• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Mar 18, 2008; 105(11): 4253–4258.
Published online Mar 12, 2008. doi:  10.1073/pnas.0706562105
PMCID: PMC2393747
Genetics

Genomic imprinting effects on adult body composition in mice

Abstract

Genomic imprinting results in the differential expression of genes, depending on which allele is inherited from the mother and which from the father. The effects of such differential gene expression are reflected in phenotypic differences between the reciprocal heterozygotes (Aa vs. aA). Although many imprinted genes have been identified and play a key role in development, little is known about the contribution of imprinting to quantitative variation in trait expression. Here, we examine this problem by mapping imprinting effects on adult body composition traits in the F3 generation of an intercross between the Large (LG/J) and Small (SM/J) inbred mouse strains. We identified eight pleiotropic imprinted quantitative trait loci (iQTL) located throughout the genome. Most iQTL are in novel locations that have not previously been associated with imprinting effects, but those on chromosomes 7, 12, and centromeric 18 lie in regions previously identified as containing imprinted genes. Our results show that the effects of genomic imprinting are relatively small, with reciprocal heterozygotes differing by ≈0.25 standard deviation units and the effects at each locus accounting for 1% to 4% of the phenotypic variance. We detected a variety of imprinting patterns, with paternal expression being the most common. These results indicate that genomic imprinting has small, but detectable, effects on the normal variation of complex traits in adults and is likely to be more common than usually thought.

Keywords: epigenetics, mouse, obesity, quantitative trait loci, organs

Genomic imprinting refers to the expression of only one of the two alleles carried by a diploid organism with expression depending on the parent of origin of the alleles (1, 2). At some imprinted loci, such as Igf2r in mice, only the maternal allele is expressed, whereas at other loci, such as Igf2, only the paternal allele is expressed (3, 4). This parent-of-origin-specific expression pattern is evinced by phenotypic differences between the ordered reciprocal heterozygotes, Aa and aA, where the first allele denotes the paternally derived copy and the second denotes the maternally derived allele. Although the two heterozygotes (Aa and aA) are genotypically equivalent, they can be phenotypically distinct because, under genomic imprinting, only one of the two parental alleles is expressed. A variety of molecular mechanisms are thought to be involved in imprinting effects, including DNA methylation, histone modification, noncoding RNAs (ncRNA), and even long-distance interchromosomal interactions (47).

When first discovered, genomic imprinting was thought to be a rare and unusual (8), but to date studies of murine translocations have identified nearly 100 imprinted genes on 11 chromosomes (www.har.mrc.ac.uk/research/genomic_imprinting; ref. 9), and the Imprinted Gene Catalog lists 144 imprinted genes across the whole genome (9). Most of these discoveries were based on the severe phenotypic effects of the overexpression of imprinted genes attributable to an unbalanced chromosome complement produced by translocation. Imprinted genes tend to be found in clusters of adjacent loci (10), but their gross phenotypic effects remain obscure. Morison et al. (9) note that there may be many imprinted genes with subtle effects that remain undiscovered. Researchers have performed genome-wide scans using biochemical techniques to identify differentially methylated regions based on parent of origin of the alleles (11). In addition, Luedi et al. (12) predicted ≈600 potentially imprinted genes by using sequence analysis to identify regions with sequence features similar to those found at known imprinted loci. The total number of imprinted genes and their range of phenotypic effects, however, remain uncertain.

Imprinting effects have been characterized as exhibiting maternal or paternal gene expression leading to the expectation that each of the heterozygotes will be similar to its paired homozygote (e.g., refs. 13 and 14). With paternal expression, the Aa heterozygote and AA homozygote would both express the A allele, and the aA heterozygote and aa homozygote would both express the a allele. However, parent-of-origin-specific expression can lead to more complex phenotypic patterns when multiple linked, imprinted genes with diverse expression patterns act in concert (R.H., J.M.C., C.R., and J.B.W., unpublished data). Hager et al. (R.H., J.M.C., C.R., and J.B.W., unpublished data) present a scheme for characterizing the potential diversity of imprinting patterns. In this scheme (summarized in Fig. 1), imprinting patterns are classified as either parental expression (paternal or maternal) or dominance (bipolar and polar) imprinting (see Fig. 1). Types of imprinting are classified according to the relative values of the imprinting, additive, and dominance genotypic values at the locus of interest (see Fig. 1) (15).

