Proteomic and structural heart changes in obese rats are incompletely restored after weight loss


 Background As a systemic disorder, obesity strongly affects the cardiovascular system, inducing cardiac overgrowth, which increases the risk of heart failure and death. Moreover, obesity is potentially curable, leading to the restoration of the heart phenotype, but it is not clear if all of the after-effects are reversed after weight loss. of the heart phenotype, but it is not clear if all of the after-effects are reversed after weight loss. Here we describe the proteomic and morphologic phenotype of the heart in a rat model of developmental obesity with an evaluation of whether the observed effects are persistent in spite of weight loss. Methods Developmental obesity with hyperlipidemia and insulin resistance was induced in young rats by exposure to a Western diet composed of human snacks. An histologic evaluation of the heart was performed to measure the size of the cardiomyocytes and amount of connective tissue discriminating the phenotype of obesity cardiomyopathy. The cardiac tissue and plasma were analyzed by global proteomic profiling. Based on these data, we targeted proteins for evaluation with the western blot. The histological and proteomic measurements were performed after weight loss to validate which features of obesity cardiomyopathy were persistent. Results Obesity cardiomyopathy was determined as cardiac hypertrophy associated with fibrosis, oversized myocytes, and mTOR upregulation. A plethora of molecular changes were observed, suggesting an effect on the utilization of metabolic substrates in the hearts of obese animals. This was confirmed by increased levels of ACSL-1, a key enzyme for fatty acid degradation and decreased GLUT-1, a glucose transporter. Immunological processes and lipid metabolism were also affected in the cardiac tissue and plasma. Weight loss led to the normalization of the heart’s size, but some after-effects of obesity such as connective tissue abundance and abnormal proteomic composition were still persistent. Conclusion In addition to morphological consequences, obesity cardiomyopathy involves many proteomic changes. Obesity treatment and weight loss provides for a partial repair of the heart’s architecture, but cardiac fibrosis and some proteomic alterations persist.

According to International Statistical Classification of Diseases and Related Health Problems (10th revision), obesity is defined as a disorder characterized by an abnormally high, unhealthy amount of body fat and is listed as E66 subgroup of "endocrine, nutritional and metabolic diseases." This pandemic disease has many severe consequences including an increased risk of death, morbidity, and accelerated aging [1] due to the many systemic changes associated with obesity. The disease is widespread globally across all age groups [2,3]. In the United States, 1 in 5 children are obese and the scale of the problem is well illustrated by reported cases of obesity among children. Obesity symptoms are potentially curable by losing weight due to reducing calorie intake and increasing energy expenditure. Other methods that are effective for weight loss include medication and surgery [4]. Reversing obesity is essential for lifespan and healthspan because it affects all the systems of the human body, especially contributing to the development of cardiovascular diseases [5]. Obesity and the Western diet are leading causes of hypertension [6]. A pressure overload leads to pathological cardiac hypertrophy and overgrown myocytes enmeshed within an abundant network of the extracellular matrix (ECM) is the hallmark of obesity cardiomyopathy and left ventricle (LV) remodeling [7]. Although pressure overload is the predominant inductor of LV hypertrophy [8], the hypertrophied heart develops independently of hypertension during obesity [9], suggesting that other mechanisms may also contribute to the cardiac overgrowth. The hypertrophy of cardiomyocytes with cardiac fibrosis and remodeling associated with the Western diet and weight gain may be a consequence of the intensification of intracellular anabolic processes and sustaination of a proinflammatory status [10,11]. Normally cardiac fibroblasts make up about 70% of cardiac cells and support cardiomyocytes by producing ECM and by regulating the proliferation and migration of other cardiac cells [12]. Thus, fibroblasts play an important role in cardiac repair. However, in certain circumstances the excessive proliferation and differentiation of fibroblasts lead to the increased deposition of fibrotic content (fibrosis) and consequently to heart failure [13].
Weight reduction and weight loss maintenance is capable of reversing many of the alterations in cardiac performance and morphology associated with obesity [14] and is essential in avoiding heart failure related to obesity cardiomyopathy [15][16][17]. However, more detailed investigations are needed in order to identify the potential changes that are persistent after weight loss and to find treatment options.
This study uses animal models to demonstrate how developmental obesity affects the heart's proteome and morphology and identifies changes that are not reversed following weight normalization. 28-day-old Long Evans male rats were used in two independent experiments. In Exp.1, animals were fed ad libitum with standard chow (control to obese cohort, ContO) or the same chow supplemented with Western diet foods (Obese). These included cheese, dry sausage made of pork (kabanos), crackers, salty potato chips, candy bars (bounty, mars, prince-polo) and a 10% saccharose solution to drink. Additionally, all animals had unlimited access to drinking water. After 10 weeks the animals were euthanized by decapitation to collect blood and heart samples.

