Isolated and combined impact of dietary olive oil and exercise on markers of health and energy metabolism in female mice

Abstract

An olive oil (OO) rich diet or high-intensity interval training (HIIT) independently improve markers of health and energy metabolism, but it is unknown if combining OO and HIIT synergize to improve these markers. This study characterized the isolated and combined impact of OO and HIIT on markers of health and energy metabolism in various tissues in C57BL/6J female mice. Nine-week-old mice were divided into four groups for a 12-week diet and/or exercise intervention including: 1) Control Diet without HIIT (CD), 2) Control Diet with HIIT (CD+HIIT), 3) OO diet (10% kcal from olive oil) without HIIT, and 4) OO diet with HIIT (OO+HIIT). Neither dietary OO or HIIT altered body weight, glucose tolerance, or serum lipids. HIIT, regardless of diet, increased aerobic capacity and HDL cholesterol levels. In liver and heart tissue, OO resulted in similar adaptations as HIIT including increased mitochondrial content and fatty acid oxidation but combining OO with HIIT did not augment these effects. In skeletal muscle, HIIT increased mitochondrial content in type II fibers similarly between diets. An RNA sequencing analysis on type I fibers revealed OO reduced muscle regeneration and lipid metabolism gene abundance, whereas HIIT increased the abundance of these genes, independent of diet. HIIT training, independent of diet, induced sWAT hypertrophy, whereas OO induced gWAT hypertrophy, an effect that was augmented with HIIT. These data highlight the pleiotropic effects of OO and HIIT, although their combination does not synergize to further improve most markers of health and energy metabolism.

1. Introduction

The Mediterranean diet is a lifestyle intervention that can improve cardiometabolic risk factors associated with chronic disease [1]. A staple of the Mediterranean diet is olive oil (OO), and consumption of OO improves insulin sensitivity [2], particularly in women [3], while also reducing inflammation [4] and the incidence of major cardiovascular events [5, 6]. Dietary OO has tissue specific effects that contribute to these health benefits, including the ability to alleviate β-cell dysfunction [2], NAFLD during high-fat diet feeding [7, 8], high-fat diet induced inflammation and ROS [7, 9], and postmeal dyslipidemia [10]. Dietary OO consists predominately (>70%) of oleic acid (18:1 cis-9), a monounsaturated fatty acid, and one mechanism contributing to the beneficial effects of OO may be due to the fact that oleate, compared to palmitate, promotes greater increases in mitochondrial biogenesis and fatty acid oxidation [11–15]. Moreover, recent data from our lab has shown that oleate derived from lipid droplets allosterically activates SIRT1 and PGC-1α to promote mitochondrial metabolism and health [14].

In addition to dietary interventions, exercise training also improves cardiometabolic risk factors [16] and promotes fatty acid metabolism [17]. For instance, aerobic exercise training, particularly high-intensity interval training (HIIT), promotes fatty acid oxidation and increases mitochondrial content in a myriad of tissues [18–22]. Similar to the Mediterranean diet, one of the mechanisms by which exercise promotes health and mitochondrial metabolism is through SIRT1 and PGC-1α signaling in various tissues [23, 24].

Although the isolated impact of either dietary OO or HIIT on markers of metabolic health and mitochondrial metabolism is established, it is not clear if combining dietary OO with HIIT would synergize to further improve markers of health and mitochondrial metabolism in various tissues. Furthermore, there is evidence that sex modifies the response to OO or exercise. For instance, one study reported higher oleate intake was associated with greater insulin sensitivity in women compared to men, whereas another study found only men experienced improved insulin sensitivity in response to the Mediterranean diet [3, 25]. Moreover, women typically have higher intramyocellular lipid numbers, higher mitochondrial volume, and oxidize more fat during exercise compared to men [26, 27]. Given these previous findings on the influence of sex in response to OO or exercise, there remains a gender disparity in the literature as the majority of studies use males as the population of choice [28, 29]. To address these gaps in knowledge, the objective of this study was to use a female murine model to characterize the isolated and combined effect of dietary OO and HIIT on markers of metabolic health and energy metabolism in various tissues.

