Treatment with αvβ3-integrin-specific 29P attenuates pressure-overload induced cardiac remodelling after transverse aortic constriction in mice

Heart failure remains one of the largest clinical burdens globally, with little to no improvement in the development of disease-eradicating therapeutics. Integrin targeting has been used in the treatment of ocular disease and cancer, but little is known about its utility in the treatment of heart failure. Here we sought to determine whether the second generation orally available, αvβ3-specific RGD-mimetic, 29P, was cardioprotective. Male mice were subjected to transverse aortic constriction (TAC) and treated with 50 μg/kg 29P or volume-matched saline as Vehicle control. At 3 weeks post-TAC, echocardiography showed that 29P treatment significantly restored cardiac function and structure indicating the protective effect of 29P treatment in this model of heart failure. Importantly, 29P treatment improved cardiac function giving improved fractional shortening, ejection fraction, heart weight and lung weight to tibia length fractions, together with partial restoration of Ace and Mme levels, as markers of the TAC insult. At a tissue level, 29P reduced cardiomyocyte hypertrophy and interstitial fibrosis, both of which are major clinical features of heart failure. RNA sequencing identified that, mechanistically, this occurred with concomitant alterations to genes involved molecular pathways associated with these processes such as metabolism, hypertrophy and basement membrane formation. Overall, targeting αvβ3 with 29P provides a novel strategy to attenuate pressure-overload induced cardiac hypertrophy and fibrosis, providing a possible new approach to heart failure treatment.


Introduction
Heart failure (HF), often the terminal aetiology of many cardiovascular diseases (CVD), is predicted to affect around 64 million people globally [1].Targeting the progression of CVD is therapeutically challenging owing to the myriad of HF symptoms.In general, HF is often accompanied by poor cardiac function, deposition of non-contractile fibrotic tissue, an insufficient coronary microvascular and pathological cardiomyocyte hypertrophic remodelling [2].Although advances in therapeutics have improved the diagnostic outcome of patients with HF, the 5-year mortality rate remains high.Therefore, there remains a need to develop novel therapeutics to improve outcome in patients with HF.
Unfortunately, due to its failure to improve outcome in clinical trials for the treatment of glioblastoma the future of Cilengitide was hindered [20].Cilengitide is administered intravenously and whilst its affinity for αvβ3 is subnanomolar (IC 50 = 0.6 nM), it also has residual affinity for αvβ5 (IC 50 = 8.4 nM) and α5β1 (IC 50 = 15 nM) integrins [13,21].To overcome these limitations, second generation RGD-mimetics were developed [21,22].One such peptide, termed 29 [c(*vRGDA*A)] exhibits a favourable preference for αvβ3 (IC 50 = 0.6 nM) over both αvβ5 (IC 50 = 145 nM) and α5β1 (IC 50 = 21 nM) integrins, docks to αvβ3 in a similar mechanism to Cilengitide and has been engineered to be orally available [21,23,24].The orally available peptide prodrug 29P [c(*vR (Hoc) 2 GD(OMe)A*A)] is cleaved into the active hexapeptide 29 in plasma [21].In vitro, lower doses of 29 promote the upregulation of VEGFR2 protein expression in HUVECs and in vivo (50 μg/kg) it can increase angiogenesis of blood vessels within a subcutaneous Lewis Lung Cell tumour model similar to the effects of ldCil [21].
Here we show that 29P can prevent adverse cardiac remodelling and restore cardiac function in a transverse aortic constriction (TAC) pressure-overload model of cardiac hypertrophy.RNA sequencing suggests that 29P can modify key genes involved in cardiac hypertrophy, fibrosis and fatty acid metabolism, suggesting a new mode of action for 29P.

Animals
Male C57Bl/6 J (Envigo) aged between 7 and 8 weeks and weighing between 21 g-28 g were subject to TAC using methods detailed previously [25,26].Mice were housed in a pathogen-free, temperature-and humidity-controlled environment (21 ± 2 • C, light:dark cycle of 12 h) and were maintained in ventilated cages with access to standard chow diet and sterile water.The procedures described in this study were performed in accordance with the U.K. Animals (Scientific Procedures) Act (1986) and institutional guidelines for laboratory animal research at The University of Manchester, UK.

