CaMKII inhibitor KN‐93 impaired angiogenesis and aggravated cardiac remodelling and heart failure via inhibiting NOX2/mtROS/p‐VEGFR2 and STAT3 pathways

Abstract Persistent cardiac Ca2+/calmodulin‐dependent Kinase II (CaMKII) activation was considered to promote heart failure (HF) development, some studies believed that CaMKII was a target for therapy of HF. However, CaMKII was an important mediator for the ischaemia‐induced coronary angiogenesis, and new evidence confirmed that angiogenesis inhibited cardiac remodelling and improved heart function, and some conditions which impaired angiogenesis aggravated ventricular remodelling. This study aimed to investigate the roles and the underlying mechanisms of CaMKII inhibitor in cardiac remodelling. First, we induced cardiac remodelling rat model by ISO, pre‐treated by CaMKII inhibitor KN‐93, evaluated heart function by echocardiography measurements, and performed HE staining, Masson staining, Tunel staining, Western blot and RT‐PCR to test cardiac remodelling and myocardial microvessel density; we also observed ultrastructure of cardiac tissue with transmission electron microscope. Second, we cultured HUVECs, pre‐treated by ISO and KN‐93, detected cell proliferation, migration, tubule formation and apoptosis, and carried out Western blot to determine the expression of NOX2, NOX4, VEGF, VEGFR2, p‐VEGFR2 and STAT3; mtROS level was also measured. In vivo, we found KN‐93 severely reduced myocardial microvessel density, caused apoptosis of vascular endothelial cells, enhanced cardiac hypertrophy, myocardial apoptosis, collagen deposition, aggravated the deterioration of myocardial ultrastructure and heart function. In vitro, KN‐93 inhibited HUVECs proliferation, migration and tubule formation, and promoted apoptosis of HUVECs. The expression of NOX2, NOX4, p‐VEGFR2 and STAT3 were down‐regulated by KN‐93; mtROS level was severely reduced by KN‐93. We concluded that KN‐93 impaired angiogenesis and aggravated cardiac remodelling and heart failure via inhibiting NOX2/mtROS/p‐VEGFR2 and STAT3 pathways.


