Heart growth and function are angiogenesis-dependent, but little is known concerning the effects of key regulators of angiogenesis on diastolic heart failure. Here, we tested the hypothesis that local vascular endothelial growth factor-B (VEGF-B) gene therapy prevents left ventricular diastolic dysfunction.
Rats were subjected to pressure overload by infusing angiotensin II (33.3 µg/kg/h) for 2 weeks using osmotic minipumps. Intramyocardial delivery of adenoviral vector expressing VEGF-B167A improved the angiotensin II-induced diastolic dysfunction compared with LacZ control virus. Local VEGF-B gene transfer increased the mean capillary area in the left ventricle in control and angiotensin II-infused animals, whereas the density of capillaries was not affected. Interestingly, significant increases were noted in Ki67+ proliferating cells, expression of interleukin1β, and c-kit+ cells in response to VEGF-B gene transfer. The increase in cardiac c-kit+ cells was not associated with an induction of stromal cell-derived factor 1α, suggesting no mobilization of cells from bone marrow. Also, the phosphatidylinositol 3-kinase/Akt pathway was activated.
VEGF-B gene transfer resulted in prevention of the angiotensin II-induced diastolic dysfunction associated with induction of the Akt pathway, increased proliferation and number of c-kit+ cells, as well as an increase in the capillary area in the left ventricle. VEGF-B may offer novel therapeutic possibilities for the prevention of the transition from compensated to decompensated cardiac hypertrophy and thereby for the treatment of heart failure.
Heart failure is one of the most common causes of cardiovascular morbidity and mortality, and its prevalence is rapidly increasing as the mean age of the population advances. Approximately half of the patients with heart failure have diastolic dysfunction with preserved left ventricular (LV) ejection fraction.1–3 In particular, hypertensive patients with LV hypertrophy are at a high risk to develop heart failure with preserved systolic function. Initially, LV hypertrophy induced by high blood pressure is an adaptive response. However, after sustained external load, hearts may evolve to a state of decompensated hypertrophy resulting in tissue fibrosis, loss of contractile function, and cardiac dilatation. The prevalence of heart failure with preserved systolic function in the community is increasing4 and is associated with an exceedingly high mortality.5
The molecular mechanisms responsible for the diastolic heart failure are poorly defined, but disruption of coordinated tissue growth and angiogenesis may contribute to the progression from adaptive LV hypertrophy to heart failure.6,7 Recently, it has been shown that a reduction in cardiac capillary density promotes contractile dysfunction in transgenic mice.7 It has also been concluded that as a response to hypertrophic stimuli both heart growth and function are angiogenesis-dependent.8 Despite these observations implicating a beneficial role for angiogenesis in diastolic heart failure, little is known concerning the effects of key regulators of angiogenesis on diastolic dysfunction. Vascular endothelial growth factors (VEGFs) regulate all types of vascular growth and have thus received much attention regarding their potential use for therapeutic vascular growth in cardiovascular diseases.8 VEGF-B, a member of the VEGF family,9,10 is known to be widely abundant in many tissues and cell types including the myocardium, and skeletal and vascular smooth muscle, and brown adipose tissue.11 Mice lacking the Vegfb gene have been found to have smaller hearts, dysfunctional coronary vasculature,12 atrial conduction abnormality,13 and VEGF-B to have a relatively restricted role in pathological angiogenesis.14
In the present study, we tested the hypothesis that whether local VEGF-B gene therapy by intramyocardial delivery of adenoviral vector expressing VEGF-B167A, the predominant isoform in humans,9,10 prevents LV diastolic dysfunction. VEGF-B was overexpressed in healthy rat hearts and in hearts in rats subjected to pressure overload by angiotensin II infusion. Since these experiments revealed that VEGF-B overexpression led to amelioration of diastolic dysfunction, we evaluated numerous potential mechanisms triggering the improvement of LV function and structure by VEGF-B gene transfer.
An expanded Methods section is available in the Supplementary material online.
