Streptozotocin-Induced Type 1 and 2 Diabetes Mellitus Mouse Models Show Different Functional, Cellular and Molecular Patterns of Diabetic Cardiomyopathy

The main cause of morbidity and mortality in diabetes mellitus (DM) is cardiovascular complications. Diabetic cardiomyopathy (DCM) remains incompletely understood. Animal models have been crucial in exploring DCM pathophysiology while identifying potential therapeutic targets. Streptozotocin (STZ) has been widely used to produce experimental models of both type 1 and type 2 DM (T1DM and T2DM). Here, we compared these two models for their effects on cardiac structure, function and transcriptome. Different doses of STZ and diet chows were used to generate T1DM and T2DM in C57BL/6J mice. Normal euglycemic and nonobese sex- and age-matched mice served as controls (CTRL). Immunohistochemistry, RT-PCR and RNA-seq were employed to compare hearts from the three animal groups. STZ-induced T1DM and T2DM affected left ventricular function and myocardial performance differently. T1DM displayed exaggerated apoptotic cardiomyocyte (CM) death and reactive hypertrophy and fibrosis, along with increased cardiac oxidative stress, CM DNA damage and senescence, when compared to T2DM in mice. T1DM and T2DM affected the whole cardiac transcriptome differently. In conclusion, the STZ-induced T1DM and T2DM mouse models showed significant differences in cardiac remodeling, function and the whole transcriptome. These differences could be of key relevance when choosing an animal model to study specific features of DCM.


Introduction
The prevalence of diabetes mellitus (DM) is continuously increasing at a frightening rate. The World Health Organization (WHO) reported that 108 million adults affected by DM in 1980 increased to 422 million in 2014, and this trend is only steeply increasing [1,2]. The main cause of morbidity and mortality in diabetic patients is cardiovascular complications [3][4][5]. DM increases the risk of heart failure (HF) up to fivefold [6][7][8][9][10]. The clinical outcomes associated with HF are considerably worse for patients with DM than for

STZ-Based T1DM and T2DM Mouse Models Affected Global Left Ventricular Function Differently
To assess whether STZ-derived mouse models of type 1 and type 2 DM cause different phenotypes of cardiomyopathy, a single high dose of STZ (200 mg/Kg) was injected into 10-week-old male and female C57BL/6J mice to cause a total depletion of pancreatic β cells, generating T1DM mice [16]. T2DM mice were generated through HFD for four weeks, starting at six weeks of age, and, during the fourth week, via daily injections of low-dose STZ (40 mg/Kg) over four consecutive days [22]. Control mice were fed with a normal chow diet (NCD) or an HFD without STZ. Four weeks after high/low doses of STZ, all treated animals were diabetic, showing altered fasting glycemia levels when compared to control mice that were fed with normal as well as high-fat diets (NCD = 150.1 ± 18.9 mg/dL; HFD = 155.5 ± 14.06 mg/dL; T1DM = 455.3 ± 69.3 mg/dL; T2DM = 345.8 ± 70.2 mg/dL; p-value < 0.0001). Hyperglycemia was maintained through eight weeks after STZ administration in both T1DM and T2DM (T1DM = 440.6 ± 57.3 mg/dL; T2DM = 351 ± 63.8 mg/dL).
To assess whether STZ-based T1DM and T2DM affected cardiac systolic and diastolic function differently or similarly, in vivo M-mode parasternal long-axis echocardiography and flow and tissue Doppler imaging were performed at eight weeks following STZ treatment. Considering that HFD per se does not alter histology or cardiac structure or function over twelve weeks (see [22]), to simplify presentation of these data, we included in this analysis only mice fed with the normal chow diet as a control (CTRL).
As previously reported [22], in the mice with T2DM, an increase in the ratio of early transmitral flow velocity (E wave) to early mitral annulus tissue velocity (E wave), an index of left ventricular (LV) filling pressure, was detected (30% increase as compared to CTRL; p = 0.0159; Figure 1C and Table 1). This was also evident in the T1DM mice, where a significant reduction in E velocity and in the ratio of early to late diastolic mitral annulus tissue velocity was observed (E /A ; T1DM: 33% decrease as compared to CTRL; p = 0.014) ( Figure 1B and Table 1). In T1DM vs. T2DM, this reduction was accompanied by a similar increase of the E/E ratio ( Figure 1C and Table 1), which represents a more reliable and reproducible index of diastolic dysfunction in animal models of cardiomyopathy. Changes in the mitral-valve early-to late-filling-velocity E/A ratio did not reach significance ( Figure 1D and Table 1). Representative traces of the mitral-flow Doppler and tissue Doppler velocities are shown in Figure 1A. These data suggest that diastolic dysfunction is a common feature of both STZ-induced T1DM and T2DM cardiomyopathy and is more severe in STZ-induced T1DM mouse models.
Systolic function was assessed through M-mode parasternal long-axis echocardiography. Despite T2DM mice developing diastolic dysfunction, this echocardiography showed that there was no significant difference in the interventricular septum or posterior wall thickness, the LV end-diastolic diameter (LVEDD) or the LV end-systolic diameter (LVESD) in the STZ-treated mice when compared to the CTRL mice (p > 0.05) ( Figure 1E-G and Table 1). Furthermore, in the T2DM mice, systolic function was preserved, as demonstrated from the normal values of ejection fraction and fractional shortening at eight weeks ( Figure 1H,I and Table 1). Overall, these data suggest that an STZ-induced T2DM mouse model leads to development of a model of HF with preserved ejection fraction.
On the contrary, the STZ-induced T1DM mouse model resulted in an overall worse cardiac remodeling, significantly increasing both the LVEDD and LVESD at 8 weeks, when compared to the CTRL mice (8% and 21% increases, respectively; Figure 1F,G and Table 1). The latter turned into a significant diabetes-induced reduction in ejection fraction (22% decrease) and fractional shortening (26% decrease) in the T1DM mice at eight weeks (Figure 1H,I and Table 1). Therefore, these data suggest that an STZ-induced T1DM mouse model leads to the development of an animal model of HF with reduced ejection fraction.

