The taming of the shrew

Abstract

Transposons are mobile genetic elements that can be harmful for the host when mobilized. However, they are also genomic reservoirs for novel genes that can be evolutionarily beneficial. There are many examples of domesticated transposases, which play important roles in the hosts. In most cases domesticated transposases have lost their endonuclease activities and the hosts utilize their DNA-binding properties. However, some other domesticated transposases perform endonuclease activities for host biological processes. Because such a catalytically active transposase is potentially harmful for the integrity of the host genome, its activity should be tightly regulated. The catalytically active domesticated piggyBac transposase Tpb2p catalyzes programmed DNA elimination in the ciliate Tetrahymena. Here, we discuss the regulatory mechanism that prevents unintended DNA cleavage by Tpb2p and compare it to another well-studied catalytically active domesticated transposase, the RAG recombinase in V(D)J recombination. The regulatory mechanisms involve the temporarily regulated expression of the transposases, the target sequence preference of the endonuclease, and the recruitment of the transposases to locally restricted chromatin environments.

Transposases are enzymes encoded by DNA transposons. They catalyze the excision of the transposon at one genomic locus and its integration at a distant position. Both of these steps can be deleterious to the host. The excision step causes DNA double-strand breaks that are harmful if not repaired properly¹ and their integration may disrupt host genes. Nonetheless, the harmfulness of transposases is only one side of the coin, and it is becoming clear that many transposases have been domesticated for the benefit of their hosts.² In most cases, the host benefits from the DNA- or chromatin-interaction ability of the transposase. Examples include Daysleeper in Arabidopsis³ and the widespread eukaryotic Cenp-b.⁴ In contrast, there are only a few examples in which the domesticated transposase is still catalytically active. Because the endonuclease activity of transposases represents a great threat to the host, tight regulatory mechanisms are necessary to prevent unintended genome rearrangements. The most famous example of a domesticated transposase with catalytic activity is the RAG (recombination-activated gene) recombinase from jawed vertebrates.⁵ The RAG enzyme complex containing RAG1 and RAG2 catalyzes the double-strand break formation in V(D)J recombination during lymphocyte development. Recently, we and others reported further examples of catalytically active transposases from the piggyBac family that play essential roles in programmed DNA elimination in the ciliated protists Tetrahymena and Paramecium.^6,7 Here, we discuss how the activity of the Tetrahymena piggyBac transposase 2 (Tpb2p) is regulated to catalyze programmed DNA elimination without causing harmful genome rearrangement events through a comparison of Tpb2p and the RAG recombinase.

Most ciliates, such as Tetrahymena thermophila, have two structurally and functionally distinct nuclei in a single cell (Fig. 1A). The large, polyploid macronucleus (Mac) provides the cell with all RNAs necessary for vegetative growth, whereas the small, diploid micronucleus (Mic) is transcriptionally silenced during vegetative growth. During sexual reproduction, the Mic acts as the germline nucleus and produces both the Mac and the Mic of the progeny⁸ (Fig. 1A). During new Mac development, two types of programmed genome rearrangement events occur in Tetrahymena. The first is chromosome breakage, by which five germline-derived chromosomes are fragmented to approximately 200 Mac chromosomes.⁹ The breaks are repaired by de novo telomere addition. The second type of genome restructuring is the elimination of more than 8000 internal eliminated sequences (IESs), followed by ligation of their flanking sequences (Fig. 1B). The elimination of IESs occurs reproducibly: invariable sets of IESs are eliminated from the Mac, and the majority of their boundaries vary by only a few to several base pairs.¹⁰ IESs are thought to be transposon remnants, with some of them still having the potential to be activated.¹¹ Therefore, IES excision is a process that removes harmful genetic elements from the transcriptionally active Mac. Moreover, it is essential for the streamlining of the somatic genome because it removes some IESs from gene-coding regions, creating functional genes.¹²

Figure 1. Conjugation and DNA elimination in Tetrahymena. (A) During conjugation, the Mic gives rise to the new Mic and the new Mac. DNA elimination occurs in the exconjugant, which has 2 Mics and 2 new Macs (middle); red = Mic, purple = Mac. (B) DNA elimination removes IESs from the developing Mac genome, and the flanking regions are ligated. (C) DNA elimination occurs in specific nuclear compartments, heterochromatin bodies, which contain histone H3 K9me3/K27me3, Pdd1p, and Tpb2p. We propose that this compartmentalization restricts Tpb2p’s action (see text). (D) Localization of Tpb2p, Pdd1p, and H3K9me3 in heterochromatin bodies, as shown by immunofluorescence staining. Asterisk = new Mac, Arrow = Mic.

