Deranged regulatory T-cells and transforming growth factor-β1 levels in type 1 diabetes patients with associated autoimmune diseases.

AIM: This study was designed to enumerate regulatory T-cells (Tregs) and estimate transforming growth factor-β1 (TGF-β1) levels in type 1 diabetic (T1D) patients with respect to disease duration and associated autoimmune diseases.
METHODS: One hundred and fifty patients and twenty healthy controls were recruited in the study. The patients were subcategorized into eight categories on the basis of disease duration (new onset [NO] and long standing [LS]) and associated diseases, i.e., celiac disease (CD) and autoimmune thyroid disease (AiTD). Treg cells were assessed as CD4+ CD25hi+, FOXP3+ cells and serum TGF-β1 levels were assessed by ELISA.
RESULTS: The frequency of Tregs and levels of TGF-β1 were significantly increased in the patients compared to the healthy controls. Among the different categories of the patients, no significant differences were seen for TGF- β1 levels, but for Tregs in patients with T1D and AiTD (P = 0.035). A significant correlation was also found between percentage count of Tregs and TGF-β1 levels in NO cases in all disease subcategories, but not in LS patients.
CONCLUSION: Thus, there was an increased percentage of Tregs and serum levels of TGF-β1 in T1D patients, irrespective of the disease duration and associated autoimmune diseases. The significant correlation in these two parameters at the onset of the disease, but not in LS disease, indicates that the immunological milieu in LS autoimmune diseases is more complicated with disease-associated conditions such as prolonged hyperglycemia, insulin therapy, and/or continued gluten in diet. Treatment and modulation of these long-term complications for improving immunological parameters require further research.

Introduction

Regulatory T-cells (Tregs) represent a subpopulation of CD4+ T-cells and comprise 5%–10% of the peripheral CD4+ T-cells. These cells were first described by Sakaguchi et al. in 1995,[1] and since then, they have been a major focus of research in various immunologic complications such as autoimmune diseases, carcinomas, and solid organ transplants. Tregs arise from thymus and require self-antigen recognition for development and then finally seed in the peripheral lymphoid organs where they modulate the immune responses by their immunosuppressive properties. Tregs are characterized by high surface expression levels of CD25, which is interleukin (IL)-2 receptor, and intracellular FOXP3, which is a master regulator of this lineage. There are also several other molecules that are associated with Tregs, these include glucocorticoid-induced tumor necrosis factor, cytotoxic T lymphocyte antigen-4, CD103, CD 127lo (IL-7R), CD62 L, tumor necrosis factor receptor 2, and transforming growth factor-β1 (TGF-β)-R1.[23]

Besides the naturally occurring cells (nTregs), Tregs can also be induced in the periphery by interaction with the antigen or polarization by TGF-β and these are generally denoted as inducible Tregs (iTregs). They can be differentiated from nTregs by T-cell-specific demethylation region.[45] While nTregs suppress immunological responses in multiple ways such as negative signaling, surface molecules, cytokine killing, and antigen-presenting cell function downregulation, iTregs mediate suppression primarily in cytokine-dependent manner, as IL-10 and TGF-β secreted by Tregs are immunosuppresive. Treg cells can even be induced to develop Th17 phenotype by exposure to IL-6. This is being speculated as a mechanism for escaping central tolerance and thus leading to the commencement of autoimmunity.[67]

There have been convincing evidences both from human and mouse model studies that support the notion that type 1 diabetes (T1D) develops due to loss of immunological tolerance that is maintained by Tregs. Mice with mutation in FOXP3 gene develop immunodysregulation polyendocrinopathy enteropathy X-linked syndrome that emphasizes the role of Tregs in subduing autoimmunity.[8] The mutations in the Aire gene that lead to the development of autoimmune polyendocrine syndrome-1 or autoimmune polyendocrinopathy candidiasis ectodermal dystrophy are also associated with reduced Treg function.[9] In addition, it has been shown that diabetes can be induced by transfer of diabetogenic T-cells from nonobese diabetic mice to syngeneic recipients and can also be prevented by co-injection of CD4+ CD25++ cells.[1011] Many studies have explored the number and function of Treg cells in T1D and there have been several conflicting reports citing no alteration in the number of Tregs, whereas others describing decreased Treg numbers in patients compared to healthy controls. In this study, we have tried to see the frequency of Tregs in peripheral blood and serum TGF-β1 levels in T1D patients with respect to duration of the disease and associated polyendocrinopathies, i.e., autoimmune thyroid disease (AiTD) and celiac disease (CD).

