Beneficial effects of cherry consumption as a dietary intervention for metabolic, hepatic and vascular complications in type 2 diabetic rats

Values are expressed as mean ± SEM, and n indicates the number of rats. Statistical analysis was performed with Student’s t test for unpaired data or ANOVA followed by LSD test after normality test validation protected least-significant difference tests, where appropriate (Statistica^® version 12, StatSoft, France). If normality was violated, we used log transform. P < 0.05 was considered to be statistically significant.

HFHF rats maintained a significantly higher weight gain, body mass index and body weight (Table 1, Fig. 1a) than ND rats until the end of the study period. This increase was associated after 4 months with higher abdominal circumference, abdominal fat (Table 1) and hyperleptinemia (37.2 ± 5.4 vs. 13.2 ± 0.9 ng/mL) (Fig. 2a). Moreover, HFHF rats developed fasting glycaemia (defined as glucose levels > 1.26 g/L; 1.35 ± 0.04 vs. 1.10 ± 0.04 g/L) and insulin resistance (defined as HOMA-IR value > 2.4; 4.84 ± 0.63 vs. 1.43 ± 0.15) while ND rates did not (Fig. 1b, c). Further, HFHF rats developed glucose intolerance with area under the curve (343 ± 29 vs. 244 ± 18) and C-peptidemia higher than in ND rats (during ipGTT: t0′: 1954 ± 267 vs. 583 ± 81; t60′: 4398 ± 299 vs. 2274 ± 284 pmol/L) at a faster rate than the ND rats (Fig. 1d, e). Finally, higher cholesterol (2.2 ± 0.3 vs. 1.4 ± 0.1 mM), triglyceride (1523 ± 185 vs. 805 ± 175 µM) and free fatty acid (234 ± 29.4 vs. 82 ± 10.6 µM) levels revealed that dyslipidaemia was induced by 4 months of HFHF diet administration (Fig. 2b–d).

The HFHFCherry diet did not have a beneficial impact on metabolic and lipidic parameters (Figs. 1, 2a–d). In fact, HFHFCherry rats had the same body weight evolution as rats maintained on HFHF diet, with a higher fasting leptin level than HFHF rats (51.26 ± 4 ng/mL). HFHFCherry rats developed hyperglycaemia (1.45 ± 0.07 g/L), dyslipidaemia (TG: 131 ± 189 µM; Chol: 2.86 ± 0.29 mM; FFA: 354 ± 86 µM), glucose intolerance (area under the curve: 391 ± 25), and insulin resistance (HOMA-IR: 6.1 ± 0.95). Moreover, they developed C-peptidemia after 2 h of ipGTT at a faster rate than HFHF rats (t0′: 2423 ± 397 and t60′: 5841 ± 275 pmol/L) (Figs. 1, 2a–d).

However, the addition of cherries to the nutritional intervention where ND replaced the HFHF diet (HFHF/NDCherry) normalized glucose tolerance (area under the curve: 243.2 ± 10.32) and C-peptide levels during ipGTT (t0′: 1029 ± 118 and t60′: 3222 ± 368 pmol/L). Moreover, fasting glycaemia was under 1.26 g/L (1.24 ± 0.03 g/L) and cherry consumption decreased HOMA-IR (2.541 ± 0.29) (Fig. 1b–e). Plasmatic FFA levels were decreased (130 ± 11.7 µM) but not normalized to ND levels. However, hypercholesterolemia and hypertriglyceridemia were eliminated (1.52 ± 0.16 mM and 650 ± 167 µM, respectively) (Fig. 2b–d). The HFHF/NDCherry diet stopped weight gain until the end of the treatment and body weight, body mass index, abdominal circumference (Table 1, Fig. 1a) and leptinemia (20.8 ± 2.7 ng/mL) were normalized to ND levels (Fig. 2a).

HFHF rats had a twofold higher TBARS level in plasma than ND rats (62.80 ± 8.9 vs. 35 ± 1.1 µM MDA) (Fig. 2e). This oxidative stress complication was associated with an increase in superoxide dismutase (SOD) activity (71.43 ± 4.22 vs. 57.52 ± 5.83% of inhibition) and a decrease in catalase activity (0.022 ± 0.004 vs. 0.052 ± 0.03 µM) (data not shown). No difference in total antioxidant capacity was observed despite plasmatic oxidative stress (5.75 ± 0.17 vs. 5.75 ± 0.18 µM eq. Trolox) (Fig. 2f).

