Proteomic analysis to identify candidate biomarkers associated with type 1 diabetes

Introduction

Type 1 diabetes mellitus (DM1) is characterized by the destruction of insulin-producing β-cells in the pancreas, with consequent insulin deficiency.¹ It is a metabolic and endocrine condition more common in childhood development and is preceded by the dysregulation of many biological pathways.² At present, no therapy exists to halt the immune-mediated destruction of β-cells,³ and DM1 has a substantial long-term impact on the quality of life, particularly in children and adolescents.

According to the International Diabetes Federation,⁴ the total population with DM was equivalent to 415 million and is estimated to increase to 471 million by 2035. A recent report of World Health Organization estimated that, globally, 422 million adults aged >18 years were living with diabetes in 2014, compared with 108 million in 1980.⁵ Following the same trend, the incidence of DM1 has increased by ~3% worldwide and represents 5%–10% of the entire population with diabetes.⁶

There has been a lot of research conducted on diabetes over the decades. Several proteomic approaches have been undertaken to identify biomarkers for monitoring the prediction^7,8 and progression⁹ of the disease. However, there are some limitations due to low specificity or late application to the disease stage.¹⁰

DM1 is a complex disease with altered expression of many genes and their products.¹¹ Therefore, different approaches were developed using current proteomic technologies to explore its clinical potential and enable the discovery of new biomarkers for DM1.

Recent advances in proteomics have triggered rapid progress in mass spectrometry (MS)-based proteomics. It enables more than simple protein identification and can accurately and reliably quantitate differences in protein abundance in an organism, a cell, or a tissue at a given time or under a particular condition.¹² The label-free technique has been shown to be more accurate in estimating absolute abundance of proteins in complex samples.¹³

In the current study, we performed a quantitative and qualitative analysis using a sample pretreatment immunodepletion of serum albumin and immunogloblin G (IgG) combined with data-independent label-free mass spectrometry (MS^E) analysis with the aim of identifying differentially expressed levels of serum proteins in DM1 patients and healthy controls (HCs).

Materials and methods

Study participants

The present study enrolled DM1 patients selected among those in a regular follow-up in the outpatient of Endocrinology and Diabetes at the Federal University Hospital of Ceara, Fortaleza, Brazil. DM1 patients aged ≥18 years, both male and female, who had attended the outpatient department of Endocrinology and Diabetes for a minimum of 6 months, with clinical and cognitive conditions to answer an interview and the capacity to stand up normally were eligible for the DM1 group. Patients using immunosuppressive medications (renal, heart, or hepatic dysfunction; type 2 diabetes; etc.); those with alcoholism and mental disorders (including dementia and Alzheimer’s disease); those diagnosed with leprosy or HIV infection; and those with autoimmune disease, amputation or ulcers on the lower limbs, acute infectious frames, or diabetic ketoacidosis in the 3 months preceding the evaluation, as well as pregnant or lactating females, were excluded. HCs were selected from the Federal University and Ceará State University, considering the same criteria used for DM1 patients, besides not having a familiar history of diabetes and with fasting blood glucose ≤99 mg/dL.

The study initially enrolled all patients attending the outpatient department as potential participants. After applying rigorous inclusion and exclusion criteria to achieve a group with highly homogeneous clinical features and accounting for loss during the selection process, a total of 60 subjects were included as study participants: 30 DM1 patients and 30 HCs. The DM1 and HC groups were matched by gender, age, body mass index (BMI), total cholesterol, high-density lipoprotein cholesterol (HDL-C), serum low-density lipoprotein cholesterol (LDL-C), total protein, urea, and creatinine. General health features, demographics, and clinical data were obtained through medical records and direct interviews. The study protocol (number 1357544) was approved by the Federal University Hospital of Ceará Ethics Committee, and the study was conducted according to the principles of the Declaration of Helsinki. Written informed consent was obtained from all the participants.

Anthropometry and health-related behaviors

Weight and height were measured according to the SISVAN (Food and Nutritional Surveillance System; 2004).¹⁴ BMI was calculated as weight (in kg) divided by height (in m²).¹⁵ The considered standard doses for alcohol consumption and categories were in accordance with World Health Organization.¹⁶ Cigarette smoking status was consistent with that shown in Vital and Health Statistics.¹⁷

Serum samples

Blood samples were collected from all subjects after an overnight fast (12 hours) for biochemical analysis. All the samples were centrifuged at 3000 rpm for 15 minutes at room temperature, and the serum obtained was kept frozen at −80°C until analyzed. Fasting blood sugar, total cholesterol, HDL-C, triglyceride, urea, and creatinine were measured by colorimetric enzymatic method using commercially available kits (Bioclin, Quibasa Quimica Basica Ltda, Belo Horizonte, Minas Gerais, Brazil) and an automated biochemical analyzer (Mindray BC-2800; Shenzhen Mindray Bio-Medical Electronics Co. Ltd., Shenzhen, China). LDL-C was calculated using the Frederickson–Friedewald equation.¹⁸ C-peptide analysis was performed by immunoassays of microparticles by quimioluminescence (Kit ARCHITECT C-Peptide) using Architec Asystema no. i2000 (Chicago, IL, USA). Samples were analyzed in triplicate.

