Increased intake of fructose causes dramatic metabolic disturbances leading to the development of insulin resistance, type 2 diabetes and the metabolic syndrome in humans and the animals(Reference de Moura, Ribeiro and de Oliveira1, Reference Tran, Yuen and McNeill2). Although the genesis of fructose-induced disorders is very complex and includes different tissues such as hepatic, adipose, etc., cardiac events are also observed(Reference Axelsen, Lademann and Petersen3, Reference Dekker, Su and Baker4). As has recently been reported, among other disturbances, a fructose-rich diet (FRD) is accompanied by cardiac insulin resistance that targets Akt/endothelial NO synthase (eNOS) signalling(Reference Miatello, Risler and Castro5, Reference Zakula, Koricanac and Tepavcevic6). Consequences of enhanced fructose intake are shown to be sex-dependent and the protective effect observed in females is referred to oestrogens(Reference Galipeau, Verma and McNeill7). It is well known that oestrogens are important regulators of cardiac function and the effects of physiological concentrations of oestrogens are beneficial for the heart(Reference Brown, Hoong and Doggrell8, Reference Patten and Karas9).
The phosphatidylinositol 3 kinase/Akt signalling pathway regulating eNOS activity and eNOS itself is very important for cardiac physiology(Reference Massion and Balligand10–Reference Matsui, Nagoshi and Rosenzweig12). To be more interesting, Akt and eNOS are also attributed to female-specific cardiac protection(Reference Camper-Kirby, Welch and Walker13, Reference Murphy and Steenbergen14), and they have been shown to be activated in the heart by oestrogens(Reference Nuedling, Kahlert and Loebbert15, Reference Patten, Pourati and Aronovitz16) in a rapid non-genomic way(Reference Haynes, Sinha and Russell17). Moreover, the phosphatidylinositol 3 kinase/Akt/eNOS axis is a backbone of insulin beneficial cardiac effects and a target of cardiac insulin resistance(Reference Yu, Gao and Ma18). Activation of the Akt/eNOS signalling pathway by insulin or otherwise requires the phosphorylation of Akt at serine 473 and threonine 308 positions(Reference Sale and Sale19), as well as the phosphorylation of eNOS at serine 1177 and dephosphorylation at threonine 495(Reference Andreozzi, Laratta and Procopio20).
In addition, there are no data that a FRD influences cardiac extracellular signal-regulated kinase (Erk 1/2) signalling(Reference Zakula, Koricanac and Tepavcevic6). However, an increased intake of fructose is blamed to be an inducer of heart hypertrophy development(Reference Al-Awwadi, Araiz and Bornet21–Reference Panchal, Poudyal and Iyer23). In contrast, oestradiol (E2) has the ability to activate cardiac mitogenic Erk 1/2 signalling(Reference de Jager, Pelzer and Muller-Botz24, Reference Koricanac, Tepavcevic and Zakula25) and to prevent cardiac hypertrophy in different pathological conditions(Reference Pedram, Razandi and Lubahn26, Reference Donaldson, Eder and Baker27).
On the basis of existing literature data indicating the detrimental effects of a FRD on cardiac Akt/eNOS signalling, as well as the crucial importance of the same pathway for oestrogen signalling in the heart, we hypothesised that E2 might attenuate cardiac disturbances related to an enhanced intake of fructose. We also analysed the effects of a FRD and E2 replacement on cardiac Erk 1/2 signalling related to their role in cardiac hypertrophy development. To validate our hypothesis, we analysed Akt, eNOS and Erk 1/2 expression and phosphorylation in the heart of ovariectomised female rats subjected to a FRD in the absence and presence of E2.