Adenoviral vectors, cardiac gene transfer, and angiotensin II infusion with osmotic minipumps
Human clinical-grade first-generation serotype 5 replication-deficient (E1, partially E3 deleted) adenoviruses were produced. Male Sprague Dawley rats were anaesthetized, and a left thoracotomy and pericardial incision were performed. Recombinant adenovirus (5 × 108pfu) in a 100 µL volume was injected into the anterior wall of the left ventricle. The heart was repositioned in the chest and the incision was closed. Angiotensin II (33.3 µg/kg/h) to induce pressure overload15,16 or vehicle (0.9% NaCl) was administered via subcutaneously implanted osmotic minipumps (Alzet model 2002, DURECT Corporation, Cupertino, CA, USA). Minipumps were implanted before the gene transfer. The experimental design was approved by the Animal Use and Care Committee of the University of Oulu. The investigation confirms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996).
Echocardiography, RNA isolation, and quantitative real-time PCR
Transthoracic echocardiography was performed using Acuson Ultrasound System (Sequoia™ 512) and a 15 MHz linear transducer (15L8;Acuson, Mountain View, CA, USA;n = 7–11/group). After echocardiography, the animals were sacrificed, the hearts were removed, and cardiac chambers were stored at −70°C. Total RNA from whole LV tissue was isolated by the guanidine thiocyanate–CsCl method (n = 8–12/group). The cDNA was synthesized from 1.0 µg of LV total RNA by using the Transcriptor First Strand cDNA synthesis kit (Roche Applied Science). The relative expression level of mRNA encoding human VEGF-B in the left ventricle was measured according to the manufacturer's protocol with quantitative RT-PCR (ABI PRISM 7700 Sequence detector, Applied Biosystems) using a specific Assays-on-Demand (Applied Biosystems) target mix (Hs00173634_m1). The expression level of human VEGF-B was normalized to 18S RNA.
Histology, immunohistochemistry, and image analysis
Left ventricles were fixed in 10% buffered formalin solution (n = 8–13/group). Transversal sections of the left ventricle were embedded in paraffin, and 5 µm-thick sections were cut from the mid-section of the heart, at the level of the papillary muscles. Masson's trichrome technique was used to define the area of fibrosis in the left ventricle by analysing the extent of positive staining using a digital image analysis system. The ApopTag in situ apoptosis detection kit (Chemicon, Temecula, CA, USA) was used for 3′-end labelling of apoptotic DNA. Primary antibodies for human VEGF-B, rat c-kit, Ki67, interleukin (IL)1β, phosphorylated Akt, endothelial nitric oxide synthetase (eNOS), stromal cell-derived factor 1α (SDF1α), SDF1β, and tumour necrosis factor α (TNFα) as well as total p38 mitogen-activated protein kinase (MAPK), phosphorylated p38 MAPK, and GATA4 were used to detect the protein expression. Lectin I was used to stain endothelial cells.
Results are expressed as the mean ± SEM. These data were analysed with a one-way analysis of variance (ANOVA) followed by a least significant difference (LSD) post hoc test. For comparisons between two groups, the Student's t-test was used. P-value of <0.05 was considered statistically significant.
VEGF-B167A gene was effectively transfected into the left ventricle by adenoviral delivery
In order to study the effects of myocardial VEGF-B167A overexpression, we first examined the efficiency of adenovirus-mediated gene transfer on mRNA levels in left ventricle in rats. The expression levels of mRNA encoding human VEGF-B167A were analysed by a quantitative real-time PCR 3, 6, and 14 days after VEGF-B167A or LacZ injection. The human VEGF-B mRNA levels were significantly increased 3 days (P < 0.01, Figure 1A) after the injection of VEGF-B167A, and the expression pattern of the VEGF-B gene in left ventricle resembled that observed following adenoviral gene transfer.17 Production of VEGF-B protein was also increased, because using a specific antibody for human VEGF-B, 41% of the cardiomyocytes in the left ventricle in proximity of the VEGF-B167A adenoviral vector injection site were positive for VEGF-B at 2 weeks (P < 0.001, Figure 1B and C). Hearts injected with LacZ control virus showed no staining for human VEGF-B. The antibody used recognizes both 167 and 186 isoforms of human VEGF-B, but does not cross-react with any isoforms of recombinant human VEGF-A, VEGF-C, or VEGF-D, nor recombinant rat VEGF164. The injection site was used to select the same location in all samples used for staining.
