Serum metabolomics reveals one-carbon metabolism differences between lean and obese cats not affected by L-carnitine or choline supplementation

Abstract

The supplementation of choline and L-carnitine in obese cats has garnered attention as a potential method for preventing and treating feline hepatic lipidosis (FHL). Providing dietary choline above the current recommended allowance to overweight and obese cats may have positive effects on one-carbon metabolism and hepatic lipid mobilization. Research on the metabolomic effects of L-carnitine supplementation in cats, however, remains limited. This study aimed to investigate the individual effects of choline and L-carnitine supplementation on the fasted serum metabolomic profiles of obese (n = 9; body condition score [BCS]: 8-9/9) and lean (n = 9; BCS: 4-5/9) adult male neutered cats fed at maintenance energy requirements. Cats were fed a commercial extruded cat food top-dressed with choline (6 x National Research Council recommended allowance: 378 mg/kg BW^0.67), L-carnitine (200 mg/kg BW), or control (no supplement) in a 3 × 3 complete Latin square design for 6 wk per treatment, with a 2-wk washout between each treatment period. The cats were fed once daily, and BW and BCS were assessed weekly. Fasted serum metabolites were analyzed at the end of each treatment period using direct infusion mass spectrometry (DI-MS) and liquid chromatography-mass spectrometry (LC-MS). The data were analyzed using SAS with proc GLIMMIX, considering group and period as random effects, and treatment, body condition, and their interaction as fixed effects. Statistical significance was set at P < 0.05, and Tukey’s post-hoc test was used for multiple comparisons when significance was observed. Obese cats had greater concentrations of s-adenosylhomocysteine, cysteine, cystine, reduced glutathione, and oxidized glutathione (GSSG), suggestive of alterations in one-carbon metabolism with obesity. The oxidation of fatty acids may have improved with both L-carnitine and choline supplementation. While choline and L-carnitine independently affected concentrations of betaine, GSSG, and decarboxylated S-adenosylmethionine, respectively, neither supplement broadly altered one-carbon metabolism. The present study suggests that dysfunction in one-carbon metabolism should be taken into consideration when examining the pathogenesis and increased FHL risk in obese cats.

Metabolomics revealed the potential effects of L-carnitine and choline supplementation on fatty acid oxidation, while distinct one-carbon metabolite profiles between lean and obese cats highlight possible metabolic differences in the one-carbon cycle that warrant further investigation into feline obesity and the pathogenesis of hepatic lipidosis.

Introduction

An estimated 63% of domestic cats in developed countries are considered overweight or obese (Cave et al., 2012; Öhlund et al., 2018; Arena et al., 2021; Chiang et al., 2022). The high prevalence of obesity in cats is a cause for concern due to the various secondary health issues that these animals are at risk of developing (Scarlett and Donoghue, 1998; Teng et al., 2018; Chiang et al., 2022). One of these health concerns is feline hepatic lipidosis (FHL), characterized by excessive accumulation of lipids in the liver (Hall et al., 1997). Although the precise physiological mechanisms leading to FHL are unclear, evidence suggests that affected cats accumulate hepatic lipids as a result of the mobilization of fatty acids (FA) from adipose tissue (Hall et al., 1997). It is believed that obese cats subjected to high degrees of dietary energy restriction are at the highest risk of developing FHL (Biourge et al., 1994; Kuzi et al., 2017). However, even during weight maintenance, obese cats already exhibit a greater accumulation of hepatic triglycerides (TAG) as compared to lean cats (Clark et al., 2013). Furthermore, the endocrinological disturbances associated with obesity, including altered insulin and glucagon dynamics, may contribute to an increased mobilization of FA from adipose tissue to the liver. Specifically, insulin resistance, in conjunction with elevated counter-regulatory hormones, can promote hormone-sensitive lipase activity and non-esterified fatty acid release, increasing hepatic FA uptake (Biourge et al., 1997; Holm, 2008; Brown et al., 2010).

Under normal post-absorptive conditions, particularly when insulin is low and glucagon is elevated, mobilized FA taken up by the liver undergo β-oxidation or are re-packed into TAG for export out of the liver as very-low-density lipoproteins (VLDL) (Fritz, 1963; Nguyen et al., 2008). However, in the case of FHL, the accumulation of TAG occurs within hepatocytes (Valtolina et al., 2005). While increased β-oxidation may help reduce hepatic lipid accumulation, if fatty acids are predominantly re-esterified into TAG that cannot be efficiently exported as VLDL, lipid accumulation and hepatic dysfunction may persist. Therefore, there is interest among researchers and the pet food industry to identify dietary supplements that can support increased β-oxidation and hepatic TAG export. Specifically, dietary carnitine and choline have been suggested as potential supplements for the prevention and/or treatment of FHL (Verbrugghe and Bakovic, 2013).

Limited research exists on the effects of dietary choline supplementation on feline hepatic health. However, greater intake of dietary choline, at five to six times the recommended allowance (RA) published by the National Research Council (NRC; 63 mg/kg BW^0.67) (NRC, 2006), in overweight and obese cats resulted in greater concentrations of serum lipids and lipoproteins (Verbrugghe et al., 2021; Rankovic et al., 2022a), which may reflect enhanced lipid packaging and export from the liver and/or reduced peripheral utilization. Similar findings have been reported in other species, including humans and rodents, where choline supplementation supports hepatic lipid export and prevents fatty liver development (Best and Hunstman, 1935; Zeisel et al., 1991; Walkey et al., 1998; Fischer et al., 2007). Choline is a precursor for the formation of phosphatidylcholine (PC), necessary for the packaging and mobilization of cholesterol (CHOL) and TAG into VLDL (Yao and Vance, 1988), and cats will develop fatty liver when consuming diets deficient in choline (da Silva et al., 1959; Anderson et al., 1979; Schaeffer et al., 1982). Furthermore, choline contributes to the production of betaine, which participates in the one-carbon cycle as a methyl donor to remethylate homocysteine to methionine (Finkelstein, 1990). This reaction generates S-adenosylmethionine (SAMe), a key methyl donor for multiple metabolic pathways, including the methylation of phosphatidylethanolamine to form PC and the synthesis of carnitine. By supporting this one-carbon pathway, dietary choline ensures adequate PC for VLDL assembly and TAG export, while also contributing to fatty acid oxidation through carnitine synthesis. Indeed, in overweight adult cats, a choline intake of six times the published RA increased serum concentrations of one-carbon metabolites, PC, and various amino acids (Rankovic et al., 2022b). Collectively, these findings suggest that choline may have beneficial effects on hepatic health and the maintenance of lean muscle mass.

The inclusion of L-carnitine in feline weight control and weight loss diets by the pet food industry is a common practice aimed primarily at supporting FA oxidation. However, there is a lack of comprehensive research on the effects of L-carnitine on the circulating metabolic profile of cats. As L-carnitine plays a crucial role in facilitating the entry of FA into the mitochondria for β-oxidation, which is the main pathway for the disposal of FA (Fritz, 1963), it has been a supplement of interest to FHL. In humans, impaired β-oxidation has been implicated as a possible cause of non-alcoholic fatty liver disease (NAFLD) (Wei et al., 2008). However, research on the involvement and benefits of L-carnitine supplementation in FHL has yielded inconsistent results (Jacobs et al., 1990; Blanchard et al., 2002; Aroch et al., 2012; Gorman et al., 2016).

To gain a better understanding and make comparisons of the pathways in which dietary choline and L-carnitine participate, direct flow injection mass spectrometry (DI-MS) and liquid chromatography-mass spectrometry (LC-MS) can be utilized for the quantification of serum metabolites, including those involved in the one-carbon and folate cycle. It was hypothesized that both choline and L-carnitine supplementation would increase FA oxidation in obese cats, and that choline would enhance one-carbon metabolism in both lean and obese cats. Therefore, the objective of this study was to investigate the differences in serum metabolites and biochemical pathways between obese and lean adult cats when consuming dietary choline, L-carnitine, or no supplement.

Materials and Methods

Samples for this study were collected as part of a previous study investigating serum lipid and lipoprotein profiles, energy expenditure, respiratory quotient, and body composition with choline supplementation in obese and lean cats (Rankovic et al., 2023). The University of Guelph Animal Care Committee (AUP#4496) approved the procedures regarding the care and use of animals for this research, according to provincial and national guidelines.

Animals and housing

Male neutered domestic shorthair (n = 18) cats were used for this research (Marshall’s Bio Resources, Waverly, NY, United States of America). The cats were 1 to 2 yr of age at the start of the trial (mean ± SD: 2.00 ± 0.33 yr; range: 1.28–2.29 yr). Half of the cats (n = 9) were classified obese at the start of the trial with an assigned body condition score (BCS) of ≥ 8/9 (Laflamme, 1997), and a mean body weight (BW) of 6.46 ± 0.45 kg. The remaining half (n = 9) were classified lean (BCS: 4-5/9), with a mean BW of 4.62 ± 0.45 kg. Apart from obesity, all cats were deemed healthy based on medical history, physical examination, and results of a complete blood count (CBC) and serum biochemistry, completed within 6 mo of the start of the trial.

The cats were group-housed in a free-living environment (23 ft × 19 ft) at the Animal Biosciences Cattery at the Ontario Agricultural College of the University of Guelph (Guelph, ON, Canada). Water was available ad libitum as both still (bowls) and flowing (open tap). Humidity and temperature within the room were controlled throughout the trial at 40% and 24 °C, respectively. The lights were controlled for a 12 h light 12 h dark cycle (on at 0700 h, off at 1900 h). Details regarding daily cleaning and enrichment were as previously described by Frayne et al. (2019).

Diet

Cats were individually placed in cages for one hour daily (08:00h) to allow for individual feeding. Cats consumed a commercial extruded diet (Nutram Total Grain-Free Chicken and Turkey Recipe, Elmira Pet Products, Elmira, ON, Canada) for 4 wk before the start of the trial (adaptation period) and during the trial. The diet was formulated for feline adult maintenance according to the published nutrient profiles of the Association of American Feed Control Officials (AAFCO, 2021). Food was provided based on diet history, at an amount determined to allow for maintenance of current BW. Individual orts were measured and recorded daily. Fasted BW and BCS were assessed and recorded weekly.

Measurements of proximate and nutrient analyses were performed as previously described in Rankovic et al. (2022a). Dietary choline and carnitine concentrations were determined via enzymatic colorimetry and liquid chromatography-mass spectrometry, respectively, as described by the Association of Official Analytical Chemists (AOAC 999.14 and 2012.17, respectively) (Horwitz et al., 1970). The commercial diet contained 4284 mg choline/kg diet dry matter basis (DMB) and 46 mg carnitine/kg diet DMB.

Study design and supplementation

The cats were divided into three groups of six, with three obese and three lean cats per group. The three groups were balanced for BW. A replicated 3 × 3 Latin square design was used, where cats received each of the three treatments (choline, carnitine, and control) for 6 wk per treatment. Between each 6-wk treatment period, there was a 2-wk washout where the cats only consumed the commercial feline diet. The cats continued to receive their food in quantities to maintain BW during the washout periods.

