Early Elevation of Systemic Plasma Clusterin after Reperfused Acute Myocardial Infarction in a Preclinical Porcine Model of Ischemic Heart Disease

Clusterin exerts anti-inflammatory, cytoprotective and anti-apoptotic effects. Both an increase and decrease of clusterin in acute myocardial infarction (AMI) has been reported. We aimed to clarify the role of clusterin as a systemic biomarker in AMI. AMI was induced by percutaneous left anterior artery (LAD) occlusion for 90 min followed by reperfusion in 24 pigs. Contrast ventriculography was performed after reperfusion to assess left ventricular ejection fraction (LVEF), left ventricular end diastolic volume (LVEDV) and left ventricular end systolic volume (LVESV) and additional cMRI + late enhancement to measure infarct size and LV functions at day 3 and week 6 post-MI. Blood samples were collected at prespecified timepoints. Plasma clusterin and other biomarkers (cTnT, NT-proBNP, neprilysin, NGAL, ET-1, osteopontin, miR21, miR29) were measured by ELISA and qPCR. Gene expression profiles of infarcted and remote region 3 h (n = 5) and 3 days (n = 5) after AMI onset were analysed by RNA-sequencing. AMI led to an increase in LVEDV and LVESV during 6-week, with concomitant elevation of NT-proBNP 3-weeks after AMI. Plasma clusterin levels were increased immediately after AMI and returned to normal levels until 3-weeks. Plasma NGAL, ET-1 and miR29 was significantly elevated at 3 weeks follow-up, miR21 increased after reperfusion and at 3 weeks post-AMI, while circulating neprilysin levels did not change. Elevated plasma clusterin levels 120 min after AMI onset suggest that clusterin might be an additional early biomarker of myocardial ischemia.


Introduction
Early diagnosis and therapy of acute myocardial infarction (AMI) are essential to reduce infarct size and improve prognosis [1][2][3]. Current gold standard of early diagnosis of AMI is high-sensitivity cardiac troponin (hs-cTn), [4] even though this marker has several limitations [5]. It has restricted specificity with an increase in non-ischemic injuries, such as pulmonary oedema, or embolisation, anticancer treatment or chronic kidney disease [6]. Therefore, further research to identify novel

Other Biomarkers Associated with Myocardial Ischemic Injury
Troponins are currently gold standard in diagnosis of AMI. We measured Troponin I, type 3 and observed an increase 120 min after AMI onset

Pro-Fibrotic Plasma miR21 and miR29 after Acute Myocardial Infarction
LV remodelling and cardiac fibrosis are highly relevant pathological processes after AMI and for a variety of miRNAs, an association with cardiac fibrosis has been observed. Relative expression of both miR21 and miR29 increased at 3 weeks compared to baseline levels and relative expression of miR21 was also significantly increased post reperfusion compared to baseline (mean ± SD: pre vs. NT-proBNP is an established marker of myocardial dysfunction and prognosis in chronic heart failure [43]. We assessed plasma NT-proBNP levels, additionally to cMRI to evaluate chronic LV dysfunction after AMI. We observed a significant increase after 3 weeks (median [IQR]: pre vs. 3 w: 112.3 [71.3; 167.9] vs. 189.9 [109.9; 387.9] pg/mL, p = 0.02, Figure 3B).

Association between Clusterin and Left Ventricular Function Parameters, Infarct Size and Biomarkers
We correlated clusterin 120 min after AMI onset with LV function parameters assessed by contrast ventriculography. Plasma clusterin post AMI was significantly associated with LVEF (r = −0.69, p = 0.0002, Figure 5A) and LVESV (r = 0.52, p = 0.0092, Figure 5B), but not LVEDV (r = 0.16, p = 0.46, Figure 5C). Significant negative correlation was also found between the changes of clusterin from pre-AMI to post-AMI with 3 d cMRI LVEF (r = −0.544, p = 0.036). Clusterin levels at day 3 and week 3 follow-up did not correlate with the 3 day and 6 week left ventricular function parameters and infarct size.