STZ-Based T1DM and T2DM Mouse Models Affected Myocardial Performance Differently
Although rodent systolic cardiac function has been classically estimated through measuring of ejection fraction and fractional shortening from 2D echocardiography, the recent introduction of speckle-tracking analysis allows for a more sensitive evaluation of cardiac dysfunction [23]. Consistently with the ejection-fraction and fractional-shortening data, in the T1DM mice, global longitudinal strain was depressed 8 weeks after injection of STZ when compared to the CTRL mice (35% reduction; p < 0.0001; Figure 1J). On the other hand, in line with the standard echocardiographic evaluation that showed no significant change in EF or FS, speckle-tracking-based strain analysis on the long-and short-axis B-mode demonstrated that STZ-induced T2DM cardiomyopathy did not reduce myocardial contractility, as indeed, global longitudinal values were unaltered at 8 weeks ( Figure 1J). Overall, these data suggest that STZ-induced T1DM but not T2DM cardiomyopathy significantly and consistently reduces myocardial performance in mice.

STZ-Based T1DM and T2DM Mouse Models Affected Left Ventricular Remodeling Differently
STZ-induced DM in animals represents a clinically relevant model to study the pathogeneses of diabetic-derived cardiomyopathy and associated complications [15]. Although the STZ-induced T1DM and T2DM mouse models shared common characteristics, considering that the two models affected cardiac function differently over an 8-week follow-up, we hypothesized that they also affected pathologic cardiac remodeling, defined based on cardiomyocyte hypertrophy, apoptosis and interstitial fibrosis, differently [24]. When compared to CTRL mice, heart sections from the T1DM and T2DM mice showed increased ventricular cardiomyocyte (CM) size (Figure 2A,B). Interestingly, the T1DM mice showed a higher CM area (hypertrophy) when compared to the T2DM counterpart (Figure 2A,B). The increase in CM size in the T1DM versus T2DM mice was further investigated through RT-PCR on freshly isolated CMs from n = 3 additional mice per group (CTRL, T1DM and T2DM mice). Expression levels of stress/hypertrophy-associated genes, such as Myh7, Acta1, Mybpc2, Gja1, Capn3, Nppa and Myl7, in adult CMs from the T1DM and T2DM mice revealed a significant increase when compared to those of adult CMs from the CTRL mice ( Figure 2C). Moreover, between the two diabetic groups, we found higher expression levels of Myh7, Mybpc2 and Gja1 in CMs from the T1DM mice when compared to those from the T2DM mice ( Figure 2C). On the other hand, Myl7 was upregulated in the T2DM versus T1DM mice ( Figure 2C).
Pathological CM hypertrophy was associated with enhanced levels of cell death and myocardial interstitial fibrosis as consequences ( Figure 3). An increase in CM death is commonly detected in the early stages of STZ-induced DM in mice [25,26]. Apoptotic DNA fragmentation, evaluated through a TUNEL-based assay, showed an increased percentage of TUNEL-positive CM nuclei in the T1DM versus T2DM mice: 1.5 ± 0.6% vs. 0.6 ± 0.5%, respectively, compared with 0.01 ± 0.01% positive CM nuclei in the CTRL mice ( Figure 3A). This finding was further assessed through RT-PCR on freshly isolated CMs from the T1DM, T2DM and CTRL mice, where the expression levels of the apoptotic gene markers Bax, Casp3, Bcl2, Foxo3 and Foxo1 were higher in the T1DM versus T2DM mice and in the T1DM and T2DM mice versus the CTRL mice ( Figure 3B).  Pathological CM hypertrophy and cell death were accompanied by enhanced levels of myocardial interstitial fibrosis, assessed with Picrosirius red staining ( Figure 3C). The comparison between the T1DM and T2DM mice revealed an increase in myocardial fibrosis in the first group of animals, as confirmed with RT-PCR on freshly isolated CMs, in which the expressions of Col1a1, Col1a2 and Col3a1 were found to be higher in the T1DM mice compared to in the T2DM mice ( Figure 3D).
