Genetic modifiers interact with Cpefat to affect body weight, adiposity, and hyperglycemia
Gayle B. Collin1,*,
Terry P. Maddatu1,*,
aunak Sen1,2 and
Jürgen K. Naggert1
1 The Jackson Laboratory, Bar Harbor, Maine
2 Department of Epidemiology and Biostatistics, University of California, San Francisco, California
 |
ABSTRACT
|
---|
Obesity and Type II diabetes are complex diseases in the human population. The existence of a large number of contributing loci and gene-gene as well as gene-environment interactions make it difficult to identify the disease genes underlying these complex traits. In mouse models of obesity and Type II diabetes such as the murine fat mutation, genetic crosses can be used to dissect the genetic complexity influencing the observed phenotypes. The underlying defect in the fat mutant is a Ser202Pro change in carboxypeptidase E (CPE), an enzyme responsible for the final proteolytic processing step of prohormone intermediates. On the HRS/J (HRS) inbred strain background, mice homozygous for the fat mutation exhibit early onset hyperinsulinemia followed by postpubertal moderate obesity without hyperglycemia. In contrast, on the C57BLKS/J (BKS) genetic background, fat/fat mice become severely obese, hyperinsulinemic, and hyperglycemic. Therefore, in the Cpefat genetic model, the fat mutation is necessary but not sufficient for the development of obesity, Type II diabetes, and related metabolic disorders. To dissect the susceptibility loci responsible for modifying obesity- and diabetes-associated traits, we characterized, both genetically and phenotypically, fat/fat male progeny from a large intercross between BKS. HRS-fat/fat and HRS-+/+ mice. Four major loci were mapped, including a locus for body weight (body weight 1) on chromosome 14; a locus for hyperglycemia (fat-induced diabetes 1) on chromosome 19; a locus for hyperglycemia, hyperinsulinemia, and hypercholesterolemia (fat-induced diabetes 2) on chromosome 5; and a locus for adiposity and body weight (fat-induced adiposity 1) on chromosome 11. The identification of these interacting genetic determinants for obesity and Type II diabetes may allow better definition of the obesity/diabetes-related hormone signaling pathways and ultimately may provide new insights into the pathogenesis of these complex diseases.
quantitative trait loci; Type II diabetes
 |
INTRODUCTION
|
---|
OBESITY AND RELATED METABOLIC DISORDERS such as insulin resistance and Type II diabetes are major health problems in the human population. Presumably, the genetic dissection of these complex traits in humans has been difficult because of their polygenic nature, the genetic heterogeneity, and gene-gene and gene-environmental interactions on the disease traits (4, 7, 23). For this reason, genetic studies of obesity and diabetes have extended to the use of mouse models. The mouse constitutes a useful model system because a large number of animals can be bred while controlling for genetic and environmental influences. These advantages are necessary to study the epistatic effects of mutated genes on susceptibility loci.
Five single-gene mutations that result in an obese phenotype have been identified in mice, one of which is the fat mutation. The fat mutation is a single nucleotide change in the carboxypeptidase E (CPE) gene resulting in a 202Ser to Pro change in the CPE protein in fat mice (22). The CPE gene (Cpe) encodes an enzyme responsible for the processing of hormones, such as insulin and proopiomelanocortin. CPE is present at high concentrations in secretory granules of the pancreatic islets and is required for the excision of residual dibasic residues remaining at the carboxy-terminal end of propeptides after processing by the endolytic prohormone convertases PCSK1 and PCSK2 (10). The mutation in Cpefat renders the enzyme unstable, and it is ultimately degraded in the endoplasmic reticulum. The reduction in CPE activity in turn leads to a decrease of biologically active hormones such as insulin. CPE defects have also been observed in other tissues, including the hypothalamus, which suggests that a number of additional prohormones may be implicated in the obesity phenotype such as neuropeptides, melanin-stimulating hormone (3), or melanin-concentrating hormone (25).
The fat mutation initially arose on the inbred HRS/J (HRS) strain (7). HRS-fat/fat mice exhibit early chronic hyperinsulinemia followed by mature onset obesity without hyperglycemia. When the fat mutation was transferred to the C57BLKS/J (BKS) background, male mice harboring the mutation in a homozygous state became hyperglycemic after the sixth backcross generation (N6). In addition, BKS-fat/fat mice exhibit elevated plasma triglyceride, high-density lipoprotein (HDL), and total cholesterol levels. As the background strain difference with respect to diabetes development demonstrates, Cpefat is necessary, but not sufficient, for the development of diabesity (obesity/diabetes) in the Cpefat model. Other susceptibility loci must interact with fat to produce the hyperglycemia and other metabolic phenotypes observed in BKS-fat/fat mice. To dissect the susceptibility loci interacting with Cpefat, we performed a genetic analysis of male mice homozygous for the Cpefat mutation from a large (BKS x HRShr/+ or +/+) F1 Cpefat/+ intercross [the inbred HRS strain also carries a mutation in the hairless gene (hr)]. For temporal phenotypes such as body weight and plasma glucose, we constructed sets of derived phenotypes to capture key aspects of longitudinal measurements such as rate, flux, and severity of each phenotype. Here, we present the outcome of a genome-wide screen for susceptibility loci that interact with Cpefat in the development of obesity- and diabetes-related disorders.