Materials And Methods Dietary induced obesity
In Exp. 2, animals from the AWL group (After Weight Loss) were treated in the same manner as in Exp. 1 to develop obesity, but this was extended for 2 more weeks (12 weeks of weight gain). After this time the rats were subjected only to standard rodent chow with an 80% calorie restriction (CR; the 100% of caloric needs was established empirically), which was reduced to 70% after two weeks until the end of the weight-loss-phase. During this period the control group from Exp. 2 (ContA) received the same isocaloric food to stop subtle weight gain in these control adult rats. The amount of food was established empirically for every cage and 20 g of standard chow/rat/day determined the criteria. After the period of weight reduction, the rats from AVL and ContA groups received an isocaloric amount of food for two weeks (stabilization phase) to equalize the calorie intake among control and experimental groups. Drinking water was provided ad libitum during all phases of Exp. 2.
Six hours prior to insulin tolerance and glucose loading tests the rats were fasted (tested animals from Exp.1 were fed by standard chow and water for 24 h before fasting), than insulin (1 U/kg in Exp.1; 0.75 U/kg in Exp.2; ActiRapid, Novo Nordisc) or glucose (2 g/kg) were injected in i.p. injections.
Blood was collected from the tip of tail at the appropriate time intervals to measure glucose levels (CardioCheck Professional).
For the histological examination of the morphology of cardiomyocytes, 6 hearts per group collected during Exp. 1 and Exp. 2 were gently squeezed to remove excess blood, then weighed and immersed in a 10% formalin solution in PBS (pH = 7.2). For proteomic examination, fresh pieces of LV were collected from another animals (n = 5 per group in Exp. 1).
After decapitation ~ 2 mL of trunk blood was collected for serum and plasma samples. For the serum collection, the blood was allowed to clot by leaving it for 30 min on ice. The samples were then centrifuged (2,000 × g for 15 minutes, 4C), and the serum was pipetted and stored at − 80 °C. Serum cholesterol and triglyceride levels were measured in the sera using a Mindray BS-200 Chemistry Analyzer (Shenzhen Mindray Bio-Medical Electronics Co.).
For the plasma (used for proteomic profiling), 1 ml of blood was collected into 1.5 ml EDTA coated Eppendorf tubes containing 10 ul of 0.5 M EDTA and immediately centrifuged (1300 × g for 10 minutes, 4C). 99 ul of plasma was pipetted to fresh tubes containing 1 ul of a protease inhibitor cocktail (#P8340, Sigma Aldrich) [18].
Histological examination of the heart The LV was dissected from formalin-fixed hearts. The tissues were dehydrated with graded concentrations of alcohol and embedded in paraffin. 5 µm paraffin slices from each tissue sample were stained with (1) Hematoxylin and Eosin (H&E) or (2) Masson's trichrome stain.
To calculate the cross-sectional area of the cardiomyocyte, the sections stained with H&E were photographed (with 40x objective) using a Nikon light microscope (Nikon ECLIPSE E600) with an Olympus Camera (Olympus DP 26). The individual cell surface area was measured by a blinded observer using cellSens Entry Imaging Software (Olympus). One hundred cell surface areas were counted per each group, and the average value was used for analysis.
Masson's stain was used to investigate LV morphology and perivascular/interstitial fibrotic changes.
The sections were photographed (with 40x objective) 6 times each for the following: LV area without visible vessels (interstitial zone) and area with visible vessels (perivascular zone) [19]. The photos were subjected to deconvolution in the NIH Fiji program [20], than the areas of blue (connective tissue) or red (myocytes) channels were automatically counted. The data is expressed as the amount of connective tissue relative to the total area (connective tissue + myocytes) and expressed as a %.