Fig. 1.
Classification of imprinting patterns (following R.H., J.M.C., C.R., and J.B.W., unpublished work). Classification as parental expression (maternal or paternal) or dominance imprinting (polar or bipolar) depends on the value and sign of the imprinting ...

Genomic imprinting studies have concentrated on effects early in mammalian life, predominantly prenatal or early postnatal effects (1618). However, it is quite possible that imprinting affects adult traits and disease susceptibility as well either through the persistent effects of early growth and development (1924) or through direct effects on adult physiology. Clearly, the occurrence of imprinting effects later in life remains an empirical problem, and, indeed, data on the occurrence of imprinting effects on such traits may ultimately suggest that an exclusive focus on early development is unwarranted for imprinting.

Here, we undertake a genome-wide mapping approach to identifying regions of the genome that have an imprinting effect on adult body composition using a three-generation intercross between inbred mouse strains. An advantage to this kind of screen is that the imprinted regions are identified by their effects on phenotypes, providing clues to their physiological roles and to their relative importance in natural phenotypic variation. Also, we are able to detect even relatively small effects so that the role of genomic imprinting on normal physiological and developmental processes and on complex diseases can be ascertained.

Results

We discovered nine quantitative trait loci (QTL) on seven different chromosomes that had significant parent-of-origin effects at the chromosome-wide level (see Table 1), with one (Bwi12.1) having an effect that also was significant at the genome-wide level. The full multivariate model including additive, dominance, and imprinting effects on all seven traits indicated that seven of the nine loci were significant at the chromosome-wide level, with four of these seven also being significant at the genome-wide level. Genotypic values for LL, LS, SL, and SS genotypes, with their standard errors, are presented in supporting information (SI) Table 2. The significant QTL on chromosome 6 (Bwm6.1) did not show significant differences between reciprocal heterozygotes among offspring of heterozygous mothers, as expected for genomic imprinting effects, but did provide strong evidence for a maternal genetic effect (25). Most individual traits mapped to multiple genomic locations at the chromosome-wide level, including liver, fat pad, heart, and kidney weight, when only one significant result is expected under the null model of no imprinting effect. There was only one chromosome-wide significant result each for spleen and body weight as expected by chance under the null model, but the result for body weight also is significant at the more stringent genome-wide level. Furthermore, several additional locations identified as significant at the chromosome-wide level for other traits were significant at the point-wise level for body weight and spleen weight. Tail length was the only trait that was never significant at the chromosome-wide level and therefore provides little or no evidence for imprinting effects in this population.

Table 1.
Identification of iQTL is defined as BwiX.Y where Bwi stands for imprinting effects on body composition, X identifies the chromosome, and Y identifies the iQTL on the chromosome (also included is one locus, Bwm6.1, where the parent-of-origin effect is ...

Imprinted QTL (iQTL) regions with a significant effect on one or more traits also had smaller effects on other traits that were substantial but failed to reach the stringent chromosome-wide significance threshold on their own. These traits are also included in Table 1 if they were significant at the point-wise level at a location affecting another trait at the chromosome-wide level. Most iQTL affected several traits. Liver weight and body weight mapped together at all six of the loci affecting either trait, whereas three of these six also affected reproductive fat pad and kidney weight. Kidney weight mapped to six iQTL, including one unique location (Bwi4.1), whereas spleen and heart weights mapped to three locations each. Generally, we find that iQTL tend to be pleiotropic, having effects on many different aspects of body composition.

The average difference between reciprocal heterozygotes (2i) is 0.25 standard deviation (SD) units, ranging from 0.16 to 0.42 SD, so these effects must be considered small. Phenotypic variance explained by these effects also is small, varying from 1% to 4% per locus. The strongest imprinting effects are at Bwi12.1 where the difference between reciprocal heterozygotes is −0.34 SD units (i = −0.17 SD) and the locus accounts for an average of 2.2% of the phenotypic variance in the affected traits. Collectively, the iQTL account for 1.4% to 4.6% of the phenotypic variance, depending on the trait. The proportion is lowest for tail length and heart and spleen weights, at intermediate levels for reproductive fat pad and kidney weights, and highest for liver and body weights.

