FttA is a CPSF73 homologue that terminates transcription in Archaea

Abstract

Regulated gene expression is largely achieved by controlling the activities of essential, multisubunit RNA polymerase transcription elongation complexes (TECs). The extreme stability required of TECs to processively transcribe large genomic regions necessitates robust mechanisms to terminate transcription. Efficient transcription termination is particularly critical for gene-dense bacterial and archaeal genomes^1-3 in which continued transcription would necessarily transcribe immediately adjacent genes and result in conflicts between the transcription and replication apparatuses^4-6; the coupling of transcription and translation^7,8 would permit the loading of ribosomes onto aberrant transcripts. Only select sequences or transcription termination factors can disrupt the otherwise extremely stable TEC and we demonstrate that one of the last universally conserved archaeal proteins with unknown biological function is the Factor that terminates transcription in Archaea (FttA). FttA resolves the dichotomy of a prokaryotic gene structure (operons and polarity) and eukaryotic molecular homology (general transcription apparatus) that is observed in Archaea. This missing link between prokaryotic and eukaryotic transcription regulation provides the most parsimonious link to the evolution of the processing activities involved in RNA 3'-end formation in Eukarya.

Extended Data Figure 1

Extended Data Figure 2

Extended Data Figure 3

Extended Data Figure 4

Extended Data Figure 5

Extended Data Figure 6

Extended Data Figure 7

Extended Data Figure 8

Figure 1.. FttA is a bona fide archaeal transcription termination factor.

a. Transcripts within intact TECs are retained in pellet (P) fractions; transcripts released from terminated TECs partition into the supernatant (S). Radiolabeled transcripts within starting material (SM) TECs₊₁₂₅ and mock treated TECs₊₁₂₅ are retained in pellet fractions (lanes 1-4), whereas FttA^WT addition results in cleavage of nascent transcripts and termination of most TECs (lanes 5-6). A catalytically deficient FttA variant (FttA^H255A) abrogates cleavage and RNA release (lanes 7-8). Lane M contains ³²P-labeled ssDNA markers. b. FttA-mediated termination is distinct from RNase treatment of intact TECs. TECs₊₁₂₅ (SM, lane 1) are resistant to repeated washes and readily resume elongation upon NTP addition to generate +225 nt transcripts (lanes 2-5). Dashed boxes and arrows denote +125 transcripts that are elongated to +225 nt transcripts; the specific activity of +225 transcripts can be increased by addition of additional ³²P-α-UTP during resumed elongation. RNase I_f digestion of nascent transcripts associated with washed TECs₊₁₂₅ results in degradation of the nascent transcript to just ~20-30 nts, but TECs with shortened transcripts remain associated with the DNA and survive repeated washing (lanes 10-11). TECs_~+25-30 resultant from RNase I_f treatment of TECs₊₁₂₅ readily resume elongation upon NTP addition to generate ~+125 nt full-length transcripts (lanes 12-13). Dashed boxes and arrows denote ~+25 transcripts that are elongated to ~+125 nt transcripts; the specific activity of ~+125 transcripts can be increased by addition of ³²P-α-UTP during resumed elongation. FttA addition to TECs+125 disrupts most TECs with nascent transcript cleavage (lanes 6-9), results in release of most TECs from the template and cleaved transcripts cannot be extended by NTP addition (lanes 8-9). c and d. Diagrams of the fate of TECs₊₁₂₅ following RNase I_f and FttA treatment, respectively. e. FttA releases RNAP from the DNA template into solution confirming dissociation of the TEC and bona fide FttA-mediated transcription termination. RNAP is tracked and quantified by Western blots (n = 3 independent replicates) with anti-HA antibodies that recognize the modified RpoL-subunit. f. FttA is not reliant on NTP hydrolysis to inactivate TECs, cleave nascent transcripts and terminate transcription. For panels a, b f: Similar results were observed in 4 independent experiments.

