Background
Obesity is defined as an increase of adipose tissue mass in the body and its accumulation in peripheral organs that leads to metabolic abnormalities such as type 2 diabetes (T2D), insulin resistance and hyperlepidemia [
1,
2]. Obesity is thus a worldwide healthcare problem increasing morbid-mortality [
2].
The visceral adipose tissue plays an important role in the regulation of postprandial lipid and glucose systemic homeostasis by targeting essential organs (adipose tissue, muscle, etc.) and systems (neuroendocrine axis, etc.) [
3‐
6]. It is considered as a secretory gland source of several bioactive peptides called adipokines [
6]. Thus, adipose tissue becomes a major protagonist of metabolic alterations triggered by lipid over-accumulation in their cytoplasmic droplet which leads to adipocytes function disorders [
7,
8]. An alteration in the adipokine secretion profile leads to insulin resistance, glucose intolerance and lipid metabolic disturbances [
8‐
10].
Consumption of High-fat diets (HFD) is a central risk factor for metabolic disorders linked to obesity [
9,
11]. Adverse effects of HFD on metabolic homeostasis are linked to adipose tissue physiology and are highly influenced by gender [
12,
13]. The imbalance between caloric intake and energy expenditure leads to hyperplasia and hypertrophy of adipocytes depending on the type as much as the amount of dietary lipids [
14,
15].
Numerous factors regulate the adipose tissue activity including adipocyte-specific genes such as peroxisome proliferator-activated receptors gamma (PPARγ) [
16]. Dietary fatty acids (FA) and their derivatives are described as PPARγ ligands that trigger physiological responses such as adipogenesis and adipokine secretion. Thus, HFD-inducing metabolic disorders act via PPARγ to induce different levels of systemic homeostatic remodelling.
Appreciating the pathogenesis of HFD-induced metabolic disorders requires a thorough knowledge of adipose tissue physiology and the regulation of adipokines secretion and action including the role of gender in response to these parameters.
Thus, significant progress has been made in our understanding of the relation between HFD feeding and adipose tissue dysfunction. However, experimental evidence for HFD-mediating metabolic alterations remains to be elucidated. Therefore, we propose a possible role for HFDs' fat nature on the establishment of metabolic disorders. Herein, we evaluated the impact of FA nature on the development of metabolic alterations through modulation of adipose tissue activity and secretion profiles. We also determined gender-specific impact on the kinetic of metabolic disorders progression in response to these diets.
Methods
Experimental protocol
The animal protocol was approved by the Animal Care and Use Committee of the Montreal Heart Institute conforming to the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Three groups of 10 male and 10 female C57BL/6J mice (Jackson Laboratory, Bar Harbor, MN, USA) were included in this study at 5 weeks of age. Each group was fed either standard diet (SD) used as control (6% fat, 57% sucrose) or one of the two low cholesterol HFD (34.9% fat, 26.3% sucrose) (detailed composition presented in Table
1). These latter were iso-caloric but differed in fat nature: VD was composed of soy and cotton oil while AD was composed of lard. Weight gain was monitored during the 20 weeks of protocol and daily food consumption was calculated by subtracting the residual quantity from the supplemented food quantity each day. Energy intake was calculated on the basis of 3.8 kcal/g for the SD and 5.2 kcal/g for both HFD.