The iQTL effects displayed a variety of imprinting patterns, including 12 effects with parental expression (3 maternal, 9 paternal), 11 with bipolar dominance imprinting (9 positive and 2 negative), four with polar overdominance imprinting (2 positive, 2 negative), and two with polar underdominance imprinting. Most traits showed multiple forms of imprinting across loci, but all three reproductive fat pad iQTL effects displayed paternal expression as did four of six body weight iQTL effects.

The pattern of imprinting varied among traits at the same iQTL. For example, Bwi3.2 displays all three kinds of imprinting: parental expression, bipolar dominance imprinting, and polar dominance imprinting. The strongest imprinting effects were for Bwi12.1, affecting liver, kidney, reproductive fat pad, and overall body weight (see Fig. 2). The difference between paternal expression for fat pad and body weight and polar overdominant imprinting for liver and kidney weights is the relative value of the SS genotype, which is higher for fat pad and body weight than for liver and kidney weights.

Fig. 2.
Imprinting patterns at Bwi12.1 for liver, heart, kidney, reproductive fat pad, and body weight in grams. Shown are the mean phenotypes for each of the four ordered genotypes along with standard error bars.

Genotypic values at the chromosome 6 QTL (Bwm6.1) that showed a significant parent-of-origin-dependent effect were not consistent with imprinting. Instead, its statistical significance seems to be attributable to a large additive maternal genetic effect that appears as imprinting only because of the correlation between additive genetic maternal effects and imprinting effects in this design (25).

Discussion

We have demonstrated that genomic imprinting affects normal variation in adult body composition phenotypes, indicating that later life features may be affected by imprinting and, therefore, a focus on imprinting as affecting early development is perhaps misplaced. Although only one imprinted locus reached a genome-wide level of significance in the test for parent-of-origin effects (Bwi12.1), a total of 14 iQTL effects were significant at the chromosome-wide level. In addition, the overall tests of significance for the nine loci (including the locus Bwm6.1, showing a maternal effect), where each locus was tested for its total effect (additive, dominance, and parent-of-origin-dependent) on the entire suite of traits, indicated that four of the nine QTL were significant at the genome level and an additional three loci at the chromosome-wide level. This overall multivariate significance test provides strong evidence supporting these QTL because it considers all seven traits jointly. The multivariate tests also suggest that the two iQTL (Bwi5.1 and Bwi18.1) that were not significant at the chromosome level should be interpreted with caution, although the fact that these loci each show an imprinting effect on multiple traits supports the hypothesis that they are likely to represent real iQTL. Chen and Storey (26) have shown that the use of per-chromosome significance thresholds balances the possibility of identifying false positive QTL with that of ignoring too many true positives in a genome scan. As with any QTL mapping study, the reliability of the imprinting locations reported here need to be evaluated by additional study allowing validation and finer mapping resolution.

The highest number of iQTL affecting any single trait was six (a number shared by liver, kidney, and body weight), which can be compared with the number (N) of significant additive (aQTL) and/or dominance (dQTL) QTL effects (genotypic values) found in the F2 population for these same body composition traits (27, 28). Overall, many more aQTL effects (n = 51) than iQTL effects (n = 29) were found for body composition. Overall, the frequency of significant dQTL effects is more similar to the frequency of iQTL effects, with a total of 25 significant dQTL effects compared with 29 iQTL effects reported here. Additive genotypic effects are clearly more common than imprinting effects in this population whereas imprinting and dominance effects are comparable in frequency. Like most aQTL and dQTL (28), the majority of iQTL affected several body composition traits, indicating the generally body-wide scope of imprinting effects.

The most prominent iQTL is Bwi12.1, with effects on the reproductive fat pad, liver, kidney, and total body weight, but even these imprinting effects must be characterized as relatively small given that the difference between reciprocal heterozygotes is distributed between 0.20 and 0.40 SD units. Dominance genotypic values for these same traits mapped in the F2 population also are typically within this range, but differences between homozygotes often are >0.40 SD units (27, 28). Thus, imprinting effects are similar in magnitude to dominance effects but small when compared with additive genotypic values.

