Advertisement

Quantitative trait loci that determine plasma lipids and obesity in C57BL/6J and 129S1/SvImJ inbred mice

Open AccessPublished:June 21, 2004DOI:https://doi.org/10.1194/jlr.M400098-JLR200
      The plasma lipid concentrations and obesity of C57BL/6J (B6) and 129S1/SvImJ (129) inbred mouse strains fed a high-fat diet containing 15% dairy fat, 1% cholesterol, and 0.5% cholic acid differ markedly. To identify the loci controlling these traits, we conducted a quantitative trait loci (QTL) analysis of 294 (B6 × 129) F2 females fed a high-fat diet for 14 weeks. Non-HDL cholesterol concentrations were affected by five significant loci: Nhdlq1 [chromosome 8, peak centimorgan (cM) 38, logarithm of odds [LOD] 4.4); Nhdlq4 (chromosome 10, cM 70, LOD 4.0); Nhdlq5 (chromosome 6, cM 0) interacting with Nhdlq4; Nhdlq6 (chromosome 7, cM 10) interacting with Nhdlq1; and Nhdlq7 (chromosome 15, cM 0) interacting with Nhdlq4. Triglyceride (TG) concentrations were affected by three significant loci: Tgq1 (chromosome 18, cM 42, LOD 3.2) and Tgq2 (chromosome 9, cM 66) interacting with Tgq3 (chromosome 4, cM 58). Obesity measured by percentage of body fat mass and body mass index was affected by two significant loci: Obq16 (chromosome 8, cM 48, LOD 10.0) interacting with Obq18 (chromosome 9, cM 65).
      Knowing the genes for these QTL will enhance our understanding of obesity and lipid metabolism.
      Cardiovascular disease is often coincident with dyslipidemia, obesity, hypertension, and diabetes, which are often clustered in some individuals and recognized as metabolic syndrome (
      • Isomaa B.
      • Almgren P.
      • Tuomi T.
      • Forsen B.
      • Lahti K.
      • Nissen M.
      • Taskinen M.R.
      • Groop L.
      Cardiovascular morbidity and mortality associated with the metabolic syndrome.
      ). These disorders are complex, multifactorial, and controlled by both environmental and genetic factors. The causal relationship between the risk of cardiovascular disease and either obesity or increased LDL cholesterol or triglyceride (TG) is definitively established. Much is known about the nature and effect of environmental factors, yet relatively little is known about the genetic basis of these disorders. Thus, knowledge of the primary genetic determinants of plasma lipoprotein levels and obesity will enhance our understanding of the pathophysiological background and may provide novel molecular targets for intervention.
      Mouse crosses have helped to localize and identify genes underlying these complex traits (
      • Allayee H.
      • Ghazalpour A.
      • Lusis A.J.
      Using mice to dissect genetic factors in atherosclerosis.
      ,
      • Smith J.
      Quantitative trait locus mapping for atherosclerosis susceptibility.
      ,
      • Brockmann G.A.
      • Bevova M.R.
      Using mouse models to dissect the genetics of obesity.
      ,
      • Wang X.
      • Paigen B.
      Quantitative trait loci and candidate genes regulating HDL cholesterol: a murine chromosome map.
      ). When exposed to a high-fat diet, mice of different inbred strains exhibit great variation in plasma lipoproteins and obesity (
      • Paigen B.
      • Svenson K.L.
      • Peters L.L.
      Diet effects on plasma lipids and susceptibility to atherosclerosis. MPD:99. Mouse Phenome Database (MPD). Bar Harbor, ME: The Jackson Laboratory.
      ). Our laboratory has used quantitative trait loci (QTL) analysis to investigate the genetics underlying lipoprotein metabolism and atherosclerosis (
      • Wang X.
      • Paigen B.
      Quantitative trait loci and candidate genes regulating HDL cholesterol: a murine chromosome map.
      ,
      • Korstanje R.
      • Paigen B.
      From QTL to gene: the harvest begins.
      ). The number of genetic loci that differ between C57BL/6J (B6) and 129S1/SvImJ (129) mice underscores the importance of strain background when evaluating the impact of a gene deficiency in targeted mutant mice. In most cases, targeted mutant mice are derived from embryonic stem cells of 129 mouse substrains. A target gene in these cells is “knocked out” by homologous recombination, and the resulting cells are microinjected into B6 blastocysts, which develop into B6/129 chimeras. These in turn are mated to B6 mice to produce mice heterozygous for B6 and 129 alleles at all loci for which these strains differ. These mice are intercrossed to generate mice homozygous for 129 alleles at the target locus (−/−) and a small region surrounding it, but the remainder of their genomes are a random mixture of B6 and 129 alleles (
      • Smithies O.
      • Maeda N.
      Gene targeting approaches to complex genetic diseases: atherosclerosis and essential hypertension.
      ). If littermates of such mixed-background targeted mutant stocks differ in their allelic combinations, they could yield different experimental results. Thus, to evaluate gene function in targeted mutant mice, the genetic background must be carefully controlled by constructing B6/129 congenic strains. This is carried out by successively backcrossing carriers of the targeted mutation to B6 mice until the only 129 alleles left on a nearly pure B6 background are the target locus (−/−) and the surrounding genetic materials (
      • Sigmund C.D.
      Viewpoint: are studies in genetically altered mice out of control?.
      ).
      We report here the results of our investigation of plasma non-HDL cholesterol levels, TG concentrations, and obesity among (B6 × 129) F2 females that had for 14 weeks consumed a high-fat diet containing 15% dairy fat, 1% cholesterol, and 0.5% of the hydrophobic bile acid cholic acid, which promotes cholesterol absorption. Previously, we reported QTL detected with this intercross that determine plasma HDL-cholesterol levels and atherosclerosis susceptibility (
      • Ishimori N.
      • Li R.
      • Kelmenson P.M.
      • Korstanje R.
      • Walsh K.A.
      • Churchill G.A.
      • Forsman-Semb K.
      • Paigen B.
      Quantitative trait loci analysis for plasma HDL-cholesterol concentrations and atherosclerosis susceptibility between inbred mouse strains C57BL/6J and 129S1/SvImJ.
      ). In this study, we identified several QTL for non-HDL cholesterol, TG, percentage of body fat mass (% fat), and body mass index (BMI).

      MATERIALS AND METHODS

      Animals and diet

      B6 and 129 mice were obtained from The Jackson Laboratory (Bar Harbor, ME) and mated to produce the (B6 × 129) F1 progeny, which were intercrossed to produce an F2 population of which the 301 female F2 progeny were used in this investigation. Mice were maintained in a temperature- and humidity-controlled environment with a 14 h light/10 h dark cycle and given unrestricted access to food and acidified water. The cages were covered with polyester filters and contained pine shavings bedding. Six week old mice were fed a high-fat diet (
      • Nishina P.M.
      • Verstuyft J.
      • Paigen B.
      Synthetic low and high fat diets for the study of atherosclerosis in the mouse.
      ,
      • Nishina P.M.
      • Lowe S.
      • Verstuyft J.
      • Naggert J.K.
      • Kuypers F.A.
      • Paigen B.
      Effects of dietary fats from animal and plant sources on diet-induced fatty streak lesions in C57BL/6J mice.
      ) containing 15% dairy fat, 1% cholesterol, and 0.5% cholic acid for 14 weeks, after which they were killed by cervical dislocation. Experiments were approved by the Institutional Animal Care and Use Committee of The Jackson Laboratory.