The serum protein concentration was determined by the Bradford method.¹⁹ The serum samples were pooled in the same group for proteomic analysis. Using 5.33 μg of protein from each individual sample, 15 samples were pooled to form each group of patients with diabetes and the HC groups. Therefore, two pools of samples from patients with DM1 (DM1a and DM1b) and two pools of HCs (HCa and HCb) were generated.

Immunodepletion of high-abundance proteins

In order to improve the chance of detecting a wide range of serum proteins of medium to low abundance, proteins of high abundance, albumins, and immunoglobulins were depleted from the serum. Aliquots of serum were filtered through 0.22-µM membrane (KASVI, São José do Pinhais, Brazil). Then, 150 µL of filtered serum was applied to a HiTrap® Albumin & IgG Depletion column (GE Healthcare Bio-Sciences Corp., Piscataway, NJ, USA) attached to an AKTA Purifier 10 fast protein liquid chromatography system (GE Healthcare). The column was preequilibrated with a solution of 20 mM Tris–HCl (pH 7.4) and 0.15 M NaCl. The elution of fraction rich in albumin and immunoglobulin was realized with 0.1 M Glycine–HCL buffer (pH 2.7) delivered at 1 mL/min, and absorbance was monitored at 216 and 280 nm.

Nano-ultra performance liquid chromatografy MS analysis

After immunodepletion, the fractions of the pools were dialyzed, concentrated, and quantified using a Vivaspin™ (GE Healthcare) and a NanoVue™ Spectrometer (GE Healthcare), respectively, at an absorbance at 280 nm. The denatured samples (100 µg) were diluted in 50 mM ammonium bicarbonate, denatured with a solution of 0.2% RapiGest™ SF (Waters Corporation, Milford, MA, USA) at 80°C for 15 minutes in a dry bath, and reduced with 100 mM dithiothreitol at 60°C for 60 minutes, and alkylation with 300 mM iodoacetamide for 30 minutes was performed in the dark at room temperature. The digestion occurred with the addition of 1 µg of modified trypsin (Promega, Madison, WI, USA) to each sample at 37°C, and the samples were incubated overnight. The reaction was stopped by adding 10 µL of 5% of trifluoroacetic acid, the samples were incubated for 90 minutes at 37°C, centrifuged, and the supernatant was transferred to vial (Waters, Manchester, UK). Tryptic peptides of yeast alcohol dehydrogenase (ADH) were added to vials containing digested proteins to a final concentration of 50 fmol/µL as internal standards. The final protein concentration was ~1 µg/µL.

The tryptic peptides were separated using a nanoACQUITY UPLC® system (Waters) equipped with an HSS T3 C18 reverse-phase column (1.8 µm, 75 µm×20 mm; Waters) for 110 minutes using 0%–40% gradient for 90 minutes and 40%–85% for 5 minutes. The column was reequilibrated for 15 minutes at 35°C. The flow rate was 0.35 µL/min, and mobile Phases A and B contained 0.1% formic acid in water and 0.1% formic acid in acetonitrile, respectively. All the samples were measured in triplicate. MS^E was performed using a Synapt HDMS (nano-electrospray ionization quadrupole /orthogonal acceleration time-of-flight mass spectrometer; Waters, Manchester, UK). The equipment was operated in “V” mode with a precursor double-charge resolution ≥10,000 full width at half maximum and in electrospray positive-ion mode nano-electrospray ionization (+). Mass spectrometer data were acquired using a NanoLockSpray probe channel infusion of Glu-fibrinogen peptide (Glu-Fib) derived from fibrinopeptide B human (M + 2H)² and final calibration of the equipment used MS/MS Glu-Fib fragments.

Data processing and protein identification and quantification

Liquid chromatography (LC)/MS^E data were processed, and the proteins were identified using the ProteinLynx Global Server (PLGS) version 2 software with the UniProt reverse Homo sapiens annotated database. For searching spectra and the database, the default parameters of PLGS were used, followed by a maximum of one missed trypsin cleavage. The absolute quantification of each run was calculated according to the three most intense peptides (label-free Hi3 method) using ADH peptides as internal standards.¹³ Relative quantification of identified proteins from each expression group was performed using alpha-1-antitrypsin (UniProt) as an internal common housekeeping protein to normalize expression levels, using PLGS Expression^E software (Waters).²⁰ The average quantitative values of all the samples were calculated, and the p-values (<0.05) were calculated using Expression^E software to refer to the differences between biological replicates.

Gene ontology analysis

The identified proteins were classified into different categories, as biological processes, cellular location, and molecular function in accordance with the information found in the Gene Ontology database, UniProtKB, and EBI GOA.

Statistical analysis

The data obtained concerning demographics and biochemical and anthropometric parameters were analyzed using R software Version 3.2.2 (2017). For the comparison of two means, Student’s t-test was used for independent samples; otherwise, Mann–Whitney U test was used. Correlations between variables were tested using Pearson and Spearman correlation tests. Data were considered significant with p-values <0.05. Figure 1 presents the experimental strategy used in this study.

Figure 1 Diagram of the study design used for research of proteomic biomarker in the serum of individuals. Abbreviations: LC–MSE, liquid chromatography––data-independent label-free mass spectrometry; IgG, immunoglobulin G.