We recently showed that the endonuclease activity of the domesticated piggyBac transposase Tpb2p is required for IES elimination in Tetrahymena,¹³ which strongly suggests that Tpb2p is the enzyme responsible for excising IESs. Indeed, our in vitro study showed a preferred activity of Tpb2p at the boundary sequences of well-studied R-IES¹³ and other elements in vitro (unpublished). Different IESs in Tetrahymena do not share any detectable common sequences, neither within themselves nor in their flanking regions. Therefore, it is expected that, in sharp contrast to canonical DNA transposases, the enzyme that excises IESs should have promiscuous substrate specificity. Consistent with this idea, we found that only the 2nd and 3rd bases downstream of the cleavage site are crucial for the precision of the cleavage at the R-IES boundary both in vitro and in vivo.¹³ However, this promiscuous substrate specificity of Tpb2p raises a question: how is the precise boundary of IESs specified? We believe one of the key components to solve this issue is the heterochromatin formation that occurs at IESs.

The IES elimination process can be divided into four distinct steps: i) small RNA-directed heterochromatin formation, ii) assembly of heterochromatin bodies (Fig. 1C), iii) DNA excision, and iv) ligation of the flanking DNA by a non-homologous end-joining (NHEJ) pathway.¹⁴ Heterochromatin formation includes the de novo establishment of tri-methylation on lysine 9 and lysine 27 of histone H3 (H3K9me3 and K27me3), specifically at IESs^15,16 (Fig. 1C). The chromodomain protein Pdd1p subsequently binds one or both of the histone modifications and promotes the assembly of IESs into sub-nuclear foci (heterochromatin bodies)¹⁷(Fig. 1D). We have shown that Tpb2p is required for the assembly of heterochromatin bodies and DNA excision in Tetrahymena.^7,13 The fact that endonucleolytically inactive Tpb2p can only promote the former step supports the longstanding idea that DNA excision occurs in heterochromatin bodies. The cysteine-rich domain (CRD) of Tpb2p, which might fold into a PHD finger, was found to preferentially interact with peptides having the histone H3 N-terminal tail sequence when they were tri-methylated at either position lysine 9 (K9me3) or lysine 27 (K27me3) in vitro.¹³ Furthermore, the mutation of two cysteines in the CRD led to the inability of Tpb2p to promote heterochromatin body formation in vivo.¹³ Therefore, Tpb2p is most probably recruited to IESs via interactions between H3K9/K27me3 and CRD, and this interaction is necessary for the assembly of heterochromatin bodies. We speculate that heterochromatin formation at IESs restricts the endonucleolytic action of Tpb2p in two ways: by specifically recruiting Tpb2p to the IESs and by preventing Tpb2p from accessing the middle region of IESs. Tpb2p may only be able to cleave euchromatic regions adjacent to heterochromatinized IESs. A potential regulation could also be mediated by cis-regulatory sequences that are found close to some IESs.^18-22 They might be able to position Tpb2p in a reproducible manner by localizing nucleosomes at precise positions and/or by arranging IESs in distinct chromatin compartments.

Although the interaction of Tpb2p with heterochromatin may restrict its action at the edges of IESs, the promiscuous substrate specificity of Tpb2p could still potentially cause undesirable DNA cleavage at non-target genomic locations. We currently do not know how such a deleterious action of Tpb2p is inhibited in Tetrahymena. To consider how Tpb2p might be regulated, it is worth recalling what is known about another catalytically active domesticated transposase, RAG recombinase. Here, we propose a potential regulatory mechanism of Tpb2p that was deduced from a comparison of Tpb2p and RAG recombinase (Fig. 2).