Our findings of imprinting effects on adult body composition traits correspond to those found for several diseases caused by imprinting defects that persist well after birth, including Beckwith–Wiedemann Syndrome, Prader–Willi Syndrome, and Albright Hereditary Osteodystrophy (29, 30). Studies have mapped imprinting effects in various human populations for adult obesity and body composition traits (3032) and imprinting marks, such as methylation and histone configuration, often persist into adulthood (31). Imprinting may play a physiological role in metabolism and body composition throughout life thereby contributing both to normal variation and the etiology of complex diseases rather than being restricted to prenatal and neonatal time periods.

Patterns of Genomic Imprinting Effects.

We have identified several different patterns of imprinting effects (Table 1), including both paternal and maternal expression, bipolar dominance imprinting, and polar dominance imprinting (see Fig. 1). The differences among these categories can be subtle, as illustrated when different traits show different imprinting patterns at the same locus (see Fig. 2). The molecular origins of these patterns can be complex (5) because they involve traits remotely related to specific gene expression phenotypes. Silencing of one parental allele may lead to simple monoallelic expression (e.g., ref. 2). However, even in the simple case of monoallelic expression, patterns of effects on higher-order phenotypes, such as body weight and obesity, may not correspond to mRNA levels in a simple linear fashion. Other imprinting patterns, such as polar and bipolar dominance imprinting, may be the result of more complex molecular mechanisms involving several different linked imprinted loci (4) (R.H., J.M.C., C.R., and J.B., unpublished work), as has been hypothesized for the callipyge mutant (33) and for bipolar imprinting (R.H., J.M.C., C.R., and J.B., unpublished work).

The diversity of imprinting patterns found for different traits at a single locus is reminiscent of the patterns of differential dominance, where a pleiotropic QTL has different dominance patterns for different traits (34). Although rarely discussed (35), it is common for traits at the same locus to variously display allele A dominant to allele a, a dominant to A, codominance, or even overdominance for different traits at the same locus. This finding is interpreted as being attributable to different genotype-to-phenotype maps for different traits affected by the same locus (34). We see a similar phenomenon here for imprinting effects, implying that the mapping of imprinting effects from gene expression to phenotype varies from trait to trait. Although differential dominance can have important effects on evolution and the maintenance of genetic diversity under selection, the consequences of differential imprinting remain unexplored.

Hager et al. (25) point out that imprinting is not unique in producing apparent parent-of-origin effects. Maternal genetic effects, genetic variation in the effects of the environment provided by the mother on her offspring's development, also can produce a parent-of-origin pattern of offspring genotypic values including differences between reciprocal heterozygotes. Here, we find that a location on chromosome 6 (Bwm6.1) produced an apparent imprinting effect in that the difference between reciprocal heterozygotes was highly significant but, on further study, was shown to be caused by a maternal genetic effect, not an imprinting effect. The conflation of maternal genetic and imprinting effects also may affect other imprinting mapping studies, raising an important issue for such studies.

Location of Detected iQTL.

Most of the iQTL mapped here are in locations (e.g., chromosomes 3, 4, and 5) where there are either no imprinted loci in some databases (www.har.mrc.ac.uk/research/genomic_imprinting) or a paucity of loci (9). Although chromosome 7 has several imprinted regions, our iQTL, Bwi7.1, mapped to chromosomal bands F1–3, whereas the documented imprinting regions are in the B band range. The Prader–Willi Syndrome maps to the F5 band on chromosome 7, which also includes Igf2. One imprinted gene has been noted in the distal portion of the Bwi7.1 region, Inositol polyphosphate-5-phosphatase F version 2 (Ipp5f_v2; ref. 36). Ipp5f_v2 is paternally expressed in the brain but does not have any known higher-order phenotypic effects. Bwi7.1 also shows paternal expression for several of its affected traits. It remains uncertain whether there are more imprinted genes nearby. Dong et al. (30) also found evidence for an imprinting effect on obesity mapping to human chromosome 11q13, a region of conserved synteny to our Bwi7.1 iQTL.