Results

Clinical features

A total of 60 individuals with an average age of 33.30±9.80 years (ranging from 18 to 59 years) were studied and grouped into 30 DM1 patients and 30 HC individuals. Table 1 shows the biochemical and anthropometric characteristics of the patients enrolled for the study. DM1 and HC groups were matched, and there were no significant differences between the groups regarding age, gender, and BMI or in the serum concentrations of total cholesterol, HDL-C, LDL-C, total proteins, and creatinine. As expected, fasting blood glucose and C-peptide were significantly different between DM1 patients and HCs (p<0.001). The DM1 group also exhibited a negative correlation between C-peptide and duration of disease (r=−532; p<0.001). Moreover, no significant differences were observed between the groups relative to alcohol consumption and cigarette smoking status.

Table 1 Demographic, anthropometric, and clinical characteristics of participants in study Notes: Values are expressed as mean (SD) and number (percentage). aStudent’s t-test; bMann–Whitney test; p-values <0.05 are considered significant. Abbreviations: BMI, body mass index; DM1, type 1 diabetes mellitus; HbA1c, glycated hemoglobin; HC, healthy controls; HDL, high-density lipoprotein; LDL, low-density lipoprotein; TG, triglycerides.

Expression of proteins

A comparative proteomic analysis was performed between DM1 and HC groups to investigate alterations in serum proteins in DM1 patients. In the current study, a total of 130 proteins were identified in the serum of the DM1 and HC groups in the total data set (Table S1). Eight proteins were differentially expressed in the serum of the DM1 group compared with the HC group (six proteins were upregulated and two were downregulated). Of the list of proteins identified, the criteria stated for the differential expression in each of the two biological replicates only considered proteins with differential expression (DM1/HC) with absolute ratios >1.5 (upregulated) and <0.66 (downregulated) and the levels of those proteins with ratios between 1.5 and 0.66 (unchanged). Based on gene ontology analysis, most of the proteins that were identified as related to molecular function were involved in binding (48%), catalytic (12.35%) and antioxidant (4.16%) activities, and other functions. They also participated in biological processes related to regulation (19.81%), cellular process (19.81%), localization (14.12%), stimulus response (9.58%), and immune system (9.75%), among others. In addition, the majority of proteins were located in the extracellular region (Figure 2). Table 2 shows the proteins identified as differentially expressed in the serum from DM1 group compared with HC group.

Figure 2 Percentage of serum proteins identified from albumin and immunoglobulin G depleted serum by LC–MSE for patients with type 1 diabetes and healthy controls based on their molecular function, biological process, and cellular components. Abbreviations: ER, endoplasmic reticulum; LC–MSE, liquid chromatography––data-independent label-free mass spectrometry.

Table 2 Characterization of differentially expressed proteins in serum of DM1 patients compared with HCs as identified by LC–MSE Notes: aAverage differences of replicate samples run to type 1 diabetes patients compared with healthy control. Proteins are shown as 1.5-fold differences either upregulated or downregulated (p<0.05). Functions were assigned according to the String databases and literature search. Abbreviations: AMBP, alpha-1-microglobulin; DM1, type 1 diabetes mellitus; HC, healthy control; LC–MSE, liquid chromatography–data-independent label-free mass spectrometry.

Discussion

Diabetes occurs as a result of a range of minor changes in protein expression that alters the stability of the health β-cell to a prediabetic phenotype of instability and possibly β-cell destruction.²¹ Thus, the proteomic analysis of protein expression is an important tool for understanding the molecular modifications associated with disease progression and may contribute to the therapeutic monitoring of this autoimmune disease.

In the present study, a proteomic approach using a label-free quantitative MS analysis identified eight differentially expressed proteins in the serum of DM1 patients compared with HCs. All the proteins that were either upregulated (alpha-2-macroglobulin [α2M], apolipoprotein A-II [ApoA-II], β2 glycoprotein I, Ig alpha-2 chain C region [IGHA2], alpha-1-microglobulin [AMBP], and prothrombin [THRB]) or downregulated (pregnancy zone protein [PZP] and complement C4 [C4]) are discussed below.

α2M belongs to the α-M family of proteins and is the largest major antiproteinase present in the plasma of vertebrates.²² In the present study, α2M expression levels were upregulated in the DM1 group compared with HCs. In support of our data, Takada et al also showed that serum levels of α2M were clearly upregulated in many individuals with DM1 and type 2 diabetes mellitus (DM2).²³ Interestingly, upregulation of α2M has been suggested as a potential biomarker for diabetic retinopathy and nephropathy,^23–25 and studies have indicated that α2M can also act as a binding and carrier protein for cytokines, including IL-Iβ and IL-6, among others,²⁶ leading to the induction of several defense reactions, such as immune response, acute phase reaction, and hematopoiesis.²² Therefore, acting as a carrier and binding protein could be the mechanism by which α2M performs its immune functions, acting in inflammatory processes, and could explain its enhanced expression in the serum of people with DM1.