Association between Clusterin and Left Ventricular Function Parameters, Infarct Size and Biomarkers
We correlated clusterin 120 min after AMI onset with LV function parameters assessed by contrast ventriculography. Plasma clusterin post AMI was significantly associated with LVEF (r = −0.69, p = 0.0002, Figure 5A) and LVESV (r = 0.52, p = 0.0092, Figure 5B), but not LVEDV (r = 0.16, p = 0.46, Figure 5C). Significant negative correlation was also found between the changes of clusterin from pre-AMI to post-AMI with 3 d cMRI LVEF (r = −0.544, p = 0.036). Clusterin levels at day 3 and week 3 follow-up did not correlate with the 3 day and 6 week left ventricular function parameters and infarct size. Additionally, elevated, clusterin concentration at 120 min post AMI-onset did not correlate with plasma levels of any other assessed biomarker, including Troponin I, type 3 (r = −0.21, p = 0.33) and miR21 (r = −0.31, p = 0.14).

Transcriptomic Profiling
Venn diagrams ( Figure 6A,B) reveal a notable overlap, but also differences between up-and down regulated genes 3 hours and 3 days after AMI onset both in AMI and remote region. Clusterin Additionally, elevated, clusterin concentration at 120 min post AMI-onset did not correlate with plasma levels of any other assessed biomarker, including Troponin I, type 3 (r = −0.21, p = 0.33) and miR21 (r = −0.31, p = 0.14).

Transcriptomic Profiling
Venn diagrams ( Figure 6A,B) reveal a notable overlap, but also differences between up-and down regulated genes 3 h and 3 days after AMI onset both in AMI and remote region. Clusterin was significantly upregulated in AMI and remote tissue both after 3 h and 3 days. Functional clustering focussed on clusterin as a central gene ( Figure 6C,D) showed a strong connection to genes associated with angiogenesis, complement activation (inflammatory response to myocardial ischemia), apoptotic processes, TGFβ signalling, amyloid beta homeostasis and other genes involved in protein stabilization and chaperoning. The majority of those genes are downregulated. A similar pattern was observed 3 h after AMI onset in the remote region ( Figure 6E). A cluster of mostly upregulated genes associated with angiogenesis, complement activation and cardiac muscle cell proliferation was identified in the remote region on day 3 ( Figure 6F). A direct comparison of genes involved in cardiac muscle function and apoptosis in AMI and remote region both after 3 h and 3 days is given in the Supplementary Material online (Table S1).

Discussion
In the present study, we demonstrate increased plasma clusterin levels after ischemic myocardial injury (120 min after AMI onset) that returned to normal levels already after 3 days, suggesting that it can be a similar and additionally early and high-sensitive biomarker of acute myocardial ischemia as hs-cTn, which increases 2-3 h after ischemia onset and remains elevated

