NEW YORK (GenomeWeb News) – A study appearing online in JAMA Psychiatry suggests methylation marks in and around dozens of genes may serve as biomarkers for schizophrenia, providing potential clues to disease subtypes and patients' past environmental exposures.
Researchers from the US and Virginia did a methylome-wide association study involving 759 individuals with schizophrenia and almost as many unaffected controls, using Invitrogen's MethylMiner kit to enrich for and assess methylation-prone sites. When they did see apparent ties to schizophrenia, methylation marks in those regions were subsequently verified by bisulfite sequencing.
All told, the researchers identified 139 methylation-related sites showing significant associations with the psychiatric disorder. Among them: a locus near the RELN gene that's been implicated in prior methylation-based studies of schizophrenia.
Those involved in the study noted that some of the schizophrenia-associated methylation marks they detected might reflect immune responses to past infections and/or epigenetic shifts triggered by environmental exposures or experiences suspected of dialing up schizophrenia risk, such as hypoxia.
"These markers, possibly from as early as embryonic development or birth, could be of considerable clinical importance as they could allow for identification of distinct schizophrenia subtypes or help predict treatment responses," first author Karolina Aberg, associate director of Virginia Commonwealth University's Center for Biomarker Research and Personalized Medicine, said in a statement.
In Nature Genetics, an international team led by investigators in the UK described findings from a sequencing-based effort to track down mutations contributing to acute lymphoblastic leukemia development in cases that include a characteristic fusion between the ETV6 gene and the RUNX1 gene.
Through low-coverage whole-genome sequencing on samples from 51 individuals with childhood ALL, the researchers verified the presence of ETV6-RUNX1 fusion in the tumors and tracked down hundreds of other structural variations.
To that, they added exome sequencing on matched tumor and remission samples from 50 of the original cases and five more cases of ETV6-RUNX1 fusion-containing ALL, which revealed nearly 800 somatic substitutions and more than a dozen small insertions and deletions across the tumor set.
In particular, the team found evidence of frequent deletions affecting the RAG endonuclease genes, which code for proteins that get targeted to variable loci during the development of some immune cells. Sites that the RAG system targets for recombination also appeared prone to excess rearrangement in the ETV6-RUNX1 fusion-containing ALL tumors, the group reported.
Together with findings from their genotyping assessment of individual cells taken during different stages of leukemia development, the researchers' results suggest RAG-mediated processes normally involved in immune cell differentiation and diversity sometimes go awry in those who acquire the ETV6-RUNX1 fusion prenatally in their B cell progenitor cells.
"It may seem surprising that evolution should have provided a mechanism for diversifying antibodies that can collaterally damage genes that then contribute to cancer," co-senior author Mel Greaves, with the Institute of Cancer Research in London, said in a statement. "But this only happens because the fusion gene (ETV6-RUNX1) that initiates the disease 'traps' cells in a normally very transient window of cell development where the RAG enzymes are active, teasing out their imperfect specificity."
A Broad Institute-led team that included members of the Multiple Myeloma Research Consortium published a study in Cancer Cell that highlighted the extensive genetic variability present in multiple myeloma tumors.
The researchers did whole-exome sequencing on matched tumor and normal samples from 177 individuals with multiple myeloma, along with whole-genome sequencing on tumor-normal pairs from another 26 individuals with the blood cancer.
Sifting through this sequence data, the team saw recurrent alterations in the tumors, including mutations in several known cancer players such as NRAS, TP53, and NRAF. KRAS alterations were common too, especially in tumor samples from individuals who'd received treatment for their cancers.
The distribution of such genetic glitches was often far from uniform, though. Instead, the investigators saw signs of pervasive heterogeneity across sub-populations of tumor cells, even within the same individual — information that may prove important for accurately characterizing and combating the disease in the future.
"[W]hen we treat an individual patient with multiple myeloma, it's possible that we're not just looking at one disease, but at many — in the same person, there could be cancer cells with different genetic make-ups," co-senior author Todd Golub, chief scientific officer at the Broad, said in a statement.
"These findings indicate a need to identify the extent of genetic diversity within a tumor as we move toward precision cancer medicine and genome-based diagnostics," added Golub, who is also affiliated with the Dana-Farber Cancer Institute and Harvard Medical School.
Researchers from Rockefeller University and the Icahn School of Medicine at Mount Sinai's Institute for Genomics and Multiscale Biology used an in vivo RNA interference screening-based strategy to systematically knock down mouse genes as part of an effort to find new tumor suppressor gene candidates in squamous cell carcinoma — work that they reported in Science.
Using a set of 1,762 small hairpin RNAs, the investigators knocked down almost 350 mouse genes, dialing in the delivery of their RNAi library to the ectoderm layer of nine-and-a-half day old mouse embryos with an ultrasound-guided method and a lentivirus-based vector.
By sequencing DNA from embryos that went on to develop squamous cell carcinoma-like lesions, the team determined which of the shRNAs were enriched in the samples. That, in turn, pointed to their gene targets as possible players in tumor suppression.
The resulting gene set included BRCA1, already known for its role in the cancer, as well as seven genes not suspected of suppressing squamous cell carcinoma or other tumor types in the past.
One of the potential tumor suppressors caught their particular attention: the mysosin IIa-coding gene Myh9, which spells out the sequence for a form of myosin IIa heavy chain that is found in non-muscle tissue.
In follow-up experiments, the team saw that mouse and human cells missing the gene had impaired p53 stability, pointing to a previously unappreciated role for myosin IIa. Levels of myosin IIa appeared lower than usual in some human head and neck squamous cell carcinomas assessed by members of the Cancer Genome Atlas, particularly those from individuals with poor survival outcomes.