Background
Ischemic heart disease is the leading cause of death globally [
1]. In the case of an acute myocardial infarction, myocardial damage caused by ischemia is exacerbated by oxygenized blood returning to the heart at reperfusion [
2]. Approaches to treat infarction should both promote reperfusion and protect myocardium from the detrimental effects of ischemia and reperfusion.
Receptor tyrosine kinases (RTK) are cell surface receptors that mediate cellular survival, proliferation, and migration. A few RTKs have been shown to be necessary for development of the heart in gene-modified mouse models. Such examples include
Erbb2 [
3,
4],
Erbb4 [
5],
Ror1 [
6], and
Ror2 [
6,
7]. Understanding of the regulation of RTK activity and expression in ischemic heart is limited to few receptors. Expression of EGFR and ERBB2 have been demonstrated to be regulated in infarcted human heart [
8,
9], and alterations in EGFR, ERBB2, ERBB4, VEGFR1, VEGFR2, IGF1R, and INSR signaling have been observed in experimental ischemia-reperfusion models [
8,
10‐
12].
Few RTKs have been investigated as targets for the treatment of experimental ischemia-reperfusion injury. Induction of constitutively active ERBB2 after infarction causes myocardial regeneration in mice [
13]. Activating ERBB4 with its ligand neuregulin-1 reduces scar size in mouse [
14] and rat [
15] infarction models. Activation of INSR by insulin infusion during reperfusion has been shown to reduce infarction size in an ischemia-reperfusion rat heart model in Langendorff perfusion system [
16]. Moreover, glucose-insulin-potassium infusion has been tested in clinical trials as a myocardial infarction treatment with mixed results [
17].
Here, we used an in silico expression analysis and a phosphoarray analysis of an in vivo pig ischemia-reperfusion injury model to screen for changes in the expression and activity of RTKs in normal vs. ischemic heart. ROR1 was identified as a receptor demonstrating activity in both screens. We show that ROR1 was expressed in human heart, in pig myocardium and in cultured mouse cardiomyocytes and rat cardiomyoblasts. In cardiomyocytes in vitro, both ROR1 expression and phosphorylation were downregulated by hypoxia. We also demonstrate that ROR1 knockdown enhanced, and treatment with its ligand, Wnt-5a, reduced the viability of cardiomyocytes. These findings suggest that ROR1 signaling may suppress survival of cardiomyocytes and that ROR1 could be further tested as a potential treatment target for the myocardial ischemic injury.
Methods
In silico transcriptomics
Affymetrix gene expression data from IST Online database (ist.
medisapiens.com; Medisapiens Ltd.) were analyzed to characterize RTK expression in samples representing healthy heart (
n = 62), acute myocardial infarction (
n = 12) or ischemic cardiomyopathy (
n = 63). Out of the 55 RTKs listed by HUGO Gene Nomenclature Committee, data were available for 49 genes in acute myocardial infarction samples and for 52 genes in ischemic cardiomyopathy samples (Additional file
1). Data were normalized by array-generation-based gene centering method [
18] and log2-transformed. Expression levels of RTKs demonstrating statistically significant differences in two-group comparisons were visualized as box plots and heatmaps using Pretty Heatmaps package (pheatmap) [
19] in RStudio [
20].
