Welcome to the laboratory of Randy Jirtle in the Department of Radiation Oncology at Duke University. The overall objective of our research program is to determine the role of genomic imprinting in carcinogenesis, embryogenesis, and toxicology. Genomic imprinting refers to an epigenetic marking of genes that results in monoallelic expression. This parent-of-origin dependent phenomenon is a notable exception to the laws of Mendelian genetics (Murphy and Jirtle, 2003). Imprinted genes are normally involved in embryonic growth and behavioral development (Murphy et al., 2001), but they also function in somatic cells as oncogenes and tumor suppressor genes (Jirtle, 1999).

We are presently investigating the role of M6P/IGF2R, DLK1, and PEG3 in carcinogenesis (De Souza et al., 1995; Yamada et al., 1997; Wylie et al., 2000); and Murphy et al., 2001). We are also testing the utility of using phylogenetic comparisons of known imprinted domains in Metatherians (i.e. opossum) and Eutherians (i.e. mouse and humans) to orthologous regions in non-imprinted Prototherians (i.e. platypus) to identify novel human imprinted genes and their cis-acting regulatory elements (Murphy and Jirtle). These studies should provide those critical bioinformatic approaches required to more effectively characterize imprinted domains known to harbor genetic and/or epigenetic mutations mechanistically involved in behavioral disorders such as schizophrenia, bipolar disease and autism.

We were the first to demonstrate that the mannose 6-phosphate/insulin-like growth factor 2 receptor (M6P/IGF2R) gene is a tumor suppressor (De Souza et al., 1995; Yamada et al., 1997). It encodes for a multifunctional receptor required for the activation of the potent growth inhibitor, transforming growth factor beta (TGFß), and the degradation of IGF2, a mitogen often overproduced in tumors, and the internalization of granzyme B involved in T-cell mediated apoptosis. We have demonstrated that the M6P/IGF2R is often mutated in dysplastic liver lesions and hepatocellular carcinomas (HCCs) in patients with or without hepatitis virus (HV) infection (Jirtle, 1999; De Souza et al., 1995; Yamada et al., 1997).

Furthermore, in HV-infected patients M6P/IGF2R inactivation has already occurred in the phenotypically normal tissue adjacent to these dysplastic lesions and HCCs (Yamada et al., 1997). This demonstrates that M6P/IGF2R inactivation occurs early rather than late in liver carcinogenesis, and that the liver of chronic HV-infected people consists of large clones of normal appearing M6P/IGF2R-mutated precancerous liver cells from which a majority of liver tumors ultimately develop. It is because of this large increase in precancerous liver cells that HV-infected patients have an enhanced risk of developing HCC.

The M6P/IGF2R has also been shown to be frequently mutated in breast and lung cancer and in colon, gastric and endometrial cancers with mismatch repair defects (Jirtle, 1999). Moreover, M6P/IGF2R loss of heterozygosity in head and neck cancer predicts for poor therapeutic outcome in patients treated with radiotherapy alone (Jamieson et al., 2003). Patients with M6P/IGF2R allelic loss also benefit most from concurrent chemotherapy. These findings raise the intriguing possibility that selection of head and neck cancer patients for concurrent chemotherapy can be based upon the intratumoral mutational status of M6P/IGF2R. The pathogenesis of M6P/IGF2R inactivation can now be systematically investigated in vivo with our development of a Cre/LoxP conditional M6P/IGF2R knockout mouse (Wylie et al., 2003).

A. Imprinting Evolution
We have identified the ancestral mammalian origin of parental imprinting of growth regulatory genes in the IGF-pathway, and demonstrated that both IGF2 and M6P/IGF2R are imprinted in Therain (i.e. marsupials and placental mammals), but not in Prototherians (i.e. monotremes) mammals (Killian et al., 2000). Thus, the M6P/IGF2R is a cancer susceptibility gene that was imprinted approximately 150 million years ago in an ancestor common to marsupials and Eutherians. This ancestral origin of imprinting was followed by a more recent lineage-restricted loss of M6P/IGF2R imprinting in all members of the Euarchonta clade that includes Primates (Killian et al., 2001).

Thus, the divergent evolution of M6P/IGF2R imprinting in primates and rodents is consistent with humans being more resistant to cancer formation because two rather than only one allele of this tumor suppressor must be mutated to inactivate its function. Moreover, the divergent evolution of M6P/IGF2R imprinting predicts that the success of in vitro embryo procedures such as cloning may be species-dependent with human embryo development resulting from either in vitro fertilization or cloning not being as adversely impacted by large offspring syndrome (LOS) as rodents and ungulates since the M6P/IGF2R is not imprinted in humans.

B. DLK1/MEG3 Genes
Using a bioinformatics-based approach, we identify a novel imprinted domain on human chromosome 14q32 that contains two reciprocally imprinted genes (Wylie et al., 2000). MEG3 (maternally expressed gene 3) is maternally expressed and appears to lack an open reading frame while DLK1 (delta, drosophila like homolog 1) is paternally expressed, and encodes for a cell-surface transmembrane protein that closely related to the invertebrate proteins delta and notch. Many of the predicted structural and regulatory features of the DLK1/GTL2 domain are highly analogous to those implicated in IGF2/H19 imprint regulation, including two hemi-methylated consensus binding sites for the vertebrate enhancer blocking protein, CTCF. These results provide the first evidence that a common mechanism and domain organization may be used for juxtapositioned reciprocally imprinted genes.