Proteomic profiling Protein extraction
The isolated brain tissues were lysed in a buffer with 1 M triethylammonium bicarbonate (TEAB) and 0.1% sodium dodecyl sulfate (SDS) and automatically homogenized using a Precellys 24 homogenizer (Bertin Technologies) in 0.5-mL tubes pre-filled with ceramic (zirconium oxide) beads (Bertin Technologies). Next, the material was subjected to a threefold cycle of freezing and thawing. Then, the tissue in the buffer was sonicated in a bath for three 1-minute cycles on ice and homogenized again using the Precellys 24 instrument. The protein concentration was measured using Pierce BCA protein assay kit (Thermo Fisher Scientific) in the isolated protein fraction according to the manufacturer's instructions.

In-solution digestion
Ten-microgram aliquots of the proteins were diluted with 15 µl of 50 mM NH4HCO3 and reduced with Data analysis of the proteins Protein identification was performed using the Swiss-Prot rat database with a precision tolerance set to 10 ppm for peptide masses and 0.08 Da for fragment ion masses. All raw data obtained for each dataset was imported into MaxQuant 1.5.3.30 version for protein identification and quantification.
Protein was considered as positively identified if at least two peptides per protein were found by the Andromeda search engine, and a peptide score reached the significance threshold FDR = 0.01.
The obtained data was exported to Perseus ver. 1.5.3.2 software (part of the MaxQuant package). The numeric data was transformed to the logarithmic scale and each sample was annotated with its group affiliation. Next, the data was filtered based on valid values -proteins, which had valid values in 70% of samples in at least one group were kept in the table. A one-way analysis of variance (ANOVA) analysis was performed on the analyzed sample data with permutation-based FDR 0.05 used for truncation and the resulting list of differentiating proteins was normalized using a Z-score algorithm for the hierarchical clustering of data.
The immunoblots were visualized by means of Clarity Western ECL Blotting Substrates (Bio-Rad) and detected with the ChemiDoc™ Touch Imaging System (Bio-Rad). The targeted proteins were quantified with ImageLab Software 6.0.1 (Bio-Rad). The results were normalized to the total protein content in the gel (Stain-free technology, Bio-Rad).

Statistical analysis
Statistical analysis was performed using GraphPad Prism 8.01 software (GraphPad Software Inc.).
Depending on the data distribution (as evaluated by the Shapiro-Wilk normality test), either the Student's T-test or U-Mann-Whitney test were used for the estimation of significant differences between the two subject groups. In the case of repeating measurements, two-way ANOVA with Sidak's multiple comparison test was applied. In all tests, significance was considered when p < 0.05.

