Blood-based signatures in type 1 diabetes

Abstract

Type 1 diabetes mellitus is one of the most common chronic diseases in childhood. It develops through autoimmune destruction of the pancreatic beta cells and results in lifelong dependence on exogenous insulin. The pathogenesis of type 1 diabetes involves a complex interplay of genetic and environmental factors and has historically been attributed to aberrant adaptive immunity; however, there is increasing evidence for a role of innate inflammation. Over the past decade new methodologies for the analysis of nucleic acid and protein signals have been applied to type 1 diabetes. These studies are providing a new understanding of type 1 diabetes pathogenesis and have the potential to inform the development of new biomarkers for predicting diabetes onset and monitoring therapeutic interventions. In this review we will focus on blood-based signatures in type 1 diabetes, with special attention to both direct transcriptomic analyses of whole blood and immunocyte subsets, as well as plasma/serum-induced transcriptional signatures. Attention will also be given to proteomics, microRNA assays and markers of beta cell death. We will also discuss the results of blood-based profiling in type 1 diabetes within the context of the genetic and environmental factors implicated in the natural history of autoimmune diabetes.

Keywords: Biomarker; Gene expression profiling; Innate immunity; Microarray; Review; Type 1 diabetes.

Fig. 1

Longitudinal analyses of diabetes progression or non-progression among HRS of type 1 diabetes probands with integration of multiple biomarkers. (a, d, g, j) Analysis of a type 1 diabetes progressor. (b, e, h, k) Analysis of an autoantibody-positive non-progressor. (c, f, i, l) Analysis of an autoantibody-negative non-progressor. (a–c) Heat maps illustrating plasma-induced transcription. Expression levels of the 307 transcripts annotated as being inflammatory and 1,067 transcripts annotated as being regulatory genes that were identified in [78] and used for temporal scoring of signatures. (d–f) Temporal plot of the composite inflammatory index (I.I._com) derived from the ontology-based scoring of the plasma-induced signature analysis described in [78]. This measure reflects the average induction of inflammatory transcripts vs the average induction of regulatory transcripts. (g–i) Temporal analysis of autoantibody titres (as described in [109]), plotted on a log₁₀ scale as fold change relative to the upper limit of normal controls. Blue shaded area <1 = antibody-negative. Blue line, GAD; red line, IA2; green line, IAA; purple line, ZnT8. (j–l) Longitudinal analyses of the percentage of activated Tregs among the total Tregs in the periphery. As described in [110], resting and activated CD4⁺ Tregs were respectively defined as CD45RA⁺/FoxP3^low and CD45RA⁻/FoxP3^high. The relationship between percentage activated Tregs and age was not significant in the type 1 diabetes progressor illustrated in (j) (slope =0.02, p=0.19, R²=0.26); but was found significant in the non-progressors illustrated in (k) (slope =0.11, p=0.01, R²=0.90) and (l) (slope =0.03, p=0.04, R²=0.62).

Fig. 2

Models of type 1 diabetes progression or non-progression. (a, d) Disease progression of an HRS, equivalent to the individual illustrated in Fig. 1a, d, g, j. (a–c) A nonlinear decay in beta cell mass following environmental triggering of autoimmunity in a genetically susceptible individual (derived from the model originally proposed by Eisenbarth [111] and subsequently modified [90, 112]). (d–f) The immune state in terms of proinflammatory activity (black line) or immune-regulatory activity (blue line) reflected by the plasma-induced signatures illustrated in Fig. 1. Shading reflects overall bias captured by I.I._com (green, regulatory; red, inflammatory). (b, e) Disease non-progression in an antibody-positive high HRS of a type 1 diabetes proband, equivalent to the individual illustrated in Fig. 1b, e, h, k. This individual experienced disruption followed by restoration of the balance between inflammatory and regulatory processes. (c, f) Disease non-progression in an antibody-negative HRS of a type 1 diabetes proband, equivalent to the individual illustrated in Fig. 1c, f, i, l. Our studies support the existence of an endogenous elevated innate inflammatory state in type 1 diabetes families that is counter-regulated in an age-dependent manner resulting in a state of regulated susceptibility. Prior to the generation of a robust immunoregulatory compartment (reflected by green shading in lower panel), there is a window of susceptibility of (viral) disease triggering, which temporally diminishes