Figure 2. Comparison of the regulatory mechanisms of RAG recombinase and Tpb2p.

First, these potentially harmful enzymes are only expressed in a particular developmental process with a short time window. Tpb2p expression is limited to the stage of new Mac development when DNA elimination occurs.⁷ The RAG proteins are only expressed during lymphocyte development and are quickly degraded after productive V(D)J recombination by the proteasome.^23,24 Therefore, it would be interesting to study whether there is a similar active protein degradation mechanism for Tpb2p in Tetrahymena.

Second, in both DNA elimination and V(D)J recombination, the chromatin environment of the target sequences regulates the interaction of the recombinase with the target sequences. As discussed above, Tpb2p is likely tethered to IESs through heterochromatin-specific histone modifications. In contrast, the heterochromatic environment inhibits the access of RAG recombinase to recombination signal sequences (RSS),²⁵ a consensus sequence at V, D and J genes. The recombinase can access an RSS only when the locus is activated for transcription, accompanied by euchromatic modifications, including H3K4me3.²⁶ Moreover, RAG’s endonuclease is activated upon binding H3K4me3.²⁶ It has been proposed that the binding of RAG2 via its PHD finger to H3K4me3 releases an auto-inhibition of the RAG1 protein.²⁷ Analogous to this mechanism, the endonuclease activity of Tpb2p may be auto-inhibited and specifically activated upon its interaction with heterochromatin. Alternatively, a protection mechanism that repels Tpb2p or inhibits its enzymatic activity in non-IES (and thus euchromatic) regions could be present. Further study is necessary to reveal whether there is such an additional layer of regulation of Tpb2p’s endonuclease activity.

Lastly, both DNA elimination and V(D)J recombination likely occur in specialized nuclear compartments. In Tetrahymena, DNA elimination is believed to occur in the heterochromatin bodies¹³ and in mammals, V(D)J recombination is thought to occur in recombination centers.²⁸ In addition, DNA looping helps to bring separated gene segments in closer proximity for recombination.²⁹ This compartmentalization may inhibit the spontaneous action of domesticated transposases outside of their targets. As it is not known how IESs are assembled into heterochromatin bodies in Tetrahymena, it would be interesting to investigate the physical configuration of the chromosomes in the developing Mac using a genome-wide chromosome conformation capture (3C) technique.

Although our knowledge of the regulation of Tpb2p is still far behind our knowledge of the RAG recombinase, the study of transposases in DNA elimination in ciliates may shed light on the evolutionary process of transposon domestication, which is one of the black boxes of the biology of domesticated transposases. Tetrahymena and its closely related ciliate Paramecium use piggyBac transposases that have undergone the domestication process for DNA elimination. These transposases are no longer encoded within the context of a transposon but are single-copy genes in the macronuclear genome.³⁰ In contrast, the more distantly related ciliate Oxytricha eliminates DNA with the help of Tc1/mariner-like transposases encoded by the germline-limited TBE transposons.³¹ Therefore, it is possible that ciliates preserve traceable “fossil” records of the transposase domestication process. The system in Oxytricha, in which transposons participate in their own excision from the somatic genome, could represent a very ancient system of DNA elimination. During the evolution of the Tetrahymena/Paramecium clade, a common ancestor could have domesticated a piggyBac transposase that was then modified in such a way that it could excise all IESs. Because ciliates are an evolutionarily very diverged group of eukaryotes, and their mode of DNA elimination varies in different species, it would be interesting to search for “missing link” species that might have both systems in parallel. We believe that further biochemical studies of Tpb2p in the model ciliate Tetrahymena and the comparison of DNA elimination pathways in different ciliate species will help in illuminating the “taming of the shrew” process of transposase domestication.

Vogt A, Mochizuki K. A domesticated PiggyBac transposase interacts with heterochromatin and catalyzes reproducible DNA elimination in Tetrahymena. PLoS Genet. 2013;9:e1004032. doi: 10.1371/journal.pgen.1004032.