t(15;17); How does it help make a normal cell leukemic?

It has proved easier to identify translocations associated with specific hematological malignancies than it has to understand how the translocation contributes to the leukemic phenotype. In a novel study, DiCroce et al [Science 295:1079-82, 2002] have examined the function of the PML-RAR gene fusion generated by the t(15;17), typical of acute promyelocytic leukemia (APL). The authors report that the leukemia-promoting PML-RAR fusion protein induces gene hypermethylation and silencing by recruiting DNA methyltransferase to the target promoters, and that hypermethylation contributes to its leukemic potential. Retinoic acid treatment is shown to induce promoter demethylation, gene re-expression and reversion of the transformed phenotype. 

COMMENT: This is an important insight into the mechanism by which a translocation contributes to the leukemic phenotype. The normal retinoic acid receptor gene product, when bound to retinoic acid, alters gene expression in immature white blood cells, causing them to mature and stop dividing. This study explains the important clinical observation that the phenotype can be reversible not only with retinoic acid but also with demethylating drugs such as deazacytidine. This important observation might lead to new or additional therapies aimed at blocking methylation of the target gene RAR-beta2 and other targets. This study will certainly lead to additional investigations of methylation in other malignancies associated with fusion proteins and alterations in methylation. 

Environmental radiation, nuclear tests and germline mutation

Considerable effort has been applied to determine if there is a role for very low doses of environmental radiation contamination from nuclear sites in childhood cancer. Uncertainty still surrounds the biological and clinical consequences of direct and parental exposure to such contamination. A study by Dubrova [Science 295:137, 2002] has addressed the frequency of germline mutation in a population exposed to nuclear testing. The Semipalatinsk nuclear test site has been the site of 470 nuclear tests performed by the Soviet Union during the period 1949-89, including atmospheric and surface explosions from 1949-63, and underground tests from 1963-81. The surrounding population was mainly exposed to the fresh radioactive fallout from four surface explosions conducted between 1949-56, and currently the radioactive contamination outside of the test site is low.  Blood samples were collected from 40 three-generation families inhabiting the rural area of the Beskaragai, district of Kazakhstan, around the Semipalatinsk nuclear test site. These areas are characterized by the highest effective doses of exposure to ionizing radiation (>1 Sv). A control group was composed of 28 three-generation non-radiated families from the geographically similar non-contaminated area of Taldy Kurgan, district of Kazakhstan. Groups were matched by ethnicity, year of birth, parental age, occupation, and whether or not they were smokers. Both parents and offspring were profiled with eight hypervariable mini-satellite probes. The data showed that exposure to radioactive fallout from the nuclear weapons tests carried out at this test site in the late 1940s to early 1950s (the period of maximum exposure) roughly doubled the germline mutation rate in the affected population. Most importantly, there was a negative correlation between mutation rate and parental year of birth, providing experimental evidence that the human germline mutation rate changed with declining exposure to ionizing radiation. This suggests that the Moscow Treaty banning nuclear weapons tests in the atmosphere (August 1963) has affected the reduced genetic risk to the affected population. 

COMMENT: This study supports a controversial 1996 report by Dubrova and a different group of colleagues that linked germline mutations to fallout from the 1986 Chernobyl explosion [Nature 380: 683, 1996]. That study described double the usual mutation rate in the children of men living in the region of Belarus heavily contaminated with cesium-137.  Although intriguing, the biological meaning of the data from both these studies, remains unclear. The mini-satellites examined were once thought to constitute “junk DNA”. However, it now seems possible that at least some minisatellites are of importance in regulation of gene expression or possibly disease, so this observation might have direct biological consequences. Alternatively, minisatellite mutation may simply be a biomarker of exposure that could be used in other studies, though it is important to realize the exposure in this instance was high, and higher than is likely to occur intentionally in current years. It remains to be seen if increased microsatellite mutation per se will translate into disease of any kind. 

Beckwith-Wiedemann Syndrome, predicting risk of malignancy?