Transverse aortic constriction surgery
TAC was performed as detailed previously [25,26].Mice were anaesthetised (3 % isoflurane) and given 0.1 mg/kg buprenorphine subcutaneously.Mice were intubated and ventilated (200 breaths per minutes, tidal volume of 0.1 ml [Minivent, Harvard Apparatus]) and the anaesthetic reduced (2 % isoflurane).A partial thoracotomy between the left second and third rib was performed and then a 27G needle was placed between the brachiocephalic artery and left common carotid artery which was ligated with a 7-0 prolene suture.The needle was removed to yield a constriction 0.41 mm in diameter.The chest and skin were then closed with a 6.0 prolene suture.Sham operated mice were subject to the same surgical procedure without ligation.All mice were given 0.1 ml/kg of saline intraperitoneally post-surgery.Upon the termination of experiment, mouse hearts were excised and drained of blood in sterile saline solution.Hearts and kidneys were dried, weighed and then roughly sectioned to provide tissue for RNA, protein analysis and histology.Lungs were excised and weighed and the tibia length measured to normalise the heart, kidney and lung weight.

Oral gavage treatment with 29P
The oral prodrug 29P was kindly provided to use by Chaim Gilon, Amnon Hoffman, Michael Weinmüller and Horst Kessler [21].Mice were randomly assigned to receive either 6 doses of 50 μg/kg 29P or volume-matched saline vehicle (Veh) as control, delivered through oral gavage.Treatment commenced 1-week following TAC surgery for weeks with each gavage at least 48 h apart.Experiments were ended at two different time points following 29P treatment: either at 3 weeks post-TAC or at 6-weeks post-TAC, i.e. 3 weeks after cessation of 29P administration.

Transthoracic two-dimensional echocardiography
Mice were subject to echocardiography 1-, 3-and 6-week's post-TAC using the Vevo770 ultrasound machine and associated 14 MHz transducer (VisualSonics, FujiFilm).Mice were anaesthetised (1 % isoflurane) and heart rate maintained at 400-450 beats per minute.A parasternal short-axis M-mode image at the level of the papillary muscles was captured and used to estimate cardiac ejection fraction and fractional shortening (please see supplementary methods for calculations).

Histology
Full staining methodologies and analysis parameters are available in the supplementary methods.Briefly, heart mid-sections were fixed in % paraformaldehyde, processed (Leica) and embedded in paraffin wax. 5 μm thick sections were obtained on poly-L-lysine coated slides.Both Haematoxylin and Eosin (H&E) staining and Masson's trichrome staining was performed using an automated stainer (Leica).Isolectin β4 (Vector) was used to visualise the microvasculature.Briefly, rehydrated slides were incubated with hydrogen peroxide block and then subject to sodium citrate (pH 6)-mediated antigen retrieval.Avidin and biotin blocking (Jackson) was performed and slides were incubated with biotinylated isolectin β4 (1:100; Vector).Slides were then incubated with streptavidin peroxidase (Abcam), developed using the DAB chromogen/ DAB substrate method (Abcam) and counterstained with haematoxylin, dehydrated, cleared and mounted.All slide imaging was performed using the 3D-Histech Pannoramic-250 microscope slide-scanner (3D-Histech) and visualised using the Caseviewer software (3D-Histech).

Histological quantification
Cardiomyocyte cross-sectional area was obtained using Caseviewer (3D-Histech).A minimum of 100 transverse cardiomyocytes within the left ventricle (LV) were measured.The percentage of fibrosis within the LV was calculated using the semi-automated Caseviewer QuantCentre (3D-Histech) plugin.Fibre phenotypes were explored using The Workflow Of Matrix BioLogy Informatics [27] (TWOMBLI) plugin for ImageJ.Blood vessel images were cropped around longitudinal vessels and vessel length obtained using the FIJI (Image J) Skeletonize (2D/3D) [28] and Analyse skeleton plugin features [29].Vessel length was then expressed as a percentage of each image area.An average of 3-4 images were used for statistical analysis.

