RNA topoisomerase is prevalent in all domains of life and associates with polyribosomes in animals

Abstract

DNA Topoisomerases are essential to resolve topological problems during DNA metabolism in all species. However, the prevalence and function of RNA topoisomerases remain uncertain. Here, we show that RNA topoisomerase activity is prevalent in Type IA topoisomerases from bacteria, archaea, and eukarya. Moreover, this activity always requires the conserved Type IA core domains and the same catalytic residue used in DNA topoisomerase reaction; however, it does not absolutely require the non-conserved carboxyl-terminal domain (CTD), which is necessary for relaxation reactions of supercoiled DNA. The RNA topoisomerase activity of human Top3β differs from that of Escherichia coli topoisomerase I in that the former but not the latter requires the CTD, indicating that topoisomerases have developed distinct mechanisms during evolution to catalyze RNA topoisomerase reactions. Notably, Top3β proteins from several animals associate with polyribosomes, which are units of mRNA translation, whereas the Top3 homologs from E. coli and yeast lack the association. The Top3β-polyribosome association requires TDRD3, which directly interacts with Top3β and is present in animals but not bacteria or yeast. We propose that RNA topoisomerases arose in the early RNA world, and that they are retained through all domains of DNA-based life, where they mediate mRNA translation as part of polyribosomes in animals.

Figure 1.

E. coli Topoisomerase I has RNA topoisomerase activity that depends on the conserved Type IA core domains but not the non-conserved CTD. (A) Schematic representation of RNA topoisomerase assay (4). A 128-base long synthetic circular RNA contains two pairs of complementary region (Red and blue) separated by single-stranded spacer region (black). The circular substrate is converted to knot after strand passage reaction where the two pairs of complementary regions form double helices. (B) A silver-stained SDS gel showing purified recombinant wild-type (wt) or Y319F mutant EcoTop1. (C) Schematic representation of EcoTop1 wildtype and various mutants; and EcoTop3 and its mutant. The conserved core domains and the non-conserved CTDs, including Zn-fingers (orange boxes), are indicated. Their RNA topoisomerase activity was shown on the right. Arrows show the location of point mutation in the respective domain. (D) Autoradiograph from RNA topoisomerase assay showing that EcoTop1, but not its catalytic mutant Y319F, has RNA topoisomerase activity. The reaction mixture contains increasing concentrations (2.5, 5, 10 and 20 nM) of proteins. The percentage of the knot product in total RNA (circle and knot) is shown below each lane. (E, F) RNA topoisomerase assay showing wild type EcoTop1, but not its mutants as indicated, has RNA topoisomerase activity. The reaction mixture contains increasing concentrations (2.5, 5, 10 and 20 nM) of proteins. 4 nM of humTop3β was used as positive control. (G) RNA topoisomerase assay showing that the CTD-deletion mutant of EcoTop1, EcoTop67, but not its point mutant, D111N, has RNA topoisomerase activity. The reaction mixture contains increasing concentration (7.5, 15, 30,and 60 nM) of proteins. 10 nM of full-length EcoTop1 was used as positive control.

Figure 2.

RNA topoisomerase activity is present in some but not all bacterial Type IA topoisomerases. (A) Schematic representation of RNA topoisomerase representatives from bacteria. The conserved core domains and the non-conserved CTDs, including a Zn-finger (an orange box), are indicated. Their RNA topoisomerase activity was summarized on the right. (B) Silver-stained SDS gels showing purified recombinant Thermotoga maritima Top1 (TmaTop1), Mycobacterium tuberculosis Top1 (MtbTop1), and Mycobacterium smegmatis Top1 (MsTop1). (C) Autoradiograph from RNA topoisomerase assay showing TmaTop1 can convert the circular RNA substrate into a knot. The reaction mixtures contain increasing concentrations (2, 4, and 8 nM) of TmaTop1 at 37°C or 50°C. (D) An autoradiograph from RNA topoisomerase assay shows that MtbTop1 and MsTop1 (5, 10 and 20 nM) has no detectable RNA topoisomerase activity. 4 nM humTop3β was used as a positive control.

Figure 3.

RNA topoisomerase activity is present in several archaeal Type IA topoisomerases. (A) Schematic representation of Nanoarchaeum equitance topoisomerase 3 (NeqTop3) and Sulfolobus solfataricus topoisomerase 3 (SsoTop3). The conserved core domains and the non-conserved CTDs, including a Zn-finger (an orange box), are indicated. Their RNA topoisomerase activity was shown on the right. (B, C) A silver-stained SDS gel showing purified recombinant wt NeqTop3, its catalytic mutant Y293F and SsoTop3. (D, E) RNA topoisomerase assay showing that two Type IA enzymes from archaea, NeqTop3 (D) and (SsoTop3) (E), can convert the circular RNA substrate into knot, whereas NeqTop3-Y293F mutant lacked the activity. The reactions include increasing concentrations (0.62, 1.25, 2.5, 5, 10 and 20 nM) of NeqTop3-wildtype and its catalytic mutant Y293F; and SsoTop3 (1.25, 2.5, 5.0 and 10 nM) at 50°C. The percentage of the knot product in total RNA (circle and knot) was indicated below each lane. Four nM humTop3β was as positive control.

Figure 4.

