J Appl Genet 51(3), 2010, pp. 353–368 Original article Gene expression profiling in peripheral blood nuclear cells in patients with refractory ischaemic end-stage heart failure S. Szmit1,2, M. Jank3, H. Maciejewski4, M. Grabowski1, R. Glowczynska1, A. Majewska3, K. J. Filipiak1, T. Motyl3, G. Opolski1 1First Department of Cardiology, Medical University of Warsaw, Poland 2Department of Oncology, Military Institute of Medicine, Warsaw, Poland 3Department of Physiological Sciences, Faculty of Veterinary Medicine, Warsaw University of Life Sciences, Warsaw, Poland 4Institute of Computer Engineering, Control and Robotics (I-6), Wroc³aw University of Technology, Wroc³aw, Poland Abstract. Functional analysis of up- and down-regulated genes might reveal whether peripheral blood cells may be considered as a material of diagnostic or prognostic value in patients with end-stage heart failure (HF). The aim of the present study was to compare the transcriptomic profile of peripheral blood nuclear cells from 6 male patients with ischaemic end-stage HF with those of 6 male patients with asymptomatic cardiac dysfunction. The expression of genes in peripheral blood nuclear cells in both groups of patients was measured using whole-genome oligonucleotide microarrays utilizing 35 035 oligonucleotide probes. Microarray analyses re- vealed 130 down-regulated genes and 15 up-regulated genes in the patients with end-stage HF. Some of the down-regulated genes belonged to the pathways that other studies have shown to be down-regulated in cardiomyopathy. We also identified up-regulated genes that have been correlated with HF severity (CXCL16) and genes involved in the regulation of expression of platelet activation factor receptor (PTAFR, RBPSUH, MCC, and PSMA7). In conclusion, the identification of genes that are differentially expressed in peripheral blood nuclear cells of patients with HF supports the suggestion that this diagnostic approach may be useful in searching for the molecular predisposition for development of severe refractory HF in patients with post-infarction asymptomatic abnormalities and remodelling of the left ventricle. These results need further investigation and validation. Keywords: asymptomatic myocardial dysfunction, end-stage heart failure, gene expression profiling, peripheral blood nuclear cells. Introduction of the results of several studies. Such meta-analyses are valuable since they provide val- The last 10 years in biological sciences can be de- idation of microarray experiments and indicate the scribed as an era of functional genomics. Unsur- potential marker genes across different forms of prisingly, it has resulted in many publications heart diseases (Kuner et al. 2008; Asakura and concerning gene expression profiling in the heart Kitakaze 2009). affected by various pathological processes The appearance of HF symptoms may be asso- (Kittelson et al. 2005; Grzeskowiak et al. 2003; ciated with ischaemia, arrhythmias, valvular dis- Barrans et al. 2002; Tan et al. 2002; Ellinghaus orders, increased filling pressure, and systemic et al. 2005; Rajan et al. 2006). An increasing num- resistance (increase in blood pressure). Chronic ber of such papers, as well as free access to the re- obstructive pulmonary disease, renal failure, dia- sults of hybridization of many microarrays, also betes, anaemia or obesity may induce or intensify provide the opportunity to perform meta-analyses the symptoms of HF. In other cases, progression of Received: October 19, 2009. Revised: March 2, 2010. Accepted: April 26, 2010. Correspondence: S. Szmit, First Department of Cardiology, Medical University of Warsaw, Banacha 1a, 02–097 Warsaw, Po- land; e-mail: [email protected] 354 S. Szmit et al. HF is not related to identifiable causes. It is possi- low-up after ACS. These patients had a good ble, however, that there are disorders at a molecu- quality of life and good exercise tolerance. lar level that are responsible for such an Group 2: six male patients with refractory unexplained progression of HF. HF seems to be an end-stage HF and symptoms of HF that appeared ideal candidate disorder for genomic experiments during 2 years of follow-up after ACS. These pa- aimed at elucidating the complex causes of disease tients had a significant exercise intolerance and progression, which might answer the question as were candidates for heart transplant. to why some patients develop refractory end-stage None of the patients had had additional indica- HF and others do not. However, the most impor- tions for coronary intervention during 2 years of tant limitation of the wider use of microarrays for follow-up after ACS. On ECG, all of the patients investigation of the aetiology and patho- had sinus rhythm without arrhythmia. None of the physiology of HF is the necessity for heart tissue patients had pulmonary disease, pulmonary vascu- sampling in