To track myocardial cell regeneration, four weeks after STZ injection, the T1DM, T2DM and CTRL mice were implanted subcutaneously (between the two scapulae) with miniosmotic pumps to systemically release BrdU (Bromodeoxyuridine/5-bromo-2 -deoxyuridine, 50 mg/Kg/Day both) for 28 days. BrdU is an analogue of the nucleoside thymidine, whose cell incorporation in vivo is widely used to identify proliferating cells and, when administered continuously, as in this study, the detection of which provides the number of cumulative newly formed cells [27,28]. Cardiac sections from the T1DM and T2DM mice displayed a significantly lower percentage of BrdU-positive CMs when compared to the CTRL mice: 0.009 ± 0.004% vs. 0.05 ± 0.003% vs. 0.12 ± 0.02%, respectively ( Figure 4A).
Overall, these data demonstrate that the conventional myocardial histopathological changes in the STZ-induced T1DM and T2DM mice involved high levels of myocardial cell hypertrophy, cell death and reactive fibrosis compared to those of the CTRL mice. Nevertheless, the T1DM mice had exaggerated pathological left ventricular remodeling, with more pronounced cell death and reactive hypertrophy and fibrosis when compared to the T2DM mice, as well as reduced myocardial cell regeneration. These findings suggest a more marked deleterious effect on left ventricular cell remodeling in STZ-based T1DM mouse models compared to those of STZ-based T2DM.

STZ-Based T1DM and T2DM Mouse Models Affected Oxidative Stress and Cell Senescence Differently
To assess whether different levels of oxidative stress and senescence underscore different effects of STZ-induced T1DM and T2DM mouse models on cardiac tissue remodeling and ventricular performance, we evaluated ROS production in the three groups of animals included in this study. Cardiac sections from the T1DM mice displayed a higher level of ROS, revealed with 3-NT immunostaining, compared to those of the T2DM and CTRL mice ( Figure 4B,C). Moreover, the T1DM and T2DM mice significantly accumulated more DNA damage than did the CTRL mice, as demonstrated from the percentage of γ-H2AX-positive CM nuclei in the cardiac sections ( Figure 4D). Interestingly, the percentage of γ-H2AXpositive CM nuclei was found to be~twofold higher in the T1DM versus T2DM mice ( Figure 4D).
Accordingly, heart sections from the T1DM and T2DM mice had an increased rate of CM-positive nuclei for classical biomarkers involved in cell-cycle inhibition and cell senescence, such as p16, p21 and p53, compared to the CTRL mice ( Figure 4E). These results were further confirmed with RT-PCR on freshly isolated CMs from the T1DM, T2DM and CTRL mice, in which the expression levels of p16, p21, p15, p19 and p53 were evaluated and shown to be consistent with immunohistochemistry data ( Figure 5A). Overall, these data demonstrated that T1DM mice develop higher levels of cellular senescence markers compared to T2DM mice ( Figures 4D,E and 5A).

