Influence of air pollutants on circulating inflammatory cells and microRNA expression in acute myocardial infarction

Air pollutants increase the risk and mortality of myocardial infarction (MI). The aim of this study was to assess the inflammatory changes in circulating immune cells and microRNAs in MIs related to short-term exposure to air pollutants. We studied 192 patients with acute coronary syndromes and 57 controls with stable angina. For each patient, air pollution exposure in the 24-h before admission, was collected. All patients underwent systematic circulating inflammatory cell analyses. According to PM2.5 exposure, 31 patients were selected for microRNA analyses. STEMI patients exposed to PM2.5 showed a reduction of CD4+ regulatory T cells. Furthermore, in STEMI patients the exposure to PM2.5 was associated with an increase of miR-146a-5p and miR-423-3p. In STEMI and NSTEMI patients PM2.5 exposure was associated with an increase of miR-let-7f-5p. STEMI related to PM2.5 short-term exposure is associated with changes involving regulatory T cells, miR-146a-5p and miR-423-3p.


Methods
Population of the study. Our tertiary University Hospital, localized in the central core of Madrid, covers an area of 350.000 inhabitants and it is part of the Regional Network for acute ST-segment elevation myocardial infarction (STEMI). We prospectively included all consecutive patients admitted to our center between March 2017 and July 2018 with the diagnosis of STEMI and non-STEMI (NSTEMI) undergoing coronary angiography in the acute phase of the disease. For comparative purposes, we included a control group of patients with stable angina who underwent cardiac catheterization in our institution during the same recruitment period. Demographic data and other relevant clinical information were prospectively collected, including cardiovascular risk factors, previous medical history, Killip-Kimball class at presentation, angiographic information, high-sensitive T-troponin and creatine kinase peak. Exclusion criteria were: MI without obstructive coronary artery disease; coronary artery events no related to acute atherosclerotic plaque destabilization (e.g. spontaneous coronary artery dissection, coronary embolism or vasospastic angina); history of chronic inflammatory disease or concomitant treatment with anti-inflammatory drugs; and lack of data about air pollutant exposure. All patients underwent systematic circulating inflammatory cell analysis. To select plasma samples for miRNAs analysis, at the end of the recruitment period, patients were sorted out according to PM2.5 exposure. Upper and lower values were sex-and age-matched and, eventually, a group of 31 patients, representing high and low exposure, were selected (14 STEMI, 9 NSTEMI and 8 stable angina).
Air pollutant data collection. Madrid  Blood samples. An arterial blood sample was collected in BD Vacutainer tube (BD Plymouth, UK) at time of catheterization, before heparin administration. Blood samples were processed up to 24 h from collection and during this time they were kept at 4 °C. Plasma samples were obtained by centrifugation at 2000g at 4 °C, aliquoted and stored at − 80 °C until total RNA extraction. Plasma samples were tested for the presence of hemolysis using the absorbance at 414 nm in a NanoDrop One spectrophotometer (Thermo Scientific).

Inflammatory cell analysis.
Peripheral Blood Leukocytes (PBLs) were isolated from human blood samples using Ficoll-Isopaque (density = 1.121 g/ml) gradient centrifugation. Human PBLs were incubated with fluorochrome-conjugated antibodies (Supplemental Table 1) for flow cytometry analysis. Membrane staining were performed in phosphate-buffered saline (PBS), 0.5% Bovine Serum Albumin (BSA), 1 mM EDTA during 15 min on ice. For T cell subsets analysis, the rest of PBLs were cultured overnight in plates coated with 3 µg/ml purified anti-CD3 (ΟΚΤ3 clone, Biolegend) in complete RPMI medium (Gibco) before cell staining. For regulatory T cell evaluation, cells were membrane-stained with anti-CD4 and anti-CD25 and then nuclear staining was performed using the Foxp3 staining buffer set (Miltenyi Biotec), according to the provider's instructions.
Cells were analyzed in a LSRFortessa Flow Cytometer and the data were processed with FlowJo v10.0.4 (Tree Star). Gating strategy is shown in Supplemental Fig. 1.
RNA isolation and retrotranscription. RNA was extracted from 200 µl of plasma using miRNeasy Serum/Plasma Advanced Kit (Qiagen), following the manufacturer's instructions. RNA was purified using RNeasy UCP MinElute spin columns, eluting with 20 µl of Rnase-free water. RNA samples were stored at − 80 °C until. Reverse transcription was performed from 2 µl of cDNA in a final reaction volume of 20 µl using miR-CURY LNA RT Kit (Qiagen) according to manufacturer's instructions. cDNA samples were stored at − 20 °C.