Proteomic analysis has shown upregulated expression levels of IGHA2 in DM1 patients. IGHA2 is an immunoglobulin that is a membrane-bound or secreted glycoprotein produced by B lymphocytes.²⁷ This immunoglobulin has an important role in immune defense, especially at the mucosal surfaces.²⁸ In our study, the high levels of IGHA2 identified are in keeping with previous reports that have shown increased concentrations of circulating IgHA2 in DM1 and DM2 patients.^29–31 In DM1 patients, the increased expression in the serum of IGHA2 could be triggered in response to antigenic stimuli, forming continuous stimulation of antibody production.²⁹ Moreover, the literature shows that the relationship between diabetes and IGHA2 has been described as involving advanced glycation products. Some studies reported a significant increase in the glycation of IGHA2 in the plasma of patients with DM1 and with complications of diabetes, especially in retinopathy, nephropathy, and neuropathy.^32,33 In addition, a cross-sectional study reported that increased levels of IgA tended to be associated with hyperglycemia.³⁴ In this context, it was suggested that high levels of IGHA2 result from an immune response to advanced glycosylation end products that increase and accumulate with persistent high blood glucose in patients with diabetes.³¹

ApoA-II expression levels were upregulated in the DM1 group compared with the HC group. APOA-II is the second most abundant apolipoprotein constituent of HDL, and the increased expression of APOA-II has been associated with several aspects of HDL, triglycerides, obesity, and insulin resistance.^35,36 DM1 patients often show lipid disorders, such as abnormalities in the regulation of plasma lipid metabolism,³⁷ in different amounts of classes of lipoproteins, and in the transfer of lipids.³⁸ Lipoprotein abnormalities might contribute to the development of atherogenesis and diabetic nephropathy.^39,40 In fact, Soedamah-Muthu et al reported that a lower APOA-I to APO-II ratio in patients with microalbuminuria or macroalbuminuria could contribute to an increased risk of cardiovascular disease.⁴¹ Increased expression of APOA-II in transgenic mice showed enhanced HDL levels and increased size of HDL particles. Larger HDL observed is caused, in part, by the inhibition of the ability of hepatic lipase (HL) to hydrolyze phospholipids and triglycerides by APOA-II as well as an increased ratio of APOA-I to APOA-II. Therefore, interactions of APOA-II and APOA-I and HL influence the HDL structure and antiatherogenic properties.^42,43 Furthermore, an excess of APOA-II in HDL may contribute to hypertriglyceridemia by inhibiting the lipolysis of triglyceride-rich lipoproteins and by modifying the lipoprotein lipase activity.⁴⁴ In diabetes, deficient function of insulin-dependent lipoprotein lipase may lead to hypertriglyceridemia as well as decreased HDL cholesterol levels due to increased lipid exchanges between triglyceride-rich lipoproteins and HDL.³⁷ Nevertheless, studies using various approaches have described APOA-II as both proatherogenic and antiatherogenic; however, the role of this protein in lipid metabolism and atherogenesis is not well understood. Nevertheless, evidence supports the hypothesis that APOA-II is antiatherogenic, and further studies are required.⁴⁵ Taken together, these data indicate that APOA-II plays an important role in regulating lipid and β-cell metabolism. Thus, in the current study, our observation supports previous reports that show an association between apolipoproteins and DM1.

β2 glycoprotein I, also known as apolipoprotein H (APOH), is involved in the inhibition of the intrinsic pathway of the activation of blood coagulation. There is evidence to support the association between APOH and lipid metabolism, inflammation, thrombosis, autoimmune disease, and atherosclerosis.^46,47 Here, expression levels of APOH were upregulated in DM1 patients. The analysis of plasma levels of APOH from people with DM2 showed that patients with hypertriglyceridemia had significantly higher APOH values.⁴⁶ Others studies have shown an association between APOH and diabetic retinopathy and found increased levels of APOH in the vitreous fluid of DM1 patients with proliferative diabetic retinopathy.^48,49 However, to our knowledge, no study has yet shown increased levels of APOH in the serum of DM1 patients. Interestingly, APOH binds apoptotic cells in vivo, and this might trigger or enhance complement activation, thus facilitating the recognition and clearance of apoptotic cells.⁵⁰ The role of APOH in the pathogenesis of diabetic retinopathy remains to be elucidated, but it could be related to complement activation or its potential role in the recognition of dying cells.⁴⁹ Although the relationship between cell apoptosis and autoimmunity remains to be fully established, there is evidence that T-cell-induced apoptosis is a mechanism in DM1.⁵¹ Here, we show for the first time the upregulation of APOH in the serum of patients with DM1.

C4 was downregulated in the serum from patients with DM1. The complement system plays an important role in innate and acquired immunity, and evidence suggests that relevant target regulation of the complement system might be efficient in controlling autoimmune diseases.⁵² Abnormalities in complement proteins in DM1 patients have been reported, and reduced levels of C4 were attributed to hypercatabolism or reduction in protein synthesis.⁵³ Moreover, decreased levels of C4 in children and adolescents with DM1 were related to impaired metabolic control.⁵⁴ High glucose interferes with the innate function of the complement system and impairs the ability of the host to combat infection.⁵⁵ Deficiencies in C4 are strongly associated with autoimmunity humoral and lead to impaired humoral responses.⁵² Thus, the upregulation of C4 observed in this study is suggestive of loss of some functions of the complement system and thus increased risk of infectious processes and impaired metabolic control in DM1.