Results
The obesity phenotype is associated with heart hypertrophy In order to evaluate the influence of developmental obesity on the heart proteomic profile, the rats were fed with a Western diet starting from the 28th day of life for 10 weeks. Starting with the 3rd week of feeding, these rats gained weight more excessively than control individuals (Fig. 1A). The obesity phenotype resulted in decreased response to insulin measured by ITT (Fig. 1B). An analysis of sera revealed higher triglycerides (p = 0.0007) and LDL (p = 0.03) concentrations in obese rats (Fig. 1C). These rats also had increased total-cholesterol-to-HDL ratio (p = 0.0006, Fig. 1C), which is recognized as a strong cardiovascular risk marker [23]. Ten weeks after introduction of dietaryinduced obesity protocol, the authors collected the hearts from animals to perform morphometric and proteomic analyses. The hearts of obese rats were about 35% heavier than in control individuals (p < 0.0001; Fig. 2A). The histological examination of H&E-stained LV sections revealed that the cardiomyocytes in obese rats had increased an area as compared to control individuals (p = 0.0015) (Fig. 2B). The microscopic evaluation of Masson-Trichrome stained slides revealed a higher amount of connective tissue immersed between the myocytes (Fig. 2C) or localized around the vessels (Fig. 2D).
The enhanced phosphorylation of mTOR on Ser2448 was observed in the LV of obese rats (p = 0.016) ( Fig. 2E), suggesting the up-regulation of this anabolic pathway. On the other hand, the phosphorylation of mTOR on Thr2446 did not differ among groups (Fig. 2F). This mTOR region is phosphorylated during AMPK activation resulting in mTOR pathway inhibition [24]. In line with this, the authors did not observe significant changes in AMPK signaling (Fig. 2G).
The proteomic status of the cardiac cells exhibited a stronger shift towards the utilization of fatty acids in obese animals acyl-CoA synthetase long-chain family member 1 (ACSL-1; 4 proteins interacted significantly, p = 0.0035). When the protein pool was analyzed separately depending on whether the proteins were overexpressed or downregulated in the LV of obese rats, the subset of 40 elevated cardiac proteins confirmed the significant interaction with ACSL-1, whereas a subgroup of the remaining proteins (decreased expression in obese animals) interacted with GLUT-4. Thus the authors decided to perform the western blot to assess the expression of these two molecules. The level of ACSL-1 was significantly increased in the LV of obese rats (Fig. 4D, p = 0.045). In spite of the fact that cardiac GLUT-4 is downregulated [25] during insulin resistance, the authors did not clearly demonstrate its decreased content in the heart of obese rats (Fig. 4E). Rather than by transcriptional mechanisms, the regulation of GLUT-4 depends on its recruitment to the sarcolemma in response to glucose delivery and prolonged hyperinsulinemia leads to the internalization and inactivation of this glucose carrier. In addition to GLUT-4, the most abundant glucose transporter in the heart is solute carrier family 2, facilitated glucose transporter member 1 (GLUT-1) [26], thus it was decided to measure its expression. The authors have observed the decreased content of GLUT-1 (Fig. 4F, p = 0.025), which is an insulin-independent glucose transporter mainly responsible for the basal needs of the muscle cells [27]. As it was shown that Acsl1 knockout mice upregulate GLUT-1 expression in skeletal [28] and cardiac [29] muscles depending on glucose metabolism, it is possible that that ACSL-1 and GLUT-1 expression depend on each other.
The biological processes involved in the regulation of the immune system and triglycerides homeostasis are changed in the plasma of obese rats Next, the authors performed global proteomic measurement in the plasma of the rats, identifying 41 molecules which differentiate between groups (Fig. 5, heatmap). Based on the results of DAVID analysis, eight biological processes (GO: BP) were changed in the plasma of obese rats, including i.e. the process that decreases the frequency, rate, or extent of endopeptidase activity (negative regulation of endopeptidase activity, p < 0.001), immunological responses (complement activation, p < 0.001; acute-phase response, p < 0.001; inflammatory response, p < 0.05), and processes that change the state or activity of a cell as a consequence of triglyceride stimulus (response to triglyceride, p < 0.05). Figure 5 presents the protein involved in these biological processes except for the negative regulation of endopeptidase activity (Supplemental Data). Due to the fact that various molecules participate in different immunologic processes, the proteins of acute and inflammatory responses have been combined (Fig. 6). Changes of four exosomal molecules detected in the LV were also significant in the plasma of obese rats (Supplemental Data). Alpha-1-inhibitor III, apolipoprotein A4 (APOA4), and apolipoprotein C3 (APOC3) were increased in both tissues. Transthyretin (TTR; thyroid hormone-binding protein) was elevated in the heart but decreased in the plasma of obese rats.
CR-induced weight loss partially restored the heart's architecture, leaving fibrotic debris Another cohort of young animals was subjected to a Western diet for 12 weeks to gain weight and then to a 6-week-long CR protocol to reduce weight (AWL group; Fig. 7A). The group was accompanied by control animals (ContA group) which were maintained exclusively on a standard diet.
This was followed by two subsequent weeks of isocaloric feeding to normalize the systemic parameters in both AWL and ContA groups. Obesity-induced insulin insensitivity and applied CR could reverse it, even resulting in higher insulin sensitivity [30], but there were no significant differences in the response to insulin (Fig. 7B) or glucose (Fig. 7C) in AWL rats. Hyperlipidemia was also reversed after weight loss (refer to the Supplemental Data). The authors observed that the weight of the collected hearts did not differ among AWL and control animals (Fig. 6D) and the histologic examination did not reveal any differences in the size of the cardiomyocytes among groups (Fig. 6E).
However, the amount of interstitial (Fig. 7F) and perivascular (Fig. 7G) connective tissue was elevated in the LV of AWL rats, suggesting persistent cardiac fibrosis in individuals which were obese in the past.
The identification of cardiac and systemic proteomic hallmarks of obesity persist after weight loss Proteomic data revealed that the concentration of 13 identified proteins and 1 unknown molecule were still altered in the hearts of the rats after weight loss (Table 1). Two proteins -adenylate kinase 1 and complement C3 -were changed in the hearts of both obese and AWL rats. Ak1 was decreased in both obese and AWL groups. However, complement C3 was elevated in the hearts of obese rats but decreased in AWL individuals (Supplemental Data). The analysis of the plasma revealed that 15 identified proteins (and one unknown) varied in the plasma of AWL and control rats (Table 2), but none of these molecules were noted to be changed in the plasma of obese rats. These proteins are mostly involved in the regulation of endopeptidase activity (8 proteins, p < 0.001), complement activation (4 proteins, p < 0.001), and chemical reactions and pathways involving hyaluronanhyaluronian (3 proteins, p < 0.01). The extracellular protein TTR, responsible for the transport of thyroxin and retinol binding protein complex to the various parts of the body, was decreased in the plasma of AWL rats (p = 0.004). In the plasma, TTR binds to adipokine known as retinol-binding protein 4 (RBP4), which prevents glomerular filtration and the subsequent catabolism of RBP4 in the kidney. However, despite TTR depletion RBP4 levels were elevated (p = 0.048). TTR may be used as a biomarker of stress disorders [31], oxidative stress [32], lean body mass, and catabolic states [33]. RBP4 was shown as a marker of insulin resistance [34], but these results are inconclusive [35]. Several human and animal studies have investigated the influence of high circulating RBP4 levels in the pathogenesis of insulin resistance associated with type 2 diabetes and obesity [35]. The current data suggest that the systemic balance between TTR and RBP4 may be disturbed after weight loss, but this phenomenon warrants further evaluation.