RT-PCR assays and miRNA expression analysis. RT-PCR assays were performed using ready-to-use miRCURY LNA miRNA serum/plasma Focus PCR Panels and miRCURY LNA SYBR Green PCR Kit (Qiagen) attending to manufacturer's instructions. Briefly, a mix containing 980 µl of Rnase-Free Water, 1 ml of 2 × miR-CURY LNA SYBR Green Master Mix and 20 µl of cDNA template was prepared, and 10 µl was dispensed per well. A CFX384 PCR detection system (Bio-Rad) was used for the assays.
Data were analyzed using the global mean normalization method 12 . Briefly, after exclusion of values above 36, Cq values were converted to relative quantities (RQ) and sample specific normalization factor (NF) was calculated as the geometric mean of the RQs of all expressed targets per sample. Normalized Relative Quantities (NRQ) were obtained by dividing the RQs by the sample specific NF. Data were expressed as NRQ. miRNA target identification. miRTarBase database was used for the unravel of miRNA targets and only those under type support "Functional miRNA-target interactions (MTI)" were selected to be subjected to the PANTHER Classification System, targets with a weak functional support were excluded. A statistical test of over- Assuming that the number of microRNAs differentially expressed among groups is very small, we ranked the miRNAs according to the fold change (high pollution/low pollution). Arbitrary we used a cut-off ≥ 1.5-fold change to select the microRNAs to be analyzed throughout the study. Differences between groups were then analyzed using Mann-Whitney U test or Kruskal-Wallis test depending on the number of groups.
In patients presenting with MI (STEMI or NSTEMI), the correlation among air pollutants was assessed by the Spearman test. Results were shown as a correlation matrix. Furthermore, in all patients, the correlation of air pollutants with immune cells was assessed by Spearman test.
Ethical approval. This study design complied with the recommendations of the Helsinki declaration for investigation with human subjects and was approved by the Ethics Committee of La Princesa University Hospital, Madrid.
Informed consent. All patients provided informed consent.

Results
Characteristic of the recruited population. A total of 249 consecutive patients (139 STEMI, 53 NSTEMI and 57 stable angina) were included. Characteristics of the population are summarized in Table 1. Compared with the other groups, patients presenting with STEMI were younger and the inclusion episode was the debut of CAD. Hypertension, dyslipidemia and diabetes were more common in the stable angina group, while active smoking was more frequent in STEMI group. Multivessel disease was more prevalent in NSTEMI group. Otherwise, no differences were observed among the groups in terms of PM 2.5 short-term exposure. Circulating inflammatory cell analysis. To determine the association of air pollutant exposure and the immune response, we performed correlation analysis with different subsets of T lymphocytes in the whole cohort. No association between total CD4 + T cells and PM 2.5 were detected. However, a negative correlation between CD4 + CD69 + T cells and PM 2.5 exposure was observed (r = − 0.18, p = 0.01) ( Fig. 2A). Interestingly, although PM 2.5 in the whole cohort was not associated with total number of T cells, we observed a negative association with the percentage of Treg CD69 + T cells (r = − 0.15, p = 0.04) ( Fig. 2A). We wonder whether these air pollution-associated changes were occurring in the different clinical presentation of atherosclerosis disease. Remarkably, the reduction of both CD4 + CD69 + and Treg CD69 + T cells was observed in NSTEMI and STEMI patients but not in stable angina (Fig. 2B). Regarding T cells producers of IL-22 and IL-17 no association was detected with PM 2.5 in the whole cohort or the different clinical presentations.
In addition, associations of NO, NO 2 , O 3 , CO and SO 2 levels with the immune response was explored. High levels of CO were associated with an increase of the percentage of peripheral blood CD4 + T cells (r = 0.27, p = 0.0002) and, specifically, with the percentage of T cells producers of IL-22 (r = 0.27, p = 0.0005) (Fig. 3A). Moreover, exposure to high levels of CO was also associated with a high number of CD4 + IL-22 + cells per ml of blood (r = 0.32, p = 0.0001) (Fig. 3A). A weaker but significant correlation of CO exposure with the numbers of CD4 + IL-17 + cells per ml of blood was observed (r = 0.17, p = 0.02 Fig. 3A). On the contrary, the expression of the anti-inflammatory CD69 receptor on total CD4 + T cells and CD4 + CD25 + Foxp3 + regulatory T cells (Treg) showed a negative correlation with CO exposure (r = − 0.22, p.003 and r = − 0.20, p = 0.007 respectively) (Fig. 3B). Regarding the SO 2 , our data showed a weak but significant negative association with the total count of peripheral blood leucocytes (r = − 0.19, p = 0.003, Fig. 3C). Nevertheless, no significant associations were detected between the analyzed populations and the rest of air pollutants evaluated (NO, NO 2 and O 3 ). miRNA analysis. A total of 31 patients were selected for this analysis, 17 of them exposed to low pollution and 14 exposed to high pollution. Eight out of 31 patients had diagnosis of stable angina, 9 NSTEMI and 14 STEMI. Assuming that the number of microRNAs differentially expressed among different groups would be small, we first calculated the fold change of miRNA expression between high and low pollution. We identify 22 microRNAs with a fold change ≥ 1.5 However, only 9 out of 22 microRNAs showed statistically significant differences between patients exposed to high levels of pollution compared to the exposure to low levels (Table 2).