Pig model of heart ischemia-reperfusion injury
Animal experiments were approved by the Laboratory Animal Care and Use Committee of the State Provincial Office of Southern Finland (license number: ESAVI/1167/04.10.03/2011). The landrace pig myocardial samples (
n = 7) were a kind gift from Drs. Christoffer Stark and Timo Savunen. Experimental procedure has been described in detail earlier [
21]. Pigs weighed 29–43 kg. Myocardial ischemia-reperfusion injury was produced by exposing anesthetized pigs (
n = 4) to cardiopulmonary bypass with aortic cross-clamping and cardioplegic arrest for 60 min, causing global myocardial ischemia. A pediatric membrane oxygenator (Dideco 905 Eos, Dideco) was used for the bypass. Procedures were performed in the laboratory of Research Center of Applied and Preventive Cardiovascular Medicine, University of Turku, Turku, Finland. For pre-anesthesia, an intramuscular injection of 100 mg xylazin (Rompun vet, Bayer Animal Health GmbH) and 25 mg midazolam (Midazolam Hameln, Hameln pharmaceuticals GmbH) was used. For anesthesia, 20 mg boluses of propofol (PropofolLipuro, B. Braun Melsungen AG) and 150 μg phentanyl (Fentanyl-Hameln, Hameln pharmaceuticals GmbH) were administered via a cannulated ear vein, and pigs were intubated and connected to a respirator (Dräger Oxylog 3000, Drägerwerk AG), the respiratory rate set to 18–22 times/min with a tidal volume of 8–10 ml/kg using 40% oxygen. Continuous infusion of propofol 15–30 mg/kg/h, phentanyl 1.5 μg/kg/h and midazolam 100 μg/kg/h was used to maintain anesthesia. A right sided thoracotomy was performed and the ascending aorta and right atrium cannulated for the bypass. 500 ml of cold (10 °C) Modified St Thomas Hospital No II cardioplegia was used to protect the hearts during the bypass, administered via a cannula to the aortic root at the time of cross-clamping and 30 min later. Antibiotic prophylaxis (Cefuroxime 750 mg, Orion Pharma) was given preoperatively and then every 8 h. 10,000 IU of heparin (Heparin, LEO Pharma) was administered as a bolus before cannulation of the heart and this was repeated every 30 min during extracorporeal circulation. 14,000 IU of protamine sulphate (Protamin, LEO Pharma) was used to neutralize the heparin. For thrombosis prophylaxis, 20 mg of enoxaparin (Klexane, Sanofi) was administered 1 and 12 h after the surgery. 100–150 mg of lidocaine (Lidocain, Orion Pharmaceuticals) and 150–225 mg of amiodarone (Cordarone, Sanofi) were used for rhythm disorders, and 5 mg boluses of ephedrine (Efedrin, Stragen Nordic) and noradrenaline infusion (80–160 μg/h) (Noradrenalin Hospira, Hospira) were used for post-operative hemodynamic support, when needed. For post-operative analgesia, 50 mg of bupivacaine (Bicain, Orion Pharmaceuticals) was infiltrated to the wound. For monitoring of adequate ventilation and perfusion, blood gases (i-STAT, Abbott Laboratories), invasive central venous pressure, ECG and oxygen saturation were followed throughout the procedure.
After the 60-min aortic cross-clamping, hearts were reperfused and the pigs were maintained anesthetized and mechanically ventilated for 29–31 h before sacrification with intravenous injection of potassium chloride. Control samples (n = 3) were obtained from pigs used as blood donors for priming of the heart-lung machine. The control pigs underwent the same anesthetic protocol as the treatment group. Transmural left ventricle samples, collected after the sacrification, were snap-frozen and stored at − 80 °C. Troponin T levels were measured from plasma samples of ischemia-reperfusion-injured pigs, collected at the baseline and 6 and 24 h after reperfusion, by the laboratory of the Turku University Central Hospital using electrochemiluminescence immunoassay (Elecsys Troponin T high sensitive, Roche). Formalin-fixed, paraffin-embedded tissue samples were stained with hematoxylin and eosin and imaged with Zeiss AxioImager M1 microscope.
Phosphoarray analysis of RTK phosphorylation
Pig myocardial samples (280 to 460 mg) were homogenized and analyzed for phosphorylation status of 49 RTKs using the Proteome Profiler Human Phospho-RTK Array Kit (R&D Systems). Five hundred μg of protein was analyzed per sample. Receptors included in the analysis are listed in Additional file
2 B.
Array blot images were quantified by densitometry with NIH ImageJ v1.50i software. Intensity values were normalized by dividing each dot’s intensity with the sum of intensities of the whole array, allowing comparison between different samples. Data were scaled to interval 0–1 by dividing all values with the highest value. Normalized values were visualized as a bargraph (mean (SD)) and as a heatmap with Pretty Heatmaps package [
19]. Receptors with at least two quantifiable results in both sample groups were included. Receptors were clustered using maximum distance method.