Table 1
Analysis of diets compositions
Sucrose
| 26,3% | 26,3% | 57% |
Protein
| 26,2% | 26,2% | 18,9% |
Lipid
| 34,9% | 34,9% | 6% |
FA composition:
|
g/100 g of diet
|
C14 Myristic | 0,2 | 0,9 | 0,006 |
C14:1 Myristoleic | 0 | 0,5 | 0 |
C16 Palmitic | 13,0 | 21,8 | 0,764 |
C16:1 Palmitoleic | 0,2 | 3,8 | 0 |
C18 Stearic | 10,2 | 12,4 | 0,15 |
C18:1 Oleic | 62,0 | 39,3 | 1,26 |
C18:2 Linoleic | 12,9 | 12,8 | 3,13 |
C18:3 Linolenic | 0,7 | 1,6 | 0,28 |
C20 Arachidic | 0,3 | 0 | 0,01 |
C20:4 Arachidonic | 0 | 1,7 | 0 |
C22 Behenic | 0,3 | 0 | 0,003 |
C24 Lignoceric | 0,1 | 0 | 0 |
Cholesterol | 0,035 | 0,030 | |
FA proportions:
|
Percentage of total FA
|
Saturated FA | 24,2% | 37,0% | 16,96% |
Mono-unsaturated FA | 62,2% | 46,0% | 22,71% |
Poly-unsaturated FA | 13,6% | 17,0% | 60,33% |
Intra-Peritoneal glucose tolerance test
The Intra-Peritoneal Glucose Tolerance Test (IPGTT) was performed after 20 weeks of diet following overnight (18 hr) fasting. After measuring the fasting glycæmia using an Accu-Check® (Roche Diagnostics, Laval, QC, Canada) glucometer, mice received an intra-peritoneal injection of glucose solution (2 g/kg). Subsequent measures of plasma glucose levels were performed 30, 60, 90, 120 and 180 minutes post-injection. The area under the glucose tolerance curve (AUC) was measured to evaluate mice glucose tolerance.
Physiological analyses
Non-invasive measurement of mice cardiac function was performed by Doppler echocardiography under anesthesia (isoflurane 2.5%). The velocities of the early mitral flow (E) over the late mitral flow (A) were measured with a Doppler Signal Processing Workstation (GE-Ultrasound System). An increase in the E/A ratio, with restrictive aspect of transmitral flow (E/A > 2) indicated a diastolic dysfunction with increased LV filling pressure. To calculate ventricular mass, left ventricular (LV) tele-diastolic diameter, posterior wall and inter-ventricular septum thickness were measured.
Biochemical analyses
At sacrifice, after an overnight fast (18 hr), blood samples were gathered by cardiac exsanguinations and plasma was collected and stored at -20°C until analysis. Plasma adiponectin and leptin concentrations were measured using mouse ELISA kits (ALPCO, Salem, NH, USA) according to the manufacturer's instructions. Plasma insulin, resistin and TNFα levels were analyzed using mouse ELISA kits (AssayPro, St. Charles, MO, USA) according to the manufacturer's instructions. All measurements were analyzed in duplicate for at least 8 animals per group. The circulating concentrations of FA in mice serum were evaluated using ADIFAB free FA indicator (Invitrogen, Burlington, ON, Canada).
Homeostasis model assessment of insulin resistance (HOMA-IR)
Insulin resistance was assessed by calculation of HOMA-IR using glucose and insulin concentrations obtained after overnight fasting (18 hr), using the following formula:
HOMA-IR is known to be correlated with the insulin sensitivity evaluated by the euglycemic hyperinsulinemic clamp [
17,
18].
Quantitative real-time PCR (Q-PCR)
At sacrifice, mouse adipose tissue was removed from the abdominal visceral region weighed to evaluate its accumulation level, and then frozen until mRNA extraction. Total RNA was isolated from approximately 30 mg of frozen white adipose tissue of the abdominal visceral region using Qiazol reagent according to the manufacturer's instructions (Qiagen, Toronto, ON, Canada). Single-strand cDNA was synthesized according to the procedure in the iScript cDNA Synthesis Kit manual (Bio-Rad Laboratories, Montreal, QC, Canada). Q-PCR reactions were carried out using the Brilliant-II SYBR® Green Master-Mix (Stratagene, Mississauga, ON, Canada) and specific primers:
-
Adiponectin primers (#GenBank: NM_009605.4):
-
Leptin primers (#GenBank: NM_008493.3):
-
TNFα primers (#GenBank: NM_013693.2):
-
PPARγ2 primers (#GenBank: NM_011146.3):
-
ERα primers (#GenBank: NM_007956.4):
The mRNA levels were normalized to Cyclophilin-A expression levels (Fwd 5'-CCG-ATG-ACG-AGC-CCT-TGG-3'; Rev 5'-GCC-GCC-AGT-GCC-ATT-ATG-3'). The targeted and referenced genes were amplified in duplicate in the same run using the Mx3000P Q-PCR System (Stratagene).
