Type 1 diabetic mellitus patients with increased atherosclerosis risk display decreased CDKN2A/2B/2BAS gene expression in leukocytes

Background Type 1 diabetes mellitus (T1DM) patients display increased risk of cardiovascular disease (CVD) and are characterized by a diminished regulatory T (Treg) cell content or function. Previous studies have shown an association between decreased CDKN2A/2B/2BAS gene expression and enhanced CVD. In the present study the potential relationship between CDKN2A/2B/2BAS gene expression, immune cell dysfunction and increased cardiovascular risk in T1DM patients was explored. Methods A cross-sectional study was performed in 90 subjects divided into controls and T1DM patients. Circulating leukocyte subpopulations analysis by flow cytometry, expression studies on peripheral blood mononuclear cell by qPCR and western blot and correlation studies were performed in both groups of subjects. Results Analysis indicated that, consistent with the described T cell dysfunction, T1DM subjects showed decreased circulating CD4+CD25+CD127− Treg cells. In addition, T1DM subjects had lower mRNA levels of the transcription factors FOXP3 and RORC and lower levels of IL2 and IL6 which are involved in Treg and Th17 cell differentiation, respectively. T1DM patients also exhibited decreased mRNA levels of CDKN2A (variant 1 p16Ink4a), CDKN2A (p14Arf, variant 4), CDKN2B (p15Ink4b) and CDKN2BAS compared with controls. Notably, T1DM patients had augmented pro-atherogenic CD14++CD16+-monocytes, which predict cardiovascular acute events and enhanced common carotid intima-media thickness (CC-IMT). Conclusions Decreased expression of CDKN2A/2B/2BAS in leukocytes associates with increased CC-IMT atherosclerosis surrogate marker and proatherogenic CD14++CD16+ monocytes in T1DM patients. These results suggest a potential role of CDKN2A/2B/2BAS genes in CVD risk in T1DM. Electronic supplementary material The online version of this article (10.1186/s12967-019-1977-1) contains supplementary material, which is available to authorized users.


Background
Type 1 diabetes mellitus is an autoimmune disease characterized by a deficiency of insulin production by the pancreas. Increased susceptibility to T1DM is determined by a combination of several genetic and environmental factors. Thus, variants of human leukocyte antigen (HLA) genes which influences antigen presentation during thymic cell selection processes and peripheral activation of the immune response have been associated with T1DM pathogenesis [1]. The disease results from the destruction of insulin-producing β cells by autoreactive effector CD4+ and CD8+ T cells, which initially infiltrate the islets producing insulitis, in response to pancreatic islet autoantigens [2]. These autoreactive T cell responses to autoantigens are present before the clinical onset of the disease. In healthy conditions T cell responses are balanced by regulatory mechanisms such as immunological tolerance, but have a reduced functional capacity during disease development [1]. One key player in immunological tolerance is a CD4+ T cell sub-population, regulatory T (Treg) cells, whose defective function has been suggested to contribute to T1DM development. It has been shown that T1DM individuals display apoptosis-prone Treg cells in the blood. Furthermore, analysis of lymphocytic islet infiltration from T1DM patient donors revealed autoreactive T cells with an effector phenotype, such as CD4+ T helper (Th) 1, an abnormal expansion of proinflammatory effector Th17 cells and diminished Treg cell content or function [3].
Despite glycemic control through restoration of insulin levels, T1DM subjects display increased cardiovascular disease (CVD) similar to that observed in type 2 diabetes mellitus (T2DM) subjects [4,5]. However, unlike T2DM there is a lack of understanding of the risk in T1DM patients in current intensive glycemic control management guidelines. Many T1DM patients might display metabolic abnormalities that promote characteristics associated with enhanced CVD risk such as a chronic proinflammatory state. On the other hand, as mentioned above, T1DM subjects exhibit deranged T cell function which is also associated with increased CVD risk.
In fact, atherosclerosis progression, the main cause of CVD, is facilitated by an unbalanced interplay of different CD4+ T cell subsets, which promote plaque lesion formation in the vascular bed. Proinflammatory effector CD4+ Th1 and Th17 subsets are proatherogenic while Treg cells, which are a minor population in plaques, suppress Th cell activity in lesions. Moreover, Treg cells promote an anti-inflammatory and pro-resolving macrophage phenotype within atheromas [6]. Thus, altered Treg/Th17 ratio in circulating leukocytes favor atherosclerosis progression and human patients with vulnerable atheroma plaques and coronary artery disease (CAD) display reduced circulating Treg cells [7][8][9].
We have recently shown that CAD complications in T2DM subjects associated with changes in immune cell homeostasis and with diminished expression of CDKN2A/2B/2BAS genes [10]. These genes encode for the tumour suppressors p16 Ink4a , ARF and p15 Ink4b and the regulatory antisense non-coding RNA named ANRIL (antisense noncoding RNA in the INK4 locus) [11]. Previous genome-wide studies linked single nucleotide polymorphisms (SNPs) in CDKN2A/2B/2BAS genes to increased risk of T2DM and CVD [11]. Functional studies also indicated that upregulation of these genes might be a therapeutic tool to modulate insulin resistance and atherosclerosis [12][13][14][15][16][17].
In light of the above research, in the present study we investigated whether deranged immune systems in T1DM individuals could be related to atherosclerosis surrogate markers and altered expression of CDKN2A/2B/2BAS genes.