order to obtain material for genomic lar disease, or any other significant internal dis- analyses. It is not always possible to perform a bi- ease: diabetes, renal insufficiency, anaemia, opsy in a patient with myocardial dysfunction or depression, obesity or cachexia. There was no dif- damage. For this reason, we sought to establish ference between the 2 groups in levels of inflam- whether any correlation exists between end-stage mation markers (white blood count and C-reactive HF and gene expression in peripheral blood nu- protein). The patients in both groups received opti- clear cells, which represent easily obtainable ma- mal pharmacological treatment, had good blood terial for further analysis. pressure control and satisfactory cholesterol and glucose levels. The individual clinical characteris- tics of the 12 patients included in the study are de- Materials and methods scribed in Table 1. Patients Ethics statement The study comprised 2 groups of patients after The study was approved by the local Ethics Com- acute coronary syndrome (ACS) treated by pri- mittee of the Medical University of Warsaw, and mary percutaneous transluminal coronary all participants gave written informed consent. Table 1. Individual clinical characteristics of heart failure (HF) in all 12 patients involved in the study Patients of group 1 (asymptomatic cardiac struc- Patients of group 2 (refractory end-stage heart Characteristics tural abnormalities, AHA/ACC stage B) failure, AHA/ACC stage D) –1 –1 VO2 peak (mL kg min ) 33.2 31.2 30.8 32.2 32.6 31.8 11.5 12.0 9.9 9.7 11.7 12.0 VE-VCO2 slope 26 25 23 21 25 25 35 42 40 34 35 38 Ejection fraction (%) 64 50 55 50 55 60 30 18 28 25 30 20 Natriuretic peptide type B 59 61 37 46 50 88 1487 560 1400 750 634 1200 (pmol/mL) Peak plasma TnI levels 78 37 80 51 36 53 59 70 48 63 42 36 during ACS (ng/mL) Body mass index (kg/m2) 28.4 24.8 26.4 25.6 28.1 25.3 25.6 28 30.9 25.1 27.1 28.6 White blood count 8.52 6.55 4.12 8.18 7.0 6.81 5.10 8.91 4.54 9.80 6.23 5.49 (103/µL) C-reactive protein (mg/dL) all < 0.3 all < 0.3 Age (years) all 50–55 all 50–55 Pharmacological treatment ACEIs, beta-blockers, antiplatelet therapy, statin ACEIs, beta-blockers, antiplatelet therapy, (identical within group) statin, diuretics (aldosterone antagonists, furosemide) ACS = acute coronary syndrome; ACEIs = angiotensin-converting enzyme inhibitors; AHA/ACC = American Heart Association / American College of Cardiology; Tnl = troponin subunit I; VO2 = oxygen consumption at peak exercise; VE-VCO2 slope= slope ratio of regression be- tween ventilation and carbon dioxide angioplasty (PTCA) of the descending left artery RNA isolation, validation, labelling, and with stent implantation: hybridization Group 1: six male patients with cardiac struc- Blood samples were obtained from the cephalic tural abnormalities or remodelling who had not vein of the forearm into PAXGene tubes developed HF symptoms during 2 years of fol- (QIAGEN, USA). Duplicate samples were ob- Blood cell gene expression in heart failure 355 tained from each patient. Total RNA from periph- based on some group comparison measures. For eral blood nuclear cells was isolated using a this purpose, samples from patients with end-stage PAXgene Blood RNA kit (QIAGEN, USA). Iso- HF were denoted as S11-S14, and samples from lated RNA samples were dissolved in RNase-free patients with asymptomatic cardiac dysfunction water, and RNA quantity was measured with the were denoted as S21-S24 (Figure 1). Prior to gene use of NanoDrop (NanoDrop Technologies, ranking, expression values were logged and each USA). Samples with an adequate amount of RNA sample S11-S24 was median-centered. In order to were treated with DNase I to eliminate DNA con- obtain reliable results, we ranked the genes using tamination, and then purified using RNeasy 3 statistical tests: Wilcoxon rank test, t-test, and MiniElute Cleanup Kit (QIAGEN, Germany). fold-change ratio. Standard gene ranking methods de- Analysis of final RNA quality and integrity was fine a gene as differentially expressed when the follow- performed with a BioAnalyzer (Agilent, USA). ing conditions are met for the gene: (1) Wilcoxon After validation, equal amounts of total RNA from P < 0.05; (2) t-test P < 0.05; and (3) fold change > 1.7, all patients within each group were pooled. Pooled total RNA (10 µg) was reverse-transcribed by us- where: ing a SuperScript Plus Indirect cDNA Labelling fold change= exp(abs(log(abs(µ1))-log(abs(µ2)))), kit (Invitrogen, USA) according to the manufac- and µ1 and µ2 are mean values of expression of turer’s protocol. Single-strand cDNA was labelled the gene in groups 1 and 2, respectively. with Alexa 555 or Alexa 647 dyes (Invitrogen, USA) in order to obtain a dye-swap approach. The Real-time PCR (Q-PCR) efficiency of dye incorporation was measured by Expression of PTAFR and RBPSUH genes was using NanoDrop.