Georgiades et al. (40) found strong evidence for imprinting on chromosome 12, especially at the telomeric end of the chromosome where the genes involved in the callipyge mutation occur (33, 38, 39). However, Bwi12.1 maps proximal to the known imprinted region so it seems to represent a unique imprinting region, although imprinting has been implicated in the effects of a mutation (Adp) caused by a transgene insertion into the region of Bwi12.1 (40, 41). Finally, there also is some evidence for imprinting in the proximal region of chromosome 18, corresponding to the Bwi18.1 region. The Impact gene is a paternally expressed gene discovered by allelic message display that is expressed in the brain but has no known function or phenotypic consequences (42).

Luedi et al. (12) used bioinformatic tools to identify 600 putative imprinted genes in the mouse genome. SI Table 3 lists the genes that map to the confidence regions of the iQTL mapped here. Both our confidence regions and Luedi et al.'s list of predicted imprinted genes are long, so some correspondence may be expected by chance. Even so, some confidence regions contained many such imprinting candidates, including 16 within Bwi7.1, 13 in Bwi18.1, and 11 in Bwi3.2. These iQTL appear to be especially rich in imprinting candidates because we would expect only about six of Luedi et al.'s (12) candidates to map to one of our iQTL regions by chance. Because the evidence for imprinting used by Luedi et al. is experimentally unrelated to the evidence we present, the enriched regions should be productive in further research to fine-map the iQTL detected here and apply molecular tests for gene expression and epigenetic modification of the positional candidate genes.

We have mapped a series of imprinted gene effects on normal variation in adult body size and composition traits in mice. Although these effects are individually small and not as common as additive genotypic effects, they are about as frequent as dominance effects in our intercross population. Whether these results are typical for other populations remains to be determined. The presence of imprinting genotypic effects on normal variation raises unresolved issues concerning their own evolution and their consequences for phenotypic evolution.

Materials and Methods

The experimental population includes animals from the F2 intercross of Large (LG/J) and Small (SM/J) inbred mouse strains and their F3 offspring (43, 44). Ten LG/J females were crossed with 10 SM/J males to produce 54 F1 hybrids. These F1 hybrids were intercrossed to produce 510 F2 animals. The F2 males and females were then randomly mated to produce 200 full-sibling F3 families with a total of 1,632 F3 animals. One hundred fifty-eight of the 200 F3 families participated in a cross-fostering protocol in which half of the pups from a pair of litters born on the same day were reciprocally exchanged between mothers (43, 45). Thus, for approximately half of the population of pups, the normal correlation between postnatal maternal and offspring genotypes was negated.

The two parental strains, LG/J and SM/J, are inbred lines derived from separate breeding experiments for large (LG/J; ref. 46) and small (SM/J; ref. 47) 60-day body weight from disparate source populations. These strains were maintained at The Jackson Laboratories and were the source for several experiments on the genetics of body size (48). In our laboratory, adult LG/J animals are 16 to 20 g heavier than SM/J animals. The strains show genetic variation for body weight, growth, skeletal morphology, obesity, glucose and insulin levels, and response to a high-fat diet (4951) among other features (52). QTL mapping studies of phenotypes used in this study have found that genetic variation is attributable to many loci (6, 7, 1114, 33, 38, 39, 53) (R.H., J.M.C., C.R., and J.B.W., unpublished work) of relatively small effect (a < 0.3 SD) with variable dominance, epistatic interactions, and modular pleiotropic gene effects (54). Imprinting effects on these traits are unknown.

Animals were killed after 70 days of age or after having produced and reared their offspring to weaning, at 3 weeks of age. At necropsy, animals were weighed to obtain an overall measure of body size, the length of the tail was measured with calipers, and then they were dissected and their organs (reproductive fat depot, heart, kidneys, spleen, and liver) were weighed to the nearest 0.01 g with a digital scale. Before gene mapping, the effects of age at necropsy, sex, and litter size at birth were removed from the data, and the residuals were used in the analysis (43).

Fresh DNA was extracted from livers in 2006 and used to score 353 polymorphic SNP markers across the autosomes using the Illumina Golden-Gate assay. These SNP markers (see SI Table 4) were chosen from the 4,200 polymorphic markers scored as part of the CTC/Oxford genotyping consortium (www.well.ox.ac.uk/mouse/INBREDS/; ref. 52), which resulted in markers being placed every 4–5 cM along the genome except in regions that seem to be nonpolymorphic between LG/J and SM/J (52). Genotypes were obtained for all F2 animals and their F3 offspring.

