l-Citrulline supplementation improves glucose and exercise tolerance in obese male mice
Funding information:
This study was supported by a New Investigator Operating Grant from Diabetes Canada to E.E.M., a Project Grant from the Canadian Institutes of Health Research to J.M.S., a Career Development Award from the American Diabetes Association to J.E.C. and a Discovery Grant from the Natural Sciences and Engineering Research Council of Canada to J.R.U. J.E.C. is a Borden Scholar and J.R.U. is Scholar of Diabetes Canada. A.E. is supported by a Scholarship from the Libyan Ministry of Higher Education.
Edited by: Rebecca Campbell
Abstract
New Findings
-
What is the central question of the study?
Does the action of l-citrulline, which has been shown to augment performance in animals and athletes, possibly via increasing mitochondrial function, translate to obese animals, and does this improve glycaemia?
-
What is the main finding and its importance?
Chronic supplementation with l-citrulline improves not only exercise capacity, but also glycaemia in obese mice, which would be beneficial as obese individuals are at increased risk for type 2 diabetes. However, l-citrulline supplementation also caused a mild impairment in insulin signalling and insulin tolerance in obese mice.
l-Citrulline is an organic α-amino acid that has been shown to have a number of salutary actions on whole-body physiology, including reducing muscle wasting and augmenting exercise and muscle performance. The latter has been suggested to arise from elevations in mitochondrial function. Because enhancing mitochondrial function has been proposed as a novel strategy to mitigate insulin resistance, our goal was to determine whether supplementation with l-citrulline could also improve glycaemia in an experimental mouse model of obesity. We hypothesized that l-citrulline treatment would improve glycaemia in obese mice, and this would be associated with elevations in skeletal muscle mitochondrial function. Ten-week-old C57BL/6J mice were fed either a low-fat (10% kcal from lard) or a high-fat (60% kcal from lard) diet, while receiving drinking water supplemented with either vehicle or l-citrulline (0.6 g l−1) for 15 weeks. Glucose homeostasis was assessed via glucose/insulin tolerance testing, while in vivo metabolism was assessed via indirect calorimetry, and forced exercise treadmill testing was utilized to assess endurance. As expected, obese mice supplemented with l-citrulline exhibited an increase in exercise capacity, which was associated with an improvement in glucose tolerance. Consistent with augmented mitochondrial function, we observed an increase in whole body oxygen consumption rates in obese mice supplemented with l-citrulline. Surprisingly, l-citrulline supplementation worsened insulin tolerance and reduced insulin signalling in obese mice. Taken together, although l-citrulline supplementation improves both glucose tolerance and exercise capacity in obese mice, caution must be applied with its broad use as a nutraceutical due to a potential deterioration of insulin sensitivity.
1 INTRODUCTION
l-Citrulline is an organic α-amino acid found naturally in high quantities in watermelons, onions and garlic. Furthermore, l-citrulline is synthesized almost exclusively in the intestine and requires arginine and glutamine for its biosynthesis in the urea cycle (Bahri et al., 2013). It is also a by-product of nitric oxide synthase (NOS), which converts l-arginine into nitric oxide (NO) and l-citrulline. It can also serve as a precursor for l-arginine biosynthesis, and thus has been used as a supplement to increase circulating arginine concentrations in situations of arginine deficiency (Bahri et al., 2013), though it also increases circulating arginine concentrations in healthy young and aged adults (Bailey et al., 2015; Churchward-Venne et al., 2014).
Of interest, studies in humans have shown that citrulline–malate supplementation may enhance performance, as it reduced muscle fatigue by promoting aerobic energy production in exercising muscle (Bendahan et al., 2002). In addition, citrulline–malate supplementation in humans has also been shown to increase performance during high-intensity anaerobic exercise (Perez-Guisado & Jakeman, 2010). Similarly, studies in rats suggest that citrulline–malate enhances gastrocnemius muscle performance, as citrulline–malate supplementation reduced both the phosphocreatine and oxidative cost of contraction following electrically induced transcutaneous stimulation (Giannesini et al., 2011). While the above studies did not discern where citrulline or malate is responsible for the effects on performance, a recent study by Villareal et al. revealed that l-citrulline supplementation increases exercise performance in mice due to upregulation of peroxisome proliferator-activated receptor-γ coactivator 1α (PGC-1α) expression in skeletal muscle (Villareal, Matsukawa, & Isoda, 2018).
