A continuous nonradioactive assay for RNA-dependent RNA polymerase activity

A continuous nonradioactive assay for RNA-dependent RNA polymerase activity

ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 325 (2004) 247–254 www.elsevier.com/locate/yabio A continuous nonradioactive assay for RNA-dependent ...

193KB Sizes 0 Downloads 33 Views

ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 325 (2004) 247–254 www.elsevier.com/locate/yabio

A continuous nonradioactive assay for RNA-dependent RNA polymerase activity Frederick C. Lahser and Bruce A. Malcolm* Department of Antiviral Therapeutics, Schering-Plough Research Institute, 2015 Galloping Hill Road, Kenilworth, NJ 07033, USA Received 21 July 2003

Abstract Current assays for the activity of viral RNA-dependent RNA polymerases (RdRps) are inherently end-point measurements, often requiring the use of radiolabeled or chemically modified nucleotides to detect reaction products. In an effort to improve the characterization of polymerases that are essential to the life cycle of RNA viruses and develop antiviral therapies that target these enzymes, a continuous nonradioactive assay was developed to monitor the activity of RdRps by measuring the release of pyrophosphate (PPi ) generated during nascent strand synthesis. A coupled-enzyme assay method based on the chemiluminescent detection of PPi , using ATP sulfurylase and firefly luciferase, was adapted to monitor poliovirus 3D polymerase (3Dpol ) and the hepatitis C virus nonstructural protein 5B (NS5B) RdRp reactions. Light production was dependent on RdRp and sensitive to the concentration of oligonucleotide primer directing RNA synthesis. The assay system was found to be amenable to sensitive kinetic studies of RdRps, requiring only 6 nM 3Dpol to obtain a reliable estimate of the initial velocity in as little as 4 min. The assay can immediately accommodate the use of both homopolymer and heteropolymer RNA templates lacking uridylates and can be adapted to RNA templates containing uridine by substituting a-thio ATP for ATP. The low background signal produced by other NTPs can be corrected from no enzyme (RdRp) controls. The effect of RdRp/RNA template preincubation was assessed using NS5B and a homopolymer RNA template and a time-dependent increase of RdRp activity was observed. Progress curves for a chain terminator (30 -deoxyguanosine 50 -triphosphate) and an allosteric NS5B inhibitor demonstrated the predicted time- and dose-dependent reductions in signal. This assay should facilitate detailed kinetic studies of RdRps and their potential inhibitors using either standard or single-nucleotide approaches. Ó 2003 Elsevier Inc. All rights reserved. Keywords: RNA-dependent RNA polymerase; RdRp; Coupled-enzyme assay; Luciferase; ATP sulfurylase; Hepatitis C virus; Poliovirus

During the replication of positive-strand RNA viruses, the syntheses of both positive- and negativestrand RNAs are catalyzed by the RNA-dependent RNA polymerase (RdRp)1, which is generally part of a larger multiprotein complex [1,2]. Due to their essential roles in the viral life cycles, RdRps have been the target of considerable interest for antiviral drug development [3–5]. Many of the published assays for RdRp activity are based on the detection of radiolabeled RNA prod*

Corresponding author. Fax: +1-908-740-3918. E-mail address: [email protected] (B.A. Malcolm). 1 Abbreviations used: RdRp, RNA-dependent RNA polymerase; SPA, scintillation–proximity assays; PNP, p-nitrophenyl; 3Dpol , poliovirus 3D polymerase; HCV, hepatisis C virus; NS5B, HCV nonstructural protein SB; APS, adenosine 50 -phosphosulfate; DTT, dithiothreitol; PMSF, phenylmethylsulfonyl flouride. 0003-2697/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2003.10.034

ucts synthesized during elongation [6–8]. Most of these assays are gel-based, in which radiolabeled RNAs are visualized following gel electrophoresis, but variations exist where products are detected by scintillation–proximity methods [9]. In scintillation–proximity assays (SPA), a biotinylated primer is annealed to a template to establish a duplex for extension by the polymerase in the presence of radiolabeled nucleotides. Terminated reactions are subsequently analyzed by the capture of the biotinylated and radiolabeled product on streptavidin beads. Although radioactive signals are easily detected and quantified, these assay systems inherently produce end-point measurements; kinetic evaluations are feasible, but the processing of multiple time-point samples for each experimental variation (time, drug concentration, etc.) is labor intensive.