STZ-Based T1DM and T2DM Mouse Models Affected Oxidative Stress and Cell Senescence Differently
To assess whether different levels of oxidative stress and senescence underscore different effects of STZ-induced T1DM and T2DM mouse models on cardiac tissue remodeling and ventricular performance, we evaluated ROS production in the three groups of animals included in this study. Cardiac sections from the T1DM mice displayed a higher level of ROS, revealed with 3-NT immunostaining, compared to those of the T2DM and CTRL mice ( Figure 4B,C). Moreover, the T1DM and T2DM mice significantly accumulated more DNA damage than did the CTRL mice, as demonstrated from the percentage of γ-H2AX-positive CM nuclei in the cardiac sections ( Figure 4D). Interestingly, the percentage of γ-H2AX-positive CM nuclei was found to be ~twofold higher in the T1DM versus T2DM mice ( Figure 4D).  To further investigate the involvement of cellular senescence, inducing chronic inflammation, in diabetes, we evaluated the senescence-associated secretory phenotype (SASP) [22,29,30] in freshly isolated CMs obtained from T1DM, T2DM and CTRL mice. The SASP has been postulated as a pathophysiological link between diabetes and senescence in cardiovascular diseases [22,23,31]. Thus, we evaluated the expression levels of Tgfβ2, IL-6, Ccl11, IL-1a and IL-1b, detecting higher levels of expression of these markers in T1DM compared to in T2DM. The only exception was PAI-1, as our results demonstrated a higher level of PAI-1 expression in the STZ-based T2DM mouse model when compared to the T1DM mice ( Figure 5B).
Overall, these data suggest that DM determines a high level of oxidative stress in cardiac tissue, in both STZ-induced T1DM and T2DM mice, that is associated with DNA damage and cellular senescence. These phenomena are exaggerated in T1DM.