It is possible that citrulline's potential beneficial actions on exercise performance involve increases in mitochondrial function. Indeed, l-citrulline-mediated increases in circulating l-arginine levels increase NO production (Bahri et al., 2013; Lee & Kang, 2018; Sureda et al., 2009), and NO has been demonstrated in numerous studies to augment mitochondrial biogenesis and subsequent mitochondrial function, possibly via augmenting PGC-1α activity (McConell et al., 2010; Nisoli et al., 2004; Tengan, Rodrigues, & Godinho, 2012). Furthermore, a number of studies have demonstrated that increasing mitochondrial function through a variety of approaches (e.g. exercise, heat therapy, genetic-based strategies) protects against obesity-induced impairments in glucose homeostasis (Dube et al., 2008; Gupte, Bomhoff, Swerdlow, & Geiger, 2009; Seth et al., 2007). It has also been demonstrated that mice with a whole-body deletion of endothelial NOS exhibit impaired mitochondrial function, which results in glucose intolerance and insulin resistance (Duplain et al., 2001; Le Gouill et al., 2007).
Taken together, as increases in skeletal muscle mitochondrial function are frequently associated with improved glucose homeostasis, we anticipated that in addition to augmenting exercise performance, citrulline may have additional salutary actions that result in improved glycaemia. We hypothesized that treatment with l-citrulline would not only improve exercise tolerance in mice subjected to experimental obesity, but also improve glucose homeostasis, both of which would be due to increases in skeletal muscle mitochondrial function.
2 METHODS
2.1 Ethics approval
All experimental procedures performed in mice were approved by the University of Alberta Health Sciences Animal Welfare Committee under user protocol number AUP00001412, in accordance with the guidelines of the Canadian Council on Animal Care. In addition, all possible steps were taken by the investigators to minimize pain and suffering of animals, and killing was conducted in accordance with approved institutionalized protocols at the Health Sciences Laboratory Animal Services Facility of the University of Alberta. All animals were killed via intraperitoneal (i.p.) injection of sodium pentobarbital (Euthansol®; 0.3 ml (kg body weight)−1), and once loss of consciousness and a lack of toe-pinch reflex was observed (within 5 min), all animals underwent cardiac puncture for collection of blood, following which their tissues were extracted for further analyses. Our studies fully comply with the ethical principles and animal ethics checklist of Experimental Physiology.
2.2 Animal care
Ten-week-old male C57BL/6J mice (The Jackson Laboratory, Bar Harbor, ME, USA) were fed either a low-fat diet (LFD; 10% kcal from lard, Research Diets D12450J; New Brunswick, NJ, USA) or a high-fat diet (HFD, 60% kcal from lard, Research Diets D12492) and supplemented with either vehicle or l-citrulline (0.6 g l−1) in their drinking water for 15 weeks. This dose of l-citrulline amounts to a mouse on average consuming ∼100–150 mg kg−1 per day based on studies in our lab indicating that C57BL/6J mice consume ∼8 ml of l-citrulline-supplemented drinking water per day (8.39 ± 0.40 ml vehicle control drinking water vs. 8.19 ± 0.78 ml l-citrulline-supplemented drinking water). All experimental groups were subjected to several physiological assessments throughout the 15 weeks, and upon study completion all animals were killed via i.p. injection of sodium pentobarbital following a 6 h fast and at 15 min post-administration of saline or insulin. Tissues (e.g. gastrocnemius muscle, liver) were subsequently extracted and immediately snap frozen in liquid N2 using Wollenberger tongs precooled to the temperature of liquid N2, and stored at −80°C.
2.3 Cell culture
All reagents were purchased from Sigma-Aldrich (Oakville, Ontario, Canada). C2C12 (American Type Culture Collection, Manassas, VA, USA) cells were cultured in six-well plates in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. Cells were incubated in a water-jacketed CO2 incubator maintained at 37°C with 95% O2–5% CO2. Upon confluency C2C12 cells were differentiated into myotubes via growth in DMEM containing 2% horse serum and 1% penicillin–streptomycin as previously described (Ussher et al., 2009). After 5 days’ differentiation, C2C12 myotubes were treated with either 0.1% FBS-supplemented DMEM containing 30 µm, 300 µm or 1 mm l-citrulline, 100 µm or 200 µm l-arginine, or saline for 24 h. In a separate cohort of C2C12 myotubes, following 24 h treatment with saline, l-citrulline or l-arginine, all cells were subsequently treated with vehicle control (saline) or insulin (0.1 or 1 nm) for an additional 30 min.
2.4 Glucose homeostasis assessment
Intraperitoneal glucose and insulin tolerance tests (GTTs/ITTs) were performed in overnight- or 6 h-fasted mice, respectively, using a glucose dose of 2 g (kg body weight)−1 for the GTT, and an insulin dose of 0.7 U (kg body weight)−1 for the ITT. A 6 h fast was chosen for the ITT, as lean C57BL/6J mice are highly insulin sensitive and may become hypoglycaemic during an ITT following an overnight fast, due to their baseline glucose levels decreasing to ∼4 mm. Blood glucose was measured at 0, 15, 30, 60, 90 and 120 min post-glucose or insulin administration from mouse tail whole-blood using the Contour Next blood glucose monitoring system (Bayer, Leverkusen, Germany) as previously described (Al Batran et al., 2018, 2019). In addition, blood samples were collected in tubes containing the anticoagulant agent (EDTA) at 0 and 30 min post-glucose administration, in order to measure plasma insulin levels via use of a commercially available kit (Alpco Diagnostics, Salem, NH, USA) as previously described (Al Batran et al., 2018).