248

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

Two nonradioactive polymerase assay systems have been described. A colorimetric RNA polymerase assay based on the incorporation of ATP or GTP with a pnitrophenyl moiety attached to the c-phosphate (PNP– NTP) has been described [10]. Incorporation of a PNP–NTP during polymerization releases a PNPpyrophosphate (PNP-PPi ), which is simultaneously hydrolyzed by calf intestinal phosphatase to generate the chromophore p-nitrophenylate, which is monitored at 405 nm. The authors demonstrated that the brome mosaic virus replicase and the bovine viral diarrhea virus and T7 bacteriophage RNA polymerases were all capable of utilizing the modified NTP substrate, albeit with lower efficiencies than those of the unmodified NTPs. This reduction of the incorporation efficiency can limit the application of this approach to permissive systems. Another nonradioactive assay in which an RNA template was covalently attached to a solid surface (Covalink module; Nalge Nunc International, USA) was described subsequently [11]. The polymerase reactions were initiated on the RNA template in the presence of the modified nucleotide biotin-16-UTP. Reactions were terminated and products with incorporated biotin were quantified by the addition of streptavidin-conjugated alkaline phosphatase and p-nitrophenolphosphate to produce a chromophore detectable at 405 nm. Although also based on p-nitrophenylate, this is intrinsically an end-point assay, with the attendant limitations discussed above with regard to the SPA methods. In addition, high concentrations of reagents are necessary to generate sufficient product for detection by colorimetric means, and the use of high enzyme and substrate levels makes analyses of potent inhibitors problematic. This report describes a sensitive continuous, nonradioactive RdRp assay in which activity is measured by a coupled-enzyme system based on the detection of free pyrophosphate (PPi ), a by-product of the polymerasemediated nucleotide incorporation reaction. In contrast to the requirement for modified nucleotides by the two nonradioactive polymerase assays described above, this new assay can utilize native NTPs. The luciferase-based approach increases the dynamic range of detection and allows for the use of lower concentrations of enzyme, templates, and nucleotides. Data in this assay are collected in real time, which allows the kinetic analysis of polymerization or even of single-nucleotide incorporation. In addition, continuous monitoring of the reaction allows progress curve evaluation of potential inhibitors, not only to provide greater statistical strength to the analysis of simple activity measurements but also to make evaluation of time-dependent inhibition straight forward. Exemplary studies were performed using the poliovirus 3D RNA polymerase (3Dpol ) [12,13] and the nonstructural protein 5B (NS5B) [9,14,15] of the hepatitis C virus (HCV) [16].

Materials and methods Materials Chemical reagents for assay-related buffers were obtained from Sigma (St. Louis, MO), USB (Cleveland, OH), GibcoBRL (Rockville, MD), and Boehringer Mannheim/Roche (Indianapolis, IN). Purified firefly luciferase and D -luciferin were from Promega (Madison, WI). Nucleotide triphosphate stocks were purchased from Promega and Ambion, Inc. (Austin, TX). Adenosine 50 -phosphosulfate (APS) and adenosine 50 -triphosphate (ATP) sulfurylase (1 U produces 1 lmol of ATP from APS and PPi per min at pH 8 at 30 °C) were obtained from Sigma. Poly C homopolymer was from Amersham (Piscataway, NJ), and oligo G12 was from Oligos, Etc. (Wilsonville, OR). Thermostable inorganic pyrophosphatase was from New England Biolabs (Beverly, MA). 30 -Deoxyguanosine 50 -triphosphate (30 dGTP) was from TriLink BioTechnologies (San Diego, CA). Compound 1, a benzo-1,2,4-thiadiazine, was purchased from ChemBridge (San Diego, CA). Pyrophosphate stock solutions and the Pi Per pyrophosphate assay kit were obtained from Molecular Probes (Eugene, OR). RNA oligonucleotides were synthesized by Dharmacon Research, Inc. (Lafayette, CO). Microlite 1 polystyrene immunoassay plates (96-well) for the polymerase/luciferase assay were from Dynex (Chantilly, VA). A microtiter plate luminometer (Lmax) was obtained from Molecular Devices (Sunnyvale, CA). Enzyme stocks of the polio 3D polymerase were prepared and stored in 50 mM Hepes (pH 7.3), 0.5 M NaCl, 5 mM DTT, 20% glycerol, 2 mg/L leupeptin, and 100 lM PMSF [9]. Enzyme stocks of the HCV NS5B with a Cterminal truncation of 55 amino acids (NS5BDCT55; referred hereafter as ‘‘NS5B’’) were prepared and maintained in enzyme storage buffer (25 mM Hepes (pH 7.5), 5 mM DTT, 0.6 M NaCl; 15% glycerol, 0.1% octylglucoside, 2 mg/ml leupeptin, 100 lM PMSF) as described previously [9]. Polymerase assay Polymerase assays were performed at room temperature using a 100-ll reaction mix in a 96-well plate. Final reaction conditions for 3Dpol [17,18] were 50 mM Hepes (pH 8.0), 2.5 mM MgCl2 , 20 U/ml RNAsin, 0.5 lg/ml oligo G12 , 5 lg/ml polyC, 1 lM GTP, 4 mM DTT, and 3Dpol , while those for NS5B were modified from those used previously [9] (20 mM Hepes, pH 7.3, 7.5 mM DTT, 20 U/ml RNAsin, 0.5 lg/ml oligo G12 , 5 lg/ml polyC, 1 lM GTP, 10 mM MgCl2 , 60 mM NaCl, 100 lg/ml bovine serum albumin, and NS5B). Both 3Dpol and NS5B assay mixtures were supplemented with components for the ATP sulfurylase and luciferase coupled-enzyme reactions (0.5 mM coenzyme A (CoA),