Interestingly, the functional profiling of the microRNA gene targets listed in Table 2 showed an enrichment mainly in biological processes associated with development and morphogenesis of cardiovascular system as www.nature.com/scientificreports/ well as in several processes of inflammatory response (Table 3). Indeed, more than 50% of the top 50 biological processes enriched in genes regulated by our list of miRNAs correspond to the cardiovascular system biology and immune response pathways.
Subsequently, we analyzed the changes in circulating miRNAs associated to pollution in each clinical presentation. Interestingly, we observed that expression of miR-let-7f-5p was increased in NSTEMI and STEMI patients exposed to high levels of pollution, while no significant changes were detected in patients with stable angina (Fig. 6A). In addition, exposure to high pollution was significant associated with higher levels of miR-423-3p and miR-146a-5p only in STEMI patients (Fig. 6B).

Discussion
This is the first study systematically assessing the biological changes in peripheral blood CD4 + T cell and circulating miRNAs, associated with short term exposure to air pollutants in patients with MI. CAD characterizing our population represents a relevant difference with the previous reports, since in most studies healthy participants had been recruited to assess the biological response to air pollutants 11 . Nevertheless, healthy participants and patients with CAD may have a different response to exogenous stressors. Besides, previous studies exploring biomarkers in MI patients do not usually include pollution as a variable 13 . www.nature.com/scientificreports/ www.nature.com/scientificreports/ www.nature.com/scientificreports/ Plaque destabilization may lead to a wide range of clinical presentations, from asymptomatic plaque rupture or erosion to occlusive atherothrombosis. Thrombogenicity, inflammation, oxidative stress and endothelial function have a large variability in response to exogenous and endogenous stimulus leading to a multifaceted vulnerability milieu that eventually explains the final clinical presentation resulting from acute plaque destabilization 14 .
Notably, air pollution appears to participate in all the stages of this vulnerable state 8 .
In the circulating white cell analysis, CO was associated both with an increased number of CD4 + cells producers of IL-17 and IL-22. These findings are consistent with a previous study which analyzed white blood cell changes in patients with chronic respiratory disease exposed to CO in the previous 24 h, observing increased lymphocytes counts 15 . Interestingly, IL-22 and IL-17 expression are induced by the activation of aryl hydrocarbon receptors, a transcription factor that is a target for pollution 16 . High SO 2 short-term exposure was associated with reduction in the total leucocytes count. In a previous animal model using inhaled SO 2 , this finding was also reported 17 . Remarkably, both PM 2.5 and CO short-term exposures were associated to CD69 + Treg cells reduction. In this regard it is important to highlight that the immunosuppressive activity of Treg cells is increased in those cells expressing CD69 18 . Moreover, PM 2.5 exposure in STEMI patients was associated with a reduction in Treg cells. These findings are relevant as numerous studies showed that Tregs deficiency or dysfunction are associated with the development of atherosclerosis 19,20 that may be related to the protective effect of Tregs on PM-induced inflammatory response 21 . There is no previous data about the effects of air pollution on Treg cell in patients with CAD. Nevertheless, a similar decreased expression of FOXP3 has been described in atopic children exposed to air pollutants 22 . Recently, our group identified the protective role of CD69 for atherosclerotic disease, and peripheral leucocytes from subclinical atherosclerosis individuals express low level of this molecule 23 . In this regard, our current data strongly suggest that exposure to air pollutants is associated with a reduction in CD69 in T cells.
In the miRNA analysis, we found several miRNA altered by PM 2.5 short-term exposure. Interestingly, all of them were linked to gene expression involved in cardiovascular or immune system processes participating in the atherosclerotic disease. Remarkably, few of them were specifically modified only in patients presenting with an acute MI: miR-let-7f-5p was increased in patients with STEMI or NSTEMI, while miR-423-3p and miR-146a-5p were only increased in the STEMI group.
The let-7 family is highly expressed in the cardiovascular system, being miR-let-7f related to angiogenesis, ischemia, arrhythmia and heart development 24 . Recently, upregulation of miR-let-7f-5p has been documented in activated platelets 25 . In a previous report, stress cardiomyopathy showed higher levels of miR-let-7f-5p compared to STEMI, arguing that the observed difference, among others, may be related to alteration of the microcirculation 13 . However, no data of pollutant exposure have been reported.