      Quantitative phenotype measurements

      After consuming the high-fat diet for 14 weeks, mice were fasted for 4 h before blood samples were collected in plasma separator tubes containing EDTA, placed on ice, and centrifuged. Plasma lipid concentrations were measured using an enzymatic assay (Beckman, Fullerton, CA) as previously described (
      • Nishina P.M.
      • Lowe S.
      • Verstuyft J.
      • Naggert J.K.
      • Kuypers F.A.
      • Paigen B.
      Effects of dietary fats from animal and plant sources on diet-induced fatty streak lesions in C57BL/6J mice.
      ). Because mice were fasted before blood was collected and because chylomicrons display very short half-lives (
      • de Faria E.
      • Fong L.G.
      • Komaromy M.
      • Cooper A.D.
      Relative roles of the LDL receptor, the LDL receptor-like protein, and hepatic lipase in chylomicron remnant removal by the liver.
      ), non-HDL was presumed to comprise predominantly VLDL and LDL. Non-HDL concentrations were obtained by subtracting HDL from total cholesterol concentrations. The % fat of each mouse was measured using peripheral dual-energy X-ray absorptiometry (PIXImus; GE-Lunar, Madison, WI), a method that has been validated in mice as an accurate measure (
      • Nagy T.R.
      • Clair A.L.
      Precision and accuracy of dual-energy X-ray absorptiometry for determining in vivo body composition of mice.
      ). The BMI of each mouse was calculated by dividing its body weight (grams) by the square of its anal-nasal length (meters).

      Genotyping

      We genotyped 294 F2 progeny initially with 88 simple sequence length polymorphic (SSLP) markers (Research Genetics, Huntsville, AL) spaced ∼20 centimorgan (cM) apart and later added 23 additional SSLP markers in the QTL regions as previously described (
      • Ishimori N.
      • Li R.
      • Kelmenson P.M.
      • Korstanje R.
      • Walsh K.A.
      • Churchill G.A.
      • Forsman-Semb K.
      • Paigen B.
      Quantitative trait loci analysis for plasma HDL-cholesterol concentrations and atherosclerosis susceptibility between inbred mouse strains C57BL/6J and 129S1/SvImJ.
      ). The average spacing between these markers (±SD) was 14 ± 12 cM. DNA isolation, PCR amplification, and subsequent gel electrophoresis have been described previously (
      • Mu J.L.
      • Naggert J.K.
      • Svenson K.L.
      • Collin G.B.
      • Kim J.H.
      • McFarland C.
      • Nishina P.M.
      • Levine D.M.
      • Williams K.J.
      • Paigen B.
      Quantitative trait loci analysis for the differences in susceptibility to atherosclerosis and diabetes between inbred mouse strains C57BL/6J and C57BLKS/J.
      ). Reported genetic map positions were retrieved from the Mouse Genome Informatics database (http://www.informatics.jax.org).

      Statistics

      One-way ANOVAs with Tukey’s correction for multiple pairwise compositions were used to determine statistically significant differences in plasma lipid levels, BMI, and % fat between mouse groups. Data were analyzed using Graphpad Prism (Windows version 3.00; GraphPad Software, San Diego, CA). Phenotypes were associated using Pearson’s correlation. As described previously (
      • Sen S.
      • Churchill G.A.
      A statistical framework for quantitative trait mapping.
      ,
      • Churchill G.A.
      • Doerge R.W.
      Empirical threshold values for quantitative trait mapping.
      ), a three-step QTL analysis was conducted to search for main effects and pairwise gene interactions and then to integrate all of the main and interacting QTL phenotype associations into a multiple regression. In the regression analysis, we combined all significant and suggestive QTL and interactions in a multiple regression model. Terms that did not meet the nominal 0.02 level in the regression were eliminated in a backward stepwise manner with the exception that main effect terms involved in a significant interaction were retained. Final models were reported for each trait. Some traits were log transformed before analysis. This resulted in approximate normality for non-HDL and BMI. QTL were deemed significant if they either met or exceeded the 95% genome-wide adjusted threshold, which was assessed by permutation analysis for each trait [logarithm of odds (LOD) ≥ 3.0 for log-transformed non-HDL, LOD ≥ 3.1 for TG and log-transformed BMI, and LOD ≥ 3.2 for % fat]; they were deemed suggestive if they either met or exceeded the 37% genome-wide adjusted threshold (LOD ≥ 1.8 for TG and LOD ≥ 1.9 for other traits) but were not significant. QTL confidence intervals were calculated according to the posterior probability density of QTL locations, as described previously (
      • Sen S.
      • Churchill G.A.
      A statistical framework for quantitative trait mapping.
      ). Variance indicates the percentage of the total F2 phenotypic variance associated with each marker. Analyses were carried out using Pseudomarker 0.9 software (Sen and Churchill; http://www.jax.org/staff/churchill/labsite).

      Naming QTL

      In accordance with the International Committee on Standardized Genetic Nomenclature for Mice (http://www.informatics.jax.org/mgihome/nomen) and the Complex Traits Consortium (
      • Biola O.
      • Angel J.M.
      • Avner P.
      • Bachmanov A.A.
      • Belknap J.K.
      • Bennett B.
      • Blankenhorn E.P.
      • Blizard D.A.
      • Bolivar V.
      • Brockmann G.A.
      • Buck K.J.
      • Bureau J.F.
      • Casley W.L.
      • Chesler E.J.
      • Cheverud J.M.
      • Churchill G.A.
      • Cooke M.
      • Crabbe J.C.
      • Crusio W.E.
      • Darvasi A.
      • de Haan G.
      • Dermant P.
      • Doerge R.W.
      • Elliot R.W.
      • Farber C.R.
      • Flaherty L.
      • Flint J.
      • Gershenfeld H.
      • Gibson J.P.
      • Gu J.
      • Gu W.
      • Himmelbauer H.
      • Hitzemann R.
      • Hsu H.C.
      • Hunter K.
      • Iraqi F.F.
      • Jansen R.C.
      • Johnson T.E.
      • Jones B.C.
      • Kempermann G.
      • Lammer F.
      • Lu L.
      • Manly K.F.
      • Matthews D.B.
      • Medrano J.F.
      • Mehrabian M.
      • Mittlemann G.
      • Mock B.A.
      • Mogil J.S.
      • Montagutelli X.
      • Morahan G.
      • Mount J.D.
      • Nagase H.
      • Nowakowski R.S.
      • O'Hara B.F.
      • Osadchuck A.V.
      • Paigen B.
      • Palmer A.A.
      • Pierce J.L.
      • Pomp D.
      • Rosemann M.
      • Rosen G.D.
      • Schalkwyk L.C.
      • Seltzer Z.
      • Settle S.
      • Shimomura K.
      • Shou S.
      • Sikela J.M.
      • Siracusa L.D.
      • Spearow J.L.
      • Teuscher C.
      • Threadgill D.W.
      • Toth L.A.
      • Toye A.A.
      • Vadasz C.
      • Van Zant G.
      • Wakeland E.
      • Williams R.W.
      • Zhang H.G.
      • Zou F.
      The nature and identification of quantitative trait loci: a community's view.
      ), we have named QTL as follows. QTL are named if significant or if suggestive but confirm a QTL reported previously. If a QTL substantially overlaps a previously discovered QTL, it is given the same name if the crosses share at least one parent in common (i.e., B6 × 129 and CAST × 129) and it is given a new name if the strains are all different.