The F2 and F3 genotypes were used to reconstruct haplotypes using the “block-extension algorithm” in the PedPhase program (55) that solves the “minimum-recombination haplotypes configuration problem” (56). With this haplotype information, it was possible to distinguish all four genotypes in the F3 population at each marker locus (LL, LS, SL, and SS) with the paternal allele listed first and the maternal allele second.

We performed mapping analyses by using marker regression (57) implemented by canonical correlation using the SAS CANCORR procedure (SAS Institute) with the 1,632 F3 animals (58) (R.H., J.M.C., C.R., and J.B.W., unpublished work). The four ordered genotypes at the marker loci (LL, LS, SL, and SS) were assigned additive (Xa; LL = +1, LS, SL = 0, SS = −1), dominance (Xd; LS, SL = 1, LL, SS = 0), and imprinting (parent of origin) (Xi; LS = +1, LL, SS = 0, SL = −1) genotypic index values following Mantey et al. (59). The set of phenotypes are regressed onto the genotypic index scores at each marker locus, and the resulting regression coefficients are the additive (a), dominance (d), and imprinting (i) genotypic values. Because we primarily focused on the imprinting effect, we obtained the probability of this effect controlling for potential additive and dominance effects at the locus and used the significance of these parent-of-origin-dependent effects to identify iQTL. To help confirm that locations identified this way are real QTL, we also tested each locus for its overall effect (a, d, and i) on the full set of traits using canonical correlation. This test can be viewed as an overall significance test for a particular locus.

Probability values (P) generated by the CANCORR procedure were transformed to a logarithmic probability ratio (LPR) to make them comparable to the LOD scores typically seen in QTL analyses [where LPR = −log10(P)]. As defined in standard quantitative genetic theory, the additive genotypic value (a) is equal to half the difference between the homozygotes, and the dominance genotypic value (d) is the deviation of the heterozygote mean from the midpoint of the homozygotes (15). The imprinting genotypic value (i) is half the difference between the reciprocal heterozygotes (59).

This measure of imprinting effects (i) has a 0.5 correlation with genetic maternal effects (am). To examine whether the mapped effect is caused by an imprinting or maternal effect, the four genotypic values for each trait at each locus were examined relative to the three maternal genotypes and a test performed for maternal genetic effects (i.e., to test whether am was zero) as half the difference between pups born of homozygous mothers. If the offspring of homozygous mothers differs significantly from each other and the observed “imprinting” effect is primarily attributable to differences between the means of heterozygous offspring of homozygous mothers, it is quite likely that the parent-of-origin effect appears as a consequence of maternal genetic effects rather than imprinting (25).

Significance testing was approached through simulation of the family structure of the observed F2 and F3 generations to control the significance threshold for both multiple comparisons and the genetic autocorrelation among the F3 siblings. This was done by simulating the pedigree from the F1 generation with the levels of recombination observed in the F2 generation to produce genotypes for F3 families with the same family size distribution as the actual population. Phenotypes then were associated with genotypes of their own simulated family so that the familial structure of the phenotypes and their heritabilities and genetic correlations remain unaltered. The simulated genotypes also have the same family structure as the original population but the simulated genotypes are wholly unassociated with the observed phenotypes. We obtained the distribution of LPR scores at each marker, for each chromosome, and across the whole genome under the null model of no association between specific genotypes and phenotypes using 1,000 simulations. Significance thresholds for the whole genome and for each chromosome were taken as the 95th percentile LPR values from the set of 1,000 simulated genome- and chromosome-level tests. A per-locus 5% significance threshold also was generated to allow for tests of all other traits at each identified locus.

Supplementary Material

Supporting Information:

Acknowledgments.

We thank Elizabeth Norgard, Joseph Jarvis, and Mihaela Pavlicev for helpful comments on an earlier draft of this manuscript. This work was supported by funding from the Biotechnology and Biological Sciences Research Council (BBSRC), an Underwood Fellowship from the BBSRC, and National Institutes of Health Grant DK055736.

Footnotes

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0706562105/DC1.

