Structural and mechanistic basis for the inhibition of Escherichia coli RNA polymerase by T7 Gp2

Abstract

The T7 phage-encoded small protein Gp2 is a non-DNA-binding transcription factor that interacts with the jaw domain of the Escherichia coli (Ec) RNA polymerase (RNAp) β' subunit and inhibits transcriptionally proficient promoter-complex (RPo) formation. Here, we describe the high-resolution solution structure of the Gp2-Ec β' jaw domain complex and show that Gp2 and DNA compete for binding to the β' jaw domain. We reveal that efficient inhibition of RPo formation by Gp2 requires the amino-terminal σ(70) domain region 1.1 (R1.1), and that Gp2 antagonizes the obligatory movement of R1.1 during RPo formation. We demonstrate that Gp2 inhibits RPo formation not just by steric occlusion of the RNAp-DNA interaction but also through long-range antagonistic effects on RNAp-promoter interactions around the RNAp active center that likely occur due to repositioning of R1.1 by Gp2. The inhibition of Ec RNAp by Gp2 thus defines a previously uncharacterized mechanism by which bacterial transcription is regulated by a viral factor.

Figure 1

RPc and RPo Formation and the Structure of the Gp2-β′ Jaw Fragment Complex (A) Cartoon depiction of RPc and RPo formation at σ⁷⁰-dependent bacterial promoters (the inspiration for the cartoon was taken from Murakami and Darst [2003]). (B) Ribbon representation of the Gp2-β′ jaw fragment complex. (C) The same as (B) but rotated by 90° along the horizontal plane. In (B) and (C), the interface region is enlarged in the insets, and the residues located at the interaction interface are shown as sticks and labeled correspondingly. See also Figure S1.

Figure 2

Interaction of the β′ Jaw Domain with dsDNA (A) Surface representation of the Ec core RNAp model (Opalka et al., 2010) color-coded as in Figure 1A. The boxed region is enlarged and looks at the DNA-binding surface (shown in ribbon representation in insets i–iv). The β′ jaw domain is shown in green as a surface representation and forms part of the DNA-binding face (i). Inset (ii) is as in (i), but showing the path of the dwDNA from the current model of the RPo (Opalka et al., 2010). Highlighted in red are residues T1169, R1174, and M1189, which undergo significant chemical shift changes in β′ jaw fragment in the presence of dsDNA (B) and Gp2 (C). Inset (iii) is as in (ii), but showing the redefined path of the dwDNA in the dwDBC. Inset (iv) is as in (iii), but with the surface representation of Gp2 shown in cyan. Note the lack of steric clash between Gp2 and the β′ insertion 6 domain, which provides further support for our composite model. (B) Overlay of 2D ¹H-¹⁵N HSQC spectra of the β′ jaw fragment with and without dsDNA recorded at pH 6.5, 303 K (see key for details). Peaks with significant chemical shift differences are indicated in red with their residue numbers (T1169, R1174, and M1189). (C) As in B, but showing the 2D ¹H-¹⁵N HSQC spectra of the β′ jaw fragment with dsDNA (i.e., the β′ jaw fragment is ¹⁵N labeled) with or without unlabeled Gp2 (see key for details).

Figure 3

Gp2 Requires R1.1 to Efficiently Inhibit RPo Formation by Eσ⁷⁰ (A) Autoradiograph of 20% (v/v) denaturing urea gels showing the synthesis of the ApApUpU transcript (underlined nucleotides are α³²P labeled) from the lacUV5 promoter by Eσ⁷⁰ (lanes 1 and 2) and Eσ⁷⁰_ΔR1.1 (lanes 3 and 4) in the absence and presence of Gp2. The percentage of ApApUpU transcript synthesized (% A) in the reactions with Gp2 with respect to reactions with no Gp2 is given at the bottom of the gel for each reaction. (B) As above, but showing the synthesis of the ApApUpU transcript in the absence (lane 1) and presence (lanes 2–5) of Gp2 under conditions in which Eσ⁷⁰_ΔR1.1 was preincubated with increasing amounts of isolated R1.1 domain added in trans to the reaction (shown as the ratio of σ⁷⁰_ΔR1.1 to R1.1). For (A) and (B), all data obtained in at least three independent experiments fell within 5% of the % A value shown. See also Figure S2.