The relative quantification of target genes was determined using the MxProTM Q-PCR software version 3.00 (Strategene). Briefly, Ct average of each duplicate was calculated for each gene and Cyclophilin-A and the ΔCT (CTgene - CTCyclo-A) was determined. The control adipose tissue sample was chosen as a reference sample and set as 100% of gene quantity. Finally, the mRNA abundance of other samples to the mRNA abundance of the control adipose tissue was calculated with use of the formula 2-ΔΔCT.
Statistical Analyses
All statistical analyses were performed separately for males and females.
Data are presented as mean ± standard deviation for continuous variables. Repeated measures analysis of covariance (ANCOVA) model was used to study weight gain across the study time and between groups (SD, VD and AD groups) adjusted for baseline weight value. The group × time interaction was also included in the ANCOVA model and comparisons between groups at a given time point were undertaken only in the presence of a significant group × time interaction. Otherwise, global conclusions were drawn based on the main time and group effects of the model.
Repeated measures analysis of variance (ANOVA) models were used to study the glycemic parameters across the IPGTT time (0, 30, 60, 90, 120 and 180 min) and between groups. Models with time, group and group × time interaction as independent variables were used. Comparisons between groups at a given time-point were undertaken only in the presence of significant group × time interaction. Otherwise, global conclusions were drawn based on the main time and group effects of the model.
The insulin, adiponectin, leptin, TNFα and FA levels were compared between groups (SD, VD and AD groups) using an ANOVA model.
In addition, the relationships among weight gain, cardiac parameters and metabolic parameters were investigated using Pearson or Spearman correlations according to the nature of the data distribution.
All analyses were done using SAS version 9.1 (SAS Institute Inc., Cary, NC, USA) and conducted at the 0.05 significance level.
Discussion
HFD-induced obesity has become widely accepted as a key factor of alteration in insulin sensitivity and metabolism [
9,
11]. However, the physiological regulation and role of HFD in mediating the unhealthy effects of increased adiposity remain not fully elucidated.
The novelty of this study is the use of two types of HFD identical in their lipid proportion but different in their fat nature. These diets were low in cholesterol in order to reduce its implication in metabolic alterations and mimic industrially-produced popular fast food known as obesity and T2D inducers [
9,
11,
19]. Both HFD promoted more weight gain than the control SD with a faster increase in males compared to females. However, the degrees of metabolic alterations differed considerably between the two HFD and were highly influenced by gender. This was not a consequence of the daily caloric uptake but was rather due to the dietary fat nature and mice gender.
For a comparable weight gain in HFD groups, males and females under AD showed a more pronounced accumulation of adipose tissue in the abdominal visceral region compared to corresponding VD groups. Abdominal visceral adipose tissue mass was slightly increased in VD group compared to control but remained non significant when reported as percentage of total body weight. Thus, despite its implication in increasing weight gain, a VD promotes a fat mass distribution different than an AD in both genders. In fact, it is well established that mono-unsaturated FA are more accumulated in the subcutaneous region and prevent central fatty acid accumulation [
20]. Furthermore, it was reported that postprandial fat oxidation as well as a meal thermic effect was higher with mono-unsaturated FA rich diet compared to a saturated FA rich diet [
21]. The increased thermic effect could reflect a pronounced storage activity of the subcutaneous adipose mass. Thus, we can speculate that our mono-unsaturated FA rich diet trigger a different adipose tissue accumulation than the saturated FA rich AD. However, more experiences are needed to confirm this hypothesis.
This differential fat accumulation plays a crucial role in metabolic alterations development. A link was established between HOMA-IR increase and the visceral adipose tissue accumulation in AD-fed groups. These results support previous findings that demonstrate a correlation between visceral abdominal adipose tissue accumulation and metabolic alterations; a correlation not reported for subcutaneous fat accumulation [
22,
23]. High concentrations of saturated FA such as in AD are associated with lipotoxicity effects (pancreas, liver, muscle, adipose tissue...) and alter cell membrane dynamics [
24]. This prevents the dimerization of cell surface receptors such as insulin and leptin receptors and inhibits the signalization of the corresponding hormone. Therefore, the increased levels of insulin and leptin in our model could reflect a defect in their respective signalling pathway. In addition, increased insulin level and HOMA-IR are obtained simultaneously with enhanced resistin levels in AD groups. Resistin secretion in adipose tissue is known to induce insulin resistance through the inhibition of its signalling pathway and the stimulation of hepatic glucose production [
25]. These data could explain the fact that AD mice developed increased insulin circulating levels. The VD, rich in unsaturated FA, did not affect insulin levels nor modulates resistin levels in corresponding groups as reported in a previous study with unsaturated oleic acid [
26]. So, at similar circulating levels, different combinations of FA promote distinctive alterations in the systemic glucose homeostasis.
