Figure 1.
Mechanical loading was applied to the rat forelimb and a central region of the ulna was analyzed.
(A) Medial view of bones in a right forelimb of a rat obtained by microCT during simulated loading (Reprinted from Journal of Biomechanics, 40, Uthgenannt BA & Silva MJ, 317–324, 2007, with permission from Elsevier). (B) The central 5 mm of the ulna and surrounding periosteum were isolated for microarray analysis. (C) Representative transverse histological sections from a previous study [5] that illustrate bone formation after loading. WBF loading leads to woven bone formation while LBF loading increases lamellar bone formation. After loading, fluorochrome labels were injected in vivo on days 3 (green) and 8 (red) prior to animal sacrifice on day 10. Plastic embedded transverse sections were taken 1 mm distal to the ulna midpoint.
Table 1.
Loading parameter summary for the 48 rats used in the study.
Figure 2.
Flow chart describing the experiment and analysis.
Three separate groups of animals were used to provide samples for WBF, LBF and Non-Loaded groups. Thus, comparisons between loading conditions are between (not within) animals.
Table 2.
Comparisons (nine total) of the number of DEGs between groups at 1 hr, day 1 and day 3; the woven vs. lamellar DEGs were further analyzed using GeneGo software.
Table 3.
Relative fold changes (loaded over normal) for gene expression analysis done using qRT-PCR.
Figure 3.
A Venn diagram depicts the commonality of differentially expressed genes.
Genes were differentially regulated at 1 hr, 1 and 3 days post-loading for WBF loading (woven) vs. LBF loading (lamellar) groups from microarray analysis.
Figure 4.
The total number of differentially regulated genes on the top 10 canonical pathway maps.
GeneGo enrichment analysis was completed for woven vs. lamellar comparisons at each of the three timepoints investigated. The analysis resulted in a top 10 list of statistically significant canonical pathway maps associated with our data. On each pathway, genes were counted if their expression was significantly different between woven and lamellar bone formation. The number of DEGs on each map (30 maps total) were counted and summed at each timepoint for both upregulated and downregulated genes separately. A total of six pathways were represented at multiple timepoints, although the data were only counted once.
Figure 5.
Percentage distribution of statistically significant pathway map folders comparing woven to lamellar bone formation at (A) 1 hr, (B) day 1 and (C) day 3.
GeneGo software was used to identify the distribution of the data as mapped onto major functional categories of human metabolism and cell signaling at each timepoint. Out of the 37 available GeneGo pathway map folders, 25 were statistically significant (p<0.05) for at least one timepoint (Table S1). The statistically significant pathway map folders were grouped by the authors for before being displayed in the pie chart. The smaller size of the pie chart at 1 hr reflects the lower number of differentially expressed genes. Within the GeneGo pathway map folders there were 486 differentially expressed genes at 1 hr, 3699 genes at day 1 and 3046 genes at day 3. A complete listing of activated maps is available in Table S2.
Figure 6.
qRT-PCR relative expression (delta CT) of genes related to (A) inflammation, (B) angiogenesis and (C) osteogenesis and matrix remodeling.
Inflammatory genes were upregulated at all timepoints for woven bone with the exception of Nfkbia on day 1. In contrast, only Nfkbia at 1 hr was upregulated for lamellar bone. The inflammation marker Il6 is increased over 500-fold 1 hr after damaging loading. Both angiogenic markers Sele and Ptgs2/Cox2 are positively related to increased vasculature with expression peaking on day 1 for woven and lamellar bone. In contrast, Cxcl10 is angiostatic and peaks on day 3 for woven bone but is not significant at any timepoint for lamellar bone. Sost, a bone formation inhibitor is downregulated at all timepoints for woven bone. Mmp13 and Ctsk are markers of bone remodeling and both are upregulated at later timepoints for woven bone. In contrast, Mmp13 is downregulated for lamellar bone. *p<0.05 vs. normal.
Figure 7.
In combination with our previous report [5], we have created an overview of the early molecular response comparing woven to lamellar bone formation.
There is an early immune response that persists through time but tends to decrease in expression. The vascular response is also a major component of woven bone formation and it precedes osteogenesis. Osteogenic indicators are differentially regulated shortly after loading, but seem to increase over time. Finally, bone remodeling markers are activated later, possibly to repair bone damage. Gene expression changes likely persist through many weeks after loading, but our data only includes early timepoints after loading.