Adenoviral Delivery of Angiotensin-(1-7) or Angiotensin-(1-9) Inhibits Cardiomyocyte Hypertrophy via the Mas or Angiotensin Type 2 Receptor

The counter-regulatory axis of the renin angiotensin system peptide angiotensin-(1-7) [Ang-(1-7)] has been identified as a potential therapeutic target in cardiac remodelling, acting via the mas receptor. Furthermore, we recently reported that an alternative peptide, Ang-(1-9) also counteracts cardiac remodelling via the angiotensin type 2 receptor (AT2R). Here, we have engineered adenoviral vectors expressing fusion proteins which release Ang-(1-7) [RAdAng-(1-7)] or Ang-(1-9) [RAdAng-(1-9)] and compared their effects on cardiomyocyte hypertrophy in rat H9c2 cardiomyocytes or primary adult rabbit cardiomyocytes, stimulated with angiotensin II, isoproterenol or arg-vasopressin. RAdAng-(1-7) and RAdAng-(1-9) efficiently transduced cardiomyocytes, expressed fusion proteins and secreted peptides, as demonstrated by western immunoblotting and conditioned media assays. Furthermore, secreted Ang-(1-7) and Ang-(1-9) inhibited cardiomyocyte hypertrophy (Control = 168.7±8.4 µm; AngII = 232.1±10.7 µm; AngII+RAdAng-(1-7) = 186±9.1 µm, RAdAng-(1-9) = 180.5±9 µm; P<0.05) and these effects were selectively reversed by inhibitors of their cognate receptors, the mas antagonist A779 for RAdAng-(1-7) and the AT2R antagonist PD123,319 for RAdAng-(1-9). Thus gene transfer of Ang-(1-7) and Ang-(1-9) produces receptor-specific effects equivalent to those observed with addition of exogenous peptides. These data highlight that Ang-(1-7) and Ang-(1-9) can be expressed via gene transfer and inhibit cardiomyocyte hypertrophy via their respective receptors. This supports applications for this approach for sustained peptide delivery to study molecular effects and potential gene therapeutic actions.


Introduction
The renin-angiotensin system (RAS) is recognized for its systemic actions, however the presence of RAS components in specific tissues (e.g. heart, brain, kidney), suggests the presence of a local RAS. Furthermore, a counter-regulatory axis of the RAS exists, which functions mainly via angiotensin converting enzyme 2 (ACE2)/angiotensin (Ang)-(1-7)/mas and inhibits many detrimental cardiovascular disease phenotypes [1,2]. Ang-(1-7) has been shown to antagonise pathological actions such as cardiac hypertrophy and fibrosis through the receptor mas [1,3]. Most recently we reported that Ang-(1-9), a poorly characterised peptide not previously reported as a receptor agonist, also had antihypertrophic effects on angiotensin II (AngII)-induced cardiomyocyte hypertrophy, as a functional ligand at the angiotensin type 2 receptor (AT 2 R) [4]. Moreover, we also demonstrated that Ang-(1-9) reduced cardiac fibrosis in stroke prone spontaneously hypertensive rats through the AT 2 R [5]. These studies highlight the potential for therapeutic application of Ang-(1-7) and Ang- (1)(2)(3)(4)(5)(6)(7)(8)(9) in cardiovascular disease applications. Though active angiotensin peptides are generated extracellularly in the plasma via renin mediated cleavage of angiotensinogen to angiotensin I, followed by ACE-mediated cleavage to AngII, methods which enable their expression through gene transfer approaches are available. Transgenic expression of AngII and Ang-(1-7) can be mediated through the use of synthetic fusion protein expression cassettes which are expressed intracellularly and result in cleavage and secretion of active peptides. Such approaches have been utilised to demonstrate organ-specific effects of individual angiotensin peptides in the heart, kidney and brain [6][7][8][9][10][11][12][13][14]. Gene therapy approaches have also been reported for Ang-(1-7) in models of both myocardial infarction and diabetic retinopathy using viral vector-mediated gene transfer, highlighting their potential in this setting [15,16]. Here, we have generated adenoviral (Ad) vectors encoding fusion proteins expressing Ang-(1-7) or Ang-(1-9) and compared their effects in models of cardiomyocyte hypertrophy. We report that adenoviral gene transfer can be used to express different angiotensin peptides and it can be shown that these peptides are secreted from cells and maintain the receptor-specific interactions that have been reported for the endogenous peptides. This highlights the general applicability of this approach and importantly for the first time demonstrates that Ang-(1-9) can be expressed via adenoviral gene transfer and mediate functional effects at the AT 2 R.

Generation of RAdAng-(1-7) and RAdAng-(1-9)
The fusion protein expression cassette consists of a signal peptide, an IgG molecule linked to Ang-(1-7) or Ang-(1-9) and a cleavage site for furin protease enabling active peptides to be secreted ( Figure 1A). Western immunoblotting of Ad transduced H9c2 cardiomyocytes demonstrated expression of each fusion protein with a size of 32 kDa as expected ( Figure 1B).