Materials and Methods

Patients

One hundred and fifty patients with T1D with or without associated AiTD and CD were enrolled in the study after obtaining written informed consent. The patients were attending Diabetes clinics at Postgraduate Institute of Medical Education and Research, Chandigarh, India. The patients were diagnosed for T1D, CD and AiTD following the American Diabetes Association,[12] European society for pediatric gastroenterology, hepatology and nutrition,[13] and American thyroid association[14] criteria, respectively. Clinical details and laboratory investigations were obtained from the medical record of the patients. The study was approved by Institute's Ethics Committee. Twenty age and sex-matched healthy voluntary controls were also enrolled in the study. The patients were initially grouped into four categories as follows: (i) patients with T1D only, (ii) patients with T1D + AiTD, (iii) patients with T1D + CD, and (iv) patients with T1D + CD + AiTD. Again, they were subgrouped on the basis of the duration of the disease as new-onset (NO) and long-standing (LS) cases. Therefore, the patients were divided into eight different categories.

Blood sample collection, serum, and peripheral blood mononuclear cells’ isolation

Eight milliliters of peripheral venous blood sample was obtained from the patients following standard phlebotomy protocol. Subsequently, serum was separated, and peripheral blood mononuclear cells were isolated by density gradient centrifugation method.

For enumeration of Tregs, the lymphocytes isolated were surface stained with anti-CD4 conjugated to fluorescein isothiocyanate, anti-CD25 conjugated to allophycocyanin, and intracellularly stained with anti-FOXP3 conjugated to phycoerythrin (Becton and Dickinson, New Jersey, USA). Acquisition was done on flow cytometer (BD FACS CANTO, Becton and Dickinson, New Jersey, USA) by acquiring 100,000 events, and analysis was done on BD FACS DIVA software (Becton and Dickinson, New Jersey, USA). The CD4+ cells that were CD25hi were further gated for FOXP3. The enumeration of CD4+ CD25hi+ FOXP3+ Tregs was done as increase in median fluorescence intensity (MFI).

Estimation of serum levels of cytokine transforming growth factor-β1

Serum levels of cytokine TGF-β1 were assessed in the patients and the controls by ELISA set procured from BD Biosciences (Becton and Dickinson, New Jersey, USA). The values of unknown samples were calculated by log-log regression.

Statistical analysis

Data were expressed as mean with standard deviation and median with interquartile range as applicable. The data in different categories were compared by nonparametric tests such as Mann–Whitney U-test, Kruskal–Wallis test and Pearson's correlation test if the data were skewed and by parametric test such as Student's t-test when data were normally distributed by Graphpad prism-4 software (GraphPad Software, Inc. California, USA). Two-tailed P < 0.05 was considered statistically significant.

Results

The grouping of the patients, their demographic profile, and clinical details have been outlined in Table 1. The mean age of onset of diabetes was similar, i.e., 10–12 years in all the subcategories of the patients except patients with T1D + AiTD, in whom it was higher, i.e., 15.3 ± 4.7 years. Male preponderance was seen in patients with T1D only and female preponderance was seen in patients with a co-associated autoimmune disorder. The incidence of diabetic ketoacidosis was highest in patients with NO T1D + CD. The incidence of diabetes-related complications, i.e., retinopathy, nephropathy, and neuropathy was higher in LS patients of all categories compared to respective NO patients, and the highest incidence was seen in patients with T1D + CD + AiTD (66.6%).

