**4. Proteomics: A powerful tool for predictive medicine**

The role of proteomic technologies in the study of autoimmune diseases can hardly be overestimated. Virtually all currently known autoimmune diseases including diabetes mellitus, multiple sclerosis, systemic lupus erythematosus and other severe autoimmune disorders have proteomic markers of their own. At the same time, the advent of efficient high-precision diagnostic technologies opened up new opportunities in the search for novel preclinical diagnostic markers. Identification of autoAbs to immunoglobulins GADA, IAA, ICA, IA-2A and ZnT8 has become a routine procedure in T1D diagnosis. In this chapter, the main emphasis will be laid on some characteristics of specific proteins for progressive T1D.

Clusterin (apolipoprotein J) holds considerable promise as a candidate biomarker; its main function is traditionally recognized as a tool to control apoptosis. It should be noted, however, that clusterin exhibits the behaviour of an antiapoptotic chaperone when used at low concentrations; at higher concentrations ( 12% of control), it causes disruption of mitochondria and initiates cell apoptosis by a mitochondrial mechanism. Recent reports highlighted a high regenerative potential of Clusterin, particularly, with respect to beta cells. The functional activity of this protein demands further verification and analysis, but its elevation always points to apoptosis of pancreatic beta cells. (Lee et al., 2011)

Transcortin (Corticosteroid-binding globulin, CBG) and Lumican are capable to induce pronounced (1.5–2-fold against control) upregulation. Transcortin fulfils the function of a glucocorticoid transporter and is strongly inhibited by insulin; hence, insulin deficiency is always associated with hyperproduction of Transcortin. (Fernández-Real et al., 1999) However, this protein is hardly effective as a selective biomarker for T1D, since it emerges exclusively at the latest stages of autoimmune aggression and its role in T1D etiogenesis is

Preclinical and Predictive Algorithms in Monitoring

I quartile 35 ± 9 1\*

MID COOH NH2 67

+ − + − 1 (0)

− + + + 1 (0)

relatives are not shown). (Buzzetti et al., 2007; Mayr et al., 2007)

antibodies

**5.2 IA-2** 

pathogenesis of T1D.

the first clinical manifestations of the disease.

years before the clinical stage of T1D.

Patients with Autoimmune Diseases and Their Relatives-at-Risks 205

Likelihood of developing diabetes depending on the combination of epitope-specific GADA

GADA

+ + + + 24 (8) 39% + + + − 12 (6) 26% + + − + 15 (5) 56% + + − − 62 (21) 45%

+ − − − 13 (3) 27%

− + − − 5 (2) 48% − − + − 6 (1) 20% − − − − 10 (3) 33% For GAD antibodies, MID refers to epitopes within GAD65 amino acids 235–442, COOH refers to epitopes within GAD65 amino acids 436–585, NH2 refers to epitopes within GAD65 amino acids 1–100, while GAD67 refers to epitopes present in GAD6 (combinations with no

IA-2 belongs to type 1 membrane-bound proteins containing extracellular NH2-terminal glycosylated, membrane-bound and СООН-terminal cytoplasmic regions. The immune epitope of IA-2 is localized exclusively in the cytoplasmic region of IA-2 where its PTP (protein-tyrosine-phosphate)-like domain is the main recognition site for auto-Abs. The fact that the dominant T cell epitope of IA-2, also localized in the PTP-like region, is structurally similar to the VP7 region (VP7 is a major immunogenic protein of rotaviruses) provides additional evidence for the crucial role of the molecular mimicry mechanism in the

IA-2β (also known as fogrin, PTP-NP, ICAAR and IAR) is structurally similar to IA-2. Its intracellular and extracellular domains are structurally identical (by 74 and 26%, respectively) to IA-2. IA-2β is predominantly localized in secretory vesicles of beta and some other neuroendocrine cells. Anti-IA-2β Abs are present in nearly 50% of patients with newly diagnosed T1D; their emergence is usually recorded several years before the appearance of

AutoAbs against the insulin antigen are detected in 70–90% of individuals as early as 10 – 12

Epitope combination n (T1D cases) 10-year risk

Autoantibody titer 10-year risk (% ± SE) HR (95% CI) P

II quartile 22 ± 9 0.8 (0.3–1.9) 0.55 III quartile 52 ± 9 1.7 (0.8–3.6) 0.18 IV quartile 43 ± 10 1.6 (0.7–3.6) 0.26

still unclear. Lumican, the key mediator in fibrosis-related processes, manifests an even higher degree of upregulation than Transcortin and is widely distributed in all body tissues. Its significant elevation may represent an acute response of renal tissues to high plasma levels of glucose and is also characteristic of nephropathies. However, being a convenient tool for predicting diabetic nephropathies, Lumican cannot predict associated diseases.

It may be concluded from the above-said that Clusterin is the only candidate for a selective protein biomarker for T1D, because it emerges at early stages of the disease and is more related to cause than effect.