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

310 lM D -luciferin, 1 nM luciferase, 5 lM APS, 0.03– 0.3 U ATP sulfurylase) (see figure and table legends). Typically, the oligo G12 and poly C components (or heteropolymer primer and template) were preincubated at room temperature for 15 min prior to the addition of NS5B polymerase, luciferase, and ATP sulfurylase (the enzyme/template mix). For 3Dpol reactions, the enzyme was coincubated along with the RNA. Reactions were initiated by the addition of GTP, APS, and D -luciferin (the substrate mix). To test the effect of inhibitors on NS5B, compounds were titrated in the appropriate diluents and added to one of the mixtures (see legend of Fig. 4 for details). Microtiter plates containing the initiated reaction samples were immediately transferred to the luminometer for detection of the light signal generated over time (using the instrumentÕs ‘‘Long Kinetic’’ parameters), taking a 0.2 s reading of each well every 30–60 s (time interval depends on the number of samples tested). Readings were monitored by comparison to a no-NS5B reaction control. Initial velocities were determined from nonlinear regression assuming pseudo-first order kinetics [P ¼ S0 ð1  ekt Þ], and kcat and Km were subsequently calculated by fitting estimated initial velocities to the Michaelis–Menten equation (Prism 3.0; GraphPad Software, San Diego, CA). Data collection was restricted to PPi concentrations <100 nM to avoid the influence of substrate depletion and/or product inhibition. To test the effect of NTPs on the ATP sulfurylase/ luciferase coupled readout system, control reactions for NS5B assays (with 0.3 U ATP sulfurylase), omitting RdRp and RNA, were assembled as above. Reactions were initiated by addition of nucleotide (1 lM final). To assess the potential contribution of free PPi in commercial NTP preparations, aliquots of the 10 mM stocks were pretreated with inorganic pyrophosphatase. A 10ll aliquot of the NTP stock was incubated with 0.5 U of thermostable inorganic pyrophosphatase for 10 min at 75 °C; untreated NTP samples were processed in parallel as controls. Treated and untreated NTPs were diluted to a final concentration of 1 lM for evaluation. To assess the effect of assay components only on the polymerase reaction and not the sulfurylase/luciferasebased detection system, RdRp reactions were assembled without these reagents. Aliquots were taken from the reactions over time and stopped by heating to 75 °C for 5 min. Pyrophosphate content was assessed using the Pi Per pyrophosphate assay kit (Molecular Probes) as described by the manufacturer, with a SpectraMAX spectrophotometer (Molecular Devices) at 565 nm. Polymerase/RNA preincubation studies The effect of preincubating the polymerase with RNA template was assessed by combining (at various time

249

points) NS5B, poly C, and oligo G12 at twice the final reaction concentration and initiated by addition of an equal volume of a 2 reaction mix containing luciferase, ATP sulfurylase, CoA, GTP, APS, and D -luciferin. Data were typically collected for 30 min.

Results Principle of the method The assay was assembled based on the reaction conditions described previously [9] with the incorporation of reagents used for the light-based detection of free pyrophosphate [19–21]. Free PPi is converted to ATP in a reaction catalyzed by ATP sulfurylase using APS. This ATP pool provides the energy for a luciferase-catalyzed reaction, producing photons in direct proportion to the amount of PPi generated by the viral polymerase. The chemical formulas describing these steps are as follows: ðslowÞ

ðRNAÞn þ NTP ! ðRNAÞnþ1 þ PPi ; RdRp

APS þ PPi

ðfastÞ

!