2.5 Exercise capacity assessment
During the assessment of exercise tolerance, mice ran on a calibrated, motor-driven treadmill (Columbus Instruments, Columbus, OH, USA) at a starting speed of 3 m min−1 for 2 min, followed by increasing speeds of 4 m min−1 for 2 min, 5 m min−1 for 2 min, 6 m min−1 for 6 min, 8 m min−1 for 10 min, 10 m min−1 for 3 min, 12 m min−1 for 3 min, 14 m min−1 for 3 min, 16 m min−1 for 3 min, 18 m min−1 for 3 min and 20 m min−1 until exhaustion. Animals ran until they reached exhaustion, which was defined as when an individual mouse would no longer run for more than 5 s at a time, and chose to remain on the apparatus shock grid even when assisted back onto the treadmill. The first 6 min is not included in our exercise capacity assessment, as this period is used for the mice to acclimatize and become familiar with the treadmill.
2.6 Indirect calorimetry
An Oxymax comprehensive lab animal monitoring system (Columbus Instruments) was used to assess in vivo energy metabolism via indirect calorimetry. After a 24 h period of acclimatization in the system, the subsequent 24 h period was utilized for data collection on whole-body oxygen consumption rates, respiratory exchange ratios, heat production and ambulatory activity as previously described (Al Batran et al., 2018).
2.7 Magnetic resonance imaging body composition analysis
An EchoMRI 4in1/700 body composition analyser was utilized to assess body composition for quantifying lean and fat mass in fully conscious mice via specialized nuclear magnetic resonance relaxometry-based technology.
2.8 Western blotting
Frozen soleus or gastrocnemius tissue (20 mg), or C2C12 myotube cells were homogenized in buffer containing 50 mm Tris–HCl (pH 8 at 4°C), 1 mm EDTA, 10% glycerol (w/v), 0.02% Brij-35 (w/v), 1 mm DTT, protease and phosphatase inhibitors (Sigma-Aldrich), following which extracted protein samples were denatured and subjected to western blotting protocols as previously described (Gopal et al., 2017). Protein kinase B (Akt) and phospho-Akt (9272S and 4060L; Cell Signaling Technology, Danvers, MA, USA) antibodies were prepared in a 1/1000 dilution in 5% bovine serum albumin.
2.9 Real-time PCR analysis
First-strand cDNA was synthesized from TRIzol-extracted RNA using the SuperScript III synthesis system (Invitrogen, Carlsbad, CA, USA). Real-time PCR was carried out with the CFX Connect real-time PCR machine (Bio-Rad Laboratories Inc., Hercules, CA, USA) using custom-designed Sybr Green primers (Table 1). Relative mRNA transcript levels were quantified with the 
method (Livak & Schmittgen, 2001) using peptidylprolyl isomerase A (Ppia/PPIA) as a housekeeping internal control gene as previously described.
| Gene name | Forward | Reverse |
|---|---|---|
| Ppargc1a | TATGGAGTGACATAGAGTGTGCT | CCACTTCAATCCACCCAGAAAG |
| Nrf1 | AGCACGGAGTGACCCAAAC | TGTACGTGGCTACATGGACCT |
| Tfam | GGAATGTGGAGCGTGCTAAAA | GCTGGAAAAACACTTCGGAATA |
| Tfbm2 | CCCAGAAAGCGTTTACAGAT | GAGATGTATGTATATGGGTG |
| Ppia | GAGCTGTTTGCAGACAAAGTTC | CCCTGGCACATGAATCCTGG |
2.10 Determination of mitochondrial respiration levels
Mitochondrial oxygen consumption was assessed in fresh saponin-permeabilized gastrocnemius muscle using a Clark oxygen electrode connected to an Oxygraph Plus recorder (Hansatech Instruments Ltd, Pentney, UK) as previously described (Kuznetsov et al., 2008). Permeabilized muscle was loaded onto a chamber containing 2 ml of respiration medium at 30°C. Saline, 30 µm l-citrulline, 1 mm l-citrulline, 100 µm l-arginine or 200 µm l-arginine was added to the chamber, following which basal respiration, complex I substrates glutamate (10 mm) and malate (5 mm) were added and oxygen consumption was recorded (State 4 respiration). Then, 1 mm ADP was added and the ADP-stimulated respiration rate was determined (State 3 respiration). Respiration rate was represented as nmol O2 consumed per mg protein per minute. The respiratory control ratio was calculated as the ratio between ADP-stimulated and basal respiration rates.
