Falls et al. (1999) — Genomic Imprinting: Implications for Human Disease

Bibliographic Reference (APA 7.0)

Falls, J. G., Pulford, D. J., Wylie, A. A., & Jirtle, R. L. (1999). Genomic imprinting: Implications for human disease. The American Journal of Pathology, 154(3), 635—647. https://doi.org/10.1016/S0002-9440(10)65309-6

Core Argument

Genomic imprinting is the epigenetic marking of a gene based on its parental origin that results in monoallelic expression — a notable exception to Mendelian genetics in which the parental complements are not equivalent. The “imprint mark” is a parental-specific methylation of CpG-rich domains established during gametogenesis, erasable in the germline when transmitted through the opposite sex but maintained during somatic cell division. Imprinted genes are involved in fetal and placental growth, cell proliferation, and adult behavior. Consequently, abnormal expression of imprinted genes results in numerous human genetic disorders including cancer. This review surveys the mechanisms, evolution, and disease consequences of genomic imprinting.

Methods

This is a comprehensive review article synthesizing evidence from multiple experimental and clinical approaches: (1) mouse embryo manipulation studies (androgenetic and parthenogenetic embryos, pronuclear transplantation) demonstrating that both parental genomes are required for normal development; (2) uniparental disomy (UPD) mapping in mice identifying imprinted chromosomal regions; (3) targeted gene disruption studies confirming parent-of-origin monoallelic expression for Igf2, M6p/Igf2r, and Mest; (4) positional cloning and subtractive hybridization for identifying novel imprinted genes; (5) clinical genetics of human imprinting disorders — Beckwith-Wiedemann syndrome (BWS), Prader-Willi syndrome (PWS), and Angelman syndrome (AS); and (6) tumor analyses examining loss of imprinting (LOI), loss of heterozygosity (LOH), and UPD in human cancers.

Key Findings

• Imprinted genes are functionally haploid. Because only one allele is expressed, an imprinted tumor suppressor gene requires only one hit — a single mutational or epigenetic event — to eliminate tumor suppressor function, increasing cancer susceptibility relative to biallelically expressed loci that require two hits.

• Loss of imprinting (LOI) at the IGF2 locus is the most common molecular event in BWS patients without cytogenetic abnormalities and is found in approximately 70% of Wilms’ tumors. This biallelic expression of the fetal growth factor IGF2, often coupled with methylation and silencing of the maternal H19 allele, is consistent with an enhancer-competition model for co-regulation. LOI of IGF2 has been documented in over 20 tumor types including rhabdomyosarcoma, hepatoblastoma, bladder, breast, colorectal, lung, prostate, and testicular germ cell cancers.

• M6P/IGF2R functions as a tumor suppressor gene. It degrades IGF2, activates the growth inhibitor TGF-beta-1, and sorts lysosomal enzymes. It is mutated in 60% of dysplastic liver lesions and hepatocellular carcinomas and in 30% of breast tumors; allelic inactivation is an early event occurring at initiation rather than progression. Unlike the mouse (where it is imprinted), human M6P/IGF2R shows polymorphic imprinting — individuals with an imprinted M6P/IGF2R locus may have increased tumor susceptibility because only one allele is active.

• Imprinted genes contribute to carcinogenesis through three distinct mechanisms. (i) LOH or UPD at an imprinted region may delete the only functional copy of a tumor suppressor gene; (ii) LOI or UPD of an imprinted growth-promoting gene may allow its expression to be inappropriately increased; (iii) mutational inactivation of an imprinting center (IC) may result in aberrant expression of multiple imprinted oncogenes and/or tumor suppressors within an imprinted chromosomal cluster.

• Additional imprinted tumor suppressors and growth regulators were identified. p57[KIP2] (CDKN1C) at 11p15.5 encodes a maternally expressed cyclin-dependent kinase inhibitor; epigenetic silencing is found in some tumors and BWS patients, though p57[KIP2] coding mutations have not been identified in tumors. NOEY2 (ARHI) at 1p31 is a maternally imprinted Ras-related putative tumor suppressor lost in the majority of breast and ovarian carcinomas. WT1 at 11p13 exhibits polymorphic imprinting.

Concepts Introduced or Used

• Genomic imprinting (gametic imprinting, parental imprinting) — epigenetic marking based on parental origin resulting in monoallelic expression
• Loss of imprinting (LOI) — reactivation of the silenced allele, producing biallelic expression; a cancer mechanism
• Uniparental disomy (UPD) — inheritance of both copies of a chromosome or region from one parent; can duplicate an active imprinted oncogene or delete the only active copy of an imprinted tumor suppressor
• Imprinting center (IC) — a chromosomal region controlling the imprinting status of multiple genes in a cluster; mutagenesis of the IC can dysregulate multiple imprinted genes simultaneously
• Enhancer-competition model — mechanism for coordinate regulation of IGF2 and H19 via competition for shared downstream enhancers
• Polymorphic imprinting — imprinting that varies among individuals in a population (M6P/IGF2R, WT1, IGF2, HTR2A in humans)
• Functional haploidy — the state of an imprinted gene where only one allele is expressed; has the same vulnerability as a hemizygous locus to single-hit inactivation
• CpG island methylation — the biochemical basis of the imprint mark; parental-specific methylation of CpG-rich domains