The Type IA topoisomerase from yeast Saccharomyces cerevisiae has RNA topoisomerase activity. (A) Schematic representation of Top3 of Saccharomyces cerevisiae (yTop3) and its partner, Rmi1. The conserved core domains and the non-conserved CTDs of yTop3 are indicated. The OB fold of Rmi1 is indicated in dark blue color. The RNA topoisomerase activity was shown on the right. (B) Silver-stained SDS gels showing purified recombinant MBP-yTop3, its catalytic mutant Y356F, and 6-Histidine tagged wildtype yTop3 (His-yTop3). (C) RNA topoisomerase assay showing that MBP-yTop3 but not its catalytic mutant Y356F, has RNA topoisomerase activity. The reaction mixture contains increasing concentrations (2.5, 5, 10 and 20 nM) of wildtype or Y356F mutant of MBP-yTop3. (D) RNA topoisomerase assay showing that yTop3-Rmi1 complex has similar activity as yTop3 alone. The reaction mixture contains increasing concentration (0.8, 1.6, 3.1. 6.2, 12.5, 25 and 50 nM) of his-tagged yTop3 (left) or yTop3-Rmi1 complex (right). At 50 nM concentration, His-tagged yTop3 proteins stably bound the RNA substrate to produce a gel-shift band with distinct mobility compared to knot. The gel-shift band can also be distinguished from knot by phenol-chloroform treatment, which abolishes the gel-shift band but does not affect the band corresponding to an RNA knot (data not shown). 4 nM humTop3β was used as a positive control.

Figure 5.

The RNA topoisomerase Top3β associate with polyribosomes in Drosophila and chicken cells. (A) Immunoblotting analysis of sucrose fractions of Drosophila S2 cell extracts shows co-sedimentation of Top3β, TDRD3 and FMRP with polyribosomes. Cytoplasmic extracts of S2 cells were prepared and treated with and without EDTA, which is known to disrupt polyribosomes. The absence or presence of EDTA is indicated on the left. Lysates were centrifuged on a 15–60% (w/w) linear sucrose gradient. The presence of ribosomal subunits, mono and polyribosomes was monitored at 254-nm absorption (Supplementary Figure S3A), and indicated as the bottom of the blots. Notably, the co-sedimentation of the above proteins with polyribosome fractions was strongly reduced by EDTA treatment, suggesting that these proteins associate with polyribosomes. (B) Immunoblotting analysis of sucrose fractions of chicken DT40 cell lysates shows that chicken Top3β, TDRD3 and FMRP associate with polyribosomes. RPS6, a small ribosomal subunit, was included as a marker for ribosomes. β-actin was included as a negative control, which was absent in polyribosome fractions. The asterisk indicates a polypeptide cross-reactive with RSP6 antibody. The red boxes highlight polyribosome fractions in which the indicated proteins display differences or no changes.

Figure 6.

Top3β homologs from animals require TDRD3 for their association with polyribosomes. (A) Immunoblotting of sucrose gradient fractions of siRNA treated HEK293 cells shows that TDRD3 but not FMRP is required for polyribosome association of Top3β. Cytoplasmic extracts were prepared from the HEK293 cell transfected with a control siRNA oligo, or siRNA against TDRD3, or siRNA targeting FMRP, as indicated on the right. The Top panels of (A) were reproduced from our previous publication (4) for the convenience of readers. (B) Immunoblotting of sucrose gradient fractions of chicken DT40 cell lysate shows that TDRD3 is required for polyribosome association of Top3β but not FMRP. Cytoplasmic extracts were prepared from the wild-type, TDRD3^−/− cells and TDRD3^−/−cells ectopically expressing human TDRD3 cDNA. (C) Immunoblotting of sucrose gradient fractions of wildtype and TDRD3-knockout Drosophila S2 cells shows that TDRD3 is essential for polyribosome association of Top3β but not FMRP. Cytoplasmic extracts were prepared from the wild-type and TDRD3^−/− S2 cells. The red boxes highlight polyribosome fractions in which the indicated proteins display differences or no changes.

Figure 7.

Type IA topoisomerases have evolved from enzymes with dual activities in microorganisms to multi-protein complexes with distinct functions in animals; RNA topoisomerases may originate in the RNA world and are conserved through evolution. (A) Schematic representation of evolution of Type IA topoisomerases in DNA and RNA metabolism. In E. coli, both Type IA enzymes (Top1 and Top3) have dual activities that can catalyze topoisomerase reactions on DNA and RNA. In yeast, the only Type IA enzyme is part of a complex (Top3-Rmi1) that also has dual activities for DNA and RNA. In human, only one of the two Type IA paralogs, Top3β, has RNA topoisomerase activity, whereas Top3α does not. Interestingly, Top3β, but not Top3α, has acquired during evolution a bona fide RNA-binding domain (RGG box) that is required for its RNA topoisomerase activity. Moreover, the two Top3 paralogs comprise two distinct complexes, with the Top3β complex containing a RNA binding protein (FMRP), whereas the Top3α complex containing a DNA helicase (BLM). These data argue that Type IA topoisomerases have evolved into two functional distinct complexes in animals, one for RNA and DNA (Top3β-TDRD3-FMRP), and one for DNA only (Top3α-Rmi1-Rmi2-BLM). (B) A model of origin and evolution of Type IA topoisomerases and their activity for RNA and DNA. It has been postulated that life starts with a pool of self-replicating RNAs; and there exists an RNA world with RNA genome prior to the current DNA world. We propose that Type IA enzymes may originate in the RNA world to solve topological problems during RNA metabolism. When the RNA world evolved and was eventually replaced by the DNA world, many of these enzymes retained their RNA topoisomerase activity while developing a new activity for DNA. This may explain the prevalence of the RNA topoisomerase activity in Type IA enzymes from all three domains.