Beckwith-Wiedemann syndrome (BWS) is a complex disorder that involves frequent alterations in the imprinting and methylation of several genes on chromosomal band 11p15, including IGF-2, H19, KvLQT-1 and LIT1. Genomic imprinting is a genetic modification that typically involves DNA methylation but may involve other factors as well.  Imprinting means that only the allele inherited from a specific parent is expressed in the cells of offspring. The phenotypic features of BWS include macrosomia, neonatal hypoglycemia, midline abdominal defects (e.g., omphalocele, diastasis recti and umbilical hernia) and macroglossia. The increased incidence of embryonal cancers in infants and young children is well known.  The most frequent cancer is Wilms' tumor, occurring in about 11% of children with BWS before 4 years of age,  followed by other embryonal tumors including hepatoblastoma and neuroblastoma, with a lower frequency of rhabdo-myosarcoma and adrenocortical carcinoma. In an analysis based on the BWS Registry established in 1994 at the Genetic Epidemiology Branch of the National Cancer Institute, DeBaun et al. [Am J Hum Genet 70:604-11, 2002] examine the correlation between the genetic basis for BWS in an individual patient and the risk of cancer.  The work shows that the frequency of altered DNA methylation of H19 in patients with cancer was significantly higher, 56% (9/16), than the frequency in patients without cancer, 17% (13/76; p=0.002). Cancer was not associated with LIT1 alterations.  Furthermore, the frequency of altered DNA methylation of LIT1 in patients with midline abdominal wall defect and macrosomia was significantly higher, 65% (41/63) and 60% (46/77), respectively, than in patients without such defects, 34% (10/29) and 18% (2/11), respectively (p=0.012 and p=0.02, respectively).  In addition, paternal uniparental disomy of 11p15 was associated with hemihypertrophy (p=0.003), cancer (p=0.03), and hypoglycemia (p=0.05). The authors conclude that these results define an epigenotype-phenotype relationship in BWS in which aberrant methylation of  H19 and uni-parental disomy is strongly associated with cancer risk.

COMMENT:  These data show the utility of registries for gathering clinical and biological information on rare diseases.  Biological samples were not available from all patients within the BWS Registry. However, the frequency of cancer and other phenotypic features did not differ significantly between those with a molecular analysis and those without, suggesting these data are applicable to the broader population. These data offer the prospect of predicting risk of malignancy more accurately for an individual patient and moderating cancer screening accordingly. The data also advance our understanding of the contribution of germline molecular abnormalities with phenotype. Of note, the abnormalities that are being examined are not mutations in the DNA sequence itself, but epigenetic alterations (i.e., alterations in methylation). 

Medulloblastoma- caused by a virus?

The human polyoma virus JC virus (JCV) is associated with progressive multi-focal leukoencephalopathy, a fatal demyelinating disease seen in immuno-compromised patients. When JCV is injected into the brain of monkeys it induces brain tumors in 100% of animals. A new study [Del Valle et al, J Natl Cancer Inst 2002; 94:267-73]) demonstrates the presence of JCV T antigen and agnoprotein DNA in around 65% of human medulloblastomas, and expression of protein in around half the cases studied. The JCV T antigen has been long favored as the cause of tumorigenesis after JCV infection as it is known to associate with p53 and Rb proteins that control cell proliferation. The authors of this study propose that their data implicate agnoprotein in the mechanism of tumorigenesis, although the biological role of this viral protein is unknown. 

COMMENT: Correctly interpreting the significance of the presence of a common virus when it turns up in tumor tissue is a challenge. An illuminating editorial by Howard Fine accompanies this paper and discusses the issue.  Infection with JCV is common, with 65% of children infected by the age of 14, and the virus is thought to circulate in B cells.  Laboratory models of JCV infection provide somewhat contradictory results regarding a role in tumorigenesis. T antigen is known to induce tumors similar to medulloblastoma in transgenic animals and is known to bind and functionally deregulate the p53 and Rb proteins, a plausible mechanism of tumorigenesis. Notably, 100% of transgenic mice that are conditionally Rb-null/p53 null develop brain tumors that are histologically identical to medulloblastoma [Marino et al, Gene Dev. 2000, 14: 994].  Patients with Li-Fraumeni syndrome (a germline p53 mutation) are at increased risk of medulloblastoma, but surprisingly p53 null mice do not get tumors. In addition, sporadic medulloblastomas rarely, if ever, have p53 or Rb gene mutations or deletions. In aggregate, the data regarding JCV actually being the cause of the malignant phenotype in medulloblastoma leaves the case unproven. It is also possible, as Dr. Fine points out that the importance of JCV is less direct, perhaps causing proliferation of a cell population that would then be vulnerable to later genetic hits. This idea is somewhat analagous to the Greaves hypothesis of early infection causing B cell proliferation and predisposing to later genetic events leading to ALL.  Indeed, Malcolm Smith has proposed that JCV might play such a role in the etiology of ALL [C3 vol 8 No. 3]. Careful (and large) epidemiological, not biological, studies would be required to demonstrate such an effect in meduloblastoma. 

C3 Quarterly Newsletter
Children's Cancer Research Fund
Epidemiology Research Unit
Division of Pediatric Epidemiology
Clinical Research
University of Minnesota
420 Delaware St. SE, Box 422
Minneapolis, MN 55455
pedsepi@umn.edu

Editors: 
Stella M. Davies, MD, PhD, and Julie A. Ross, PhD