Discussion
Here, we discriminate the proteomic and morphological consequences of obesity in the hearts of rats accompanied by an evaluation of whether the changes are reversed by CR-induced weight loss. The presented data shows that the altered proteomic status and excessive amount of connective tissue can be found in the LV after returning to a normal body weight. This conclusion has not been recognized thus far.

Morphological and molecular characteristics of the heart in obesity
The phenotype of obesity cardiomyopathy includes heart hypertrophy with an increased amount of interstitial and perivascular connective tissue [36]. The upregulation of the mTOR pathway is predominantly involved in this cardiac remodeling [37] as a consequence of overfeeding and insulin resistance [38]. Therefore, we observed enlarged cardiomyocytes and increased mTOR phosphorylation suggesting the upregulation of this anabolic process in the cardiac cells of obese rats. We assumed that decreased AMPK signaling, as a possible consequence of overfeeding, may be partially involved in the mTOR control, but AMPK phosphorylation and its targeted site on mTOR (Thr2446, leading to its inhibition when phosphorylated [24]) seem to be unaffected, rejecting the assumption above. Further, the observed cardiomyopathy was reflected in a plethora of molecular changes. The proteomic machinery responsible for the transport of triglycerides was boosted in obese animals both in the plasma and the heart muscle. In order to metabolize long-chain fatty acids, they must first be converted to acyl-CoA by ACSL proteins [39]. Among the five mammalian ACSL isoforms, ACSL-1 predominates in the cardiac tissue [40]. The hearts of obese rats expressed more ACSL-1 but had reduced amounts of GLUT-1. These findings suggest that in obesity cardiomyopathy, glucose uptake may be decreased in favor of lipid metabolism. Whereas energetic metabolism in the normal heart depends primarily on ß-oxidation and less on glucose utilization, in the diabetic heart ATP production becomes almost completely reliant on fatty acids oxidation [41,42], causing the inability to switch between energetic substrates in case of trauma. The inability of efficient ATP generation from glucose under specific pathological conditions contributes to heart failure [43].
The results of some previous studies, where non-targeted proteomic methods were applied for the screening of cardiac tissue in animal model of dietary induced obesity, are mostly consistent with our observations. A map of the cardiac proteome in obese rats showed that the proteins involved in regulation of metabolism were predominantly changed [44]. The proteomic machinery responsible for fatty acids uptake and oxidation was upregulated in obese animals without hypertrophied hearts [44].
Thus the role of boosted fatty acids turnover in the cardiac remodeling is doubtful. In the mice with dietary induced obesity the expression of the proteins involved in mitochondrial metabolism was affected, resulting in production of reactive oxygen species [45][46][47]. This is probably a consequence of the excessed delivery of fats which are intensively metabolized in cardiomyocytes [48]. Although substantial protein expression changes the in obesity the cardiac proteome may be also affected in a different way, by modifying pattern of acetylation [49].
Most of the LV proteins affected by obesity originated from exosomal vesicles, raising the importance of these molecules in the pathomechanisms of obesity cardiomyopathy. Exosomes are the smallest extracellular vesicles (diameter, 30-100 nm) next to microvesicles (100-1000 nm) and apoptotic bodies (500-4000 nm). Six main types of cardiac cells, i.e., fibroblasts, cardiomyocytes, endothelial cells, cardiac progenitor cells, adipocytes, and cardiac telocytes, release exosomes carrying different cargo and enforcing various responses [50]. All of these cell types use extracellular molecules to communicate with each other and any alteration either in the composition of the exosomal cargo or its machinery can affect cardiac homeostasis, leading to disease [13]. Cardiac exosomes are oversecreted as a consequence of stressing factors such as angiotensin II production, hypoxia, inflammation, or injury [51]. Recent literature has shown that exosomes may modify cardiac repair and the fate of fibrosis via the modulation of fibroblast function and likely contribute to obesity cardiomyopathy in which microRNAs seem to play a crucial role [13]. For example, it was proposed that cardiac fibrosis in diabetic cardiomyopathy may be a consequence of the increased exosomal secretion of microRNA-320, a negative regulator of Heat shock protein 20 expression. Moreover, the overexpression of heat shock protein 20 remarkably attenuated diabetes-induced cardiac dysfunction and adverse remodeling [52]. Here we have observed an obesity-related decrease in cardiac heat shock 70 kDa protein, which normally protects cells against various types of stress [53] and is downregulated in obesity [54]. Its role can be considered in targeting obesity cardiomyopathy.