Entities Referenced

• IGF2 — paternally expressed fetal growth factor at 11p15.5; LOI found in >20 tumor types
• H19 — maternally expressed untranslated RNA at 11p15.5; involved in silencing IGF2
• M6P/IGF2R — mannose 6-phosphate/insulin-like growth factor 2 receptor at 6q26; tumor suppressor mutated in liver and breast cancer
• p57[KIP2] (CDKN1C) — maternally expressed cyclin-dependent kinase inhibitor at 11p15.5
• WT1 — Wilms’ tumor suppressor at 11p13 with polymorphic imprinting
• NOEY2 (ARHI) — maternally imprinted Ras-related putative tumor suppressor at 1p31
• KvLQT1 — maternally expressed gene at 11p15.5 spanning a BWS common breakpoint region; proposed to maintain regional imprint control
• PEG1/MEST — paternally expressed mesoderm-specific hydrolase at 7q32; targeted deletion demonstrated role in embryonic growth and maternal nurturing behavior
• SNRPN, IPW, ZNF127, NDN (necdin) — paternally expressed candidate genes for Prader-Willi syndrome at 15q11-q13
• UBE3A — maternally expressed ubiquitin protein ligase at 15q11-q13; mutated in Angelman syndrome; imprinted in brain
• Beckwith-Wiedemann syndrome (BWS) — overgrowth disorder at 11p15.5 with 10—20% risk of embryonal tumors; Wilms’ tumor rate 1000-fold higher than normal
• Prader-Willi syndrome (PWS) — paternal deficiency disorder at 15q11-q13
• Angelman syndrome (AS) — maternal deficiency disorder at 15q11-q13

Limitations (as stated by authors)

• The mechanism of imprinting is “complex and not completely understood” — the precise molecular events of imprint establishment, maintenance, and reading remain areas of active investigation.
• Estimates based on mouse models indicate 100—200 imprinted genes may exist, but only “more than 25” had been identified at the time of writing, leaving the majority of imprinted genes undiscovered.
• Polymorphic imprinting in humans (M6P/IGF2R, WT1, IGF2, HTR2A) complicates generalizability — an individual’s imprinting status at a given locus may determine disease susceptibility, making population-level predictions difficult.
• The putative tumor suppressor function of p57[KIP2] “remains to be clarified” since coding mutations had been found in approximately 5% of BWS patients but not in tumors.
• The involvement of genomic imprinting in behavioral disorders (bipolar affective disorder, schizophrenia, autism) “remains to be elucidated” — observed parent-of-origin effects may or may not reflect imprinting mechanisms.
• The functional significance of polymorphic imprinting — whether it “functions in determining individual and/or species differences in susceptibility to diseases remains to be determined.”

Relevance to Clonal Evolution

This review provides a critical mechanistic link between epigenetics and the core principles of clonal-evolution.

Imprinted tumor suppressors as single-hit targets. Imprinted genes are functionally haploid — only one allele is expressed. For an imprinted tumor suppressor, loss of function requires only one hit (mutation or epigenetic silencing of the sole active copy), rather than the two hits required for a biallelically expressed locus. This means imprinted tumor suppressor loci are intrinsically more vulnerable to inactivation. From the perspective of driver-mutation kinetics, the waiting time for a loss-of-function driver is effectively halved at imprinted loci. The paper explicitly states: “an imprinted tumor suppressor gene would be predicted to increase cancer susceptibility since the inactivation of only one allele would eliminate tumor suppressor function” (p. 640).

LOI as an epigenetic driver of clonal-expansion. Biallelic expression of IGF2 via LOI provides a constitutive growth signal — IGF2 is a potent fetal growth factor. The paper documents that 70% of Wilms’ tumors show this biallelic expression, which can occur as an early event (detectable in phenotypically normal kidney tissue surrounding the tumor). LOI of IGF2 is thus an epigenetically mediated driver-mutation — it does not alter the DNA sequence of IGF2 itself but produces the same functional consequence as a gain-of-function mutation by doubling gene dosage. This is an important demonstration that heritable changes in gene expression (not just DNA sequence changes) can initiate clonal-expansion.

Imprinting centers as master regulators of genome stability. The discovery of imprinting centers (ICs) — chromosomal regions that control imprinting across an entire cluster — introduces the concept that a single mutational event (e.g., microdeletion at the SNRPN IC) can simultaneously dysregulate multiple imprinted oncogenes and tumor suppressors. This is an epigenetic parallel to genetic-instability: rather than increasing the mutation rate itself, IC disruption increases the phenotypic mutation rate by simultaneously altering the expression dosage of multiple functionally haploid cancer genes. The paper notes that the minimally deleted regions for PWS and AS are distinct, suggesting a bipartite IC structure with separable elements.

UPD as a recombination-based mechanism with clonal consequences. The paper describes how UPD at 11p15 can arise through postfertilization mitotic recombination events, producing somatic mosaicism in BWS. This is a concrete mechanism for generating intratumor-heterogeneity: a recombination event in a single cell can create a lineage with paternal duplication of IGF2 (biallelic expression without LOI) and maternal loss of p57[KIP2] and H19 simultaneously. Such an event functions as a compound driver, producing multiple fitness-affecting changes in a single mitotic event.

Polymorphic imprinting and differential cancer susceptibility. The finding that M6P/IGF2R imprinting is polymorphic in humans — most individuals have biallelic expression, but some have imprinted (monoallelic) expression — introduces the concept of inherited epigenetic variation affecting cancer risk. An individual with an imprinted M6P/IGF2R has only one active allele of this tumor suppressor, making them constitutionally more susceptible to its loss. This is analogous to a germline heterozygous mutation but arises epigenetically. The paper cites Xu et al. (1997) reporting partial imprinting of M6P/IGF2R in 50% of Wilms’ tumor patients as supporting evidence.

Relevance to intermediate-clone dynamics. The finding that LOI of IGF2 and inactivation of H19 occur early in development and are detectable in phenotypically normal kidney surrounding Wilms’ tumors (Okamoto et al., 1997, cited in the paper) is consistent with the intermediate-clones concept: the epigenetically altered cell population expands as a phenotypically normal but predisposed field (a “first hit” clone), within which a subsequent genetic event triggers frank neoplasia.