Discussion
In the present study, we demonstrate increased plasma clusterin levels after ischemic myocardial injury (120 min after AMI onset) that returned to normal levels already after 3 days, suggesting that it can be a similar and additionally early and high-sensitive biomarker of acute myocardial ischemia as hs-cTn, which increases 2-3 h after ischemia onset and remains elevated several days after AMI. Furthermore, we observed a correlation of plasma clusterin 120 min after AMI onset with LVEF and LVESV. The failure of correlation between clusterin (a molecular chaperone) and TnI levels might be explained by the different intracellular localization, subcellular compartments and molecular interactions, and warrants further investigations.
In the current clinical setting, biomarkers play an important role in diagnosis and early treatment modalities of AMI. Even though hs-cTn has improved diagnosis and management of AMI patients, they still feature several drawbacks such as limited specificity and a delay in measurably increased values [6]. Diagnosis of AMI with hs-cTn might be difficult in patients suffering from chronic renal failure, subarachnoid haemorrhage, acute pulmonary embolism, chronic obstructive pulmonary disease, acute noncardiac critical illness and after strenuous exercise. Furthermore, an elevation of hs-cTn by other cardiac causes such as advanced heart failure, direct myocardial trauma, acute pericarditis, acute inflammatory myocarditis and tachycardia has been observed [44]. This clearly indicates that there is a dearth of biomarkers that could identify myocardial ischemia within the first few hours after the onset of AMI additionally to hs-cTn. In our study, acute myocardial ischemia did not alter the plasma neprilysin and osteopontin concentration, while NGAL, ET-1 and miR29 increased at week 3 post-AMI, suggesting a role in development of cardiac remodelling. Interestingly, similar to clusterin and TnI, miR21 increased also immediately after infarction. miRNAs are key regulators of cardiovascular diseases and a variety of plasma miRNAs have been identified as stable circulatory biomarkers [45][46][47]. miR21 has previously been reported to be elevated in AMI patients and significantly correlate with cTnI and CK-MB [48]. Even though miRNAs are highly stable and rapidly released from damaged cells, the majority of miRNAs is neither disease nor organ specific. However, using a miRNA panel or combining (individual) miRNAs with well-known biomarkers such as hs-cTn might improve their diagnostic accuracy [49]. Interestingly, no correlation could be found between any of the early ischemiaor late remodelling-related biomarkers with clusterin.
In contrast with the other circulating factors (except TnI and miR21), clusterin proved to be a good marker for acute ischemia already 120 min after ischemia onset. This hypothesis is supported by the correlation of post AMI clusterin with LVEF and LVESV assessed by contrast ventriculography 30 min after reperfusion onset. Nevertheless, the mechanism of clusterin release can only be speculated. Clusterin is a heat shock protein-like intra-and extracellular chaperone and its expression is stimulated by cellular stress and tissue injury (e.g., ischemia, inflammation, apoptosis, oxidative stress, heat stress and ionising radiation) [7,9,50]. Extracellular clusterin stabilizes stressed proteins in a folding-competent state [8]. By clearing aggregating protein species and dead cells clusterin exerts anti-inflammatory and cytoprotective effects [9]. In the myocardium it protects the cardiomyocytes against apoptosis and promotes angiogenesis [51]. Our RNASeq data showed a significant upregulation of clusterin in the heart after 3 h and 3 days. Whereas gene expression compared to control decreased after 3 days in AMI tissue, we could observe an increase after 3 days in the remote region. Functional clustering of deregulated genes revealed an association of genes associated with apoptosis, inflammatory response to myocardial ischemia and angiogenesis. Immunohistochemical staining of human hearts after AMI showed increased expression of clusterin in the infarct zone at an early time point and increased expression in the peri-infarct zone in older infarct tissue, although not in healthy hearts [52]. Pavo et al. described increased clusterin expression in the infarct zone already five hours after the onset of myocardial ischemia and varying expression in the remote area, indicating a role in cardioprotection and restoring of cell function with a possible mediator role for intrinsic remote conditioning [51].
Clusterin expression is also associated with diabetes type II and high cholesterol levels, both being well known risk factors for atherosclerosis and AMI [8]. In our AMI model (in pigs not suffering from atherosclerosis, high cholesterol or diabetes) we could examine clusterin expression in acute myocardial ischemia without any confounding factors. Our results indicate that the increase in clusterin concentration immediately after AMI might be strongly associated to myocardial necrosis and not be caused by any underlying factors such as atherosclerosis and diabetes. However, a contributing part of those risk factors in humans cannot be excluded. This may also be the reason for sustained elevated clusterin levels in AMI patients that was previously reported [14]; however, this was not observed in our experiments with healthy animals. As already stated, hs-cTn levels are strongly influenced by a variety of cardiac and non-cardiac diseases that are common in patients at risk of AMI. Using hs-cTn as a single marker in patients with acute thoracic pain symptomatic as a tool to diagnose AMI might result in inconclusive results. We believe that plasma clusterin may serve as an additional biomarker in these patients further improving diagnostic accuracy of diagnostic standards.
Even though we provide evidence that plasma clusterin levels are regulated during controlled myocardial ischemia, several limitations to our study should be mentioned. Firstly, it is an observation study and no causality effects can be concluded, a mechanistic explanation of clusterin in AMI needs to be elaborated in further studies. Secondly, we measured clusterin levels in a small group of pigs; clinical applicability needs to be evaluated in a larger patient cohort. Third, we investigated clusterin dynamics only during a brief period after myocardial infarction; long term dynamics and a possible prognostic factor for mortality should be the subject of additional studies. Fourth, the current commercially available clusterin ELISA kits showed a relatively large scatter of data, which should be further refined.