The Global Transcriptome Profile Showed Different Gene-Expression Signatures in the STZ-Based T1DM and T2DM Mouse Models
To assess whether STZ-induced T1DM and T2DM cardiomyopathy are characterized based on different patterns of gene expression, RNA extracted from cardiac sections from the T1DM and T2DM mice (n = 3 for each group) were processed for whole-heart transcriptome analysis through RNA sequencing. Cardiac sections from age-and sex-matched mice were used as the control (CTRL, n = 3). Once the libraries were obtained, the adaptertrimmed, high-quality reads aligned to the murine genome (mm10) data were processed to identify up-and downregulated gene sets, grouped with Gene Ontology resource tools, in different samples.
Principal component analysis (PCA) of global mRNA expression revealed the main axes of variance in the three cardiac samples and located the mRNA profiles of T1DM and T2DM at opposite poles ( Figure 6A), showing at the same time a homogeneous clustering organization between the replicates of the two groups. A volcano plot enabled quick visual identification of genes, with large fold changes that were statistically significant. When a fold change of |FC| ≥ 1.5 was considered to be significant, the comparison of gene expression in the three samples revealed that 894 genes were upregulated in the T1DM versus CTRL mice ( Figure 6B) and 857 were found to be downregulated in the same comparison ( Figure 6B).
On the other hand, 303 upregulated and 426 downregulated genes were found in the T2DM samples when compared to the CTRL samples ( Figure 6B). In comparison of the T1DM versus T2DM samples, the gene-expression analysis revealed 1920 upregulated and 2013 downregulated genes, indicating transcriptomic differences between the two DM animal models ( Figure 6B). Distribution of the numbers of common, downregulated and upregulated genes in the T1DM versus CRTL and T2DM versus CRTL samples is shown in a Venn diagram ( Figure 6C). We found 50 common upregulated genes and 26 common downregulated genes in the comparisons of the T1DM versus CRTL and T2DM versus CRTL samples ( Figure 6C). The common gene expression between the two diabetic groups displayed similar fold changes when compared to the same gene expression in the CTRL samples ( Figure 6C,D).
When the mRNA level expression was considered, the deregulated genes found in the T1DM versus CTRL mice were related to biological processes that were involved with mitochondrial dysfunction; calcium signaling; senescence pathways; and cardiac remodeling, hypertrophy, fibrosis, inflammation, oxidative stress and hypoxia ( Figure 6E).
In the comparison between the T2DM and CTRL samples, the deregulated genes were involved in inflammation (CXCR4 signaling), cardiac contraction and relaxation (protein kinase A signaling, nitric oxide signaling and renin-angiotensin signaling), cardiac hypertrophy signaling and glucose metabolism signaling (glycolysis, gluconeogenesis and oxidative phosphorylation) ( Figure 6E).
We also compared the deregulated genes in T1DM versus T2DM and found that they were related to biological processes involved with mitochondrial dysfunction, oxidative phosphorylation, inflammation, cardiac hypertrophy, cardiac fibrosis and senescence pathways ( Figure 6E and Supplementary Figure S1). We also compared the deregulated genes in T1DM versus T2DM and found that they were related to biological processes involved with mitochondrial dysfunction, oxidative phosphorylation, inflammation, cardiac hypertrophy, cardiac fibrosis and senescence pathways ( Figure 6E and Supplementary Figure S1).
RNA-seq analysis revealed modulation in the genes involved in calcium handling in both T1DM and T2DM mice when compared to CTRL mice ( Figure 6E). Remarkably, a number of genes involved with Ca 2+ metabolism presented a more pronounced dysregulation in T1DM compared to T2DM ( Figure 7A,B). Accordingly, the expression levels of the ryanodine receptor (Ryr2), SERCA2a (Atp2a), phospholamban (Pnl) and the Na (+) /Ca (2+) exchanger (NCX) were significantly lower in T1DM mice compared to the type 2 counterpart, as assessed with RT-PCR in freshly isolated CMs from the T1DM, T2DM and CTRL mice ( Figure 7C). These data demonstrate that the aforementioned genes were downregulated in the T1DM versus T2DM mice, confirming different modulations of calcium signaling in the two diabetic mouse models. RNA-seq analysis revealed modulation in the genes involved in calcium handling in both T1DM and T2DM mice when compared to CTRL mice ( Figure 6E). Remarkably, a number of genes involved with Ca 2+ metabolism presented a more pronounced dysregulation in T1DM compared to T2DM ( Figure 7A,B). Accordingly, the expression levels of the ryanodine receptor (Ryr2), SERCA2a (Atp2a), phospholamban (Pnl) and the Na (+) /Ca (2+) exchanger (NCX) were significantly lower in T1DM mice compared to the type 2 counterpart, as assessed with RT-PCR in freshly isolated CMs from the T1DM, T2DM and CTRL mice ( Figure 7C). These data demonstrate that the aforementioned genes were downregulated in the T1DM versus T2DM mice, confirming different modulations of calcium signaling in the two diabetic mouse models. Overall, global transcriptome data demonstrated that the STZ-based T1DM and T2DM mouse models displayed different gene-expression signatures. The deregulated genes involved in the STZ-induced type 1 and type 2 diabetic mice were involved in mitochondrial dysfunction, glucose metabolism, pathological cardiac remodeling, inflammation and oxidative stress. Remarkably, the genes involved in the aforementioned pathways were found to be more severely deregulated in the T1DM mice compared to the T2DM mice.