Table 1

Different categories of the patients and their respective demographic and clinical profile

Parameter	Patient category (n=150)
	T1D (n=78)		T1D + CD (n=34)		T1D + AiTD (n=22)		T1D + AiTD + CD (n=16)
	NO (n=44)	LS (n=34)	NO (n=9)	LS (n=25)	NO (n=7)	LS (n=15)	NO (n=4)	LS (n=12)
Mean age (years)	10.8±4.7	21.9±8	12.4±6.0	17.7±7.8	15.1±4.7	22.7±9.1	11.5±2	19.6±5.9
Mean age of onset of diabetes (years)	10.6±4.6	11.9±5.5	12.4±5.8	10.9±3.9	15.3±4.7	13.3±6.4	11.5±2	11.5±5.9
Sex ratio (male/female)	2.14/1	1.43/1	1/2	1/2.57	1/1.3	1/1.5	3/1	1/2
Mean c peptide	0.27±0.12	0.28±0.14	0.26±0.10	0.28±1.2	0.24±0.07	0.22±0.11	0.30±0.21	0.26±0.11
Mean glycated hemoglobin (%)	11.32±3.5	9.84±2.7	11.12±3.1	10.36±3.1	12.65±4.6	9.4±2.2	12.4±2.5	9.23±2.1
Family history of diabetes (%)	29	29	22	24	43	20	50	0
DKA incidence (%)	59	62	88	40	43	53	50	25
Diabetes related complications incidence (%)	9.1	44.1	22.2	36	14.28	53.3	0	66.6

T1D: Type 1 diabetes, CD: Celiac disease, AiTD: Autoimmune thyroid disease, NO: New onset, LS: Long standing, DKA: Diabetic ketoacidosis

Tregs were analyzed in the patients and the controls and it was seen that the frequency of Tregs was significantly increased in the patients compared to the controls (P = 0.0003). The enumeration of Tregs was also done in different categories of the patients as shown in Figure 1. No significant difference was found in the Tregs number among the different categories of the patients. Then, the analysis was done on the basis of duration of the disease, a significant reduction was only found in patients with NO T1D + AiTD compared to LS patients with T1D + AiTD (P = 0.035), but in none of the other disease categories.

Figure 1

Enumeration of regulatory T-cells (median fluorescence intensity) in different subcategories of the patients

Serum levels of TGF-β1 were analyzed in the patients and the controls and similar to the number of Tregs; the levels of TGF-β1 were also significantly increased in the patients compared to controls (P = 0.0076). Further analysis among the different disease subcategories showed no significant difference in the levels of TGF-β1 as shown in Figure 2. However, it was seen that the levels of TGF-β1 were lower in patients with AiTD. Tregs frequency and levels of TGF-β1 were also analyzed with respect to age of onset of diabetes and gender, but no significant difference was found in either of the disease category.

Figure 2

Estimation of serum transforming growth factor-β1 levels (pg/ml) in different subcategories of the patients

A correlation analysis was done in the MFI of Tregs and levels of TGF-β1 in different categories of the patients as shown in Figure 3. A significant correlation was seen in the NO cases in all the categories as indicated by Pearson's correlation coefficient and P values (NO T1D: r = 0.3108, P = 0.04, NO T1D+CD: r = 0.8568, P = 0.0032, NO T1D+AiTD: r = 0.8898, P = 0.0073, NO T1D+AiTD+CD: r = 0.9712, P = 0.028), but no correlation was seen in LS patients in any of the disease category. Therefore, these results indicate that Tregs number and the respective cytokine levels were increased in T1D patients but did not vary among the group of patients who had associated polyendocrinopathy. Association analysis was also done to see if any of the clinical findings, hematological parameters, or laboratory features of the patients correlate with the immunological parameters; however, no association was found.

Figure 3

Correlation of regulatory T-cells (median fluorescence intensity) and serum transforming growth factor-β1 levels (pg/ml) in different subcategories of the patients. Plot a shows all patients, plot b shows new onset T1D patients, plot c shows long standing T1D patients, plot d shows new onset T1D + CD patients, plot e shows long standing T1D + CD patients, plot f shows new onset T1D + AiTD patients, plot g shows long standing T1D + AiTD patients, plot h shows new onset T1D + CD + AiTD patients, plot i shows long standing T1D + CD + AiTD patients