Prior to daily feeding, the L-carnitine supplement (48.5% L-carnitine tartrate, Bill Barr & Company, Overland Park, KS, USA) was pre-measured and top-dressed with a small volume of water onto ¼ of the cats’ daily food intake and soaked for 20 min. The L-carnitine was supplemented at 200 mg/kg BW daily (Center et al., 2012; Gooding et al., 2016). The commercial diet used was not supplemented with L-carnitine, and the measured amount present in the diet was negligible (46 mg carnitine/kg diet); therefore, only the supplemented L-carnitine was considered in intake calculations. The choline supplement (PuraChol, 70% aqueous choline chloride, 52.2% choline ion; Balchem Corporation, Montvale, NJ, United States of America) was similarly pre-measured, top-dressed, and soaked for 20 min with the first ration daily. The choline was supplemented at six times the RA for dietary choline published by the NRC [6 × 63 mg/kg BW^0.67 (378 mg/kg BW^0.67)] (NRC, 2006; Rankovic et al., 2022a, 2022b), taking into account the estimated choline intake from the commercial extruded food. Daily choline and L-carnitine intakes and calculations are as previously described by Rankovic et al. (2023). The doses of choline and L-carnitine fed in the present study were based on previously published research in cats. Dietary choline at five and six times the NRC RA fed to overweight and obese cats increased serum lipid and lipoprotein concentrations, thought to represent increased hepatic lipid mobilization (Verbrugghe et al., 2021; Rankovic et al., 2022a, 2023). The L-carnitine dose of 200 mg/kg BW was selected based on prior studies showing beneficial effects on body composition and fat oxidation in overweight cats (Center et al., 2000, 2012; Shoveller et al., 2014). Supplemental L-carnitine accelerated fat loss during energy restriction, supported lean mass retention, and increased fatty acid oxidation, particularly in overweight cats. Once the first ration (1/4 daily food intake) with the assigned treatment was consumed, cats were provided the second ration (3/4 daily food intake), which consisted of only kibble. Cats on the control treatment and during washout received their food as-is (no supplement or top-dressing) divided into two rations similar to the L-carnitine and choline treatment groups.

Blood collection and serum metabolomic profile analyses

Cats were administered both acepromazine (0.04 mg/kg BW) and hydromorphone (0.05 mg/kg BW) intramuscularly as pre-medication, and midazolam (0.3 mg/kg BW) and alfaxalone (1–3 mg/kg BW) intravenously via cephalic catheter for anesthetic induction, following a 24-h fast. Cats were anesthetized for the collection of blood and tissue samples. However, said tissue samples were not analyzed as part of the present study and are further described in Dobberstein et al. (2024). Following induction, jugular blood samples were collected (5 mL). The collected whole blood was immediately syringed into serum-separating tubes and stored at 5 °C until centrifugation (up to 10 h). Whole blood was centrifuged at 2500 g × 15 min at 4 °C (LegendRT, Kendro Laboratory Products 2002, Germany). Serum was separated, aliquoted, and stored at −80 °C until shipped on dry ice to The Metabolomics Innovation Center (TMIC) at the University of Alberta (Edmonton, AB, Canada) and the University of Victoria (Victoria, BC, Canada).

Analyses of serum metabolites occurred at the University of Alberta (Edmonton, AB, Canada) by direct-injection mass spectrometry (DI-MS) with a reverse-phase liquid chromatography-tandem mass spectrometry (LC-MS) custom assay, as previously described by Foroutan et al. (2020). This assay, in combination with a 4000 QTrap mass spectrometer (Applied Biosystems/MDS Analytical Technologies, Foster City, CA, United States of America) with an Agilent 1260 series ultra-high performance liquid chromatography (UHPLC) system (Agilent Technologies, Palo Alto, CA, United States of America), detected and quantified 134 metabolites within the serum samples (100 µL). Metabolites were grouped by class using the Human Metabolome Database (HMDB: http://www.hmdb.ca) into the following: biogenic amines; amino acids, amino acid derivatives and ammonium compounds; acylcarnitines; phosphatidylcholine diacyl (PC aa) and phosphatidylcholine acyl-alkyl (PC ae); lysophosphatidylcholines (LPC); sphingomyelins (SM) and hydroxysphingomyelins (HSM); and organic acids and sugars. A sum of the metabolite concentrations within each class was calculated to obtain a total for each class. In addition to total acylcarnitines, the following were also calculated: total short-chain (sum of C2 through C5), total medium-chain (sum of C6 through C12), and total long-chain acylcarnitines (sum of C14 through C18).

Detection and analysis of one-carbon and folate pathway metabolites was performed at the University of Victoria (Victoria, BC, Canada). For quantification of folate, methyltetrahydrofolic acid (5-MTHF), tetrahydrofolic acid (THF), and dihydrofolic acid (DHF), serum samples (40 μL) were vortexed for 1 min with an internal standard solution containing 13C4-folic acid and 13C4-5-MTHF (20 μL) and acetonitrile (140 μL). Following vortex-mixing, the sample solutions were placed on ice for 30 min and then centrifuged at 21,000 g × 10 min at 5 °C. The supernatant was collected, diluted 3-fold with water, and 10 μL of resultant solution was injected into a Waters Acquity UPLC system coupled to a Sciex QTRAP 6500 Plus mass spectrometer operated in positive-ion mode. A C18 UPLC column (2.1 × 100 mm, 1.7 μm) was used for LC separation, with 0.1% formic acid in water and 0.1% formic acid in acetonitrile as the binary solvents for the gradient elution at 40 °C and 0.3mL/min. For quantification of the other one-carbon metabolites, 40 μL of serum was vortexed for 1 min with a 20-μL of an internal standard solution containing isotope-labelled cysteine, glutathione, glutathione oxidized (GSSG), cysteine, and s-adenosylhomocysteine (SAH), and 140 μL of acetonitrile. The sample solutions were placed on ice for 30 min before being centrifuged at 21,000 g × 10 min. Following a 4-fold dilution with water, 10 μL aliquots of said solution were injected onto an Agilent 1290 UHPLC system coupled to an Agilent 6495B triple-quadrupole mass spectrometer operated in positive-ion mode. Similarly, LC separation was done on a C18 UPLC column (2.1 × 100 mm, 1.6 μm), with an ammonium formate buffer and methanol as the binary solvents for the gradient elution at 40 °C and 0.25 mL/min. Concentrations of the identified metabolites were calculated by internal standard calibration by interpolating the constructed linear-regression curves of individual metabolites with the analyte-to-internal standard peak ratios measured from the sample solutions. The detected and quantified metabolites included: 5-MTHF, acetylcholine, cobalamin, cystathionine, cystine, cysteine, folate, GSSG, glutathione reduced (GSH), homoserine, pyridoxal, pyridoxamine, pyridoxine, riboflavin, SAH, SAMe, decarboxylated SAMe (dcSAMe), and THF.

Statistical analyses

Statistical analysis of the serum metabolomic profile was performed in SAS (SAS Studio 3.8, SAS Institute, Cary, NC, United States). The normality of the residuals was assessed using the Shapiro-Wilk test and the visual assessment of scatter plots. Homogeneity of variance was evaluated by visual inspection of residual versus predicted value plots for patterns of heteroscedasticity. Acetylcholine, alpha ketoglutaric acid, betaine, choline, creatine, C7DC, glutamine, histidine, isobutyric acid, methionine, pyridoxal, pyridoxine, total amino acids, and total medium-chain acylcarnitines were not normally distributed, and a log transformation was applied. The log-transformed data were back-transformed to obtain the least square means (LSM) for each metabolite.

The GLIMMIX procedure within SAS was used to analyze serum metabolite concentrations. Treatment, “body condition” (obese or lean), and treatment × body condition were included as the fixed effects, period and group as the random effects, and cat as the subject. The covariance matrix resulting in the smallest Akaike information criterion value was used. A P-value < 0.05 was considered significant, and a P-value of < 0.10 was considered a trend. When a significant effect of treatment, body condition, or body condition × treatment interaction was present, a Tukey’s post hoc test was performed for multiple comparisons. Results are expressed as LSM ± SEM.

Two-way heatmaps with Euclidean distance measures and Ward clustering algorithm were created in Metaboanalyst 5.0 of serum metabolites with a significant effect of treatment, body condition, and/or treatment and body condition interaction. Mathematical transformation (log transformation) and data normalization using mean centering were applied to the data within Metaboanalyst 5.0 for normalization and were visually inspected via density and box plots.

Results

One obese cat was removed during period three due to a medical condition unrelated to the trial. Data from said cat were not included for period three (choline treatment). All other cats (n = 17) remained healthy on the assigned treatments and throughout the trial. Supplements were accepted by the cats with no known adverse effects.

As presented in Rankovic et al. (2023), food intake and energy intake were greater in the obese cats as compared to the lean cats (PCondition < 0.0001), as expected. Obese cats consumed an average of 66 (± 1.98) grams and 255 (± 7.60) kcal, in comparison to the lean cats, who consumed 53 (± 1.57) grams and 203 (± 6.05) kcal. Choline supplementation lowered food and energy intake when compared to the control treatment (225 vs. 232 kcal and 59 vs. 60 grams, respectively; PTreatment = 0.025). However, there was no effect of treatment × body condition interaction on food intake (PCondition × Treatment = 0.669).

Metabolomic differences can be observed with body condition, treatment, and/or body condition × treatment interaction in the generated heatmaps (Figures 1 and 2) and are described in the text below.

Figure 1.

Heatmap with Euclidean distance and Ward clustering of mean serum metabolites determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving dietary choline supplementation (378 mg/kg BW^0.67), dietary L-carnitine supplementation (200 mg/kg BW) and control (no additional dietary supplement), in a 3 × 3 Latin square design for 6-wk periods, with a significant effect of body condition (*), treatment (**), and/or body condition × treatment (***) interaction (P < 0.05). Metabolites are separated into a) biogenic amines, amino acids, amino acid derivatives, ammonium compounds, organic sugars and acids; b) acylcarnitines; c) phosphatidylcholines, lysophosphatidylcholines, sphingomyelins and hydroxysphingomyelins.

Figure 2.

Heatmap with Euclidean distance and Ward clustering of mean serum one-carbon and folate metabolites determined by LC-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving dietary choline supplementation (378 mg/kg BW^0.67), dietary L-carnitine supplementation (200 mg/kg BW) and control (no additional dietary supplement), in a 3 × 3 Latin square design for 6-wk periods, with a significant effect of body condition (*), treatment (**), and/or body condition × treatment (***) interaction (P < 0.05). dcSAMe: S-adenosylmethionine decarboxylated; SAH: s-adenosylhomocysteine; GSH: glutathione reduced; GSSG: glutathione oxidized.

DI-MS

Metabolites quantified by DI-MS are presented in Tables 1–7. Heatmaps of the 37 DI-MS metabolites with a significant effect of body condition, treatment, and/or body condition × treatment interaction (P < 0.05) are presented in Figure 1.

Table 1.

Mean serum concentrations (μM) of biogenic amines determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

Amines	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
Acetylornithine	Lean	32.4 ± 4.4	36.2 ± 4.4	36.9 ± 4.4	0.408	0.712	0.465
	Obese	40.9 ± 4.4	34.9 ± 4.6	40.5 ± 4.4
Alpha amino adipic acid	Lean	2.93 ± 0.23	2.45 ± 0.23	2.57 ± 0.23	0.219	0.750	0.365
	Obese	2.85 ± 0.23	3.04 ± 0.24	2.89 ± 0.23
Asymmetric dimethylarginine	Lean	1.74 ± 0.14	1.67 ± 0.14	1.75 ± 0.14	0.349	0.770	0.856
	Obese	1.79 ± 0.14	1.84 ± 0.15	1.93 ± 0.14
Carnosine	Lean	43.9 ± 6.7	53.5 ± 6.7	33.2 ± 6.7	0.041	0.919	0.025
	Obese	26.8 ± 6.7	22.6 ± 7.1	41.6 ± 6.7
Creatinine	Lean	102 ± 7	108 ± 7	94.0 ± 7.0	0.488	0.332	0.014
	Obese	85.6 ± 7.0	105 ± 7	120 ± 7
Kynurenine	Lean	3.69 ± 0.41	3.81 ± 0.41	3.40 ± 0.41	0.968	0.973	0.347
	Obese	3.53 ± 0.41	3.38 ± 0.43	3.95 ± 0.41
Methionine sulfoxide	Lean	3.46 ± 0.42	3.26 ± 0.42	4.32 ± 0.42	0.976	0.416	0.115
	Obese	4.14 ± 0.42	3.52 ± 0.44	3.42 ± 0.42
Putrescine	Lean	0.91 ± 0.08	1.01 ± 0.08	0.89 ± 0.08	0.331	0.977	0.514
	Obese	1.01 ± 0.08	0.95 ± 0.09	1.02 ± 0.08
Sarcosine	Lean	5.14 ± 0.67	5.57 ± 0.67	5.90 ± 0.67	0.544	0.883	0.749
	Obese	5.89 ± 0.67	6.08 ± 0.71	5.66 ± 0.67
Serotonin	Lean	4.61 ± 1.71	6.01 ± 2.23	4.86 ± 1.80	0.686	0.872	0.462
	Obese	6.45 ± 2.39	3.77 ± 1.45	4.53 ± 1.68
Spermidine	Lean	0.17 ± 0.01	0.17 ± 0.01	0.16 ± 0.01	0.154	0.080	0.200
	Obese	0.21 ± 0.01	0.16 ± 0.02	0.20 ± 0.01
Spermine	Lean	0.13 ± 0.01	0.13 ± 0.01	0.13 ± 0.01	0.360	0.190	0.258
	Obese	0.14 ± 0.01	0.12 ± 0.01	0.14 ± 0.01
Total dimethylarginine	Lean	2.33 ± 0.19	2.83 ± 0.19	2.39 ± 0.19	0.274	0.458	0.693
	Obese	2.36 ± 0.19	2.47 ± 0.20	2.75 ± 0.19
Trans-4-hydroxyproline	Lean	22.4 ± 3.0	19.6 ± 3.0	19.5 ± 3.0	0.480	0.799	0.708
	Obese	18.7 ± 3.0	19.5 ± 3.1	18.6 ± 3.0
Trimethylamine N-oxide	Lean	33.7 ± 18.2	31.9 ± 18.2	39.6 ± 18.2	0.105	0.051	0.103
	Obese	23.0 ± 18.2	61.3 ± 19.3	106 ± 18
Total biogenic amines	Lean	263 ± 33	277 ± 33	251 ± 33	0.470	0.175	0.053
	Obese	236 ± 33	267 ± 35	354 ± 33

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Table 7.

