NEW YORK (GenomeWeb) – Jumping genes are more common in esophageal cancers than previously thought and may have a role in disease development, according to a University of Cambridge research team.
While analyzing whole-genome paired-end sequencing data from esophageal tumors as part of the International Cancer Genome Consortium, the team noticed a class of rearrangements that intrigued them. As several tumors had rearrangement breakpoints at about the same spot, the team suspected that these apparent rearrangements might actually be mobile element insertions.
As Cambridge's Paul Edwards and his colleagues reported in BMC Genomics yesterday, they confirmed their suspicions and found these mobile element insertions appear to be common in esophageal adenocarcinomas, uncovering an average 100 insertions in each tumor, and one tumor with some 700 mobile element insertions.
"These jumping genes play hopscotch across our genetic code in cancer cells more than in normal cells," Edwards said in a statement. "When one of these mobile genetic sequences plants itself in the middle of a gene that controls the cell's growth it radically alters how the cell behaves, which can sometimes cause cancer."
In their paired-end sequencing of esophageal adenocarcinomas, Edwards and his colleagues found that 17 out of 22 tumors in their discovery set had 61 seeming translocation breakpoints, all within a 1.4 kilobase region of chromosome 22, and other tumors similarly shared other translocation breakpoints. This pattern, they said, doesn't resemble that of true recurrent translocations as the breakpoints were clustered too close together.
Instead, they said these breakpoints were likely where L1 mobile elements inserted.
There are a handful of ways in which paired-end read mapping could mistake a mobile element insert for a rearrangement, the researchers said.
The first instance is when an insert includes a unique sequence transduced by the mobile element, and that, they noted, accounted for a number of observed insert clusters, including the one at chromosome 22.
The second scenario is one in which the insert includes a sequence that can be matched to a mobile element that's included in the reference genome. Some 84 rearrangement junctions in nine esophageal tumors could be mapped to the 3' end of an L1 on chromosome 7, the researchers reported.
Lastly, a third type of junction, they noted, was only seen with the Novoalign aligner, and not with the BWA aligner. In this case, reads from within one mobile element insert were mapped to different spots on the reference genome, yielding a split read.
Using a PCR approach, Edwards and his colleagues confirmed 42 out of the 45 tumor inserts they tested were from mobile elements and noted that the verified inserts all followed the expected structure.
To estimate just how many mobile element inserts are in the tumor samples, the researchers scoured their data for tumor-specific poly-A runs to uncover some 5,300 candidate elements, a subset of which they confirmed.
From this, they estimated that the tumor-specific inserts ranged from five to about 700, though with an average 100 inserts per tumor. Overall, 40 of the 43 tumors harbored 10 or more candidate inserts.
Theses insertions, depending on their locations, could activate or inactivate genes, the researchers said. Nearly a third of the elements uncovered by analyzing discordant read pairs were in a gene listed by Ensembl, they further reported. For instance, inserts in the coding sequences of MRPL13, ZAN, SYC1, and ZDHHC14 seemed likely to affect gene function, the researchers said, and inserts in ROBO2, CNBD1, and RSPO2, which have all been linked to cancer before, could be drivers of disease.
Overall, the researchers said their study highlights difficulties in using paired-end sequencing to identify mobile elements and distinguish them from rearrangements, as well as the reverse: mobile element inserts make it challenging to interpret structural rearrangements in the genome.