2. Materials and Methods

2.1. Mouse Groups and Diet

All animal protocols used in this study were approved by the Institutional Animal Care and Use Committee at the University of Minnesota. Four to five-week-old C57Bl/6J mice were purchased from Jackson Labs and group housed in cages located in a temperature (20–22° C) and light (12 h light and dark cycle) controlled animal facility. Female mice were used in this study because 1) females are underrepresented in biomedical research [29], and 2) there is a significantly higher prevalence of the metabolic syndrome in women compared to men [30, 31], thus there is a need to better understand how various lifestyle modalities impact metabolism and health in females, an underrepresented group. The dark cycle started at 12 p.m. so that mice could perform high-intensity interval exercise (HIIT) during the dark cycle (details described below) at a time of the day that was convenient for the investigators. Mice consumed standard rodent chow until 8 weeks of age when they were switched to their assigned purified diet, which they consumed for the remainder of the study. The diets consisted of a Teklad (Envigo) custom made control diet (TD.170820) or olive oil diet (TD.170821) with each diet containing identical energy from protein (18% kcal), carbohydrate (67% kcal), and fat (14.9% kcal). The control diet contained 9% soybean oil, 5% lard, and 1% fish oil while the olive oil diet contained 4% soybean oil, 10% OO, and 1% fish oil. We chose a 10% OO diet as previous work from our lab, albeit in male mice, has shown this dose increases hepatic mitochondrial content [14]. Additionally, at nine weeks of age mice were randomized to a 12-week treadmill exercise regimen (described below). Thus, there were a total of four groups including: 1) Control diet without HIIT (CD, n = 12), 2) Control diet with HIIT (CD+HIIT, n = 11), 3) Olive oil (OO) diet without HIIT (OO, n = 11), and 4) OO diet with HIIT (OO+HIIT, n = 11).

2.2. High-Intensity Interval Exercise Training

Given that mice are nocturnal and primarily active during the dark cycle, the mice performed the treadmill exercise sessions beginning ~1 h after the dark cycle started to avoid disrupting their normal circadian rhythms. The investigators used a red light to allow for visualization of training sessions. At eight weeks of age all mice were acclimated to the treadmill (1055 M Exer-6M Open Treadmill for Mice with Shock, Columbus Instruments) by being placed on the treadmill and walking/running between 0.6 – 14 m/min for 5–10 min per day for 5 straight days. The following week the mice were randomly assigned to either an exercise group or sedentary group. Mice were fasted for ~1 h prior to exercise while food was given back to the animals after exercise. Control mice were fasted for the same duration. The exercise training involved a HIIT protocol that was based on previously published protocols [32–34] and consisted of a high intensity interval (2 minutes: 14 m/min and was increased by 1 m/min every week until the end of the study) immediately followed by a low intensity interval (2 minutes: 0.6 m/min), and this cycle was repeated for 1 h. Mice were encouraged to run by using a shock grid. The mice exercised 4 d per week on Monday, Tuesday, Thursday, and Friday and at the same time of day for 12 weeks. The mice were sacrificed 48 h after the last exercise session.

2.3. Aerobic Capacity Test

Aerobic capacity was determined with a progressive treadmill test where exercise started at 10 m/min for 5 min and thereafter the speed was increased by 4 m/min every 4 min until the mice were exhausted (defined as inability to stay off the shock grid). This test was performed during week 10 of the diet and exercise program, after a ~1 h fast, and ~1 h after the start of the dark cycle. Mice in the exercise groups performed this test 24 h after the last exercise session, and this test served as their exercise session for that day.

2.4. Oral Glucose Tolerance Test and Serum Lactate

All mice underwent oral glucose tolerance testing during week 11 of the diet and exercise program. After a 4 h fast, mice were orally gavaged with glucose (3 mg/g body weight) dissolved in sterile water. Tail blood samples were taken at baseline and 15, 30, 45, and 60 min after the oral gavage and glucose was measured with a glucometer. Fasting serum lactate levels were measured with a commercially available colorimetric assay (Pointe Scientific, Inc., Canton, MI). The coefficient of variation (CV) for glucose with the glucometer used in our lab is <3%. The CV for serum lactate was 3.9%.

2.5. Serum Lipids

Mice were anesthetized and a cardiac stick was performed to collect ~4 h fasting blood samples prior to the mice being sacrificed. Blood was allowed to clot on ice and then was spun at 5,000 × g for 5 min and serum was collected. Serum triglyceride (Stanbio Laboratory, Boerne, TX), ketone bodies (FUJIFILM Wako Diagnostic U.S.A. Corporation, Mountain View, CA), NEFA (Wako), and cholesterol (Thermo Fisher Scientific Inc., Middletown, VA) were all measured with commercially available colorimetric assays. The CV for triglyceride, ketone bodies, NEFA, total cholesterol, and HDL cholesterol was 4.0%, 7.7%, 5.7%, 4.9%, and 3.7%, respectively.

2.6. H&E Staining

Fresh tissue samples harvested from the animals were immediately placed in 10% formalin and fixed for 24 h. Following fixation, the samples were submerged in a 70% ethanol solution for at least 24 h prior to subsequent paraffin embedding for hematoxylin and eosin (H&E) staining. Images were taken with a Leica Digital Microscope, and lipid droplet size was quantified using ImageJ software.

2.7. Tissue Triacylglycerol Content

A piece of frozen sample was obtained and weighed, and lipids were extracted using the chloroform:methanol method. Lipids were dried down using nitrogen and subsequently resuspended into isopropanol with 1% Triton-X. Triglyceride levels were measured using a commercially available colorimetric assay (Stanbio Laboratory, Boerne, TX) and then normalized to tissue weight.