The cardiac outcomes of obesity reflect systematic disturbances The proteomic composition of the plasma in obese rats strongly supports the notion that the systemic consequences of obesity and the Western diet are reflected in the cardiac muscle. Increased angiotensinogen levels led to the suspicion that obese rats developed hypertension with the involvement of the renin-angiotensin system [55], and this could contribute to the overgrowth of cardiomyocytes [56]. Apolipoproteins are master regulators of plasma lipoprotein particles (HDL, LDL and chylomicrons), the carriers of fats and cholesterol in the body. APOA4 and APOC3 apolipoproteins were increased in the plasma of obese rats. APOC3 overproduction is especially noteworthy because of driving the risk of cardiovascular disturbances [57] and can be postulated as a therapeutic target to reduce the disease's development [58]. Moreover, four apolipoproteins were elevated in the heart (apolipoprotein A1, APOA4, APOC3, and apolipoprotein E), suggesting the high vulnerability of cardiac tissue towards the lipidomic changes accompanying dietary-induced obesity.
Apoa4 is expressed almost exclusively in the intestine and released into the blood in chylomicrons transporting dietary-derived fatty acids. Apolipoprotein E is more widely expressed, but predominantly by the adrenal gland, brain, and liver. These proteins were elevated in the heart muscle but also in plasma of obese rats possibly as a consequence of its higher concentration in the blood. Apolipoprotein A1 and APOC3 levels were elevated in the heart, but they are expressed in the liver and intestine and apparently their cardiac accumulation accompanied the influx of lipids and cholesterol to the heart [59][60][61]. Decreased levels of gelsoline indicates pathological conditions as this protein normally occurs in high amounts and is extensively degraded when binding to filamentous actin (released upon cell death or rupture) occurs [62]. This protein mitigates the detrimental effects of systemic inflammation [63], thus gelsoline downregulation could potentiate a greater impact of obesity-related inflammation on the system. Protein clustering revealed that obese animals developed contradictory inflammatory status. On the one hand, chronic inflammation elevated pro-inflammatory molecules such as fibronectin [64] and haptoglobin [65] in obese animals. Inconsistently, α1-inhibitor-3 [66] and kininogens [67] known as pro-inflammatory factors were decreased. This was associated with the downregulation of anti-inflammatory α1-antitrypsin [68] and transferrin, which is negatively regulated during the acute phase [69]. On the other hand, the defense response could be diminished, which was recognized as the decreased expression of the complement system proteins being a part of the immune system launched during a pathogen attack [70].
Weight loss reverse cardiac overgrowth but not fibrotic deposition Obesity evokes cardiac remodeling, which seems to be rescued after weight loss [17,36]. However, a significant finding of our study was that after weight loss the heart is still abundantly composed of connective tissue. We suspect that this could be a residue of obesity as well as a consequence of CR applied for weight reduction. Whereas intracellular compartments in the cardiomyocytes are degraded via autophagy, allowing for tissue rearrangement, the redundant ECM is proteolyzed and then digested by the phagocytic cells of the complement immune system [71]. The proteins of the complement cascade were downregulated in the heart collected after weight loss, suggesting that inflammatory response and the recruitment of phagocytic cells is rather repressed than mobilized.
Such a cardiac immunosuppression could diminish the complete restoration of tissue during weight loss. Due to the novelty of this thesis, more research is required to support its authenticity. However, there may be an additional reason for the observed elevation of the heart's connective tissue after weight reduction. CR impacts cardiac morphology in an age-dependent manner by stimulating the fibrotic deposition in young-adult mice (age-matched to our rats), but not in older animals, emphasizing the possible cardiac residue of CR [72].
The different proteomic composition of the LV in AWL rats may be a consequence of cardiac fibrosis and ECM excess, which is illustrated by the lower amount of cardiac tubulin as this protein is abundantly expressed in the myocytes but scarce in ECM [73]. Conversely, the pool of lumican was elevated in the hearts of AWL rats. Lumican is a proteoglycan which binds collagens, the main components of cardiac ECM [74], therefore its level correlates with the deposition of fibrocytes representing tissue fibrosis. Lumican expression is increased in hepatic [75], pulmonary [76], and cardiac [74] fibrosis. On the one hand, experimental and clinical findings revealed that the overexpression of lumican in cardiac fibroblasts is evoked during heart failure [74]. This protein regulates cardiac remodeling following LV pressure overload [77] and has been shown to have a role in ECM remodeling and fibrosis in different cardiovascular diseases [78][79][80]. On the other hand, lumican is required for cardiac remodeling by ensuring the structural integrity of connective tissue and survival following pressure overload [77]. We propose that lumican may play a role in obesityrelated cardiac remodeling and we observed that its levels are elevated after weight loss, carrying the possible risk of fibrosis-related disturbances.