Discussion
The main findings emanating from this study are that: (i) Streptozotocin (STZ)-induced T1DM and T2DM affected left ventricular function and myocardial performance differently in mice, resulting in cardiomyopathy with heart failure with reduced ejection fraction for T1DM and with heart failure with preserved ejection fraction for T2DM; (ii) Cardiomyocytes from the STZ-induced T1DM mice displayed exaggerated apoptotic death and reactive fibrosis and hypertrophy, along with increased cardiac oxidative stress and DNA damage, compared to those of the T2DM mice; (iii) The STZ-induced T1DM mice also displayed a higher level of senescent cardiac cells and SASPs while showing severely reduced cardiomyocyte regeneration; (iv) The STZ-based T1DM and T2DM mouse models differently affected the whole cardiac transcriptome, whereby several molecular pathways were commonly modulated, but at different levels, and other biological processes were distinctively activated.
DM is a chronic disease created via insufficient insulin production/secretion from the pancreas or via insulin resistance [32]. It is marked with uncontrolled hyperglycemia and accompanying metabolic derangements, eventually leading to severe damage to numerous tissue/organs, including the heart [32]. DM affects cardiac structure, tissue and function independently from other cardiovascular risk factors, causing, per se, a cardiomyopathy that leads to heart failure [8]. Diabetic cardiomyopathy is the focus of active basic and clinical medicine to understand the cellular and molecular basis of DM as well as to find effective therapeutic strategies [33]. Animal models of DM have been instrumental to understanding the pathogenesis and progression of this cardiomyopathy and extrapolating it to humans, but no ideal model exists, with several of them accounting for only specific features of the complexity that underlies DM cardiomyopathy [34,35]. DM in small animals may be developed through two principal mechanisms: use of specific drugs or genetic manipulation [36]. Streptozotocin has been widely used to create models of T1DM and T2DM [15,16], and these models have been generally used interchangeably to study DM cardiomyopathy, despite the fact that they have numerous clinical, immunological and genetic differences [11]. Therefore, we compared the anatomy, function, histology and whole transcriptomes of mouse hearts from STZ-induced DM models.
DM is associated with heart failure with both preserved ejection fraction (HFpEF) and reduced ejection fraction (HFrEFs) [37,38]. Here, we show that STZ-based T1DM and T2DM mouse models affected global left ventricular function and myocardial performance differently. The T2DM mice had normal ejection fraction but displayed diastolic dysfunction with significant increases in the E and E/E ratios. On the other hand, the T1DM mice had both diastolic and systolic dysfunction with reduced ejection fraction.
Cardiac remodeling, the key process that underlies heart failure, is classically defined based on ongoing cardiac death and reactive myocyte hypertrophy and interstitial fibrosis [39,40]. These cellular modifications are present in both the STZ-induced T1DM and T2DM mouse models. However, the levels of these pathological cellular events were significantly higher in the T1DM mice when compared to the T2DM mice, which may explain how T1DM-related cardiomyopathy reduces ejection fraction while the STZ-induced T2DM model causes prevalent diastolic cardiomyopathy.
The pathogenic effect of hyperglycemia in DM is classically mediated with an increased production of ROS that leads to tissue damage through activation of several stress-sensitive cellular pathways [39][40][41]. Experimental evidence has highlighted a direct link between oxidative stress and DM cardiomyopathy and has persuasively pointed to increased ROS production, which generates cardiac complications in diabetic patients [40,42]. Since the heart has low levels of free radical scavenging mechanisms, excessive formation of ROS results in induction of cardiovascular complications as central mechanisms for diabetesassociated inflammation and pathologic remodeling in the heart [39,40,43]. Defects in the antioxidant defense system further