Subjects
The study was performed in accordance with the ethical principles of the Helsinki Declaration, and was approved by the Hospital Clínico Universitario de Valencia ethics committee. All subjects gave written informed consent. For the study 90 unrelated individuals attending the outpatient clinic and selected over 18 months by the opportunistic sampling method, were divided into controls and T1DM subjects. T1DM was diagnosed when fasting plasma glucose was ≥ 126 mg/dL or when HbA1C was ≥ 6.5% or when a patient had classic symptoms of hyperglycemia or hyperglycemic crisis, with a random plasma glucose ≥ 200 mg/dl, in accordance with the ADA criteria for T1DM [18] together with positivity of autoantibodies.

Human carotid artery ultrasound evaluation for common carotid intima-media thickness (CC-IMT)
A standardized imaging protocol was used for common carotid intima-media thickness (CC-IMT) measurements in agreement with the Mannheim consensus [19]. B-mode ultrasound imaging of the right and left carotid arteries was performed using a Siemens Sonoline G40 instrument equipped with 7 to 10-MHz broadband linear array transducers as previously described [20]. With the carotid dilatation and flow dividers as anatomic landmarks, the sonographer obtained high-resolution images of the common carotid (1 cm proximal to the bifurcation), the bifurcation (between dilatation and flow divider), and the internal carotid (1 cm distal to the flow divider). Carotid arteries and bifurcation were examined for presence of atherosclerotic plaques defined as described [20]. An experienced sonographer (S.M.-H.) performed all examinations. Intraobserver variability was examined in 20 subjects. The coefficient of variability of mean CC-IMT was 5.2%.

Human plasma parameters analysis and enzyme-linked immunosorbent assay (ELISA)
Samples were collected after 12-14 h fasting from an antecubital vein in tubes containing EDTA (BD Vacutainer) and centrifuged within 4 h. Plasma was stored at 4 °C and biochemical parameters were measured by standard clinical methodology. Insulin was determined by radioimmunoassay. For ELISA heparinized human whole blood (10 U heparin/ml) was obtained, plasma was isolated and cytokines were measured using the human MCP1, TNFα, TGFβ, IL2, IL6 and IL17 DuoSet ELISA (R&D Systems, UK).

Gene expression analysis by quantitative real-time PCR (qPCR)
Human peripheral blood mononuclear cells (PBMCs) were isolated from whole blood of patients with Lymphoprep (Axis Shield PoC, Norway) to obtain RNA using TRIzol (Invitrogen). RNA (0.5-1 μg) was retrotranscribed with cDNA Synthesis kit and amplified with Luminars Color-qPCR Master MIX (Fermentas) using a thermal Cycler 7900Fast System (Applied Biosystems

Statistical analysis
Quantitative variables appear as mean ± sem and correlations as individual points. Differences were assessed by Student´s t-test or Mann-Whitney U test and correlation studies by the Spearman correlation coefficient (Graph-Pad Prism 5.03, CA, USA). Statistical significance was set at p ≤ 0.05.

