Introduction

Both C57BL/6 and Swiss Webster (SW) mice have been extensively used for decades as general purpose model for research and drug safety testing. Interestingly, there are several strains of Swiss Webster outbred mice sold by various breeders, which over time, these mice may have acquired slight genetic drift despite their common origin. In contrast, C57BL/6 mice are the most widely used inbred strain due to its permissive background for maximal expression of most mutations. Although there are several breeder specific strains available, due to this mouse being an inbred strain, it generally has less genetic drift between commercial sources than SW mice. Genetic variations between mouse strains are important and have been shown to affect phenotype. Therefore, we are interested in differences between tissue uptake and trafficking of fatty acids in these two commonly used strains.

Metabolism is significantly affected by mouse strain differences. Several studies support the hypothesis that different strain variants have an effect on phenotype. Despite similar insulin sensitivity and secretion in several inbred strains, C57BL/6 mice are the least glucose tolerant of mice studied [1, 2]. However, conflicting glucose tolerance data was found when C57BL/6 mice are fed a high-fat diet and compared to a different 129 substrain, 129T2 compared to 129X1, suggesting that differences in strains used to compare to C57BL/6 mice can affect data interpretation [3]. Furthermore, C57BL/6 mice subjected to a high-fat diet have increased body fat mass and decreased oxygen consumption, coupled with lower endurance capacity and decreased β-oxidation in the liver, which correlates to increased obesity compared to other inbred strains [4, 5]. These data suggest liver and muscle metabolism differences contribute to physical performance and susceptibility to diet-induced obesity.

Interestingly, substrain differences in lipid metabolism in C57BL/6J compared to C57BL/6N mice have also been observed, with a high-fat diet inducing an increase in plasma insulin and blood glucose in the C57BL/6N substrain, resulting in more severe hepatic steatosis and inflammation [6]. Similar results are found in an ob/ob obesity model in inbred strains of C57BL/6 mice which more rapidly cleared circulating triacylglycerol (TAG), resulting in a more severe hepatic steatosis [7]. Another study comparing C57BL/6 substrains found a functional deletion in the nicotinamide nucleotide transhydrogenase gene in C57BL/6J mice [8]. Similarly, a study in 129 substrains found genetic variations have a considerable impact on strategies that involve targeted mutagenesis [9]. These data support the observation that studies between strains cannot be treated interchangeably without mention of genetic variations that could contribute to alterations in interpretation of results.

SW and C57BL/6 mice are extensively used for the study of lipid metabolism. Brain- and heart-fatty acid binding protein (FABP) were isolated from the brain of SW mice [10], while C57BL/6 mice are often used to study n-3 lipid metabolism and its potential benefits after ischemia and spinal cord injury [11–13] Further, hearts of C57BL/6 mice have decreased incorporation of exogenous radiolabeled oleic acid as compared to several other inbred strains as a result of increased mRNA expression of acyl-CoA oxidase 1 (Acox1) [14]. Interestingly, several inbred and outbred mouse strains including C57BL/6 and CD-1:SW have a mutation in the gene encoding for secretory group II phospholipase A2, with C57BL/6 mice having a homozygous frameshift mutation in the sPLA2 gene, while CD-1:SW mice were heterozygous for this same mutation [15]. However, what impact mouse strain has on tissue fatty acid uptake and trafficking into specific lipid pools is unknown.

To address this gap in knowledge, we determined the impact of strain differences in fatty acid uptake and trafficking in SW and C57BL/6 mouse strains using a steady-state radiotracer kinetic model [16, 17]. Strain influences on [1-14C]16:0, [1-14C]20:4n-6, or [1-14C]22:6n-3 uptake and trafficking in liver, heart, and brain in vivo was determined. These data show for the first time, that mouse strain impacts whole body 20:4n-6 metabolism as indicated by increased radiotracer in plasma in C57BL/6 mice as compared to SW mice. We also found an increased incorporation and trafficking of [1-14C]20:4n-6 in a number of heart phospholipids of C57BL/6 mice. Further in heart, the decreased incorporation of [1-14C]16:0 and [1-14C]22:6n-3 into the total phospholipid pool was accounted for by a profound reduction in esterification into the ethanolamine glycerophospholipids (EtnGpl) and choline glycerophospholipids (ChoGpl) in C57BL/6 as compared to SW mice. From these data we conclude that there are significant differences in lipid metabolism mainly in heart between these mice, which suggests that genetic background plays a substantial role in fatty acid trafficking.