2.8. Western Blotting

Tissues were homogenized in lysis buffer and the protein concentration was assessed using a commercially available kit (ThermoFisher, Waltham, MA), with a CV of 4.8%. The primary antibodies SIRT1 (3931S), ATP-citrate lyase (4332S), Cox4 (4844S), and UCP2 (89326S) were purchased from Cell Signaling Technology (Danvers, MA). The rodent OxPhos (ab110413) primary antibody cocktail was purchased from Abcam (Cambridge, MA). All primary antibodies were diluted 1:1000. The appropriate LI-COR (Lincoln, NE) secondary antibodies (diluted 1:5000) were used with a LI-COR Odyssey Fc imaging system to detect fluorescent bands on each blot.

2.9. qRT-PCR and RNA Sequencing

RNA was isolated with Trizol reagent and cDNA was synthesized using BioRAD’s Iscript kit and thermal cycler. Quantitative real time PCR with SYBR green select reagent was used to detect variations in gene abundance with select mouse primers. RNA from samples used for RNA sequencing was purified using a Qiagen ® RNeasy Mini Kit. The University of Minnesota Genomics Center performed the RNA sequencing analysis using TruSeq Stranded mRNA LP, NovaSeq 150 PE, with 20M reads per sample.

2.10. Oleate Oxidation Assays

A protocol described previously, with minor modification, was adapted herein to measure fatty acid oxidation with [1-¹⁴C]oleate (Perkin Elmer, Cat # NEC317250UC) [35]. Briefly, freshly isolated quadriceps muscle was homogenized with a Teflon pestle in ice-cold STE (0.25 M sucrose, 10 mM Tris-HCL, 1 mM EDTA, pH 7.4) buffer. The homogenate was then centrifuged at 420 × g for 10 min at 4° C, and the supernatant containing mitochondria was used in the oleate oxidation assays. 30 μl of supernatant was added to 370 μl of the oxidation reaction mixture (final concentration: 100 mM sucrose, 10 mM Tris-HCL at pH 7.4, 5 mM KH₂PO₄, 0.2 mM EDTA at pH 8.0, 90 mM KCL, 1 mM MgCl₂, 2 mM L-carnitine, 0.1 mM malate, 0.05 mM Coenzyme A, 2 mM ATP, 1 mM DTT, 0.7% BSA, 250 μM cold oleate, 0.4 μCi / ml 1-¹⁴C oleate) at 37° C for ~1 h. Next, 200 μl of 1 M perchloric acid was added to the tube to release [¹⁴C]CO₂, which was collected in filter paper containing 1 M NaOH at the top of the tube. The radioactivity of the filter paper (complete fatty acid oxidation) and acid soluble metabolites (incomplete fatty acid oxidation) was measured using scintillation counting.

2.11. Citrate Synthase Activity

The activity of citrate synthase within tissue homogenates was determined using a commercially available kit (Cayman Chemical MitoCheck, Cat# 701040). Briefly, tissues were homogenized, the samples and standards were loaded into a 96 well plate, and the absorbance of the Citrate Synthase Developing Reagent was measured at 412 nm every ~1 min for 20 min.

2.12. Statistical Analysis

The statistical analyses were performed using GraphPad Prism 8.3.0. Two-way ANOVA was used to test for significant differences, and for significant main effects and interaction effects a follow-up Tukey post-hoc test was used to test for significant differences between groups. Statistical significance was set at P ≤ 0.05. For the RNA-sequencing analysis, soleus muscle group samples were compared using EdgeR Software Package in R version 4.0.2. First, genes were filtered such that only genes that have a minimum normalized count of 10 across 70% of samples in a group were kept. Sample counts were normalized in EdgeR using the trimmed mean of M-values method, and the EdgeR exact Test was performed to test for differences between groups. Genes were considered significantly different between comparisons if a FDR < 0.05 and LOG₂ fold change > 1X was obtained.

3. Results

3.1. Neither dietary olive oil or HIIT training modifies body weight, but HIIT, regardless of diet, increases aerobic capacity and HDL cholesterol levels.

Body and tissue weight, glucose tolerance, aerobic capacity, and serum metabolite concentrations are markers that can be used to gauge metabolic health. Herein, dietary OO, HIIT, or their combination did not significantly (P > 0.05) modify body weight gain (Figure 1A–B), the relative weight of various muscles including the heart, soleus, and gastrocnemius (gastroc) muscles (Figure 1C), or the relative weight of other tissues including the liver, gonadal white adipose tissue (gWAT), and subcutaneous WAT (sWAT) (Figure 1D) (P > 0.05). The HIIT protocol increased aerobic capacity 83% (P < 0.001) in the CD+HIIT group and 88% (P < 0.001) in the OO+HIIT group compared to the CD and OO groups, respectively (Figure 1E). Glucose tolerance was not significantly different (P > 0.05) between groups (Figure 1F–G). Likewise, the concentrations of various serum metabolites including lactate, triglycerides, NEFA, and ketone bodies were not significantly different between groups (Figure 1H–K, all P > 0.05). The CD+HIIT and OO+HIIT groups had 25% (P = 0.009) and 19% (P = 0.033) higher total cholesterol levels compared to the CD and OO groups, respectively. This effect of HIIT on cholesterol was due to an increase in HDL cholesterol of 23% (P = 0.036) and 21% (P = 0.047) in the CD+HIIT group and OO+HIIT groups compared to the CD and OO groups, respectively (Figure 1L).