Limitations of the study and transferal of the results
The rat model of dietary-induced obesity was introduced to reflect developmental obesity. Thus the animals were young (28 days old) at the time the Western diet was introduced, reflecting ~ 8 months of human age [81] when infants begin receiving solid food [82]. The overconsumption of protein and sugars especially induce developmental obesity [83], and the global intake of the latter range from 1.9-13.4% in humans before 2 years of age [84] underlies the high dietary risk factors for the development of obesity in infants. Comprehensive proteomic evaluation was performed in the plasma and cardiac tissue of obese and CR-cured rats. In the second group, the tissue material was collected after two weeks of isocaloric intake, where AWL rats and control companions received the same amount of calories, allowing for weight maintenance without significant weight gain or reduction. This approach seemed to be the most reasonable and made it somewhat possible to eliminate the systemic effects of CR [85]. The animals could not be fed by standard chow in an ad libitum manner because from our experience, AWL rats regain weight more extensively than control individuals. Thus, during the "stabilization state" both groups received ~ 100% of calorie needs, which was set empirically. We did not find any justification to compare the effects observed in obese animals with AWL rats and adequate controls (by comparing the four groups in parallel), especially considering the different age of the animals. In such a comparison we would not estimate the impact of body weight only, but also the age factor, thereby falsifying the conclusion.
The prevalence of obesity actually increased with hordes of young people predisposed to weight-loss therapies. If adequate approaches will be applied to reduce body weight, these patients could be rescued from obesity within a few years [86]. The restoration of a lean phenotype masks the obesity experience, possibly shaping future diagnosis in adulthood or the elderly. However, the persistent effects of obesity are not well explored. What if by disregarding the bygone obesity the true reason for the current disability will be obscured? We suggest that clinicians should consider the possibility of the persisting cardiac consequences of obesity in lean patients during diagnosis.