VEGF-B167A gene transfer prevents angiotensin II-induced diastolic dysfunction
To evaluate the effect of VEGF-B167A gene transfer on cardiac function, we performed echocardiography. Angiotensin II-induced pressure overload for 2 weeks decreased the E/A ratio (3.9 ± 0.5 in the LacZ group vs. 1.7 ± 0.3 in LacZ + angiotensin II group, P < 0.01, Figure 2A) and prolonged LV isovolumic relaxation time (IVRT, 20.9 ± 1.0 vs. 26.1 ± 0.7, P < 0.01, Figure 2B). VEGF-B167A gene transfer ameliorated angiotensin II-induced diastolic dysfunction; the E/A ratio was almost normalized to 3.2 ± 0.6 (P < 0.05 vs. LacZ + angiotensin II) and IVRT to 21.5 ± 1.3 (P < 0.01 vs. LacZ + angiotensin II). Angiotensin II infusion, VEGF-B gene transfer, or their combination had no effect on LV systolic function (fractional shortening or ejection fraction, Figure 2C and D). Angiotensin II infusion was associated with a small decrease in LV diameter in diastole (6.8 ± 0.4 vs. 8.2 ± 0.2 in LacZ control, P < 0.01), which was normalized by VEGF-B gene transfer (P < 0.05, Figure 2E). Angiotensin II-induced thickening of the interventricular septum in diastole (P < 0.01) was not influenced by local intramyocardial VEGF-B gene transfer (Figure 2F). Angiotensin II infusion caused LV hypertrophy, as reflected by 26% increase in the LV weight/body weight ratio (3.4 ± 0.2 in LacZ group vs. 4.5 ± 0.2 in LacZ + angiotensin II, P < 0.01) at 2 weeks. VEGF-B gene transfer did not significantly change the angiotensin II-induced increase in the LV weight/body weight ratio (data not shown).
Local VEGF-B167A gene transfer increases mean capillary area in left ventricle
To determine the effect of VEGF-B on angiogenesis, histological sections were immunohistochemically stained against lectin 3 days and 2 weeks after gene transfer. The number of capillaries was counted from five representative fields of LV anterior wall choosing three corresponding fields from the epicardial and two from the endocardial side of the left ventricle. A statistically significant 1.3-fold increase in the capillary cross-sectional area was noted after VEGF-B gene transfer both with and without angiotensin II infusion (P < 0.05, Figure 3A and C), whereas treatments had no effect on capillary density (Figure 3B). The capillary density and mean capillary area remained unchanged 3 days after VEGF-B gene transfer (data not shown).
Apoptosis, cardiomyocyte cross-sectional area, and fibrosis
To examine the possible role of structural changes in mediating the beneficial functional effects of VEGF-B gene transfer, we measured apoptosis, cardiomyocyte cross-sectional area, and fibrosis in the left ventricle. VEGF-B gene transfer decreased the number of apoptotic cells at 3 days (58 ± 14 TUNEL+ cells/mm2, LacZ vs. 30 ± 4, VEGF-B, P < 0.05, Figure 4A), but not at 2 weeks (Figure 4C) after injections. Double immunofluorescence staining of TUNEL+ cells showed that they were not positive for cardiomyocyte marker α-actinin (Figure 4B). The cardiomyocyte cross-sectional area remained unchanged by treatments (Figure 4D). Angiotensin II increased the percentage of the fibrotic area of the LV anterior wall at 2 weeks (P < 0.01). VEGF-B gene transfer did not have a significant effect on fibrosis, although there was a trend for the decreased fibrotic area by VEGF-B gene transfer (Figure 4E and F). There was no change in the degree of fibrosis in response to angiotensin II or VEGF-B treatments at 3 days (data not shown).
