Cystoviral polymerase complex protein P7 uses its acidic C-terminal tail to regulate the RNA-directed RNA polymerase P2

Abstract

In bacteriophages of the cystovirus family, the polymerase complex (PX) encodes a 75-kDa RNA-directed RNA polymerase (P2) that transcribes the double-stranded RNA genome. Also a constituent of the PX is the essential protein P7 that, in addition to accelerating PX assembly and facilitating genome packaging, plays a regulatory role in transcription. Deletion of P7 from the PX leads to aberrant plus-strand synthesis suggesting its influence on the transcriptase activity of P2. Here, using solution NMR techniques and the P2 and P7 proteins from cystovirus ϕ12, we demonstrate their largely electrostatic interaction in vitro. Chemical shift perturbations on P7 in the presence of P2 suggest that this interaction involves the dynamic C-terminal tail of P7, more specifically an acidic cluster therein. Patterns of chemical shift changes induced on P2 by the P7 C-terminus resemble those seen in the presence of single-stranded RNA suggesting similarities in binding. This association between P2 and P7 reduces the affinity of the former toward template RNA and results in its decreased activity both in de novo RNA synthesis and in extending a short primer. Given the presence of C-terminal acidic tracts on all cystoviral P7 proteins, the electrostatic nature of the P2/P7 interaction is likely conserved within the family and could constitute a mechanism through which P7 regulates transcription in cystoviruses.

Keywords: NMR; RNA-directed RNA polymerase; cystovirus; polymerase complex; transcription.

Fig. 1

(A) Chemical shift perturbations for the C-terminal resonances of P7, determined from ¹⁵N,¹H HSQC spectra (600 MHz) of ¹⁵N-labeled full-length P7 in the absence or the presence of P2 (P7:P2 molar ratio 1:2). Δδ_amide values, calculated using Eq. (3A), in 0 mM NaCl (green bars), 50 mM NaCl (black bars) and 150 mM NaCl (red bars) are shown. Negative bars (for 0 mM NaCl) represent residues for which the resonances are broadened below the threshold of detection. (B) Δδ_amide values for the C-terminal resonances of full-length wild-type P7 (black), P7^D160N (blue) and P7^D168N (yellow) in the presence of a 2-fold excess of P2. The buffer contained 50 mM NaCl in all cases. (C) The C-terminal sequences of the P7 proteins from the ϕ6, ϕ8, ϕ12, ϕ13 and ϕ2954 cystoviruses are shown. Sequence alignment was performed using the entire sequences of the five P7 proteins (ϕ6:1-161, ABF19537.1; ϕ8:1-175, NP_652745.1; ϕ12:1-169, NP_690822.1; ϕ13:1-156, AF261668_1; ϕ2954:1-177, YP_002600762.1) though only the C-terminal sequences are depicted. While the overall conservation is low, all the sequences are characterized by a large concentration of acidic residues (shown in red letters; basic residues are shown in blue letters) at the C-terminal ends. The P7⁰ and P7⁺ mutants (see inset) were generated by replacing all the acidic residues of C-terminus of ϕ12 P7 by corresponding neutral residues or by positively charged residues, respectively. In addition, peptides encompassing residues 155 through 169 of P7 (P7pep) and P7⁰ (P7pep⁰) were synthesized.

Fig. 2

(A) Residues for which the Δδ_amide values (for a P2:P7pep molar ratio 1:10) belong to the 1σ class (0.024 ppm > Δδ_amide ≥ 0.016 ppm), the 2 σ class (0.031 ppm > Δδ_amide ≥ 0.024 ppm) and the 3 σ class (Δδ_amide ≥ 0.031 ppm) are depicted on the structure of P2 and colored yellow, orange and red respectively. The arrow indicates the mouth of the template tunnel in all cases. (B) Residues for which Δδ_met values (for a P2:P7pep molar ratio 1:8) belong to the 2 σ class (0.017 ppm > Δδ_met ≥ 0.013 ppm) or the 3 σ class (Δδ_met ≥ 0.017 ppm) are depicted as in (A). (C) Surface representation of P2 showing the 2 σ and 3 σ class perturbations in the presence of P7pep (P2:P7pep molar ratio 1:8 as in B, left) or 5’FL-UUUUC-3’ (P2:ssRNA molar ratio 1:4; right). A portion of the front surface has been removed to show the inside of the template tunnel. Note the similarities in the overall pattern of perturbations seen in the two cases. Key residues discussed in the text are labeled. (D) Charge distribution within the template tunnel. The negatively charged surface at the base of the tunnel is created by the catalytic aspartates (D349, D469 and D470).

Fig. 3

(A) Binding isotherms determined by measuring the change in fluorescence anisotropy for two 5-nt non-genomic ssRNA constructs (5’FL-AAGCC-3’, black; 5’FL-UUUUC-3’, red) in the absence (solid lines, filled circles) or the presence (dotted lines, open circles) of P7pep (also see Table S3). Circles represent the experimental data and the lines represent fits to Eq. (2). Also shown as blue circles are the anisotropy values measured for increasing concentrations of P7pep in the absence of P2. (B) ¹³C, ¹H HMQC spectra (600 MHz) of ILV, U-[¹⁵N,²H] P2 (black) in the presence of 5’-UUUUC-3’ (P2:ssRNA molar ratio 1:2; green) and after the addition of P7pep (P2:ssRNA:P7pep molar ratio 1:2:12; orange). As shown in the insets, positions of the resonances in the case of P2+ssRNA+P7pep seem to shift towards those for P2+P7pep.

Fig. 4

(A) Inhibition of de novo RNA synthesis on the U-template using only GTP in the presence of increasing concentrations of either P7pep or P7pep⁰. (B) Inhibition of the elongation of a pGpG primer on the U-template using CTP in the presence of increasing concentrations of either P7pep or P7pep⁰. Circles represent the experimental data and the lines represent fits to Eq. (5).