Discussion

Although T1D is the most studied and explored autoimmune disease, yet the pathogenetic mechanisms have not been completely delineated. Whereas, in several other autoimmune diseases such as autoimmune hepatitis and systemic lupus erythematous, the antigenic targets that are in 100% consensus with the diagnosis are known, and also impaired Treg characteristics and augmented inflammatory pathways have been described.[1516] This clear-cut paradigm is not seen in T1D because it presents a complex milieu due to the presence of chronic hyperglycemia-related complications in the backdrop of autoimmune state. This further becomes intricately complex, when T1D is accompanied with other autoimmune diseases such as CD and AiTD. As it is known that 15%–30% of the patients with T1D develop other autoimmune diseases,[17] these accomplices act as confounding factors in delineating immunological characteristics of the disease such as number and function of Tregs. Contrasting reports pertaining to the number of Tregs in T1D patients are available in literature. Many studies have described decrease in the number of CD4+, CD25hi in various autoimmune disorders including T1D.[181920] These have been contradicted by other studies that show that there is no difference in the Tregs number in T1D patients and the controls[2122] and have indicated that age mismatched participants may be the reason for yielding false results in the previous studies. Furthermore, enumeration of Tregs with additional and more specific markers such as FOXP3, CD127low, and Helios rather than just CD25 might help in getting more relevant and authentic results.

We have found that the number of Tregs was increased in the patients compared to the controls. These findings have been supported by studies by El Gamal et al. and Oling et al. who reported that autoantibody positive individuals testing positive for three to four autoantibodies had a significantly higher frequency of CD4+ CD25hi human leukocyte antigen DR− and CD4+ CD25hi CD69− cells compared to controls.[2324] Alonso et al., have also observed increased Treg number in peripheral blood of T1D patients with chronic atrophic gastritis.[25] Brusko et al., suggested that the altered peripheral blood frequencies of Tregs, as defined by expression of FOXP3 are not specifically associated with T1D but are because of age, which is an important variable in analysis of immune regulation.[26] In our study we found no significant difference in Treg number with reference to age and sex. Lindley et al. and Ryba-Stanislawowska et al. have also found no correlation in Treg number and age.[2728] Kivling et al., have shown higher expression of FOXP3 mRNA in a CD patient cohort.[29] Frisullo et al., have reported increased CD4+ CD25+ FOXP3+ T-cells in peripheral blood of CD patients and have shown the correlation between the frequency of these cells and dietary treatment.[30] A recent study by Perri et al., has shown an increased expression of programed death-1 on Tregs in LS patients indicating that they are defective in proliferation and function.[31] Defective Treg functions have also been described in CD.[32] Insulin which is prescribed exogenously to T1D patients have also been described to have an effect on immune cells including Tregs cells and IL-10 production as described by Han et al.,[33] however more studies are needed to evaluate the effect of therapy on immune cells. There is a possibility that immune cells become exhausted or refractory to immune regulation due to chronic inflammation, therapy or disease duration.

TGF-β is a pleiotropic cytokine that has role in immunosuppression as well as fibrosis. The detection of this cytokine at significantly higher levels in the patients reinstates the complex disease scenario. High levels of TGF-β have been described to exhibit immunosuppressive effect by inhibiting Th17 cells differentiation and favor iTreg differentiation from naïve CD4+ cells.[34] Many studies have also described high levels of TGF-β in T1D patients with associated microvascular complications like retinopathy and nephropathy.[3536] We saw a correlation between Tregs number and TGF β1 levels in NO cases. This may be indicative of Tregs functional efforts in the initial stages of the disease onset. It has also been postulated by several authors that the effector cells become nonresponsive to regulation by Tregs in autoimmune diseases,[37] so the increased Tregs number and TGF β1 levels may be indicative of enhanced Treg function to combat the awry effector cells. In LS cases, Tregs and TGF β1 remain high but no correlation is seen. This signifies a loss of feedback control of TGF β1, which over time along with advanced glycated end products, contributes to microvascular complications and induces fibrosis in end organs. In CD, there is an added insult to the microvilli in the jejunum with resultant atrophy and fibrosis which further adds to this complicated milieu.

Our study shows increased Tregs and TGF β1 secretion in T1D patients irrespective of associated autoimmune conditions like CD and AiTD. Modulating long-term complications of the disease would require more understanding of Treg heterogeneity and dynamics of TGF β1 secretary pathways.

Financial support and sponsorship

The authors acknowledge institute research grant and Indian Council of Medical Research, New Delhi for funding.

Conflicts of interest

There are no conflicts of interest.