2.11 Statistical analysis
All values are presented as means ± standard deviation (SD). Significant differences were determined by the use of an unpaired, two-tailed Student's t test, or a two-way analysis of variance (ANOVA) followed by a Bonferroni post hoc analysis. Differences were considered significant when P < 0.05.
3 RESULTS
3.1 l-Citrulline supplementation improves obesity-induced glucose intolerance but worsens insulin sensitivity
C57BL/6J mice were fed either a LFD or a HFD supplemented with either vehicle or l-citrulline (0.6 g l−1) in their drinking water for 15 weeks. This had no effect on food intake (data not shown), body weight (Table 2) or lean body mass (Table 3). While l-citrulline supplementation in ad libitum-fed mice had no impact on blood glucose levels of either lean or obese mice, it did result in a lowering of fasting blood glucose levels in obese mice (Table 4). Moreover, glucose tolerance was improved only in the obese mice following 11 weeks of l-citrulline supplementation (Figure 1a,b), which was associated with a trend to increased circulating insulin levels, when comparing the change in insulin levels at the 30 min vs. 0 min time points of the GTT (Figure 1c).
| Body weight (g) | |||
|---|---|---|---|
| Week 1 | Week 7 | Week 15 | |
| LFD control | 24.7 ± 0.5 | 28.2 ± 0.6 | 29.2 ± 1.5 |
| LFD l-citrulline | 23.9 ± 0.8 | 27.3 ± 1.1 | 28.4 ± 2.8 |
| HFD control | 25.5 ± 0.7 | 45.6 ± 1.4* | 50.6 ± 1.4* |
| HFD l-citrulline | 25.6 ± 0.6 | 41.6 ± 3.0* | 49.0 ± 4.5* |
- Body weight was assessed in mice fed a LFD or HFD supplemented with either vehicle control or l-citrulline for 15 weeks (n = 5–8). Values are means ± SD. *P < 0.05, significantly different vs. the respective LFD counterpart. HFD, high fat diet; LFD, low fat diet.
| Total lean body mass (g) | |
|---|---|
| LFD control | 22.6 ± 2.6 |
| LFD l-citrulline | 22.6 ± 2.4 |
| HFD control | 25.0 ± 1.3 |
| HFD l-citrulline | 25.4 ± 1.6 |
- Lean body mass was assessed in mice fed a LFD or HFD supplemented with either vehicle control or l-citrulline for 15 weeks (n = 4–8). Values are means ± SD. HFD, high fat diet; LFD, low fat diet.
| Blood glucose (mm) | ||
|---|---|---|
| Fasting | Ad libitum feeding | |
| LFD control | 4.77 ± 0.66 | 7.48 ± 1.17 |
| LFD l-citrulline | 4.57 ± 0.70 | 7.05 ± 0.81 |
| HFD control | 8.77 ± 1.06† | 11.50 ± 1.82† |
| HFD l-citrulline | 7.12 ± 1.40*† | 12.57 ± 4.10† |
- Blood glucose levels were assessed in mice fed a LFD or HFD supplemented with either vehicle control or l-citrulline for 11 weeks, following an overnight fast or with ad libitum feeding (n = 8–10). Values are means ± SD. *P < 0.05, significantly different vs. the respective vehicle control-treated counterpart. †P < 0.05, significantly different vs. the respective LFD counterpart. HFD, high fat diet; LFD, low fat diet.
l-Citrulline supplementation improves glucose homeostasis in obese mice. (a) Glucose tolerance in lean and obese mice treated with either vehicle or l-citrulline for 11 weeks (n = 12–15). (b) Area under the curve (AUC) during the GTT (n = 12–15). (c) Plasma insulin levels during the glucose tolerance test at 0 and 30 min post-glucose administration (n = 5–11). Values are means ± SD. Differences were determined using a two-way ANOVA, followed by a Bonferroni post hoc analysis. *P < 0.05, significantly different vs. the respective vehicle control-treated counterpart
Conversely, l-citrulline supplementation for 9 weeks worsened insulin tolerance in obese mice (Figure 2a,b). This worsening was associated with alterations in insulin signalling, as baseline Akt phosphorylation appeared elevated in muscles from lean mice supplemented with l-citrulline, whereas in obese mice, insulin-stimulated Akt phosphorylation in gastrocnemius but not soleus muscles was non-existent when compared to vehicle supplementation (Figure 2c,d). Of interest, these adverse actions of l-citrulline on skeletal muscle insulin tolerance/signalling may be indirectly mediated, since direct treatment of differentiated C2C12 myotubes with pharmacological concentrations of l-citrulline during serum starvation (0.1% FBS) did not worsen insulin-stimulated Akt phosphorylation (Figure 3).