Li J, Jiang T, Proceedings of the Seventh Annual International Conference on Research in Computational Molecular Biology, April 18–21, 2002, Washington, DC.

References

1. Bartolomei MS, Tilghman SM. Annu Rev Genet. 1997;31:493–525. [PubMed]
2. Reik W, Walter J. Nat Rev Genet. 2001;2:21–32. [PubMed]
3. DeChiara TM, Robertson EJ, Efstratiadis A. Cell. 1991;64:849–859. [PubMed]
4. Wood AJ, Oakey RJ. PLoS Genet. 2006;2:1677–1685.
5. Allis CD, Jenuwein T, Reinberg D. Epigenetics. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press; 2007.
6. Kiefer JC. Devel Dynam. 2007;236:1144–1156. [PubMed]
7. Pauler FM, Barlow DP. Genes Dev. 2006;20:1203–1206. [PubMed]
8. Haig D. Proc R Soc London B. 1997;264:1657–1661. [PMC free article] [PubMed]
9. Morison IM, Ramsay JP, Spencer HG. Trends Genet. 2005;21:457–465. [PubMed]
10. Lewis A, Reik W. Cytogenet Genome Res. 2006;113:81–89. [PubMed]
11. Smith RJ, Dean W, Konfortova G, Kelsey G. Genome Res. 2003;13:558–569. [PMC free article] [PubMed]
12. Luedi PP, Hartemink AJ, Jirtle RL. Genome Res. 2005;15:875–884. [PMC free article] [PubMed]
13. Spencer HG. Genetics. 2002;161:411–417. [PMC free article] [PubMed]
14. Wolf JB, Hager R. PLoS Biol. 2006;4:2238–2243.
15. Falconer DS, Mackay TFC. Introduction to Quantitative Genetics. New York: Longman; 1996.
16. Moore T, Reik W. Rev Reprod. 1996;1:73–77. [PubMed]
17. Itier JM, Tremp GI, Leonard JF, Multon MC, Ret G, Schweighoffer F, Tocque B, Bluet-Pajot MT, Cormier V, Dautry F. Nature. 1998;393:125–126. [PubMed]
18. Li LL, Keverne EB, Aparicio SA, Ishino F, Barton SC, Surani MA. Science. 1999;284:330–333. [PubMed]
19. Kajantie E. Ann NY Acad Sci. 2006;1083:11–27. [PubMed]
20. Barker DJP. Obesity Rev. 2007;8(Suppl 1):45–49. [PubMed]
21. Hager R, Johnstone RA. Biol Lett. 2006;2:253–256. [PMC free article] [PubMed]
22. Waterland RA, Jirtle RL. Nutrition. 2004;20:63–68. [PubMed]
23. Meaney MJ, Szyf M. Trends Neurosci. 2005;28:456–463. [PubMed]
24. Dolinoy DC, Das R, Weidman JR, Jirtle RL. Pediat Res. 2008;61(Suppl):1–8.
25. Hager R, Cheverud JM, Wolf JB. Genetics. 2008;178:1–8.
26. Chen L, Storey LD. Genetics. 2006;173:2371–2381. [PMC free article] [PubMed]
27. Cheverud JM, Vaughn TT, Pletscher LS, Peripato AC, Adams ES, Erikson CF, King-Ellison KJ. Mamm Genome. 2001;12:3–12. [PubMed]
28. Kenney-Hunt JP, Vaughn TT, Pletscher LS, Peripato AC, Routman EJ, Cothran K, Durand D, Norgard E, Perel C, Cheverud JM. Mamm Genome. 2006;17:526–537. [PubMed]
29. Constancia M, Kelsey G, Reik W. Nature. 2004;432:53–57. [PubMed]
30. Dong C, Li WD, Geller F, Lei L, Li D, Gorlova OY, Hebebrand J, Amos CI, Nicholls RD, Price RA. Amer J Hum Genet. 2005;76:427–437. [PMC free article] [PubMed]
31. Gorlova OY, Amos CI, Wang NW, Shete S, Turner ST, Boerwinkle E. Eur J Hum Genet. 2003;11:425–432. [PubMed]
32. Lindsay RS, Kobes S, Knowler WC, Bennett PH, Hanson RL. Diabetes. 2001;50:2850–2857. [PubMed]