Figure 1. Body weight and metabolic characteristics.

A) Body weight time course of each group during the 12-week diet and/or HIIT intervention. B) Body weight gain calculated as difference between body weight at week 12 and week 1. C-D) Various tissue weights relative to total body mass (expressed as % of body weight). E) Time to exhaustion during a progressive aerobic capacity test. F-G) Glucose time course (F) and incremental area under the curve (G) during an oral glucose tolerance test. H) Serum fasting lactate levels. I) Serum fasting triglyceride levels. J) Serum non-esterified fatty acid levels. K) Serum fasting total ketone body levels. L) Serum fasting total, HDL, and LDL cholesterol levels. a: P < 0.001, b: P < 0.001, c: P = 0.003, d: P < 0.001, e: P = 0.004, f: P < 0.001, g: P = 0.044, h: P = 0.001, i: P < 0.001, j: P < 0.001.

3.2. Dietary olive oil induces hepatic adaptations that are similar to HIIT training.

H&E staining revealed that lipid droplet size within the liver was greater in the OO (P < 0001), CD+HIIT (P < 0.001), and OO+HIIT groups (P < 0.001) compared to the CD group, suggesting OO and HIIT have a similar impact on liver lipid levels but their combination is not additive (Figure 2A–B). To confirm these observations, in a subset of samples liver lipids were extracted and triglyceride levels were measured, which revealed a similar trend where the OO (P = 0.047), CD+HIIT (P < 0.001), and OO+HIIT groups (P = 0.006) had higher triglyceride levels compared to the CD group (Figure 2C). These alterations in triglyceride levels were not associated with changes in markers of lipogenesis, as there was no difference in the protein abundance of ATP-citrate lyase (ATP-CL) or mRNA abundance of Fas between groups (Figure 2D–F). However, both the OO diet and HIIT had robust effects on mitochondria. Specifically, the OO diet increased markers of mitochondrial content including COX 4 (P = 0.046), complex V (P = 0.003), complex IV (P= 0.004), complex II (P = 0.045), and complex I (P < 0.001) protein abundance compared to the CD group and to a similar level as that observed within both HIIT treatment groups (Figure 2D–E). Additionally, the mRNA abundance of Pnpla2 (adipose triglyceride lipase), a marker of lipolysis, was higher in the CD+HIIT (P = 0.002) and OO+HIIT (P = 0.015) groups compared to the CD and OO groups, respectively. Similarly, Cpt1a mRNA abundance, a marker of fatty acid oxidation, was higher in the CD+HIIT (P < 0.001) and OO+HIIT (P = 0.026) groups compared to the CD and OO groups, respectively (Figure 2F). Consistent with increased markers of oxidative metabolism, the OO diet increased the complete oxidation of [1-¹⁴C]oleate (P = 0.005, Figure 2G), and this level of fatty acid oxidation was comparable to both HIIT groups, while incomplete fatty acid oxidation (Figure 2H) was not different (P > 0.05) between any groups. Similar to fatty acid oxidation, citrate synthase activity (Figure. 2I) in liver homogenates was elevated in the OO group compared to the CD group (P = 0.005), and this elevation was comparable to the levels observed in both HIIT groups. Other markers (SIRT1 and UCP2 protein and Ppara and Ppargc1a mRNA) of mitochondrial biogenesis and oxidative metabolism were not different (P > 0.05) between groups. Collectively, these data suggest that an OO diet produces metabolic adaptations in the liver that somewhat mirror HIIT, but OO does not augment HIIT induced increases in hepatic lipid content or mitochondrial biogenesis.

Figure 2. Dietary olive oil and HIIT modify liver metabolism.

A) H&E images of liver tissue sections. B) Relative size of lipid droplets quantified from H&E images. C) Liver triglyceride levels. D-E) Western blot images (D) and densitometry quantification (E) of ATP-citrate lyase, SIRT 1, COX 4, complex I-V of electron transport chain, and UCP2. F) Relative mRNA abundance of Pnpla2, Fas, Ppara, Ppargc1a, and Cpt1a. G) Complete oxidation of [1-¹⁴C]oleate in liver tissue homogenate. H) Incomplete oxidation (acid soluble metabolites) of [1-¹⁴C]oleate in liver tissue homogenate. I) Citrate synthase activity of liver tissue homogenate. ^aCD vs. OO: P = 0.046, ^bCD vs. CD+HIIT: P = 0.020, ^cCD vs. OO: P = 0.003, ^dCD vs. CD+HIIT: P = 0.003, ^eCD vs. OO: P = 0.004, ^fCD vs. CD+HIIT: P = 0.007, ^gCD vs. OO: P = 0.045, ^hOO vs. OO+HIIT: P = 0.040, ⁱCD vs. OO: P < 0.001, ^jCD vs. CD+HIIT: P = 0.008.