Fructose was purchased from API-PEK (Becej). Anti-phospho-Ser473 Akt, anti-phospho-Thr308 Akt, anti-phospho-Ser1177 eNOS, anti-phospho-Thr495 eNOS, anti-actin and secondary anti-mouse antibody were obtained from Santa Cruz Biotechnology, Inc. Anti-Akt antibody was purchased from Affinity Bioreagents, Inc. Anti-eNOS antibody, anti-Erk 1/2 and anti-phospho-Thr202/Tyr204 Erk 1/2 were products of Cell Signaling. Reagents for the bicinchoninic acid assay were purchased from Pierce. Secondary anti-rabbit and anti-goat antibodies, insulin and 17β-E2 were purchased from Sigma-Aldrich Corporation.
Animal experiments were conducted in accordance with standards approved by the official Vinca Institute's Ethical Committee for Experimental Animals.
Female Wistar rats (21 d old) were separated from their mothers and divided into the control (normal diet; ND) group and FRD group. The control animals had free access to tap water and standard laboratory chow. The animals on the FRD were fed by the same food, but instead of tap water they drank 10 % (w/v) fructose solution. The body weight of rats in the two experimental groups was the same at the beginning of the diet. All animals were kept under standard temperature and dark–light conditions. The diet regimen prolonged for 9 weeks. Body mass was recorded during the diet regimen, while mass of the heart was measured after killing.
At 2 weeks before killing, all animals were bilaterally ovariectomised, under ketamine (40 mg/kg, intraperitoneally)–xylazine (5 mg/kg, intraperitoneally) anaesthesia, to remove endogenous oestrogens and to prevent their effects(Reference Vasudevan, Xiang and McNeill28). Then, half of the fructose-fed rats (FFR) were subjected to E2 replacement treatment (40 μg/kg, subcutaneously, every second day) after ovariectomy (OVX), which continued until the day before killing. To study the fructose and E2 effects on the insulin regulation of the analysed molecules, the experimental groups were divided into two subgroups, one of which was treated with insulin (12 IU/kg of body weight, 40 min before killing, intraperitoneally). Each group contained nine animals in three independent experiments.
The dose and the way of E2 administration were taken over from the literature as a replacement protocol, in order to achieve the concentration of E2 near physiological(Reference Busserolles, Mazur and Gueux29, Reference Sales, Ureshino and Pereira30). The dose and timing of insulin treatment were also estimated on the basis of literature data, adjusted to intraperitoneal administration(Reference Carvalheira, Calegari and Zecchin31), and optimised to observe the peak of changes in the phosphorylation of Akt, Erk and eNOS.
The ovariectomised FFR were compared with the ovariectomised animals subjected to the normal diet to assess the effect of the fructose diet. The FFR treated with E2 were compared with the ovariectomised FFR to analyse the effect of E2 in the context of the fructose diet, as well as with ovariectomised rats on the normal diet, to estimate E2 capacity to counteract to the expected detrimental effect of the FRD. The equivalent comparisons were performed for insulin-treated counterparts of all three groups.
Preparation of cardiac cell lysate
After killing, the heart of rats was removed from the body, washed with cold saline, dried and measured, and the tissue was homogenised on cold with an Ultra-turrax homogeniser in a buffer (pH 7·4) containing 10 mm-Tris, 150 mm-NaCl, 1 mm-EDTA, 1 mm-ethylene glycol tetraacetic acid (EGTA), 1 % Triton X-100, protease inhibitors and phosphatase inhibitors(Reference Dupont, Derouet and Simon32). The homogenates were centrifuged at 600 g for 20 min at 4°C and the supernatants obtained were ultracentrifuged for 60 min at 100 000 g. Protein concentration was determined by the bicinchoninic acid method(Reference Smith, Krohn and Hermanson33) and after boiling in Laemmli sample buffer, the supernatants were used as cardiac cell lysate for Western blot analysis.