Increased number of c-kit+ cells and proliferation following VEGF-B167A gene transfer
Recent studies have shown that c-kit+ cardiac stem cells may regenerate functional myocardium.18–20 The number of c-kit+ cells was significantly increased 2 weeks after VEGF-B gene transfer with (2.4-fold, P < 0.05) and without angiotensin II infusion (3.8-fold, P < 0.001, Figure 5A and B). Double immunofluorescence staining with c-kit and phospho-histone H3 antibodies, performed to characterize proliferating cells, showed that a small portion of c-kit+ cells were proliferating (Figure 5C). Stress signals such as tissue injury or inflammation cause upregulation of SDF1α, which promotes the recruitment of stem cells into the heart.21 No changes were observed in SDF1α or SDF1β proteins, analysed by immunohistochemistry, with angiotensin II infusion, VEGF-B gene, or their combination at 3 days or 2 weeks (data not shown).
A significant increase by VEGF-B gene transfer was also observed in Ki67+ proliferating cardiomyocytes in control (1.8-fold, P < 0.01, Figure 6A and C) and angiotensin II-infused rats (2.8-fold, P < 0.05). Staining performed with phospho-histone H3 and α-actinin antibodies confirmed that part of the proliferating cells were cardiomyocytes (Figure 6D). In other non-specified cells, the increase in Ki67 positivity in response to VEGF-B gene transfer was not statistically significant (Figure 6B). IL1β has previously been shown to re-initiate myocyte DNA synthesis,22 and a significant increase was observed in the expression of interstitial/cytoplasmic IL1β protein in the left ventricle after VEGF-B gene transfer at 2 weeks (2.4-fold, P < 0.01, Figure 6E). In combination with angiotensin II-induced pressure overload, a 1.6-fold non-significant increase was noted.
Increased expression of phosphorylated Akt after VEGF-B167A gene transfer
The number of cells positive for phosphorylated Akt was clearly induced by VEGF-B gene transfer at 2 weeks (1.5-fold, P < 0.001, Figure 6F), whereas the expression of phosphorylated Akt was not altered at 3 days (data not shown). The expression of both total and phosphorylated p38 MAPK decreased by 2 weeks angiotensin II treatment, whereas VEGF-B167A gene transfer had no effect (see Supplementary material online, Figure S1A and B). Transcription factor GATA4 has been proposed to function as a stress-responsive regulator of cardiac angiogenesis.23 In the present study, the expression of GATA4 was not altered by angiotensin II treatment and/or VEGF-B gene transfer (see Supplementary material online, Figure S1C). We also studied the expression of eNOS and TNFα by immunohistochemistry, but did not observe any significant changes with angiotensin II treatment and/or VEGF-B167A at 3 days or 2 weeks after injections (data not shown).
Approximately half of patients with a diagnosis of heart failure have a normal or near-normal LV systolic function.4 These patients with a preserved ejection fraction differ from those with heart failure and a low LV ejection fraction in many important ways: they tend to be older and female, and their condition is more likely to be associated with hypertension than with myocardial ischaemia. The rates of death and hospitalizations among these patients are high and have not declined, as they have in patients with heart failure and a low LV ejection fraction.24 Unfortunately, no pharmacological therapy has been shown to be effective in improving outcomes in patients with diastolic dysfunction with preserved LV ejection fraction. The angiotensin II-mediated hypertension has been used extensively in the past and in key studies in the development of antihypertensive agents. Angiotensin II administration by osmotic minipumps increases mean arterial pressure and causes cardiac hypertrophy15,16,25 and diastolic dysfunction,15 as reflected by a decreased E/A ratio and prolonged LV isovolumic relaxation time. The novel finding of the present study is that local VEGF-B167A adenoviral gene transfer into the left ventricle of rat hearts improved LV diastolic dysfunction associated with induction of the Akt pathway, increased proliferation and number of c-kit+ cells, as well as dilatation of pre-existing capillaries.