      RESULTS

      Inheritance of plasma non-HDL and TG levels, % fat, and BMI

      Plasma non-HDL and TG concentrations, % fat, and BMI were measured after animals had been fed the high-fat diet for 14 weeks (Table 1). Compared with 129, B6 mice displayed significantly increased non-HDL levels. The F1 mice displayed non-HDL levels intermediate between and significantly different from those of the parental strains; thus, high non-HDL cholesterol levels were inherited in an additive manner. Compared with 129, B6 mice displayed significantly decreased plasma TG levels. The F1 mice displayed TG levels comparable to those of strain B6 and significantly lower than those of strain 129; thus, high TG levels were inherited in a recessive manner. B6 mice displayed significantly lower % fat than did 129 mice, and the F1 mice displayed intermediate values between those of the parental strains. The BMI distribution was similar to that for % fat; B6 mice displayed significantly lower BMI compared with 129 mice. The F1 mice displayed BMI values intermediate between the parents but closer in value to those of strain 129. We started with 301 F2 females and quantified each trait of 294 females after the 14 week high-fat diet. Log-transformed non-HDL, TG, % fat, and log-transformed BMI were normally distributed among the F2 progeny (Fig. 1A–D). The log BMI was positively correlated with TG and % fat but negatively correlated with log-transformed non-HDL (Table 2).
      TABLE 1Plasma lipid concentrations, % fat, and BMI of 10 female B6, 10 female 129, 8 female F1, and 294 female F2 progeny fed a high-fat diet for 14 weeks
      MicenPlasma
      Plasma lipid concentrations were measured in mice fasted for 4 h.
      % FatBMI
      Non-HDLTG
      mg/dl%×104 g/m2
      B610140 ± 9
      Significant difference (P < 0.001, by ANOVA) versus 129.
      ,
      Significant difference (P < 0.01, by ANOVA) versus F1.
      45 ± 2
      Significant difference (P < 0.01, by ANOVA) versus 129.
      19 ± 0
      Significant difference (P < 0.001, by ANOVA) versus 129.
      ,
      Significant difference (P < 0.001, by ANOVA) versus F1.
       23 ± 1
      Significant difference (P < 0.001, by ANOVA) versus 129.
      ,
      Significant difference (P < 0.01, by ANOVA) versus F1.
      1291071 ± 3
      Significant difference (P < 0.01, by ANOVA) versus F1.
      64 ± 536 ± 1
      Significant difference (P < 0.01, by ANOVA) versus F1.
       29 ± 0
      F18107 ± 638 ± 1
      Significant difference (P < 0.001, by ANOVA) versus 129.
      29 ± 2 27 ± 1
      F2294
      The number of F2 mice is 292 for plasma lipid concentrations and 291 for % fat. Because it is the distribution and not the mean among the F2 population that is most important for detecting genetic linkage to a phenotype, we did not test for significant differences between F2 progeny and either the parental strains or F1 progeny.
      123 ± 551 ± 125 ± 0 27 ± 0
      BMI, body mass index; % fat, percentage of body fat mass; TG, triglyceride. Data are presented as means ± SEM.
      a Plasma lipid concentrations were measured in mice fasted for 4 h.
      b Significant difference (P < 0.001, by ANOVA) versus 129.
      c Significant difference (P < 0.01, by ANOVA) versus F1.
      d Significant difference (P < 0.01, by ANOVA) versus 129.
      e Significant difference (P < 0.001, by ANOVA) versus F1.
      f The number of F2 mice is 292 for plasma lipid concentrations and 291 for % fat. Because it is the distribution and not the mean among the F2 population that is most important for detecting genetic linkage to a phenotype, we did not test for significant differences between F2 progeny and either the parental strains or F1 progeny.
      Figure thumbnail gr1
      Fig. 1Distributions and genome-wide scans for the traits. A–D: Log-transformed plasma non-HDL concentrations, plasma triglyceride (TG) concentrations, percentage of body fat mass (% fat), and log-transformed body mass index (BMI), respectively, in (B6 × 129) F2 progeny fed a high-fat diet for 14 weeks. The number of mice is 292 for plasma lipid concentrations, 291 for % fat, and 294 for BMI. Chromosomes 1 through X are represented numerically on the ordinate. The relative width of the space allotted for each chromosome reflects the relative length of each chromosome. The abscissa represents the logarithm of odds (LOD) score, the traditional metric of genetic linkage. The significant (P < 0.05) and suggestive (P < 0.63) levels of linkage were determined by permutation testing (
      • Churchill G.A.
      • Doerge R.W.
      Empirical threshold values for quantitative trait mapping.
      ).
      TABLE 2Pearson correlation coefficients among plasma lipid concentrations, % fat, and BMI in the F2 progeny fed a high-fat diet for 14 weeks
      VariableLog Non-HDLTG% Fat
      TG0.14
      P < 0.05.
      % fat−0.23
      P < 0.0001.
      0.21
      P < 0.001.
      Log BMI−0.23
      P < 0.0001.
      0.22
      P < 0.001.
      0.40
      P < 0.0001.
      The number of mice for each analysis was from 289 to 292.
      a P < 0.05.
      b P < 0.0001.
      c P < 0.001.

      Identification of genetic loci affecting non-HDL and TG concentrations, % fat, and BMI

      The genome-wide scans for single QTL are presented in Fig. 1 and summarized in Table 3, which provides the QTL peak, 95% confidence interval, LOD score, allele conferring the high value, nearest SSLP marker to QTL peak, overlapping QTL reported previously, and candidate genes for each QTL. The QTL were named if they were significant either as single QTL or interacting QTL. Suggestive QTL in this cross that were found previously were also named. We named the loci Nhdlq for non-HDL QTL, Tgq for TG QTL, and Obq for obesity QTL, in each case followed by a number. Figure 2shows the allele effects, which demonstrate the magnitude of the effect and the inheritance pattern (dominant, recessive, or additive).
      TABLE 3QTL identified for single gene or pairwise genome-wide scans of 294 (B6 × 129) F2 females
      TraitsChromosomal (Chr)

      Location
      95% Confidence

      Interval
      Locus

      Name
      Logarithm of

      Odds Score
      High

      Allele
      Nearest

      Marker
      Overlapping QTL
      Overlapping QTL identified in previous studies.


      (Reference)
      Candidate

      Genes (cM)
      cM
      Non-HDL Chr 8 (38)
      Interacting QTL.
       15–52Nhdlq1 4.4B6D8Mit248Nhdlq1 (19)Cpe (32.6), Lpl (33.0)
       Chr 10 (70)
      Interacting QTL.
       65–70Nhdlq4 4.0129D10Mit35Pltp2 (25)Apof (73.0)
       Chr 6 (0)
      Interacting QTL.
       0–24Nhdlq5 2.4
      Suggestive QTL.
      B6D6Mit86
       Chr 7 (70) 50–80
      We did not name this QTL because it is below the statistically significant level and overlaps no previously discovered QTL.
       2.4
      Suggestive QTL.
      129D7Mit141
       Chr 7 (10)
      Interacting QTL.
       5–20Nhdlq6D7Mit294Unnamed QTL (19)Apoc2 (4.0)
       Chr 15 (0)
      Interacting QTL.
       0–20Nhdlq7D15Mit13Unnamed QTL (19)
      TG Chr 18 (42) 37–44Tgq1 3.2B6/129D18Mit50
       Chr 9 (66)
      Interacting QTL.
       44–68Tgq2 2.2
      Suggestive QTL.
      B6D9Mit281Unnamed QTL (28)
       Chr 14 (14) 6–48
      We did not name this QTL because it is below the statistically significant level and overlaps no previously discovered QTL.
       2.0
      Suggestive QTL.
      B6D14Mit60
       Chr 4 (58)
      Interacting QTL.
       30–90Tgq3D4Mit308Lepr (46.7), Angptl3