Mean serum concentrations (μM) of organic acids and sugars determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
3-Hydroxyphenyl-hydracrylic acid	Lean	0.010 ± 0.001	0.011 ± 0.001	0.011 ± 0.001	0.813	0.106	0.742
	Obese	0.011 ± 0.001	0.010 ± 0.001	0.012 ± 0.001
4-Hydroxyhippuric acid	Lean	0.052 ± 0.005	0.045 ± 0.005	0.048 ± 0.005	0.625	0.059	0.637
	Obese	0.050 ± 0.005	0.046 ± 0.005	0.044 ± 0.005
Alpha-ketoglutaric acid	Lean	2.54 ± 0.18	2.48 ± 0.18	2.71 ± 0.18	0.402	0.091	0.615
	Obese	2.63 ± 0.18	2.45 ± 0.19	3.03 ± 0.18
Beta-hydroxybutyric acid	Lean	78.8 ± 14.5	81.9 ± 14.5	94.8 ± 14.5	0.614	0.911	0.467
	Obese	82.5 ± 14.5	83.8 ± 15.0	75.3 ± 14.5
Butyric acid	Lean	0.90 ± 0.15	0.71 ± 0.15	0.76 ± 0.15	0.896	0.424	0.832
	Obese	0.81 ± 0.15	0.73 ± 0.15	0.78 ± 0.15
Citric acid	Lean	240 ± 15	224 ± 15	227 ± 15	0.570	0.615	0.248
	Obese	205 ± 15	221 ± 16	245 ± 15
Fumaric acid	Lean	1.35 ± 0.23	1.52 ± 0.23	1.38 ± 0.23	0.830	0.262	0.189
	Obese	1.16 ± 0.23	1.25 ± 0.24	1.94 ± 0.23
Glucose	Lean	4939 ± 360	4576 ± 360	5218 ± 360	0.199	0.186	0.841
	Obese	5559 ± 360	4736 ± 383	5512 ± 360
Hippuric Acid	Lean	3.21 ± 0.69	3.27 ± 0.69	3.58 ± 0.69	0.707	0.956	0.837
	Obese	3.50 ± 0.69	3.03 ± 0.72	3.02 ± 0.69
Homovanillic Acid	Lean	0.030 ± 0.004	0.033 ± 0.004	0.029 ± 0.004	0.178	0.519	0.622
	Obese	0.032 ± 0.004	0.038 ± 0.004	0.039 ± 0.004
Indole acetic acid	Lean	0.35 ± 0.05	0.33 ± 0.05	0.25 ± 0.05	0.347	0.767	0.320
	Obese	0.22 ± 0.05	0.30 ± 0.06	0.29 ± 0.05
Isobutyric acid	Lean	0.78 ± 0.55	0.74 ± 0.52	0.98 ± 0.69	0.637	0.827	0.986
	Obese	0.85 ± 0.60	0.99 ± 0.73	1.26 ± 0.89
Lactic acid	Lean	1423 ± 254	1254 ± 2254	1671 ± 254	0.178	0.018	0.384
	Obese	1574 ± 254	1294 ± 269	2372 ± 254
Methylmalonic acid	Lean	0.066 ± 0.008	0.051 ± 0.008	0.056 ± 0.008	0.194	0.845	0.373
	Obese	0.044 ± 0.008	0.051 ± 0.008	0.056 ± 0.008
Propionic acid	Lean	1.12 ± 0.19	1.27 ± 0.19	1.15 ± 0.19	0.496	0.924	0.463
	Obese	1.15 ± 0.19	1.07 ± 0.19	1.12 ± 0.19
Pyruvic acid	Lean	43.8 ± 7.4	34.4 ± 7.4	53.2 ± 7.4	0.058	0.010	0.751
	Obese	53.7 ± 7.4	44.2 ± 7.9	72.3 ± 7.4
Succinic acid	Lean	0.94 ± 0.12	1.16 ± 0.12	1.05 ± 0.12	0.392	0.082	0.085
	Obese	0.94 ± 0.12	0.94 ± 0.13	1.47 ± 0.12
Uric acid	Lean	10.4 ± 2.0	10.9 ± 2.0	11.1 ± 2.0	0.348	0.349	0.390
	Obese	12.7 ± 2.0	9.62 ± 2.1	14.9 ± 2.0

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Biogenic amines

Serum biogenic amines are presented in Table 1. Total concentration of serum biogenic amines tended to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.053) Carnosine concentrations were affected by a body condition × treatment interaction (P_{Condition x Treatment} = 0.025), but no differences occurred when a Tukey’s posthoc adjustment was applied. Similarly, obese cats consuming L-carnitine had the highest creatinine concentrations (P_{Condition x Treatment} = 0.014), yet there were no differences following a Tukey’s posthoc adjustment. There was a tendency for treatment to differ spermidine and trimethylamine N-oxide (TMAO) concentrations (P_Treatment = 0.080, and 0.051, respectively). However, no effect of body condition or body condition × treatment interaction was noted on either spermidine or TMAO (P_Condition = 0.154, and 0.105, respectively; P_{Condition x Treatment} = 0.200, and 0.103, respectively). No differences occurred with body condition, treatment, and/or body condition × treatment interaction in the serum concentrations of the other biogenic amines determined by DI-MS (P > 0.10), including acetylornithine, alpha amino adipic acid, asymmetric dimethylarginine, kynurenine, methionine sulfoxide, putrescine, sarcosine, serotonin, spermine, total dimethylarginine, and trans-4-hydroxyproline.

Amino acids, amino acid derivatives, and ammonium compounds

Concentrations of serum amino acids, amino acid derivatives, and ammonium compounds are presented in Table 2. Overall, serum amino acid concentrations tended to differ with body condition (P_Condition = 0.060), but did not differ with treatment or body condition × treatment (P_Treatment = 0.900, P_{Condition X Treatment} = 0.881). Of the individual amino acids, aspartic acid was higher in obese cats (P_Condition = 0.007), with a tendency to differ with body condition × treatment interaction (P_{Condition X Treatment} = 0.067). Additionally, serum alanine, glycine, and methionine tended to differ with body condition (P_Condition = 0.077, 0.055, and 0.092, respectively) but did not differ with treatment or body condition × treatment interaction (P_Treatment = 0.415, 0.372, and 0.379, respectively; P_{Condition X Treatment} = 0.188, 0.547, and 0.117, respectively). There was a tendency for glutamic acid to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.061). The remaining serum amino acids analyzed by DI-MS were not affected by body condition, treatment, and/or body condition × treatment interaction (P > 0.10), including arginine, asparagine, citrulline, glutamine, histidine, homocysteine, isoleucine, leucine, lysine, ornithine, phenylalanine, proline, serine, taurine, threonine, tyrosine, tryptophan, and valine.

Table 2.

Mean serum concentrations (μM) of amino acids, amino acid derivatives, and ammonium compounds determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
Amino Acids
Alanine	Lean	440 ± 52	445 ± 52	418 ± 52	0.077	0.415	0.188
	Obese	500 ± 52	452 ± 54	622 ± 52
Arginine	Lean	134 ± 11	122 ± 11	129 ± 11	0.102	0.238	0.574
	Obese	151 ± 11	141 ± 11	134 ± 11
Asparagine	Lean	46.5 ± 4.7	50.5 ± 4.7	44.7 ± 4.7	0.178	0.495	0.955
	Obese	51.6 ± 4.7	54.2 ± 4.9	50.8 ± 4.7
Aspartic Acid	Lean	27.6 ± 3.1	18.2 ± 3.1	21.2 ± 3.1	0.007	0.581	0.067
	Obese	26.2 ± 3.1	30.1 ± 3.3	33.2 ± 3.1
Citrulline	Lean	18.5 ± 1.4	18.7 ± 1.4	18.8 ± 1.4	0.996	0.319	0.289
	Obese	19.6 ± 1.4	16.6 ± 1.4	19.7 ± 1.4
Glutamic Acid	Lean	61.4 ± 4.2	61.3 ± 4.2	58.1 ± 4.2	0.203	0.297	0.061
	Obese	64.2 ± 4.2	58.6 ± 4.5	74.4 ± 4.2
Glutamine	Lean	600 ± 42	673 ± 47	681 ± 47	0.242	0.182	0.919
	Obese	648 ± 45	737 ± 54	706 ± 49
Glycine	Lean	307 ± 20	268 ± 20	270 ± 20	0.055	0.372	0.547
	Obese	320 ± 20	314 ± 21	316 ± 20
Histidine	Lean	92.2 ± 5.1	97.7 ± 5.4	99.0 ± 5.5	0.153	0.792	0.713
	Obese	104 ± 6	103 ± 6	104 ± 6
Homocysteine	Lean	9.54 ± 0.87	10.27 ± 0.87	10.11 ± 0.87	0.431	0.473	0.833
	Obese	9.87 ± 0.87	10.66 ± 0.87	11.39 ± 0.87
Isoleucine	Lean	48.3 ± 5.1	57.2 ± 5.1	58.2 ± 5.1	0.985	0.767	0.209
	Obese	57.4 ± 5.1	53.1 ± 5.3	53.5 ± 5.1
Leucine	Lean	88.1 ± 9.7	104 ± 10	109 ± 10	0.974	0.773	0.147
	Obese	107 ± 10	97.4 ± 10.2	97.4 ± 9.7
Lysine	Lean	157 ± 14	163 ± 14	159 ± 14	0.179	0.715	0.526
	Obese	190 ± 14	173 ± 15	168 ± 14
Methionine	Lean	27.8 ± 2.6	32.8 ± 3.1	35.5 ± 3.3	0.092	0.379	0.117
	Obese	37.6 ± 3.5	34.9 ± 3.9	36.0 ± 3.4
Ornithine	Lean	19.8 ± 1.8	19.3 ± 1.8	18.4 ± 1.8	0.174	0.528	0.975
	Obese	22.2 ± 1.8	21.8 ± 1.9	20.2 ± 1.8
Phenylalanine	Lean	56.2 ± 4.4	60.8 ± 4.4	57.4 ± 4.4	0.523	0.697	0.607
	Obese	58.3 ± 4.4	58.8 ± 4.7	64.3 ± 4.4
Proline	Lean	129 ± 11	122 ± 11	123 ± 11	0.385	0.866	0.804
	Obese	130 ± 11	129 ± 11	136 ± 11
Serine	Lean	114 ± 11	110 ± 11	99.8 ± 11.2	0.423	0.195	0.838
	Obese	124 ± 11	110 ± 12	108 ± 11
Taurine	Lean	63.4 ± 7.0	62.3 ± 7.0	60.2 ± 7.0	0.543	0.318	0.165
	Obese	59.6 ± 7.0	57.3 ± 7.4	79.7 ± 7.0
Threonine	Lean	104 ± 8	102 ± 8	111 ± 8	0.302	0.733	0.633
	Obese	118 ± 8	110 ± 9	111 ± 8
Tyrosine	Lean	40.7 ± 4.1	43.8 ± 4.1	40.5 ± 4.1	0.558	0.599	0.350
	Obese	40.9 ± 4.1	41.5 ± 4.3	49.0 ± 4.1
Tryptophan	Lean	49.9 ± 4.1	51.9 ± 4.1	49.5 ± 4.1	0.452	0.752	0.359
	Obese	53.8 ± 4.1	48.3 ± 4.3	56.3 ± 4.1
Valine	Lean	131 ± 13	143 ± 13	143 ± 13	0.838	0.735	0.519
	Obese	145 ± 13	133 ± 14	147 ± 13
Total Amino Acids	Lean	2759 ± 146	2832 ± 150	2815 ± 149	0.060	0.900	0.881
	Obese	3043 ± 162	2996 ± 168	3113 ± 165
Amino Acid Derivatives
Betaine	Lean	239 ± 59	452 ± 105	251 ± 59	0.717	0.031	0.867
	Obese	258 ± 64	400 ± 99	213 ± 50
Creatine	Lean	8.37 ± 1.54	11.3 ± 2.1	8.33 ± 1.53	0.841	0.803	0.345
	Obese	11.0 ± 2.0	8.55 ± 1.66	9.13 ± 1.68
Methylhistidine	Lean	26.7 ± 2.4	27.2 ± 2.4	26.6 ± 2.4	0.210	0.375	0.288
	Obese	28.1 ± 2.4	27.9 ± 2.6	33.9 ± 2.4
Ammonium Compounds
Choline	Lean	7.68 ± 1.07	9.96 ± 1.38	7.52 ± 1.04	0.639	0.856	0.272
	Obese	7.84 ± 1.09	7.14 ± 1.05	8.59 ± 1.19