3.3. Dietary olive oil enhances HIIT induced increases in skeletal muscle oleate oxidation.

Neither OO nor HIIT training modified the histological appearance (Figure 3A) or the individual fiber size (Figure 3B) of the soleus muscle between groups (P > 0.05). Neither OO nor HIIT modified gastrocnemius triglyceride levels between groups (all P > 0.05) (Figure 3C) consistent with unchanged markers (ATP-CL protein and Fas mRNA) of lipogenesis (Figure 3D–F). The CD+HIIT (P = 0.016) and OO+HIIT (P = 0.021) groups had greater gastrocnemius protein abundance of complex IV compared to the CD and OO groups, respectively (Figure 3D–E), while the CD+HIIT group also had greater COX 4 (P = 0.018) and complex V (P = 0.032) protein abundance vs. the CD group. These changes in protein abundance coincided with increased plantaris mRNA abundance of Ppara (P = 0.016) and Ppargc1a (P = 0.035), but not Pnpla2, in the CD+HIIT group vs. the CD group (Figure 3F). In quadricep muscle tissue homogenates, the OO+HIIT intervention increased the complete oxidation of [1-¹⁴C]oleate compared to the OO (P = 0.0002) and CD+HIIT (P = 0.004) groups (Figure 3G). Likewise, the incomplete oxidation of [1-¹⁴C]oleate was greater in the OO+HIIT group compared to the OO (P < 0.001) or CD+HIIT (P = 0.025) groups (Figure 3H). Additionally, in plantaris muscle, citrate synthase activity was significantly higher in the OO+HIIT group (P = 0.001) compared to the OO group (Figure 3I). Together, these data suggest a synergy between dietary OO and HIIT to amplify the ability of skeletal muscle to oxidize oleate.

Figure 3. Synergy of dietary olive oil and HIIT to increase skeletal muscle oleate oxidation.

A) H&E images of soleus tissue sections. B) Relative myofiber size determined from H&E images. C) Gastrocnemius triglyceride levels. D-E) Western blot images (D) and densitometry quantification (E) of ATP-CL, SIRT 1, COX 4, complex I-V of electron transport chain, and UCP2. F) Relative mRNA abundance of Pnpla2, Fas, Ppara, Ppargc1a, and Cpt1a. G) Complete oxidation of [1-¹⁴C]oleate in quadriceps muscle tissue homogenate. H) Incomplete oxidation (acid soluble metabolites) of [1-¹⁴C]oleate in quadriceps muscle tissue homogenate. I) Citrate synthase activity of plantaris muscle tissue homogenate. ^aCD vs. CD+HIIT: P = 0.018, ^bCD vs. CD+HIIT: P = 0.032, ^cCD+HIIT vs. CD: P = 0.016, ^dOO vs. OO+HIIT: P = 0.022.

3.4. RNA sequencing reveals reciprocal effects of OO and HIIT on gene abundance in type I skeletal muscle fibers.

To this point the metabolic adaptations in skeletal muscle were determined in either mixed gastrocnemius, quadriceps, or plantaris muscles, all of which are known to express predominately type IIB (fast, glycolytic) muscle fibers [36]. To examine metabolic changes induced by OO and/or HIIT in type I / IIA (slow/intermediate, oxidative) muscle fibers, we performed RNA-sequencing using soleus muscle, which in mice is predominately a mix of type I and type IIA muscle fibers [36]. Volcano plots (Supplemental Figure 1A) revealed that the CD+HIIT group, compared with the CD group, had 355 genes upregulated and 129 genes that were downregulated, while the OO+HIIT group, compared with the OO group, had 208 genes upregulated and 70 downregulated. No genes were differentially expressed between the OO+HIIT and CD+HIIT groups (data not shown). Meanwhile, the OO diet, compared to the CD diet, upregulated 72 genes and downregulated 223 genes. Next, a dot plot (Supplemental Figure 1B) was generated to illustrate which Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were modified by each condition. Common KEGG pathways upregulated with HIIT, independent of diet, included the complement and coagulation cascade, retinol metabolism, peroxisome proliferator-activated receptor (PPAR) signaling, and biosynthesis of amino acids, to name a few, with many of these genes involved in muscle regeneration and lipid metabolism. Genes within these same pathways were next plotted in heat maps (Figure 4), which revealed that overall HIIT, independent of diet, increased the gene abundance of most genes within these pathways (i.e., promoted muscle regeneration and lipid metabolism), whereas the OO diet alone, compared to the CD, had reciprocal effects by reducing the abundance of most genes within each of these same KEGG pathways. Given that HIIT increased the abundance of numerous genes, independent of diet, and the fact there were no differentially expressed genes between OO+HIIT vs. CD+HIIT, these data suggest that HIIT may override the effect of dietary OO on gene expression changes on markers associated with muscle regeneration and lipid metabolism in skeletal muscle.

