between BL6 23M versus 6M and LG/J 23M versus 6M. (d) Venn diagram of downregulated DEGs from BL6 23M versus 6M and LG/J 23M versus 6M, FDR ≤ 0.05. (e) Representative GO processes of the common downregulated genes between BL6 23M versus 6M and LG/J 23M versus 6M. (f) Volcano plot of DEGs from LG/J 23M versus BL6 23M, used for GO Process enrichment analysis, Fold Change > 2, FDR ≤ 0.05. (g) Representative GO processes of upregulated genes in LG/J 23M versus BL6 23M. (h) Representative GO processes of downregulated genes in LG/J 23M versus BL6 23M. (i‐k) Representative DEGs from selected GO processes from 23 M LG/J versus 23M BL6. (l) TUNEL (arrows) and Ki67 staining of 23M LG/J and BL6 discs. t test or Mann–Whitney test was used as appropriate. n = 6 mice/strain were analysed

FIGURE 6

Degenerated human discs show distinct transcriptomic profiles and cluster 1 and 7 present similar enriched pathways with LG/J 23M. (a) Schematic showing PCA clustering of http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE70362 deposited microarray data based on histological grades (b) Hierarchical clustering of DEGs, p ≤ .05, showed 7 degenerative groups—red boxes, and 3 healthy groups—green boxes, with dissimilarly < 0.5 cut‐off Euclidean distance. (c) PCA of clusters obtained in (B). (d) Representative GO processes of upregulated genes between cluster 1 (degenerated) and cluster A (healthy), FDR ≤ 0.1. (e) Representative GO processes of upregulated genes between cluster 7 (degenerated) and cluster A (healthy), FDR ≤ 0.1 (f) Representative GO processes of downregulated genes between cluster 1 and cluster A FDR ≤ 0.1. (g) Representative GO processes of downregulated genes between cluster 7 and cluster A, FDR ≤ 0.1 (h) Representative DEGs in comparisons from cluster 1 versus cluster A and cluster 7 versus cluster A (i) Representative table of the cluster details used for enrichment analysis