Figure 2.. FttA-mediated termination shares mechanistic requirements of rho-mediated bacterial transcription termination.

a. Promoter-directed transcription of biotinylated templates encoding G-less or C-less cassettes permits formation of TECs with increasing length A-, C-, and U-rich, or A-, G-, and U-rich nascent transcripts, respectively. FL = full-length; all templates permit elongation for 100 nts beyond the G- or C-less cassette. b. TECs remain stably associated and transcripts are primarily recovered in the pellet (P) fraction in the absence (−) of FttA. When FttA is present (+), transcripts are cleaved and primarily recovered in the supernatant (S) fraction. Cleavage releases ~20 – 30 nt shorter transcripts (boxed). The left-most lane contains ³²P-labeled ssDNA markers. c. Addition of Spt4-Spt5 largely abrogates the RNA sequence-requirements of FttA-mediated transcription termination. T= total reaction = P+S. The left-most lane contains ³²P-labeled ssDNA markers. d. Transcript release was quantified with and without FttA addition for TECs with increasing length transcripts on G-less (pink/salmon) and C-less cassettes (mint/green), with and without Spt4-Spt5 addition for TECs₊₁₂₅ formed on G- and C-less cassettes. Error bars were calculated as standard deviation from the mean (n = 3 independent experiments). For panels b, c: Similar results were observed in 3 independent experiments.

Figure 3.. FttA-mediated transcription termination is competitive with transcription elongation.

a and b. Washed, NTP-deprived TECs₊₁₂₅ were assembled on biotinylated templates with a +125 nt G-less cassette. Resumed elongation upon differential [NTP] addition permits transcription to generate +225 nt transcripts, albeit at different rates. c. FttA readily terminates stalled or slowly elongating TECs (lanes 17-24) and FttA-mediated termination becomes competitive with transcription elongation even at high [NTP] in the presence of Spt4-Spt5 (lanes 25-32). Similar results were observed in 3 independent experiments.

Figure 4.. Inhibition of FttA activity abolishes transcription termination in vitro and reduced FttA-expression or activity alters steady-state RNA 3’-termini in vivo.

a. and b. TECs₊₁₂₅ (lane 3) resume elongation upon NTP addition to generate +225 nt transcripts −/+ 25 mM DPA (lanes 4-11). FttA addition results in transcript cleavage and release of most TECs to the supernatant - DPA (lanes 12-13), inhibiting resumed elongation upon NTP addition (lanes 14-15). Pre-incubation of FttA with 25 mM DPA inhibits FttA-mediated termination (lanes 16-17), permitting TECs₊₁₂₅ to resume elongation (lanes 18-19). Lanes M1 and M2 contain ³²P-labeled 10- and 50-nt ssDNA markers, respectively. Similar results were observed in 5 independent experiments. c. Inhibition of metallo-beta-lactamase/beta-CASP protein activity impairs growth of T. kodakarensis. A mid-log culture of T. kodakarensis strain TS559 was split into nine cultures, with three biological replicates exposed to 0 mM (peach), 12.5 mM (green) or 25 mM DPA (purple). d, e and f. RNAs recovered one-hour post DPA-addition to cultures of TS559, or from cultures of IR5 grown in the absence of NaF display altered 3'-termini. TRIzol extracted RNAs were reverse transcribed with primers complementary to nascent transcript sequences of TK1146, TK0222 and TK0676 to generate cDNAs that were quantified and normalized to internal controls. Inhibiting FttA-activity with DPA or lowering steady-state FttA levels by riboswitch-mediated controlled expression impacts the abundance of RNAs with extended 3’-termini in vivo. RNA abundance in untreated TS559 cultures (open bars) was set to 1.0, and fold changes in the abundance of amplicons reflecting RNA transcripts with extended 3’-sequences at increasing distances from the translation stop site (purple, green, orange and red) are shown for strain IR5 (solid bars), TS559 + 12.5 mM DPA (wide stripes) and TS559 + 25 mM DPA (narrow stripes). Errors were calculated at a 95% confidence interval with the center value as the mean of 3 biological replicates. g. Maps of the TK1428 locus in parental (TS559), N-terminally tagged (TK1428D) and riboswitch-regulated expression (IR5) strains. h. Western blots demonstrate the reduction in FttA protein levels in strain IR5 upon removal of NaF from the medium; n = 3 independent replicates. Size standards are identified by MW (left).