SDS–PAGE and Western blotting
Cardiac lysate proteins (50 μg/lane) were separated on 7·5 % or 10 % SDS–PAGE and transferred to polyvinylidene fluoride membranes(Reference Laemmli34). The membranes were blocked with 5 % bovine serum albumin and blotted with an antibody against phospho-Akt, anti-phospho-eNOS or phospho-Erk 1/2. After extensive washing, the membranes were incubated with the appropriate secondary horseradish peroxidase-conjugated antibody and used for subsequent detection with ECL reagents. Thereafter, the membranes were stripped, washed and reprobed with antibodies against Akt, eNOS or Erk, respectively. To ensure that protein loading was equal in all samples, the blots were stripped again and incubated with actin antibody. Films were scanned and analysed using ImageJ software (NIH).
RNA isolation and quantitative real-time RT-PCR
Immediately after removal, animal tissue was snap-frozen in liquid N2 and then stored at − 70°C. Total RNA was isolated with TRI Reagent (Ambion, Inc.) according to the manufacturer's instructions. The quantity of mRNA was assessed by a NanoDrop® ND-1000 spectrophotometer (Thermo Scientific). Structural integrity of RNA was confirmed by formaldehyde agarose gel electrophoresis. Total RNA from the tissue specimens with satisfactory quality was converted to complementary DNA. Then, 1 μg of RNA was treated with DNase I (Fermentas) and reverse transcription was performed using a First Strand cDNA Synthesis Kit, with oligo-dT18 primers (Fermentas), according to the manufacturer's instructions. Mock reactions lacking RT were performed during the complementary DNA synthesis step in order to additionally exclude genomic contamination. Real-time PCR was performed in duplicate in an ABI Real-time 7500 System (ABI). Detection of eNOS gene expression was done by amplification in a total volume of 25 μl in EvaGreen qPCR Mastermix-R (Applied Biological Materials, Inc.). Final concentrations of primers were 6·25 μmol/l each. The reaction conditions were as follows: an initial denaturation at 95°C for 10 min followed by forty cycles of a denaturation step at 95°C for 45 s, an annealing step at 60°C for 60 s and an elongation step at 72°C for 60 s. All samples were normalised to the housekeeping GAPDH transcript levels. Primer sequences of GAPDH and eNOS genes are available upon request. Differences in mRNA expression between the groups were tested by REST 09 software (Corbett Life Science).
Values are expressed as means and standard deviations (except for mRNA expression results, where standard errors were used instead of standard deviations). We performed the same experiment three times independently and pooled the results for all nine animals per group for one-way ANOVA. A value of P< 0·05 was considered as statistically significant. The power analysis of the study was performed by ‘PS power and sample size program’ software(Reference Dupont and Plummer35). The power value for nine animals per group, with an observed effect size and standard deviations and a significance level of 0·05, was >0·8. In addition, differences in mRNA expression between the groups were tested by a pairwise randomisation and bootstrapping technique using the relative expression software tool, REST 09 (Corbett Life Science, http://rest.gene-quantification.info)(Reference Pfaffl, Horgan and Dempfle36).
Mass of the heart of fructose-fed rats related to oestradiol status
Based on the data that the FRD and the E2 treatment influence the mass of the heart and are included in the genesis or protection of heart hypertrophy, respectively, we measured the mass of the heart and expressed it as an absolute value or relative to the body mass. As presented in Fig. 1, neither diet nor hormone replacement treatment altered the mass of the heart (Fig. 1(a)) and the heart:body ratio (Fig. 1(b)).