The HFHFCherry diet had no beneficial impact on plasmatic oxidative stress. In fact, HFHFCherry rats had the same levels of TBARS (52.21 ± 9.5 µM MDA) and SOD activity (78 ± 3.67% of inhibition) as HFHF rats, a great variability of catalase activity (neither significant vs. ND rats nor HFHF rats, 0.031 ± 0.014 µM) and also decreased total antioxidant capacity (4.36 ± 0.49 µM eq. Trolox) (Fig. 2e, f and data not shown).

However, the HFHF/NDCherry diet normalized plasmatic TBARS levels (33.41 ± 2.08 µM MDA) and catalase activity (0.043 ± 0.01 µM), had a higher SOD activity (81.87 ± 1.08% of inhibition) and increased total antioxidant capacity (6.62 ± 0.09 µM eq. Trolox) (Fig. 2e, f and data not shown).

Rats fed an HFHF diet showed a decrease in hepatic glycogen content (0.024 ± 0.003 vs. 0.042 ± 0.004 mg/mg of liver) and a maximal steatosis score after 4 months (3.0 ± 0 vs. 0.67 ± 0.21) (data not shown). Many cell nuclei were shifted from a location in the centre of the hepatocyte to the periphery, apparently due to interference from normal cell structures because of the presence of numerous large fat globules (Fig. 3a, b). These structural disorders were associated with a higher content of hepatic TG (19.9 ± 2.74 vs. 1.83 ± 0.08 nmol/mf of liver) and cholesterol (32.7 ± 9.87 vs. 7.82 ± 0.68 mg/mg protein) (Fig. 3c). These complications were associated with inflammation with macrophages infiltration (1.02 ± 0.12 vs. 0.57 ± 0.14% of area) and oxidative stress (222 ± 31.3 vs. 100 ± 7.7%) (Fig. 3a, b).

The HFHFCherry diet had no beneficial impact on plasmatic oxidative stress (Figs. 2f, 3). In fact, HFHFCherry rats had a great variability in glycogen content (neither significant vs. ND rats nor HFHF rats, 0.036 ± 0.005 mg/mg of liver), a maximal steatosis score (3.0 ± 0) with hepatic TG (8.52 ± 1.45 nmol/mf of liver) and cholesterol (32.85 ± 6.54 mg/mg protein) associated with a threefold macrophages infiltration (1.55 ± 0.1% of area) and oxidative stress (200 ± 18.6%).

However, the HFHF/NDCherry diet normalized macrophages infiltration (0.42 ± 0.13% of area), oxidative stress (93 ± 7.1%) and TG (2.58 ± 0.45 nmol/mf of liver) (Fig. 3a, b). HFHF/NDCherry rats presented an intermediate appearance and steatosis score (1.67 ± 0.2) associated with hepatic cholesterol (26 ± 4.76 mg/mg protein) and lower glycogen content than ND rats (0.0267 ± 0.0035 mg/mg of liver) (Fig. 3 and data not shown).

The HFHF diet induced hepatic increase of p22phox expression (0.85 ± 0.15 vs. 0.40 ± 0.06 a.u) (Fig. 4a), an NADPH oxidase (Nox) subunit and a major source of glucose-induced ROS production in liver [50, 53]. Moreover, Nrf2 polyubiquitination was increased 1.5-fold in HFHF rats in comparison to ND rats (0.45 ± 0.04 vs. 0.32 ± 0.03 a.u) (Fig. 4b), all leading to excessive hepatic ROS formation and oxidative stress as shown before. The HFHFCherry and HFHF/NDCherry diets preserved a physiological level of p22phox (respectively: 0.51 ± 0.10 and 0.66 ± 0.16 a.u) and Nrf2 expression and ubiquitination (respectively: 0.27 ± 0.03 and 0.30 ± 0.03 a.u) (Fig. 4a, b).