Figure 4

Gp2 Requires R1.1 of σ⁷⁰, but Not the Consensus Promoter DNA Sequences, to Fully Inhibit RPo Formation by Eσ⁷⁰ (A) Autoradiograph of a 20% (v/v) denaturing urea gel showing the synthesis of RNA-U from the MS probe by Eσ⁷⁰ and Eσ⁷⁰_ΔR1.1 in the absence (condition I) and presence (conditions II and III) of Gp2. The percentage of RNA-U synthesized (% A) in the reactions with Gp2 with respect to reactions with no Gp2 is given at the bottom of the gel for each reaction. (B) As in (A), except that the reaction was conducted with Eσ⁷⁰_ΔR1.1 in the absence and presence (at ∼8-fold molar excess over σ⁷⁰_ΔR1.1) of isolated R1.1 domain added in trans. For (A) and (B), all data obtained in at least three independent experiments fell within 5% of the % A value shown. See also Figure S3.

Figure 5

Inhibition of RPo Formation by Gp2 Involves Long-Range Antagonistic Effects on Eσ⁷⁰-Promoter Interactions (A) Autoradiograph of a 4% (v/v) native polyacrylamide gel comparing binding of Eσ⁷⁰ to the +20/+20 (lanes 1 and 2) and −7/−7 (lanes 3 and 4) probes in the presence (lanes 2 and 4) and absence (lanes 1 and 3) of Gp2. (B) Graph showing the percentage of DNA bound by Eσ⁷⁰ in the presence of Gp2 compared with reactions with no Gp2. The dsDNA probes with different downstream end points are indicated in the x axis of the graph. (C) Autoradiograph of 4% (v/v) native polyacrylamide gels comparing binding of Eσ⁷⁰ with the +20/+20 (lanes 1–3) and +1/+1 (lanes 4–6) probes in the absence (condition I) and presence (conditions II and III) of Gp2. (D) Autoradiograph of a 4% (v/v) native polyacrylamide gel comparing binding of Eσ⁷⁰ (lanes 1 and 2) and Eσ⁷⁰_ΔR1.1 (lanes 3–6) in the presence (lanes 2, 4, and 6) and absence (lanes 1, 3, and 5) of Gp2 to the +1/+1 probe. In lanes 5 and 6, the isolated R1.1 domain is present in trans. (E) Autoradiograph of 4% (v/v) native polyacrylamide gels comparing binding of Eσ⁷⁰ to promoter probes with either the nontemplate or template strand ending at the −7 position (−7/+20 and +20/−7) in the absence (lane 2 and 5) and presence (lanes 3 and 6) of Gp2. In (A–E), the percentage of DNA bound by Eσ⁷⁰ or Eσ⁷⁰_ΔR1.1 (% C) in the reactions with Gp2 with respect to reactions with no Gp2 is given at the bottom of the gels. The data obtained in at least two independent experiments fell within 3% of the % C value shown. (F) Autoradiographs of 20% (v/v) denaturing urea gels showing the synthesis of the ApApUpU transcript (underlined nucleotides are α³²P labeled) from the lacUV5 promoter +20/+20 and −7/+20 probes by Eσ⁷⁰ in the absence (lanes 1 and 3) and presence (lanes 2 and 4) of Gp2. (G) As in F, but using the −7/+20 probe comparing the activity of Eσ⁷⁰ and Eσ⁷⁰_ΔR1.1 in the absence (lanes 1 and 3) and presence (lanes 2, 4, and 5) of Gp2. In lane 5, the isolated R1.1 domain is present in trans (at ∼8-fold molar excess over σ⁷⁰_ΔR1.1). In (F) and (G), the percentage of ApApUpU transcript synthesized (% A) in the reactions with Gp2 with respect to reactions with no Gp2 is given at the bottom of the gel for each reaction. The data obtained in at least two independent experiments fell within 5% of the % A value shown. See also Figure S4.

Figure 6

Gp2 Appropriates R1.1 to Efficiently Inhibit RPo Formation by Eσ⁷⁰ (A and B) Measurement of FRET between fluorescein incorporated into σ⁷⁰ at position 36 (σ^70∗) and Rif during RPo formation in the absence (A) and presence (B) of Gp2. The fluorescence emission spectra are recorded with 482 nm excitation. (C) The FRET efficiency values and distance calculations are tabulated (see also Figure S5) and the values presented are averages obtained from two to three individual experiments; the estimated error in R₀ is ∼10%. (D) Cartoon (as in Figure 1A) depicting the mechanism by which RPo formation is inhibited by Gp2 at σ⁷⁰-dependent promoters.