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Of the amino acid derivatives and ammonium compounds, only serum betaine differed with treatment (P_Treatment = 0.031). Serum betaine concentrations were higher with the choline treatment compared to the L-carnitine treatment, but did not differ from control. Similarly, serum betaine did not differ between the L-carnitine treatment and control. Body condition and body condition × treatment interaction did not affect betaine concentrations (P_Condition = 0.717, P_{Condition X Treatment} = 0.867). No differences occurred in serum creatine, methylhistidine, or choline with body condition, treatment, or body condition × treatment interaction (P_Condition = 0.841, 0.210, and 0.639, respectively; P_Treatment = 0.803, 0.375, and 0.856, respectively; P_{Condition X Treatment} = 0.345, 0.288, and 0.272).

Acylcarnitines

Concentrations of serum acylcarnitines identified by DI-MS are presented in Table 3. The calculated total concentration of serum acylcarnitines differed with treatment (P_Treatment < 0.0001), with higher total acylcarnitines with L-carnitine as compared to both choline and control, but no difference between the choline treatment and control. However, body condition and body condition × treatment interaction did not affect total serum acylcarnitine concentrations (P_Condition = 0.723, P_{Condition X Treatment} = 0.120). Similarly, free acylcarnitine (C0) concentrations were greatest with the L-carnitine treatment (P_Treatment < 0.0001), compared to choline and control; but did not differ with body condition or body condition × treatment interaction (P_Condition = 0.705, P_{Condition x Treatment} = 0.124). L-carnitine resulted in higher concentrations of total short-chain acylcarnitines compared to choline or control (P_Treatment < 0.0001). However, body condition and body condition × treatment interaction did not affect total short-chain acylcarnitine concentrations (P_Condition = 0.707; P_{Condition x Treatment} = 0.165). Of the individual short-chain acylcarnitines, C2, C3, C4, and C5:1 were similarly greatest with L-carnitine, when compared to both choline and control (P_Treatment = < 0.0001, 0.008, 0.0001, and 0.004, respectively), yet again body condition and body condition × treatment interaction did not affect C2, C3, C4, and C5:1 (P_Condition = 0.682, 0.532, 0.917, and 0.531, respectively; P_{Condition x Treatment} = 0.181, 0.676, 0.868, and 0.191, respectively). Concentrations of serum C5 were highest with the L-carnitine treatment, compared to the other two treatments (P_Treatment = 0.004). Specifically, C5 was greater in obese cats consuming L-carnitine than in obese cats consuming the other two treatments (P_{Condition x Treatment} = 0.040). Serum C4:1 concentrations tended to differ with body condition (P_Condition = 0.079), but were not affected by treatment or body condition × treatment interaction (P_Treatment = 0.620; P_{Condition x Treatment} = 0.186). No differences were noted for the remaining short-chain acylcarnitines with body condition, treatment, or body condition × treatment interaction (P > 0.010), including C3 OH, C3:1, C4 OH, C5 OH, C5 DC, C5 MDC, and C5:1 DC.

Table 3.

Mean serum concentrations (μM) of acylcarnitines determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
Free acylcarnitines (C0)	Lean	16.8 ± 4.5	16.7 ± 4.5	26.9 ± 4.5	0.705	<0.0001	0.124
	Obese	12.5 ± 4.5	16.2 ± 4.7	37.6 ± 4.5
Short Chain Acylcarnitines
C2	Lean	2.28 ± 0.76	2.20 ± 0.76	4.36 ± 0.76	0.682	<0.0001	0.181
	Obese	1.48 ± 0.76	2.06 ± 0.78	6.26 ± 0.76
C3	Lean	0.15 ± 0.027	0.12 ± 0.027	0.20 ± 0.027	0.532	0.008	0.676
	Obese	0.10 ± 0.027	0.11 ± 0.028	0.20 ± 0.027
C3 OH	Lean	0.050 ± 0.003	0.050 ± 0.003	0.051 ± 0.003	0.929	0.700	0.488
	Obese	0.053 ± 0.003	0.048 ± 0.003	0.049 ± 0.003
C3:1	Lean	0.029 ± 0.001	0.028 ± 0.001	0.026 ± 0.001	0.873	0.551	0.333
	Obese	0.028 ± 0.001	0.028 ± 0.001	0.028 ± 0.001
C4	Lean	0.13 ± 0.03	0.10 ± 0.03	0.21 ± 0.03	0.917	0.0001	0.868
	Obese	0.11 ± 0.03	0.10 ± 0.03	0.22 ± 0.03
C4 OH	Lean	0.033 ± 0.002	0.033 ± 0.002	0.038 ± 0.002	0.901	0.190	0.842
	Obese	0.033 ± 0.002	0.034 ± 0.002	0.036 ± 0.002
C4:1	Lean	0.035 ± 0.002	0.036 ± 0.002	0.034 ± 0.002	0.079	0.620	0.186
	Obese	0.038 ± 0.002	0.035 ± 0.002	0.041 ± 0.002
C5	Lean	0.19 ± 0.04Ac	0.13 ± 0.04Ac	0.20 ± 0.04Ac	0.741	0.004	0.040
	Obese	0.11 ± 0.04Ad	0.15 ± 0.04Ad	0.30 ± 0.04Ac
C5 OH	Lean	0.043 ± 0.002	0.041 ± 0.002	0.040 ± 0.002	0.986	0.555	0.116
	Obese	0.037 ± 0.002	0.043 ± 0.002	0.045 ± 0.002
C5 DC	Lean	0.032 ± 0.002	0.030 ± 0.002	0.032 ± 0.002	0.661	0.841	0.492
	Obese	0.028 ± 0.002	0.032 ± 0.002	0.031 ± 0.002
C5 MDC	Lean	0.047 ± 0.002	0.044 ± 0.002	0.046 ± 0.002	0.203	0.320	0.439
	Obese	0.043 ± 0.002	0.041 ± 0.002	0.046 ± 0.002
C5:1	Lean	0.041 ± 0.002	0.036 ± 0.002	0.041 ± 0.002	0.531	0.004	0.191
	Obese	0.037 ± 0.002	0.033 ± 0.002	0.044 ± 0.002
C5:1 DC	Lean	0.026 ± 0.003	0.027 ± 0.003	0.022 ± 0.003	0.613	0.460	0.450
	Obese	0.026 ± 0.003	0.026 ± 0.003	0.26 ± 0.003
Total short chain acylcarnitines	Lean	3.08 ± 0.88	2.89 ± 0.88	5.30 ± 0.88	0.707	<0.0001	0.165
	Obese	2.13 ± 0.88	2.77 ± 0.93	7.33 ± 0.88
Medium Chain Acylcarnitines
C6	Lean	0.057 ± 0.005Ac	0.055 ± 0.005Ac	0.055 0.005Ac	0.495	0.087	0.034
	Obese	0.046 ± 0.005Ad	0.049 ± 0.005Acd	0.063 0.005Ac
C6:1	Lean	0.035 ± 0.002	0.034 ± 0.002	0.031 ± 0.002	0.303	0.607	0.698
	Obese	0.032 ± 0.002	0.031 ± 0.002	0.032 ± 0.002
C7 DC	Lean	0.18 ± 0.04Ac	0.075 ± 0.017Acd	0.068 ± 0.016Ad	0.411	0.063	0.038
	Obese	0.081 ± 0.019Ac	0.066 ± 0.016Ac	0.10 ± 0.02Ac
C8	Lean	0.046 ± 0.004Ac	0.042 ± 0.004Ac	0.046 ± 0.004Ac	0.889	0.005	0.016
	Obese	0.039 ± 0.004Ad	0.035 ± 0.004Ad	0.063 ± 0.004Ac
C9	Lean	0.038 ± 0.003	0.035 ± 0.003	0.036 ± 0.003	0.320	0.754	0.830
	Obese	0.039 ± 0.003	0.039 ± 0.003	0.037 ± 0.003
C10	Lean	0.11 ± 0.003	0.11 ± 0.003	0.10 ± 0.003	0.101	0.752	0.045
	Obese	0.12 ± 0.003	0.11 ± 0.003	0.12 ± 0.003
C10:1	Lean	0.20 ± 0.007	0.19 ± 0.007	0.21 ± 0.007	0.355	0.261	0.243
	Obese	0.20 ± 0.007	0.19 ± 0.007	0.20 ± 0.007
C10:2	Lean	0.11 ± 0.007	0.10 ± 0.007	0.11 ± 0.007	0.209	0.914	0.847
	Obese	0.11 ± 0.007	0.11 ± 0.007	0.11 ± 0.007
C12	Lean	0.074 ± 0.005	0.066 ± 0.005	0.075 ± 0.005	0.482	0.048	0.628
	Obese	0.073 ± 0.005	0.070 ± 0.005	0.084 ± 0.005
C12 DC	Lean	0.11 ± 0.004Ac	0.093 ± 0.004Ad	0.096 ± 0.004Acd	0.744	0.165	0.017
	Obese	0.096 ± 0.004Ac	0.099 ± 0.004Ac	0.099 ± 0.004Ac
C12:1	Lean	0.074 ± 0.003	0.069 ± 0.003	0.074 ± 0.003	0.380	0.122	0.655
	Obese	0.069 ± 0.003	0.067 ± 0.003	0.074 ± 0.004
Total medium chain acylcarnitines	Lean	1.12 ± 0.07	0.89 ± 0.05	0.91 ± 0.05	0.308	0.095	0.067
	Obese	0.91 ± 0.05	0.86 ± 0.05	0.99 ± 0.06
Long Chain Acylcarnitines
C14	Lean	0.045 ± 0.003	0.041 ± 0.003	0.047 ± 0.003	0.431	0.009	0.361
	Obese	0.043 ± 0.003	0.043 ± 0.003	0.053 ± 0.003
C14:1	Lean	0.064 ± 0.006	0.065 ± 0.006	0.079 ± 0.006	0.150	0.026	0.912
	Obese	0.071 ± 0.006	0.071 ± 0.007	0.090 ± 0.006
C14:1 OH	Lean	0.030 ± 0.002	0.029 ± 0.002	0.032 ± 0.002	0.361	0.006	0.181
	Obese	0.029 ± 0.002	0.030 ± 0.002	0.037 ± 0.002
C14:2	Lean	0.022 ± 0.002	0.018 ± 0.002	0.022 ± 0.002	0.812	0.046	0.273
	Obese	0.018 ± 0.002	0.018 ± 0.002	0.025 ± 0.002
C14:2 OH	Lean	0.030 ± 0.001	0.030 ± 0.001	0.029 ± 0.001	0.779	0.886	0.733
	Obese	0.029 ± 0.001	0.029 ± 0.001	0.030 ± 0.001
C16	Lean	0.16 ± 0.01	0.16 ± 0.01	0.20 ± 0.01	0.673	0.006	0.801
	Obese	0.17 ± 0.01	0.16 ± 0.02	0.20 ± 0.01
C16 OH	Lean	0.11 ± 0.004	0.11 ± 0.004	0.12 ± 0.004	0.330	0.448	0.652
	Obese	0.11 ± 0.004	0.11 ± 0.004	0.12 ± 0.004
C16:1	Lean	0.059 ± 0.005	0.064 ± 0.005	0.072 ± 0.005	0.206	0.004	0.583
	Obese	0.055 ± 0.005	0.065 ± 0.005	0.082 ± 0.005
C16:1 OH	Lean	0.037 ± 0.002	0.036 ± 0.002	0.038 ± 0.002	0.340	0.041	0.119
	Obese	0.037 ± 0.002	0.034 ± 0.003	0.045 ± 0.002
C16:2	Lean	0.029 ± 0.002	0.026 ± 0.002	0.029 ± 0.002	0.645	0.106	0.731
	Obese	0.028 ± 0.002	0.027 ± 0.002	0.031 ± 0.002
C16:2 OH	Lean	0.018 ± 0.001	0.020 ± 0.001	0.018 ± 0.001	0.614	0.728	0.791
	Obese	0.019 ± 0.001	0.020 ± 0.001	0.019 ± 0.001
C18	Lean	0.082 ± 0.008	0.083 ± 0.008	0.10 ± 0.008	0.377	0.014	0.863
	Obese	0.092 ± 0.008	0.085 ± 0.008	0.11 ± 0.008
C18:1	Lean	0.11 ± 0.012	0.13 ± 0.012	0.15 ± 0.012	0.364	0.037	0.756
	Obese	0.13 ± 0.012	0.14 ± 0.013	0.15 ± 0.012
C18:1 OH	Lean	0.031 ± 0.002	0.031 ± 0.002	0.034 ± 0.002	0.452	0.322	0.067
	Obese	0.034 ± 0.002	0.029 ± 0.002	0.029 ± 0.002
C18:2	Lean	0.045 ± 0.005	0.051 ± 0.005	0.063 ± 0.005	0.328	0.026	0.862
	Obese	0.053 ± 0.005	0.056 ± 0.006	0.065 ± 0.005
Total long chain acylcarnitines	Lean	0.87 ± 0.05	0.90 ± 0.053	1.02 ± 0.053	0.415	0.008	0.884
	Obese	0.93 ± 0.05	0.91 ± 0.056	1.09 ± 0.053
Total acylcarnitines	Lean	22.0 ± 5.3	21.3 ± 5.3	34.1 ± 5.3	0.723	<0.0001	0.120
	Obese	16.5 ± 5.3	20.4 ± 5.6	47.0 ± 5.3