Figure 4. Heat map of differentially expressed genes between each comparison within select KEGG pathways.

Heatmap of differentially expressed genes within the complement and coagulation cascade, retinol metabolism, PPAR signaling, linoleic/arachidonic acid metabolism, bile secretion, and biosynthesis of amino acid KEGG pathways between groups. The data is expressed as LOG of counts per million (CPM). CPM is expressed as the raw expression of each gene (i.e., total number of reads mapped to the gene’s exon) normalized by the total number of reads mapped to all gene exons.

3.5. Dietary olive oil or HIIT increases mitochondrial content in heart tissue.

Heart tissue sections revealed that neither OO nor HIIT training significantly modified gross tissue morphology (Figure 5A). Likewise, heart triglyceride levels and the mRNA abundance of markers of triglyceride metabolism, including Pnpla2 and Fas (Figure 5C), or the protein abundance of ATP-CL (Figure 5D–E) were not significantly different (P > 0.05) between conditions, suggesting that OO, HIIT, or their combination had minimal effects on heart triglyceride metabolism. The mRNA abundance of Ppara was significantly reduced (P = 0.003) in the CD+HIIT group compared to the CD group, whereas Ppargc1a or Cpt1a mRNA abundance was not modified (P > 0.05) by OO or HIIT training (Figure 5C). On the other hand, the OO diet significantly increased COX 4 (P < 0.001), complex IV (P = 0.049), and complex III (P = 0.048) protein abundance compared to the CD group, while the CD+HIIT group had higher protein abundance of COX 4 (P < 0.001), complex V (P < 0.001), complex IV (P < 0.001), complex III (P < 0.001), complex II (P = 0.009), and complex I (P = 0.003) compared to the CD group (Figure 5D–E). Additionally, the OO+HIIT group had significantly higher protein abundance of COX 4 (P = 0.029) compared to the OO group (Figure 5D–E), suggesting that although the OO diet increases mitochondrial content in the heart, HIIT training amplified this response while diet did not impact the effect of HIIT. In line with this data, citrate synthase activity tended (P = 0.055) to be higher in the OO group and was significantly higher (P = 0.001) in both HIIT groups (P < 0.001) compared to their respective CD groups (Figure 5F).

Figure 5. Dietary olive oil increases heart mitochondrial content to a similar extent as HIIT.

A) H&E images of heart tissue sections. B) Heart triglyceride levels. C) Relative mRNA abundance of Pnpla2, Fas, Ppara, Ppargc1a, and Cpt1a. D-E) Western blot images (D) and densitometry quantification (E) of markers ATP-CL, SIRT 1, COX 4, complex I-V of electron transport chain, and UCP2. F) Citrate synthase activity of heart muscle tissue homogenate. ^aCD vs. OO: P = 0.0002, ^bCD vs. CD+HIIT: P < 0.0001, ^cOO vs. OO+HIIT: P = 0.029, ^dCD vs. OO, P = 0.035, ^eCD vs. CD+HIIT: P < 0.0001, ^fCD vs. OO: P = 0.049, ^gCD vs. CD+HIIT: P < 0.0001, ^hCD vs. OO: P = 0.048, ⁱCD vs. CD+HIIT, P < 0.0001, ^jCD vs. CD+HIIT, P = 0.010, ^kCD vs. CD+HIIT, P = 0.004.

3.6. Neither dietary olive oil nor HIIT modify brown adipose tissue (BAT).

There were no gross morphological differences in the appearance of BAT tissue between the groups (Figure 6A). Likewise, BAT lipid droplet size (P = 0.608) was not different between the CD vs. OO groups and was similar among the HIIT groups (Figure 6B). Additionally, triglyceride levels (P = 0.133) (Figure 6C), the protein abundance of ATP-CL, SIRT 1, COX 4, complex V, complex IV, complex III, complex II, complex I, UCP2 (all P > 0.05) (Figure 6D–E), or citrate synthase activity (P = 0.959) (Figure 6F) were not significantly different between the groups, suggesting that neither OO nor HIIT modify BAT metabolism.

Figure 6. Neither dietary olive oil nor HIIT modify BAT lipid metabolism.

A) H&E images of BAT tissue sections. B) Relative lipid droplet size within H&E images. C) BAT triglyceride levels. D-E) Western blot images (D) and densitometry quantification (E) of markers ATP-CL, SIRT 1, COX 4, complex I-V of electron transport chain, and UCP2. F) Relative citrate synthase activity of BAT tissue homogenate.

3.7. OO and HIIT effects on adipose tissue varies by depot.

Tissue sections of sWAT stained for H&E revealed that the CD+HIIT group (P = 0.002) and OO+HIIT group (P = 0.004) had larger lipid droplets compared to the CD group (Figure 7A–B). These alterations in sWAT adipocyte size between the CD vs. CD+HIIT groups were not associated with alterations in the protein abundance of markers of lipogenesis including ATP-CL or mitochondrial metabolism including SIRT 1, COX 4, complex V, complex IV, complex III, complex II, complex I of the electron transport chain, or UCP2 (all P > 0.5) (Figure 7C–D). Additionally, citrate synthase activity (Figure 7E) was not modified (P = 0.354) between the CD vs. CD+HIIT group, suggesting that some other unknown factor has a role in mediating the impact of HIIT on adipocyte size.