Effects of the fructose-rich diet and oestradiol status on the expression and phosphorylation of Akt in the heart of ovariectomised female rats
Akt kinase is regulated by numerous phosphorylations of the molecule, but those at positions Ser473 and Thr308 are assigned as crucial for enzyme activity. As presented in Fig. 2(a) and (b), the results of the present study confirmed that insulin increases the phosphorylation of both residues. The FRD decreased the phosphorylation of both amino acid residues in the heart of rats, in basal conditions (with no additional insulin treatment) and/or insulin-stimulated conditions (P< 0·001, OVX-FRD v. OVX-ND for Ser473; P< 0·01, OVX-FRD v. OVX-ND; P< 0·05, OVX-FRD+insulin v. OVX-ND+insulin for Thr308; Fig. 2(a) and (b)). The E2 treatment of FFR reversed phospho-Ser473 Akt content to the control level in basal conditions. However, E2 replacement was effective in reversing the diet effects on Akt phosphorylation at Thr308 in insulin-stimulated conditions, while failed to reverse the FRD effect in basal condition (P< 0·01, OVX-FRD+E2 v. OVX-ND). Neither the diet nor hormone replacement changed the protein expression of Akt in the heart (Fig. 2(c)), determined as Akt content in cardiac lysate.
Expression and phosphorylation of the cardiac endothelial nitric oxide synthase effects of the fructose-rich diet and oestradiol replacement
Enzyme activity of eNOS is also regulated by a number of serine/threonine phosphorylations. Phosphorylation at serine 1177 is accompanied by the activation of eNOS, while phosphorylation at threonine 495 is found to be constitutive and inhibitory. Insulin itself stimulated phosphorylation at Ser1177 only (Fig. 3(a)). As presented in Fig. 3(a), a significant decrease in Ser1177 phosphorylation of eNOS was observed in the heart of FFR, in insulin-stimulated conditions (P< 0·05, OVX-FRD+insulin v. OVX-ND+insulin). E2 over-reverted the negative effects of the fructose diet (P< 0·05, OVX-FRD+E2 v. OVX-FRD; P< 0·05, OVX-FRD+E2 v. OVX-ND; P< 0·01, OVX-FRD+E2+insulin v. OVX-FRD+insulin; P< 0·05, OVX-FRD+E2+insulin v. OVX-ND+insulin) and elevated the eNOS serine 1177 phosphorylation level above the level detected in rats fed on the normal diet.
In contrast to the serine 1177 site, the FRD caused an increase in the phosphorylation of eNOS at threonine 495 in basal (P< 0·001, OVX-FRD v. OVX-ND) and insulin-stimulated conditions (P< 0·05, OVX-ND+insulin v. OVX-ND+insulin) (Fig. 3(b)). E2 replacement had no effect on this phosphorylation in FFR.
The protein content of eNOS determined in heart lysate was unaffected by the FRD and did not depend on the presence of E2 (Fig. 3(c)). In contrast to protein expression, as presented in Fig. 3(d), the steady-state mRNA level of eNOS was surprisingly elevated in FFR (OVX-FRD v. OVX-ND, expression mean factor 5·765, SE range 2·355–12·795, 95 % CI 1·904, 19·051, P= 0·000). It is interesting to note that OVX and the fructose diet separately acted as strong negative regulators of cardiac eNOS mRNA expression (ovariectomised v. intact rats, on the normal diet: mean factor 0·154, P= 0·018; intact rats on the fructose diet v. intact rats on the normal diet: mean factor 0·077, P= 0·016). In FFR, E2 replacement returned the eNOS steady-state mRNA level to the control value (OVX-FRD+E2 v. OVX-ND, expression mean factor 0·995, SE range 0·290–3·144, 95 % CI 0·193, 4·620, P= 0·901).
Effects of the fructose-rich diet and oestradiol replacement on the phosphorylation and expression of extracellular signal-regulated kinase 1/2 in the heart of ovariectomised female rats
The absence of the potential effects of the fructose diet and E2 treatment on heart mass initiated the analysis of their effects on mitogenic signalling through Erk 1/2 kinases. In accordance with the results obtained for the absolute and relative mass of the heart, we did not observe any changes in Erk 1/2 phosphorylation at Thr202/Tyr204 and protein expression, caused by the diet or hormone treatment (Fig. 4(a) and (b)).