Values expressed as LSM ± SEM. Down a column, different upper case letter superscripts (A,B), represent significant difference between body conditions within a treatment, where a p-value of less than 0.05 is considered significant; Across a row, different lower case letter superscripts (c,d), represent significant difference between treatments within a body condition, where a p-value of less than 0.05 is considered significant; Repeated measures ANOVA with Tukey post-hoc test. DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Concentrations of total medium-chain acylcarnitines tended to differ with body condition × treatment interaction (P_Treatment = 0.095; P_{Condition x Treatment} = 0.067). Concentrations of C8 were greatest in obese cats consuming L-carnitine, compared to obese cats consuming control or choline (P_{Condition x Treatment} = 0.016). Similarly, serum C6 was greater in obese cats consuming L-carnitine than in obese cats consuming control (P_{Condition x Treatment} = 0.034). A body condition × treatment interaction effect was also observed for serum C7:DC, C12:DC, and C10. Serum concentrations of C7:DC were lower in lean cats when consuming L-carnitine as compared to control (P_{Condition x Treatment} = 0.038). Lean cats consuming choline had lower concentrations of serum C12: DC concentrations than lean cats consuming control (P_{Condition x Treatment} = 0.017). Although numerically C10 was greatest in obese cats consuming control and in obese cats consuming L-carnitine, there was no difference between body condition × treatment means when a Tukey’s posthoc adjustment was applied (P_{Condition x Treatment} = 0.045). Additionally, serum C12 was lower with choline than L-carnitine treatment (P_Treatment = 0.048). However, serum C12 concentrations did not differ between choline and control, or L-carnitine and control. Overall, body condition alone did not affect serum concentrations of any individual medium-chain acylcarnitines (P_Condition > 0.10). Additionally, there were no differences in the serum concentrations of C6:1, C9, C10:1, C10:2, and C12:1 with treatment or body condition × treatment interaction (P_Treatment = 0.607, 0.754, 0.261, 0.914, and 0.122, respectively; P_{Condition x Treatment} = 0.698, 0.830, 0.243, 0.847, and 0.655, respectively).

Total long-chain acylcarnitine concentrations were higher with the L-carnitine as compared to the choline and control treatments (P_Treatment = 0.008); but were not affected by body condition or body condition × treatment interaction (P_Condition = 0.415_;P_{Condition x Treatment} = 0.884). Similarly, concentrations of C14, C14:1, C14:1 OH, C14:2, C16, C16:1, C16:1 OH, C18, C18:1 and C18:2 differed with treatment (P_Treatment < 0.05). Serum C14:1 OH, C16, C16:1, and C18 were higher with L-carnitine, as compared to both choline and control (P_Treatment = 0.006, 0.006, 0.004, and 0.014, respectively). Concentrations of serum C14:1, C18:1, and C18:2 were greater with L-carnitine than control (P_Treatment = 0.026, 0.037, and 0.026, respectively). L-carnitine also resulted in higher concentrations of serum C14 and C14:2, when compared with the choline treatment, but not compared to control (P_Treatment = 0.009, and 0.046, respectively), and also not between the choline treatment and control. None of the aforementioned individual long-chain acylcarnitines were affected by body condition or body condition × treatment interaction (P_Condition and P_{Condition x Treatment} > 0.10). However, there was a tendency for C18:1 OH to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.067). Concentrations of C14:2 OH, C16 OH, C16:2 and C16:2 OH did not differ with treatment, body condition or body condition × treatment interaction (P_Treatment = 0.886, 0.448, 0.106, and 0.728, respectively; P_Condition = 0.779, 0.330, 0.645, and 0.614, respectively; P_{Condition x Treatment} = 0.733, 0.652, 0.731, and 0.791, respectively).

Phosphatidylcholines

Concentrations of serum diacyl (PC aa) and acyl-alkyl (PC ae) PC are presented in Table 4. Total serum PC aa was greatest in obese cats (P_Condition < 0.0001) but did not differ with treatment (P_Treatment = 0.190) or body condition × treatment interaction (P_{Condition x Treatment} = 0.399). Of the PC aa metabolites, C32:2, C36:0, C38:6, and C40:6 concentrations were similarly higher in the obese cats, as compared to the lean cats (P_Condition = 0.017, 0.019, < 0.0001, and 0.0001, respectively) but did not differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.147, 0.135, 0.285, and 0.872, respectively). Of the aforementioned PC aa metabolites, only PC aa 38:6 had a tendency to differ with treatment (P_Treatment = 0.082). Serum PC aa C36:6, PC aa C38:0, PC aa C40:1 and PC aa C40:2 were unaffected by treatment, body condition, or body condition × treatment interaction (P_Treatment = 0.780, 0.650, 0.448, and 0.743, respectively; P_Condition = 0.229, 0.457, 0.357, and 0.398, respectively; P_{Condition x Treatment} = 0.368, 0.411, 0.191, and 0.148, respectively). Serum total PC ae, PC ae C36:0, and PC ae C40:6 concentrations all tended to differ with body condition (P_Condition = 0.051, 0.082, and 0.086, respectively), regardless of treatment (P_Treatment = 0.587, 0.809, and 0.477, respectively) or body condition × treatment interaction (P_{Condition x Treatment} = 0.325, 0.202, and 0.549, respectively).

Table 4.

Mean serum concentrations (μM) of PC aa and PC ae determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
PC aa
PC aa C32:2	Lean	3.05 ± 0.18	3.60 ± 0.18	3.22 ± 0.18	0.017	0.556	0.147
	Obese	3.77 ± 0.18	3.52 ± 0.19	3.46 ± 0.18
PC aa C36:0	Lean	13.5 ± 0.9	14.8 ± 0.9	15.2 ± 0.9	0.019	0.970	0.135
	Obese	17.2 ± 0.9	16.2 ± 0.9	15.3 ± 0.9
PC aa C36:6	Lean	0.41 ± 0.03	0.45 ± 0.03	0.45 ± 0.03	0.229	0.780	0.368
	Obese	0.48 ± 0.03	0.48 ± 0.03	0.44 ± 0.03
PC aa C38:0	Lean	1.92 ± 0.13	2.09 ± 0.13	1.97 ± 0.13	0.457	0.650	0.411
	Obese	2.17 ± 0.13	2.03 ± 0.14	1.92 ± 0.13
PC aa C38:6	Lean	34.45 ± 2.56	34.83 ± 2.56	33.03 ± 2.56	<0.0001	0.082	0.285
	Obese	44.37 ± 2.56	41.81 ± 2.65	35.48 ± 2.56
PC aa C40:1	Lean	0.49 ± 0.029	0.51 ± 0.029	0.51 ± 0.029	0.357	0.448	0.191
	Obese	0.58 ± 0.029	0.51 ± 0.031	0.50 ± 0.029
PC aa C40:2	Lean	0.93 ± 0.063	0.98 ± 0.063	1.00 ± 0.063	0.398	0.743	0.148
	Obese	1.11 ± 0.063	0.98 ± 0.067	0.95 ± 0.063
PC aa C40:6	Lean	26.97 ± 3.30	27.03 ± 3.30	24.24 ± 3.30	0.0001	0.279	0.872
	Obese	34.81 ± 3.30	33.36 ± 3.40	29.24 ± 3.30
Total PC aa	Lean	81.70 ± 6.37	84.27 ± 6.37	79.66 ± 6.37	<0.0001	0.190	0.399
	Obese	105 ± 6	98.96 ± 6.56	87.31 ± 6.37
PC ae
PC ae C36:0	Lean	1.63 ± 0.091	1.75 ± 0.091	1.72 ± 0.091	0.082	0.809	0.202
	Obese	1.98 ± 0.091	1.78 ± 0.097	1.76 ± 0.091
PC ae C40:6	Lean	1.83 ± 0.15	1.97 ± 0.15	1.83 ± 0.15	0.086	0.477	0.549
	Obese	2.15 ± 0.15	2.03 ± 0.15	1.85 ± 0.15
Total PC ae	Lean	3.45 ± 0.21	3.71 ± 0.21	3.54 ± 0.21	0.051	0.587	0.325
	Obese	4.12 ± 0.21	3.82 ± 0.22	3.61 ± 0.21

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. PC = phosphatidylcholine; aa = diacyl; ae = acyl-alkyl; DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Lysophosphatidylcholines

Serum concentrations of LPC analyzed by DI-MS are presented in Table 5. Total serum LPC concentrations were greatest in the obese cats (P_Condition = 0.048), but did not differ with treatment or body condition × treatment interaction (P_Treatment = 0.446; P_{Condition x Treatment} = 0.632). Obese cats had higher concentrations of LPC C16:1, LPC C18:0, and LPC C20:3 (P_Condition = 0.042, 0.020, and 0.027, respectively). Serum LPC C16:0, LPC C17:0, and LPC C28:1 concentrations also had a tendency to differ with body condition (P_Condition = 0.056, 0.061, and 0.098, respectively). Only serum LPC C20:3 and LPC C20:4 differed with treatment (P_Treatment = 0.028 and 0.007, respectively), where choline resulted in greater concentrations than L-carnitine. Serum LPC C24:0 also tended to differ with treatment (P_Treatment = 0.086). Additionally, there was no effect of body condition × treatment interaction for any of the LPC metabolites analyzed (P_{Condition x Treatment} > 0.10).