Cell culture and Wnt-5a ligand treatment
HL-1 mouse atrial cardiomyocytes were a kind gift from Dr. Pasi Tavi (University of Eastern Finland). HL-1 cells were maintained in Claycomb medium (Sigma) supplemented with 10% FBS, 0.1 mM norepinephrine, 50 U/ml penicillin, 50 U/ml streptomycin, and 2 mM UltraGlutamine (Lonza). Culture plates were coated at 37 °C with a solution containing 0.02% gelatin and 10 μg/ml fibronectin. Seeding densities were 140,000 cells/6-well plate well, 100,000 cells/12-well plate well, and 5,000 cells/96-well plate well. H9c2 rat cardiomyoblasts were purchased from ATCC and maintained in DMEM with 1.5 g/l NaHCO3, supplemented with 10% FBS, 50 U/ml penicillin, 50 U/ml streptomycin, and 2 mM UltraGlutamine. Seeding density was 100,000 cells/6-well plate well. Cells were routinely checked for mycoplasma infection using MycoAlert Mycoplasma Detection Kit (Lonza). For hypoxia-reoxygenation experiments, the cells were cultured in 1% O2 in a hypoxic work station (InVivo2, Ruskinn Technology Ltd.) for the indicated periods of time, and returned to normal cell incubator (21% O2) for reoxygenation. In ROR1 ligand activation experiments, 200–400 ng/ml of recombinant human/mouse Wnt-5a (R&D Systems) was added to medium at the time of plating (for MTT assays) or 24 h after plating for 30–60 min (for Western analyses).
RNA interference
One day after plating, HL-1 cells were transfected with siRNAs (Qiagen) targeting ROR1 (siRNA #1, SI01404655; siRNA #2, SI01404662), or ROR2 (siRNA #1, SI01404683; siRNA #2, SI01404690) or with AllStars Negative Control siRNA at a concentration of 100 nM, using Lipofectamine 2000 (Invitrogen). Immediately prior to transfection, medium was changed to antibiotic- and norepinephrine-free Claycomb medium. Four to six hours after transfection, medium was replaced with antibiotic-free, norepinephrine-supplemented Claycomb medium.
RNA extraction and real-time RT-PCR
RNA was extracted from pig myocardium samples using TRIsure reagent (Bioline). Samples were treated with 10 units of DNAse I (Roche). cDNA was synthesized with SensiFAST cDNA Synthesis Kit (Bioline), using 1 μg of total RNA/sample. Real-time RT-PCR was carried out using QuantStudio 12 K Flex Real-Time PCR System thermal cycler (Thermo Fisher Scientific). For PCR reactions, 5 μl of TaqMan Universal Master Mix II (Thermo Fisher Scientific) and 10 ng of template cDNA were used in a reaction volume of 10 μl. Primer concentrations were 0.3 μM and probe concentration 0.1 μM. Primers were acquired from Eurofins Genomics and probes from Universal Probe Library (Roche). GAPDH was used as the reference gene [8]. ROR1 was analyzed using the primers 5’-GCGGCTCGCAATATTCTC-3′ and 5’-GAAAGCCCAAGGTCTGAAATC-3′, and the probe #108. GAPDH was analyzed using the primers 5’-ACAGACAGCCGTGTGTTCC-3′ and 5’-ACCTTCACCATCGTGTCTCA-3′, and the probe #28.