The HFHF diet increased the sterol regulatory element-binding protein-1c (SREBP-1c) (0.09 ± 0.01 vs. 0.06 ± 0.01 a.u), the carbohydrate-responsive element-binding protein (ChREBP) (0.16 ± 0.04 vs. 0.05 ± 0.02 a.u) and the master regulator of intracellular cholesterol homeostasis (SREBP2) (0.40 ± 0.08 vs. 0.12 ± 0.02 a.u), three major transcription factors implicated in liver lipogenesis [54, 55]. The HFHFCherry diet decreased SREBP1 (0.035 ± 0.01 a.u), tended to increase ChREBP (0.12 ± 0.05 a.u; P = 0.08) and normalized SREBP2 (0.09 ± 0.02 a.u) expressions in comparison to the ND whereas the HFHF/NDCherry diet preserved a physiological level of SREBP-1c (0.05 ± 0.01 a.u), ChREBP (0.08 ± 0.03 a.u) and SREBP2 (0.21 ± 0.08 a.u) (Fig. 4c–e).

HFHF rats had a decrease of NO-mediated relaxation in the mesenteric artery associated with oxidative stress and decreased expression of eNOS in the endothelium of the vessel (Fig. 5). In fact, acetylcholine, which induced relaxation via a calcium dependent pathway [52], caused NO-mediated concentration-dependent relaxations in mesenteric artery rings in ND rats (at 10⁻⁵ M, 67 ± 7%) associated with eNOS expression in the vessel (100 ± 10.73%). However, blunted NO-mediated relaxation was observed in HFHF rats (at 10⁻⁵ M, 40 ± 11%), associated with a twofold decrease in eNOS (53.35 ± 7.17%) and a threefold ROS formation in all the vasculature, as observed with DHE fluorescence (339 ± 56.3% vs. 100 ± 7.14%.). Decrease in NO was not correlated with peroxintrite formation (association of ROS and NO) as observed by physiological nitrotyrosine fluorescence levels (113.5 ± 14.97 vs. 100 ± 8.32%) (Fig. 5b).

HFHFCherry and HFHF/NDCherry rats experienced an intermediate NO-mediated relaxation (respectively at 10⁻⁵ M, 59.77 ± 16.22%; 58.09 ± 18.80%) over 2 months, similar to ND and HFHF rats. However, HFHFCherry rats presented excessive ROS (338 ± 44.1%) in all the vasculature associated with endothelial peroxinitrite (194.5 ± 26.45%) and a compensatory increase of eNOS expression (156.3 ± 26.3%). In contrast, the HFHF/NDCherry diet prevented ROS and nitrotyrosine formation (respectively: 98.3 ± 18.6%; 129.46 ± 14.43%) and presented a physiological level of eNOS (108.3 ± 11.7%) (Fig. 5).

The HFHF diet induced an increase in ROS in the entire pancreas (212.30 ± 25.2% vs. 100 ± 12%). The HFHFCherry and HFHF/NDCherry diets had no effect on pancreatic oxidative stress because of the formation of ROS in the pancreas, including islets, as observed in HFHFCherry rats (172.6 ± 13.3%) and HFHF/NDCherry rats (234 ± 32.7%) (Additional file 2: Figure S1).

Figure 6 presents all disorders observed in HFHF rats and highlights the beneficial impacts of cherry consumption and nutritional intervention on blood, vessels, and liver.

Firstly, we have demonstrated that cherry consumption decreased oxidative stress in plasma. While diabetes induces a decrease in catalase activity and TBARS production without any modulation of TAOC, cherry consumption was able to increase catalase activity and TAOC leading to a decrease in TBARS complications. Accordingly, studies in healthy human subjects reported that cherry consumption increases plasma TAOC [43, 56]. Traustadottir et al., in a double blind placebo-controlled crossover design in older adults, showed that consumption of tart (sour) cherry juice improves antioxidant defences by increasing the capacity to constrain an oxidative challenge and reducing oxidative damage to nucleic acids [57]. Moreover, cyanidin-3-rutinoside, present in our cherry extract [35], displays a wide range of biological activities, including antioxidant and anti-inflammatory [58]. Recently, an in vitro study confirmed the beneficial impact of cyaniding-3-rutinoside on oxidative stress damage and the inhibition of TBARS formation in bovine serum albumin [59]. Recent epidemiological studies highlighted that people with acatalasemia develop T2D [60] and Hait et al. have demonstrated that catalase deletion promotes obesity associated with the impairment of glucose tolerance and insulin sensitivity, increased plasmatic TGs and induced steatosis and inflammation in the liver of Cat^−/− mice [61]. Our results are in accordance with these data and suggested that cherry consumption could prevent alterations in lipids mobilization and utilization thought their beneficial effect on catalase, thus avoiding excess circulating lipids. Therefore, all these data highlight not only the improvement of systemic oxidative balance with cherry consumption but also an improvement of lipids profiles.