AMBP is a serum protein filtered by the glomerular membrane in the proximal tubule.^56,57 Increased urinary excretion of this protein was observed in DM1 patients in the presence of microalbuminuria and normoalbuminuria.⁵⁷ In this sense, new research has indicated some biomarkers that precede albuminuria and have been detected in both DM1 and DM2 patients with early renal dysfunction.⁵⁸ AMBP has been proposed as a biomarker of tubular dysfunction in the early stages of diabetic nephropathy and in the late stages of chronic kidney disease.^59–62 Increased excretion urinary of AMBP has been frequently identified in patients with diabetes, compared with the control group.⁵⁷ Increased levels of AMBP suggest that serum AMBP may play a role in the regulation of inflammatory processes,⁶² such as that occurs in diabetes. Interestingly, our study also showed that increased levels of AMBP occur in the serum proteome of DM1 patients, and to the best of our knowledge, this is the first study to observe this protein in serum from DM1 patients.

THRB is the protein precursor of thrombin, which plays a key role in thrombosis by converting fibrinogen to fibrin.⁶³

Diabetes is associated with a number of changes in thrombotic and fibrinolytic coagulation factor level/activity, which collectively increase the risk of thrombus formation.⁶³ Thrombin generation is enhanced in both DM1 and DM2 individuals, secondary to low-grade activation of the coagulation system.⁶⁴ It has been suggested that elevated THRB levels can contribute both to thrombotic risk and to shortened clotting times in patients with diabetes.⁶⁵ In this sense, hyperglycemia has been associated with enhanced thrombin production, and controlling glucose levels results in reduced thrombin generation, suggesting that hyperglycemia is prothrombotic.⁶⁴ Conversely, hypoglycemia is also associated with enhanced clot formation.⁶³ Here, we identified that THRB was upregulated in the serum of DM1 patients, which is an important component in the pathway of coagulation. This is in accordance with previous studies showing that this protein is suggestive of poorly controlled diabetes.

PZP was shown to be downregulated in the serum of DM1 patients compared with HCs. PZP is one of the most abundant pregnancy glycoprotein-associated plasma proteins with described immune inhibitory activities and is remarkably homologous to α2M.^66,67 Several studies have suggested that PZP has immunosuppressive properties in pregnancy, indicating a role for this protein in maintaining immune tolerance to the fetus and equilibrating the maternal immune response.^66,68 In fact, serum levels increase during gestation (up to 1,000 mg/L).⁶⁹ Furthermore, high serum levels of PZP have been observed in relation to some diseases, including inflammatory conditions, such as rheumatoid arthritis and also in people who later developed Alzheimer’s disease.^70,71 In the current study, we observed downregulation of PZP expression in the DM1 group compared with HCs. This finding is very relevant because, to our knowledge, this is the first report of low levels of PZP in the serum of DM1 patients. The reduction of this protein in DM1 patients suggests that there could be a suppression mechanism of it, thereby avoiding inhibition of the activation of T lymphocytes against the pancreatic β-cells. It has been suggested that PZP contributes to immune regulation by noncovalently sequestering a variety of other ligands such tumor necrosis factor-α, IL-2, and IL-6.⁷² In addition, Skornicka et al demonstrated that both native and activated PZP isoforms inhibited the activation of T-cell growth and IL-2 production.⁶⁶ DM1 is thought to involve chronic inflammation, leading to the release and increased activity of various immune system cells with a central role for T cells and macrophages.⁷³ Studies have shown that some inflammatory markers are enhanced in individuals with DM1 compared with those without.^74,75 Taken together, our results suggest that PZP could be a target for a potential DM1 therapy.

Conclusion

In conclusion, we used a proteomic approach with label-free MS^E analysis as a strategy to discover differential protein profiles between the serum of patients with DM1 and HCs.

We identified proteins that were expressed in differential levels in the serum of patients with DM1 compared with HCs. In addition, to our knowledge, we described the first evidence for changes in expression levels of PZP and APOH in the serum of DM1 patients.

Our results showed differential expression of proteins related to immune system, lipid metabolism, and coagulation pathways, which may be correlated to the development and progression of DM1. Accordingly, these findings need to be fully investigated and validated in future studies.

Acknowledgments

This work was supported by funding from Biotechnology and Molecular Biology Laboratory from Ceará State University and the National Research Council for graduate student scholarships. The authors thank the scholars of scientific initiation: PS, LT, JC, and MH.

Author contributions

All authors contributed toward data analysis, drafting and revising the paper and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.