 |
MATERIALS AND METHODS
|
---|
Mice
Mice in this study were ad libitum fed on a NIH31 diet with 6% fat (PMI Feeds; St. Louis, MO) and provided unlimited access to water (HCl acidified, pH 2.83.2) in a temperature- and humidity-controlled setting with a 12:12-h light-dark cycle. Because fat/fat mice have difficulty breeding, ovaries from hyperglycemic female insipient congenic BKS.HRS-Cpefat/Cpefat mice (BKS-fat/fat, cohorts I and II at backcross N6 and cohort III at backcross N10) were transplanted into NOD.CB17-Prkdascid/J homozygous female mice, and the recipients were mated to HRShr/+ or +/+ male mice (the HRS stock is maintained at The Jackson Laboratory by hr/+ x hr/+ matings). F1 offspring were intercrossed to produce a total of 2,838 F2 mice in 3 cohorts. Although obesity is observed in both male and female BKS-fat/fat mice, hyperglycemia in females is rare. Therefore, we selected only F2 fat/fat male progeny for the quantitative trait loci (QTL) analysis.
Phenotypic Measurements
Body weights.
Mice were weighed monthly until termination at 24 (cohorts II and III) or 30 wk (cohort I) of age. See Table 1 for a schedule of phenotypic measurements.
Biochemical assays.
Whole blood was collected via the orbital sinus biweekly (between 7 and 10 AM) using an EDTA-coated capillary tube. Samples were centrifuged, and plasma was used for all biochemical assays. Plasma glucose, cholesterol, and triglyceride levels were measured using commercial colorimetric assays (Sigma; St. Louis, MO; and Roche; Indianapolis, IN). To determine HDL-cholesterol levels, low-density and very-low-density lipoprotein particles were precipitated with polyethylene glycol (15). To determine plasma triglyceride levels, free glycerol was subtracted from the total glycerol content. Plasma insulin levels were measured using a radioimmunoassay kit (Linco; St. Louis, MI). Plasma glucose levels were monitored biweekly, whereas lipid and insulin levels were determined at 16 wk of age and at termination.
Necropsies.
Subcutaneous (rostral and caudal), retroperitoneal, and epididymal fat pads were isolated and weighed at termination. Total fat pad weights were defined as the sum of the fat pads described here. Adiposity indexes were calculated by dividing the total fat pad weight by carcass weight and multiplying by 100. In addition to fat pad weights, body lengths (nose to tail) and girths of all animals were measured.
Allele-specific PCR.
DNA was isolated from tail tissue by using proteinase K digestion and salt precipitation (28). Because Cpefat and hr were segregating in the F2 intercross progeny, DNAs were PCR amplified with allele-specific primers to determine the genotypes at both loci. PCR conditions for Cpe primers have been previously described (19). Cycling conditions for PCR-amplification of hr were as follows: 94°C for 2 min followed by 49 cycles of 94°C for 20 s, 50°C for 30 s, and 72°C for 30 s. A final extension was done at 72°C for 7 min. PCRs were done in a 10-µl volume containing 1x PCR buffer (Roche), 100 ng genomic DNA, 1.5 mM MgCl2, 200 nM dNTPs, 0.25 units Taq polymerase, 0.2 µM primer F1 (5'-agggaggtcagagcctctg-3'), 0.1 µM primer R1(5'-ggtcctcctgtttgcttgg-3'), and 0.1 µM primer F2 (5'-gccactccttgaacctgtg-3'). DNAs were separated on a 3% metaphor-1% agarose gel (BMA; Rockland, ME), stained with ethidium bromide, and examined under ultraviolet light. DNA from heterozygous hr/+ animals yielded two fragments of 910 and 581 bps, respectively.
Genome scan.