      (48.5), Cpt2 (54.4)
      % fat Chr 8 (48)
      Interacting QTL.
       42–53Obq16 10.0129D8Mit248
       Chr 12 (2) 0–16
      We did not name this QTL because it is below the statistically significant level and overlaps no previously discovered QTL.
       2.9
      Suggestive QTL.
      129D12Mit182Pomc1 (4.0)
       Chr 6 (0) 0–10Mob2 2.6
      Suggestive QTL.
      129D6Mit86Mob2 (20)Lep (10.5)
       Chr 1 (74) 48–108Obq17 2.3
      Suggestive QTL.
      129D1Mit495Obq8 (21), Obq9 (21)
       Chr 9 (65)
      Interacting QTL.
       0–75Obq18D9Mit281Dob2 (30), Mob8 (31),

      Obq5 (32), Adip5 (33)
      BMI Chr 17 (8) 0–25Obq19 2.9
      Suggestive QTL.
      B6D17Mit143Unnamed QTL (21), Obq4 (22)Igf2r (7.35), Acat2 (7.5), Ppard (13.5)
       Chr 8 (52) 38–72Obq16 2.5
      Suggestive QTL.
      129D8Mit248Obq16 (this study)
       Chr 1 (102) 56–108Obq17 2.4
      Suggestive QTL.
      129D1Mit406Obq8 (21), Obq9 (21)
      cM, centimorgan; QTL, quantitative trait loci. The number of F2 mice is 292 for plasma lipid concentrations and 291 for % fat.
      a Overlapping QTL identified in previous studies.
      b Interacting QTL.
      c Suggestive QTL.
      d We did not name this QTL because it is below the statistically significant level and overlaps no previously discovered QTL.
      Figure thumbnail gr2
      Fig. 2Genome-wide scans (solid lines) and posterior probability densities (broken lines) for the quantitative trait loci (QTL). A, C, E, and G represent the QTL, posterior probability densities, and 95% confidence intervals; B, D, F, and H provide the contributions at the peak of each QTL. Posterior probability density is a likelihood statistic that gives rise to the 95% confidence intervals indicated by gray bars (
      • Sen S.
      • Churchill G.A.
      A statistical framework for quantitative trait mapping.
      ). Homozygosity for B6 alleles is represented by B6/B6, homozygosity for 129 alleles is represented by 129/129, and heterozygosity at a locus is represented by B6/129. Marker locations for each QTL are in parentheses. Error bars represent SEM. Chr, chromosome; cM, centimorgan.
      For plasma non-HDL concentrations, the genome scan is shown in Fig. 1A. The significant chromosome 8 QTL (Fig. 2A; peak LOD 4.4), named Nhdlq1, had a dominant B6 allele for increased non-HDL concentrations (Fig. 2B). This locus confirmed a QTL identified earlier using strains CAST and 129 (
      • Lyons M.A.
      • Wittenburg H.
      • Li R.
      • Walsh K.A.
      • Korstanje R.
      • Churchill G.A.
      • Carey M.C.
      • Paigen B.
      Quantitative trait loci that determine lipoprotein cholesterol levels in an intercross of 129S1/SvImJ and CAST/Ei inbred mice.
      ). Nhdlq4, on chromosome 10 (Fig. 2C; peak LOD 4.0), caused higher non-HDL when homozygous for a recessive 129 allele (Fig. 2D). Two suggestive QTL were discovered at the D6Mit86 locus and the D7Mit141 locus. The pairwise genome scan revealed three significant interactions. Nhdlq1 interacted with the D7Mit294 locus, which we named Nhdlq6. Nhdlq6 did not affect non-HDL concentrations by itself, but its effect in combination with Nhdlq1 on non-HDL was strong (Fig. 3A). When the Nhdlq1 genotype was B6/B6, homozygosity for a recessive allele from strain 129 at Nhdlq6 contributed significantly to increase non-HDL. A second significant interaction was found between Nhdlq4 and the D6Mit86 locus, named Nhdlq5, which was suggestive as a single QTL (Fig. 3B). When the Nhdlq4 genotype was 129/129, the contribution of a recessive B6 allele for increased non-HDL at Nhdlq5 became significant. A third interaction was found between Nhdlq4 and the D15Mit13 locus, which we named Nhdlq7. Nhdlq7 did not affect non-HDL by itself, but its combined effect with Nhdlq4 on non-HDL was dramatic (Fig. 3C). When the Nhdlq4 genotype was 129/129, homozygosity for a recessive B6 allele at Nhdlq7 significantly increased plasma non-HDL.
      Figure thumbnail gr3
      Fig. 3The effects of gene interactions detected by the pairwise genome scan. Homozygosity for B6 alleles is represented by B6/B6, homozygosity for 129 alleles is represented by 129/129, and heterozygosity is represented by B6/129. Y axes show mean values of log non-HDL (A, B, and C), TG (D), and % fat (E). Error bars represent SEM.
      For plasma TG concentrations, we found a significant locus on chromosome 18, which we named Tgq1 (Fig. 2E; peak LOD 3.2) and two suggestive loci on chromosomes 9 and 14. At Tgq1, the heterozygous B6/129 genotype was associated with significantly increased TG concentrations (Fig. 2F). The pairwise genome scan revealed that an interaction at D9Mit281 and D4Mit308, which we named Tgq2 and Tgq3, respectively, affected plasma TG concentrations with statistical significance (Fig. 3D). When the Tgq3 genotype was 129/129, homozygosity for a recessive strain B6 allele at Tgq2 contributed significantly increased TG.
      For obesity measured by % fat, the genome scan is shown in Fig. 1C. The significant chromosome 8 QTL named Obq16 (Fig. 2G; peak LOD 10.0) had an additive 129 allele for higher % fat (Fig. 2H). Three suggestive QTL were discovered on chromosomes 1, 6, and 12. The D6Mit86 locus confirmed a QTL, Mob2, identified earlier using strains B6 and SPRET (
      • Warden C.H.
      • Fisler J.S.
      • Shoemaker S.M.
      • Wen P.Z.
      • Svenson K.L.
      • Pace M.J.
      • Lusis A.J.
      Identification of four chromosomal loci determining obesity in a multifactorial mouse model.
      ). We named this locus Mob2 in the present cross, which shared the parental strain B6 in common with the earlier cross. The D1Mit495 locus confirmed adjacent QTL, Obq8 and Obq9, identified earlier using strains NZO and SM (
      • Taylor B.A.
      • Wnek C.
      • Schroeder D.
      • Phillips S.J.
      Multiple obesity QTLs identified in an intercross between the NZO (New Zealand obese) and the SM (small) mouse strains.
      ). We named this locus Obq17 in the present study. The pairwise genome scan revealed a significant interaction between Obq16 and the D9Mit281 locus, which we named Obq18. Obq18 was not shown to affect % fat by itself, but its combined effect with Obq16 on % fat was dramatic (Fig. 3E). When the Obq16 genotype was 129/129, an additive/codominant allele for higher % fat from strain B6 at Obq18 contributed a significant effect.
      For obesity measured by BMI, the genome scan is shown in Fig. 1D. Three suggestive QTL were discovered on chromosomes 1, 8, and 17. The D17Mit143 locus confirmed QTL identified earlier using either strains AKR/J and C57L/J (
      • Taylor B.A.
      • Phillips S.J.
      Obesity QTLs on mouse chromosomes 2 and 17.
      ) or strains NZO and SM (
      • Taylor B.A.
      • Wnek C.
      • Schroeder D.
      • Phillips S.J.
      Multiple obesity QTLs identified in an intercross between the NZO (New Zealand obese) and the SM (small) mouse strains.
      ). The D8Mit248 locus confirmed a significant QTL, Obq16, identified for % fat in this cross. Thus, we gave this locus the same name, Obq16. The D1Mit406 locus colocalized a QTL, Obq17, identified for % fat in this cross and was given the same name. Two loci, the D2Mit285 locus and the D18Mit4 locus, exceeded the 37% genome-wide adjusted threshold (LOD ≥ 1.9), but their terms did not meet the nominal level in the regression, so we did not report these loci as suggestive QTL (Fig. 1D).
      The multiple regression analyses (Tables 4 and 5) show the effect of each QTL and interactions when considered together. The percent of the total phenotypic variance in F2 mice is best estimated by a multiple regression analysis. For non-HDL, this multiple regression analysis confirmed six QTL and three interactions identified for single gene or pairwise genome-wide scans. Taken together, these QTL and their interactions explained 49.9% of the total F2 phenotypic variance; Nhdlq1 and Nhdlq4 contributed ∼10% each and the others each contributed 2–5% of the total variance. For TG, four single QTL and one interaction explained 24.3% of the total variance. For % fat, five single QTL and one interaction explained 34.3% of the total variance; Obq16 contributed approximately half of the genetic variance. For BMI, three QTL explained 11.1% of the total variance, with each QTL contributing 3–4% of the total variance.
      TABLE 4Multiple regression analyses of variance for non-HDL and TG in 292 (B6 × 129) F2 females
      TraitsChromosomal (Chr)

