NEW YORK (GenomeWeb News) – Researchers from Washington University School of Medicine are complementing their ongoing human cancer genome sequencing studies by sequencing tumors from mouse models of cancer.
In a paper appearing online today in the Journal of Clinical Investigation, the team described how they used a mouse carrying a human gene fusion linked to a sub-type of acute myeloid leukemia known as acute promyelocytic leukemia. After waiting for the mouse to develop cancer, the team sequenced the mouse's tumor genome, narrowing in on mutations that seem to be particularly relevant to human forms of APL.
"If we find mutations that occur in mouse models and we see those same mutations, however rare, in human cancers, they are highly likely to be relevant," senior author Timothy Ley, an oncology researcher at Washington University, said in a statement.
Past studies have shown that a fusion oncogene created by a translocation between the promyelocytic leukemia gene PML on chromosome 15 and the retinoic acid receptor alpha gene RARA on chromosome 17 can produce the APL subtype of AML, researchers explained. But other genetic changes involved in the blood cancer are less well understood.
And while sequencing studies of human cancers are providing insights into disease-related mutations, they added, it can be difficult to discern which genetic alterations are involved in cancer pathogenesis, progression, and treatment resistance and which represent background changes.
In an effort to explore this in more detail, Ley and his co-workers slipped the human PML-RARA oncogene into an untranslated gene region on chromosome 14 in inbred mice from the 129/SvJ strain. These mice were subsequently backcrossed with mice from the B6 strain.
The gene fusion is expressed in mouse myeloid progenitor cells, they explained, and typically causes leukemia in the animals after seven to 18 months.
Once the mice developed APL after nearly a year, the team used massively parallel sequencing with the Illumina GAIIx to get some 15.5 times haploid coverage of the APL tumor genome from a male mouse.
Because comparisons between this genome and the B6/Jackson mouse reference genome uncovered more than 110,000 heterozygous and homozygous SNPs, the team decided to generate additional reference sequence using pooled DNA from six wild type male 129/SvJ strain mice.
That reference genome, which covered the wild type 129/SvJ haploid genome nearly 29 times, contained almost five million SNPs compared to the B6 reference genome, they noted.
After tossing out variants found in the 129/SvJ reference genome and other apparent background changes, the researchers were able to narrow in on a few dozen suspicious SNPs in the mouse APL genome that they subsequently assessed by Sanger sequencing.
Of these, three somatic, non-synonymous mutations affecting Jak1, Jarid2, and Capns2 genes seemed to be tumor-specific, they reported, and one — the Jak1 mutation — was recurrent, turning up in half a dozen of the 89 APL mouse tumors that the team screened.
Subsequent experiments indicated that Jak1 mutations can ramp up APL progression, since mice carrying both the human PML-RARA gene fusion and a mutated version of the human JAK1 gene got the disease within a month instead of a year, they found.
Alterations in the same gene have also turned up in recent studies of human APL and another form of leukemia called acute lymphoblastic leukemia, fueling speculation that Jak1 gene changes may be driver mutations for APL and related cancers.
"By establishing that the mutation occurs in other mouse tumor samples and in patients with leukemia, that tells us this mutation is a driver," Ley said in a statement, "it almost certainly is relevant for the progression of cancer."
Moreover, the researchers' findings in mice hint that Jak1 inhibitors — compounds currently being tested for the treatment of other types of cancer — could curb APL development in 10 animals tested.
Their analyses of the APL tumor genome also uncovered a somatic deletion in the lysine-specific demethylase 6A gene Kdm6a. Similar deletions were detected in three of the 14 other APL mouse tumors that they screened using NimbleGen comparative genomic hybridization custom-tiling arrays and in one of 150 human AML tumors tested.
Given their findings so far, those involved in the study expressed optimism about using genomic information from mouse models to provide meaningful information about human disease.
"[T]his mouse model is remarkably similar to the human disease," Ley said. "This gives us a new way to use whole-genome sequencing to rapidly identify the most relevant mutations in human cancers."
"We expect this to expedite our ability to determine whether mutations in patients are important for disease progression," co-author Richard Wilson added. "If we find the same mutations in human cancers and in a mouse model of the disease, then we know they are likely to be relevant, even if we've only seen the mutations in a small fraction of patients."