Female mice remained normo-glycemic under HFD confirming less advancement in metabolic alterations compared to males. This female-specific resistance to T2D development was even more striking with VD compared to AD, which induced hyper-insulinæmia and slight glucose intolerance. However, VD as well as AD induced T2D with an increased fasting glycæmia and glucose intolerance in male mice and that starting the 12
th week of HFD feeding. These alterations could be correlated with adipocytes death demonstrated in HFD-fed male mice [
27]. This parameter was however never been investigated in HFD-fed female mice. On the other hand, our preliminary study didn't show any increase in the glycaemia of the HFD-fed female mice even after 16 weeks of diet. We have thus extended our protocol to 20 weeks. It is well established that females show less extensive metabolic alterations due to the oestrogen action that protect adipocytes in female from insulin resistance and inflammation [
12,
13,
28]. In our model, we have demonstrated that females had a higher level of ERα mRNA in their adipocytes. Its mRNA expression level could reflect a higher activity of ERα in corresponding tissue. Furthermore, recent evidence shows that FA accumulation is more oriented to the visceral fat pad in male while it is more directed to the peripheral region [
29]. Thus, since android fat accumulation is more associated to metabolic alterations development [
22], these observations could explain the prevention of T2D development in female groups.
Both HFD increased leptin levels in male and female mice with a significant difference between VD and AD groups. Correlated with body weight increase, high circulating leptin levels, known to improve glucose homeostasis [
30], suggests leptin resistance in obese mice. Leptin resistance affects glucose homeostasis and contributes directly to hyperglycaemia [
30,
31]. Male mice showed increased leptin levels concurrently with increased AUC values of the IPGTT. In female groups, the apparent lack of negative leptin action on glycaemia could be due to a counterbalanced effect by the high adiponectin levels compared to male groups. Adiponectin, an insulin-sensitizer adipokine and an inhibitor of hepatic glucose production, contributes to improved glucose homeostasis [
32,
33].
To evaluate the direct impact of diets on adipose tissue, additional analyses were performed at the mRNA level. TNFα is a key regulator of the adipogenesis that decreases insulin sensitivity and promotes free FA production by stimulating lipolysis and inhibiting the antilipolytic effects of insulin [
34,
35]. In our model, TNFα mRNA levels in visceral abdominal adipose tissue were increased in males under AD and, to a much lesser extent, in VD-fed males. This effect could be due to enhanced macrophage infiltration in the adipose tissue [
27,
35,
36]. In the AD group, it was associated with leptin and resistin circulating levels known to augment TNFα production [
37,
38]. In the VD group, oleic acid may have contributed to limit TNFα up-regulation. This FA was found to be effective in reversing the inflammatory status in adipose tissue responsible for decreased insulin sensitivity [
36]. In the AD female group, despite their increased levels of leptin and resistin, TNFα mRNA levels remained unchanged in the visceral adipose tissue. A high level of circulating adiponectin is known to inhibit TNFα effects [
39,
40]. Adiponectinemia in HFD-fed group was reduced in female mice compared to control. However, these levels of adiponectin seem to still be enough to prevent TNFα mRNA up regulation in the adipose tissue. Moreover, it could also have contributed to the delay of T2D development regardless of high leptin levels [
40]. However, decreased mRNA expression levels of adiponectin in the adipose tissue of females under HFD could be the first sign of progression toward a metabolic alteration cascade.
Thus, adiponectin modulation seems to play a critical role in our model and could contribute to differences between males and females, with delayed metabolic alterations in the latter.
Hyperplasia and hypertrophy of adipocytes are involved in obesity [
41]. The balance between the two processes is controlled by FA nature that governs nuclear receptors activity [
42]. Therefore, inhibition of adipogenesis through high levels of circulating FA triggers adipocytes hypertrophy and leads to insulin resistance [
43], a situation encountered in the AD mice group. In contrast with the oleic-rich VD-fed groups, metabolic alterations were less extensive with an absence of hyper-insulinæmia. In this case, adipocyte hypertrophy and hyperplasia occurred. Hyperplasic obesity is accepted as being less harmful regarding metabolic alterations and adipose tissue inflammation [
35,
36].