      Location
      Nearest MarkerDF
      DF indicates degrees of freedom; it includes main effect and any interactions.
      Type III SS
      SS, sum of squares.
      Variance (%)
      Variance indicates the percentage of the total F2 phenotypic variance associated with each marker.
      F ValueP ValueLocus Name
      cM
      Non-HDLChr 8 (38)
      Interacting QTL.
      D8Mit248 60.78410.06.611.58 × 10−6Nhdlq1
      Chr 10 (70)
      Interacting QTL.
      D10Mit35 101.06313.55.382.99 × 10−7Nhdlq4
      Chr 6 (0)
      Interacting QTL.
      D6Mit86 60.3804.83.210.0047Nhdlq5
      Chr 7 (70)D7Mit141 20.1782.34.510.0118
      Chr 7 (10)
      Interacting QTL.
      D7Mit294 60.2763.52.330.0329Nhdlq6
      Chr 15 (0)
      Interacting QTL.
      D15Mit13 60.3754.83.160.0051Nhdlq7
      Chr 8 (38):Chr 7 (10)D8Mit248:D7Mit294 40.2513.23.170.0144Nhdlq1:Nhdlq6
      Chr 10 (70):Chr 6 (0)D10Mit35:D6Mit86 40.2683.43.390.0101Nhdlq4:Nhdlq5
      Chr 10 (70):Chr 15 (0)D10Mit35:D15Mit13 40.3444.44.350.0020Nhdlq4:Nhdlq7
      Totals 2917.87149.9
      TGChr 18 (42)D18Mit50 21,3513.45.590.0042Tgq1
      Chr 9 (66)
      Interacting QTL.
      D9Mit281 62,8507.13.930.0009Tgq2
      Chr14 (14)D14Mit60 21,0552.64.370.0135
      Chr 4 (58)
      Interacting QTL.
      D4Mit308 62,5966.53.580.0019Tgq3
      Chr 9 (66):Chr 4 (58)D9Mit281:D4Mit308 41,9004.73.930.0040Tgq2:Tgq3
      Totals 29140,24924.3
      a DF indicates degrees of freedom; it includes main effect and any interactions.
      b SS, sum of squares.
      c Variance indicates the percentage of the total F2 phenotypic variance associated with each marker.
      d Interacting QTL.
      TABLE 5Multiple regression analyses of variance for % fat and BMI in 294 (B6 × 129) F2 females
      TraitsChromosomal (Chr)

      Location
      Nearest MarkerDF
      DF indicates degrees of freedom; it includes main effect and any interactions.
      Type III SS
      SS, sum of squares.
      Variance (%)
      Variance indicates the percentage of the total F2 phenotypic variance associated with each marker.
      F ValueP ValueLocus Name
      cM
      % fatChr 8 (48)
      Interacting QTL.
      D8Mit24861,60616.210.601.36 × 10−10Obq16
      Chr 12 (2)D12Mit18224004.07.910.0005
      Chr 6 (0)D6Mit8622292.34.520.0117Mob2
      Chr 1 (74)D1Mit49522722.85.390.0050Obq17
      Chr 9 (65)
      Interacting QTL.
      D9Mit28165045.13.320.0036Obq18
      Chr 8 (48):Chr 9 (65)D8Mit248:D9Mit28143553.63.520.0081Obq16:Obq18
      Totals2909,90934.3
      BMIChr 17 (8)D17Mit14320.0364.36.910.0012Obq19
      Chr 8 (52)
      Interacting QTL.
      D8Mit24820.0273.25.190.0061Obq16
      Chr 1 (102)
      Interacting QTL.
      D1Mit40620.0313.65.880.0031Obq17
      Totals2930.84411.1
      The number of F2 mice is 291 for % fat.
      a DF indicates degrees of freedom; it includes main effect and any interactions.
      b SS, sum of squares.
      c Variance indicates the percentage of the total F2 phenotypic variance associated with each marker.
      d Interacting QTL.

      DISCUSSION

      In the present study, we describe two inbred mouse strains, B6 and 129, that display different plasma levels of non-HDL cholesterol and TG and degrees of obesity when fed a high-fat diet. Three-step QTL analyses on 294 (B6 × 129) F2 progeny resulted in the localization of six QTL for non-HDL, four QTL for TG, five QTL for % fat, three QTL for BMI, and five gene interactions.
      For plasma non-HDL concentrations, we identified four main-effect QTL (Nhdlq1, Nhdlq4, Nhdlq5, and the D7Mit141 locus) and two additional QTL by gene interactions (Nhdlq6 and Nhdlq7). Previously, our group mapped a chromosomal locus in a (CAST × 129) F2 intercross that determines non-HDL levels to chromosome 8 (cM 20–60) and named it Nhdlq1 (
      • Lyons M.A.
      • Wittenburg H.
      • Li R.
      • Walsh K.A.
      • Korstanje R.
      • Churchill G.A.
      • Carey M.C.
      • Paigen B.
      Quantitative trait loci that determine lipoprotein cholesterol levels in an intercross of 129S1/SvImJ and CAST/Ei inbred mice.
      ). Because the present study confirmed the previously reported QTL using a cross having a parental strain, 129S1/SvImJ, in common with the cross used in the earlier study, we named the locus Nhdlq1 in accordance with the International Committee on Standardized Genetic Nomenclature for Mice. Potential candidate genes for Nhdlq1 are the gene (Lpl; cM 33.0) coding for LPL and the gene (Cpe; cM 32.6) coding for carboxypeptidase E, which produces biologically active forms of proinsulin and proopiomelanocortin. Mice possessing the fat mutation (a spontaneous mutation in the Cpe gene) exhibited prominent obesity and higher plasma non-HDL concentrations relative to controls after consuming a high-fat diet (
      • Bouchard G.
      • Johnson D.
      • Carver T.
      • Paigen B.
      • Carey M.C.
      Cholesterol gallstone formation in overweight mice establishes that obesity per se is not linked directly to cholelithiasis risk.
      ). Nhdlq1 coincidentally colocalizes with a QTL for % fat, Obq16, in the present cross. Nhdlq6 maps to the region containing the apolipoprotein gene cluster (Apoc1, Apoc2, Apoc4, and Apoe; cM 4.0). Interestingly, APOC2 is a cofactor for LPL (
      • Merkel M.
      • Eckel R.H.
      • Goldberg I.J.
      Lipoprotein lipase: genetics, lipid uptake, and regulation.
      ). The gene interaction between Nhdlq1 and Nhdlq6 makes Lpl and Apoc2 excellent candidates for genes underlying these QTL. Likewise, Nhdlq4 interacted independently with Nhdlq5 and Nhdlq7. These gene interactions may give clues to the candidate genes. Nhdlq4 colocalized with a QTL for phospholipid transfer protein (PLTP) activity, Pltp2, found previously using an (SM × NZB) F2 intercross (
      • Korstanje R.
      • Albers J.J.
      • Wolfbauer G.
      • Li R.
      • Tu A.Y.
      • Churchill G.A.
      • Paigen B.J.
      Quantitative trait locus mapping of genes that regulate phospholipid transfer activity in SM/J and NZB/BlNJ inbred mice.
      ). PLTP is responsible for the transfer of phospholipids from VLDL to HDL (
      • Tall A.R.
      • Krumholz S.
      • Olivecrona T.
      • Deckelbaum R.J.
      Plasma phospholipid transfer protein enhances transfer and exchange of phospholipids between very low density lipoproteins and high density lipoproteins during lipolysis.
      ), suggesting that the gene underlying Nhdlq4 might determine plasma non-HDL levels by regulating PLTP activity. An excellent candidate gene for Nhdlq4 is the gene (Apof; cM 73.0) coding for a lipid transfer inhibitor protein, Apo F (
      • Wang X.
      • Driscoll D.M.
      • Morton R.E.
      Molecular cloning and expression of lipid transfer inhibitor protein reveals its identity with apolipoprotein F.
      ). Previously, our group reported a single nucleotide polymorphism that causes an amino acid change in the protein between B6 and 129 strains (
      • Korstanje R.
      • Albers J.J.
      • Wolfbauer G.
      • Li R.
      • Tu A.Y.
      • Churchill G.A.
      • Paigen B.J.
      Quantitative trait locus mapping of genes that regulate phospholipid transfer activity in SM/J and NZB/BlNJ inbred mice.
      ).
      For TG levels, we identified three main-effect QTL (Tgq1, Tgq2, and the D14Mit60 locus) and, by gene interactions, an additional QTL (Tgq3). Potential candidate genes for Tgq3 are the gene (Lepr; cM 46.7) coding for the receptor of leptin; the gene (Cpt2, cM 54.4) coding for a mitochondrial fatty acid transporter, carnitine palmitoyltransferase 2; and the gene (Angptl3, cM 48.5) coding for angiopoietin-like 3. Tgq2 colocalized with a QTL previously identified in a (B6 × KK-Ay) F2 intercross (
      • Suto J.
      • Matsuura S.
      • Yamanaka H.
      • Sekikawa K.
      Quantitative trait loci that regulate plasma lipid concentration in hereditary obese KK and KK-Ay mice.
      ) but did not map near any genes known to play a prominent role in lipoprotein or lipid metabolism. Alternatively, genes underlying Tgq2 are entirely novel genes that might otherwise not have been considered. The interaction between Tgq2 and Tgq3 may give clues to the underlying genes’ identities.
      To identify genetic loci that affect the development of obesity in response to the high-fat diet, we measured two different traits that reflect obesity, % fat and BMI. For % fat, we identified four main-effect QTL (Obq16, Obq17, Mob2, and the D12Mit182 locus) and one additional QTL (Obq18) by its interaction with Obq16. For BMI, we identified three suggestive main-effect QTL (Obq16, Obq17, and Obq19). Because log BMI was positively correlated with % fat (P < 0.0001), two of three QTL for BMI, the D8Mit248 locus and the D1Mit406 locus, colocalized with QTL for % fat and for Obq16 and Obq17, respectively. The discrepancy between QTL obtained from these traits might be reflected from variations in body length. The suggestive D12Mit182 locus maps to the region of the gene (Pomc1; cM 4.0) coding for proopiomelanocortin-α, mutation of which causes monogenic obesity through the liptin-melanocortin signaling pathway (
      • Cummings D.E.
      • Schwartz M.W.
      Genetics and pathophysiology of human obesity.
      ). The suggestive QTL on chromosome 6 colocalized with a QTL for fat pad weight, Mob2, found previously using a (B6 × SPRET) × B6 backcross (
      • Warden C.H.
      • Fisler J.S.
      • Shoemaker S.M.
      • Wen P.Z.
      • Svenson K.L.
      • Pace M.J.
      • Lusis A.J.
      Identification of four chromosomal loci determining obesity in a multifactorial mouse model.
      ). This genetic locus maps to the region of the gene (Lep; cM 10.5) coding for leptin. The Obq17 locus colocalized with an adjacent QTL for obesity, Obq8 and Obq9, which were discovered previously using progeny of (SM × NZO) F2 intercrosses (
      • Taylor B.A.
      • Wnek C.
      • Schroeder D.
      • Phillips S.J.
      Multiple obesity QTLs identified in an intercross between the NZO (New Zealand obese) and the SM (small) mouse strains.
      ). The Obq18 locus has been found repeatedly using (SWR × AKR) F2, (B6 × CAST) F2, and (B6 × KK) F2 intercross progeny fed a high-fat diet (
      • West D.B.
      • Goudey-Lefevre J.
      • York B.
      • Truett G.E.
      Dietary obesity linked to genetic loci on chromosomes 9 and 15 in a polygenic mouse model.
      ,
      • Mehrabian M.
      • Wen P.Z.
      • Fisler J.
      • Davis R.C.
      • Lusis A.J.
      Genetic loci controlling body fat, lipoprotein metabolism, and insulin levels in a multifactorial mouse model.
      ,
      • Taylor B.A.
      • Tarantino L.M.
      • Phillips S.J.
      Gender-influenced obesity QTLs identified in a cross involving the KK type II diabetes-prone mouse strain.
      ). Obq19 colocalized with a QTL found previously using progeny of (AKR × C57L) and (SM × NZO) F2 intercrosses fed a high-fat diet (
      • Taylor B.A.
      • Wnek C.
      • Schroeder D.
      • Phillips S.J.
      Multiple obesity QTLs identified in an intercross between the NZO (New Zealand obese) and the SM (small) mouse strains.
      ,
      • Taylor B.A.
      • Phillips S.J.
      Obesity QTLs on mouse chromosomes 2 and 17.
      ). Potential candidate genes for Obq19 are the gene Ppard (cM 13.5), coding for an important transcriptional factor that regulates glucose and fatty acid metabolism; the gene Igf2r (cM 7.35), coding for an insulin-like growth factor II receptor; and the gene Acat2 (cM 7.5), coding for an acetyl-CoA acetyltransferase 2 that catalyzes the synthesis of cytosolic acetoacetyl-CoA, a precursor of cholesterol and other steroids. However, high-fat diets in the most previous reports do not contain substantial amounts of cholesterol or any cholate, and this difference in diets might affect the identities of colocalizing QTL between this study and previous studies.
      Reed and colleagues (
      • Reed D.R.
      • Li X.
      • McDaniel A.H.
      • Lu K.
      • Li S.
      • Tordoff M.G.
      • Price R.A.
      • Bachmanov A.A.
      Loci on chromosomes 2, 4, 9, and 16 for body weight, body length, and adiposity identified in a genome scan of an F2 intercross between the 129P3/J and C57BL/6ByJ mouse strains.
      ) carried out a study of both male and female progeny of an F2 intercross between mice of the C57BL/6ByJ and 129P3/J strains to identify QTL for body weight, body length, and adiposity. Of the QTL found in the present study, only Obq18 colocalized with a QTL for adiposity, Adip5, reported by Reed et al. However, these investigators reported that the effect of Adip5 is limited to F2 males and is not found in females (peak LOD < 1.0). Whereas a QTL on chromosome 16, Adip9, was found to interact with Adip5 by Reed et al., Obq18 interacted with Obq16 on chromosome 8 in the present study, suggesting that the gene underlying Adip5 is not identical to the one underlying Obq16. The QTL identified by Reed et al. in the (C57BL/6ByJ × 129P3/J) F2 cross are based on analyses of mice fed a normal chow diet, and it is unlikely that these loci would similarly affect the development of obesity in response to a high-fat diet. Indeed, obesity is a complex trait, reflecting the effect of a network of genes, and it is affected by diet, age, gender, and exercise (
      • Brockmann G.A.
      • Bevova M.R.
      Using mouse models to dissect the genetics of obesity.
      ).
      Several spontaneous single-gene mutations causing obesity, such as agouti yellow (Ay), obese (Lepob), diabetes (Leprdb), fat (Cpe fat), tubby (Tubtub), and mahogany (Atrnmg), have been identified in inbred mice (
      • Justice M.J.
      Capitalizing on large-scale mouse mutagenesis screens.
      ,
      • Butler A.A.
      • Cone R.D.
      Knockout models resulting in the development of obesity.
      ). These mutations, however, do not account for the wide variation of obesity in the general human population. Some human pedigree studies provide clear genetic evidence of oligogenic or polygenic predisposition for obesity, indicating that obesity is a complex trait (
      • Perusse L.
      • Rice T.
      • Chagnon Y.C.
      • Despres J.P.
      • Lemieux S.
      • Roy S.
      • Lacaille M.
      • Ho-Kim M.A.
      • Chagnon M.
      • Province M.A.
      • Rao D.C.
      • Bouchard C.
      A genome-wide scan for abdominal fat assessed by computed tomography in the Quebec Family Study.
      ). Indeed, mice of different inbred strains exhibit wide variation in body weight and predisposition to spontaneous or diet-induced obesity. Multiple modifier genes likely contribute to the variation. Crosses between mice of various strains have identified ∼100 chromosomal loci that contribute to obesity (
      • Brockmann G.A.
      • Bevova M.R.
      Using mouse models to dissect the genetics of obesity.
      ).
      The present study discovered five epistatic interactions for plasma non-HDL and TG concentrations and % fat (Fig. 3). The pairwise genome scans revealed four significant QTL, Nhdlq6, Nhdlq7, Tgq3, and Obq18, that did not affect traits by a single locus but affected the traits with the counterpart locus. When a genotype of one locus was B6/B6, homozygosity for a 129 allele at the counterpart locus contributed significantly to the effect on the trait. This evidence might facilitate in vitro assays to test candidate genes.
      In summary, by performing a QTL analysis of a (B6 × 129) F2 female cohort, we identified chromosomal regions that affect plasma non-HDL and TG concentrations and obesity in mice with backgrounds that are a combination of B6 and 129. Knowledge of the primary genetic determinants of plasma lipid concentrations and obesity will enhance our understanding of lipoprotein metabolism and likely provide novel molecular targets for metabolic obesity.