ATPsulfurylase

ATP þ SO4 ; and

ð1Þ ð2Þ

d-luciferin þ ATP þ O2 %hv

! oxyluciferin þ AMP þ PPi þ CO2 :

luciferase

ð3Þ

The PPi produced by the ATP-luciferase reaction (step 3) feeds back into the ATP sulfurylase reaction (step 2) to maintain the ATP level. As the RdRp produces more PPi (step 1), the concentration of ATP continues to rise over time with a resulting increase in the production of light. As opposed to a DNA sequencing methodology using this cascade (pyrosequencing) [22], in which the reaction scheme is terminated at each elongation step by the nucleotide-degrading enzyme apyrase, the absence of apyrase allows this RdRp assay to monitor enzyme activity continuously. As with all enzyme-catalyzed reactions there is the potential for product inhibition. The product of the luciferase reaction (step 3), oxyluciferin, is a known inhibitor of luciferase. Coenzyme A relieves this inhibition by displacing oxyluciferin from luciferase, thereby extending the duration of light production [23,24], and was included routinely in the assay buffer. Homopolymer assay Initial experiments used the poly C/oligo G12 system with polio 3Dpol . Using reaction conditions comparable to those for DNA pyrosequencing (see Materials and methods) [20,21], the reaction rate was linear with respect to enzyme concentration up to the maximum tested (100 nM) (Figs. 1A,B). A measurable signal could

250

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

Fig. 1. Characterization of poliovirus 3Dpol using the continuous assay format. (A) Titration of 3Dpol . Poliovirus 3Dpol was serially diluted in storage buffer and assayed ðn ¼ 2Þ at room temperature with 1 lM GTP, 0.03U ATP sulfurylase, and the poly C/oligo G12 in a 96-well plate with a final reaction volume of 100 ll. Light was captured for 0.2 s every 45 s during a 30-min reaction. Initial velocity values were derived from progress curve analyses, fitted by nonlinear regression to a simple first order model [i.e., P ¼ S0 ð1  ekt Þ]. (B) Linearity of reaction with RdRp concentration. Initial velocity estimates from the data in (A) were plotted against [3Dpol ]. Values were fitted by linear regression. (C) Effect of primer on activity. Various concentrations of oligo G12 primer were added to 5 lg/ml polyC, with other conditions and analysis parameters as described for (A). (D) Linearity of reaction with primer concentration. Initial velocity estimates from data in (C) were plotted against [oligo G12 ]. Values were best fit to the Michaelis– Menten model. Results shown are the average of two (A, B) or three (C, D) independent experiments, and SE 65% for all points.

be obtained with 6 nM 3Dpol in as little as 1 min. For the subsequent studies, reactions used 10 nM 3Dpol . A low level of PPi could be detected from reactions lacking primer (Fig. 1C). Polymerase activity increased with primer concentration, with indications of saturation at the maximum tested (5 lg/ml) (Fig. 1D). Heteropolymeric templates To develop an assay using nonhomopolymer templates, the ability of 3Dpol to utilize a natural heteropolymer template and primer was tested. A 75-mer RNA oligonucleotide (Fig. 2A) was synthesized based on a sequence previously used as a template for NS5Bcatalyzed reactions (D-RNA) [14]. Due to the strict primer dependence of 3Dpol [12,25], a primer complementary to the 30 12 bases of D-RNA was also generated. Assays utilizing the heteropolymer template contained 100 nM of both template and primer (100 nM duplex) and 60 nM 3Dpol . Higher concentrations of RdRp were used in these reactions for improved signal

to noise ratio (data not shown). In contrast to the homopolymer assay, reactions using the heteropolymer RNA template included GTP, CTP, UTP, and ATPaS, a nucleotide that is a poor substrate for luciferase [21]; (see below). ATP and ATPaS utilization by RdRps was compared in assays that did not use the coupled-enzyme detection system. Estimates of initial velocities were derived from autoradiography of radiolabeled products for NS5B and pyrophosphate end-point measurements for 3Dpol . For both RdRps, the apparent initial velocities of reactions in which ATPaS replaced ATP were about one third those of ATP-containing reactions (data not shown). As shown in Fig. 2B, the 3Dpol reaction produces a strong signal compared with the no-primer control reaction (signal to noise of 10:1). The APS component, an ADP mimic, was not found to inhibit RdRp activity at the concentration used in the assay (data not shown). Low levels of light can be detected from the reaction in the absence of primer (Fig. 2B). There are three possible explanations for this time-dependent signal:

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

251

Table 1 Background signal from NTPs [NTP] (1 lM final)

Untreated initial velocity (  SD) [nM PPi/min]

PPase treated initial velocity (  SD) [nM PPi/min]

GTP CTP UTP ITP ATPaS

0.01  0.01 0.04  0.01 0.04  0.01 0.1  0.02 0.024  0.004

0.011  0.004 0.04  0.01 0.04  0.01 0.1  0.01 0.03  0.001

Aliquots of 10 mM NTP were treated with inorganic pyrophosphatase as described under Materials and methods for this experiment. Reactions testing treated and untreated NTPs were performed in the absence of NS5B and RNA. RLU values were converted to nM PPi and analyzed by nonlinear regression to obtain initial velocity estimates.