l-Citrulline supplementation worsens insulin tolerance in obese mice. (a) Insulin tolerance in lean and obese mice treated with either vehicle or l-citrulline for 9 weeks (n = 7–17). (b) Changes in plasma glucose levels normalized to starting baseline glucose levels during the ITT (n = 7–17). (c) Insulin signalling (Akt phosphorylation) in soleus muscles from both lean and obese mice treated with either vehicle or l-citrulline during the ITT (n = 4). (d) Insulin signalling (Akt phosphorylation) in gastrocnemius muscles from both lean and obese mice treated with either vehicle or l-citrulline during the ITT (n = 4). Values are means ± SD. Differences were determined using a two-way ANOVA, followed by a Bonferroni post hoc analysis. *P < 0.05, significantly different vs. the respective vehicle control-treated counterpart. †P < 0.05, significantly different vs. the LFD vehicle control no insulin group. ‡P < 0.05, significantly different vs. the HFD vehicle control + insulin group
3.2 l-Citrulline supplementation increases exercise capacity in both lean and obese mice
To evaluate the effect of l-citrulline supplementation on exercise performance, both lean and obese mice supplemented with l-citrulline for 8 weeks ran on a forced exercise treadmill at gradually increasing speeds until exhaustion. As we expected, l-citrulline improved exercise performance, as we observed significant increases in the time and distance that both l-citrulline-supplemented lean and obese mice were able to run on the forced exercise treadmill (Figure 4).
3.3 l-Citrulline supplementation increases whole-body oxygen consumption rates but does not impact substrate preference in obese mice
We next assessed in vivo energy metabolism in lean and obese mice following 10 weeks of l-citrulline supplementation via indirect calorimetry, which revealed significant increases in whole-body oxygen consumption rates in obese mice during the initial hours of the dark cycle (Figure 5a,b). This increase was associated with increased ambulatory activity (Figure 5c), but was not associated with changes in substrate preference, as respiratory exchange ratios were similar in both lean and obese mice supplemented with l-citrulline (Figure 6a,b). To further support our observations that l-citrulline supplementation does not modify substrate preference, and to assess whether l-citrulline has direct actions on muscle mitochondria that enhance respiration, we quantified respiratory control ratios in permeabilized fibres from gastrocnemius muscles of lean mice. Direct treatment of permeabilized fibres with pharmacological concentrations of l-citrulline had no impact on mitochondrial respiration rates or the respiratory control ratio (Table 5). Because l-citrulline can be recycled into l-arginine, which may enhance mitochondrial respiration by increasing NO formation, we also treated permeabilized gastrocnemius fibres with pharmacological concentrations of l-arginine, though once again we saw no change in the respiratory control ratio (Table 5).
| Respiration (nmol O2 min−1 mg−1) | |||
|---|---|---|---|
| Basal | ADP-stimulated | Respiratory control ratio | |
| Vehicle control | 0.512 ± 0.251 | 1.667 ± 0.627 | 3.795 ± 1.861 |
| 30 µm l-citrulline | 1.447 ± 1.702 | 5.259 ± 4.143 | 4.536 ± 2.361 |
| 1 mm l-citrulline | 1.097 ± 0.652 | 5.044 ± 2.603 | 5.470 ± 4.251 |
| 100 µm l-arginine | 2.485 ± 2.036† | 10.387 ± 7.898* | 4.673 ± 1.682 |
| 200 µm l-arginine | 1.587 ± 1.413 | 6.334 ± 4.781 | 4.395 ± 1.584 |
- Oxygen consumption was assessed using a Clark electrode connected to an Oxygraph Plus recorder where malate and glutamate were used to stimulate basal respiration. Rates are presented as the respiratory control ratio, which is a ratio of ADP-stimulated to basal respiration. Values are means ± SD (n = 5–6). *P < 0.05, significantly different vs. the respective vehicle control-treated counterpart. †P = 0.19, vs. the respective vehicle control-treated counterpart.
