Differential gene expression in skin RNA of horses affected with degenerative suspensory ligament desmitis

Abstract

Background: Equine degenerative suspensory ligament desmitis (DSLD) is a systemic connective tissue disorder first identified in Peruvian Paso horses but afflicting other horse breeds as well. Inappropriate accumulation of proteoglycans in connective tissues, most prominently in tendons and ligaments, leads to progressive and debilitating lameness and pain. It is largely unknown what drives the overproduction of proteoglycans, but our previous studies suggest involvement of bone morphogenetic protein 2 (BMP2), a member of the transforming growth factor-β (TGFβ) family, impacting synthesis of proteoglycans. To identify potential players in pathogenesis of DSLD a new approach utilizing next generation sequencing was undertaken.

Methods: Next generation sequencing was performed using RNA extracted from skin biopsies of six control Peruvian Pasos and six horses with DSLD (4 Peruvian Pasos and 2 warmbloods). The CuffDiff result sets were validated with algorithms used to run them. This was based on the determined false discovery rates derived from the P values adjusted for multiple testing for any given result.

Results: Bioinformatics analysis of transcriptomes revealed differential expression of over 1500 genes, including increased expression of genes for several growth factors (most prominently BMP2, FGF5, CTGF, many members of the EGF family), and mediators of signaling (Fos, Myc, MAPK system), and keratins. Two genes encoding for enzymes involved in synthesis of hyaluronan were also overexpressed. Gene expression was decreased for protein cores of many proteoglycans, several growth factors, most collagens, and many peptides with immune function.

Conclusions: The overexpression of BMP2 correlates well with our previous data. However, the decrease in expression of numerous proteoglycans was unexpected. A mutation in a gene of a less characterized proteoglycan and/or glycosyltransferase with subsequent increased production of hyaluronan and/or a proteoglycan(s) undetected in our study could account for the systemic proteoglycan deposition. Decreased collagen gene expression indicates abnormal connective tissue metabolism. The increased expression of keratin genes and FGF5 supports reports of skin abnormalities in DSLD. Underexpression of immune function genes corresponds with lack of inflammation in DSLD tissues. Finally, though the proteoglycan and/or glycosaminoglycan abundant in DSLD has not been identified, we validated our previous data, including overexpression of BMP2, and systemic nature of DSLD due to disturbed metabolism of the extracellular matrix.

Keywords: BMP2; Collagens; Differential expression of genes; Equine degenerative suspensory ligament desmitis; Keratins; Next generation sequencing; Proteoglycans.

Fig. 1

Principal component analysis: full matrix: PCA analyses of both the full and filtered matrices showed that the samples from control and experimental grouped together well with expected variance among individuals. a This analysis was run using the R-code in file DESeq2. Outlier CTL3 is marked by a black circle. b Using the R-code in file EdgeR this analysis showed very similar results

Fig. 2

PANTHER GO analysis of DEG of (slim) molecular functions in DSLD: The pie chart shows changes in gene expression identified by Cufflinks against genes in the Equus caballus genome database. The following gene categories were evaluated: red, RNA binding (GO:0003723); blue, calcium ion binding (GO:0005509); green, metal ion binding (GO:0046872); yellow, molecular function regulator (GO:0098772); orange, protease binding (GO:0002020); violet, protein binding (GO:0005515); light blue, signaling receptor binding (GO:0005102); pink, single-stranded RNA binding(GO:0003727)

Fig. 3

Expression of growth factor genes relevant to DSLD. Genes differentially expressed in DSLD were identified based on log fold change of 2

Fig. 4

Expression of selected genes encoding for signaling mediators. Genes differentially expressed in DSLD were identified based on log fold change of 2

Fig. 5

Expression of genes for core proteins of proteoglycans and related molecules. Genes differentially expressed in DSLD were identified based on log fold change of 2

Fig. 6

Expression of collagen and keratin genes. Genes differentially expressed in DSLD were identified based on log fold change of 2

Fig. 7

Expression of selected ECM genes. Genes differentially expressed in DSLD were identified based on log fold change of 2

Fig. 8

Expression of genes for selected molecules of immune function. Genes differentially expressed in DSLD were identified based on log fold change of 2