| INTRODUC TI ON
The prevalence of heart failure (HF) continued to rise over time and was a global difficult problem, and projections showed that the prevalence of HF would increase 46% from 2012 to 2030, resulting in >8 million people ≥18 years of age with HF in the United States. 1 Although great progress has been made in drug therapy of HF, the effect was limited, and it was necessary to explore new and effective drug targets.
Cardiac remodelling was an important pathological basis of HF and was characterized by myocardial hypertrophy, apoptosis, interstitial fibrosis and reduced heart function. 2 In recent years, calcium/ calmodulin-dependent protein kinase II (CaMKII) caused wide attention. Studies found that the activation and expression of CaMKII were increased in HF human and animal model, 3,4 and CaMKII signallings were essential to cardiac remodelling in pressure-overload mice, or mediated cardiac hypertrophy of rats induced by isoproterenol (ISO). HF was exacerbated in cardiac-specific CaMKIIdeltaC transgenic mice, inhibition of CaMKII reversed cardiac dysfunction in a genetic model of dilated cardiomyopathy, [5][6][7][8][9] and some strategies which inhibited CaMKII suppressed cardiac hypertrophy, 10,11 and a recent study indicated that chronic CaMKII inhibition with KN-93, a widely used inhibitor of CaMKII, improved cardiac function and cardiac remodelling in HF mice induced by pressure-overload. 12 It seemed that CaMKII inhibition was an effective drug to reverse cardiac remodelling and improve heart function.
Of note, CaMKII was an important mediator for the ischaemiainduced coronary angiogenesis, in which process, CaMKII played a key role in the trigger of vascular endothelial cell growth factor (VEGF) expression, and the inhibitor of CaMKII KN-93 suppressed the proliferation and migration of cardiac endothelial cells and inhibited angiogenesis. 13 A newest study demonstrated that cross-talk between STAT3 and CaMKII mediated endothelial angiogenesis, 14 and STAT3 was also a crucial factor to promote angiogenesis. 15,16 Recently, a lot of new evidence confirmed that angiogenesis was contributed to the inhibition of cardiac remodelling and improvement of heart function, beta-AR blocker metoprolol promoted angiogenesis in hypertrophied myocardium and reduced hypertrophy, and cardiac function of TAC-induced HF mice was improved by promoting angiogenesis, 17 ivabradine and a microRNA attenuated cardiac hypertrophy and remodelling via promoting angiogenesis after myocardial infarction, 18,19 and some conditions which impaired angiogenesis thus promoted or aggravated ventricular remodelling and HF. [20][21][22] It seemed that inhibition of CaMKII would aggravate cardiac remodelling and HF progression.
So, did inhibitor of CaMKII reverse myocardial remodelling? If not, what were the underlying mechanisms?
The mechanisms of angiogenesis were complex, excepted for STAT3 pathway, mitochondrial reactive oxygen species (mtROS) was a critical activating factor for VEGF signalling, and promoted angiogenesis in endothelial cells, NADPH oxidase 2 (NOX2) sensed Nox4-derived ROS to promote mtROS production and caused tyrosine phosphorylation of VEGF receptor type 2 (p-VEGFR2) and induced angiogenesis. 23 There were four known isozymes of CaMKII (α, β, γ and δ) 24 ; in myocardial tissue, CaMKIIδ had been linked to arrhythmia, 25,26 as well as remodelling due to atrial fibrillation and myocardial infarction. 27 A protective role for CaMKIIδ had been implicated in ischaemia/reperfusion. 28 Given these latter associations, CaMKIIδ remains a target of high interest in cardiovascular disease, 29 KN-93 had inhibitory effects on CaMKIIδ, and it was widely used as an inhibitor of CaMKII, so, we select KN-93 as tool in our study.
In this study, by using cardiac hypertrophy rat model induced by ISO, we found that KN-93 reduced the myocardial microvessel density and aggravated cardiac remodelling and HF. In order to reveal the underlying mechanisms, we cultured the HUVECs and studied the effects of KN-93 on angiogenesis, and we found that KN-93 suppressed cell proliferation, migration and tubule formation via inhibiting NOX2/mtROS/ p-VEGFR2 and STAT3 pathways, and also promoted apoptosis of HUVECs. University). Rat model of HF was established as described previously by us. 30 ISO (7 mg/kg) was administered once daily by intraperitoneal injection for 2 weeks (defined as ISO group, n = 12), the dosage was according to our previously study, 31 but was reduced. KN-93 (10 ug/g) was administered once daily in the same manner at 30 min before ISO was treated (defined as KN-93 group, n = 14), lasted for 2 weeks. Control animals were administrated with 0.9% NaCl (Ctrl group, n = 12).

| Echocardiographic measurements
Echocardiography measurements were performed to evaluate the heart function and structure as described previously by us. 32

| Heart weight to body weight ratio (HW/BW)
Rats were weighed before anaesthetization, fresh hearts were isolated and washed, vessels and other redundant tissues were removed, and then, the heart was weighed; HW/BW (mg/100 g) was calculated.

| Histological staining
Hearts were isolated from rats under anaesthesia with chloral hydrate, left ventricular tissues were dissected and fixed in 4% paraformaldehyde, embedded in paraffin, and then, the samples were sectioned into 5 μm-thick slices. Haematoxylin-eosin (HE) staining and Masson's trichrome were performed. The cardiac collagen volume fraction (CVF), perivascular fibrosis ratio and the microvascular density (MVD) were measured by using Image-pro Plus 6.0 (Media Cybernetics, Bethesda). Six fields were randomly selected under 200× microscope, counted the number of microvessels and averaged, and then converted into counts/100 μm 2 .

| TUNEL staining
Left ventricular tissues were collected, and the TUNEL assay was performed as described previously by us. 30

| Transmission electron microscope
Preparation of specimens was described previously by us 30

| Cell Culture and treatment
HUVEC cells were purchased from BeNa Culture Collection and cultured in ECM medium containing 10% FBS (1.5 mM L-glutamine, 100 U/ml penicillin and 100 μg/ml Streptomycin) and incubated in 37°C 5% CO 2 saturated humidity incubator.

| Migration detection
The proliferation of HUVECs was quantified by MTT assay. When the cells were cultured in 6-well plate for 24 h, a 200 ul tip of pipettor was used to draw evenly straight lines vertically through the holes in the 6-well plate, 2 lines for each hole, then washed and took pictures at 0 h. Added KN-93, ISO, the concentration was same as above and cultured for 24 h, took pictures again.

| Tubule formation experiments
Five hundred microliter matrix gel (diluted with serum-free base medium 1:8) was absorbed into a 12-well plate and put it into the incubator and gel for 30 min; cells were digested to prepare cell suspension, added to 12-well plate and added KN-93, ISO; the concentration was same as above. Then put 12-well plate in an incubator with 37°C and 5% CO 2 , cultured for 48 h and take photographs.