3.4 l-Citrulline supplementation reverses obesity-induced impairments in the expression of key regulators of mitochondrial function/biogenesis
We next quantified mRNA expression for a number of factors associated with the regulation of mitochondrial function and/or biogenesis, as we posited these factors may contribute to how l-citrulline improves muscle performance and aerobic energy metabolism. Experimental obesity resulted in significant reductions or trends to reductions in the mRNA expression of peroxisome proliferator activated receptor γ coactivator-1 α (Ppargc1a), nuclear respiratory factor 1 (Nrf1) and mitochondrial transcription factor B2 (Tfbm2) in gastrocnemius but not soleus muscles (Figure 7a,b). Of interest, l-citrulline supplementation prevented the obesity-induced reduction in gastrocnemius muscle Ppargc1a, Nrf1 or Tfbm2 expression, while also increasing mitochondrial transcription factor A (Tfam) expression, but had no effect on this expression profile in soleus muscle (Figure 7a,b). To determine whether the actions of l-citrulline on gastrocnemius muscle mRNA expression profiles were due to direct actions on the muscle, we treated differentiated C2C12 myotubes during serum starvation (0.1% FBS) with increasing concentrations of l-citrulline. Similar to our in vitro observations assessing insulin signalling, direct treatment of C2C12 myotubes with l-citrulline had negligible influence on the mRNA expression of Ppargc1a, Nrf1, Tfam or Tfbm2 vs. their saline-treated counterparts (Figure 7c). Likewise, the in vivo changes in gastrocnemius Ppargc1a, Nrf1, Tfam and Tfbm2 mRNA expression are not due to l-citrulline-mediated increases in l-arginine levels, as direct treatment of C2C12 myotubes with l-arginine failed to increase Ppargc1a, Nrf1, Tfam or Tfbm2 mRNA expression (data not shown).
4 DISCUSSION
Our current study supports previous findings from others demonstrating that l-citrulline supplementation may improve exercise performance, as both lean and obese mice supplemented with l-citrulline in the drinking water were able to run ∼28% and ∼47% longer, respectively, on a forced exercise treadmill. In addition, we report here for the first time novel actions of l-citrulline supplementation on obesity-induced dysglycaemia, where l-citrulline supplementation improves glucose homeostasis but surprisingly worsens insulin tolerance and insulin signalling in obese mice. Such observations should be taken into consideration for individuals choosing to consume l-citrulline as a nutraceutical supplement in attempts to improve their exercise performance.
The increase we observed in exercise performance in both lean and obese mice supplemented with l-citrulline was expected, as previous studies in humans have reported similar observations: citrulline-malate co-supplementation reduced muscle fatigue in 18 men performing finger flexions at 1.5 s intervals lifting a 6 kg weight (Bendahan et al., 2002). It was postulated that this benefit was due to augmented aerobic energy production in exercising muscle, as 31P magnetic resonance spectroscopy (31P-MRS) studies revealed a 34% increase in oxidative ATP production and 20% increase in the rate of phosphocreatine recovery within the exercising muscle. Likewise, citrulline–malate co-supplementation reduced muscle soreness and repetition number during flat barbell bench presses in 41 male volunteers (Perez-Guisado & Jakeman, 2010). In preclinical studies, citrulline–malate supplementation improved muscle efficiency in electrically stimulated gastrocnemius muscles from anaesthetized male Wistar rats, as seen by decreases in both the phosphocreatine and oxidative costs of contraction during 31P-MRS studies (Giannesini et al., 2011). Furthermore, studies in rats have demonstrated that citrulline supplementation enhances endurance capacity, potentially via activating muscle protein synthesis and modulating substrate flux, though independent of a direct improvement of mitochondrial function (Goron et al., 2017). Conversely, a single dose of l-citrulline failed to improve maximum number of repetitions over five sets, time to exhaustion and maximal oxygen consumption during chest press exercise in 22 volunteer athletes (11 males/11 females) (Cutrufello, Gadomski, & Zavorsky, 2015). The reasons for the discrepancy between these studies remain unknown, though a possible explanation could stem from the fact that the latter study utilized l-citrulline supplementation alone vs. citrulline–malate co-supplementation. Nonetheless, Takeda and colleagues have shown that l-citrulline supplementation improves time to exhaustion and reduces circulating lactate levels during swimming exercise in ICR mice (Takeda, Machida, Kohara, Omi, & Takemasa, 2011).