Fig. 2. Assay using a heteropolymer RNA template and 3Dpol . (A) Predicted secondary structure of the 75-nucleotide-long D-RNA, as predicted by Mfold [37]. The annealing site for a complementary 12mer primer is shown (black arrow). (B) Primer dependence of 3Dpol on a heteropolymer RNA template. Duplicate reactions contained 60 nM 3Dpol , 100 nM each D-RNA and 12-mer primer (100 nM template RNA), 1 lM each nucleotide (GTP, CTP, UTP, and ATPaS), and 0.03 U ATP sulfurylase. RdRp and RNA were coincubated at room temperature for 15 min prior to the addition of NTPs to initiate the reactions. Progress curve data were fit using nonlinear regression analysis, as described under Materials and methods.

(1) APS may nonenzymatically react with D -luciferin, forming a compound that activates luciferase and produces light [19],(2) 3Dpol may be capable of inefficient de novo (primer-independent) synthesis, as described for NS5B [6,26], or (3) some amount of self-priming may take place with this particular template (Fig. 2A). Such background signals need to be assessed for each RdRp and substrate, using appropriate control reactions. Another possible source of background is direct utilization of the NTPs by the ATP sulfurylase/luciferase cascade. To address this possibility, reactions without either RdRp or template were examined. As shown in Table 1, NTPs can act as substrates for the enzyme cascade with varying efficiencies. The Km for ATP, the normal substrate for the luciferase reaction, is 0.65  0.04 lM (under these experimental conditions, data not shown). GTP, the substrate used in the homopolymer studies reported above, had the lowest level of signal. ATPaS was also found to be a poor substrate

for luciferase, which is consistent with the report of Ronaghi et al. [21], who described the use of dATPaS as a substitute for dATP in pyrosequencing reactions. Inosine triphosphate (ITP), however, was a relatively good luciferase substrate, about 10-fold better than GTP. To eliminate contaminating PPi as a source of the signal, NTP stocks were treated with inorganic pyrophosphatase. No significant changes in the initial velocity estimates were observed when NTPs were pretreated with pyrophosphatase (Table 1). Taken together, these data suggest that, under these conditions, background signals are produced as a result of utilization of NTPs by luciferase, necessitating the use of appropriate control reactions. Polymerase assays using NS5B To extend the use of the assay to other RdRps, the HCV NS5BDCT55 (NS5B) enzyme was tested using appropriate reaction buffer conditions (see Materials and methods). Results from a titration of NS5B (Fig. 3A) showed an increase in released PPi in a timeand enzyme concentration-dependent manner, similar to those obtained for 3Dpol (Fig. 1A). The NS5B reaction was linear with enzyme concentration but showed timedependent curvature at 100 nM, most likely the effect of: (1) substrate (e.g., NTP) depletion or (2) product inhibition by oxyluciferin [23]. In subsequent studies of this enzyme, reactions used 20 nM NS5B. Effect of NS5B/RNA preincubation The impact of preincubation of NS5B with template– primer was examined using the poly C/oligo G12 assay, described above. NS5B was coincubated with template– primer prior to the initiation of RNA synthesis by addition of the remaining reaction components. As shown in Fig. 3B, there is a 10-fold increase in signal with 24-h preincubation (i.e., v0 of 10 vs v0 of 100). This suggests that, for this system, formation of the initiation complex was rate limiting; plateauing of the preincubation time

252

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

Fig. 3. Characterization of NS5B using the continuous assay format. (A) Increase in PPi concentration due to increasing NS5B concentration. Polymerase was serially diluted in storage buffer (see Materials and methods) and assayed at room temperature in the presence of 1 lM GTP, 0.3 U ATP sulfurylase, 0.5 lg/ml oligo G12 , and 5 lg/ml polyC. Light was captured for 0.2 s every 60 s during a 30-min reaction. Initial velocity values were derived from progress curve analyses, fitted by nonlinear regression as described for Fig. 1. (B) Preincubation of NS5B with RNA template increases apparent polymerase activity. HCV NS5B was preincubated with poly C and oligo G12 for various times (24, 6, 3, 2, 1, 0.5, 0.25 h) prior to initiating the reaction as described under Materials and methods. Data were collected from duplicate reactions and fit to the simple first order model as described for Fig. 1. Initial velocity estimates were plotted against time of preincubation and fit to a simple first order model.