To identify polymorphic regions between the BKS and HRS strains across the genome, DNA from parental BKS-fat/fat N6, BKS-fat/fat N10, and HRS-+/+ mice was typed for single stranded length polymorphism (SSLP) markers (Research Genetics) using PCR assays as previously described (14). Initially, a total of 87 SSLP markers at an average spacing of 20 cM was tested for the genome scan, with the largest gap being 47 cM between markers D1Mit26 and D1Mit17 (marker information available upon request). Additional markers were tested for chromosomal regions showing significant linkage. At this resolution, we detected (1/87 markers tested in the scan) a region on chromosome 17 that differed between BKS-fat/fat N6 and N10 animals. N6 animals were all HRS genotype at D17Mit142, whereas N10 animals were segregating with HRS and BKS alleles. Because these areas and other HRS contaminated regions may exist [up to
2% of the genome (2)], a BKS-modifying QTL may have been missed. Of the 2,838 F2 mice generated from the intercross, DNA from 282 fat/fat F2 male progeny was PCR amplified. Only 10% of fat/fat male mice instead of the expected 12.5% were available because some mice died early or were lost before the completion of the study. Approximately, 83% of all possible genotypes were complete. To determine whether a QTL modified a trait in absence of the Cpe defect, a genome scan was performed on DNA from 48 F2-fat/+ and 24 F2-+/+ mice.
Statistical Analysis
Before data analysis, our data were thoroughly examined for mice exhibiting spontaneous, extraneous ailments such as malocclusions. These animals were removed from the data set before analysis. The phenotypic profiles of the F2 progeny were obtained using one-way ANOVA. In examining the longitudinal data of the male F2-fat/fat mice, we noted different patterns of disease progression. For example, some mice became hyperglycemic early and sustained their hyperglycemia, others were transiently hyperglycemic, and others showed a slow but moderate rise in plasma glucose levels. To address the longitudinal nature of our phenotypes, we created derived phenotypes for body weight and plasma glucose. This allowed us to obtain results that have a clear biological interpretation, i.e., to detect modifiers that act early or late in the disease process (change), that affect the rate of disease progression (curvature), that affect disease severity (final), or that remain constitutively expressed throughout the time course (average). This also makes our results comparable with future studies that may measure body weight progression at different time points or even at irregular time intervals. We constructed the following sets of derived phenotypes.
Body weight.
Let bi be the log body weight at week i.
 | (1) |
 | (2) |
 | (3) |
 | (4) |
Plasma glucose.
Let gi be the log plasma glucose level of the mouse at week i.
 | (5) |
 | (6) |
 | (7) |
 | (8) |
See Fig. 1 for a visual explanation. Termination age, the time at which necropsies are performed, differed between cohorts (24 or 30 wk of age). We statistically examined whether there was a difference in mean values between necropsies done at 24 versus 30 wk. The adiposity indexes measured at 24 and 30 wk of age were not significantly different (P = 0.097). Therefore, in our genome scan analysis, we combined adiposity data for all cohorts to increase the power to detect linkage. To adjust for this, we included an additive effect for cohort in our analyses of adiposity.
The multiple-imputation method of Sen and Churchill (26) was used to perform one-dimensional and two-dimensional genome scans for each of the derived phenotypes and adiposity. A one-dimensional genome scan, also commonly referred to as a genome scan, is a survey of all single-locus models explaining the phenotype. This amounts to performing an ANOVA at every locus when complete genotype information is available. We examine how much phenotypic variation can be explained by genetic variation at every locus and summarize the statistical evidence that a locus affects the phenotype by the LOD score. A two-dimensional genome scan is a survey of all two-locus models explaining the phenotype. Here, we examined how much phenotypic variation can be explained by genetic variation at every pair of loci in the genome. For both one-dimensional and two-dimensional scans, we included an additive cohort effect for the three cohorts of mice. This method of analysis is able to combine the overall evidence in favor of a locus in all three cohorts while respecting their differences. The rationale for the combining of crosses has been previously reported and used in other studies (1, 8, 17, 31). Genotypes from chromosome 8 (linked to the fat mutation) and distal chromosome 17 (cohort I and II animals) were all HRS genotype and hence excluded from the genome analysis. After the genome scans, a two-stage model selection procedure was followed.
Locus and locus pair selection.
Locus and locus pair selection have been previously described (29). All loci that cleared the 5% permutation test threshold for the one-dimensional scans were selected. We selected locus pairs as follows. All pairs that cleared the 5% permutation test threshold in the two-dimensional scan for the full two-locus model were subjected to secondary tests. To determine whether both loci in a pair were acting additively, we compared the two-locus additive model to both single locus models dropping each locus in turn. For the comparison, we used a nominal P value of 0.002. This corresponds approximately to the 5% permutation test threshold for the one-dimensional genome scan. We then tested the interaction component of the two-locus model using a nominal P value of 0.0001. This corresponds to a 5% permutation test threshold for the interaction test only.
Backward elimination.