RNA sequencing of left ventricle apex at the 3-week experimental endpoint
The full method for RNA extraction and sequencing is detailed in the supplementary methods.

RNA extraction
RNA from snap frozen left ventricular apex tissue was extracted using the TRIzol (ThermoFisher)/chloroform method.

RNA sequencing
RNA sequencing was performed by the genomics facility at Queen Mary University London.
RNA quantification was performed using the Nanodrop, followed by qualitative RNA analysis using the Agilent 2100 Bioanalyser according to manufacturer's guidance (Agilent).The mRNA Library was prepared using NEBNext Ultra II (New England Biolabs) and quantification of mRNA libraries was performed using Qubit 2.0 fluorometer.mRNA fragment size was checked using the Agilent 4200 Tapestation (Agilent).All samples were processed for individual Illumina next generation sequence library denaturation and loading.Following the sequencing run, 2 out of 20 samples had a read per million value of <10 million and were excluded.The data was aligned to the mouse genome (Mm10 genome assembly) using Partek Flow (Star 2.3.7), with an alignment rate of 98 %.Genes and transcripts were annotated with Genocode 25 and the number of reads within exons on average 84 %.Normalisation of data was performed using counts per million (CPM) + 0.0001.

Single cell expression of Itgb3 in the heart following TAC
The publicly available Gene Expression Omnibus (GEO) database was searched for single cell RNA-Seq (https://www.ncbi.nlm.nih.gov/geo/).This identified the database GSE18072 (subseries of GSE180794 [32]) which contained data from endothelial cells, cardiomyocytes and fibroblasts at 1-week and 8-weeks post-TAC.The Log 2 [CPM] and the P-value were obtained from the spreadsheets available for download.

Statistics
All statistical analysis was performed using GraphPad Prism, each statistical test used is detailed within the figure legends (GraphPad, v9.2.0).

29P treatment ameliorates pressure overload induced cardiac hypertrophic remodelling and impeded cardiac function at 3 weeks post-TAC but not post-29P treatment cessation
To determine the cell types in which αvβ3 integrin is expressed post-TAC, a search of the GEO repository was performed.Utilising data from Froese et al., 2022 (GSE180720 [32]), at 1-week post-TAC, the gene encoding β3 integrin, Itgb3, was upregulated in isolated endothelial cells, cardiomyocytes and fibroblasts (Fig. 1A).By 8-weeks post-TAC, Itgb3 expression decreased to that of Sham control mice in endothelial cells and fibroblasts but remained significantly elevated in cardiomyocytes (Fig. 1A).These data indicate that within the heart, upregulated Itgb3 expression is transient depending on the cell type.
Treatment with 50 μg/kg 29P by oral gavage 1-week following surgery for 2 weeks had no significant effect on survival or body weight throughout the study (Fig. 1B and Supplementary Fig. 1A and B), suggesting no overt toxic effects, similar to Cilengitide.TAC initiates cardiac hypertrophy and chronic maladaptive remodelling can result in heart failure (HF).At the 3-week endpoint, increased heart weight to tibia length (Hw:Tl) and lung weight to tibia length (Lw:Tl) ratios were observed in Veh treated TAC mice when compared to its Sham control, indicative of cardiac hypertrophy and lung congestion, respectively (Fig. 1C).However, 29P-treatment partially prevented the increase in Hw:Tl ratio and lung congestion (Lw:Tl) was absent when compared to Veh TAC controls (Fig. 1C).We have previously shown that intervention at 3 weeks post-AAC with 3 weeks of ldCil, restores its cardioprotective effects at 6-weeks thus reversing maladaptive remodelling; effects which are sustained at 12-weeks [16].As such, we sought to determine if our model of TAC and 29P treatment could sustain its cardioprotective effects (Fig. 1B).However, at the 6-week experimental endpoint, i.e. 3 weeks following treatment cessation, 29P treatment showed no difference in Hw:Tl or Lw:Tl (Supplementary Fig. 1C).Overall, these data suggest that immediately after treatment cessation, but not 6 weeks post-TAC i.e., 3 weeks post-treatment cessation, 29P confers protection against cardiac remodelling and prevents early signs of lung oedema, a phenotype associated with cardiac failure.
Progressive remodelling ultimately affects cardiac function.Although initially there may be some element of compensation, over time, decompensation results in loss of function which can be characterised in vivo using transthoracic echocardiography (Fig. 1D).Parameters obtained pre-treatment showed reduced cardiac function (assessed Fig. 2. Analysis of upregulated and downregulated DEGs following TAC reveals 29P may impact expression of genes involved in the response to hypoxia and angiogenesis Venn diagram illustrating downregulated (A) and upregulated (B) genes following 29P treatment when compared to changes observed between Veh TAC and its respective Sham control.Overlapping genes are represented in the heat map with associated Z-score.Gene ontology enrichment analysis of these genes was performed and biological pathways of interest for each subset of DEGs is displayed as a box plot for downregulated (C) and upregulated (D) genes.through fractional shortening [FS] and ejection fraction [EF]) was present following 1-week TAC (Fig. 1E).However, at the 3-week experimental timepoint, 29P treatment of TAC mice significantly, albeit partially, prevented the deterioration of cardiac function when compared to Veh treated TAC controls (Fig. 1F).This effect was not sustained significantly at the 6-week experimental endpoint, 3 weeks following 29P treatment cessation (Supplementary Fig. 1D).
Changes to kidney mass have been reported following TAC when the aorta is similarly constricted with a 27G needle [33].In this study, following TAC, we observed a trend towards a reduced kidney mass and decreased kidney mass:TL ratio in the absence of any phenotypic differences (Supplementary Fig. 2A-D).At 3-week post-TAC, Veh treated TAC mice exhibit altered expression of key genes involved in the reninangiotensin aldosterone system, namely Ace (angiotensin converting enzyme) and Mme (membrane metalloendopeptidase) and treatment with 29P partially restored their expression levels towards those seen in Sham controls (Fig. 1G).However, at the 6-week experimental endpoint, the expression of Ace and Mme were comparable between Veh-and 29Ptreated TAC mice (Supplementary Fig. 2E).These data show that 29P treatment is unable to alleviate the loss of kidney mass following TAC.