Altered T cell subsets in PBMCs from human T1DM patients
T1DM is characterized by deranged balance of CD4+ Th cellular subsets, whose fate is determined by the expression of their corresponding transcription factors and activation of the different cytokine-induced JAK/ STAT(signal transducer and activator of transcription)signaling pathways. Expression of the above were therefore analyzed in T1DM subjects and controls.
Expression analysis showed a marked reduction in FOXP3 expression, a key transcription factor for Treg cell function, in T1DM subjects compared with controls ( Fig. 3d, p < 0.0001). Similarly, analysis of the mRNA levels of RORC, the transcription factor expressed in Th17 cells was also diminished in T1DM subjects (Fig. 3c, p < 0.02). No changes were observed in TBET or GATA3 (Fig. 3a, b). Given that JAK-dependent pathways can be directly inhibited by SOCS1 and SOCS3 proteins [23], these were also analyzed. T1DM patients showed a marked decrease in the expression of SOCS1 mRNA (Fig. 3e, p < 0.007) and no changes in SOCS3 mRNA levels (Fig. 3f ) compared with controls.
Inverse correlations were found between mRNA levels of FOXP3 and SOCS1 and HbA1c % (Additional file 1: Table S2 p < 0.007 and p = 0.05 respectively) as well as with glucose levels (Additional file 1: Table S2, p < 0.05 and p < 0.004, respectively).