Methods

Surgery and Fatty Acid Infusion

Samples were collected from 3 month old Swiss Webster and C57BL/6 male mice (National Institute of Cancer, Frederick, MD, USA) under an approved protocol by the University of North Dakota (Grand Forks, ND, USA) Animal Care and Use Committee (protocol 0110-1). During catheter insertion mice were anesthetized with (1–3 %) Halothane and polyethylene10 catheters were inserted into the right femoral vein and artery. Mice were allowed to recover from anesthesia for three hours following surgery to allow equilibration of metabolism [18, 19]. After recovery, mice were infused with 170 µCi/kg of either [1-14C]16:0, [1-14C]20:4n-6, or [1-14C]22:6n-3 at a rate of 50 µl/min for 10 min to achieve steady state plasma radioactivity. Blood was collected at set intervals and plasma separated to assess intravascular radioactivity. Mice were then euthanized at 10 min with pentobarbital (i.v.) and immediately subjected to head focused microwave irradiation to heat denature enzymes in situ [17, 20]. Whole brain, liver, and heart were removed, flash frozen in liquid nitrogen, and then pulverized into a homogeneous powder under liquid nitrogen conditions. Because of the contribution of residual blood to tissue radioactivity, whole blood was extracted using a two-phase extraction method [21] as previously described [22]. Residual blood in liver, heart, and brain was estimated to be 17, 22, and 2 %, respectively, based upon previously published data [23–26].

Tissue Lipid Extraction

Following pulverization, lipids were extracted from the tissues using a two-phase extraction method [21]. Powdered tissue was kept on dry ice until added to a tared Tenbroeck homogenizer, to determine the mass (g ww) of the tissue added. To the Tenbroeck homogenizer 17 vol. of chloroform and methanol (2:1, by vol) was added. Once tissue was homogenized, the extract was removed, saved and then the homogenizer was washed with 3 vol. of chloroform and methanol (2:1, by vol) and the rinse added to the initial extract. The tissue residue was saved for protein quantification. After homogenization, 4 vol of 0.9 % KCl was added to extract, mixed by vortexing, and allowed to separate into two phases overnight at −20 °C under a N2(g) atmosphere. The following morning, the top phase was removed and saved. To the lower phase, 4 vol. of chloroform, methanol, and water (3:48:47, by vol) was added, the sample vortexed, chilled, and subjected to centrifugation to facilitate phase separation. The upper phase was removed and combined with the previous upper phase to determine aqueous fraction radioactivity using liquid scintillation counting (LSC). The lower phase was dried down with N2 (g) and dissolved in hexane:2-propanol (3:2, by vol) and a portion used to determine organic radioactivity using LSC. The unused portion was stored under N2 (g) atmosphere at −80 °C in hexane:2-propanol (3:2, by vol).

Lipid Separation by Thin Layer Chromatography

Extracted lipids were separated using thin layer chromatography (TLC) on heat-activated (110 °C) Whatman silica gel-60 plates. Neutral lipids were separated into individual classes using petroleum either, diethyl ether, and acetic acid (75:25:1.3, by vol.) [27]. Phospholipids were separated into individual classes using chloroform, methanol, acetic acid, and water (50:37.5:3:2, by vol.) [28]. Lipids were visualized using iodine (phospholipids) or 6-(p-toluidino)-2-naphthalenesulfonic acid (TNS) (neutral lipids) and identified using commercially available standards (Avanti, Polar Lipids, Alabaster, AL, USA). Once lipids were visualized, the silica was scraped into glass scintillation vials and 0.5 mL ddH2O was added to facilitate desorption of lipids from the silica. Then 10 mL of Scintiverse BD cocktail (Fisher Scientific, Pittsburgh, PA, USA) was added and samples were mixed by vortexing and then allowed to settle for 1 h. Samples were counted on a Beckman LS 6500 liquid scintillation counter equipped with low-level detection software (Fullerton, CA, USA).