Figure 7. Dietary olive oil or HIIT modify sWAT and gWAT morphology.

A) H&E images of sWAT tissue sections. B) Relative lipid droplet size within H&E images from A. C-D) Western blot images (C) and densitometry quantification (D) of markers ATP-CL, SIRT 1, COX 4, Complex I-V of electron transport chain, and UCP2. E) Relative citrate synthase activity of sWAT tissue homogenate. F) H&E images of gWAT tissue sections. G) Relative lipid droplet size within H & E images from F. H-I) Western blot images (H) and densitometry quantification (I) of markers ATP-CL, SIRT 1, COX 4, complex I-V of electron transport chain, and UCP2. J) Relative citrate synthase activity of gWAT tissue homogenate. ^aOO vs. OO+HIIT: P = 0.044, ^bOO vs. OO+HIIT: P = 0.018.

In gWAT, H&E images revealed that adipocyte size was significantly greater in the OO group (P = 0.001) compared to the CD group (Figure 7F–G). These modifications in adipocyte size between diets was not associated with changes in the protein abundance of ATP-CL, SIRT 1, COX 4, complex V, complex IV, complex III, and complex I of the electron transport chain (all P > 0.05) (Figure 7H–I). HIIT augmented the impact of OO on adipocyte size, as the OO+HIIT group had greater (P = 0.002) adipocyte size compared to the OO group (Figure 7F–G). The increase in gWAT adipocyte size with the OO+HIIT group was associated with a modest increase (P = 0.044) in complex II protein abundance, but lower (P = 0.018) UCP2 protein abundance, compared to the OO group (Figure 7H–I), while citrate synthase activity was not significantly different between these groups (P = 0.096) (Figure 7J).

4. Discussion

This study makes several novel observations about the isolated and combined effects between dietary OO and HIIT on metabolic outcomes across organ systems. We show that consuming OO results in metabolic adaptations in liver and heart that are similar to HIIT, although combining OO+HIIT was not additive. Combining OO+HIIT results in some additive effects in skeletal muscle and minor effects on morphology of gWAT. In skeletal muscle the capacity for oleate oxidation was greatest in the OO+HIIT group, an effect that was independent of mitochondrial content. In gWAT, combining OO+HIIT increased adipocyte size to a greater extent than OO alone. Additionally, RNA sequencing revealed that OO reduced the abundance of genes involved in muscle regeneration and lipid metabolism, while HIIT, independent of diet, had the opposite effect by increasing the abundance of genes involved in muscle regeneration and lipid metabolism. These data highlight the pleotropic effects of dietary OO and/or HIIT training to modify lipid metabolism and mitochondrial content in various tissues in females.

Herein in female mice, neither OO and/or HIIT training modified body weight or composition, and this finding is consistent with previous studies [21, 37–40]. Typically, the lifestyle modalities used in the current study are more effective at reducing body weight during hypercaloric high-fat diet feeding conditions in rodents, whereas when feeding diets with moderate fat content, such as that used in the current study, body weight is not as significantly modified. For instance, HIIT training modestly, but significantly, reduced body weight in male mice fed a high-fat diet, but not a low-fat diet [21, 37]. A high-fat diet containing OO modestly reduced body weight gain compared to an isocaloric high-fat diet containing butter [39], whereas a low-fat diet containing OO did not modify body weight compared to mice consuming iso-caloric diets with canola oil, sesame oil, or soybean oil [38]. Furthermore, previous work from our lab showed that an OO rich diet did not modify body weight gain of mice that had carcinogen initiated liver tumorigenesis [40].

The saturation index of fatty acids will differentially modify hepatic lipid metabolism [14, 41–43]. For example, the monounsaturated fatty acid oleate, in contrast to the saturated fatty acid palmitate, does not induce cytotoxicity and apoptosis in hepatocytes [41, 42], while oleate promotes greater increases in triglyceride levels compared to palmitate [43]. Additionally, previous data from our lab has shown that lipid droplet derived monounsaturated fatty acids, such as oleate, allosterically activate SIRT1 and promote oxidative metabolism specifically by enhancing PGC-1α/PPARα signaling in hepatocytes [14]. Similar to these previous findings, the present study also shows that OO promotes increases in hepatic triglyceride levels, mitochondrial content, and fatty acid oxidation. However, the addition of HIIT to the OO diet was not additive, as similar hepatic adaptations were observed between OO and HIIT groups.