Besides the beneficial impact on redox homeostasis, cherry consumption normalized glucose tolerance, insulin resistance, dyslipidaemia, hyperleptinemia, decreased hyperglycaemia, and hyperinsulinemia. These beneficial impacts of cherry consumption could be closely linked to its ability to decrease obesity and inhibit adipocyte dysfunction, two disorders strongly associated with the development of insulin resistance, cell impairment and T2D [62, 63]. In fact, anthocyanins are considered modulators of adipose tissue metabolism which improve adipocytes dysfunction and adipocytokines secretion in insulin resistance, increase β-oxidation and decrease fat accumulation on adipocytes [64]. Hypertrophic adipocytes, which release rather than store FFAs, are linked to insulin resistance [65], but were decreased by cherry consumption, which could explain the normalization of glucose tolerance, insulin sensitivity, leptinemia and dyslipidaemia. All these beneficial effects were observed only when cherry consumption was associated with ND and not with the HFHF diet. Then, our results suggest that cherry consumption may cause lipid trafficking away from the abdomen and hence reduce the associated complications, mainly NASH and cardiovascular dysfunction. These findings are in accordance with some in vivo studies which demonstrated that cherries and their bioactive food components decrease body weight and abdominal fat [66], blood lipids [67, 68] and fasting blood glucose [64, 66]. More precisely, Cherian et al. demonstrated that a single dose of anthocyanins decreases fasting glycaemia by 19% and improves glucose tolerance by 29% in moderately-diabetic rats. Moreover, 4 weeks of treatment dropped the pre-treatments levels of fasting blood glucose by 50% and increased glucose tolerance by 41% [69]. Similar results were observed in high fat diet-rats with 5-caffeoylquinic acid [70], one of the compounds in Regina cherries [35]. Another therapeutic approach to treat diabetes is to delay the absorption of glucose via inhibition of enzymes, such as α-glucosidase, in the digestive organs. It has been confirmed that α-glucosidase activity in vitro can be inhibited by berry extracts rich in polyphenols [71] and by cyanidin-3-rutinoside [72], a derivate of anthocyanin present in the cherry extract used in our study [35]. All these data highlight the fact that bioactive food compounds found in cherries are responsible for an improvement in systemic metabolic balance.