Protein Quantification

Protein content in the tissue residue was quantified using a modified dye-binding assay utilizing bovine serum albumin as a standard [29]. The tissue residue was dried of residual solvent and then subjected to hydrolysis with 0.2 KOH at 65 °C overnight [30]. Samples were then mixed with dye binding reagent and allowed to equilibrate for 10 min prior to reading on a spectrophotometer at 595 nm.

Statistics

Statistical analysis was done using the Instat® statistical program (Graphpad, San Diego, CA). Multiple comparisons were assessed using a one-way ANOVA with a Tukey–Kramer post hoc test, with p < 0.05 considered as significant, n = 3–4. A two-tailed Student’s t test was used to determine significance between treatment groups, with p < 0.05 considered to be significant, n = 3–4.

Results

Plasma Area Under the Curve was Significantly Different for [1-14C]20:4n-6 Infusions

The effect of strain differences on fatty acid tracer content in plasma in SW and C57BL/6 mice was determined by incremental plasma sampling during the steady state radiotracer infusion. The tracer content in plasma is vital for the calculation of the coefficient of incorporation, k i *, where radioactivity in a given compartment is divided by the integrated area under the curve for plasma radioactivity for each individual mouse. There was a significant difference (p < 0.05) in plasma curve area, 1334 ± 100 (n = 4) for C57BL/6 and 1024 ± 121 (n = 4) for SW mice (Fig. 1). Plasma areas for [1-14C]16:0, 945 ± 80 (C57BL/6) versus 925 ± 85 (SW) and for [1-14C]22:6n-3, 1212 ± 75 (C57BL/6) versus 1154 ± 91 (SW), were not significantly different between strains (Fig. 1). The reduced [1-14C]20:4n-6 tracer concentration in plasma of SW mice could be a result of enhanced uptake of 20:4n-6 into tissues in these mice as compared to C57BL/6.

Fig. 1
figure 1

Plasma curves for [1-14C]16:0, [1-14C]20:4n-6, and [1-14C]22:6n-3 comparing infusion of radiotracer fatty acids into SW and C57BL/6J mice. SW (unfilled box) and C57BL/6 (filled circle) were infused with [1-14C]16:0 (upper panel), [1-14C]20:4n-6 (middle panel), and [1-14C]22:6n-3 (lower panel). Values are expressed as nCi/mL as found in plasma collected during radiotracer infusion and represent mean ± SD (n = 3–4). The asterisk indicates a statistically significant difference from SW and C57BL/6 strains using one-way ANOVA and a Tukey–Kramer post hoc test was used for plasma curves and a Student’s t test was used to compare the area under the curve between strains (p < 0.05)

Full size image

C57BL/6 Mice Have Increased Total Uptake into Heart Tissue

Tissue fatty acid uptake and incorporation into the organic and aqueous fractions was determined (Table 1). In heart, there was a significant 1.7-fold increase in total uptake of [1-14C]20:4n-6, accounted for by a 1.8-fold increase in incorporation into the organic fraction of C57BL/6 mice (Table 1). Although, the total uptake of [1-14C]22:6n-3 into heart was not significantly different, the incorporation into the organic fraction was reduced 33 % in C57Bl/6 compared to SW mice (Table 1). There was no difference in [1-14C]16:0 total uptake or in incorporation into organic or aqueous fractions. For all of the fatty acid tracers, there was no significant difference in the heart aqueous fraction, which represents products of β-oxidation [18, 31, 32].