Discussion
We have generated adenoviral vectors which selectively overexpress Ang-(1-7) or Ang-(1-9). RAdAng-(1-7) and RAdAng-(1-9) and transduce neonatal rat cardiomyoycte cell lines and adult primary rabbit cardiomyocytes and secrete RAS peptides. Transducing each respective cell type with viral vector doses which achieved 50 or 100% transduction efficiency resulted in secreted peptides inhibiting cardiomyocyte hypertrophy in the same manner as described for exogenous peptides, via mas and AT 2 R respectively [3,4], confirming over-expression and release of active peptides via adenoviral transduction.
Currently, there is great interest in studying the counterregulatory ACE2/Ang-(1-7)/mas axis and our own recent publications have also highlighted the importance of studying other peptides such as Ang-(1-9), which may have, as yet, relatively unexplored functions [4,5]. Most studies of angiotensin peptide function utilise systemic delivery of peptides via osmotic minipumps, however, transgenic models directing tissue-specific production of angiotensin peptides to avoid confounding systemic effects have also been developed [6][7][8][9][10][11][12][13][14]. These studies have proved invaluable in our understanding of endogenous tissuespecific RAS. However, the possibility of developmental abnormalities caused by embryonic over-expression following transgenic model generation cannot be discounted. Such effects may explain the discrepancies between transgenic studies and gene overexpression studies. For example, ACE2 null mice have been reported to have increased blood pressure with no cardiac changes [2] while ACE2 overexpression by adeno-associated virus (AAV) 6-mediated gene delivery leads to severe cardiac fibrosis [19]. Furthermore, Ang-(1-7) has previously been investigated as a gene therapeutic approach utilising viral vector mediated delivery. In a rat model of myocardial ischaemia lentiviral gene transfer of Ang-(1-7) 5 weeks before induction of myocardial infarction prevented decreased myocardial performance as indicated by improvements in fractional shortening and decreased myocardial wall thinning [15]. In a model of diabetic retinopathy local ocular delivery of adeno-associated virus (AAV) expressing Ang-(1-7) prevented diabetes-induced retinal vascular damage and inflammation [16]. Furthermore, adenoviral gene transfer of AngII to vascular smooth muscle cells in vitro resulted in secretion of AngII and functional effects on smooth muscle cell hypertrophy [20]. These studies highlight the potential for gene transfer of angiotensin peptides for molecular investigations and therapeutic approaches and our study builds on these findings. Adenoviral vectors mediate acute overexpression of genes in adult cardiomyocytes in vitro and in vivo which may be applicable for gene therapy applications in acute conditions such as myocardial infarction (MI) and furthermore, their tropism is efficient for other cell types in other organs such as kidney, brain and skeletal muscle, which are also important to study tissuespecific RAS effects. Moreover, for long term expression of peptides in the heart AAV, particularly serotypes such as 1, 6 and 9 [21], can produce stable expression in myocardium following non-invasive intravenous delivery which may be beneficial to modulate long term detrimental remodelling in the heart following an MI, in order to counter progression to heart failure. Therefore, there are a range of delivery vectors which will enable assessment of Ang-(1-7) and Ang-(1-9) in gene therapy approaches.

Cardiomyocyte Hypertrophy and Adenoviral Transduction
H9c2 cardiomyocytes were obtained from ECACC (Wiltshire, UK). The isolation of primary rabbit cardiomyocytes was approved by the University of Glasgow Animal Procedures and Ethics Committee and performed in strict accordance with UK Home Office guidelines under the Animals Scientific Procedures Act 1986. Isolation was performed as previously described [4]. Expression of the fusion proteins in H9c2 cardiomyocytes transduced with 500 or 1000 plaque forming units (pfu)/cell of each RAd for 48 h was confirmed by western immunoblotting as described previously [22] using a mouse anti-IgG2b antibody (rabbit polyclonal 1:250; Abcam, Cambridge, UK), which detects expression of the IgG portion of the fusion protein ( Figure 1A). Cardiomyocytes were stimulated with AngII, arginine-vasopressin or isoproterenol to induce hypertrophy as described previously [4,17,18]. Cell size was measured using Image ProPlus 4.1 image software (Media Cybernetics, Basingstoke, UK ). For each condition we measured 100 cells (10 fields of view in each condition). In H9c2 cardiomyocytes cell length was measured. In primary cardiomyocytes midpoint width was measured as previously described [4]. To assess RAdAng-(1-7) or RAd-(Ang-1-9) in cardiomyocyte hypertrophy H9c2 cardiomyocytes were transduced with 500 or 1000 plaque forming units (pfu)/cell and primary rabbit left ventricular cardiomyocytes with 50, 100 or 300 pfu/cell of RAdAng-(1-7), RAdAng-(1-9) or RAd60 (negative control). After 24 h incubation, cells were stimulated with AngII (100 nM; Sigma, Dorset, UK), isoproterenol (1 mM; Sigma, Dorset, UK) or arg-vasopressin (1 mM; Sigma, Dorset, UK) and incubated for 96 hours (H9c2) or 24 h (rabbit). To assess receptor use experiments were performed in the presence or absence of the Mas antagonist A779 (10 mM; Bachem, Rhein, Germany) or the AT 2 R antagonist PD123,319 (500 nM; Sigma, Dorset, UK).

Conditioned Media Assay
To provide evidence that the peptides were secreted using a conditioned media assay. HeLa cells were transduced with 100 pfu/cell of either RAdAng1-7, RAdAng1-9 or RAd60 and incubated in serum free media for 48 hours at 37uC to allow the secreted peptides to accumulate in the media (conditioned media). The conditioned media then was collected and added to previously quiesced H9c2 cardiomyocytes 30 minutes before addition of AngII (100 nM). H9c2 cardiomyocytes were incubated for 96 hours before measuring cell size as already mentioned.

Statistical Analysis
Experiments were performed in triplicate on 3 different occasions. Mean 6 standard error of the mean (S.E.M) is presented. One way ANOVA with Bonferroni's correction for multiple comparisons were applied and statistical difference was considered with p values ,0.05.