The paternal expression, chromosomal localization, and biological function of DLK1 also make it a likely candidate gene for the callipyge (Greek: calli-, beautiful; -pyge, buttocks) gene (CLPG) whose mutation in sheep results in animal leanness and fast twitch muscular hypertrophy. Nevertheless, sequence analysis of these and other candidate genes in the linkage interval failed to identify mutations. Further sequencing of the entire linkage region using sheep identical-by-descent to the founder animal revealed a single A/G transition mutation that perfectly correlated with the callipyge phenotype (Freking et al., 2002). This mutation was not in a previously recognized gene or regulatory region; however, comparison of human, mouse, bovine, and ovine sequence showed that a 144 bp region encompassing the mutation was highly conserved. Further investigation led to the identification of a novel transcript produced from the region containing the mutation, lending credence to the power of comparative genomics to identify otherwise elusive regulatory elements and genes.

C. PEG3 Gene
The Krüppel-type zinc finger protein, PEG3 (paternally expressed gene 3), is implicated in critical cellular and behavioral functions including growth, apoptosis and maternal nurturing behavior. The PEG3 promoter in humans is encompassed within a large CpG-rich region that is differentially methylated (Murphy et al., 2001). We have also shown that PEG3 is ubiquitously imprinted throughout development and postnatally, and that its multiple isoforms are paternally expressed. This is the first evidence of an imprinted gene residing on human chromosome 19q13.4. The imprinted status of PEG3 throughout life coupled with its neural expression and putative roles in regulating cell growth suggest that PEG3 in humans may be a susceptibility locus for cancer as well as neurobehavioral deficits such as maternal nurturing.

We have identified the ancestral mammalian origin of parental imprinting of growth regulatory genes in the IGF-pathway, and demonstrated that both IGF2 and M6P/IGF2R are imprinted in Therian but not in Prototherian mammals. The subsequent divergence of primates and their close relatives from artiodactyls, rodents and marsupials in M6P/IGF2R imprinting infers that male and female mammalian ancestors to humans resolved an epigenetic struggle at the M6P/IGF2R locus that would otherwise have resulted in enhanced cancer susceptibility (Jirtle, 1999; De Souza et al., 1995; Yamada et al., 1997) and many more visits to the clinic (Killian et al., 2001).

The identification of the CLPG gene in the DLK1/MEG3 imprinted domain further demonstrates the power of phylogenetic comparisons in delineating novel imprinted genes and their cis-acting regulatory elements (Freking et al., 2002; Murphy and Jirtle, 2003). Our finding that PEG3 is imprinted not only in mice but also in humans indicates that a single mutation in the expressed paternal allele has the potential of compromising one of the most fundamental of all human behaviors, maternal nurturing (Murphy et al., 2001). We will surely be treated to many more surprises as we identify additional imprinted genes, and decipher the mechanisms of imprinting regulation and the process by which this novel form of gene regulation evolved.

Imprinting Publications

> Murphy and Jirtle, 2003 Imprinting Evolution and the Price of Silence. BioEssays 25: 2003 (In Press) > Freking et al., 2002 Identification of the single base change causing the callipyge muscle hypertrophy phenotype, the only known example of polar overdominance in mammals. Genome Res. 12: 1496-1506, 2002. [Journal Cover] [PDF] > Killian et al., 2001 Divergent evolution in M6P/IGF2Rimprinting from the Jurassic to the Quaternary. Hum. Mol. Genet. 10: 1721-1728, 2001. [Journal Cover] [PDF] > Murphy et al., 2001 Imprinting of PEG3,the human homolog of a mouse gene involved in nurturing behavior. Genomics 71: 110-117, 2001. [Journal Cover] [PDF] > Wylie et al., 2000 Novel Imprinted DLK1/GTL2 Domain on Human Chromosome 14 Contains Motifs that Mimic Those Implicated in IGF2/H19 Regulation. Genome Res. 10: 1711–1718, 2000. [PDF] > Killian et al., 2000 M6P/IGF2R imprinting evolution in mammals. Mol. Cell 5: 707-716, 2000. [PDF] > Jirtle, 1999 Genomic imprinting and cancer. Exp. Cell Res. 248: 18-24, 1999. [PDF]

M6P/IGF2R Publications

> Jamieson, T.A. et al., 2003 M6P/IGF2R loss of heterozygosity in head and neck cancer associated with poor patient prognosis. BMC Cancer 3: 4, 2003. [PDF] > Wylie et al., 2003 Tissue-specific inactivation of murine M6P/IGF2R. Am. J. Pathol. 162: 321-328, 2003. [PDF] [Commentary] > Yamada, et. al., 1997 Loss of the gene encoding mannose 6-phosphate/insulin-like growth factor II receptor is an early event in liver carcinogenesis. Proc. Natl. Acad. Sci. USA 94: 10351-10355, 1997. [PDF] > De Souza, et. al, 1995 M6P/IGF2R gene is mutated in human hepatocellular carcinomas with loss of heterozygosity. Nat. Genet. 11: 447-449, 1995.