By Monica Heger
This article has been updated from a version posted April 14 to include outside comment.
Researchers at the Genome Center at Washington University have sequenced the primary tumor, a metastatic tumor, and a xenograft tumor from a patient with basal-like breast cancer to gain surprising insights into the nature of metastasis and xenograft tumors.
In a description of the work published last week in Nature, the team notes that 26 out of 48 confirmed somatic point mutations and indels in all three tumor samples were enriched in one or both of the metastatic and xenograft tumors, despite the fact that cells from the xenograft tumor — grown in a mouse from implanted cells from the patient's primary tumor — were harvested before the patient had received any treatment, and the metastatic tumor developed eight months after the patient had undergone both chemotherapy treatment and radiation therapy.
"That was a little unsuspected," said Elaine Mardis, co-director of the Genome Center and a senior author on the paper. "Xenografts are thought to be a good way to study primary tumors, but maybe they are better for studying the process of metastasis."
Mardis and her team sequenced a "quartet" of DNA samples from a 44-year-old African-American woman with basal-like breast cancer, including DNA from peripheral blood, the primary tumor, a metastatic tumor from the brain, and a mouse xenograft derived from the primary tumor.
They used a paired-end sequencing strategy with read lengths of 75 base pairs on the Illumina Genome Analyzer and generated between 111 and 150 gigabases of sequence data for each sample, achieving haploid coverage of between 23- and 39-fold.
The team confirmed 50 point mutations and small indels in the coding sequences, splice sites, and RNA genes, of the three tumor samples, 48 of which were present in all three tumors. Of the 48 shared mutations, 26 were enriched in the metastatic and/or xenograft tumors, 20 were present at similar frequencies in all three tumors, and two were enriched in the primary tumor.
The team also found two de novo mutations in the metastatic tumor that were not present in the primary or xenograft tumors, but "because the xenograft line, without these two mutations, exhibits metastatic lesions in ovarian, lymphoid and subcutaneous tissue, it is unlikely that these mutated genes are essential to the metastatic process," the authors wrote.
Of the mutations common to all three tumors, the researchers identified several that had been previously associated with cancer. In particular, they found a missense mutation in a gene that has been associated with poor survival in breast cancer, and has also been found to be frequently deleted in colorectal adenocarcinoma and head and neck carcinoma.
Another interesting finding, said Mardis, was a mutation in a gene that had been previously linked to metastasis in breast cancer. The team detected that mutant allele at 28-percent frequency in the metastatic tumor, but only at 7- and 10-percent frequency in the primary and xenograft tumors. The gene where the mutation resides has "been shown to modulate mesenchymal stem cell differentiation," the authors wrote. And overexpression of the gene has "also been implicated in promoting the migration, invasion and tumorigenesis of breast cancer cells."
Aside from point mutations and indels, the team found that the metastatic and xenograft tumors had more copy number alterations than the primary tumor, and that the copy number alterations observed in the primary tumor were conserved in the metastatic and xenograft tumors.
One deletion in particular that affected all three tumors has been previously associated with increased tumorigenic characteristics, said Mardis.
Last year, a group from the BC Cancer Agency in Vancouver published a study comparing the genomes of a primary tumor from lobular breast cancer to a metastatic tumor from the same patient (IS 10/13/2009). In that study, the researchers found 32 somatic point mutations in the metastatic tumor, 19 of which were undetectable in the primary tumor.
Sam Aparicio, an author of the study and a chair of breast cancer research at the BC Cancer Agency, told In Sequence last week that the results of the Wash U study more or less confirmed the prior study, showing that tumors evolve.
He emphasized, however, that the two studies looked at different subtypes of cancer, and also the interval between the primary tumor and the metastatic tumor was much longer in the BC study — nine years, as opposed to eight months — which could explain why his group found so many more new mutations in the metastatic tumor than the Wash U group.
"These are still two cases, so to generalize the specific findings would be difficult, but what it does tell us is that comparing a tumor before and after [metastatis] is important because you can look at the mutations that changed and that gives you an idea about which mutations may have been more important in the metastatic process," Aparicio said.
He added that an important aspect of the Wash U study was the characterization of the xenograft tumor. "The striking thing is that the tumor appears to have evolved in a way that makes it look more like a metastatic tumor than a primary tumor," he said. "A lot of work is done studying [xenograft] breast tumors in immuno-deficient mice. But now it's evident that putting the tumors in the mice changes the nature of the tumor — it's evolved into something else. That makes it more complicated to interpret those studies."
n order to understand the importance of the mutations that the Wash U team characterized, many more samples will need to be sequenced, said Mardis. She said that while the team was able to identify several mutations affecting genes that had been previously linked to cancer, the vast majority of the mutations have not yet been characterized. "It will take more samples before we really understand these mutations," she said. "What we're trying to do is enrich the clue set for understanding tumors, and if nothing else, I think there are a number of really good clues here."
She added that sequencing primary, metatstatic, and xenograft tumors would be important for understanding the process of metastasis. For instance, she said, the fact that a subset of mutations was enriched in the metastatic tumor supports the hypothesis that certain cells break off from the primary tumor, settle elsewhere, and grow into a second tumor. Also, the comparison of the metastatic and xenograft tumors supports the idea that the genetic makeup of the metastatic tumor was not affected by the chemotherapy and radiation treatments the patient received before that tumor developed.
Mardis added that she and her colleagues are now in the process of analyzing primary, metastatic, and xenograft tumors from other patients with basal-like breast cancer, but with metastatic tumors in different body sites.