References

1.		Milech A, Angelucci AP, Golbert A, et al. Diretrizes Da Sociedade Brasileira de Diabetes. 2015–2016th ed. Oliveira JEP de, Vencio S, editors. São Paulo: AC Farmacêutica; 2016. Available from: http://www.diabetes.org.br/profissionais/images/docs/DIRETRIZES-SBD-2015-2016.pdf.
2.		Petersen KF, Shulman GI. Etiology of insulin resistance. Am J Med. 2006;119(5 Suppl 1):S10–S16.
3.		Atkinson MA, Eisenbarth GS, Michels AW. Type 1 diabetes. Lancet. 2014;383(9911):69–82.
4.		International Diabetes Federation. IDF Diabetes Atlas; 7th ed. 2015. Available from: http:/www.diabetesatlas.org. Accessed February 1, 2018.
5.		World Health Organization. Global Report on Diabetes. 2016. Available in http://www.who.int/iris/handle/10665/204871. Accessed February 1, 2018.
6.		American Diabetes Association (ADA). Standard of medical care in diabetes - 2017. Diabetes Care. 2017;40(Suppl 1):s4–s5.
7.		Purohit S, She JX. Biomarkers for type 1 diabetes. Int J Clin Exp Med. 2008;1(2):98–116.
8.		Sosenko JM, Krischer JP, Palmer JP, et al. A risk score for type 1 diabetes derived from autoantibody-positive participants in the diabetes prevention trial-type 1. Diabetes Care. 2008;31(3):528–533.
9.		Chen CC, Lu YC, Chen YW, et al. Hemopexin is up-regulated in plasma from type 1 diabetes mellitus patients: role of glucose-induced ROS. J Proteomics. 2012;75(12):3760–3777.
10.		Zhi W, Sharma A, Purohit S, et al. Discovery and validation of serum protein changes in type 1 diabetes patients using high throughput two dimensional liquid chromatography-mass spectrometry and immunoassays. Mol Cell Proteomics. 2011;10(11):M111.012203.
11.		Sundsten T, Ortsater H. Proteomics in diabetes research. Mol Cell Endocrinol. 2009;297(1–2):93–103.
12.		Neilson KA, Ali NA, Muralidharan S, et al. Less label, more free: approaches in label-free quantitative mass spectrometry. Proteomics. 2011;11(4):535–553.
13.		Silva JC, Gorenstein MV, Li GZ, Vissers JP, Geromanos SJ. Absolute quantification of proteins by LCMSE: a virtue of parallel MS acquisition. Mol Cell Proteomics. 2006;5(1):144–156.
14.		SISVAN. Orientações Básicas Para a Coleta, Processamento, Análise de Dados E Informação Em Serviços de Saúde; 2004.
15.		WHO Expert Committee on Physical Status: the use and interpretation of anthropmetry. Physical status: the use of and interpretation of anthropometry: report of a WHO expert committee. 1995;854:1–4520.
16.		World Health Organization. Global Status Report on Alcohol. Department of Mental Health and Substance Abuse. 2004.
17.		Schoenborn CA, Adams PE. Health behaviors of adults: United States, 2005-2007. Vital Health Stat 10. 2010;(245):1–132.
18.		Friedewald WT, Levy RI, Fredrickson DS. Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge. Clin Chem. 1972;18(6):499–502.
19.		Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem. 1976;72:248–254.
20.		Li GZ, Vissers JP, Silva JC, Golick D, Gorenstein MV, Geromanos SJ. Database searching and accounting of multiplexed precursor and product ion spectra from the data independent analysis of simple and complex peptide mixtures. Proteomics. 2009;9(6):1696–1719.
21.		Sparre T, Larsen MR, Heding PE, Karlsen AE, Jensen ON, Pociot F. Unraveling the pathogenesis of type 1 diabetes with proteomics: present and future directions. Mol Cell Proteomics. 2005;4(4):441–457.
22.		Rehman AA, Ahsan H, Khan FH. Alpha-2-macroglobulin: a physiological guardian. J Cell Physiol. 2013;228(8):1665–1675.
23.		Takada T, Kodera Y, Matsubara M, et al. Serum monomeric α2-macroglobulin as a clinical biomarker in diabetes. Atherosclerosis. 2013;228(1):270–276.
24.		Caseiro A, Barros A, Ferreira R, et al. Pursuing type 1 diabetes mellitus and related complications through urinary proteomics. Transl Res. 2014;163(3):188–199.
25.		Lu CH, Lin ST, Chou HC, Lee YR, Chan HL. Proteomic analysis of retinopathy-related plasma biomarkers in diabetic patients. Arch Biochem Biophys. 2013;529(2):146–156.
26.		Athippozhy A, Huang L, Wooton-Kee CR, et al. Differential gene expression in liver and small intestine from lactating rats compared to age-matched virgin controls detects increased mRNA of cholesterol biosynthetic genes. BMC Genomics. 2011;12:95.
27.		Schroeder HW, Cavacini L. Structure and function of immunoglobulins. J Allergy Clin Immunol. 2010;125(2 Suppl 2):S41–S52.
28.		Woof JM, Russell MW. Structure and function relationships in IgA. Mucosal Immunol. 2011;4(6):590–597.
29.		Triolo G, Giardina E, Zarcone MR, Giordano C, Rinaldi A, Bompiani GD. Contribution of secretory IgA, polymeric IgA and IgA/secretory component-containing circulating immune complexes to the serum hyper-IgA in diabetes mellitus. Diabetologia. 1984;27(Suppl):157–159.
