"This discovery would not have been possible without theunbiased approach taken by the
Pediatric Cancer Genome Project," Dr. Suzanne Baker,corresponding author of the study, said in a press release. Baker is amember of St. Jude's Department of Developmental Neurobiology and co-leader ofthe St. Jude Neurobiology and Brain Tumor Program. "The mutations had not beenreported in any other tumor, so we would not have searched for them in DIPGs. Yetthe alterations clearly play an important role in generating this particulartumor."
The cancer and normal genomes of seven patients with DIPGwere sequenced for the study originally, with another 43 being sequenced afterresearchers discovered the high frequency of the mutations. After sequencingall 16 related genes that make related variants of histone H3 proteins in theadditional 43 patients, it was discovered that many of the tumors have the samemistakes in only two of those genes.
Researchers discovered that 78 percent of the DIPG tumorsstudied had mutations in one of two genes belonging to the histone H3 familythat contain instructions for making proteins that have similar roles inpackaging DNA inside cells. DNA is wrapped around histones so it is compactenough to fit in nuclei, and the packaging influences which genes are switchedon or off, the repair of DNA mutations and DNA stability. If any of theseprocesses are disrupted, it can contribute to cancer.
"It is amazing to see that this particular tumor typeappears to be characterized by a molecular 'smoking gun' and that thesemutations are unique to fast-growing pediatric cancers in the brain," Dr. RichardK. Wilson, director of The Genome Institute at Washington UniversitySchool of Medicine in St. Louis and one of the corresponding authors of thestudy, said in a press release. "This is exactly the type of result one hopesto find when studying the genomes of cancer patients."
Sixty percent of the 50 DIPG tumors sequenced had a singlealteration in the H3F3A gene, which, when translated into a protein, led to thesubstitution of methionine for lysine as the 27th amino acid, while another 18percent had the same mistake in the HIST1H3B gene.
"The amino acid that is mutated most frequently in theDIPGs, which is lysine 27 of H3F3A, is a residue that is modified by a numberof different enzymes, and that modification determines whether or not there aretranscriptional activators or transcriptional repressors recruited to the DNAof selective regions where that modification happens on this amino acid, tohistone 3," says Baker. "And that specific modification is associated, and hasbeen shown previously to be associated, with genes that are really important indifferentiation and in development. And so there's a possibility that thesemutations have a very specific birth advantage only in the context of aparticular developmental stage.
"I think that particularly makes sense for DIPGs, becausethese tumors are seen almost exclusively in children…I think that this reflectsthe context of the developing brain, where childhood gliomas develop comparedto that in the adult brain," she adds.
While mutations in enzymes that target histone H3 have beenseen in other cancers, this is the first case of a specific alteration ofhistones. H3F3A and HIST1H3B were also mutated in glioblastoma tumors, with 36percent of the 36 glioblastoma tumors carrying one of three distinct pointmutations in the genes, as well as a single change in H3F3A's makeup not seenin DIPG tumors. Otherwise, the H3 alterations are not seen in other cancers,not even adult glioblastoma.
"The high frequency of these mutations, 78 percent, suggeststhat this extremely important for the tumor to have this mutation in order tobecome a DIPG," says Baker. "So that makes us hopeful that if we could devise away to target or counteract this mutation, that it might have a very, very bigeffect on the survival or the growth of DIPG tumors. Now exactly what thatintervention would be is not yet clear, so really the next avenue of researchwill be to try to understand exactly how these mutations contribute to tumorformation and growth."
The first authors of the study are Gang Wu, AlbertoBroniscer and Troy McEachron, and the other corresponding authors are JinghuiZhang and James Downing, all of St. Jude. Additional study others includeCharles Lu, Li Ding and Elaine Mardis, from Washington University, and BarbaraPaugh, Jared Becksfort, Chunxu Qu, Robert Huether, Matthew Parker, JunyuanZhang, Amar Gajjar, Michael Dyer, Charles Mullighan, Richard Gilbertson andDavid Ellison, from St. Jude.