Akt and eNOS play a very important role in the regulation of cardiac functions(Reference Massion and Balligand10, Reference Matsui, Nagoshi and Rosenzweig12). Cardiac Akt/eNOS signalling is a target of insulin resistance(Reference Zakula, Koricanac and Tepavcevic6) and E2 regulation as well(Reference Patten, Pourati and Aronovitz16). Phosphorylation of eNOS by Akt with a subsequent increase in NO production is an important downstream effector in the anti-apoptotic signalling by insulin. Moreover, increasing evidence suggests that eNOS acts as a metabolic sensor in cardiomyocytes, implying that defective NO production might be linked to cardiomyocyte metabolism dysfunction(Reference Yu, Gao and Ma18). On the basis of published data, we performed the study of the effects of oestrogen replacement on this signalling pathway in ovariectomised female rats fed the FRD. In addition, we analysed heart mass and the mitogenic Erk pathway, which also might be modulated by the diet regimen and oestrogen treatment.
In contrast to studies implicating fructose effects on the development of heart hypertrophy(Reference Al-Awwadi, Araiz and Bornet21–Reference Panchal, Poudyal and Iyer23), we did not observe any changes in absolute and relative heart mass in FFR, which could be attributed to the duration of the diet regimen and/or the content of fructose in the applied diet(Reference Al-Awwadi, Araiz and Bornet21, Reference Panchal, Poudyal and Iyer23). Hormone replacement treatment was also ineffective in terms of the changes in the heart mass of FFR. These data are in line with the results obtained for the mitogenic Erk 1/2 signalling pathway regulating cell proliferation and growth, where diet- or hormone-related changes were absent. The lack of the fructose effect on Erk 1/2 phosphorylation is in accordance with our recently reported results(Reference Zakula, Koricanac and Tepavcevic6), but the absence of the E2 effect is in contrast to the results obtained in cell culture(Reference de Jager, Pelzer and Muller-Botz24) or in acute hormone treatment(Reference Koricanac, Tepavcevic and Zakula25). It could be a consequence of the way of hormone administration, but also might implicate some fructose diet-induced disturbances in the activation of Erk 1/2 signalling by E2.
The results of the present study concerning the deregulation of cardiac Akt/eNOS signalling in FFR are mostly in accordance with literature data. As we already reported, the fructose-enriched diet decreased cardiac Akt phosphorylation at Ser473 and Thr308 and eNOS at Ser1177, especially in insulin-stimulated conditions(Reference Zakula, Koricanac and Tepavcevic6). On the other hand, increased intake of fructose stimulated inhibitory phosphorylation of cardiac eNOS at threonine 495. It was well documented that heart eNOS activity is decreased in FFR(Reference Miatello, Risler and Castro5, Reference Miatello, Vazquez and Renna22, Reference Vazquez-Prieto, Rodriguez Lanzi and Lembo37, Reference Vazquez-Prieto, Renna and Lembo38). Myocardial Akt protein expression and its Ser473 phosphorylation, as well as eNOS expression and phosphorylation at Ser1177, were decreased in FFR(Reference Prakash, Khanna and Singh39). However, in contrast to the present results, expression of eNOS protein and mRNA was shown to be reduced in the heart of FFR(Reference Nyby, Matsumoto and Yamamoto40). It was astounding that we observed no effect of the fructose diet on eNOS protein content, but a strong increase in the gene expression. We have no physiological explanation particularly regarding that OVX and fructose, in the present study, separately caused a strong decrease in cardiac eNOS mRNA level.