In a large Chinese cohort of general population in primary prevention, lower levels of circulating miRNA-423-3p predicted acute MI in the follow up, performing better than hs-CRP. Unfortunately, the study was missing pollution data. In addition, an in vitro research in rat cardiac fibroblasts documented a possible involvement of miRNA-423-3p in the ischemia-reperfusion injury 26 .
The observed upregulation of miR-146a was previously reported in steelworkers after short term exposure to PM 2. 5 27 . In addition, miR-146a-5p was previously proposed as a biomarker of PM-induced impaired inflammatory response 28 . miR-146a is a cytokine-responsive miRNA induced by TNF-α and interleukin-1β. In experimental atherosclerosis its overexpression inhibits cytokine responsiveness of endothelium, suggesting that it could be part of a negative feedback mechanism limiting endothelial cell inflammatory signaling 29 . Remarkably, miR-146a is a crucial regulator of T reg suppressive function preventing the conversion of Tregs in IFNγ-producing Th1-like cells 30 . In addition, miRNA-146a regulates the maturation process and pro-inflammatory cytokine secretion by targeting CD40L in oxLDL-stimulated dendritic cells 31 . Despite the miR-146a upregulation, its anti-inflammatory effect may be ineffective in polluted areas as PM 2.5 short-term exposure can silence genes by DNA methylation of CpG islands of promoters 32,33 . At any rate, the concomitant Tregs reduction and miR-146a increase observed in STEMI patients with short-term exposure to PM 2.5 seems to be strictly related and may represent a characterizing pattern of pollution-associated STEMI. . Target genes of differentially expressed genes in CAD patients exposed to high levels of PM 2.5 . Functional miRNA targets associated to both Cardiovascular System and Immune System are shown. Different colors are used to indicate the targets that are regulated by each miRNA. From all genes identified in miRTarBase as targets of microRNAs listed in Table 2, only those with functioonal support were selected to perform the enrichment analysis. Image created with BioRender.com. www.nature.com/scientificreports/ Altogether, our results strongly suggest a modulating effect of short-term exposure to air pollutants on circulating immune cells and miRNA expression in patients with CAD. These changes may participate in the increased risk of STEMI and worse outcomes in people exposed to air pollutants.
Our study has several limitations that should be acknowledged. Despite the significant differences, due to the small sample size, our study should only be considered as hypothesis generating and results should be confirmed in larger studies. Moreover, a selection bias cannot be excluded. In addition, confounder effects of lipids, diabetes status, smoking, nutrition, social status and cardiovascular drugs use, could not be rule out.
Furthermore, PM 2.5 components may vary significantly depending on the different sources of pollution and the specific climate conditions of the geographical area 34 . Madrid is the most populous city in Spain (6 million people in the urban area) with a climate of transition between the Mediterranean and the cold semi-arid Figure 5. Differential expression of miRNAs in plasma samples from CAD patients exposed to low and high levels of PM 2.5 . Box and whiskers Min to Max plots showing plasma levels of microRNAs from CAD patients (n = 31) exposed to low or high levels of PM 2.5 . Including all clinical presentations in the analysis, high PM 2.5 short-term exposure was associated with (A) increased and (B) decreased miRNA expression. Differences were analyzed using Mann-Whitney U test. www.nature.com/scientificreports/ climate. Like other capitals in developed countries, its main sources of air pollution are motorized road traffic, apart from commercial and residential heating. Therefore, our results could not be extrapolated to areas with different characteristics. In addition, polymorphisms, haplotypes and variability in plasma levels of C-reactive protein, fibrinogen, IL-6 may have altered the pro-inflammatory response of air pollution in myocardial infarction patients 35 . Finally, because their role in atherosclerosis, innate immune cells should also be explored in further studies. Furthermore, since all participants had an established coronary artery disease, the observed inflammatory changes in response to air pollutants cannot be generalized to healthy individuals. Despite these limitations, our study provides novel unique insights on the mechanisms involved in the pathogenesis of acute MI associated with short-term exposure to air pollutant. Further studies are warranted for a more complete understanding of the physiopathology of this process in order to inform clinical decisions and develop prevention strategies aimed to reduce the risk of MI in patients exposed to air pollution.
In conclusion, our study identifies circulating inflammatory cells and the miRNA changes of acute MI related to short-term exposure to air pollutants. Specifically, STEMI related to PM 2.5 short-term exposure is associated to specific changes involving CD4 + CD25 + Foxp3 + Treg cells and miR-146a-5p.

Data availability
Data sharing will be considered upon reasonable request including a detailed research plan with the corresponding author.