30.		Figueredo A, Ibarra JL, Rodriguez A, et al. Increased serum levels of IgA antibodies to hsp70 protein in patients with diabetes mellitus: their relationship with vascular complications. Clin Immunol Immunopathol. 1996;79(3):252–255.
31.		Rodriguez-Segade S, Camiña MF, Carnero A, et al. High serum IgA concentrations in patients with diabetes mellitus: agewise distribution and relation to chronic complications. Clin Chem. 1996;42(7):1064–1067.
32.		Jaleel A, Halvatsiotis P, Williamson B, Juhasz P, Martin S, Nair KS. Identification of Amadori-modified plasma proteins in type 2 diabetes and the effect of short-term intensive insulin treatment. Diabetes Care. 2005;28(3):645–652.
33.		Singh AK, Kingston JJ, Gupta SK, Batra HV. Recombinant bivalent fusion protein rVE induces CD4+ and CD8+ T-cell mediated memory immune response for protection against yersinia enterocolitica infection. Front Microbiol. 2015;6:1407.
34.		Gonzalez-Quintela A, Alende R, Gude F, et al. Serum levels of immunoglobulins (IgG, IgA, IgM) in a general adult population and their relationship with alcohol consumption, smoking and common metabolic abnormalities. Clin Exp Immunol. 2007;151(1):42–50.
35.		Castellani LW, Goto AM, Lusis AJ. Studies with apolipoprotein A-II transgenic mice indicate a role for HDLs in adiposity and insulin resistance. Diabetes. 2001;50(3):643–651.
36.		Castellani LW, Nguyen CN, Charugundla S, et al. Apolipoprotein AII is a regulator of very low density lipoprotein metabolism and insulin resistance. J Biol Chem. 2008;283(17):11633–11644.
37.		Feitosa AC, Feitosa-Filho GS, Freitas FR, Wajchenberg BL, Maranhão RC. Lipoprotein metabolism in patients with type 1 diabetes under intensive insulin treatment. Lipids Health Dis. 2013;12:15.
38.		Pérez A, Caixàs A, Carreras G, et al. Lipoprotein compositional abnormalities in type I diabetes: effect of improved glycaemic control. Diabetes Res Clin Pract. 1997;36(2):83–90.
39.		Vergès B. Lipid disorders in type 1 diabetes. Diabetes Metab. 2009;35(5):353–360.
40.		Overgaard AJ, Hansen HG, Lajer M, et al. Plasma proteome analysis of patients with type 1 diabetes with diabetic nephropathy. Proteome Sci. 2010;8:4.
41.		Soedamah-Muthu SS, Colhoun HM, Taskinen MR, Idzior-Walus B, Fuller JH; EURODIAB IDDM Complications Study Group. Differences in HDL-cholesterol:apoA-I + apoA-II ratio and apoE phenotype with albuminuric status in Type I diabetic patients. Diabetologia. 2000;43(11):1353–1359.
42.		Hedrick CC, Castellani LW, Wong H, Lusis AJ. In vivo interactions of apoA-II, apoA-I, and hepatic lipase contributing to HDL structure and antiatherogenic functions. J Lipid Res. 2001;42(4):563–570.
43.		Boucher J, Ramsamy TA, Braschi S, Sahoo D, Neville TA, Sparks DL. Apolipoprotein A-II regulates HDL stability and affects hepatic lipase association and activity. J Lipid Res. 2004;45(5):849–858.
44.		Julve J, Muñoz García C, Santos D, Blanco-Vaca F, Escolà-Gil JC. La apolipoproteína A-II altera la composición apolipoproteica de HDL y su capacidad para activar la lipoproteína lipasa. Clin Investig Arterioscler. 2010;22(5):192–197.
45.		Pownall HJ, Gillard BK, Gotto AM. Setting the course for apoAII: a port in sight? Clin Lipidol. 2013;8(5):551–560.
46.		Castro A, Lázaro I, Selva DM, et al. APOH is increased in the plasma and liver of type 2 diabetic patients with metabolic syndrome. Atherosclerosis. 2010;209(1):201–205.
47.		Gómez-Cardona EE, Hernández-Domínguez EE, Velarde-Salcedo AJ, et al. 2D-DIGE as a strategy to identify serum biomarkers in Mexican patients with type-2 diabetes with different body mass index. Sci Rep. 2017;7:46536.
48.		Ouchi Y, Yoshikawa E, Sekine Y, et al. Microglial activation and dopamine terminal loss in early Parkinson’s disease. Ann Neurol. 2005;57(2):168–175.
49.		Simó R, Higuera M, García-Ramírez M, Canals F, García-Arumí J, Hernández C. Elevation of apolipoprotein A-I and apolipoprotein H levels in the vitreous fluid and overexpression in the retina of diabetic patients. Arch Ophthalmol. 2008;126(8):1076.
50.		Chonn A, Semple SC, Cullis PR. Beta 2 glycoprotein I is a major protein associated with very rapidly cleared liposomes in vivo, suggesting a significant role in the immune clearance of “non-self” particles. J Biol Chem. 1995;270(43):25845–25849.
51.		Lee SC, Pervaiz S. Apoptosis in the pathophysiology of diabetes mellitus. Int J Biochem Cell Biol. 2007;39(3):497–504.