      Acknowledgments

      This work was funded by AstraZeneca, Sweden, and the National Institutes of Health (Grant CA-34196). The authors thank Eric F. Taylor for excellent technical assistance and Ray A. Lambert and Jennifer L. Smith for helping to prepare the manuscript. N.I. is supported by the Japan Heart Foundation/Bayer Yakuhin Research Grant Abroad and by a Japan Heart Foundation/Pfizer Grant for Research on Hypertension, Hyperlipidemia, and Vascular Metabolism.

      References

        • Isomaa B.
        • Almgren P.
        • Tuomi T.
        • Forsen B.
        • Lahti K.
        • Nissen M.
        • Taskinen M.R.
        • Groop L.
        Cardiovascular morbidity and mortality associated with the metabolic syndrome.
        Diabetes Care. 2001; 24: 683-689
        • Allayee H.
        • Ghazalpour A.
        • Lusis A.J.
        Using mice to dissect genetic factors in atherosclerosis.
        Arterioscler. Thromb. Vasc. Biol. 2003; 23: 1501-1509
        • Smith J.
        Quantitative trait locus mapping for atherosclerosis susceptibility.
        Curr. Opin. Lipidol. 2003; 14: 499-504
        • Brockmann G.A.
        • Bevova M.R.
        Using mouse models to dissect the genetics of obesity.
        Trends Genet. 2002; 18: 367-376
        • Wang X.
        • Paigen B.
        Quantitative trait loci and candidate genes regulating HDL cholesterol: a murine chromosome map.
        Arterioscler. Thromb. Vasc. Biol. 2002; 22: 1390-1401
        • Paigen B.
        • Svenson K.L.
        • Peters L.L.
        Diet effects on plasma lipids and susceptibility to atherosclerosis. MPD:99. Mouse Phenome Database (MPD). Bar Harbor, ME: The Jackson Laboratory.
        at http://www.jax.org/phenome
        Date accessed: December , 2003
        • Korstanje R.
        • Paigen B.
        From QTL to gene: the harvest begins.
        Nat. Genet. 2002; 31: 235-236
        • Smithies O.
        • Maeda N.
        Gene targeting approaches to complex genetic diseases: atherosclerosis and essential hypertension.
        Proc. Natl. Acad. Sci. USA. 1995; 92: 5266-5272
        • Sigmund C.D.
        Viewpoint: are studies in genetically altered mice out of control?.
        Arterioscler. Thromb. Vasc. Biol. 2000; 20: 1425-1429
        • Ishimori N.
        • Li R.
        • Kelmenson P.M.
        • Korstanje R.
        • Walsh K.A.
        • Churchill G.A.
        • Forsman-Semb K.
        • Paigen B.
        Quantitative trait loci analysis for plasma HDL-cholesterol concentrations and atherosclerosis susceptibility between inbred mouse strains C57BL/6J and 129S1/SvImJ.
        Arterioscler. Thromb. Vasc. Biol. 2004; 24: 161-166
        • Nishina P.M.
        • Verstuyft J.
        • Paigen B.
        Synthetic low and high fat diets for the study of atherosclerosis in the mouse.
        J. Lipid Res. 1990; 31: 859-869
        • Nishina P.M.
        • Lowe S.
        • Verstuyft J.
        • Naggert J.K.
        • Kuypers F.A.
        • Paigen B.
        Effects of dietary fats from animal and plant sources on diet-induced fatty streak lesions in C57BL/6J mice.
        J. Lipid Res. 1993; 34: 1413-1422
        • de Faria E.
        • Fong L.G.
        • Komaromy M.
        • Cooper A.D.
        Relative roles of the LDL receptor, the LDL receptor-like protein, and hepatic lipase in chylomicron remnant removal by the liver.
        J. Lipid Res. 1996; 37: 197-209
        • Nagy T.R.
        • Clair A.L.
        Precision and accuracy of dual-energy X-ray absorptiometry for determining in vivo body composition of mice.
        Obes. Res. 2000; 8: 392-398
        • Mu J.L.
        • Naggert J.K.
        • Svenson K.L.
        • Collin G.B.
        • Kim J.H.
        • McFarland C.
        • Nishina P.M.
        • Levine D.M.
        • Williams K.J.
        • Paigen B.
        Quantitative trait loci analysis for the differences in susceptibility to atherosclerosis and diabetes between inbred mouse strains C57BL/6J and C57BLKS/J.
        J. Lipid Res. 1999; 40: 1328-1335
        • Sen S.
        • Churchill G.A.
        A statistical framework for quantitative trait mapping.
        Genetics. 2001; 159: 371-387
        • Churchill G.A.
        • Doerge R.W.
        Empirical threshold values for quantitative trait mapping.
        Genetics. 1994; 138: 963-971
        • Biola O.
        • Angel J.M.
        • Avner P.
        • Bachmanov A.A.
        • Belknap J.K.
        • Bennett B.
        • Blankenhorn E.P.
        • Blizard D.A.
        • Bolivar V.
        • Brockmann G.A.
        • Buck K.J.
        • Bureau J.F.
        • Casley W.L.
        • Chesler E.J.
        • Cheverud J.M.
        • Churchill G.A.
        • Cooke M.
        • Crabbe J.C.
        • Crusio W.E.
        • Darvasi A.
        • de Haan G.
        • Dermant P.
        • Doerge R.W.
        • Elliot R.W.
        • Farber C.R.
        • Flaherty L.
        • Flint J.
        • Gershenfeld H.
        • Gibson J.P.
        • Gu J.
        • Gu W.
        • Himmelbauer H.
        • Hitzemann R.
        • Hsu H.C.
        • Hunter K.
        • Iraqi F.F.
        • Jansen R.C.
        • Johnson T.E.
        • Jones B.C.
        • Kempermann G.
        • Lammer F.
        • Lu L.
        • Manly K.F.
        • Matthews D.B.
        • Medrano J.F.
        • Mehrabian M.
        • Mittlemann G.
        • Mock B.A.
        • Mogil J.S.
        • Montagutelli X.
        • Morahan G.
        • Mount J.D.
        • Nagase H.
        • Nowakowski R.S.
        • O'Hara B.F.
        • Osadchuck A.V.
        • Paigen B.
        • Palmer A.A.
        • Pierce J.L.
        • Pomp D.
        • Rosemann M.
        • Rosen G.D.
        • Schalkwyk L.C.
        • Seltzer Z.
        • Settle S.
        • Shimomura K.
        • Shou S.
        • Sikela J.M.
        • Siracusa L.D.
        • Spearow J.L.
        • Teuscher C.
        • Threadgill D.W.
        • Toth L.A.
        • Toye A.A.
        • Vadasz C.
        • Van Zant G.
        • Wakeland E.
        • Williams R.W.
        • Zhang H.G.
        • Zou F.
        The nature and identification of quantitative trait loci: a community's view.
        Nat. Rev. Genet. 2003; 4: 911-916
        • Lyons M.A.
        • Wittenburg H.
        • Li R.
        • Walsh K.A.
        • Korstanje R.
        • Churchill G.A.
        • Carey M.C.
        • Paigen B.
        Quantitative trait loci that determine lipoprotein cholesterol levels in an intercross of 129S1/SvImJ and CAST/Ei inbred mice.
        Physiol. Genomics. 2004; 17: 60-68
        • Warden C.H.
        • Fisler J.S.
        • Shoemaker S.M.
        • Wen P.Z.
        • Svenson K.L.
        • Pace M.J.
        • Lusis A.J.
        Identification of four chromosomal loci determining obesity in a multifactorial mouse model.
        J. Clin. Invest. 1995; 95: 1545-1552
        • Taylor B.A.
        • Wnek C.
        • Schroeder D.
        • Phillips S.J.
        Multiple obesity QTLs identified in an intercross between the NZO (New Zealand obese) and the SM (small) mouse strains.
        Mamm. Genome. 2001; 12: 95-103
        • Taylor B.A.
        • Phillips S.J.
        Obesity QTLs on mouse chromosomes 2 and 17.
        Genomics. 1997; 43: 249-257
        • Bouchard G.
        • Johnson D.
        • Carver T.
        • Paigen B.
        • Carey M.C.
        Cholesterol gallstone formation in overweight mice establishes that obesity per se is not linked directly to cholelithiasis risk.
        J. Lipid Res. 2002; 43: 1105-1113
        • Merkel M.
        • Eckel R.H.
        • Goldberg I.J.
        Lipoprotein lipase: genetics, lipid uptake, and regulation.
        J. Lipid Res. 2002; 43: 1997-2006
        • Korstanje R.
        • Albers J.J.
        • Wolfbauer G.
        • Li R.
        • Tu A.Y.
        • Churchill G.A.
        • Paigen B.J.
        Quantitative trait locus mapping of genes that regulate phospholipid transfer activity in SM/J and NZB/BlNJ inbred mice.
        Arterioscler. Thromb. Vasc. Biol. 2004; 24: 155-160
        • Tall A.R.
        • Krumholz S.
        • Olivecrona T.
        • Deckelbaum R.J.
        Plasma phospholipid transfer protein enhances transfer and exchange of phospholipids between very low density lipoproteins and high density lipoproteins during lipolysis.
        J. Lipid Res. 1985; 26: 842-851
        • Wang X.
        • Driscoll D.M.
        • Morton R.E.
        Molecular cloning and expression of lipid transfer inhibitor protein reveals its identity with apolipoprotein F.
        J. Biol. Chem. 1999; 274: 1814-1820
        • Suto J.
        • Matsuura S.
        • Yamanaka H.
        • Sekikawa K.
        Quantitative trait loci that regulate plasma lipid concentration in hereditary obese KK and KK-Ay mice.
        Biochim. Biophys. Acta. 1999; 1453: 385-395
        • Cummings D.E.
        • Schwartz M.W.
        Genetics and pathophysiology of human obesity.
        Annu. Rev. Med. 2003; 54: 453-471
        • West D.B.
        • Goudey-Lefevre J.
        • York B.
        • Truett G.E.
        Dietary obesity linked to genetic loci on chromosomes 9 and 15 in a polygenic mouse model.
        J. Clin. Invest. 1994; 94: 1410-1416
        • Mehrabian M.
        • Wen P.Z.
        • Fisler J.
        • Davis R.C.
        • Lusis A.J.
        Genetic loci controlling body fat, lipoprotein metabolism, and insulin levels in a multifactorial mouse model.
        J. Clin. Invest. 1998; 101: 2485-2496
        • Taylor B.A.
        • Tarantino L.M.
        • Phillips S.J.
        Gender-influenced obesity QTLs identified in a cross involving the KK type II diabetes-prone mouse strain.
        Mamm. Genome. 1999; 10: 963-968
        • Reed D.R.
        • Li X.
        • McDaniel A.H.
        • Lu K.
        • Li S.
        • Tordoff M.G.
        • Price R.A.
        • Bachmanov A.A.
        Loci on chromosomes 2, 4, 9, and 16 for body weight, body length, and adiposity identified in a genome scan of an F2 intercross between the 129P3/J and C57BL/6ByJ mouse strains.
        Mamm. Genome. 2003; 14: 302-313
        • Justice M.J.
        Capitalizing on large-scale mouse mutagenesis screens.
        Nat. Rev. Genet. 2000; 1: 109-115
        • Butler A.A.
        • Cone R.D.
        Knockout models resulting in the development of obesity.
        Trends Genet. 2001; 17: S50-S54
        • Perusse L.
        • Rice T.
        • Chagnon Y.C.
        • Despres J.P.
        • Lemieux S.
        • Roy S.
        • Lacaille M.
        • Ho-Kim M.A.
        • Chagnon M.
        • Province M.A.
        • Rao D.C.
        • Bouchard C.
        A genome-wide scan for abdominal fat assessed by computed tomography in the Quebec Family Study.
        Diabetes. 2001; 50: 614-621