| Measurement of mtROS
Mitochondria were isolated and purified of HUVECs using Mitochondria Isolation Kit (Sigma) according to the manufacturer's instructions. In addition, mtROS level was performed as described previously by us. 31

| Statistical analysis
All data were presented as means ± SEM. Comparisons between two groups were performed with Student's t-test and among three groups were compared with one-way ANOVA followed by Tukey post hoc test for significance. All statistics were determined using SPSS15.0 software (SPSS Inc.). A probability value of p < 0.05 is considered significant.

| KN-93 aggravated cardiac remodelling and HF of ISO rats
As shown in Figure 1A, B, hearts of ISO rats were obvious hypertrophy and the hypertrophy was aggravated by treatment with KN-93.
Echocardiography data showed that LVEF and LVFS were reduced and LVIDd, LVIDs and LVPWTd were increased in ISO group. However, LVIDd, LVIDs, IVSTd and LVPWTd were increased further in KN-93 group, and LVEF and LVFS also were reduced further by KN-93, there was no statistical significance in IVSTd between Ctrl and ISO groups, and there was no significant difference in HR among three groups, and HW/ BW was also further increased by KN-93, as shown in Figure 1C-J. It suggested that KN-93 aggravated cardiac remodelling and HF of ISO rats.

| KN-93 aggravated cardiac remodelling
Results of HE staining indicated that KN-93 caused more serious inflammatory cell infiltration, myocardial hypertrophy, arranged disorder, degeneration necrosis, and interstitial hyperemia and oedema compared with ISO group and Ctrl groups, as shown in Figure 2A.  Results of Masson staining indicated that left ventricular collagen content in KN-93 rats was much more increased compared with ISO and Ctrl, and CVF was increased obviously by KN-93 compared with Ctrl and ISO, as shown in Figure 2B and 2F. The expression of collagen Ⅰ and Collagen Ⅲ were upregulated in ISO and were highest in KN-93 group, as shown in Figure 2G,H. The results confirmed that KN-93 promoted collagen deposition and aggravated myocardial interstitial fibrosis.

| KN-93 aggravated apoptosis of myocardial cells
The results of TUNEL staining indicated that KN-93 significantly increased the number of TUNEL-positive cells compared with ISO and Ctrl, as shown in Figure 3A The results were consistent with TUNEL, as shown in Figure 3B-E.

| KN-93 aggravated changes of myocardial ultrastructure in ISO rats
Myocardial ultrastructure was observed and analysed by transmission electron micrographs. The myocardial fibres were arranged regularly, the sarcomere structure was clear, mitochondria were neatly arranged, showed uniform morphology and intact structure in Ctrl group. In ISO group, myocardial cells were arranged in a disorderly manner, the plasma membrane of the myocardium was blurred, and the muscle filaments were loose, fragmentation and swelling, with absence of the Z line, increase of the mitochondria, mitochondria were varied in size and not neatly arranged, structure were unclear, swelling and crista fragmentation, or dissolved into vacuole.
However, KN-93 aggravated the changes of myocardial ultrastructure compared to ISO group, we observed severe denaturation, oedema and destruction in most of myofibrils, although enhanced mitochondrial density, but majority of mitochondria were damaged, they were collapse or clumping and dissolved into vacuole. The statistic results of density of mitochondria and percentage of vacuolization in mitochondria were represented as statistical histograms as shown in Figure 4.

| KN-93 reduced MVD, caused apoptosis of vascular endothelial cells and increased prevascular fibrosis ratio in myocardial tissue
The MVD and prevascular fibrosis ratio were measured, the statistic data indicated that MVD was much more reduced in myocardial tissue by KN-93 compared with ISO and Ctrl, prevascular fibrosis

| KN-93 suppressed proliferation, migration and tubule formation of HUVECs
We found that 1 uM KN-93 had no effect on migration and proliferation of HUVECs; however, there was no tubule formation in cells

| KN-93 promoted apoptosis of HUVECs
The data of flow cytometry indicated that the number of apoptotic cells was the highest in KN-93 group; it demonstrated that KN-93 evidently promoted apoptosis of HUVECs as shown in Figure 7A, B.

| KN-93 severely limited the production of mtROS
The level of mtROS in ISO group was obviously increased; however, KN-93 severely reduced the mtROS compared with Ctrl group and ISO group, as shown in Figure 7C.