52.		Mizuno M, Suzuki Y, Ito Y. Complement regulation and kidney diseases: recent knowledge of the double-edged roles of complement activation in nephrology. Clin Exp Nephrol. 2018;22(1):3–14.
53.		Charlesworth JA, Timmermans V, Golding J, et al. The complement system in type 1 (insulin-dependent) diabetes. Diabetologia. 1987;30(6):372–379.
54.		Liberatore RR, Barbosa SF, Alkimin Md, et al. Is immunity in diabetic patients influencing the susceptibility to infections? Immunoglobulins, complement and phagocytic function in children and adolescents with type 1 diabetes mellitus. Pediatr Diabetes. 2005;6(4):206–212.
55.		Turina M, Fry DE, Polk HC. Acute hyperglycemia and the innate immune system: clinical, cellular, and molecular aspects. Crit Care Med. 2005;33(7):1624–1633.
56.		Ekström B, Peterson PA, Berggárd I. A urinary and plasma α1-glycoprotein of low molecular weight: Isolation and some properties. Biochem Biophys Res Commun. 1975;65(4):1427–1433.
57.		Soggiu A, Piras C, Bonizzi L, Hussein HA, Pisanu S, Roncada P. A discovery-phase urine proteomics investigation in type 1 diabetes. Acta Diabetol. 2012;49(6):453–464.
58.		Gluhovschi C, Gluhovschi G, Petrica L, et al. Urinary biomarkers in the assessment of early diabetic nephropathy. J Diabetes Res. 2016;2016:1–13.
59.		Pfleiderer S, Zimmerhackl LB, Kinne R, Manz F, Schuler G, Brandis M. Renal proximal and distal tubular function is attenuated in diabetes mellitus type 1 as determined by the renal excretion of alpha 1-microglobulin and Tamm-Horsfall protein. Clin Investig. 1993;71(12):972–977.
60.		Varghese SA, Powell TB, Budisavljevic MN, et al. Urine biomarkers predict the cause of glomerular disease. J Am Soc Nephrol. 2007;18(3):913–922.
61.		Aregger F, Uehlinger DE, Witowski J, et al. Identification of IGFBP-7 by urinary proteomics as a novel prognostic marker in early acute kidney injury. Kidney Int. 2014;85(4):909–919.
62.		Korenkova V, Jones A, Hoy WE, Morais C, Cooper MA, Gobe GC. Urinary biomarkers for detection of early and advanced chronic kidney disease – a pilot study. Med Chem (Los Angeles). 2015;5(2):96–103.
63.		Alzahrani S, Ajjan R. Review article: coagulation and fibrinolysis in diabetes. Diabetes Vasc Dis Res. 2010;7(4):260–273.
64.		Ceriello A, Esposito K, Ihnat M, Zhang J, Giugliano D. Simultaneous control of hyperglycemia and oxidative stress normalizes enhanced thrombin generation in type 1 diabetes. J Thromb Haemost. 2009;7(7):1228–1230.
65.		Sauls DL, Banini AE, Boyd LC, Hoffman M. Elevated prothrombin level and shortened clotting times in subjects with type 2 diabetes. J Thromb Haemost. 2007;5(3):638–639.
66.		Skornicka EL, Kiyatkina N, Weber MC, Tykocinski ML, Koo PH. Pregnancy zone protein is a carrier and modulator of placental protein-14 in T-cell growth and cytokine production. Cell Immunol. 2004;232(1–2):144–156.
67.		Devriendt K, Van den Berghe H, Cassiman JJ, Marynen P. Primary structure of pregnancy zone protein. Molecular cloning of a full-length PZP cDNA clone by the polymerase chain reaction. Biochim Biophys Acta. 1991;1088(1):95–103.
68.		Stimson WH. Studies on the immunosuppressive properties of a pregnancy-associated alpha-macroglobulin. Clin Exp Immunol. 1976;25(2):199–206.
69.		Folkersen J, Teisner B, Grunnet N, Grudzinskas JG, Westergaard JG, Hindersson P. Circulating levels of pregnancy zone protein: normal range and the influence of age and gender. Clin Chim Acta. 1981;110(2–3):139–145.
70.		Horne CHW, Schoultz B, Towler CM, Damber MG, Stigbrand T, Gibb LM. Serum pregnancy-associated α2-glycoprotein levels in autoimmune diseases. Vox Sang. 1978;34(6):329–334.
71.		IJsselstijn L, Dekker LJ, Stingl C, et al. Serum levels of pregnancy zone protein are elevated in presymptomatic Alzheimer’s disease. J Proteome Res. 2011;10(11):4902–4910.
72.		Wu SM, Pizzo SV. Mechanism of hypochlorite-mediated inactivation of proteinase inhibition by alpha 2-macroglobulin. Biochemistry. 1999;38(42):13983–13990.
73.		Bending D, Zaccone P, Cooke A. Inflammation and type one diabetes. Int Immunol. 2012;24(6):339–346.
74.		Snell-Bergeon JK, West NA, Mayer-Davis EJ, et al. Inflammatory markers are increased in youth with type 1 diabetes: the SEARCH case-control study. J Clin Endocrinol Metab. 2010;95(6):2868–2876.
75.		Xiang Y, Zhou P, Li X, et al. Heterogeneity of altered cytokine levels across the clinical spectrum of diabetes in China. Diabetes Care. 2011;34(7):1639–1641.

Supplementary material

Table S1 List of serum proteins between type 1 diabetes patients and healthy controls as identified by LC–MSE analysis Abbreviations: AMBP, alpha-1-microglobulin; LC–MSE, liquid chromatography–data-independent label-free mass spectrometry.