Animals and Experimental Design
Domestic pigs (n = 39, weight 30-35 kg, female) underwent percutaneous coronary intervention (PCI) in order to induce catheter-based reperfused AMI. According to ESC guidelines [53,54] animals were premedicated with 250 mg aspirin and 300 mg clopidogrel and received daily doses of 100 mg aspirin and 75 mg clopidogrel during the follow up period. Functional assessment of the left ventricle and serial biomarker measurements were performed in 24 pigs, while myocardial gene expression of selected biomarker was performed 3 h and 3 days after AMI onset in five pigs of each time point.
Prior to left anterior descending artery (LAD) occlusion, the pigs received intramuscular injection of 12 mg/kg ketamine hydrochloride, 1 mg/kg xylazine and 0.04 mg/kg atropine as anaesthesia. Anaesthesia were deepened with isoflurane and O 2 via mask and maintained with 1.5-2.5 vol% isoflurane, 1.6-1.8 vol% O 2 and 0.5 vol% N 2 O via intratracheal tube. After induction of general anaesthesia, access to the right femoral artery was obtained through surgical preparation of the artery under sterile conditions and a 6-F introducer sheath (Medtronic, Minneapolis, MN, USA) was inserted. In total, 10,000 IU of heparin sodium were administered via the femoral sheath, and baseline haemodynamics were recorded. Selective angiography of the left coronary artery was performed by using a 6F guiding catheter (Medtronic, Minneapolis, MN, USA) with regular contrast media (Ultravist, Bayer, Leverkusen, Germany). After a baseline angiogram was analysed, a balloon catheter (2.75 m diameter, 8 mm length) (Abbot Vascular) was placed after the origin of the second diagonal branch. To induce AMI, the balloon was inflated with 5 atm for 90 min followed by deflation of the balloon resulting in reperfusion. Wounds were closed and anaesthesia was terminated by withdrawal of isoflurane. In this study, 1 g metamizole was applied intramuscularly (i.m.) as analgesia. Furthermore, 100 mg benzathine benzylpenicilline, 100 mg procaine benzylpenicillin and 200 mg dihydrostreptomycin-sulphate was given i.m. as antibiotic shielding. Heart rate, arterial blood pressure, electrocardiography, O 2 saturation and temperature were monitored throughout the procedure.
The experiments were conducted at the Institute of Diagnostics and Oncoradiology, University of Kaposvar, Hungary. All animal facilities met the standards of the American Association for Accreditation of Laboratory Animal Care. Animal investigations were executed in accordance with the "Position of the American Heart Association on Research Animal Use" as adopted by the American Heart Association (AHA) on 11 November 1984. The study was approved by the Ethics Committee on Animal Experimentation at the University of Kaposvar, Hungary (EC: SOI/31/26-11/2014, approval date: 25 February 2014).

Blood Sampling
Peripheral blood samples were collected before occlusion (pre-AMI), and post-AMI (immediately before recovery from the anaesthesia, e.g., 120 min after start of coronary occlusion and 30 min after start of reperfusion), at 3 days and 3 weeks post AMI. Blood was centrifuged at 2000× g for 10 min and stored at −20 • C until further analyses were performed.