| DISCUSS ION
Angiogenesis was a physiological and pathophysiological process that initiated vascular growth from pre-existing blood vessels in response to a lack of oxygen. In recent years, therapeutic angiogenesis had been shown to revascularize ischemic heart tissue, reduced the progression of tissue infarction, and evaded the need for invasive surgical procedures or tissue/organ transplants, 34 and it was considered as a potential treatment for ischemic heart disease (IHD) and was widely studied. Therefore, researchers had done a lot of exploration to reveal the correlation angiogenesis and cardiac remodelling and HF, including cardiac hypertrophy, fibrosis and heart function, they found that cardiac hypertrophy, remodelling and HF were inhibited by promoting myocardial angiogenesis, 17,35 and cardiac fibrosis also had a negative correlation with cardiac angiogenesis. 36,37 Similarly, some strategies which impaired angiogenesis thus promoted or aggravated ventricular remodelling and cardiac dysfunction. 20,21,38 So, angiogenesis should be protective against cardiac remodelling and HF.
CaMKII was an important mediator for the ischaemia-induced coronary angiogenesis. 13 However, persistent cardiac CaMKII activation was considered to promote heart failure development, 3 and some studies believed that CaMKII was a target for the treatment of heart failure. 39 Whether inhibition of CaMKII could reverse or aggravate cardiac remodelling confused us. In this study, we demonstrated in vivo and in vitro that inhibition of CaMKII by KN-93 impaired myocardial angiogenesis and aggravated cardiac remodelling and HF for the first time.
In tumour tissue, cell apoptosis was induced by inhibiting angiogenesis, 40 and cardiac fibrosis had a negative correlation with cardiac angiogenesis 36,37 ; in this study, we demonstrated KN-93 suppressed myocardial angiogenesis, thereby cardiomyocyte apoptosis and cardiac fibrosis were all aggravated by KN-93, and the results were consistent with the previous studies.
We further revealed the underlying mechanism of action of KN-93.
VEGF induced angiogenesis by stimulating EC migration and proliferation primary through the VEGFR2, and VEGF/VEGFR2 was a classical pathway to regulate angiogenesis in many tissue. 41 However, mtROS was an indispensable factor to active this pathway, VEGFR2 had to be phosphorylated by mtROS and activated, and the production of mtROS was promoted by NOX2 and NOX4, NOX2 sensed Nox4-derived H 2 O 2 to promote mtROS production and caused tyrosine phosphorylation of VEGFR2 (p-VEGFR2) and angiogenesis, the production of mtROS and phosphorylation of VEGFR2 were suppressed when NOX2 was blocked. 23 We found that KN-93 significantly down-regulated expression of NOX2 and NOX4, thereby markedly reduced the production of mtROS and p-VEGFR2, but KN-93 had no effect on expression of VEGF and VEGFR2. The results proved that KN-93 inhibited HUVECs tubule formation via reducing production of mtROS. In addition, STAT3 was also a crucial factor to promote angiogenesis; we found that STAT3 was also down-regulated by KN-93. However, mtROS excessive accumulation could inhibit STAT3; this accounted for the suppression of angiogenesis in ISO group, which induced excessive emission of mtROS. 42 So, we concluded that KN-93 impaired myocardial angiogenesis via inhibiting NOX2/mtROS/p-VEGFR2 and STAT3 pathways.
We also found by TUNEL staining that KN-93 caused apoptosis of vascular endothelial cells, and this was confirmed in cell experiments, KN-93 significantly induced apoptosis of HUVECs, probably because STAT3 was a protective factor of apoptosis, 43 So, inhibition of CaMKII by KN-93, which was a widely used inhibitor of CaMKII, impaired myocardial angiogenesis and aggravated cardiac remodelling and HF via inhibiting NOX2/mtROS/p-VEGFR2 and STAT3 pathways. This study provided evidence for us to further understand the mechanisms of cardiac remodelling and HF.
In the end, our study has certain limitations, we only study , there are other inhibitors of CaMKII, such as KN-62/CK59 or some novel inhibitor, tatCN2, RA360 and so on. Perhaps, some of them can promote angiogenesis and be benefit to cardiac remodelling, and are worth further study.

ACK N OWLED G EM ENT
This work is supported by Shaanxi Provincial Key Research and Development Projects (NO. 2019SF-218). In addition, we thank for the financial assistance.

CO N FLI C T O F I NTE R E S T
The authors declare that they have no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.