Western blotting and immunoprecipitation
Cells were lysed with lysis buffer [
22] supplemented with Pierce Protease Inhibitor Mini Tablets (Thermo Fisher Scientific). Lysates were centrifuged for 15 min at 16,000 g and the supernatants were collected. Snap-frozen pig heart tissue samples were dissolved in ice-cold Lysis Buffer 17 supplied with the Proteome Profiler Human Phospho-RTK Array Kit. Samples were separated on 8–10% polyacrylamide gels. Protein amounts loaded on the gel were 20–35 μg for cell samples and 100 μg for pig tissue samples. Separated samples were transferred to nitrocellulose membrane which was blocked with 5% non-fat milk or bovine serum albumin in 10 mM Tris-HCl (pH 7.4), 150 mM NaCl and 0.05% Tween-20 (blocking solution) for 1 h at room temperature. Membranes were incubated with primary antibodies overnight at 4 °C in the blocking solution. The following primary antibodies and dilutions were used: anti-ROR1 (sc-83033 and sc-130386, Santa Cruz Biotechnology; 1:250 and 1:125, respectively), anti-actin (sc-1616, Santa Cruz Biotechnology; 1:1,000), anti-α-tubulin (sc-5546, Santa Cruz Biotechnology; 1:1,000), anti-phospho-Akt (#4060, Cell Signaling Technology; 1:1,000), anti-Akt (#9272, Cell Signaling Technology; 1:1,000), anti-phospho-p38 (#9211, Cell Signaling Technology; 1:500), anti-p38 (#9212, Cell Signaling Technology; 1:500), anti-phospho-tyrosine (4 g10, Upstate; 1:500), and anti-GAPDH (G8795, Sigma-Aldrich; 1:1,000). Incubation with secondary HRP-conjugated antibodies (Santa Cruz Biotechnology; 1:5,000) or IRDye secondary antibodies (LI-COR, 1:10,000) was carried out for 1 h at room temperature in the blocking solution. The immunosignals were visualized with WesternBright ECL HRP substrate reagent (Advansta) and imaged with ImageQuant LAS 4000 (GE Healthcare Life Sciences) or with Odyssey imaging system (LI-COR). Densitometric analysis of the Western signals was carried out with ImageJ.
For immunoprecipitation analyses, approximately 2 mg of total protein from HL-1 cell lysates was incubated overnight at 4 °C with 4 μg of mouse monoclonal anti-ROR1 antibody (sc-130386, Santa Cruz Biotechnology). Immunoprecipitated lysates were incubated with Protein G PLUS-Agarose beads (sc-2002, Santa Cruz Biotechnology) for one hour at 4 °C. After washing the beads three times with lysis buffer without protease inhibitors, samples from the supernatants were loaded onto 10% polyacrylamide gels for subsequent Western analysis.
MTT cell viability assay
CellTiter 96 Non-Radioactive Cell Proliferation Assay (MTT) (Promega) was used to measure viability of HL-1 cells. The assay was performed on 96-well plates with plating density of 5,000 cells/well. Before addition of the MTT dye solution, the culture medium was replaced by norepinephrine-free Claycomb medium (100 μl/well). Wells only containing the medium were used for background subtraction. Absorbances at 570 nm were detected with EnSight Multimode Plate Reader (PerkinElmer) and results were normalized to untreated control sample level.
Protein sequence alignment
Human and pig RTK protein sequences from UniProt database were aligned using EMBOSS Needle tool for global alignments and EMBOSS Water tool for local alignments by the European Bioinformatics Institute (
http://www.ebi.ac.uk/Tools/psa/) [
23]. The longest pig sequences available were used.
Statistical analyses
For in silico transcriptomics analyses, Student’s t-test was used to assess the significance of differences between the expression levels of an RTK between sample groups. Correction for multiple testing was performed using false discovery rate (FDR) method [
24]. Genes with significant differences in expression were selected for visualizations. Two-tailed Student’s t-test was used for testing the significance of differences in phosphorylation array data (for RTKs with at least three quantifiable results in both treatments) and in expression levels in Western analyses. For in vitro experiments, data are represented as box plots depicting median (black horizontal line), first and third quartile (box) and the range of the data (whiskers). Multiple group comparisons were performed with Kruskal-Wallis test with Dunn’s post-hoc test using FDR method for
P-value adjustments.
P-values < 0.05 were considered significant. Analyses were performed with RStudio.
Discussion
To address the potential of RTKs as therapeutic targets in myocardial ischemia, expression and phosphorylation of RTKs was systematically analyzed in human and porcine ischemic heart samples. A subgroup of RTKs, both present in phosphorylated form in the ischemic myocardium, and differentially regulated at the expression level between the ischemic and normoxic samples, was identified. This subgroup included ALK, AXL, EGFR, EPHB2, ERBB2, FGFR2, KIT, ROR1, ROR2 and TIE1. As the role of ROR1 in ischemic heart has not previously been addressed, it was selected for further analyses. ROR1 expression and phosphorylation were found to be downregulated in cardiomyocytes in response to hypoxia. Moreover, functional in vitro experiments with RNA interference and Wnt-5a stimulation indicated that targeting of ROR1 enhances cardiomyocyte viability.