29P alters genes associated with cardiac metabolism
Although 29P has been successfully administered in mouse cancer models prior to this study [21], its potential mechanism of action(s) have not yet been explored in the heart.To investigate changes to gene expression following treatment with 29P, left ventricle apical tissue from all groups was subject to bulk RNA-Sequencing.According to the exclusion criteria detailed in the methodology, 2 libraries corresponding to one 29P Sham and one Veh TAC were removed.Furthermore, following initial principal component analysis, a further Veh TAC mouse sample library was removed owing to a clustering association with Sham mice (data not shown).Removal of these three libraries followed by GSA analysis resulted in the identification of 13,800 genes.Hierarchical clustering and principal component analysis of the remaining 17 libraries revealed a clear clustering based on type of surgery and not to treatment type (Supplementary Fig. 3A and B).Changes to gene expression after TAC have been explored previously and more recently through interrogation of murine TAC with reference to human aortic stenosis datasets, all of which are available through the GEO repository [34][35][36][37].For this study, we focus on changes to gene expression caused by the administration of 29P.Regarding 29P sham mice, RNA sequencing revealed 942 altered transcripts between Veh Sham and the 29P treated Sham group (Supplementary Table 1).Gene ontology biological process (GOBP) enrichment analysis of these DEGs revealed enrichment of genes in processes involved in cytoplasmic translation, ribosomal small subunit assembly and various mitochondrial functions (data not shown).However, as we observed no differences to survival or the overall phenotype and cardiac function in 29P Sham mice, the following analysis will focus on changes to gene expression in 29P TAC mice.
Firstly, we obtained DEGs between Veh TAC and Veh Sham and then compared this gene list with DEGs downregulated following 29P treatment (Fig. 2A).This identified 129 genes of interest, GOBP enrichment analysis identified pathways key to pathological features associated with TAC such as ECM function(s), angiogenesis and various metabolic processes (Fig. 2C).We then performed the opposing analysis and identified downregulated DEGs between TAC Veh and Sham Veh and interrogated these against DEGs upregulated following 29P treatment.From this analysis, 79 genes were identified (Fig. 2B).Of these genes, GOBP enrichment terms were related to fatty acid and metabolic processes and the response to insulin (Fig. 2D).To determine if the genes within the GOBP pathways interacted and to what extent, Cytoscape STRING network analysis of DEGs between Veh-and 29P-treated TAC mice was performed.STRING identified a large network with a subset of genes forming a large interacting subcluster encompassing both upregulated and downregulated transcripts (Fig. 3A).Analysis of the subcluster revealed an enrichment to pathways important in metabolic processes (Fig. 3A).TAC-induced changes to several metabolic genes were small but partially rescued with 29P treatment (Fig. 3B).