Keeping a state of oxidative, carbohydrate and lipid homeostasis is essential to ensuring vascular function. In fact, the endothelium, the internal layer of vessels, is in constant interaction with the blood, subjected to mechanical and chemical stresses, and plays a pivotal role in vascular homeostasis. A better understanding of the cellular basis of the pathophysiological processes and better strategies to treat damage are clearly an important goal because diabetes-associated vascular complications are responsible for 75% of the deaths associated with diabetes [73]. We have demonstrated that HFHF induced diabetes is associated with a decrease of relaxation in the mesenteric artery involving a decrease of eNOS expression and, thus, blunted NO-mediated relaxation, in addition to ROS formation. This endothelial dysfunction has been associated in several regions of the vasculature in animals and humans with T2D due to defects in NO-derived vasodilation [74‐77]. A number of studies have suggested that ROS play an important role in the pathogenesis of diabetic vasculopathy, affecting both the macro- and the microvascular systems [63, 78]. All metabolic disorders observed in blood in our model could be linked to disturbance of vessels [76]: high cholesterol and FFA, obesity and visceral fat distribution, insulin resistance, impaired fasting glucose and glycaemic fluctuations [7, 79, 80]. They have also been associated with ROS formation and exacerbated oxidative stress [7, 12, 63, 81]. However, as shown before in our study, cherry consumption was able to assure blood homeostasis, leading to decreases in oxidative stress in vessel vasculature, to increases eNOS expression and thus to assure good NO-bioavailability and relaxation in HFHF/NDCherry rats. Much of research suggests, in fact, the cardio-protective effects of cherry consumption and anthocyanins appear to have some vasoprotective effects in humans [82]. Endothelial cells from bovine arteries exposed for several hours to cyanidins increased NO output and reduced local oxidative stress [83], decreased inflammation and indirectly reduced the risk of atherosclerosis plaque formation [84]. In fact, polyphenols, such as citrus flavonoids or isoflavones from red clover, could increase flow mediated dilatation and improve vascular function after 3–4 weeks in patients with metabolic syndrome [85, 86] and anthocyanins, including cyanidin-3-rutinoside as shown in our cherry composition, have potent platelet-inhibitory properties and are considered inhibitors of platelet cell signalling and thrombus formation [87]. Moreover, cherry consumption inhibited free radical formation which could prevent the onset and development of long-term diabetic complications [88]. This vascular protection of cherry consumption against oxidative stress has been closely correlated to the improvement of metabolic and lipidic profiles in blood. In fact, the HFHFCherry diet has no beneficial impact on these plasmatic parameters and thus exhibited ROS in all the vasculature. However, in this nutraceutical strategy, eNOS was highly increased by cherry supplementation to counteract ROS and peroxinitrite formation and thus to assure NO bioavailability and relaxation in vessels, as shown in HFHF/Cherry-rats. Changes in endothelial function hasten the development of micro- and macroangiopathy and thus target the cellular basis of endothelial dysfunction and promote NO bioavailability. Cherry supplementation, in addition to lifestyle measures, should provide benefits to the overall therapeutic management of diabetes.

In addition to its vascular beneficial effect, NO derived from eNOS appears to have both antiobesogenic and insulin-sensitizing properties. These effects, discovered in recent years, are due to its ability to increase fat oxidation in peripheral tissues, such as liver and adipose tissue, to decrease lipid synthesis in the liver, increase insulin and glucose transports to key peripheral tissues and to regulate gluconeogenesis [89]. These metabolic effects of NO bioavailability could explain in part the beneficial impact of cherry consumption on the liver. In fact, massive hepatic lipid accumulation observed in HFHF rats in our study and in T2D patients [90] was eliminated in HFHF/NDCherry rats. Our findings are in line with an earlier study in HFD-mice and consumption of a mixture of pure anthocyanins [66]. Seymour et al. also reported a decrease in hyperlipidemia, hyperinsulinemia, fatty liver and hepatic steatosis [68] with 90-day administration of sour cherry. Moreover, supplementation of the HFD-rats with 5-caffeoylquinic acid, one of our cherry compounds [35], reduced macrophage infiltration and steatosis [70] via PPARγ and NFκB signaling pathways. Today, these pathways, such as antioxidants inhibiting NADPH oxidase, receive a lot of attention in the treatment of atrial fibrillation [91]. So, our results, in addition to others, highlighted the anti-steatosic effect of cherry consumption.

While many mechanisms can explain the improvement of hepatic steatosis and NASH complication, normalization of FFAs and oxidative stress by cherry consumption seem to be involved in our study. Li et al. demonstrated that FFAs cause hepatic insulin resistance, resulting in overproduction of glucose and hyperglycemia, and initiate inflammatory processes in the liver, thus resulting in the development of steatohepatitis [65]. Additionally, Pereira et al. linked FFAs to NADPH oxidase and oxidative stress in impaired hepatic insulin signalling [92]. Numerous disorders stimulate NADPH oxidase activity: elevated glucose, hyperinsulinemia, lipids, and cytokines [93]. As shown before, only cherry consumption in association with lifestyle measures (ND) normalized them in plasma and in the liver cherry consumption decreased HFHF-induced p22phox expression (NADPH oxidase subunits), ROS formation and NASH (steatosis plus inflammation). Previous studies demonstrated that anthocyanins reduce ROS generation in human HepG2 cells exposed to a high glucose environment [94], increase activity of the antioxidant enzymes SOD (liver, blood) and Gpx (liver) and decrease lipid peroxidation [95]. One of the major defence systems against stress-related injury is the Nrf2 system [10, 96] which activates the antioxidant response elements. Activation of Nrf2 by a number of polyphenols increases expression of phase II detoxifying enzymes and antioxidant enzymes [97, 98], which can directly act to eliminate free radicals and oxidative damaged molecules. Recently, Nrf2 was highlighted as a real target against diabetic nephropathy [10]. Our results showed that the HFHF diet increased Nrf2 degradation in the liver and was clearly associated with the presence of oxidative stress. But cherry supplementation avoided this degradation, in addition to decreasing p22phox expression and ROS formation.