Discussion
T1DM patients have an increased risk of CVD but the causes of this risk are not fully understood. In the present study, we found diminished mRNA expression levels of CDKN2A (p16 Ink4a ), CDKN2A (p14 Arf ), CDKN2B (p15 Ink4b ) and CDKN2BAS genes in circulating leukocytes of T1DM patients, who also displayed increased atherosclerosis risk measured as CC-IMT. Consistent with the well-known deranged immune T-cell function, T1DM subjects exhibited lower circulating Treg (CD4+CD25+CD127−) cell percentages and reduced mRNA levels of transcription factors related to CD4+ Th differentiation, RORC and FOXP3, which are determinants for Th17 and Treg differentiation, respectively. In agreement with lower expression of these transcription factors, IL2 and IL6 cytokine levels were also decreased in T1DM patients. Notably, T1DM subjects had enhanced percentages of the proinflammatory CD14++CD16+ monocyte subpopulation, a cellular subset that has been reported to predict acute cardiovascular events. Altogether, our study indicates an association between increased atherosclerosis in T1DM and reduced expression of CDKN2A/2B/2BAS in circulating leukocytes which displayed proatherogenic phenotypes such as enhanced proinflammatory monocytes and reduced Treg content. Therefore, the present data suggest for the first time a potential role of CDKN2A/2B/2BAS through a decreased expression in leukocytes in atherosclerosis  Former studies have shown an increased CVD risk in T1DM patients despite intensive glycemic control [4]. However, ours is the first investigation that associates this CVD risk with a markedly reduced expression of CDKN2A/2B/2BAS genes in circulating leukocytes from T1DM patients. Previous genome-wide studies have linked SNPs in CDKN2A/2B/2BAS genes with enhanced risk of CVD and with T2DM [11]. In addition, functional human and murine studies have shown a protective role of these genes against metabolic diseases. An atheroprotective function has therefore been suggested for these genes, since genetic inactivation of Cdkn2a variants in mice increases atherosclerosis [12,15] and reduced expression of the transcripts has been found in atherosclerosis patients [16]. Other studies suggested a role of these genes in carbohydrate metabolism [24], in insulin secretion [14] and β-cell islet biology [25]. Enhanced expression of Cdkn2a/2b in mice also prevents insulin resistance associated with aging [13] and delays hepatic steatosis produced by insulin resistance [17]. In line with these investigations, our study indicates an association between diminished expression of CDKN2A/2B/2BAS genes in leukocytes and increased atherosclerosis risk in T1DM.
One limitation in our study is the higher frequency of statin use as lipid-lowering strategy which might exert anti-inflammatory actions [26]. However, the use of these did not interfere with the study as T1DM subjects displayed enhanced inflammatory status with increased proinflammatory monocytes and decreased Treg cells.
In our study enhanced CC-IMT was also accompanied by augmented BMI and glucose levels and decreased circulating HDL-C which is consistent with the well-known association between altered metabolism and CVD. These metabolic abnormalities could modify CVD risk in T1DM patients as seen in T2DM subjects. However, this risk could be further enhanced by the intrinsic dysregulation of T cells of T1DM subjects. T1DM is characterized by impaired Treg function [27]  coronary atherosclerosis, acute coronary syndromes and plaque rupture [8,9]. Therefore, reduced circulating Treg cells in our T1DM subjects could reasonably contribute to increase atherosclerosis. On the other hand, deranged balance of CD4 + T cell subsets whose lineage commitment is specified by cytokine environment and activation of their corresponding transcription factors [28] might also promote atherosclerosis [9,29,30]. T1DM patients exhibited decreased levels of IL2 and IL6 cytokine, which in the presence of TGFβ, promote Treg differentiation by inducing FOXP3 expression and Th17 differentiation through RORC expression, respectively. FOXP3 and RORC were consistently diminished in leukocytes in our T1DM patients thus indicating reduced levels of Treg like before, as well as decreased Th17, whose role in atherosclerosis is controversial [30]. Moreover, SOCS1, which negatively regulates Treg but is necessary for its function [23] was also significantly reduced in T1DM subjects. Besides modulation of self-tolerance, Treg cells promote an anti-inflammatory macrophage phenotype and cytokines [6], hence defective function or number of these could contribute to generate proinflammatory monocytes/macrophages. In fact, a major finding of our study is that increased atherosclerosis in T1DM was associated with a prevalence of circulating proinflammatory CD14++CD16+ monocytes which have been shown to predict cardiovascular events independently of other risk factors [21] and are associated with coronary plaque vulnerability in coronary disease patients [31]. We did not observe changes in the pro-inflammatory CD14++ monocytes which is consistent with another a study showing that CD14++CD16− monocytes do not associate with atherosclerosis measured as CC-IMT in subjects without CV acute events [32]. To our knowledge, this is the first study to indicate a relationship between CD14++CD16+ monocyte prevalence and atherosclerosis in T1DM. Pro-atherogenic and proinflammatory leukocyte phenotypes observed in T1DM subjects, although promoted by the autoimmune cell derangement characteristic of these subjects, could be aggravated by the diminished expression of CDKN2A/2B/2BAS genes. Supporting this, a positive correlation was observed between CDKN2A/2B/2BAS expression and the transcription factors FOXP3, RORC and SOCS3 mRNA levels in leukocytes, suggesting a potential relationship between them. Furthermore, several lines of evidences from previous research support a role of these genes in immune cell modulation. Thus, an antiinflammatory role has been attributed to Cdkn1a (variant 1, p16 Ink4a ) by promoting IRAK1 degradation and diminishing cytokine secretion in macrophages [33]. Increased expression of Cdkn2a/2b genes in mouse models decreased circulating proinflammatory Ly6C hi monocytes, reduced activated T cells and prevented macrophage infiltration into tissues [17]. Genetic inactivation of Cdkn2a in the myeloid lineage enhanced Ly6C hi -monocytes [15], promoted megakaryopoiesis and increased platelet activity [34]. Local infusion of Cdkn2a (variant 1) in bone joints consistently impaired the expression of proinflammatory cytokines in a mouse model of autoimmune rheumatoid arthritis [35]. Notably, we recently reported that coronary artery disease in T2DM (T2DM-CAD subjects) is accompanied by decreased Treg cells and diminished expression of CDKN2A/2B/2BAS [10]. Furthermore, treatment of human lymphocytes in vitro and of atherosclerotic mouse models in vivo with PD0332991, a p16 Ink4a /p15 Ink4b mimetic drug and a proven selective inhibitor of CDK4 (the main target of these proteins), augmented Treg levels and diminished atherosclerosis and vulnerable plaque in the mouse model [10]. Comparison of both investigations revealed a much lower CDKN2A/2B/2BAS gene expression in our former study in T2DM and T2DM with CAD compared with controls, which were up to 60-80% reductions while T1DM subjects displayed 40% reductions compared with controls. In addition, CV risk measured as CC-IMT were 0.46, 0.594 and 0.829 for T1DM, T2DM and T2DM-CAD respectively which correlates well with reduced expression of CDKN2A/2B/2BAS genes and is consistent with the protective role of these genes. Altogether, these findings suggest that CDKN2A/2B/2BAS gene modulation might modify inflammatory cell phenotype in diverse experimental settings. Hence, decreased expression of CDKN2A/2B/2BAS genes in the present study could potentially contribute to the altered leukocyte phenotype and promote atherosclerosis in T1DM subjects.

Conclusions
Our study shows for the first time that increased atherosclerosis risk in T1DM patients associates with decreased expression of CDKN2A/2B/2BAS in leukocytes, which display pro-atherogenic phenotypes consisting of reduced circulating Treg cells and augmented proinflammatory CD14++ CD16+ monocytes. Expansion and stabilization of Tregs have been cited as promising therapies to treat both atherosclerosis [7] and autoimmune diseases such as T1DM [27]. Altogether, this suggests that therapies targeted at increasing CDKN2A/2B/2BAS to restore Treg cells, such as the ones we have previously described [10], could potentially be used to reduce CVD in T1DM patients.