VEGF-A is a key regulator of physiological and pathological angiogenesis via the VEGF receptor (VEGFR)-2,9,10 whereas both VEGF-B167 and VEGF-B186 isoforms, generated by alternative splicing,26 bind to VEGFR-1, and neuropilin-1 receptor.27,28 In the present study, the VEGF-B167A gene transfer significantly increased the capillary cross-sectional area and ameliorated diastolic dysfunction suggesting an important role for VEGF-B in preventing the progression to heart failure. A similar enlargement of vessels, likely increasing blood flow and thereby oxygen and nutrient transfer, associated with an increase in the number of endothelial cells per vessel cross-section, has been reported in mice hearts with α-myosin heavy-chain VEGF-B167 transgene.29 Moreover, adenovirus-mediated gene transfer of VEGF-B186 isoform has been shown to induce myocardium-specific angiogenesis and arteriogenesis via VEGFR-1 and neuropilin-1 receptor-dependent mechanisms after myocardial ischaemia in pigs and rabbits.30 In a mouse study, VEGF-B186 isoform also induced therapeutic angiogenesis and proliferation of cardiomyocytes.31 Interestingly, there seems to be a difference between the two VEGF-B isoforms, VEGF-B167 being only weakly angiogenic and proliferative, but associated with altered energy metabolism in myocardium.29,30 In a recent study by Zhang et al.32 with VEGF-B deficient mice, VEGF-B was suggested to be critical in survival of blood vessels under pathological conditions, but dispensable for their growth and angiogenesis.
Akt is a serine threonine kinase that is involved in regulating multiple biological processes, such as stimulation of cell proliferation and survival.33 Also endothelial cell survival by VEGF is mediated through the phosphatidylinositol 3-kinase (PI3K)/Akt signal transduction pathway.34 Interestingly, an activation of Akt signalling within endothelial cells has been associated with a 33% increase in capillary density within the myocardium.35 VEGF-induced angiogenesis and mobilization of endothelial progenitor cells (EPCs) have been shown to be impaired in Akt1 KO mice.36 It has also been shown that VEGFR-1, to which VEGF-B167A binds,26 regulates angiogenesis by modulating the VEGF/Akt signalling pathway.37 However, in the present study, we observed increased capillary area by VEGF-B gene transfer without elevated Akt phosphorylation suggesting that Akt signalling is not involved in VEGF-B-mediated changes in the capillary structure under these experimental conditions.
Induction of the PI3K/Akt pathway is linked to increased embryonic cardiomyocyte proliferation as well38 and likewise, Akt kinase has been suggested to be responsible for the promotion of proliferation in the post-natal heart.39 The role of Akt signalling is, however, unclear in adult myocardial cell proliferation. In this study, we noted a small but significant increase in proliferation, in agreement with the studies using a pig infarction model, in which a slight increase in endothelial and other cell proliferation was observed.30 In a recent study40 with pressure overload caused by aortic banding, the stretch-mediated IL1β release promoted synthesis of insulin-like growth factor (IGF)-1 leading to stimulation of the PI3K/Akt pathway. In our present study, an induction of IL1β, associated with activation of the Akt pathway, was noted after VEGF-B167A gene transfer suggesting that the Akt signalling pathway may be important also in adult myocardial cell proliferation. The question whether the adult heart contains capacity for cell proliferation is still debated (for a review see Lee41), but our present results support the view that adult rat myocardium after injury may have a limited potential for proliferation.
