Runaway amplification: 800 copies and counting
Massive amplification of genes is a desperate strategy taken by stressed populations adapting to an environment that has become inhospitable. Such amplifications can give an underperforming gene a much-needed boost in productivity simply by increasing its copy number. But counterintuitively, research reported in the May issue of G3 implies these amplifications may arise even in the absence of selective pressure and more frequently than expected.
The discovery was made in haploid Saccharomyces cerevisiae that the researchers weakened by replacing a critical tRNA synthetase gene with a version from a distant relative. This made the yeast replicate sluggishly until, after several generations, a few cells recovered their former vigor. These rejuvenated yeast must have adapted to using the foreign gene—but how? The researchers noticed that some of the revertants possessed enormous, grossly extended variants of chromosome VIII along with a slightly altered, but almost normal-sized copy.
Despite the hazards of major chromosomal rearrangements like these, the revertants thrived. In fact, they used the chromosomal catastrophe to their advantage. The enlarged chromosome VIII was formed by a large-scale amplification of the segment containing the tRNA synthetase gene. Along with the whole-chromosome duplication, this expansion increased the gene’s copy number three- to five-fold—enough to compensate for the suboptimal performance of the tRNA synthetase. As a side-effect, all the other genes in the same segment also increased in copy number by the same amount—except one. Seventy kilobases upstream of the tRNA synthetase, the CUP1 locus was copied an astonishing 800 times.
Strangely, CUP1 isn’t involved with tRNA synthesis. It encodes a metallothionein, a protein that increases resistance to copper salts. Amplification of CUP1 has been observed in the lab, but only in yeast subjected to stress-inducing concentrations of copper salts, which the researchers didn’t include. Even more bizarrely, the amplification in this experiment was more than ten times greater than typically seen in copper-exposed yeast.
To determine how this amplification occurred, the researchers analyzed the early subclones that gave rise to each revertant. They found that chromosome doubling and large-scale amplification of a segment of chromosome VIII could rescue the slow-growth phenotype, and these changes always preceded the increase in CUP1 copy number. The researchers propose that the initial large-scale chromosomal amplification that saved the yeast by increasing the copy number of the tRNA synthetase gene also happened to disrupt the nearby CUP1 locus, inadvertently creating a quasi-palindromic repeat.
There was little sequence homology among the quasi-palindromic repeats in each strain, leading the researchers to believe that each of these junctions was generated by a replication fork reversal that mistakenly joined lagging and leading strands. The fact that every expanded locus contained a single, unique quasi-palindromic unit while the CUP1 number varied suggests that that the CUP1 repeats arose as a chance consequence of the larger-scale amplification, possibly due to interference between neighboring replication bubbles. This mechanism explains how some cells ended up with many copies of CUP1—in one case, almost three orders of magnitude more than normal—even though there was no specific selective pressure acting on CUP1 copy number.
Interestingly, the authors report that the 1002 Yeast Genomes project has also identified natural yeast strains with large CUP1 copy numbers, some of which approach the scale reported in this study. This implies that the group’s findings aren’t merely a curiosity restricted to some quirky lab yeast—the same mechanism may exist in the wild, and amplification without selection may be more common than we know.
Thierry, A., Khanna, V., Dujon, B. (2016). Massive Amplification at an Unselected Locus Accompanies Complex Chromosomal Rearrangements in Yeast.
G3: Genes|Genomes|Genetics, 6(5), 1201–1215. DOI: 10.1534/g3.115.024547