Table 1 Distribution of fatty acid tracer into liver, heart, or brain organic and aqueous fractions in SW and C57BL/6 mice
Full size table

Liver and brain tissue did not have significant differences in uptake or incorporation into organic or aqueous fractions, suggesting the observed differences between strains were specific to [1-14C]20:4n-6 uptake into heart tissue and incorporation of [1-14C]22:6n-3 into the organic fraction.

Metabolic Targeting into Phospholipid and Neutral Lipid Pools are Differentially Modulated in Heart

Phospholipid and neutral lipid fractions were separated to determine fatty acid targeting differences in SW and C57BL/6 mice (Table 2). For liver, [1-14C]16:0, uptake into neutral lipids decreased 18 % in C57BL/6 mice (Table 2), but no other changes were observed. In heart, there was a significant reduction in the incorporation of [1-14C]16:0 (45 %) and of [1-14C]22:6n-3 (49 %) into phospholipids in C57BL/6 compared to SW mice (Table 2). Furthermore, C57BL/6 mice have decreased [1-14C]16:0 (45 %) fractional distribution into heart phospholipids (Table 2). In contrast, incorporation of [1-14C]20:4n-6 into C57BL/6 heart phospholipids increased 2.2-fold compared to SW mice (Table 2). We observed no difference in 20:4n-6 and 22:6n-3 trafficking into heart neutral lipids between groups (Table 2). There were no significant changes in [1-14C]16:0, [1-14C]20:4n-6, or [1-14C]22:6n-3 fatty acid tracer uptake or trafficking in brain (Table 2). These data indicate that mouse strain impacts specific tissue fatty acid uptake and trafficking and predominately impacts targeting into phospholipids in the heart.

Table 2 Incorporation of fatty acid tracer into liver, heart, or brain phospholipid and neutral lipid pools in SW and C57BL/6 mice
Full size table

Strain Specific Differences in Fatty Acid Targeting into Individual Lipid Classes

To determine fatty acid targeting into individual lipid classes in SW and C57BL/6 mice, phospholipid and neutral lipid fractions were separated. In liver there was a significant 17 % decrease in [1-14C]16:0 incorporation into triacylglycerol pools (TAG) (Table 3), consistent with the reduction observed in the neutral lipid fraction of C57BL/6 mice compared to SW mice (Table 2). A 1.2-fold increase in esterification of [1-14C]20:4n-6 into EtnGpl was found in liver but, there were no differences in esterification of [1-14C]22:6n-3 into liver between strains (Table 3).

Table 3 Targeting of fatty acid tracer into individual liver lipid classes in SW and C57BL/6 mice
Full size table

In heart, esterification of [1-14C]16:0 into ethanolamine glycerophospholipids (EtnGpl) and choline glycerophospholipids (ChoGpl) was reduced by 53 and 37 %, respectively (Table 4). A 51 % reduction in targeting of [1-14C]16:0 into the diacylglycerol (DAG) fraction suggests a rapid esterification of DAG into TAG, which is consistent with the 1.5-fold increase in [1-14C]16:0 targeting to TAG in C57BL/6 mice compared to SW mice (Table 4). Esterification of [1-14C]20:4n-6 into heart was significantly increased in C57BL/6 mice. Increased targeting of [1-14C]20:4n-6 into phospholipids is consistent with increased uptake into heart (Table 1) and into the phospholipid fraction (Table 2). Esterification of [1-14C]20:4n-6 into all major phospholipids were significantly increased in heart of C57BL/6 mice; cardiolipin (Ptd2Gro) 2.4-fold, EtnGpl 1.9-fold, phosphatidylinositol (PtdIns) 2.9-fold, phosphatidylserine (PtdSer) 3.6-fold, ChoGpl 2.1-fold, and sphingomyelin (CerPCho) 2.1-fold (Table 4). Esterification of [1-14C]20:4n-6 into TAG was also increased 1.4-fold in heart of C57BL/6 mice (Table 4). In heart there was a significant decrease in esterification of [1-14C]22:6n-3 into EtnGpl and ChoGpl, 44 and 52 % in C57BL/6 mice compared to SW mice, respectively, similar to that observed for 16:0 (Table 4).