Enrichment analysis of DEGs suggests a putative role for 29P in the response to hypoxia and regulation of angiogenesis
Changes to the cardiac microvasculature have been widely reported following TAC, with an initial reduction vessel density followed by a recovery period over 7-14 days [38].We have reported previously that ldCil increased the density of blood vessels within the myocardium following AAC [16] and that 29P increased tumour vasculature in vivo [21].DEGs identified following RNA-sequencing, suggest that 29P may alter genes involved in the regulation of angiogenesis and hypoxia (see Fig. 2C).Specifically, when compared with Veh TAC mice, 29P-treated TAC mice exhibit a reduction in the expression of Hif1α (Hypoxia inducible factor 1-α), a known regulator of hypoxia; Serpine1 (Serpin Family E Member 1), a negative regulator of angiogenesis [39]; Amot (Angiomotin), which is known to have both pro-and anti-angiogenic effects [40]; Sulf1 (Sulfatase 1), an inducer of postinfarct angiogenesis [41] and Xdh (Xanthine Dehydrogenase), an inhibitor of angiogenesis [42] (Fig. 4A).Furthermore, in 29P-treated TAC mice there was an increase in the expression of Itgb3, the gene encoding β3 integrin, when compared to its respective Sham control but no significant changes were found between Veh groups (Fig. 4B).However, at the 6-week timepoint there were no changes to Hif1α or Itgb3 expression between groups (Supplementary Fig. 4A and B).
Owing to the proposed mode of action of 29P and the DEGs identified in this study, we sought to determine if changes to the cardiac microvasculature were present in the heart following TAC (Fig. 4C).Comparing analysis of the vascular area at the 3-week timepoint between Veh treated Sham and Veh treated TAC showed that there was no significant loss of vessels in this model.29P had no apparent effect on vessel area after TAC when compared with Veh treated mice (Fig. 4D).At the 6-week timepoint i.e. 3 weeks post-treatment cessation, no change to vessel area was observed between any of the groups (Supplementary Fig. 4C).

29P treatment reduced left ventricular interstitial fibrosis and cardiomyocyte hypertrophy at 3 weeks post-TAC
The induction of pressure-overload causes reactive fibrosis, in which collagen is deposited in the absence of widespread cardiomyocyte death [43].At 3 weeks post-TAC, both TAC groups showed an increase in the percentage of interstitial fibrosis and branchpoints when compared to their respective Sham controls.However, at 3 weeks post-TAC we observed reduced fibrosis and fewer branchpoints per total length of fibrosis in the 29P TAC treatment group when compared to Veh TAC controls (Fig. 5A), suggesting 29P treatment can reduce the occurrence of reactive fibrosis.
In addition, utilising RNA-seq, we identified a reduction in Postn (Periostin), Ccn5 (cellular communication network 5) and Adamts1 (ADAM metallopeptidase with thrombospondin type 1 motif 1) in 29P treated TAC mice; these genes have also been implicated in collagen fibrillogenesis and ECM organisation (Fig. 5D).These data suggest that 29P treatment can alter the expression of some key genes involved in the deposition of fibrosis and ECM reorganisation which are associated with a reduction in interstitial fibrosis.
Another hallmark of pressure-overload induced cardiac hypertrophy is the concomitant increase in cardiomyocyte cross-sectional area (CSA).At 3 weeks post-TAC, cardiomyocyte CSA increases in both Veh-and 29P-treated TAC mice when compared with their respective Sham controls (Fig. 6A).Interestingly, 29P treatment partially rescued the increase in cardiomyocyte CSA when compared to Veh-treated TAC mice (Fig. 6A).To determine if changes to genes associated with cardiac hypertrophy were observed following 29P treatment, we again interrogated DEGs identified in our RNA-seq database using annotated human hypertrophic pathways derived from MSigDB (v7.5.1 [M35413] and [M14043 [45]]) (Fig. 6B).Genes in these pathways were explored due to their annotation with pathological changes including an increase width of the LV wall with loss of elasticity and enlargement of LV mass.29P treatment rescued the expression of some genes from these annotated pathways and thus these data suggest that 29P may be involved in the regulation of these processes (Fig. 6C).Furthermore, TAC resulted in an increased expression of additional well-characterised pro-hypertrophic genes such as Nppa, Nppb (data not shown) and Xirp2 (Xin Actin Binding Repeat Containing 2).Treatment with 29P partially restored the expression level of Xirp2 towards those observed in Sham controls (Fig. 6D).

Cardioprotective properties conferred by 29P treatment are lost at 6 weeks post-TAC i.e. 3 weeks following treatment cessation
As discussed briefly above, 3 weeks following treatment cessation (i.e. at 6 weeks post-TAC) functional data derived from echocardiography showed that there was no difference between TAC groups, suggesting that the cardioprotective functional effects conferred by 29P at 3-weeks post-TAC are lost following treatment withdrawal.At the 3-week timepoint we identified structural changes in TAC mice treated with 29P; however, 6-weeks post-TAC, there was no difference to percentage fibrosis in the left ventricle, TWOMBLI-assessed number of fibrosis branchpoints (Supplementary Fig. 5A and B) or Col4a3 and Col4a4 transcripts between Veh-and 29P-treated TAC mice (Supplementary Fig. 5C).We then evaluated cardiomyocyte CSA and observed an increased cell area after TAC but no difference between Veh and 29P treatment groups (Supplementary Fig. 5D).Overall, the beneficial effects of 29P treatment are lost 3 weeks post-treatment cessation.

Discussion
Despite advances to our knowledge and understanding of HF, there remains an unmet clinical need for novel therapeutics which can both prevent and reverse established cardiac remodelling with minimal to no side effects.Integrins are well characterised interactors of the extracellular matrix, making them therapeutically targetable [46].Previous studies have implicated β3 integrin and its downstream signalling as a disease-modifying mediator of cardiac remodelling [47,48].These studies established a role for low doses of the RGD-mimetic, αvβ3/αvβ5 integrin-specific, cyclic peptide Cilengitide in preventing adverse cardiac remodelling and increasing BV density in an AAC model of pressure-overload cardiac hypertrophy [16].Given these promising results in vivo, we sought to determine if the second generation orally available RGD-mimetic prodrug 29P, which we have shown harbours a pro-VEGF stimulated angiogenic effect in subcutaneous tumours and in ex vivo aortic ring sprouting assays [21], would confer similar cardioprotective properties to ldCil in a TAC model of pressure-overload induced cardiac hypertrophy and HF.
The data in this study demonstrate the cardioprotective capability of 29P.Specifically, at 3 weeks post-TAC following 2 weeks of treatment, 29P can improve cardiac function, attenuate cardiomyocyte hypertrophy and reduce interstitial fibrosis.The ability of 29P to confer these different cardioprotective affects correlate with increased β3 integrin transcript expression after TAC.Itgb3 is detectable in cardiac endothelial cells, cardiomyocytes and cardiac fibroblasts and its expression is elevated in all these cell types at 1-week post-TAC [32], suggesting Itgb3 may play a potential role in angiogenesis, hypertrophy and reactive fibrosis in the acute phase of pressure overload, respectively.ldCil treatment was shown to alter the profile of putative and known cardioprotective regulators [16].Similarly, here, apical left ventricle RNA-seq at 3 weeks post-TAC reveals changes to the expression profile of known and putative genes involved in cardiac metabolism, hypertrophy, deposition of the basement membrane and angiogenesis.Treatment with 29P restores key transcriptomic profiles towards levels observed in the Sham controls.Direct involvement between the metabolism genes identified in this study and αvβ3 has not been well documented.However, in glioblastoma cells, αvβ3 interaction with the ECM protein osteopontin promotes aerobic glycolysis [49], and β3 knockout mice have a higher serum triglyceride level caused by disrupted lipoprotein lipase secretion [50].Therefore, αvβ3 in other models can help regulate aspects of metabolism but in the heart a function for αvβ3 has not been elucidated.
In this study, 29P significantly reduces the deposition of collagen fibres and cardiomyocyte hypertrophy throughout the LV.In accordance Fig. 4. 29P treatment alters some angiogenic modulators but does not impact vessel area.A Expression of transcripts involved in hypoxia and angiogenesis which are differentially expressed between Veh-and 29P-treated TAC mice.Gene panel: Hif1α (Hypoxia inducible factor 1-α), Serpine1 (Serpin Family E Member 1), Amot (Angiomotin), Sulf1 (Sulfatase 1) and Xdh (Xanthine Dehydrogenase).B Itgb3 is upregulated in 29P treated TAC mice when compared to its respective Sham control.Sham Veh n = 5, Sham 29P n = 3, TAC Veh n = 3, TAC 29P n = 6, data analysed using DeSeq2 (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).C Representative images of isolectin stained left ventricular myocardium at 3 weeks post-TAC cropped to longitudinal vessels used for analysis in (D).Scale bar = 200 μm.Inset scale bar = 50 μm.D Percentage of vessels per tissue area is reduced in 29P treated TAC mice when compared to their respective sham controls.Sham Veh n = 7, Sham 29P n = 7, TAC Veh n = 7, TAC 29P n = 7, data analysed using two-way ANOVA with Tukey's post-hoc multiple comparison test (*p ≤ 0.05).with this, expression of collagen genes such as Col4a3 and Col4a4 as well as the gene encoding the ECM protein NID1 are all downregulated.NID1 is an important component of the vascular basement membrane and acts as a linker protein of COL4 and laminin [51].Recent in vitro studies have implicated NID1 in the transdifferentiation of fibroblasts into ECM depositing myofibroblasts [52] and in the activation of the mitogen activated protein kinase pathway in cardiomyocytes through αvβ3 binding [53].This potential role of NID1 signalling offers a potential mechanism of action of 29P treatment in non-endothelial cells.RNA-seq also revealed a reduction in the cardiac stress marker Xirp2 in 29P TAC mice.Xirp2 has been identified through single cell sequencing to be upregulated in sub-clusters of cardiomyocytes from human hypertrophic cardiomyopathy cardiac samples [54] and mutations within Xirp2 are linked to dilated cardiomyopathy [55] highlighting a further potential   cardioprotective mechanism of 29P.
There is a growing body of evidence that cardiac microvasculature rarefaction in the failing heart leads to worsening cardiac perfusion and poor prognosis [56], indeed HF onset can be delayed in preclinical models when angiogenesis is stimulated [57].We have shown that 29P increases tumour microvasculature in vivo [21], despite this we fail to observe a change to microvascular area in the heart at 3 weeks post-TAC, following 2 weeks of 29P treatment.This is also in contrast to our previous study in which ldCil treatment after AAC induced an increase in the cardiac microvasculature [16]; this disparity could be due to a number of reasons: 1) the model of HF utilised differs i.e.TAC vs AAC, 2) the time points at which treatment was commenced and its duration, 3) the experimental time point at which BV were assessed, 4) the antifibrotic and anti-hypertrophic effect of 29P reduces the overall hypoxic environment in the LV myocardium.In keeping with the latter point, Hif1α levels are reduced in 29P-treated TAC mice when compared to Veh TAC controls.
In this present study, 29P was cardioprotective at 3 weeks post-TAC, following 2 weeks of treatment.Whilst the features described in detail above indicate that 29P can help prevent the typical progression of TAC, they do not address whether the cardioprotective benefits were a permanent feature or reversible.To investigate this, we withdrew treatment at 3 weeks post-TAC and continued to monitor TAC progression through to 6 weeks.Here, we showed that when 29P treatment is halted, its cardioproctective effects are lost.The apparent loss of cardioprotection could be due to the simple fact that the aorta remains banded, therefore the driving force behind cardiac decompensation i.e. pressure-overload, is still present.As such, we could speculate that 29P confers its affects in a temporal, reversible manner and sustained administration of 29P would be required for prolonged cardioprotective effects.
Compound 29P has been engineered to be orally available and binds more with more affinity to αvβ3 over any other integrin [21] thus overcoming one of the key challenges of peptide/antibody promiscuity in integrin targeting.The first generation RGD mimetic pentapeptide Cilengitide has been administered in around 30 clinical trials for cancer and, although largely well-tolerated, does have some known adverse events (haematological, neuropathological, liver toxicity) [58].In this study, treatment with 29P did not result in any notable adverse events in either control (sham) or TAC treated groups.Control sham mice were subject to the same 29P treatment and importantly, did not display any changes to cardiac structure, function or overall survival.However, whole RNA sequencing did reveal changes to transcripts under these baseline conditions which could warrant further investigation under long-term conditions to ensure these changes do not lead to any adverse events.αvβ3 integrin has also been targeted by a variety of other therapeutics with some progressing to clinical trials.These include the small molecules SFO166, Risuteganib and SB-273005 for the treatment of retinopathies and osteoporosis and a humanised monoclonal antibody, VPI-2690B, for the treatment of diabetic nephropathy [59].Of the aforementioned αvβ3-targeting therapies only the monoclonal antibody is specific for αvβ3; however, all were reported to be safe, further suggesting that 29P would be well tolerated.29P has been previously utilised by our group in a murine model of Lewis Lung carcinoma and was well-tolerated with no observed side-effects or off-target effects when administered either orally or given in its protracted 29 form [21]. Overall, our results suggest that utilising 29P for the treatment of pathological hypertrophic remodelling has potential and warrants further long-term investigation.Although beyond the scope of this present study, given that 29P can reduced fibrosis and improve cardiac function, it would be interesting to explore the use of 29P for treatment of post-myocardial infarction.

Limitations
One of the main aims of this study was to determine the use of the orally available prodrug 29P over several weeks after TAC surgery.
However repeated oral gavage in mice can cause adverse effects over several weeks that could affect the wellbeing of the animals.Thus sustained long-term studies would likely require the use of a mini-pump device to either administer the cleaved product 29 subcutaneously or 29P into the stomach through a catheter attached to the mini-pump.However, these also have a limited capacity of maximum 6 weeks before requiring further surgery for replacement.The study has been limited to 3-and 6-week time points.Future studies would determine effects between these time points in order to pinpoint when the beneficial effects of 29P treatment stop after treatment cessation.

Declaration of competing interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Kairbaan Hodivala-Dilke reports a relationship with Ellipses Pharma that includes: consulting or advisory.Kairbaan Hodivala-Dilke reports a relationship with Vasodynamics that includes: consulting or advisory.Kairbaan Hodivala-Dilke reports a relationship with RGDscience Ltd that includes: consulting or advisory.Horst Kessler and Amnon Hoffman and Michael Weinmuller has patent #WO2019058374A1 pending to YISSUM RES DEV CO OF HEBREW UNIV JERUSALEM LTD [IL]; UNIV MUENCHEN TECH [DE].Kairbaan Hodivala-Dilke has patent #WO2021032955A1 pending to UNIV LONDON QUEEN MARY [GB].If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.tissue at 3 weeks post-TAC is available at the GEO repository, GSE247309.
AN British Heart Foundation (PG/18/75/34096).LL and EJC British Heart Foundation (FS/18/62/34183).AC and EJC British Heart Foundation (FS/4yPhD/20/34131).NS British Heart Foundation (RG/F/21/ 110055).JH, MZ and SP are employed by the University of Manchester.CG, AH, MW and HK did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.EM and JW acknowledge the support from the CRUK City of London Major Centre core funding to Barts Cancer Institute.JW is HEFCE funded by Queen Mary University of London.KHD is employed by Queen Mary University of London and is HEFCE funded by Queen Mary University of London, Barts Cancer Institute, CRUK City of London Major Centre; LW was employed at Queen Mary University London, Barts Cancer Institute and supported by the CRUK City of London Major Centre.