Backward elimination has been previously described (5). All loci and locus pairs selected in the previous stage were put together in a multiple-QTL model. The additive cohort effect was always included in the multiple-QTL model. The multiple-QTL model was formed by selecting the marker closest to the locus. If there were genotypes missing at the markers, the multiple-imputation method (26) was used to account for the missing genotypes. The basic idea is that the missing genotypes can be probabilistically inferred from the genotypes of nearby markers. This is done by simulating the missing genotype at the marker conditional on the flanking typed markers. To account for the uncertainty in our knowledge of the true genotypes, a number of imputations (26) were used. Each term in the model (including the cohort effect) was dropped, and the change in the LOD score was noted. (This corresponds to a type III ANOVA.) If the nominal P value of the change in the LOD score was >0.002 for the main effects and >0.0001 for the interactions, that term was dropped. This procedure was continued until no term could be dropped.
To test whether the major QTLs were specific to the fat/fat animals, we performed an ANOVA test on the fat/fat and non-fat/fat male progeny. We selected the nearest typed marker in both fat/fat and non-fat/fat progeny. These were D19Mit66 (26.0 cM) for initial plasma glucose [fat-induced diabetes 1 (find1)], D5Mit95 (68.0 cM) for average plasma glucose (find2), D11Mit35 (47.6 cM) for adiposity index [fat-induced adiposity 1 (fina1)], and D14Mit5 (22.5 cM) for final body weight [body weight 1 (bwt1)]. Because plasma glucose was not measured at 20 wk in non-fat/fat mice, we redefined the average glucose measurement for all animals as the average of the plasma glucose levels at 4, 8, 12, 16, and 24 wk (for this test only). The phenotype definitions of the three other phenotypes were left unchanged. Individuals with missing phenotypes or genotypes were discarded. ANOVAs were then performed on the QTL marker genotype separately for fat/fat and non-fat/fat animals.
Mutation Analysis
Total RNA was prepared from the whole brain of a C57BLKS/J and a HRS/J mouse. Tissues were homogenized, and RNA was isolated by treatment with TRIzol (Life Technologies) according to the manufacturer's protocol. cDNA was generated using the Superscript One-Step RT-PCR kit (Life Technologies; Gaithersburg, MD). PCR primers for mutation analysis of orexin/hypocretin (Hcrt) and carboxypeptidase D genes (Cpd) were designed from sequences in the public database (National Center for Biotechnology Information). PCR amplification of cDNA was performed using the Expand Template system (Roche). Amplified products were separated on a 11.2% gel, excised and purified using Nucleospin columns (Clontech; Palo Alto, CA), and sequenced (ABI Prism 3700) using both forward and reverse primers.
 |
RESULTS
|
---|
Segregation of Traits in F2 Male Progeny
To investigate the effects of the Cpefat mutation on each of the traits, phenotypic data from 76 fat/fat, 48 fat/+, and 61 +/+ intercross progeny from cohort I were collected, and the means were compared between the different genotypic groups. Table 2 summarizes the mean values of metabolic features examined in F2 progeny from cohort I. Increased adiposity, hyperglycemia, hyperinsulinemia, and hypercholesteremia were all observed in fat/fat homozygotes compared with fat/+ heterozygous and wild-type mice. The significant difference in the means of the physiological parameters observed between F2 fat/fat, fat/+, and +/+ progeny and the large variation in the F2 fat/fat progeny suggest that background genes are playing an important role in modifying the trait phenotypes. In addition, the fact that HRS-fat/fat mice do not exhibit elevated plasma glucose (7) further supports the idea that Cpefat is necessary but not sufficient for the development of diabesity in mouse models carrying a Cpe defect.
View this table:
[in this window]
[in a new window]
|
Table 2. Phenotypic profiles of parental BKS-fat/fat (N10 and N6), HRS-fat/fat, and F2 (fat/fat, fat/+, and +/+) mice from cohort I
|
|
Body weight and adiposity.
In general, fat/fat mice had significantly higher body weights than the fat/+ and +/+ mice. A heterozygous effect for body weight in fat/+ mice was not evident over the time period examined. As expected, adiposity indexes in fat/fat mutant mice were much higher than in nonobese (+/+) controls. Whereas fat/+ mice tended toward higher adiposity levels, they were not significantly different from their normal (+/+) littermates. Interestingly, significant differences in body length and girth were also observed between the three subgroups.
Hyperglycemia.
All F2 fat/+ and +/+ mice were normoglycemic. Of the total 282 F2 fat/fat males (cohorts IIII) monitored for plasma glucose levels, 253 had abnormal levels (plasma glucose >225 mg/dl). Fifty-eight percent were defined as diabetic or chronically diabetic (plasma glucose >300 mg/dl at 2 or more timepoints) before termination. Plasma glucose levels ranged between 61 and 1,020 mg/dl, with a mean of 312.0 ± 169.9 mg/dl. Different forms of diabetes onset were observed in the F2 generation: some mice were transiently diabetic early on, whereas others became diabetic later in life.
Hyperlipidemia and hyperinsulinemia.
F2 fat/fat mice developed chronic hyperinsulinemia (plasma insulin
5 ng/ml) and mild to moderate hyperlipidemia. At 30 wk of age, plasma insulin levels were elevated >30-fold, and total cholesterol levels nearly doubled in F2 fat/fat mice compared with their nonobese littermates.