Table 4 Targeting of fatty acid tracer into individual heart lipid classes in SW and C57BL/6 mice
Full size table

There were no significant differences in esterification of [1-14C]16:0 or [1-14C]22:6n-3 into any brain lipid pools. However, esterification of [1-14C]20:4n-6 into brain DAG pool was decreased 43 % in C57BL/6 mice compared to SW mice (Table 5).

Table 5 Targeting of fatty acid tracer into individual brain lipid classes in SW and C57BL/6 mice
Full size table

Discussion

Although, resting metabolic rate and tissue turnover are not different between several strains of mouse [33], the impact of mouse strain on tissue fatty acid uptake and trafficking is poorly understood. Our results clearly indicate both similarities and differences in the uptake and trafficking of fatty acids into tissues of two commonly used mouse strains. Although there is a significant difference in fatty acid uptake and trafficking in heart between these two strains, surprisingly only minor differences were observed in brain and liver. Appreciating differences in characteristics between mouse strains to leverage the advantages and minimize potential pitfalls is an important consideration in experimental design [34]. For instance, C57BL/6 mice are commonly utilized in a substantial share of the scientific literature and are often used to study lipid metabolism, including studies on the efficacy of n-3 fatty acids on inflammation and injury [11, 12, 14, 35, 36]. On the other hand, SW mice are have been used in the pharmacology of fatty acid metabolism such as effects of pyridine exposure and high-fat diets on lipid metabolism [37–39]. Thus, we determined the differences in fatty acid uptake and trafficking in three tissues commonly used in experimental design using three fatty acids commonly found at the sn-1 position (16:0) and sn-2 position (20:4n-6 and 22:6n-3) of phospholipids.

In our study, one important observation is the differences between strains in the integrated plasma curve area in [1-14C]20:4n-6 infused mice, but not for [1-14C]16:0 and[1-14C]22:6n-3. Such an observation using this steady-state kinetic model is not without precedence. In H-FABP (Fabp3) gene-ablated mice on a 129 × Balb/c background, there is a significant increase in average integrated plasma area of 16:0 in the gene-ablated mice, consistent with the profound reduction in 16:0 uptake into the heart [18] and presumably muscle tissue where H-FABP is also highly expressed [40, 41]. These data suggest that H-FABP can affect total body metabolism for fatty acids commonly used in β-oxidation, which is consistent with the reduced targeting of 16:0 into heart neutral lipids, e.g. TAG [18]. In rodents, 16:0 incorporation and turnover in TAG is very rapid indicating its use in β-oxidation [42]. Further, in rat lipid metabolism, 16:0, a saturated fatty acid, is mainly targeted towards β-oxidation [19, 43], whereas 20:4n-6 is mainly esterified into phospholipids [19], suggesting substantially different roles in the heart. Although we did not examine H-FABP levels between strains, the lack of change in the plasma curve for [1-14C]16:0 does not suggest a difference in H-FABP levels as a potential mechanism, especially as the differences observed are for only 20:4n-6. However, a global change in 20:4n-6 metabolism, similar to that observed for 16:0 in H-FABP gene-ablated mice, may occur between mouse strains.

In C57BL/6 mice there is a homozygous disruption in the secretory group II phospholipase A2 gene, while in CD-1:SW mice this is a heterozygous mutation when compared to several other normal genotype inbred mouse strains; including BALB/c, C3H/HE, DBA/1, DBA/2, NZB/B1N, and MRL lpr/lpr mice [15, 44]. This frameshift mutation occurs in the Ca2+ binding domain, rendering the group II sPLA2 enzymatically inactive in C57BL/6 mice. The sPLA2 family, I/II/V/X and otoconin-90, are closely related enzymes with a highly conserved Ca2+ binding loop that hydrolyze sn-2 fatty acids from phospholipids [45–47]. sPLA2 enzymes are vital in signal transduction as they often release 20:4n-6, which is enriched in the sn-2 position of phospholipids [48–50].