There are no published data concerning the effects of E2 on cardiac Akt/eNOS signalling in increased fructose intake conditions or other animal models of insulin resistance. The only relevant results that can be cited are those describing the effects of E2 treatment in normal rats. We recently reported a rapid stimulatory effect of a single injection of E2 on cardiac Akt Ser473 and Thr308, as well as on eNOS Ser1177 phosphorylation(Reference Tepavcevic, Koricanac and Zakula41). Others have also observed that E2 treatment increased Akt phosphorylation and Akt-mediated eNOS phosphorylation (Ser1177)(Reference Tagashira, Bhuiyan and Shioda42). A number of studies have shown that E2 increased cardiac levels of eNOS, and phosphorylated eNOS (Ser1177), thus exerting potentially beneficial cardiovascular effects of NO(Reference Kan, Hsu and Ba43, Reference Ricchiuti, Lian and Oestreicher44). In ovariectomised rats, E2 replacement restored eNOS activity and expression(Reference Wang and Abdel-Rahman45). E2 stimulated the expression of eNOS in both neonatal and adult cardiac myocytes(Reference Nuedling, Kahlert and Loebbert15). In the present study, we observed no effect of E2 on the expression of eNOS in the heart of FFR and its threonine 495 phosphorylation, while the hormone effect on serine 1177 phosphorylation was strongly inducible.
We noted that the effect of the FRD and the ability of E2 to exert an effect on phosphorylation of different amino acid residues in Akt and eNOS are selective. Observed differences in the effects of fructose and E2 on the phosphorylation level of different amino acids in the protein structure of Akt and eNOS actually implicate the different effects of these agents on kinase/phosphatase pathways responsible for the phosphorylation/dephosphorylation of these sites. It is well known that phosphorylation of Akt Thr308 is attributed to phosphoinositide-dependent protein kinase 1 (PKD1) kinase, while Rictor-mammalian target of repamycin (mTOR) kinase is probably responsible for Ser473 phosphorylation(Reference Sale and Sale19). On the other hand, eNOS is a substrate of Akt kinase that phosphorylates the Ser1177 position(Reference Michell, Griffiths and Mitchelhill46). The phosphatase to which is referred insulin-induced dephosphorylation of eNOS at the negative regulatory site Thr495 is probably protein phosphatase 1(Reference Andreozzi, Laratta and Procopio20).
To the best of our knowledge, no published literature contains any data concerning the effects of enhanced fructose intake on kinases/phosphatases involved in the phosphorylation/dephosphorylation of Akt and eNOS molecules on amino acid residues analysed in the present study, except Akt as a kinase responsible for eNOS phosphorylation at Ser1177. The FRD-induced decrease in Akt phosphorylation points to the decreased activity of responsible kinases or the increased activity of phosphatases that counteract PDK1 and Rictor-mTOR. Inhibition of Akt could be a direct explanation of the observed decrease in the phosphorylation of eNOS at Ser1177 in FFR. The increase in the phosphorylation of eNOS at Thr495 in animals fed on the FRD indicates changes in the activity of protein phosphatase 1 or kinases responsible for the phosphorylation of this amino acid residue.
Data concerning the effects of E2 on these kinases and phosphatases, particularly in cardiac tissue, are also extremely scarce and we can only speculate, on the basis of the present results, that E2 activates cardiac Rictor-mTOR kinase and probably PDK1, but has no effect on protein phosphatase 1 in the heart.
In conclusion, there was a consistent suppression of cardiac Akt/eNOS signalling at the level of protein phosphorylation in FFR. It is interesting to note that E2 managed to abrogate some, but not all, of fructose-induced damage at the basal level or in insulin-stimulated conditions, implicating its selective effects on kinase and phosphatase pathways involved in the activation of Akt/eNOS signalling. As a matter of future perspectives, it would be very intriguing to conduct an additional study the effects of E2 on PDK1, Rictor-mTOR and protein phosphatase 1 in cardiac tissue.
This study was supported by project grant no. 41009 and project grant no. 175085, from the Ministry of Education and Science, Republic of Serbia. S. R. conducted the experiments and analysed the data. G. K. and Z. Z. designed the study, conducted the experiments, analysed the data and wrote the paper. S. T., T. M., M. S. and M. P. conducted the experiments and statistical analysis. M. Z. and A. S. conducted the experiments and revised the manuscript. All authors contributed to and approved the final manuscript. The authors declare that there are no conflicts of interest.