Degradation of Nrf2 was also implicated in the development of NASH [99, 100] because of its role in lipid catabolism [101]. SREBPs (-1 and -2) and ChREBP regulate the gene expression of enzymes involved in lipogenesis and cholesterol synthesis [54, 102]. Our present study demonstrated that the HFHF diet induced an abnormal expression profile of these hepatic lipogenic transcription factors, increasing de novo lipogenesis due to hyperglycemia and hyperinsulinemia [55, 103] as observed in our model. However, cherry consumption was able to normalize them, decrease hepatic TGs and steatosis and thus decrease plasmatic dyslipidemia and hyperglycemia. Polyphenols, including anthocyanins, significantly reduce tissue lipid accumulation and the activity of enzymes that promote fat storage [104], and also lower body weight, fat mass and TGs through enhancing energy expenditure, fat utilization and modulating glucose hemostasis [105]. Little data are available on the in vivo effect of anthocyanins or cherry consumption on these hepatic targets. Anthocyanin from mulberry extract decreases SREBP1c and SREBP2 on human hepatocyte (HepG2) cultured with high fatty acid, suppressing fatty acid synthesis and enhancing fatty acid oxidation, all contributing to amelioration of lipid accumulation [106]. Additionally anthocyanins from purple sweet potato decrease SREBP1c in the same in vitro model but also in vivo in HDF-mice [107]. Musso et al. reviewed cellular mechanisms of cholesterol toxicity involved in liver injury and NASH and highlighted the therapeutic impact of anthocyanin through a decrease SREBP2 and lipogenesis [104]. So, all these data highlight the fact that the cherry is responsible for an improvement of hepatic complications associated with a decrease of oxidative stress and inflammation.

Our recent data showed that the consumption of cherry without any metabolic disorders, so on healthy rats, leads to opposite effects. For example, hepatic p22phox expression was increased, leading to oxidative stress and associated to hepatic dysfunction (unpublished data). These results highlight the difficulty to work with antioxidant compounds, which could be pro-oxidant sometime. We worked also with cherry which contained fructose as the major source of sugar; future study could be done with free-sugar cherry extract, but only for the purpose of a consumption of modified food and not the promotion of the consumption of natural healthy food. Identifying specific polyphenolic compound(s) in cherry extract leading to the beneficial effect could be also a strategy, but as highlighted by Snyder et al. [108] ‘a challenge for future research is not only to describe the improvements produced by the intake of specific healthful foods or phytochemicals, but also to determine what beneficial synergies may be produced by consuming complementary healthy foods containing a variety of bioactive compounds, acting on multiple and molecular-level regulatory pathways’. Several studies in animal models and human subjects have demonstrated that phenols are bioavailable and exert a protective role against oxidative stress and free radical damage [30, 40, 82]. Moreover epidemiological studies suggest that consumption of fruits, vegetables and plants [30] may be associated with a reduced risk of diabetes or have a protective effect [109]. Recently, Pickering et al. [10] clearly reviewed the feasibility of emerging new therapies to combat oxidative stress and inflammation in the diabetic milieu. The use of therapy like cherry brings a real asset thanks to its broad-spectrum effects on: (1) the regulation of carbohydrate and lipid metabolisms, (2) the attenuation of oxidative damage and scavenging of free radicals, (3) the improvement of endothelial function and vascular tone through the enhancement vasodilation factor production, and (4) the decrease of NASH with macrophages and ROS inhibition. All the bibliography available today on the subject brings hope on using antioxidants in future hepatitis and antidiabetic therapeutics.