increase oxidative stress during the later stages of left ventricular dysfunction in DM cardiomyopathy [44,45]. Indeed, the hyperglycemic state leads to an increase in levels of oxidative-stress-induced DNA damage, leading to altered expressions of markers such as 3-Nitrotyrosine (3-NT), 8-hydroxy-2 -deoxyguanosine (8-OHdG) and γ-H2AX [22]. These markers have also been correlated to cardiac cellular senescence events in DM [22,46]. Cardiac oxidative stress and DNA damage were both found to be significantly higher in the mice with T1DM and T2DM when compared with euglycemic controls, and even higher in T1DM compared to T2DM. The latter was accompanied by a resultant pronounced cardiac cell senescence, leading to an exaggerated senescence-associated secretory phenotype (SASP), which could be key in the increased inflammatory state, as revealed from RNA-seq data of T1DM vs. T2DM. On the other hand, our data show that plasminogen activator inhibitor-1 (PAI-1) was specifically dysregulated in the SASP of T2DM. An emerging body of evidence has implicated plasminogen activator inhibitor-1 (PAI-1) in development of T2DM [33]. Studies in PAI-1 null-allele mice have highlighted better effects on insulin and glycemic measures when mice were fed a high-fat diet, as well as protective effects against development of obesity and insulin resistance [47,48]. Moreover, PAI-1 has been demonstrated to contribute to insulin resistance that in turn stimulates PAI-1 secretion from fat cells [49].
Accordingly, the RNA-seq analysis of the data of the whole cardiac transcriptome was in line with the anatomical, functional and histological data and documented that the T1DM and T2DM animal models displayed different expressions of genes involved in several biological processes and molecular pathways of glucose metabolism, inflammation, oxidative stress, the cell-death process and cardiac contraction and hypertrophy. Furthermore, the RNA-seq bioinformatics analysis highlighted that DM severely affects the biological pathways involved with Ca 2+ handling. Abnormality of the latter, resulting from DMinduced cardiac renin-angiotensin system (RAS) activation, is involved in the pathogenesis of LV dysfunction [50]. LV relaxation and contraction are mediated with cytosolic Ca 2+ handling and the sarcoplasmic reticulum through the involvement of modulation of key genes, including Ryr2, Atp2a, Pnl and NCX [50,51], which were all significantly modulated via STZ-induced DM, in the T1DM model in particular. Furthermore, RNA-seq analysis pointed at significant differences in the mTOR and autophagy pathways in cardiomyopathy from the T1DM versus T2DM models. Numerous studies that also employed STZ-derived DM models have demonstrated that autophagy, an intracellular system for protein degradation that depends on mTOR signaling, is impaired in the DM heart, suggesting that autophagy is a potential target to reduce cardiac maladaptive alterations in patients with DM [52,53].
In conclusion, the present study offers a head-to-head comparison of the two classical models of STZ-induced DM in mice, providing a cellular, molecular and functional fingerprint of the relative cardiomyopathy. These differences should be taken in account when choosing an animal model of DM to study diabetic cardiomyopathy, and could be of key relevance when addressing the bases of and potential therapies for specific features that underlie diabetic heart disease. relative humidity and a 12 h light (6:00-18:00) and 12 h dark cycle, with water and food (containing 18.5% protein) available ad libitum. All mice received human care and all efforts were made to minimize animal suffering. Before any invasive procedure, the mice were anesthetized with i.p. injections of tiletamine/zolazepam (80 mg/kg) or inhaled isoflurane (isoflurane, 1.5%; oxygen, 98.5%; Iso-Vet Piramal Healthcare, Aurora, ON, Canada).
Three animals from each group were used for adult cardiomyocyte isolation. All of the other mice were used for immunohistochemistry and RNA-seq analysis (see below).
Eight weeks after the STZ injections, all animals were sacrificed, and the hearts were processed either for immunohistochemistry analysis and RNA isolation or for cardiomyocyte (CM) isolation [23].

Mouse Cardiomyocyte Isolation
CMs were isolated as established and reproduced in our laboratory through standard enzymatic dissociation from the hearts of each group of mice (CTRL, T1DM and T2DM mice), as previously described [28,54,55].

Echocardiography
Mice that underwent echocardiographic evaluation were prepared as previously described in detail [22,23,27]. All echo images and videos were obtained from the mice at heart rates > 400 b.p.m. Echocardiographic images and videos were obtained with a Vevo 3100 system (Visualsonics, Inc., Toronto, Canada) equipped with a MX550D ultrahigh-frequency linear-array transducer   [22,23,27]. B-mode, M-mode and speckle-tracking images were analyzed through Vevo LAB analysis software Version 3.2.0 (VisualSonics, Amsterdam, The Netherlands) as previously described [23,27]. The n-value for each experimental group is specified in the figure legends.

Tissue Harvesting, Histology and Immunohistochemistry
For immunohistochemistry analysis, the abdominal aorta was cannulated and the heart arrested in diastole using a cadmium chloride/potassium solution. Mouse tissue specimens were fixed and embedded in an optimal cutting temperature (OCT) compound for immunohistochemical analysis.
Tissues were cut into 5 µm cross-sections, respectively, and processed for both fluorescent as well as chromogenic immunohistochemistry according to specific analysis.
For bioquantification of fibrosis, OCT sections were stained with Picrosirius red. Staining was performed as per the manufacturer's instructions (Bioptica, Milan, Italy). Hematoxylin and eosin (H&E, Bioptica) were used to evaluate the cellular architectures of the samples. (1:100 dilution; Santa Cruz) antibodies. Positive reactions were visualized using a labeled polymer-HRP complex and a 3,3 -diaminobenzidine tetrahydrochloride (DAB) chromogen (EnVision + Dual Link System-HRP, DAKO, Santa Clara, CA, USA). Sections were then counterstained with hematoxylin and examined with light microscopy (LEICA, Wetzlar, Germany, DMI3000B). The numbers of p16-, p21-, p53-and γ-H2AX-positive CMs were expressed as percent fractions of the total CM nuclei.
An In Situ Cell Death Detection Kit (TdT, Sigma-Aldrich, ST. Luis, MO, USA) was used as per the manufacturer's instructions for detection of apoptosis-positive CM nuclei. BrdU and TdT fluorescence quantifications were obtained through manual counting of respective histological samples, and the numbers of BrdUpos and TdTpos CMs were, respectively, expressed as percent fractions of the total CM nuclei [23,27].
CM cross-sectional area was measured through immunofluorescence staining for wheat germ agglutinin (WGA) of the Alexa Fluor 647 conjugate (1:200 dilution; Invitrogen, Waltham, MA, USA) and digital analysis of acquired cardiac cross-section images. CM diameter was measured across the nucleus on three transverse sections (~500 myocytes/animal were sampled), as previously described [27]. All immunofluorescence staining was acquired and analyzed using confocal microscopy (LEICA, Wetzlar, Germany, TCS SP5 and SP8).

Quantitative RT-PCR (qPCR)
RNA was extracted from CMs using the TRIzol Reagent (Ambion, Waltham, Ma, USA) and quantified using a Nanodrop 2000 Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Reverse transcription was performed with 0.5-1 µg of RNA, using the HighCapacity cDNA Kit (Applied Biosystems, Waltham, MA, USA). Quantitative qPCR was performed using TaqMan Primer or Probe sets (Applied Biosystems or Eurofins Genomics, Ebersberg, Germany) (see Table 2) using the StepOne Plus Real-Time PCR System (Applied Biosystems, Waltham, MA, USA). All reactions were carried out in triplicate. Table 2. List of primers.

Library Preparation
Libraries were generated using depleted RNA obtained from 1 µg of total RNA with a TruSeq Sample Preparation RNA Kit (Illumina, Inc., San Diego, CA, USA), according to the manufacturer's protocol without further modifications, as previously described [56,57].

Sequencing
All libraries were sequenced on the Illumina HiSeq 1000, generating 100 bp paired-end reads. The libraries were divided into two groups depending on how they were prepared.

RNA-Seq Data Analysis
All FastQ files were quality checked using FastQC software (v0.11.9) [58]; then, adapter sequences were removed and low-quality reads were filtered out using Cutadapt software (version 1.18) [56] with parameters set as follows: quality cutoff, 20; minimum length, 20. The resulting high-quality reads were then mapped to the mouse reference genome (GRCm39). This alignment was performed using default parameters with STAR software (version 2.7.10b) [57]. Then, the number of reads that mapped to each transcript within the reference was computed with FeatureCounts (v2.0.1) [59]. The counts were then imported in in R package DESeq2 (v1.38.1) (R version 3.6.3) [60], and differential gene-expression analysis was performed through comparison of each experiment condition with the controls. Differential expression was reported as |fold change| ≥ 1.5 along with associated adjusted p values (FDR ≤ 0.05), computed according to Benjamini-Hochberg [61] as described in Salvati et al., 2019 [62]. For Gene Ontology (GO) analysis of DE genes, Ingenuity Pathway Analysis Software (IPA 84978992, Ingenuity ® Systems, www.ingenuity.com, accessed on 20 December 2022) was used, and only functions and pathways that showed −log(B-H p-value) ≥ 1.3 were considered.

Statistical Analysis
Data are reported as mean ± SD. Significance between any 2 groups was determined with Student's t-test and in multiple comparisons with analysis of variance (ANOVA), using GraphPad Prism version 9.4.0 for Windows (GraphPad Software Version 9.4.0, San Diego, CA, USA). In the event that ANOVA justified posthoc comparisons between group means, these were corrected using the Tukey multiple comparison test. Differences of p < 0.05 were considered statistically significant.