Table 5.

Mean serum concentrations (μM) of LPC determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
LPC
LPC C14:0	Lean	0.63 ± 0.04	0.59 ± 0.04	0.59 ± 0.04	0.173	0.807	0.866
	Obese	0.64 ± 0.04	0.64 ± 0.04	0.63 ± 0.04
LPC C16:0	Lean	39.4 ± 2.5	38.7 ± 2.5	38.2 ± 2.5	0.056	0.452	0.719
	Obese	44.0 ± 2.5	42.8 ± 2.6	39.1 ± 2.5
LPC C16:1	Lean	1.19 ± 0.11	1.20 ± 0.11	1.31 ± 0.11	0.042	0.768	0.866
	Obese	1.37 ± 0.11	1.42 ± 0.11	1.41 ± 0.11
LPC C17:0	Lean	1.07 ± 0.07	1.04 ± 0.07	1.04 ± 0.07	0.061	0.582	0.581
	Obese	1.17 ± 0.07	1.21 ± 0.07	1.08 ± 0.07
LPC C18:0	Lean	47.5 ± 3.0	45.9 ± 3.0	47.3 ± 3.0	0.020	0.613	0.659
	Obese	54.6 ± 3.0	51.8 ± 3.1	49.3 ± 3.0
LPC C18:1	Lean	13.8 ± 1.0	14.6 ± 1.0	14.3 ± 1.0	0.225	0.547	0.632
	Obese	15.5 ± 1.0	16.0 ± 1.1	14.2 ± 1.0
LPC C18:2	Lean	20.2 ± 1.6	20.6 ± 1.6	19.8 ± 1.6	0.410	0.278	0.604
	Obese	21.1 ± 1.6	23.0 ± 1.7	19.5 ± 1.6
LPC C20:3	Lean	0.89 ± 0.18	0.98 ± 0.18	0.61 ± 0.18	0.027	0.028	0.894
	Obese	1.04 ± 0.18	1.15 ± 0.18	0.86 ± 0.18
LPC C20:4	Lean	5.22 ± 0.44	5.57 ± 0.44	4.48 ± 0.44	0.058	0.007	0.704
	Obese	6.04 ± 0.44	6.48 ± 0.46	4.73 ± 0.44
LPC C24:0	Lean	0.31 ± 0.02	0.30 ± 0.02	0.27 ± 0.02	0.869	0.086	0.817
	Obese	0.31 ± 0.02	0.32 ± 0.02	0.27 ± 0.02
LPC C26:0	Lean	0.11 ± 0.01	0.12 ± 0.01	0.12 ± 0.01	0.188	0.685	0.787
	Obese	0.12 ± 0.01	0.14 ± 0.01	0.13 ± 0.01
LPC C26:1	Lean	0.064 ± 0.008	0.060 ± 0.008	0.067 ± 0.008	0.689	0.845	0.722
	Obese	0.069 ± 0.008	0.066 ± 0.008	0.063 ± 0.008
LPC C28:0	Lean	0.20 ± 0.01	0.17 ± 0.01	0.19 ± 0.01	0.990	0.410	0.337
	Obese	0.19 ± 0.01	0.19 ± 0.02	0.17 ± 0.01
LPC C28:1	Lean	0.19 ± 0.01	0.20 ± 0.01	0.19 ± 0.01	0.098	0.370	0.360
	Obese	0.24 ± 0.01	0.22 ± 0.02	0.20 ± 0.01
Total LPC	Lean	131 ± 8	130 ± 8	129 ± 8	0.048	0.446	0.632
	Obese	146 ± 8	145 ± 8	132 ± 8

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. LPC = lysophosphatidylcholine; DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Sphingomyelines and hydroxysphingomyelines

Concentrations of serum SM and HSM analyzed by DI-MS are presented in Table 6. Overall, neither body condition, treatment, or body condition × treatment interaction affected total serum HSM concentrations (P_Condition = 0.208; P_Treatment = 0.932; P_{Condition x Treatment} = 0.187). Individually, only HSM C14:1 tended to differ with body condition (P_Condition = 0.063). Treatment and body condition × treatment interaction did not affect any HSM metabolites analyzed (P_Treatment > 0.10; P_{Condition x Treatment} > 0.10). Total SM tended to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.075) but was unaffected by body condition or treatment individually (P_Condition = 0.718; P_Treatment = 0.502). Only SM C18:1 was affected by body condition × treatment interaction (P_{Condition x Treatment} = 0.047); however, there were no differences following a Tukey’s post hoc. Additionally, SM C18:1 did not differ with body condition or treatment (P_Condition = 0.296; P_Treatment = 0.365). The remaining SM metabolites did not differ with body condition, treatment, or body condition × treatment interaction (P_Condition > 0.10; P_Treatment > 0.10; P_{Condition x Treatment} > 0.10).

Table 6.

Mean serum concentrations (μM) of HSM and SM determined by DI-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
HSM
HSM C14:1	Lean	16.5 ± 1.2	19.0 ± 1.2	17.8 ± 1.2	0.063	0.521	0.245
	Obese	20.6 ± 1.2	19.8 ± 1.2	18.3 ± 1.2
HSM C16:1	Lean	6.66 ± 0.51	8.00 ± 0.51	7.04 ± 0.51	0.437	0.544	0.287
	Obese	7.79 ± 0.51	7.47 ± 0.54	7.43 ± 0.51
HSM C22:1	Lean	26.0 ± 1.6	28.3 ± 1.6	28.2 ± 1.6	0.282	0.916	0.188
	Obese	31.1 ± 1.6	27.4 ± 1.7	27.8 ± 1.6
HSM C22:2	Lean	11.3 ± 0.6	12.2 ± 0.6	12.0 ± 0.6	0.160	0.771	0.161
	Obese	13.7 ± 0.6	12.3 ± 0.7	11.9 ± 0.6
HSM C24:1	Lean	4.94 ± 0.32	5.24 ± 0.32	5.28 ± 0.32	0.322	0.650	0.285
	Obese	5.22 ± 0.32	4.44 ± 0.34	4.98 ± 0.32
Total HSM	Lean	65.3 ± 4.0	71.4 ± 4.2	70.3 ± 4.0	0.208	0.932	0.187
	Obese	78.4 ± 4.0	71.0 ± 4.2	40.5 ± 4.0
SM
SM C16:0	Lean	182 ± 11	213 ± 11	193 ± 11	0.846	0.542	0.114
	Obese	213 ± 11	193 ± 11	186 ± 11
SM C16:1	Lean	8.88 ± 0.51	10.57 ± 0.51	9.85 ± 0.51	0.103	0.580	0.662
	Obese	11.38 ± 0.51	10.31 ± 0.54	9.89 ± 0.51
SM C18:0	Lean	43.1 ± 3.1	52.7 ± 3.1	44.7 ± 3.1	0.449	0.432	0.116
	Obese	51.4 ± 3.1	47.5 ± 3.3	47.4 ± 3.1
SM C18:1	Lean	8.27 ± 0.61	10.62 ± 0.61	9.11 ± 0.61	0.296	0.365	0.047
	Obese	10.48 ± 0.61	9.62 ± 0.65	9.50 ± 0.61
SM C20:2	Lean	0.44 ± 0.04	0.48 ± 0.04	0.48 ± 0.04	0.153	0.896	0.463
	Obese	0.54 ± 0.04	0.51 ± 0.04	0.47 ± 0.04
Total SM	Lean	242 ± 15	287 ± 15	257 ± 15	0.718	0.502	0.075
	Obese	288 ± 15	260 ± 16	253 ± 15

Values expressed as LSM ± SEM. Repeated measures ANOVA with Tukey post-hoc test. HSM = Hydroxysphingomyelines; SM = sphingomyelines; DI-MS = direct infusion mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Organic acids and sugars

Organic acid and sugar concentrations are presented in Table 7. An effect of treatment was present for both lactic acid and pyruvic acid (P_Treatment = 0.018 and 0.010, respectively). L-carnitine treatment increased concentrations of both as compared to the choline treatment, but not compared to control. Lactic acid and pyruvic acid concentrations did not differ between the choline treatment and control. Additionally, pyruvic acid tended to differ with body condition (P_Condition = 0.058). Lactic acid did not differ with body condition (P_Condition = 0.178), and treatment × body condition interaction did not affect either lactic acid or pyruvic acid (P_{Condition x Treatment} = 0.384 and 0.751, respectively). Serum 4-hydroxyhippuric acid, alpha-ketoglutaric acid, and succinic acid all tended to differ between treatments (P_Treatment = 0.059, 0.091, and 0.082, respectively). Body condition did not affect serum 4-hydroxyhippuric acid, alpha-ketoglutaric acid, and succinic acid concentrations (P_Condition = 0.625, 0.402, and 0.392, respectively). However, succinic acid tended to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.085). Serum 4-hydroxyhippuric and alpha-ketoglutaric acid were unaffected by treatment × body condition interaction (P_{Condition x Treatment} = 0.637 and 0.615, respectively). Neither body condition, treatment or treatment × body condition interaction had an effect on the remaining serum organic acids and sugars (P > 0.10), including 3-hydroxyphenyl-hydracrylic acid, beta-hydroxybutyric acid, butyric acid, citric acid, fumaric acid, glucose, hippuric acid, homovanillic acid, indole acetic acid, isobutyric acid, methylmalonic acid, propionic acid and uric acid.

LC-MS

One-carbon and folate metabolites quantified by LC-MS are presented in Table 8. A heatmap of the 18 LC-MS metabolites with a significant effect of body condition, treatment, and/or body condition × treatment interaction (P < 0.05) is presented in Figure 2.

Table 8.

Mean serum concentrations (μM) of one-carbon and folate metabolites determined by LC-MS in lean cats (n = 9) and obese cats (n = 9 for control and L-carnitine; n = 8 for choline) receiving control (no additional dietary supplement), dietary choline supplementation (378 mg/kg BW^0.67) and dietary L-carnitine supplementation (200 mg/kg BW) in a 3 × 3 Latin square design for 6-wk periods.