QTL Mapping of Loci Interacting with Cpefat by Genome Scan Analysis
For QTL analysis, all data were logarithmically transformed to stabilize variance. The resulting phenotypic distributions were much more symmetric and are shown in Fig. 2. The frequency distribution of body weight varies with age, whereas the frequency distribution of plasma glucose remains homogenous between 8 and 24 wk of age. When the disease progressions were examined in the animals individually, different patterns were observed for body weight and plasma glucose. To make these observations biologically meaningful, we derived the data of each phenotype into different categories to detect modifiers affecting rate, flux, and severity of disease progression (see MATERIALS AND METHODS).

View larger version (25K):
[in this window]
[in a new window]
|
Fig. 2. Frequency distributions of BWs (A) and PG levels (B) in F2 fat/fat male mice at biweekly time intervals. The length of the horizontal lines corresponds to the number of observations.
|
|
A total of 87 SSLP markers was used to genotype a total of 282 fat/fat male progeny. The results of the one- and two-dimensional scans are shown in Fig. 3. In Table 3, a summary of the major QTLs and their positions is presented, whereas in Table 4 the type III ANOVA table corresponding to the multiple-QTL model selected for each trait is shown.

View larger version (32K):
[in this window]
[in a new window]
|
Fig. 3. A: genome scan plots (one dimensional) of traits. Chromosomes containing significant markers are shown in bold. B: allelic plots of significant quanitative trait loci (QTL) for adiposity, BW, PG, plasma insulin, total cholesterol, and derived phenotypes. C: allelic plot of a significant interactive QTL for BW from the two-dimensional scan.
|
|
View this table:
[in this window]
[in a new window]
|
Table 3. Summary of major QTL for adiposity, body weight, PG, plasma insulin, and blood cholesterol detected in F2 (BKS x HRS)-fat/fat mice
|
|
Adiposity.
A major QTL for Cpefat-induced adiposity (fina1) was identified on chromosome 11 in F2 fat/fat male progeny with a peak LOD score at D11Mit242. QTL analysis of the individual fat pad weights detected QTLs for rostral, caudal, and epididymal fat pads in the same region on chromosome 11. Mice homozygous for the BKS-derived allele at D11Mit242 had significantly higher fat depot weights (Fig. 3B) than their littermates.
Body weight.
In all F2 male progeny, a major locus for body weight (bwt1) was identified on chromosome 14, tightly linked to hr. To determine the effects of hr on various traits in F2 intercross progeny, fat/fat, fat/+, and +/+ mice were genotyped with allele-specific primers for the hr mutation, and mean values for all traits were compared (Table 5). In all three genetic subgroups, mice homozygous for the hr mutation had significantly reduced body weight measurements, which suggests that bwt1 is acting independently of fat on body weight. In addition, fat/fat, fat/+, and +/+ hr mice all had lowered total cholesterol and plasma insulin levels (not shown).
We identified a locus on chromosome 18 that interacts with bwt1 to lower body weight by intra-allelic interaction or heterosis. That is, heterozygosity on chromosome 18 had the strongest epistatic effect on body weight in mice homozygous for the HRS-derived allele of bwt1 (Fig. 3C). The chromosome 18 locus only showed a significant effect on body weights when interacting with bwt1 and had no independent effect on body weight measurements. A major locus for weight gain (maximum LOD score of 5.0 at D11Mit23) that overlapped with the adiposity locus fina1 was also detected. Mice homozygous for BKS-derived alleles (BB) had the largest weight gains compared with their heterozygous (BH) and homozygous (HH) HRS littermates (Fig. 3B).
Hyperglycemia, hyperinsulinemia, and hypercholesterolemia.
Two loci for Cpefat-induced diabetes (find1 and find2) were detected on chromosomes 19 and 5, respectively (Fig. 3, A and B). For find2, mice homozygous for BKS-derived alleles (BB) exhibited higher plasma glucose levels than their BH and homozygous HH littermates, whereas for find1, the presence of a HRS allele resulted in elevated plasma glucose levels. The find2 locus was found to affect other metabolic features besides plasma glucose such as insulin and total cholesterol levels. With the use of the backward elimination, multiple-QTL model, no locus, including find2, was identified as having an interaction with find1. The fact that no significant interaction was detected with find1 was surprising because HRS-fat/fat mice do not become diabetic. Because hr is segregated in the crosses, we hypothesized that the find1 effect on plasma glucose may be dependent on the hr genotype on chromosome 14. We compared the plasma glucose means of mice homozygous for the HRS-derived allele at find1 with two genotypic cohorts, BB and HH, at D14Mit115, a marker tightly linked to the hr locus (Table 6). As suspected, mice homozygous for the HRS-derived allele at find1 and homozygous for the BKS-derived allele at D14Mit115 had significantly higher plasma glucose levels than mice homozygous for the HRS-derived alleles at both find1 and D14Mit115.