Although, group II sPLA2 are sn-2 esterified fatty acid specific, they are not intrinsically 20:4n-6 specific [48, 51, 52]. Therefore, the preponderance of 20:4n-6 at the sn-2 position may result in less fatty acid turnover in the C57BL/6 mice. We did not however, see a significant reduction of esterification into C57BL/6 mouse heart phospholipids in any of the fatty acids observed as one would expect due to the lack of specificity toward esterified fatty acids in group II sPLA2. While the reduction in targeting of [1-14C]16:0 and [1-14C]22:6n-3 to EtnGpl and ChoGpl may be a result of reduced sPLA2 activity, the lack of a reduction in 20:4n-6 incorporation into these mice suggest the decreased activity of sPLA2 may not be critical for the changes observed between strains. However, in denervated rat heart, a reduction in general PLA2 activity results in decreased 20:4n-6 turnover, which causes a reduction of 20:4n-6 incorporation into stable lipid compartments [42]. This is in contrast to our data as we observed a broad increase in esterification of 20:4n-6 into heart phospholipids in C57BL/6 mice, suggesting the difference in uptake and incorporation are not a result of this mutation in sPLA2. However, because of the selective processes for the uptake of 20:4n-6 from plasma into heart phospholipids [19], we might have observed the significant increase in esterification into heart phospholipids due to increased 20:4n-6 availability in the plasma (Fig. 1). Therefore, changes in expression of some other gene may be altering uptake and/or metabolism of 20:4n-6, resulting in the observed differences between C57BL/6 and SW mice.

C57BL/6 mice are the most commonly used mouse strain to study metabolic diseases and they are one of the most susceptible to the development of diet-induced obesity and insulin resistance [53–55]. While insulin resistance is most associated with glucose intolerance, lipid metabolism is severely affected by this disease. Interestingly, when 129S6 mice are fed a high-fat diet there is increased obesity, hyperlipidemia, enhanced insulin secretion, glucose intolerance and fatty liver relative to control diet fed mice. However, BALB/c mice fed the same diet did not result in disease, but these mice had both enhanced insulin secretion and significantly improved glucose tolerance, suggesting insulin levels between strains can have a significant impact on phenotype [56]. Furthermore, when 129S6 are fed a high-fat diet, gene expression of acyl-CoA oxidase 1 (Acox1) is reduced, a protein which traffics fatty acids towards β-oxidation, however, under the same dietary conditions in BALB/C mice there is no change in expression [56]. Interestingly, in an ex vivo working heart model, C57BL/6 mice had increased Acox1 expression as compared to 129/SvEvTac mouse heart, and as a result had decreased incorporation of oleic acid into TAG pools and more was utilized for β-oxidation than in 129/SvEvTac mouse heart [14]. Collectively, these studies demonstrate that observations in one mouse strain are not directly applicable to another strain, and support our supposition that for lipid metabolism, observations between mouse strains are not interchangeable.

Prior to this study, it was unknown what impact, if any, the strain of mouse had on tissue fatty acid uptake and trafficking into individual lipid classes. Hence, we report that two commonly used strains of mice had significant differences in heart fatty acid uptake and trafficking. While each of the fatty acids used were altered, the major effect was on the uptake and trafficking of 20:4n-6. This is important as 20:4n-6 is an important signaling molecule in the heart [57–61]. This is an important consideration because there is an enormous body of literature in which these mouse strains have been used interchangeably. Our data suggest that investigators must be careful when comparing results involving lipid metabolism between these strains and this is more likely applicable for studies involving lipid metabolism between different mouse strains in general.