	Body Condition (BC)	Treatment (T)			P-Values
		Control (n = 18)	Choline (n = 17)	L-Carnitine (n = 18)	BC	T	BC X T
5-methyltetrahydrofolic acid	Lean	0.024 ± 0.001	0.024 ± 0.001	0.023 ± 0.001	0.063	0.581	0.913
	Obese	0.027 ± 0.001	0.026 ± 0.001	0.027 ± 0.001
Acetylcholine	Lean	0.006 ± 0.002	0.007 ± 0.001	0.007 ± 0.001	0.526	0.038	0.098
	Obese	0.004 ± 0.001	0.009 ± 0.002	0.006 ± 0.001
Cobalamin	Lean	0.003 ± 0.0003	0.003 ± 0.0003	0.003 ± 0.0003	0.358	0.409	0.756
	Obese	0.003 ± 0.0003	0.003 ± 0.0003	0.003 ± 0.0003
Cystathionine	Lean	22.2 ± 2.0Ac	20.8 ± 2.0Acd	17.8 ± 2.0Ad	0.658	0.168	0.007
	Obese	19.4 ± 2.0Ac	22.8 ± 2.0Ac	21.8 ± 2.0Ac
Cystine	Lean	25.0 ± 2.1	21.9 ± 2.1	22.4 ± 2.1	0.024	0.197	0.438
	Obese	27.2 ± 2.1	27.1 ± 2.1	25.7 ± 2.1
Cysteine	Lean	17.3 ± 1.0	16.6 ± 1.0	18.7 ± 1.0	0.009	0.863	0.184
	Obese	19.4 ± 1.0	20.7 ± 1.0	19.0 ± 1.0
Folate	Lean	5.3 × 10−4 ± 1.5 × 10−5	5.0 × 10−4 ± 1.5 × 10−5	4.8 × 10−4 ± 1.5 × 10−5	0.064	0.132	0.799
	Obese	5.4 × 10−4 ± 1.5 × 10−5	5.2 × 10−4 ± 1.6 × 10−5	5.2 × 10−4 ± 1.5 × 10−5
Glutathione oxidized	Lean	0.029 ± 0.004	0.025 ± 0.004	0.029 ± 0.004	0.0003	0.034	0.198
	Obese	0.045 ± 0.005	0.043 ± 0.004	0.055 ± 0.004
Glutathione reduced	Lean	0.94 ± 0.13	0.91 ± 0.13	1.19 ± 0.13	0.0001	0.182	0.463
	Obese	1.39 ± 0.13	1.47 ± 0.13	1.49 ± 0.13
Homoserine	Lean	24.2 ± 1.1	25.0 ± 1.1	25.6 ± 1.1	0.670	0.437	0.807
	Obese	25.2 ± 1.1	25.2 ± 1.1	25.7 ± 1.1
Pyridoxal	Lean	0.081 ± 0.016	0.10 ± 0.02	0.080 ± 0.016	0.474	0.791	0.173
	Obese	0.12 ± 0.02	0.081 ± 0.017	0.099 ± 0.02
Pyridoxamine	Lean	0.002 ± 0.0001	0.001 ± 0.0001	0.002 ± 0.0001	0.042	0.021	0.351
	Obese	0.002 ± 0.0001	0.001 ± 0.0001	0.001 ± 0.0001
Pyridoxine	Lean	0.048 ± 0.017	0.044 ± 0.015	0.030 ± 0.010	0.463	0.201	0.755
	Obese	0.052 ± 0.018	0.076 ± 0.028	0.30 ± 0.011
Riboflavin	Lean	0.054 ± 0.006	0.051 ± 0.006	0.063 ± 0.006	0.025	0.368	0.417
	Obese	0.069 ± 0.006	0.069 ± 0.001	0.069 ± 0.006
S-adenosylhomocysteine	Lean	0.094 ± 0.018	0.11 ± 0.018	0.12 ± 0.018	0.022	0.317	0.802
	Obese	0.14 ± 0.018	0.14 ± 0.019	0.17 ± 0.018
S-adenosylmethionine	Lean	0.092 ± 0.009	0.090 ± 0.009	0.093 ± 0.009	0.247	0.931	0.874
	Obese	0.10 ± 0.01	0.10 ± 0.01	0.099 ± 0.009
S-adenosylmethionine decarboxylated	Lean	0.059 ± 0.006	0.056 ± 0.006	0.048 ± 0.006	0.220	0.008	0.613
	Obese	0.062 ± 0.006	0.066 ± 0.006	0.052 ± 0.006
Tetrahydrofolic acid	Lean	0.044 ± 0.007	0.037 ± 0.007	0.036 ± 0.007	0.064	0.608	0.449
	Obese	0.052 ± 0.007	0.056 ± 0.007	0.052 ± 0.007

Values expressed as LSM ± SEM. Down a column, different upper case letter superscripts (A,B), represent significant difference between body conditions within a treatment, where a p-value of less than 0.05 is considered significant; Across a row, different lower case letter superscripts (c,d), represent significant difference between treatments within a body condition, where a p-value of less than 0.05 is considered significant; Repeated measures ANOVA with Tukey post-hoc test. LC-MS = liquid chromatography-tandem mass spectrometry; BW = body weight; T = effect of treatment; BC = effect of body condition (lean or obese); BC X T = effect of body condition by treatment interaction.

Obese cats had higher concentrations of cystine, cysteine, GSH, GSSG, riboflavin, and SAH, as compared to lean cats (P_Condition = 0.024, 0.009, 0.0001, 0.0003, 0.025, and 0.022, respectively). Conversely, serum pyridoxamine was lower in obese cats than in lean cats (P_Condition = 0.042). Concentrations of serum 5-MTHF, folate, and THF tended to differ with body condition (P_Condition = 0.063, 0.064, and 0.064, respectively). Of the aforementioned metabolites, only serum GSSG and pyridoxamine differed between treatments (P_Treatment = 0.034 and 0.021, respectively). Serum concentrations of GSSG were greater with the L-carnitine treatment as compared to the choline treatment, but not compared to control, and did not differ between the choline treatment and control. Serum pyridoxamine concentrations were lower with the choline treatment than with the control treatment. There was no effect of body condition × treatment for any of the aforementioned one-carbon and folate metabolites (P_{Condition x Treatment} > 0.10). Serum acetylcholine concentrations differed between treatments, where choline produced higher concentrations than control (P_Treatment = 0.038). Additionally, acetylcholine concentrations tended to differ with body condition × treatment interaction (P_{Condition x Treatment} = 0.098). An effect of treatment was observed for dcSAMe, where concentrations were overall lower with L-carnitine than with choline or control (P_Treatment = 0.008). Lean cats consuming L-carnitine had lower concentrations of serum cystathionine as compared to lean cats consuming control (P_{Condition x Treatment} = 0.007). There were no differences with body condition, treatment, or body × condition treatment interaction for any of the remaining metabolites analyzed by LC-MS (P > 0.10), including cobalamin, homoserine, pyridoxal, pyridoxine, or SAMe.

Discussion

Previously, Rankovic et al. (2022b) found that overweight adult cats receiving dietary choline at six times the published RA for adult maintenance had greater serum concentrations of one-carbon metabolites, PC and several amino acids, as compared to choline at the published RA; suggesting the benefits of supplemental choline towards improving methyl status, hepatic lipid mobilization and protein synthesis in overweight cats. In contrast, while several studies have investigated the effects of dietary L-carnitine on energy expenditure and macronutrient utilization in cats (Center et al., 2000, 2012; Shoveller et al., 2014), there has been limited investigation into the metabolomic effects of L-carnitine (Blanchard et al., 2002). To the authors’ knowledge, there have been no published investigations into the serum metabolomic profiles of lean or obese cats with dietary L-carnitine supplementation.

Despite the high dietary choline intake in the choline treatment group, serum choline concentrations were not greater in the present study. This finding was expected as dietary choline is quickly metabolized (Fischer et al., 2010). The same has been reported in overweight cats, growing kittens and mink receiving supplemental dietary choline (Hedemann et al., 2012; Godfrey et al., 2022; Rankovic et al., 2022b). Upon absorption, choline can participate in one of four metabolic pathways: it can (i) be oxidized to betaine to participate as a methyl group donor in the one-carbon cycle; (ii) be phosphorylated to the phospholipid PC; (iii) be acetylated to the neurotransmitter acetylcholine, or (iv) participate in a base-exchange reaction that substitutes choline for the head groups of serine or ethanolamine on endogenous phospholipids (Zeisel, 1981). In the present study, serum betaine and acetylcholine were greatest with the choline treatment. However, total PC and LPC concentrations did not differ, except for LPC20:3 and LPC20:4, which were individually greater with choline supplementation. This contrasts previous findings where choline supplementation resulted in greater total and multiple individual species concentrations of PC and LPC in growing kittens and overweight cats (Godfrey et al., 2022; Rankovic et al., 2022b) and greater circulating VLDL in overweight and obese adult cats (Verbrugghe et al., 2021; Rankovic et al., 2022a). The lack of increase in LPC and PC with choline supplementation in the present study may have been due to differences in the duration of choline supplementation. In the present study, choline was supplemented for 6 wk, as compared to previous research where individual choline doses were supplemented for 3 wk per dose (Rankovic et al., 2022a). Chronic (150 mg/kg BW for 14 d) and acute (300 mg/kg BW for 1 d) choline supplementation in dogs resulted in different phospholipid synthesis (Di Luzio and Zilversmit, 1959). Hepatic phospholipid synthesis in dogs receiving the acute dose of choline increased by 108%, as compared to chronic choline supplementation. Di Luzio and Zilversmit (1959) suggested that the observed differences in phospholipid synthesis could be attributed to the concentrations of hepatic TAG. When hepatic TAG concentrations are not abundant, choline may not exert the same lipotropic benefits. As hepatic TAG concentrations were not evaluated herein, we can not conclude whether the same relationship exists in cats. Additionally, the metabolic fate of choline may have been influenced by differences in dietary composition, particularly variations in sulfur amino acids, methyl group donors, and B vitamins. Therefore, the impact of differing nutrient profiles on the lipotropic activity of choline in cats should be further investigated. Lastly, the cats in the present study received a different protocol of drugs preceding blood collection to allow for tissue sampling as part of a related study (results not included here). This protocol included alfaxalone and midazolam for anesthetic co-induction. The use of this combination in cats has seldom been studied (Lagos-Carvajal et al., 2019; Wheeler et al., 2021), and its impact on metabolomic profiles, either in cats or in any other species, has not to our knowledge, been investigated.

As expected, L-carnitine supplementation influenced pathways associated with β-oxidation, as evidenced by the greater concentrations of free, short-chain, and long-chain acylcarnitines, related to choline and control. Similar increases in free, short-chain, and long-chain acylcarnitine have been observed in rats supplemented with L-carnitine (Bacurau et al., 2003), and reflect the role of L-carnitine in transporting long-chain FA across the mitochondrial membrane for β-oxidation (McGarry and Foster, 1980). The β-oxidation of long-chain acylcarnitines eventually results in the production of short-chain acylcarnitines and coenzyme A (CoA) (Rebouche and Seim, 2002). Therefore, the greater long-chain acylcarnitine production may have facilitated greater production of CoA, in addition to the short-chain acylcarnitines. Additionally, the greater concentrations of GSSG with L-carnitine supplementation, as compared to choline, may be suggestive of greater ROS generation during β-oxidation. However, as no differences were observed relative to control, and given the sample size, the biological significance of this increase is unclear. Although ROS production during FA oxidation has been linked to the pathogenesis of NAFLD in humans (Rolo et al., 2012), there is no evidence thus far that it plays a role in the development of FHL, as reviewed by Ling et al. (2025).

L-carnitine increases enzyme activity within the tricarboxylic acid cycle (TCA) cycle in rodents (Ruiz-Pesini et al., 2001; Kumaran et al., 2005), along with increasing resting energy expenditure in overweight cats (Shoveller et al., 2014). The alterations in metabolic intermediates observed with L-carnitine supplementation, including elevated concentrations of free and long-chain acylcarnitines, suggest potential changes in β-oxidation activity, which may have contributed to the greater pyruvate and lactate concentrations. Enhanced β-oxidation generates acetyl-CoA, which could saturate the TCA cycle’s capacity under certain conditions, particularly if oxaloacetate availability is limited. An accumulation of acetyl-CoA may have inhibited pyruvate dehydrogenase activity, reducing the conversion of pyruvate to acetyl-CoA and leading to greater pyruvate. The greater pyruvate could then have been redirected to lactate via lactate dehydrogenase activity, contributing to the greater lactate concentrations observed. While these shifts in metabolic pathways highlight the intricate interplay between glycolysis and mitochondrial metabolism, no significant increases in TCA cycle intermediates, energy expenditure, or respiratory quotient were detected (Rankovic et al., 2023), suggesting that the overall metabolic impact of L-carnitine under these conditions requires further investigation.

