The college sophomores in my genetics class become visibly uncomfortable during my lectures on transposable elements. “You mean to tell me there are things jumping around my genome?” they ask when I explain that much of their genetic material is made up of DNA segments that can move about their chromosomes. The mobility of transposable elements is understandably disconcerting, because it would be expected to lead to chromosome rearrangements and mutations that could cause disease. In the long term, however, these elements are probably beneficial, serving as a powerful force for genome change and evolution.

Although the long-term view may do little to appease a 20-year-old, my students can take some comfort in the findings of Cam et al.1 reported in this issue (page 431). The authors show that, in the fission yeast Schizosaccharomyces pombe, proteins called CENP-Bs are drafted in to quell transposon activity. What is unusual about this particular choice is that CENP-Bs themselves are derived from transposable elements.

CENP-Bs were first discovered because they bind to repetitive DNA sequences called centromeres2. During cell division, centromeres assemble proteins to serve as anchor points for the spindle fibres that move chromosomes to the daughter cells. Fission yeast has three closely related CENP-Bs that bind to centromeres and that are crucial for centromere function3.

Cam et al. wanted to know whether CENP-Bs also associate with genomic regions other than centromeres. They found that, indeed, CENP-Bs occur at dispersed DNA repeats, particularly at the Tf2 transposable elements that litter the S. pombe genome. The authors show that CENP-Bs bind to Tf2 and recruit enzymes known as histone deacetylases, which turn off Tf2 expression and prevent transposition. The silenced Tf2 elements are then packaged into subdomains of the nucleus called Tf bodies (Fig. 1).

Figure 1: Under the watchful eyes of CENP-B proteins.
figure 1

a, Cam et al.1 show that, as well as binding to centromeric DNA repeats (purple), CENP-Bs (green) bind to interspersed Tf elements (blue). CENP-Bs thus promote clustering of Tf elements into Tf bodies and recruit histone deacetylases (orange) that prevent the transcription of Tf elements. b, In rare instances, such as exposure to oxidative stress, the Tf body disassembles, Tf-element expression is activated, and Tf elements can become mobile and transpose to new chromosomal locations (red). c, Once the stress subsides, newly transposed and preexisting Tf elements are packaged into Tf bodies.

But transposition can on rare occasions be advantageous, particularly in times of stress, as it has the ability to create potentially beneficial genetic variants4. When Cam and colleagues exposed S. pombe cells to oxidative stress, Tf2 elements became unbundled from Tf bodies and were expressed (Fig. 1). Furthermore, in cells lacking certain CENP-Bs, the Tf2-related element Tf1 was mobilized, making new insertions that were then packaged into Tf bodies. These observations indicate that CENP-Bs regulate Tf activity through selective packaging.

The CENP-B proteins are an unexpected choice for combating mobile elements because they originate from transposase5, the enzyme necessary for cutting and pasting DNA during transposition. The transposase progenitor of CENP-Bs belongs to a class of mobile element that is unrelated to Tf2. The authors argue that this unrelatedness suggests a conflict between different classes of mobile element, but this view is probably unnecessarily provocative. As transposases normally bind to DNA, they are suited to being moulded over time into genome regulators. In the case of CENP-B, the vestigial DNA-binding domain of the transposase recognizes centromeric repeats and Tf2 elements. So, whereas CENP-Bs can no longer cut and paste DNA, they can recruit histone deacetylases to regulate Tf2. Other transposase-derived proteins also act as gene regulators. For example, two transcriptional activators used by plants in their response to light are derived from transposases6, and it is speculated that many such transposase-derived proteins serve crucial roles in gene regulation7.

Vertebrates also have CENP-Bs, which, like their S. pombe counterparts, bind to centromeres, and were only recently shown8 to contribute to centromere function. The vertebrate proteins also act on centromeric DNA repeats integrated on chromosomes outside the centromere. At these non-centromeric sites, the repeats become tightly packaged into transcriptionally inert protein–DNA complexes called heterochromatin in a way that is analogous to the action of CENP-Bs on the Tf2 elements of S. pombe. It is not known whether CENP-Bs also regulate the activity of vertebrate transposable elements, or how the different roles of CENP-Bs in centromere formation and heterochromatin assembly are assigned. Nonetheless, this emerging story suggests a general role for CENP-Bs in helping the cell to manage its burden of repetitive DNA.

The packaging of transposable elements into transcriptionally inactive chromatin is not a defensive strategy limited to CENP-Bs. RNA interference also serves as a surveillance mechanism for double-stranded RNAs that result from the transcription of clustered or scrambled arrays of transposable elements9. This process leads to the degradation of messenger RNA transcripts of mobile elements, thus diminishing their activity. RNA interference also assembles mobile elements into heterochromatin, similar to the action of CENP-Bs on centromeric repeats and transposable elements.

Of course, the war between transposable elements and their host genome has been going on for millennia, allowing defensive strategies to assume other roles. The RNA-interference machinery in S. pombe, for example, has only a modest impact on Tf2 expression10, yet it is essential for the formation of heterochromatin at centromeric repeats, and is thus important for centromere function11. Also, a particularly remarkable finding of Cam et al.1 is that CENP-Bs regulate the expression of some genes adjacent to sites of Tf2 insertion. This indicates that CENP-Bs and transposable elements collaborate to create genetic diversity, which could benefit the cell. The relationship between transposable elements and their host cell is made more complex by combative strategies that affect regular cellular functions. Undoubtedly, the future holds more surprises about the conflicts and collaborations between mobile elements and their host cell that contribute to genome function.