Significant loci for both plasma insulin (24 wk) and total cholesterol (16 wk) were detected on chromosome 5 (find2) (Fig. 3, AC). F2 mice homozygous for the BKS-derived allele at D5Mit95 had significantly higher total cholesterol levels but lower plasma insulin levels.
Evaluation of Major QTL on Fat/+ and +/+ Progeny
To determine whether the major QTL identified acted as modifiers of the Cpefat mutation, we performed ANOVAs on the QTL marker genotype separately for fat/fat and non-fat/fat male intercross progeny (Table 7). As in the fat/fat mice, a significant effect was also observed with bwt1 in fat/+ and +/+ mice with D14Mit5, a marker linked to the hr locus. However, none of the other QTLs (find1, find2, and fina1) showed a strong effect in non-fat/fat male progeny, which is consistent with the modifier hypothesis.
Evaluation of Candidate Genes
The region spanning fina1 includes attractive biological candidates such as Hcrt and Cpd. Orexin-positive neurons have been localized in the lateral hypothalamus and project throughout the brain. Orexin knockout mice exhibit a narcoleptic phenotype, which suggests that orexin may function in the regulation of the arousal and sleep-wake cycle (6). However, intraventricular injection of orexin peptide increases feeding in C57BL/6J mice (18). Carboxypeptidase D has a similar enzymatic activity as CPE and might functionally complement a CPE deficit in some cell types. Mutation analysis in the coding regions of Hcrt and Cpd mRNA in HRS and BKS mice detected no sequence differences. Although the promotor regions were not examined, mRNA abundance did not differ as judged by the amount of ethidium bromide-stained RT-PCR product obtained from the two strains.
 |
DISCUSSION
|
---|
Obesity is a complex, multifactorial disease in the human population often associated with increased susceptibility to comorbidities such as Type II diabetes, atherosclerosis, and hypertension (30). The susceptibility of BKS.HRS-fat/fat mice and resistance of HRS-fat/fat mice to Type II diabetes make the mutant fat mouse a useful model to study gene interactions in the development of diabesity. In this study, two major loci for hyperglycemia (find2 and find1) were identified on chromosomes 5 and 19, respectively. F2 fat/fat mice that are homozygous for the BKS-derived allele of find2 have an increased susceptibility and mice that are homozygous for the HRS-derived allele are resistant to hyperglycemia. Hence, we hypothesize that find2 may be the locus that protects HRS-fat/fat mice from developing diabetes. It has been known previously that the BKS genetic background is diabetogenic (7). In a C57BLKS/J x C57BL/6J-Leprdb/Leprdb intercross, two putative loci for hyperglycemia were identified on chromosomes 8 and 17, but no QTL was detected on chromosome 5 (21). This disparity may be due to differences in how the db and fat mutations are interacting with the BKS background. Alternatively, the chromosome 8 locus may lie close to the Cpe locus, and the BKS.HRS mice used here may not carry it. In addition, only a subset of mice was analyzed in the BKS x B6 intercross (20% of highest plasma glucose and lowest plasma glucose levels), and thus the find2 locus on chromosome 5 locus may have been missed.
(BKS x HRS) F2 fat/fat mice that are homozygous for the HRS allele at find1 have elevated plasma glucose levels. Because HRS-fat/fat mice do not become hyperglycemic, this implies that find1 is interacting with additional loci from the BKS background. Our multiple-QTL model did not detect a locus interacting with find1; however, we did observe an association between plasma glucose levels of mice homozygous (HH) at find1 and their genotype near the hr locus on chromosome 14. This suggests that the hr gene (or a closely linked gene) may have an effect on find1 expression. In addition, loci closely linked to the Cpe locus on chromosome 8 cannot be ruled out, because all mice analyzed were homozygous Cpefat. Diabetes susceptibility loci have been found in the same interval as find1 on distal mouse chromosome 19 in several rodent experiments (11, 12, 16, 27). Whether all of these loci are due to mutations in the same gene remains to be determined.
A major QTL for body weight (bwt1) was identified on chromosome 14. The LOD peak of bwt1 is in close vicinity to hr, making hr a plausible candidate. Allele typing of the hr allele in the F2 progeny showed that hr/hr mutant mice had significantly lower mean body weight values than haired (+/+) mice. Hairless HRS mice (HRS-hr/hr mice) also have lower mean adipose depot and body weights than their normally haired albino (HRS-+/+) littermates (20). It has previously been shown that hairless mice have a higher metabolic rate than normally haired albino mice (9). This was attributed to greater energy requirement for thermoregulation in mice lacking body hair. However, it was recently discovered that the hr gene product acts as a corepressor to the thyroid hormone receptor (24), and the hr mutation might, therefore, mimic some of the effects of thyroid hormone on energy metabolism. Altogether, our data suggest that hr is responsible for the lowered body weight in HRS-hr/hr mice and that the hr gene itself explains some of the variance observed in F2 fat/fat males and acts independently of the fat locus. There are currently no published reports on adiposity in human patients with congenital alopecia or artrichia carrying mutations in the Hr gene, so that it is unclear whether hr mutations have the same effect on metabolism in humans as in mice.