The inclusion of malate as a co-supplement is thought to account for the potential improvement in aerobic energy metabolism, since malate is a key intermediate of the Krebs cycle. Our results support the findings of Takeda and colleagues, however, since l-citrulline supplementation alone was sufficient to improve aerobic capacity on a forced exercise treadmill. Moreover, we observed increases in whole-body oxygen consumption rates only in mice subjected to experimental obesity, which impairs whole-body oxygen consumption (Ussher et al., 2010), suggesting that mitochondrial function may potentially be improved via l-citrulline supplementation. Conversely, mitochondrial flux under submaximal conditions, as would primarily be the case for both our lean and obese mice, reflects the cellular demand for ATP generation (Holloszy, 2009), and hence the increase in oxygen consumption may not necessarily indicate an improvement in mitochondrial function. Nonetheless, we did observe reductions in mRNA expression of numerous factors linked to the regulation of various aspects of mitochondrial function in response to obesity, which were prevented by l-citrulline supplementation. Indeed, PGC-1α is a key regulator of mitochondrial function/energy metabolism in skeletal muscle (Handschin & Spiegelman, 2008), and l-citrulline supplementation was associated with enhanced Ppargc1a expression in obese mouse gastrocnemius muscle, though no differences were observed in lean mouse gastrocnemius muscle. Such observations contrast with studies in lean mice supplemented for 15 days with l-citrulline, as Villareal and colleagues demonstrated l-citrulline-mediated increases in swimming exercise performance, which were associated with an upregulation of PGC-1α expression in gastrocnemius muscles (Villareal et al., 2018). The incompatibility of our observations with those of Villareal and colleagues could be due to the fact that our mice were not frequently exercising and we simply assessed performance during a single aerobic exercise challenge. In addition, our daily dose of l-citrulline supplementation was significantly lower, which could also explain why we did not observe increased gastrocnemius muscle Ppargc1a expression. Although l-citrulline did not have direct actions in lean mice resulting in increased Ppargc1a expression, l-citrulline supplementation reversed the impairment of experimental obesity on gastrocnemius but not soleus muscle Ppargc1a, Nrf1, Tfam or Tfbm2 mRNA expression. Because these genes all represent key regulators of mitochondrial function and subsequent energy metabolism (Dillon, Rebelo, & Moraes, 2012; Handschin & Spiegelman, 2008), preventing downregulation of this mRNA expression profile in skeletal muscle may explain why whole-body oxygen consumption rates were not decreased in response to l-citrulline supplementation in obesity. Conversely, we observed mild increases in ambulatory activity in obese mice supplemented with l-citrulline, which could also explain the increase in whole-body oxygen consumption rates, thereby affecting the mRNA expression of these genes as a secondary response.
The reasons why obesity does not decrease the expression of these genes in soleus muscle is not clear, but it could be due to the fact that as a much more oxidative red muscle, the soleus has adaptive mechanisms in place to offset the detrimental actions of obesity on oxidative gene expression. It should be noted that the l-citrulline-mediated increase in whole-body oxygen consumption rates was relatively mild despite being statistically significant, and was only seen during the initial hours upon the transition to the dark cycle of the mouse. As skeletal muscle accounts for ∼20–30% of resting energy expenditure (Zurlo, Larson, Bogardus, & Ravussin, 1990), we posit that l-citrulline's actions on skeletal muscle, in particular white skeletal muscle (e.g. gastrocnemius), may contribute to the increase in whole-body oxygen consumption rates we observed in obese mice.
We initially surmised that our in vivo observations in mice were due to direct actions on skeletal muscle, as studies have shown that l-citrulline acts directly on skeletal muscle myocytes in vitro, where it can enhance muscle protein synthesis and prevent wasting (Ham et al., 2015; Le Plenier et al., 2017). However, we observed no direct effects of l-citrulline treatment on mitochondrial respiration in permeabilized gastrocnemius muscle fibres. It is worth noting that all mice in these specific experiments were killed with sodium pentobarbital (Euthanyl), and it has been demonstrated that perfusion of isolated rat hearts with sodium pentobarbital can inhibit mitochondrial oxygen consumption (Bhayana, Alto, & Dhalla, 1980). Conversely, it has been demonstrated that high-dose pentobarbital for inducing death does not depress mitochondrial energetics in rat hearts (Takaki, Nakahara, Kawatani, Utsumi, & Suga, 1997). Nonetheless, direct comparisons between isolated hearts perfused with sodium pentobarbital vs. animals injected with high-dose sodium pentobarbital for the purpose of rapid killing and tissue extraction is challenging. Moreover, all mice from all experimental groups received a similar dose of sodium pentobarbital, and hence the potential depressive actions of sodium pentobarbital on mitochondrial respiration would be similar between groups.
In further support of l-citrulline lacking direct actions on skeletal muscle to influence mitochondrial function, differentiated C2C12 myotubes treated with l-citrulline demonstrated no changes in Ppargc1a, Nrf1, Tfam or Tfbm2 mRNA expression. This suggests that perhaps indirect actions account for the skeletal muscle phenotype we observed in obese mice supplemented with l-citrulline. l-Citrulline supplementation has been proposed as a novel approach to increase circulating arginine concentrations since arginine can be recycled from citrulline (Bahri et al., 2013), while l-arginine is a precursor for NO synthesis, and NO has been shown in numerous studies to improve mitochondrial function (Le Gouill et al., 2007; Tengan et al., 2012). We therefore assessed whether increases in l-arginine could account for our in vivo observations following l-citrulline supplementation of obese mice, but treatment with pharmacological levels of l-arginine also had no effect on Ppargc1a, Nrf1, Tfam or Tfbm2 mRNA expression. In contrast, by increasing l-arginine levels and augmenting NO synthesis, l-citrulline could improve exercise tolerance via vasodilatory mechanisms that increase blood flow and nutrient delivery to the skeletal muscle (Clark et al., 2003). Future studies are needed to discern the indirect mechanisms by which l-citrulline prevents the obesity-mediated decline in the expression of genes regulating mitochondrial function in white muscle, and whether this affects aerobic exercise performance.