course may reflect attainment of equilibrium or the gradual inactivation of the enzyme. Assuming no loss of activity over the course of the preincubation, an apparent t1=2 for activation of the reaction was estimated to be 3.7  0.3 h (Fig. 3B). The slow activation may reflect conformational changes required by the RdRp or the RdRp–RNA complex to generate a polymerizationcompetent form as noted with other systems (e.g., T7 RNA polymerase [27,28]). Inhibition of NS5B by a chain terminator and an allosteric inhibitor To demonstrate the utility of this assay for the evaluation of various inhibitors, a chain terminator (30 -

deoxyguanosine-50 triphosphate) and a recently described allosteric inhibitor (Compound 1; [29]) were evaluated using the poly C/oligo G12 form of the assay. Dose- and time-dependent reduction in signal was observed for the compound (Fig. 4A), consistent with chain termination (i.e., once elongation is terminated by 30 -dGTP incorporation, no further PPi can be produced). In a second study, Compound 1 was characterized for inhibition. When Compound 1 was added with NS5B to the poly C/oligo G12 template prior to initiation of the reaction with GTP, polymerization was slowed in a dose-dependent manner (Fig. 4B). The progress curves obtained in Fig. 4B are consistent with a time-dependent decrease in the rate of the RdRp activity, reminiscent of those observed with the chain terminator (Fig. 4A),

Fig. 4. Inhibitor studies using NS5B. (A) Inhibition of NS5B by 30 -deoxyguanosine 50 triphosphate. Duplicate reactions contained NS5B (20 nM), GTP substrate (1 lM), and 0.3 U ATP sulfurylase, with varying concentrations of inhibitor (0.1–2.5 lM in water). Luminescence was measured for 0.2 s every 40 s during a 30-min reaction. (B) Inhibition of NS5B by Compound 1 [29]. Duplicate reactions, with 20 nM NS5B and 1 lM GTP, included various concentrations of Compound 1 (2.5–20 lM in 1% dimethyl sulfoxide final). Light was captured for 0.2 s every 45 s during a 30-min reaction. Progress curve data in both panels were fit using nonlinear regression analysis, as described under Materials and methods.

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

suggesting that this inhibitor most probably interacts only with the RdRp prior to formation of the initiation (i.e., reinitiation) complex (or at some other sui generis state in the replication cycle).

Discussion Currently existing RdRp assays are, for the most part, inherently end-point measurements of product, requiring the use of artificial or radiolabeled nucleotides for detection. Aside from the obvious issues concerning the use of modified nucleotides with the RdRp of intent (and the usually lower rates of incorporation), data collection over time is labor intensive. To accommodate even a simple time course using a radiolabel incorporation method, a large number of reactions have to be set up and processed, with the attendant variability. The simple and sensitive nonradioactive RdRp assay described above can be monitored in real time, yielding improved estimation of reaction rates and the potential for progress curve analysis. The 96-well format described here allows for throughput equivalent to current end-point methods such as SPA with no postreaction handling. Furthermore, because a radiolabeled (or chemically modified) nucleotide is not necessary for product detection, there need be no ‘‘limiting’’ substrate. This flexibility is of value for the development of assays using heteropolymer RNA templates (see discussion below) and allows use of NTPs at their Km s. Real-time collection of data is important for the analysis of polymerase inhibition as highlighted in the studies above. The behavior of progress curves allows immediate inferences concerning the nature of the inhibition event for novel inhibitors, as observed for Compound 1, a known initiation inhibitor [29,30] (Fig. 4B). Assays using heteropolymer templates obviously necessitate the use of multiple NTPs, requiring preassessment of background signal. As demonstrated above, by use of ATPaS, even natural heteropolymers using all four bases can be utilized (Fig. 2). The efficiency of ATPaS incorporation by the RdRp was estimated to be about three-fold lower than that for ATP (data not shown). Preliminary testing of a particular RdRp will be essential to determine whether this strategy is feasible or whether a synthetic heteropolymer devoid of U residues will be necessary. Although newly generated ATP (from the APS/PPi /sulfurylase reaction) may compete with ATPaS for incorporation into the polymerization process, use of a sufficiently large excess of the latter nucleotide should effectively prevent removal of the former during the reaction (i.e., high ATPaS levels should prevent any significant amount of ATP from being incorporated into the product strand before it is used by luciferase; see below). As this approach enables inference of polymerization activity through PPi generation, other activities that