A major locus (fina1) on chromosome 11 was strongly associated with adiposity in the genome scan of the F2 fat/fat mice. The fact that fina1 was not observed in lean controls implies that the fina1 locus is epistatically interacting with fat, thereby, masking the effects of hr on adipose depot distribution in F2 fat/fat mutant mice. The interval for fina1 is still fairly large but includes intriguing candidates such as Hcrt and Cpd. Although no sequence variations were observed in the coding regions of both genes, the possibility of promoter defects needs further exploration.
In this study, we carried out a genetic analysis of the complex phenotype caused by the prohormone processing defect in fat mutant mice. We use a novel method of analyzing modifier genes involved in different patterns of disease progression such as rate, fluctuation, and severity. We also show that many of the phenotypes observed in fat mice can be ameliorated by the action of major modifier genes. Because obesity is often associated with alterations in neuropeptide levels, the identification of these diabesity-associated modifier QTL should also lead to a better understanding of the biological pathways and neuropeptides that may be involved in common forms of obesity.
 |
GRANTS
|
---|
This study was supported by National Institute of Diabetes and Digestive and Kidney Diseases Grants DK-46977 and DK-53278. Institutional shared services were supported by National Cancer Institute Cancer Center Support Grant CA-34196.
 |
ACKNOWLEDGMENTS
|
---|
We gratefully acknowledge the technical assistance of Theolyn Gilley and Jay Young.
 |
FOOTNOTES
|
---|
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).
Address for reprint requests and other correspondence: J. K. Naggert, The Jackson Laboratory, 600 Main St., Bar Harbor, ME 04609 (E-mail: jkn{at}jax.org).
10.1152/physiolgenomics.00208.2003.
* G. B. Collin and T. P. Maddatu contributed equally to this work. 
 |
REFERENCES
|
---|
- Almind K, Kulkarni RN, Lannon SM, and Kahn CR. Identification of interactive loci linked to insulin and leptin in mice with genetic resistance. Diabetes 52: 15351543, 2003.[Abstract/Free Full Text]
- Bailey D. How pure are inbred strains of mice? Immunol Today 3: 210214, 1982.[CrossRef][ISI]
- Berman Y, Mzhavia N, Polonskaia A, and Devi LA. Impaired prohormone convertases in Cpe(fat)/Cpe(fat) mice. J Biol Chem 276: 14661473, 2001.[Abstract/Free Full Text]
- Brockmann GA and Bevova MR. Using mouse models to dissect the genetics of obesity. Trends Genet 18: 367376, 2002.[CrossRef][ISI][Medline]
- Broman KW and Speed TP. A model selection approach for the identification of quantitative trait loci in experimental crosses. J Royal Stat Soc B 64: 641656 and 731775, 2002.[CrossRef][ISI]
- Chemelli RM, Willie JT, Sinton CM, Elmquist JK, Scamell T, Lee C, Richardson JA, Williams SC, and Xiong Y. Narcolepsy in orexin knockout mice: molecular genetics of sleep regulation. Cell 98: 437451, 1999.[CrossRef][ISI][Medline]
- Coleman DL and Eicher EM. Fat (fat) and tubby (tub): two autosomal recessive mutations causing obesity syndromes in the mouse. J Hered 81: 424427, 1990.[ISI][Medline]
- DiPetrillo K, Tsaih SW, Sheehan S, Johns C, Kelmenson P, Gavras H, Churchill GA, Paigen B. Genetic analysis of blood pressure in C3H/HeJ and SWR/J mice. Physiol Genomics 17: 215220, 2004.[Abstract/Free Full Text]
- Donnelly HT. Oxygen consumption, activity and body fat in normal and hairless mice. Lab Anim 16: 167171, 1982.[ISI][Medline]
- Fricker LD, Berman YL, Leiter EL, and Devi LA. Carboxypeptidase E activity is deficient in mice with the fat mutation. Effect on peptide processing. J Biol Chem 271: 3061930624, 1996.[Abstract/Free Full Text]
- Galli J, Li LS, Glasser A, Ostenson CG, Jiao H, Fakhrai-rad H, Jacob HJ, Lander ES, and Lutherman H. Genetic analysis of non-insulin dependent diabetes mellitus in the GK rat. Nat Genet 12: 3137, 1996.[CrossRef][ISI][Medline]