Conclusion
Global proteomic profiling with morphological evaluation was performed in the hearts of obese rats and after weight loss. We conclude that obesity cardiomyopathy is highly complex, integrating anabolic, metabolic, and immunogenic complications in the cells. The systemic status of overfeeding, hyperlipidemia, and insulin resistance may contribute to the cardiac adjustment developed during weight gain. After losing weight, the heart's phenotype can be ostensibly restored to normal.
However, some abnormalities still occur at the morphological (fibrosis) and proteomic levels. transthyretin.

Declarations
Ethics approval and consent to participate: Not applicable.
Consent for publication: Not applicable.

Figure 1
Characteristics of the obesity phenotype in rats fed with a human-mimicking Western diet for 10 weeks. As a consequence of high-calorie feeding, the animals gained weight intensively (a) and developed insulin resistance (b) with hyperlipidemia and hypercholesterolemia (c). Two-way ANOVA with Sidak's multiple comparison test (a, b) and the Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, **p<0.01, ***p<0.001).

Figure 1
Characteristics of the obesity phenotype in rats fed with a human-mimicking Western diet for 10 weeks. As a consequence of high-calorie feeding, the animals gained weight intensively (a) and developed insulin resistance (b) with hyperlipidemia and hypercholesterolemia (c). Two-way ANOVA with Sidak's multiple comparison test (a, b) and the Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, **p<0.01, ***p<0.001).

Figure 2
Hallmarks of obesity cardiomyopathy in rats. The hearts of obese animals were heavier (a) with enlarged myocytes (b). The oversized hearts contained higher levels of fibrotic tissue both in the non-vessel (c) and vessel containing (d) areas. The expression of phospho-mTOR on ser2448 (e) but not on Thr2446 (f) was higher in the LV of obese rats, suggesting the upregulation of this anabolic pathway. No significant differences were observed with regard to of AMPK activity (g). The Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, **p<0.01, ***p<0.001).

Figure 2
Hallmarks of obesity cardiomyopathy in rats. The hearts of obese animals were heavier (a) with enlarged myocytes (b). The oversized hearts contained higher levels of fibrotic tissue both in the non-vessel (c) and vessel containing (d) areas. The expression of phospho-mTOR on ser2448 (e) but not on Thr2446 (f) was higher in the LV of obese rats, suggesting the upregulation of this anabolic pathway. No significant differences were observed with regard to of AMPK activity (g). The Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, **p<0.01, ***p<0.001).     Characteristics of rats after weight loss. After 12 weeks of feeding with a human-mimicking Western diet, rats intensively gained weight (a), developing the obesity phenotype. Obese animals were subjected to CR for 6 weeks to lose weight. For the next two weeks, animals from the experimental (After Weight Loss, AWL) and control (ContA) groups received an isocaloric amount of calories (stabilization period) to ensure stable body weight. After weight loos, the rats did not show features of insulin resistance (b, c). The mass of cardiac muscle (d) and size of the cardiomyocytes (e) were restored after weight loss in rats which were previously obese. However, the interstitial (f) and perivascular (g) amount of connective tissue was elevated in these animals. Two-way ANOVA with Sidak's multiple comparison test (a, b, c) and the Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, ***p<0.001).

Figure 7
Characteristics of rats after weight loss. After 12 weeks of feeding with a human-mimicking Western diet, rats intensively gained weight (a), developing the obesity phenotype. Obese animals were subjected to CR for 6 weeks to lose weight. For the next two weeks, animals from the experimental (After Weight Loss, AWL) and control (ContA) groups received an isocaloric amount of calories (stabilization period) to ensure stable body weight. After weight loos, the rats did not show features of insulin resistance (b, c). The mass of cardiac muscle (d) and size of the cardiomyocytes (e) were restored after weight loss in rats which were previously obese. However, the interstitial (f) and perivascular (g) amount of connective tissue was elevated in these animals. Two-way ANOVA with Sidak's multiple comparison test (a, b, c) and the Student's T-test or Mann-Whitney test were used for analyzing the data (*p<0.5, ***p<0.001).

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.