Measurement of Myocardial Necrosis and Functional Parameters by Contrast Ventriculography and Cardiac MRI
Biplane contrast ventriculography was performed 30 min after start of reperfusion, before the end of anaesthesia. Fifty mL contrast medium was infused by an injection pump at a rate of 12 mL/s via a 5F pig-tail catheter. LV volumina (LVEDV, LVSESV and LVEF) were calculated off-line by using the area-length methods (Quantcor LVA, Siemens, Germany). Magnification correction was calculated from the known internal diameter of the pig-tail catheter and the known distance between the mid chest of the animal and the radiography equipment, documented during the procedure. The end-diastolic and end-systolic contours were digitalized and traced automatically and LVEF was calculated.
At 3 days and 6 weeks after artificial myocardial infarction, cMRI + late enhancement (LE) was performed to assess myocardial necrosis, LVEF, LVEDV and LVESV. The cMRI + LE acquisition method and analyses have been described previously [55]. In accordance with the ethical principle of the 3 R (replace, reduce, refine) pre-AMI cMRI has not been performed as baseline left ventricular function parameters in pigs are similar to healthy humans and have been published previously [56,57].

Transcriptomic Profiling
Detailed information is described in the Supplementary Material online and has been published previously [58]. Myocardial samples were obtained from the AMI and remote region, the latter was obtained from the opposite wall of the AMI region (mid lateral wall). Briefly, extracted total RNA of myocardial samples obtained on 3 h and 3 days after myocardial infarction onset were subjected to mRNA deep sequencing using the Illumina platform (San Diego, CA, USA). For mRNA fragmentation and enrichment NEB Next Poly(A) mRNA Magnetic Isolation Module (NEB, Ipswich, MA, USA) was used. Fragmented and primed mRNAs were reverse transcribed to cDNA. The NEBNext Ultra Directional RNA Library Kit (NEB, Ipswich, MA, USA) was used for cDNA library synthesising and enrichment. Finally, sequencing was performed on the HiSeq 2500 platform (mean depth: 15-20 million paired-end reads per sample) at the Core Facility Genomics (Medical University of Vienna, Vienna, Austria). Results were mapped to the pig transcriptome and analysed for statistically significant changes of individual genes. For analysis of biological relevance, groups were compared and significantly upor downregulated genes were functionally clustered.

PCR
Total RNA was isolated from plasma using the miRNeasy Serum/Plasma Kit (Qiagen, Hilden, Germany). The RNA quantity and quality were measured with a nanodrop machine (Witec AG, Sursee, Switzerland). MiRNA was reverse transcribed to cDNA (Qiagen, Hilden, Germany) and expression was quantified by rtPCR (Applied Biosystems 7500 Real-Time PCR System, Life Technologies, Carlsbad, CA, USA). The primers for the target sequences were designed using Primer3 software version 4.1.0 (http://primer3.wi.mit.edu/primer3web_help.htm; Microsynth, Balgach, Switzerland). The relative gene expression level was calculated using the ∆Ct method (i.e., expression level relative to an endogenous control). The expression changes were calculated relative to median expression at baseline.

Statistics
Data obtained were evaluated statistically using GraphPad Prism 6 software (GraphPad Software Inc., LA Jolla, CA, USA) and IBM SPSS Statistics version 23 (SPSS Inc., Chicago, IL, USA). Mixed linear models were used to compare parametric variables and for non-parametric variables after logistic transformation. Parametric variables were expressed as mean ± standard deviation (SD) and compared Student's paired t-test. A Wilcoxon test and Friedman test were used to compare non-parametric, paired variables and expressed as median and interquartile range (IQR). Bonferroni correction was applied for multiple testing. For correlation of non-parametric variables Spearman's rank correlation was used. All tests were performed in a two-sided manner. p-values equal or below 0.05 were considered statistically significant.

Conclusions
In conclusion, we have shown that plasma clusterin levels are associated with AMI in the early phase. In contrast to previous work, we did not observe sustained elevation of clusterin; however, this may be due to the fact that we could examine clusterin dynamics after AMI isolated from any concomitant diseases that are well known to be associated with altered clusterin expression.

Conflicts of Interest:
The authors declare no conflict of interest.