Aust et al., 673–677 Although both sporadic and ulcerative colitis (UC)-related colorectal cancers arise through a multi-step process, different genomic changes may be involved in each type because of the origin of UC-related cancers in an inflammatory environment. For example, the frequency of KRAS mutations is lower in UC-related than in sporadic colorectal tumors but nothing is known about the prevalence of BRAF mutations in UC-related cancers. Mutations in RAS GTPases or the BRAF Serine/Threonine kinase, a downstream RAS effector, contribute to the initiation and progression of cancer by altering the normal regulation of the mitogen-activated protein kinase (MAPK) cascade. To clarify the contribution of these genetic alterations in UC-related tumors the authors took a closer look at the BRAF and RAS status in 33 UC-related colorectal cancers. In their series, 9% had a mutation in BRAF and 18% in KRAS. Since both mutations affect the same pathway, it did not come as a surprise that they were mutually exclusive. In line with earlier reports that linked mutations in BRAF with sporadic colorectal cancers showing high level microsatellite instability mostly caused by hypermethylation of the hMLH1 promoter, Aust et al. found a similar pattern in UC-related colorectal carcinomas. However, unlike in sporadic tumors, BRAF mutations in UC-related cancer are not associated with early tumor stages and are absent from the corresponding nondysplastic mucosa. Immunohistochemistry showing loss of expression of the MLH1 mismatch-repair protein in the tumor of patient 9 who carries a mutation in codon 11 of BRAF. Thus, differences in the frequency and timing of individual chromosomal alterations suggest that genetic progression in these 2 tumor groups may indeed follow multiple pathways. Kamikihara et al., 690–700 Numerous imprinted genes show abnormal changes in gene expression and alterations in the normal pattern of methylation in cancer. One of the most common findings has been a “loss of imprinting,” which refers to epigenetic modifications that result in the activation of gene expression on the normally silenced allele. However, some CpG islands become hypermethylated, leading to transcriptional silencing of critical tumor suppressor genes. ZAC, a paternally expressed zink finger protein, seems to belong to the latter category, report Kamikihara et al. Earlier studies found hypermethylation of the ZAC locus in ovarian cancer but failed to demonstrate a clear link between increased methylation and decreased mRNA expression levels. Now, the authors of the latest study show just that. In addition to a number of ovarian cancer cells lines, they could correlate DNA hypermethylation in tissue samples from 28 primary tumors with the loss of gene expression in both localized and advanced tumor groups. This suggests that loss of ZAC function is a relatively early event in ovarian tumorigenesis. Forced expression of a mouse Zac1 cDNA in an ovarian cancer cell line with silenced ZAC resulted in reduced proliferation and a marked induction of apoptotic cell death. The ZAC-mediated apoptosis signal was p53-independent and eliminated by inhibitors of caspase 3, 8 and 9, identifying ZAC as a potential tumor suppressor gene. Unlike genetic alteration, the reversible nature of imprinting lesions may render them amenable to therapeutic interventions. This exciting potential for treatment underscores the urgent need for further studies like this one to understand the mechanisms by which imprinted genes contribute to tumor development. Valery et al., 799–806 Ewing's sarcoma family of tumors (ESFT) represents a group of rare neoplasms that include Ewing's sarcoma and peripheral primitive neuroectodermal tumors of the bone and soft tissue afflicting mainly children and adolescents. Perhaps not surprisingly, not much is known about the etiology of Ewing's sarcoma. In the past, a small number of studies associated three major groups of factors with Ewing's sarcoma: some pregnancy-related factors; a subject's medical history; and paternal occupation in agriculture and other farm-related factiors such as farm residency and exposure to cows and herbicides. However, the results were not consistent across studies and analyses relied on relatively small sample sizes. To overcome the limitations of these previous studies, Valery et al. pooled 3 case-controlled studies, arriving at 199 cases of ESFT and 1,451 controls. Focusing their analysis on parental occupation they found a pooled odds ratio for parental farm work during the periconception and gestation periods of 2.3 for children whose fathers had worked on farms and 3.9 for those whose mothers had farmed. The risk also increased with the number of years of parental postnatal exposure to farms. In general, the associations were stronger in younger study participants. Just like this childhood cancer study, many others have hypothesized a relationship between parental occupation, particularly farm-related work and the concomitant exposure to pesticides, and the subsequent development of cancer in their children. The challenge ahead is to reveal the biological mechanism linking farming and the risk of ESFTs in offspring.