The identification of gene expression profiles associated with progression of human diabetic neuropathy

Abstract

Diabetic neuropathy is a common complication of diabetes. While multiple pathways are implicated in the pathophysiology of diabetic neuropathy, there are no specific treatments and no means to predict diabetic neuropathy onset or progression. Here, we identify gene expression signatures related to diabetic neuropathy and develop computational classification models of diabetic neuropathy progression. Microarray experiments were performed on 50 samples of human sural nerves collected during a 52-week clinical trial. A series of bioinformatics analyses identified differentially expressed genes and their networks and biological pathways potentially responsible for the progression of diabetic neuropathy. We identified 532 differentially expressed genes between patient samples with progressing or non-progressing diabetic neuropathy, and found these were functionally enriched in pathways involving inflammatory responses and lipid metabolism. A literature-derived co-citation network of the differentially expressed genes revealed gene subnetworks centred on apolipoprotein E, jun, leptin, serpin peptidase inhibitor E type 1 and peroxisome proliferator-activated receptor gamma. The differentially expressed genes were used to classify a test set of patients with regard to diabetic neuropathy progression. Ridge regression models containing 14 differentially expressed genes correctly classified the progression status of 92% of patients (P < 0.001). To our knowledge, this is the first study to identify transcriptional changes associated with diabetic neuropathy progression in human sural nerve biopsies and describe their potential utility in classifying diabetic neuropathy. Our results identifying the unique gene signature of patients with progressive diabetic neuropathy will facilitate the development of new mechanism-based diagnostics and therapies.

Figure 1

Primary and secondary biopsy selection. Primary and secondary biopsies of 36 patients with diabetic neuropathy (DN) were included in this study. Samples with a minimum RNA integrity number (RIN) of 6.5 were used for microarray hybridization. Microarrays of 12 primary (five progressors and seven non-progressors) and 35 secondary (18 progressors and 17 non-progressors) samples were used for differential gene expression analysis and diabetic neuropathy progression modelling. btw = between; P = progressor; NP = non-progressor.

Figure 2

A network of over-represented biological concepts identified by ConceptGen. The concepts (gene sets) over-represented in the upregulated genes (A) and downregulated genes (B) in progressors. The centre nodes in violet, titled as ‘GP-CI-Common-SP-SN…’ refer to the differentially expressed genes. DEGs = differentially expressed genes; MeSH = Medical Subject Headings.

Figure 3

Gene co-citation network of differentially expressed genes generated by BiblioSphere. A literature-derived gene network of the differentially expressed genes was created by BiblioSphere using sentence level co-citations of differentially expressed genes. The network is composed of five subnetworks centred on the five most connected genes: JUN, PPARγ, LEP, SERPINE1 and APOE.

Figure 4

Gene co-citation network clustered by fast-greedy community structuring algorithm. The complete co-citation network of the differentially expressed genes was clustered based on network topology by Fast Greedy algorithm implemented in the Cytoscape GLay plug-in (Supplementary Fig. 1). The figure contains only the nodes with a minimum of five connections (edges) in the complete network. The size of node represents the number of edges. Nodes (genes) highlighted in red or yellow refer to the highly connected genes; red for the central genes in the initial BiblioSphere co-citation network (Fig. 3) and yellow for additionally identified highly connected genes in the GLay-identified clusters.

Figure 5

Biopsy selection for ridge regression modelling of diabetic neuropathy. The secondary samples, excluding those samples with paired primary samples, were used as the training set (13 progressors and 11 non-progressors) for diabetic neuropathy classification modelling, and the primary samples were used as the testing set (five progressors and seven non-progressors). P = progressor; NP = non-progressor.