Analysis of RTK expression in the acute myocardial infarction samples revealed 4 RTKs with significantly reduced expression (
EGFR,
ERBB2,
ERBB3 and
EPHA2) when compared to healthy heart. Of these,
ERBB2 has also previously been shown to be downregulated in hypoxic human heart [
8]. While the other two ERBB family members,
EGFR and
ERBB3, were down-regulated in our analyses, the previous report by Munk et al. [
8] indicates upregulation of
EGFR and no change for
ERBB3 expression in hypoxia. Interestingly, EPHA2 has been shown to have cardioprotective potential in mouse models of myocardial ischemia and ischemic cardiomyopathy [
32,
33].
Analysis of RTK expression in the ischemic cardiomyopathy samples revealed 3 RTKs with significantly enhanced expression (ROR1, KIT and TIE1) and 11 with significantly reduced expression (EPHA2, LTK, PDGFRB, ERBB3, FGFR2, AXL, ALK, ROR2, EPHA8, EPHB2 and EPHB4) when compared to healthy heart. ROR1 expression was most significantly upregulated. However, the roles for most RTKs identified in our analyses in ischemic cardiomyopathy remain to be elucidated in future studies.
The phosphoarray analysis of porcine myocardial samples indicated the highest phosphorylation level for EGFR, ROR2, INSR, EPHB3, EPHB2, TEK, RYK, ROR1 and EPHB6. For ROR1 and VEGFR2, difference in phosphorylation between control and ischemia-reperfusion samples was statistically significant. While details about the antibodies included in the phospho-RTK array are not publicly available, the kit has been designed to detect human receptors. A conservation analysis between the human and porcine RTK protein sequences indicated high conservation for most receptors (median global similarity = 93.1%) and global similarity of 92.4% for ROR1 (Additional file
7). However, the specificity of the antibodies in the array could not be directly controlled and, especially, the phosphorylation status of ROR1 in ischemic heart remains to be studied in future analyses. Nevertheless, a reproducible set of active RTKs in the ischemic heart was detected, including both novel receptors (e.g. ROR1, ROR2) and ones formerly known to be active in the heart (EGFR, ERBB2, INSR, VEGFR2).
Both ROR1 expression and phosphorylation were found to be down-regulated in cardiomyocytes in vitro. Although ROR1 phosphorylation was downregulated also in the pig model of acute ischemia-reperfusion injury in vivo, ROR1 expression in the pig model demonstrated a nonsignificant trend for increase. Moreover, ROR1 mRNA expression was upregulated in the ischemic cardiomyopathy samples. These findings could reflect the intrinsic differences in measuring protein phosphorylation vs. expression, and the associated feed-back regulation, but also the duration of hypoxia in the different models. For example, ischemic cardiomyopathy is a chronic ischemic disease involving heart failure, while the ischemia-reperfusion-model in the pig and the in vitro analyses are models for more acute hypoxic conditions.
ROR1 signaling was demonstrated to inhibit cardiomyocyte survival, as its knockdown increased, while its ligand activation decreased, cellular viability. ROR1 and ROR2 comprise the Receptor tyrosine kinase-like Orphan Receptor (ROR) family. While ROR signaling in the ischemic heart has not previously been addressed, both receptors are known to regulate heart development during mouse embryogenesis [
6,
7,
34]. Interestingly, in our in vitro analyses, hypoxia affected expression of ROR1, but not of ROR2, and knockdown of
ROR1, but not of
ROR2, enhanced cardiomyocyte viability, implying that the two receptors have overlapping but not fully redundant biological functions in these cells. The pathways regulating ROR1 functions in cardiomyocytes may involve p38, as our phospho-Western analyses indicated p38 regulation both in response to
ROR1 knockdown as well as to Wnt-5a ligand stimulation. Indeed, inhibition of p38 signaling has been shown to enable adult cardiomyocyte proliferation [
35] and promote cardiac regeneration [
36]. Interestingly, a ROR1-targeting monoclonal antibody, cirmtuzumab, has been developed for the treatment of chronic lymphatic leukemia [
37] allowing future analysis of its effect on normal and ischemic heart.