HIIT training increased hepatic mitochondrial content and fatty acid oxidation, and in line with our findings prior work has consistently shown that exercise promotes hepatic mitochondrial biogenesis and fatty acid oxidation under various dietary regimens [44, 45]. Additionally, we found that HIIT increases hepatic triglyceride levels, independent of dietary fat content. The impact of exercise on hepatic triglyceride levels is not straightforward, as some studies report exercise reduces [37, 46], does not change [47], or even increases hepatic triglyceride levels [48–50]. Given the relatively dynamic nature of hepatic lipid content, this discrepancy in findings between studies may be due to putative factors such as species, sex, feeding status (fed vs. fasted, and length of fast), diet composition, mode, duration and intensity of exercise training, or timing of measurement of liver triglyceride levels since last exercise session.

Oleate, compared to palmitate, is more readily incorporated into triglycerides in muscle cells [11], although in mice a high-fat diet containing OO, in which the predominate fatty acid is oleate, does not differentially modify skeletal muscle triglyceride levels compared to high-fat diets containing other fatty acids [51]. In line with this latter study, we show that the OO diet did not modify skeletal muscle triglyceride levels compared to the CD. Moreover, HIIT training, regardless of diet, did not modify skeletal muscle triglyceride levels 48 h after the last exercise session. This finding contrasts with some studies [50], but is in line with other studies that show prior exercise does not modify [52, 53] skeletal muscle triglyceride levels in rodents.

In C2C12 skeletal muscle cells oleate stimulates β-adrenergic signaling, PGC-1α activity, and fatty acid oxidation [13]. In the present study, the OO diet alone did not modify fatty acid oxidation in type II (glycolytic) muscle, but the HIIT protocol, regardless of diet, increased mitochondrial content. In contrast, the OO+HIIT group displayed greater oleate oxidation compared to the CD, OO, and CD+HIIT groups. Given that HIIT increased mitochondrial content similarly in both diet groups suggests that HIIT and dietary OO may synergize to specifically enhance the capacity for oleate oxidation in type II fibers, although the mechanisms are not known.

PPAR’s are nuclear hormone receptors that have a central role in lipid metabolism [54]. Prior work from our lab showed that dietary OO enhances hepatic Ppara signaling and fatty acid oxidation [14], while herein we show that dietary OO increased skeletal muscle Pparg gene abundance (Supplemental Table 1), without impacting Ppara, indicating OO impacts PPAR signaling differentially in various tissues. Other members of the PPAR family including ApoA1–5, Fabp1, Slc27a2, and Slc27a5 were all reduced with OO, but increased with exercise, independent of diet, suggesting that exercise overrides the impact of OO on PPAR signaling.

Unexpectedly, the RNA-sequencing analysis on type I / IIA (oxidative) soleus muscle showed that mice consuming the OO diet had a lower abundance of genes that are within numerous KEGG pathways that are linked to skeletal muscle inflammation and regeneration [55, 56] including complement and coagulation cascades [57], retinol metabolism [58], bile secretion [59], inflammatory mediator regulation of TRP channels, and arginine biosynthesis [60]. In particular, the OO diet reduced the gene abundance of the alpha (Fga gene), beta (Fgb), and gamma (Fgg gene) component of fibrinogen, a protein that plays a key role in the acute phase inflammatory response caused by tissue injury [61], which is an important component of muscle regeneration following exercise. On the other hand, acute exercise of sufficient intensity and duration is known to acutely induce muscle damage and promote an acute skeletal muscle regeneration response that is modulated by inflammation [55] up to 48 h post exercise [62], including genes within the complement and coagulation cascades [63, 64]. Consistent with this, the data herein shows that 48 h post HIIT there was an increase in abundance of several genes associated with inflammation and muscle regeneration, including Fga, Fgb, and Fgg gene abundance. Furthermore, there were no differentially expressed genes between OO+HIIT vs. CD+HIIT, suggesting that HIIT induced inflammatory and muscle regeneration responses in skeletal muscle are not attenuated by dietary OO.

In conclusion, OO+HIIT did not synergistically enhance increases in lipid metabolism and mitochondrial content to a greater extent than either intervention alone. However, OO+HIIT was additive in two respects, including in skeletal muscle, where OO+HIIT increased the capacity for oleate oxidation to a greater extent than HIIT or OO alone, and in gWAT, where OO+HIIT increased adipocyte hypertrophy to a greater extent than either OO or HIIT alone. However, the contribution of these additive effects to metabolic health in unclear. Additionally, OO alone increased mitochondrial content to a similar extent as HIIT in both liver and heart tissue, suggesting that the underlying mechanisms of OO diets to improve the health of liver and heart tissue may overlap with HIIT. Lastly, dietary OO reduced muscle regeneration and lipid metabolism gene abundance in skeletal muscle, while HIIT increased the abundance of these genes, an effect that could not be suppressed with dietary OO. Overall, these data highlight the efficacy of OO diets and HIIT to increase lipid metabolism and mitochondrial content in various tissues in female mice.

Supplementary Material

Heden et al. Supplemental Figure 1.

Supplemental Table 1. RNA sequencing data. Excel file showing differential expression results for each comparison made. There were no significant differences found between CD+HIIT vs. OO+HIIT so this comparison is not shown in the Excel file.