Dietary FA modulate PPARγ activity which controls adipokines secretion [
43,
44]. Superior adipogenesis potential in white adipose tissue of the abdominal visceral region of females over males is suggested by the pattern of expression levels of the PPARγ2 gene. Our results showed decreased levels of PPARγ2 mRNA in HFD groups and remained higher in females compared to males. The highest levels were found in the VD female group protected against HFD-mediated negative metabolic impacts. These results support the hypothesis about the role of PPARγ2 in preventing adipocytes hypertrophy that leads to decrease adiponectin circulating levels and insulin sensitivity.
In previous studies, crosstalk between the estrogens receptors and PPARγ regulatory pathways has been demonstrated with sex hormone regulation of adipokine production [
16,
42]. Expression of ERα, a main mediator of oestrogen effects, was investigated as it could be involved in the gender regulation and/or diet-specific response of adipose tissue. In this study, ERα mRNA levels in adipocytes were decreased with HFD but remained higher in female groups. Such higher expression levels of ERα in females may favour PPARγ2 mRNA expression under HFD when compared to males [
16].
Signs of cardiac diastolic dysfunction linked to weight gain and metabolic alterations were also detected in the HFD groups. An increase of the E/A ratio, with restrictive aspects of transmitral flow (E/A > 2), indicated a diastolic dysfunction with increased LV filling pressure. A significant elevation of the E/A ratio occurred in males fed with the VD or AD (63% and 59%, respectively) and a 90% increase in the AD female group only. Females on VD had no modification of the E/A ratio despite being overweight. Obesity can be associated with impaired LV diastolic function [
45] though the exact reason is still unclear. Leptin regulation of the hypothalamic-pituitary-adrenal axis responsible for blood pressure regulation [
31], is disturbed in obese subjects. Thus, high leptin concentration leads to diastolic dysfunction associated with higher cardiac sympatic nervous system activity and increased LV mass [
45]. This dysfunction with a reduction in cardiac compliance was associated with LV dilatation and an increased LV mass in HFD groups.
Hyperglycaemia and hyper-insulinæmia have also been suggested to be additive stimuli to LV hypertrophy [
46]. Thus, the fasting hyperglycaemia in male mice under HFD and the elevated level of circulating insulin in mice under AD could have aggravated cardiac hypertrophy and alteration. In fact, in glucose diabetic and insulin resistant mice, the myocardium consumes more FA to produce energy leading to more LV hypertrophy. Female mice under VD increased their LV mass without diastolic dysfunction. The normal glycemic and insulin rates in the presence of higher adiponectin concentration compared to the male group contributed to maintain normal diastolic function. Furthermore, the VD contains 27% more oleic acid, which is known to prevent cardiac dysfunctions [
47], than the AD. These differences in diet composition could explain the prevention of diastolic dysfunction in the female VD group but not in the female AD group.
Metabolic alterations development is different between males and females. However, the majority of studies on HFD-induced metabolic disorders are restricted to males [
48]. The strength of this work was therefore the evaluation of the respective sensitivity of both sexes to two types of low-cholesterol HFD, differing in their dietary FA nature. Thus, we have shown that combinations of FA have gender-related effects on visceral fat distribution and metabolic consequences. Therefore, susceptibility to develop HFD-linked T2D is strongly reflected by sex hormone-associated modulation of adiponectin production, TNFα and PPARγ regulation in visceral adipose tissue. However, further investigation will be required to explain the differences in response to the two types of HFD among females.
Understanding gender-specific adipose tissue adaptations underlying metabolic disorders linked to HFD and unhealthy lifestyles will considerably contribute to the development of improved strategies for the prevention and treatment of metabolic and cardiovascular diseases.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
SEA and VL carried out the physiological follow up of the mice and assays in vivo. Molecular and biochemical studies and the drafting of the manuscript were completed by SEA. IC as well was JFT participated in the design and the coordination of the study and helped to draft the manuscript. All authors read and approved the final manuscript.