- Gauguier D, Froguel P, Oarent V, Bernard C, Bihoreau MT, Portha B, James MR, Penicaud L, Lathrop M, and Ktorza A. Chromosomal mapping of genetic loci associated with non-insulin dependent diabetes in the GK rat. Nat Genet 12: 3843, 1996.[CrossRef][ISI][Medline]
- Gloyn AL and McCarthy MI. The genetics of type 2 diabetes. Best Pract Res Clin Endocrinol Metab 15: 293308, 2001.[CrossRef][ISI][Medline]
- Ikeda A, Zheng QY, Rosenstiel P, Maddatu T, Zuberi AR, Roopenian DC, North MA, Naggert JK, Johnson KR, and Nishina PM. Genetic modification of hearing in tubby mice: evidence for the existence of a major gene (moth1) which protects tubby mice from hearing loss. Hum Mol Genet 8: 17611767, 1999.[Abstract/Free Full Text]
- Izzo C, Grillo F, and Murador E. Improved method for determination of high-density lipoprotein cholesterol. I. Isolationof high-density lipoproteins by use of polyethylene glycol 6000. Clin Chem 27: 371374, 1981.[Abstract/Free Full Text]
- Kim JH, Sen S, Avery CS, Simpson E, Chandler P, Nishina PM, Churchill GA, and Naggert JK. Genetic analysis of a new mouse model for non-insulin-dependent diabetes. Genomics 74: 273286, 2001.[CrossRef][ISI][Medline]
- Li R, Lyons MA, Wittenburg H, Paigen B, and Churchill GA. Combining data from multiple inbred line crosses improves the power and resolution of quantitative trait loci mapping. Genetics 169: 16991709, 2005.[Abstract/Free Full Text]
- Lubkin M and Stricker-Krongrad A. Independent feeding and metabolic actions of orexins in mice. Biochem Biophys Res Commun 253: 241245, 1998.[CrossRef][ISI][Medline]
- Maddatu TP and Naggert JK. Allele-specific PCR assays for the tub and cpefat mutations. Mamm Genome 8: 857858, 1997.[CrossRef][ISI][Medline]
- Mount LE. Metabolic rate and thermal insulation in albino and hairless mice. J Physiol 217: 315326, 1971.[ISI][Medline]
- Mu JL, Naggert JK, Svenson KL, Collin GB, Kim JH, McFarland C, Nishina PM, Levine DM, Williams KJ, and 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 40: 13281335, 1999.[Abstract/Free Full Text]
- Naggert JK, Fricker LD, Varlamov O, Nishina PM, Rouille Y, Steiner DF, Carroll RJ, Paigen BJ, and Leiter EH. Hyperproinsulinaemia in obese fat/fat mice associated with a carboxypeptidase E mutation which reduces enzyme activity. Nat Genet 10: 135142, 1995.[ISI][Medline]
- Pomp D. Genetic dissection of obesity in polygenic animal models. Behav Genet 27: 285306, 1997.[CrossRef][ISI][Medline]
- Potter GB, Beaudoin IIIGMJ, DeRenzo CL, Zarach JM, Chen SH, and Thompson CC. The hairless gene mutated in congenital hair loss disorders encodes a novel nuclear receptor corepressor. Genes Dev 15: 26872701, 2001.[Abstract/Free Full Text]
- Rovere C, Viale A, Nahon J, and Kitabgi P. Impaired processing of brain proneurotensin and promelanin-concentrating hormone in obese fat/fat mice. Endocrinology 137: 29542958, 1996.[Abstract]
- Sen S and Churchill GA. A statistical framework for quantitative trait mapping. Genetics 159: 371387, 2001.[Abstract/Free Full Text]
- Stoehr JP, Nadler ST, Schueler KL, Rabaglia ME, Yandell BS, Metz SA, and Attie AD. Genetic obesity unmasks nonlinear interactions between murine type 2 diabetes susceptibility loci. Diabetes 49: 19461954, 2000.[Abstract]
- Strauss WM. Preparation of genomic DNA from mammalian tissue. In: Current Protocols in Molecular Biology. New York: Wiley, 1998, p. 2.2.12.2.3.
- Sugiyama F, Churchill GA, Higgins DC, Johns C, Makaritsis KP, Gavras H, and Paigen B. Concordance of murine quantitative trait loci for salt-induced hypertension with rat and human loci. Genomics 71: 7077, 2001.[CrossRef][ISI][Medline]
- Watkins PJ. Cardiovascular disease, hypertension, and lipids. BMJ 326: 874876, 2003.[Free Full Text]
- Zou F, Yandell BS, and Fine JP. Statistical issues in the analysis of quantitative traits in combined crosses. Genetics 158: 13391346, 2001.[Abstract/Free Full Text]