33. Georges M, Charlier C, Cockett NE. Trends Genet. 2003;19:248–252. [PubMed]
34. Ehrich T, Vaughn TT, Koreishi SF, Linsey RB, Pletscher LS, Cheverud JM. J Exp Zool (Mol Devel Evol) 2003;296B:58–79. [PubMed]
35. Van Dooren TJM. Evolution (Lawrence, Kans) 2006;60:1991–2003.
36. Choi JD, Underkoffler LA, Collins JN, Williams PT, Golden JA, Shuster EF, Jr, Loomes KM, Oakey RJ. Mol Cell Biol. 2005;25:5514–5522. [PMC free article] [PubMed]
37. Georgiades P, Watkins M, Surani MA, Ferguson-Smith AC. Development. 2000;127:4719–4728. [PubMed]
38. Lewis A, Redrup L. Curr Biol. 2005;15:R291–R294. [PubMed]
39. Takeda H, Caiment F, Smit M, Hiard S, Tordoi X, Cockett NE, Georges M, Charlier C. Proc Natl Acad Sci USA. 2006;103:8119–8124. [PMC free article] [PubMed]
40. Watanabe T, Tarttelin E, Neubiser A, Kimura M, Solter D. Mamm Genome. 1994;5:768–770. [PubMed]
41. DeLoia JA, Solter D. Development. 1990;(Suppl):73–79. [PubMed]
42. Hagiwara Y, Hirai M, Nishiyama K, Kanazawa I, Ueda T, Sakaki Y, Ito T. Proc Natl Acad Sci USA. 1997;94:9249–9254. [PMC free article] [PubMed]
43. Kramer MG, Vaughn TT, Pletscher LS, King-Ellison K, Adams E, Erickson C, Cheverud JM. Genet Mol Biol. 1998;21:211–218.
44. Vaughn TT, Pletscher LS, Peripato AC, King-Ellison K, Adams E, Erikson C, Cheverud JM. Genet Res. 1999;74:313–322. [PubMed]
45. Wolf JB, Vaughn TT, Pletscher LS, Cheverud JM. Heredity. 2002;89:300–310. [PubMed]
46. Goodale HD. J Heredity. 1938;29:101–112.
47. MacArthur J. Amer Nat. 1944;78:142–157.
48. Chai C. Genetics. 1956;41:157–164. [PMC free article] [PubMed]
49. Ehrich TH, Kenney-Hunt JP, Pletscher LS, Cheverud JM. Genet Res Camb. 2005;85:211–222. [PubMed]
50. Cheverud JM, Ehrich TH, Kenney JP, Pletscher LS, Semenkovich CF. Diabetes. 2004;53:2700–2708. [PubMed]
51. Kenney-Hunt JP. Pleiotropic patterns of quantitative trait loci for seventy murine skeletal traits. Genetics. 2008 in press. [PMC free article] [PubMed]
52. Hrbek T, de Brito RA, Wang B, Pletscher LS, Cheverud JM. Mamm Genome. 2006;17:417–429. [PubMed]
53. Cockett NE, Jackson SP, Shay TL, Farnir F, Berghmans S, Snowder GD, Nielsen DM, Georges M. Science. 1996;273:236–238. [PubMed]
54. Cheverud JM. In: Evolutionary Genetics: Concepts and Case Studies. Fox CW, Wolf JB, editors. Oxford: Oxford Univ Press; 2006.
55. Li J, Jiang T. J Bioinfo Comp Biol. 2003;1:41–69. [PubMed]
56. Qian D, Beckman L. Am J Hum Genet. 2002;70:1434–1445. [PMC free article] [PubMed]
57. Haley CS, Knott SA. Heredity. 1992;69:315–324. [PubMed]
58. Leamy L, Routman EJ, Cheverud JM. Am Nat. 1999;153:201–214.
59. Mantey C, Brockmann GA, Kalm E, Reinsch N. J Hered. 2005;96:329–338. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • MedGen
    MedGen
    Related information in MedGen
  • PubMed
    PubMed
    PubMed citations for these articles
  • SNP
    SNP
    PMC to SNP links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...