Concentrations of serum C8 and C12 were lowest with choline treatment in the present study, whereas obese cats consuming L-carnitine had the highest concentrations of C6, C7:DC, and C8. The lower specific medium-chain acylcarnitine species parallels previous findings in kittens receiving supplemental dietary choline during growth (Godfrey et al., 2022). Medium-chain acylcarnitines are considered products of incomplete β-oxidation (Adams et al., 2009), and have been considered proinflammatory (Rutkowsky et al., 2014). In humans, plasma concentrations of C6, C8, and C12 were elevated in patients with type 2 diabetes, as compared to healthy non-diabetic controls (Adams et al., 2009; Mihalik et al., 2010). Similarly, overweight individuals with a higher visceral fat area had higher concentrations of C12 and C12:1, as compared to overweight individuals with a lower visceral fat area (Baek et al., 2017). In cats, Grant et al. (2024) did not observe a difference in C12 between lean and obese cats at weight maintenance. However, no other medium-chain acylcarnitines were analyzed. Therefore, apart from C12, the correlation between medium-chain acylcarnitines in obese cats has not been investigated. A positive correlation between obese, insulin-resistant rodents, and incomplete β-oxidation has been published (Koves et al., 2008). The significance of the lower concentrations of C8 and C12 with choline treatment in relation to FA oxidation is unclear, as the total concentration of medium-chain acylcarnitines tended to differ with treatment but did not reach statistical significance.

mApart from the effects of dietary treatment in the present study, lean and obese cats had differences in their serum metabolomic profiles, specifically in relation to one-carbon metabolism (Figure 3). Although statistical differences in metabolite concentrations were observed, the biological relevance of these differences could not be confirmed in the present study. It is important to note that statistically significant differences do not necessarily imply biological significance, and conversely, biologically meaningful differences may not reach statistical significance. These metabolomic differences have not previously been reported in cats (Grant et al., 2024), possibly due to the broader range of one-carbon and folate-related metabolites assessed in the present study. Obese cats in the present study had higher concentrations of SAH, the immediate precursor for homocysteine production (Selhub, 1999). It has been suggested that an elevation of SAH may be indicative of the inhibition of methylation, as SAH is a potent inhibitor of most methyltransferases (Carmel and Jacobsen, 2001). However, the present study did not assess metabolic flux or protein abundance and enzyme kinetics, which would be required to draw conclusions regarding functional activity within the one-carbon cycle. In cats, species-specific adaptations in one-carbon metabolism may influence these observations. Previous work by Ruaux et al. (2001) demonstrated that cats with cobalamin deficiency had elevated methylmalonic acid, indicative of impaired methylmalonyl-CoA mutase activity, but homocysteine concentrations remained within reference ranges. Similarly, McMichael et al. (2000) found that cats with cardiomyopathy and arterial thromboembolism had reduced vitamin B12 and B6 concentrations without hyperhomocysteinemia, suggesting that excess homocysteine is likely diverted through the transsulfuration pathway to produce cysteine and taurine. Similar to homocysteine, a greater concentration of SAH is considered a factor in the development of cardiovascular diseases in humans (Valli et al., 2008). There is currently no evidence of a similar association in cats. Elevated concentrations of SAH in humans are also associated with hepatocyte apoptosis and may be a factor in the development of alcohol-related liver diseases (Kharbanda et al., 2005; Arumugam et al., 2020, 2021). Similar to the obese cats in the present study, healthy obese children had greater concentrations of SAH, cysteine, cystine, and GSSG, as compared to healthy lean children (Barbosa et al., 2021). Homocysteine can undergo transsulfuration to produce cystathionine, which can later be converted into cysteine by cystathionase, using pyridoxal 5’-phosphate as cofactor (Selhub, 1999). The authors suggested that the observed differences in the obese children were demonstrative of metabolic disturbances with one-carbon metabolism, specifically regarding transmethylation and transsulfuration. Additionally, the greater circulating GSSG and cystine were presented as markers of oxidative stress and reduced antioxidant capacity in the obese children. These differences may have been a result of higher fasting insulin levels, which can reduce the mRNA expression of enzymes involved in one-carbon metabolism (Chiang et al., 2009). Although fasting insulin concentrations have previously been published to be greater in obese cats (Hoenig et al., 2007), insulin concentrations were not analyzed within the present study.

Figure 3.

Observed differences in mean serum metabolite concentrations related to the folate and one-carbon cycles between obese (n = 9) and lean adult cats (n = 9). Red arrows represent higher or lower serum metabolite concentrations in obese cats when compared to lean cats. Green arrows represent a hypothesized greater activity of a pathway in obese cats when compared to lean cats.

Additionally, obese cats in the present study had greater concentrations of several PC species. Specifically, concentrations of serum PC aa C32:2, PC aa C36:0, PC aa C38:6, and PC aa C40:6 were greater. Grant et al. (2024) similarly published greater concentrations of serum PC aa C32:1 and PC aa C38:5 in obese cats as compared to lean cats maintaining BW. Greater concentrations of circulating phospholipids are a common finding in obese humans (Anjos et al., 2019; Michael et al., 2021), where circulating PC aa C32:2 was positively associated with BW and total fat mass (Papandreou et al., 2021), and PC aa C40:6 was positively associated with waist circumference in women (Rauschert et al., 2016). Concentrations of PC aa C32:2 and PC aa C38:6 were also higher in obese children than in overweight children (Michael et al., 2021). Additionally, plasma PC aa C32:2 concentrations were considered a predictive biomarker of hepatic lipidosis in dairy cows, as concentrations were higher in cows with hepatic lipidosis than in healthy reference cows (Imhasly et al., 2014). Concentrations of PC aa C32:2 have not been evaluated in healthy cats as compared to cats diagnosed with FHL. Therefore, the relationship between this metabolite and hepatic lipid concentrations in cats remains unknown. However, as obese cats have previously been reported to have higher hepatic TAG concentrations as compared to lean cats (Clark et al., 2013), this potential association warrants further research. The higher concentrations of PC in the obese cats in the present study are not surprising, as these cats also had higher circulating concentrations of the lipoprotein VLDL as compared to the lean cats (Rankovic et al., 2023). Lipoproteins are considered to be the main source of circulating phospholipids, with PC specifically being the most abundant subclass of lipid within VLDL (Vance and Adeli, 2008). Indeed, plasma concentrations of PC aa were positively associated with VLDL concentrations in obese individuals (Boulet et al., 2015).

Given the metabolic differences observed between lean and obese cats, specifically regarding one-carbon metabolism, future studies should aim to enroll a larger number of obese cats to maximize statistical power and better assess the potential benefits of these dietary supplements, while adhering to the 3Rs of research. Although some metabolites in the present study showed treatment effects, the absolute changes for certain metabolites, such as pyridoxamine, were minimal, and their biological relevance warrants further investigation. Additional metabolomics research in cats is needed to understand how feline-specific metabolism compares to that of other species, particularly where the one-carbon cycle is concerned. Preliminary evidence suggests that some metabolite concentrations, such as serum acetylcholine, may differ between cats and other species such as mice (Wang et al., 2019), but broader, standardized comparisons are needed before conclusions can be drawn.

Additionally, serum in the present study was sampled only once in the fasted state. To better understand how choline and L-carnitine impact the biochemical pathways in cats, the addition of frequent postprandial blood samples should be considered, along with assessing metabolite concentrations within different tissues. Gene expression and enzyme activity, particularly within the TCA and one-carbon cycles, should be assessed to better understand which biochemical pathways were up- or down-regulated with choline or L-carnitine supplementation. It is unclear how the anesthetic agents used before blood collection in the present study may have affected the serum metabolomic profile of these cats, as compared to previous research investigating the effects of choline or obesity on serum metabolomics (Godfrey et al., 2022; Rankovic et al., 2022b; Grant et al., 2024). Therefore, direct comparisons between the present study and existing feline research should be done with caution.

Based on the results presented herein, both L-carnitine and choline supplementation may have individually improved FA oxidation in the present study, warranting further investigation during weight loss in obese cats. However, no clear differences in one-carbon metabolism were evident with either supplement. Although choline resulted in greater serum concentrations of the methyl group donor betaine, concentrations of methionine, SAMe, or any other one-carbon metabolites did not increase. While this may suggest that methylation and one-carbon metabolism were not upregulated with choline treatment, this cannot be confirmed without assessing metabolite flux and enzyme kinetics, including Vmax and Km, within the one-carbon cycle. It is important to note that the metabolomic analyses in this study were performed on serum, which provides a snapshot of extracellular metabolite concentrations but may not accurately represent intracellular metabolic activity or tissue-specific processes. Several key metabolites in one-carbon metabolism, including methionine and SAMe, may be more tightly regulated and variable within tissues, and differences in intracellular pools may not be apparent in extracellular fluids, such as serum. While direct tissue sampling in companion animals is challenging, practical alternatives could include performing similar studies in unsedated obese cats using serial blood sampling to assess metabolic dynamics or using stable isotope tracer studies to investigate metabolic flux in vivo. Similarly, while L-carnitine may have influenced DNA methyltransferase activity, the biological relevance of this in feline obesity and FHL remains unclear. Without knowing how gene expression and enzyme activity were up- or down-regulated by these dietary supplements, neither can be conclusively recommended for the prevention and/or treatment of FHL at this time. The present study identified potential metabolic differences in one-carbon metabolism between lean and obese cats that have previously been identified in obese children (Barbosa et al., 2021), but not in cats. Future research should investigate enzyme activity, gene expression, and the influence of insulin and inflammatory cytokines on one-carbon metabolism in obese cats. These insights are critical for understanding the role of one-carbon metabolism in the pathogenesis of FHL for evaluating the potential of L-carnitine and choline as therapeutic interventions.

Glossary

Abbreviations

5-MTHF

5-methyltetrahydrofolic acid

AAFCO

American Association of Feed Control Officials

AOAC

Association of Official Analytical Chemists

AUP

Animal Utilization Protocol

BCS

Body condition score

BW

Body weight

CBC

Complete blood count

CHOL

Cholesterol

CoA

Coenzyme A

dcSAMe

Decarboxylated s-adenosyl methionine

DHF

Dihydrofolic acid

DI-MS

Direct infusion mass spectrometry

DMB

Dry matter basis

FA

Fatty acids

FHL

Feline hepatic lipidosis

GSH

Glutathione reduced

GSSG

Glutathione oxidized

HMDB

Human Metabolome Database

HSM

Hydroxysphingomyelin

LC

Liquid chromatography

LC-MS

Liquid chromatography-mass spectrometry

LPC

Lysophosphatidylcholine

LSM

Least square means

NAFLD

Non-alcoholic fatty liver disease

NRC

National Research Council

PC

Phosphatidylcholine

PC aa

Phosphatidylcholine diacyl

PC ae

Phosphatidylcholine acyl-alkyl

RA

Recommended allowance

SAH

S-adenosyl homocysteine

SAMe

S-adenosyl methionine

SM

Sphingomyelin

TAG

Triglycerides

TCA

Tricarboxylic acid cycle

THF

Tetrahydrofolic acid

TMAO

Trimethylamine-N-oxide

TMIC

The Metabolomics Innovation Center

UHPLC

Ultra-high-performance liquid chromatography

UPLC

Ultra-performance liquid chromatography

VLDL

Very-low-density-lipoproteins

Funding

This research was funded by a Natural Sciences and Engineering Research Council (NSERC) Collaborative Research and Development grant (CRDPJ #472710-16), in partnership with Elmira Pet Products (Elmira, ON, Canada). The choline chloride and L-carnitine supplements were kindly donated by Balchem (Montvale, NJ, United States of America).

Author Contributions

Alexandra Rankovic (Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing - original draft, Writing - review & editing), Anna Shoveller (Conceptualization, Methodology, Resources, Writing - review & editing), Marica Bakovic (Conceptualization, Funding acquisition, Methodology, Writing - review & editing), Gordon Kirby (Conceptualization, Funding acquisition, Writing - review & editing), and Adronie Verbrugghe (Conceptualization, Funding acquisition, Investigation, Project administration, Supervision, Writing - review & editing)

Conflict of interest statement. The authors disclose that this manuscript is based on data that was previously published as part of AR’s doctoral dissertation from the University of Guelph. AKS is the Champion Petfoods Chair in Canine and Feline Nutrition, Physiology and Metabolism, consults for Champion Petfoods, was previously employed by P&G and Mars Pet Care, serves on the Scientific Advisory Board for Trouw Nutrition, and has received honoraria and research funding from various commodity groups, pet food manufacturers, and ingredient suppliers. AV is the Royal Canin Veterinary Diets Endowed Chair in Canine and Feline Clinical Nutrition, declares that she serves on the Health and Nutrition Advisory Board for Vetdiet and consults for Kismet. AV has received honoraria and research funding from various pet food manufacturers and ingredient suppliers. AR, MB and GK have no conflicts of interest to disclose.