253

release pyrophosphate, such as terminal transferase activity [31,32] or a nontemplated synthesis activity (i.e., primase) [33,34], will produce signal that must be corrected for if only the rate of templated polymerization is desired. Alternatively, the magnitude of these activities can be studied using no-primer reactions and compared with those of ‘‘dedicated’’ transferases and primases using this approach. RdRps used in these studies have relatively slow turnover rates. To employ this approach with more active enzymes, lower amounts of RdRps (or transferases or primases) should be used to ensure that generation of PPi is rate limiting. The activity of the light generation cascade (ATP sulfurylase, luciferase) may also vary under the reaction conditions for a given enzyme and should be carefully evaluated as part of adapting this approach to other PPi -generating reactions. Pyrophosphate released by the polymerase reacts rapidly with the sulfurylase/luciferase cascade. Even at a 10 pmol concentration of PPi , half-maximal light intensity occurs in approximately 3 s [19], the rate-limiting step being generation of ATP by sulfurylase (the luciferase reaction with an equivalent amount of ATP under NS5B assay conditions occurs in less than 1 s; data not shown), [35,36]. In summary, a new method for assessing and characterizing RdRp activities amenable to 96-well format/ high-throughput has been described. In contrast to a number of similar assays that generate end-point results [6–11], this method can measure polymerase activity in real time, improving the quality of activity measurements and allowing facile evaluation of time-dependent phenomena.

Acknowledgments We thank Drs. H.C. Huang, Charles Lesburg, and Michael Cable, Ms. Nancy Butkiewicz, and Mr. Eric Ferrari for valuable discussions and reagents and Dr. Xiao Tong for critical reading of the manuscript.

References [1] D. Egger, B. Wolk, R. Gosert, L. Bianchi, H.E. Blum, D. Moradpour Bienz, K. Bienz, Expression of hepatitis C virus proteins induces distinct membrane alterations including a candidate viral replication complex, J. Virol. 76 (2002) 5974–5984. [2] N.L. Teterina, D. Egger, K. Bienz, D.M. Brown, B.L. Semler, E. Ehrenfeld, Requirements for assembly of poliovirus replication complexes and negative-strand RNA synthesis, J. Virol. 75 (2001) 3841–3850. [3] J.G. McHutchison, K. Patel, Future therapy of hepatitis C, Hepatology 36 (2002) S245–S252. [4] S.L. Tan, A. Pause, Y. Shi, N. Sonenberg, Hepatitis C therapeutics: current status and emerging strategies, Nat. Rev. Drug Disc. 1 (2002) 867–881.

254

F.C. Lahser, B.A. Malcolm / Analytical Biochemistry 325 (2004) 247–254

[5] M.P. Walker, Z. Hong, HCV RNA-dependent RNA polymerase as a target for antiviral development, Curr. Opin. Pharmacol. 2 (2002) 1–7. [6] W. Zhong, A.S. Uss, E. Ferrari, J.Y. Lau, Z. Hong, De novo initiation of RNA synthesis by hepatitis C virus nonstructural protein 5B polymerase, J. Virol. 74 (2000) 2017–2022. [7] S. Reigadas, M. Ventura, L. Sarih-Cottin, M. Castroviejo, S. Litvak, T. Astier-Gin, HCV RNA-dependent RNA polymerase replicates in vitro the 30 terminal region of the minus-strand viral RNA more efficiently than the 30 terminal region of the plus RNA, Eur. J. Biochem. 268 (2001) 5857–5867. [8] V. Lohmann, A. Roos, F. Korner, J.O. Koch, R. Bartenschlager, Biochemical and structural analysis of the NS5B RNA-dependent RNA polymerase of the hepatitis C virus, J. Viral Hepatol. 7 (2000) 167–174. [9] E. Ferrari, J. Wright-Minogue, J.W. Fang, B.M. Baroudy, J.Y. Lau, Z. Hong, Characterization of soluble hepatitis C virus RNAdependent RNA polymerase expressed in Escherichia coli, J. Virol. 73 (1999) 1649–1654. [10] W. Vassiliou, J.B. Epp, B.B. Wang, A.M. Del Vecchio, T. Widlanski, C.C. Kao, Exploiting polymerase promiscuity: a simple colormetric RNA polymerase assay, Virology 274 (2000) 429–437. [11] C. Park, Y. Kee, J. Park, H. Myung, A nonisotopic assay method for hepatitis C virus NS5B polymerase, J. Virol. Methods 101 (2002) 211–214. [12] J.B. Flanegan, D. Baltimore, Poliovirus-specific primer-dependent RNA polymerase able to copy poly(A), Proc. Natl. Acad. Sci. USA 74 (1977) 3677–3680. [13] T.A. Van Dyke, J.B. Flanegan, Identification of poliovirus polypeptide P63 as a soluble RNA-dependent RNA polymerase, J. Virol. 35 (1980) 732–740. [14] S.-E. Behrens, L. Tomei, R. De Francesco, Identification and properties of the RNA-dependent RNA polymerase of hepatitis C virus, EMBO J. 15 (1996) 12–22. [15] V. Lohmann, F. Korner, U. Herian, R. Bartenschlager, Biochemical properties of hepatitis C virus NS5B RNA-dependent RNA polymerase and identification of amino acid sequence motifs essential for enzymatic activity, J. Virol. 71 (1997) 8416–8428. [16] G.M. Lauer, B.D. Walker, Hepatitis C virus infection, N. Engl. J. Med. 345 (2001) 41–52. [17] J.J. Arnold, C.E. Cameron, Poliovirus RNA-dependent RNA polymerase (3Dpol) is sufficient for template switching in vitro, J. Biol. Chem. 274 (1999) 2706–2716. [18] C. Ward, M. Stokes, J.B. Flanegan, Direct measurement of the poliovirus RNA polymerase error frequency in vitro, J. Virol. 62 (1988) 558–562. [19] P. Nyren, A. Lundin, Enzymatic method for continuous monitoring of inorganic pyrophosphate synthesis, Anal. Biochem. 151 (1985) 504–509. [20] P. Nyren, Enzymatic method for continuous monitoring of DNA polymerase activity, Anal. Biochem. 167 (1987) 235–238. [21] M. Ronaghi, S. Karamohamed, B. Pettersson, M. Uhlen, P. Nyren, Real-time DNA sequencing using detection of pyrophosphate release, Anal. Biochem. 242 (1996) 84–89. [22] M. Ronaghi, M. Uhlen, P. Nyren, A sequencing method based on real-type pyrophosphate, Science 281 (1998) 363–365.