EPCs release paracrine factors such as VEGF, SDF1, and IGF-1.21 As a functional consequence, a conditioned medium of EPCs stimulates a migratory effect on cardiac resident c-kit+ progenitor cells.21 Cardiac progenitor cells (CPCs), which express c-kit, MDR1, and/or Sca-1 markers, are self-renewing, clonogenic and multipotent in vitro and differentiate into cardiomyocytes, smooth muscle cells, and endothelial cells in vivo.18–20 The c-kit+ cells appear to have the greatest potential for cardiac regeneration.19 In our study, increased expression of VEGF-B gene did not lead to an induction of migration-involved SDFs, but was associated with an increase in the number of resident c-kit+ cells. Thus, it seems likely that the increase in the number of c-kit+ cells may indicate proliferation of resident c-kit+ cells rather than mobilization of c-kit+ cells from bone marrow. In a recent study of CPCs, which were activated with hepatocyte growth factor and IGF-1 before injection to rat heart to site of experimental myocardial infarction, the cells were shown to regenerate conductive, intermediate-sized, and small coronary arteries and arterioles as well as capillary structures.42 Likewise, post-infarction angiogenesis was severely impaired in mice with abnormal c-kit function.43 These findings are in agreement with our results of increased proliferation and the number of c-kit+ cells in myocardium after VEGF-B gene transfer leading to capillary dilatation.
Both VEGF-B and neuropilin-1 have recently been implicated in antiapoptotic signaling.44,45 The number of apoptotic cells has also been shown to decrease after VEGF-B186 gene transfer, and also conditioned media from both VEGF-B186- and VEGF-B167-treated endothelial cells decrease the expression of apoptotic factors Bik1, Bmf1, and Bad1 in cardiomyocytes.30 Very recently, VEGF-B167 gene delivery was shown to exert potent antiapoptotic effect after myocardial infarction.46 Likewise, we observed a significant decrease in cardiac TUNEL+ cells 3 days after VEGF-B167 gene transfer indicating an antiapoptotic effect of VEGF-B167. On the other hand, the role of VEGF in regulating fibrosis remains unclear. In the present study, we did not observe a significant effect on angiotensin II-induced fibrosis by VEGF-B gene transfer. Similarly, there was no significant change in the cardiomyocyte cross-sectional area suggesting that neither of these structural parameters were contributing to the prevention of diastolic dysfunction by VEGF-B transfer.
Although no pharmacological therapy has been shown to be effective in improving outcomes in patients with heart failure with a preserved LV ejection fraction (for a review see Verma and Solomon47), several drugs have been reported to influence diastolic dysfunction in animal models. In particular, angiotensin-converting enzyme inhibitors and angiotensin receptor blockers or their combination are beneficial in a rat model of heart failure with preserved ejection fraction, as are also mineralocorticoid receptor blockers and long-acting loop diuretics (reviewed in Yamamoto et al.48). Moreover, β-blockers may improve the survival rate in a rat model of heart failure with preserved ejection fraction through inhibition of oxidative stress, inflammatory changes, LV hypertrophy, and fibrosis.49,50 At present, the control of hypertension—irrespective of the type of antihypertensive agent used—appears to be the most effective strategy in patients with diastolic dysfunction.47 Yet, despite treatment of high blood pressure LV hypertrophy occasionally persists and patients are still at risk for heart failure. In the present study, VEGF-B167A gene transfer activated the PI3K/Akt pathway, increased proliferation and number of c-kit+ resident cardiac cells in myocardium, and resulted in dilatation of pre-existing capillaries in the left ventricle in pressure overloaded heart associated with prevention of LV diastolic dysfunction. VEGF-B could thus offer novel therapeutic possibilities for the treatment of heart failure by preventing the transition from compensated to decompensated cardiac hypertrophy induced by pressure overload.
This work was supported by grants from the Academy of Finland (Center of Excellence, 1105818, 1211096), Finnish Foundation for Cardiovascular Research and the Sigrid Juselius Foundation.
We thank Hanna Leskinen for performing the echocardiographical measurements, Tuulikki Kärnä and Sirpa Rutanen for their expert technical assistance and Tuula Mäkinen for expert animal care.
Conflict of interest: none declared.