Importantly, it has been suggested in numerous studies that interventions leading to increased mitochondrial function can protect against obesity-induced dysglycaemia (Dube et al., 2008; Gupte et al., 2009; Seth et al., 2007). Our results support this premise, as we observed that obese mice supplemented with l-citrulline had a significant decrease in blood glucose levels in response to a GTT. These findings are consistent with previous studies in Zucker Diabetic fatty rats, whereby 4 weeks of supplementation with 63% watermelon pomace juice lowered blood glucose levels, though changes in glucose and insulin tolerance were not assessed, nor was it confirmed whether these actions were due to increased l-citrulline levels (Wu et al., 2007). Our observed improvements in glucose tolerance were associated with increased circulating insulin levels, consistent with previous findings demonstrating that both l-citrulline and l-arginine can act directly on β-cells to induce insulin secretion (Giugliano et al., 1997; Henquin & Meissner, 1981; Nakata & Yada, 2003). While these actions of l-citrulline/l-arginine in response to l-citrulline supplementation may explain the improved glucose tolerance we observed in obese mice, other actions on peripheral tissues (e.g. liver) contributing to glycaemic control also need to be considered. For example, it has been demonstrated in healthy volunteers that those individuals who experienced the largest increase in circulating citrulline levels in response to oral l-arginine administration also had the greatest suppression of hepatic glucose production (Apostol & Tayek, 2003). Although we did not assess the hepatic actions of l-citrulline supplementation, we did assess whether l-citrulline may improve glucose homeostasis in obese mice via improving muscle insulin sensitivity, but to our surprise, l-citrulline supplementation worsened insulin tolerance in obese mice. This worsening was associated with impaired insulin-stimulated Akt phosphorylation in gastrocnemius but not soleus muscle, which was also unexpected since gastrocnemius muscle was where we observed a prevention of the obesity-induced decline in Ppargc1a, Nrf1, Tfam or Tfbm2 mRNA expression. It is worth noting that despite numerous studies postulating that enhancing mitochondrial function can improve insulin sensitivity and glycaemic control, others have suggested that increasing oxidative metabolism in the absence of elevated energy demand can actually overload mitochondria and impair insulin sensitivity and overall glucose homeostasis (Koves et al., 2008; Muoio & Neufer, 2012). Although the mechanism by which l-citrulline supplementation impairs muscle insulin signalling/sensitivity remains unknown, it may involve mechanistic target of rapamycin complex 1 (mTORC1) signalling, as l-citrulline has been shown to activate mTORC1 and its downstream target p70S6 kinase (Le Plenier et al., 2017). Moreover, it has been reported that amino acid infusion increases activation of p70S6 kinase and inhibits insulin receptor substrate 1 (IRS-1) via phosphorylation on multiple serine residues, which prevents Akt activation and induces insulin resistance (Tremblay et al., 2005). Similar to what we observed with our mRNA expression profiles, l-citrulline's action on muscle insulin signalling/sensitivity is likely indirectly mediated and not the result of increasing arginine concentrations, as we did not observe impaired insulin-stimulated Akt phosphorylation in differentiated C2C12 myotubes treated with either l-citrulline or l-arginine.
Taken together, our study supports findings from previous studies implicating the nutraceutical l-citrulline as an aerobic performance enhancer, with new findings demonstrating that these actions are preserved in obesity, and that l-citrulline also attenuates obesity-induced dysglycaemia. Nevertheless, the deterioration in insulin sensitivity following l-citrulline supplementation in obese mice suggests that its broad use as a potential nutraceutical should be minimized, particularly in obese subjects.
ACKNOWLEDGEMENTS
The authors are grateful for the technical expertise and assistance of Mrs Amy Barr, who operates the Oxymax comprehensive lab animal monitoring system in the Animal Physiology Core Facility of the Cardiovascular Research Centre at the University of Alberta.
COMPETING INTERESTS
The authors have no conflicts to disclose.
AUTHOR CONTRIBUTIONS
A.E. and J.R.U. contributed to conception and design of the work, the acquisition, analysis and interpretation of the data, and drafted and revised the work for important intellectual content. R.A., K.L.H., A.M.D., K.G., A.A.G., I.Z., H.A., F.E., E.E.M., J.E.C. and J.M.S. contributed to the acquisition, analysis and interpretation of the data, and revised the work for important intellectual content. J.R.U. is the guarantor of this work and had full access to all the data, and takes responsibility for the integrity of the data and accuracy of the data analysis. All authors have read and approved the final version of this manuscript and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All persons designated as authors qualify for authorship, and all those who qualify for authorship are listed.