[23] S.R. Ford, M.S. Hall, F.R. Leach, Enhancement of firefly luciferase activity by cytidine nucleotides, Anal. Biochem. 204 (1992) 283–291. [24] R.L. Airth, W.C. Rhodes, W.D. McElroy, The function of coenzyme A in luminescence, Biochem. Biophys. Acta 27 (1958) 519–532. [25] V. Rodriguez-Wells, S.J. Plotch, J.J. DeStefano, Primer-dependent synthesis by poliovirus RNA-dependent RNA polymerase, Nucl. Acids Res. 29 (2001) 2715–2724. [26] M. Kim, H. Kim, S.P. Cho, M.K. Min, Template requirements for De Novo RNA synthesis by hepatitis C virus nonstructural protein 5B polymerase on the viral X RNA, J. Virol. 76 (2002) 6944–6956. [27] G.M.T. Cheetham, T.A. Steitz, Structure of a transcribing T7 RNA polymerase initiation complex, Science 286 (1999) 2305– 2309. [28] G.M.T. Cheetham, D. Jeruzalmi, T.A. Steitz, Structural basis for initiation of transcription from an RNA polymerase-promoter complex, Nature 399 (1999) 80–83. [29] D. Dhanak, K.J. Duffy, V.K. Johnston, J. Lin-Goerke, M. Darcy, A.N. Shaw, B. Gu, C. Silverman, A.T. Gates, M.R. Nonnemacher, D.L. Earnshaw, D.J. Casper, A. Kaura, A. Baker, C. Greenwood, L.L. Gutshall, D. Maley, A. DelVecchio, R. Macarron, G.A. Hofmann, Z. Alnoah, H.Y. Cheng, G. Chan, S. Khandekar, R.M. Keenan, R.T. Sarisky, Identification and biological characterization of heterocyclic inhibitors of the hepatitis C virus RNA-dependent RNA polymerase, J. Biol. Chem. 277 (2002) 38322–38327. [30] B. Gu, V.K. Johnston, L.L. Gutshall, T.T. Nguyen, R.R. Gontarek, M.G. Darcy, R. Tedesco, D. Dhanak, K.J. Duffy, C.C. Kao, R.T. Sarisky, Arresting initiation of hepatitis C virus RNA synthesis using heterocyclic derivatives, J. Biol. Chem. 278 (2003) 16602–16607. [31] H. Guan, A.E. Simon, Polymerization of nontemplated bases before transcription initiation at the 30 ends of templates by an RNA-dependent RNA polymerase: an activity involved in 30 end repair of viral RNAs, Proc. Natl. Acad. Sci. USA 97 (2000) 12451–12456. [32] C.T. Ranjith-Kumar, J. Gajewski, L.L. Gutshall, D. Maley, R.T. