Sex‐related differential susceptibility to ponatinib cardiotoxicity and differential modulation of the Notch1 signalling pathway in a murine model

Abstract Ponatinib (PON), a tyrosine kinase inhibitor approved in chronic myeloid leukaemia, has proven cardiovascular toxicity. We assessed mechanisms of sex‐related PON‐induced cardiotoxicity and identified rescue strategies in a murine model. PON+scrambled siRNA‐treated male mice had a higher number of TUNEL‐positive cells (%TdT+6.12 ± 0.17), higher percentage of SA‐β‐gal‐positive senescent cardiac area (%SA‐β‐gal 1.41 ± 0.59) and a lower reactivity degree (RD) for the survival marker Bmi1 [Abs (OD) 5000 ± 703] compared to female (%TdT+3.75 ± 0.35; %SA‐β‐gal 0.77 ± 0.02; Bmi1 [Abs (OD) 8567 ± 2173]. Proteomics analysis of cardiac tissue showed downstream activation of cell death in PON+siRNA scrambled compared to vehicle or PON+siRNA‐Notch1‐treated male mice. Upstream analysis showed beta‐oestradiol activation, and downstream analysis showed activation of cell survival and inhibition of cell death in PON+scrambled siRNA compared to vehicle or PON+siRNA‐Notch1‐treated female mice. PON+scrambled siRNA‐treated mice also had a downregulation of cardiac actin—more marked in males—and vessel density—more marked in females. Female hearts showed greater cardiac fibrosis than their male counterparts at baseline, with no significant change after PON treatment. PON+siRNA‐scrambled mice had less fibrosis than vehicle or PON+siRNA‐Notch1‐treated mice. The left ventricular systolic dysfunction showed by PON+scrambled siRNA‐treated mice (male %EF 28 ± 9; female %EF 36 ± 7) was reversed in both PON+siRNA‐Notch1‐treated male (%EF 53 ± 9) and female mice (%EF 52 ± 8). We report sex‐related differential susceptibility and Notch1 modulation in PON‐induced cardiotoxicity. This can help to identify biomarkers and potential mechanisms underlying sex‐related differences in PON‐induced cardiotoxicity.


| INTRODUC TI ON
Tyrosine kinase inhibitors (TKIs) are a group of drugs capable of selectively inhibiting tyrosine kinases, 9 the constitutive expression of which leads to de-regulated cell proliferation and tumour formation. 3 In chronic myelogenous leukaemia (CML), a reciprocal translocation (9; 22) (q34; q11), known as the 'Philadelphia chromosome' (Ph), results in the formation of a mutated tyrosine kinase protein, called Abelson-Breakpoint Cluster Region (BCR/ABL). Ponatinib (PON, trade name: Iclusig, from Incyte Biosciences, Epalinges, Switzerland) is the only TKI capable of inhibiting BCR/ABL3, thus leading to growth arrest and apoptosis of the neoplastic clone. 17 However, a significant number of patients treated with PON, asin general-with all TKIs, develops vascular and cardiac toxicity, with evidence of congestive heart failure and electrocardiographic abnormalities. 39 Although several mechanisms have been hypothesized for PON-induced cardiotoxicity, the in vivo effects of PON on the heart have not been fully characterized.
Notch1 is a master regulator of B-and T-cell development. 32,34 While in some types of leukaemia and T-cell lymphoma the aberrant activation of the Notch1 pathway due to 5' promoter deletions is associated with atypical myloproliferative disease, 36 in human CML cells Notch1 signalling plays a tumour suppressive role, and its overexpression may inhibit tumour proliferation. 1,43 Therefore, targeting Notch1 may represent a therapeutic approach for CML. However, Notch1 plays also a central role in vascular development. 11 The hyperactivation of Notch1 in vessels leads to abnormal vascular development and vascular dysfunction. 1 We have previously shown that selective blockade of Notch1 can prevent PON-induced vascular toxicity in human aortic endothelial cells. 24 Therefore, activation of Notch1 in CML cells by PON can be considered as the 'on-target effect' on the tumour.
The extent of PON-induced cardiotoxicity is quite variable, and this is a confounding aspect for identifying its mechanisms. 6,38 In particular, little is known about the molecular mechanisms of PONinduced cardiotoxicity and the possible differences between males and females. Such type of research can help to identify biomarkers and strategies to counteract PON-induced cardiotoxicity. Here, we report a sex-related differential susceptibility to PON-induced cardiotoxicity and identify Notch1 as the differentially affected signalling pathway.

| Animals and mice randomization
Male and female C57BL/6 mice (body weight: 30 ± 4 g, 24 months old) were purchased from Charles River Italia (Lecco, Italy). Mice were housed under a 12-hour light/dark cycle (7 am-7 pm) in temperature-and humidity-controlled rooms and were provided with ad libitum rodent chow (Teklad 7001, 4

| Preparation of transit TKO-siRNA complexes and siRNA delivery in vivo
Non-viral siRNA delivery in vivo was done by tail vein injection.

| Tissue preparation and analysis of fibrosis
After induction of anaesthesia, the hearts were removed and embedded in OCT without fixation. The blocks from each heart were cut transversely to obtain 30 sections (each 5 μm thick) from the middle of the ventricles. All histological sections were analysed blindly. Total cardiac fibrosis (including interstitial, perivascular and coronary arterial fibrosis) was assessed by Masson's trichrome staining. Digital colour images were obtained, and the extent of left ventricular (LV) fibrosis was evaluated with the NIH Image J software (Media Cybernetics, Rockville, MD, USA, http://rsb.info.nih.gov/ij) in 5 cardiac areas randomly selected for each heart section and expressed as % of the total LV area.

| Terminal Deoxyribonucleotidyl Transferasemediated dUTP Nick End Labeling (TUNEL) Assay
Detection of nuclei with fragmented DNA by TUNEL was performed using the HRP-DAB TUNEL assay kit (Abcam, Cambridge, UK) according to the manufacturer's instructions and a previously published protocol. 21 Methyl green (DakoCytomation) was used as a counterstain. The myocardial apoptotic index was calculated as the number of positive cardiac cells for field/total cardiac cell number for field x 100.

| Senescence-Associated β-galactosidase Assay
The effect of PON with/without siRNA-Notch1 or scrambled siRNA on cardiac senescence was evaluated by SA-b-gal Activity Senescence-Associatedβ-Galactosidase Staining (Cell Biolabs, Inc, San Diego, CA, USA), as previously described. 24 Cardiac frozen sections were post-fixed in 4% paraformaldehyde at room temperature for 15 min, rinsed with sterile PBS and incubated overnight with fresh senescence-associated β-galactosidase staining solution (1 mg/mL X-gal in 40 mM citric acid/sodium phosphate, pH 6.0, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, 150 mM sodium chloride, 2 mM magnesium chloride) at 37 °C. Then, the staining solution was removed, sections were mounted in 70% glycerol, and the development of blue colour was assessed under standard light microscope. The extent of the blue-stained area was evaluated with the Image J software in 5 randomly chosen cardiac areas per each heart section and expressed as % of total LV area.

| Histological evaluation of sarcomeric organization and vessel density
For immunofluorescence staining of sarcomeres and arterioles, 5 μm OCT sections were permeabilized, blocked for 30 min in PBS containing 1% bovine serum albumin and incubated for 1 hour at 4 °C with a primary antiα-smooth muscle actin (ASMA, Sigma Aldrich, St. Louis, MI, USA) or an anti-sarcomeric α-actinin (Sigma) antibody. A non-immune IgG (Becton, Dickinson and Co., Franklin Lakes, New Jersey, USA) was used as isotype control. After washing with PBS, cardiac sections were stained with a fluorescein isothiocyanate (FITC)-conjugated secondary antibody, washed, mounted and viewed under an immunofluorescence microscope. Arterioles were counted blindly in 5 randomly selected fields at 10× magnification. Vascular images were captured by using an inverted light microscope (Olympus IX71) and analysed by using Image J software.
Vessel density was evaluated in 5 randomly chosen cardiac areas per each heart section and expressed as % of total LV area.

| Immunoblotting
To examine the effects of PON with/without siRNA-Notch1 or scrambled siRNA on the levels of cardiovascular and fibrosis markers, total proteins from cardiac samples were isolated in an ice-cold RadioImmuno Precipitation Assay (RIPA) lysis buffer containing 10 mmol/litre Tris, pH 7.4, 1% SDS and 1× protease inhibitor

| Echocardiography
We performed transthoracic echocardiography 1 month after treatments to assess the functional effects of each treatment using a portable ultrasound apparatus (Esaote; Genoa, Italy) equipped with a 21-MHz linear probe according to detailed protocols described in the Online Supplement.

| Statistical Analysis
Data are expressed as mean ±standard deviation (SD). Two-group comparisons were performed by using the Student's t test for unpaired values. Multiple-group comparisons were performed using analysis of variance and the Gabriel or Tukey Honestly Significant Difference (HSD) post-hoc test to determine statistical significance within and between groups. P-values <0.05 were considered statistically significant.

| Ponatinib induces differential apoptosis, senescence and expression of the survival marker Bmi1 in male and female hearts
We examined whether there is evidence of apoptosis and senescence of cardiac cells in our ponatinib-induced cardiotoxicity model and whether there is a differential effect according to sex. In the ventricles of PON+scrambled siRNA mice, the total number of TUNEL-positive cells was fivefold and threefold higher in males and females, respectively, than in vehicle-treated controls, with a statistically significant sex-related difference. Indeed, male mice featured twofold greater numbers of TUNEL-positive cells than females ( Figure 1A and B). These effects were reversed by co-treatment with siRNA-Notch1, suggesting that PON acts on cardiac apoptosis differently in both sexes via the Notch-1 signalling pathway ( Figure 1A and B).
Senescence was identified by the expression of senescenceassociated β-galactosidase (SA β-gal). In the ventricular myocardium of PON+scrambled siRNA mice, the percentage of SA β-gal-positive senescent cardiac area was onefold higher in male mice than in vehicle-treated controls (Figure 2A, B). Female mice, however, exhibited a constant SA β-gal-positive senescent cardiac area with no statistically significant PON effect compared to vehicle-treated controls (Figure 2A, B). The effect of PON on cardiac senescence was reversed by co-treatment with siRNA-Notch1, which is in line with the results obtained with TANEL staining (Figure 2A, B).
Immunohistochemical staining showed a prominent effect of PON in female mice with a fourfold increase in expression of the Bmi1 prosurvival marker compared to vehicle, while the increase was mild in the male counterpart ( Figure 3A, B). In female mice, these effects were almost totally reversed by co-treatment with siRNA-Notch1, suggesting that PON acts on endogenous cardiac expression of Bmi1 differently in both sexes via the Notch-1 signalling pathway.

| Ponatinib induces differential myofibre structure disruption and cardiac actin reduction in male and female hearts
We monitored the anti-cardiac alpha-actinin immune reactivity to ascertain the impact of PON on myofibre structure ( Figure 4A).

| Proteomics analysis reveals the differential up-and downregulation of cell death and survival in male and female hearts exposed to ponatinib
We carried out label-free proteomics analysis in the hearts of male  Table 1 as a heatmap visualization of predictive z-score generated by IPA. In detail, the orange colour indicates the predicted activation (z-score ≥2.00), while the blue one highlights the predicted inhibition (z-score ≤-2.00). The colour intensity is directly proportional to the significance of the predicted activation or inhibition.
On the other hand, the protein cargo of male hearts treated with PON+scrambled siRNA was able to trigger more cellular functions related to activation of 'necrosis of muscle' (p-value =1.81x10 −17 , z-score =2.23) and 'cell death muscle cells' (p-value =2.16x10 −4 , zscore =2.28). Moreover, as shown in Figure 4, the protein cargo of male hearts treated with PON+scrambled siRNA was able to significantly activate the 'production of reactive oxygen species' (p-value =1.15x10 −32 , z-score =2.14) together with their 'synthesis' (p-value =2.88x10 − 25, z-score =3.12) and 'metabolism' (p-value =2.42x10 − 29, z-score =2.75) compared to male mice treated with PON+siRNA-Notch1. These data show the sex-related differential susceptibility to PON-induced cardiotoxicity, which could be reversed by the Notch1 knockdown. We used differential proteins obtained from male and female hearts treated with PON+scrambled siRNA or ve-  Figure 5B). These data could further confirm that PON acts on the heart via the Notch-1 signalling pathway and that both NOS2 and β-oestradiol are differentially involved in the negative regulation of this pathway.  (Figure 6 A, B) with no statistically significant differences between female and male mice. Female mice, however, showed constantly lower expression of Flt1 then male counterparts, with no statistically significant PON effect compared to vehicletreated controls ( Figure 6A, B). The effects of PON were reversed by PON-siRNA-Notch1 co-treatment, suggesting that PON acts on angiogenesis via the Notch1 signalling pathway ( Figure 6A, B).

TA B L E 1 Differentially activated and inhibited diseases and biofunctions Diseases and Biofunctions
Female Note: Differentially activated and inhibited diseases and biofunctions in female and male hearts treated with PON+siRNA-scrambled compared to those co-treated with PON+siRNA-Notch1. Table reports the predictive z-score (orange activation z-score ≥2.00; blue inhibition, z-score ≤ −2.00). Concentrations reached of reactive oxygen species (ROS) refers to the final output of reactive oxygen species (ROS) that is the result of synthesis/production and degradation.

F I G U R E 5
Upstream network analysis of male and female mice exposed to ponatinib and Notch-1 signalling inhibition. Panel A-B: The image shows upstream activation of the β-oestradiol gene (A) in female PON-treated compared to female vehicle-treated mice, resulting in a z-score of 4.28, and the upstream activation of inducible nitric oxide synthase (NOS2) gene (B) in male PON-treated compared to male PON+siRNA-Notch1-treated mice, resulting in a z-score of 2.03 respectively. Blue colour keys indicate inhibited regulators, while orange ones indicate the activated regulators. The intensity of colours is proportional to the score prediction value. z scores >2.0 indicate that a molecule and/or function is activated

| Ponatinib exerts anti-fibrotic effect in male and female hearts independently from Notch1 signalling pathway
Cardiac remodelling in aged hearts is a multifaceted process that includes activation of fibroblasts and a complex immune response. 27 In our experiments, total fibrosis was lower in vehicle-treated male mice than in female counterparts, which confirms previous observations, 12 and PON further attenuated collagen deposition ( Figure 7A, B). This decrease in fibrosis was paralleled by a decrease in type I collagen expression ( Figure 7C). There were no significant changes in collagen deposition and type I collagen expression after co-treatment with PON+siRNA-Notch1 ( Figure 7A-C), suggesting that Notch1 is not a direct mediator of the inhibitory effect of PON on fibrosis.

| Ponatinib induces systolic but not diastolic dysfunction in female and male mice
To

| DISCUSS ION
In the present study, we show a sex-related different susceptibility to PON-induced cardiotoxicity, higher in male than female mice, as revealed by (1)  The TUNEL assay is commonly used to detect DNA fragmentation, resulting from apoptotic signalling cascades. 23 We here  We acknowledge some limitations of our study such as the low number of mice per group, due to the higher mortality in the PONtreated groups of mice, and the short follow-up limited to 1 month.
In conclusion, we identified a sex-related different susceptibility to PON-induced cardiotoxicity in old C57BL/6 mice and delineated Notch1 as the signalling pathway for PON-induced cardiotoxicity in male and female mice. This may help to identify biomarkers and potential mechanisms underlying sex-related differences in PONinduced cardiotoxicity, which might enable significant clinical improvements in the effectiveness of PON therapies.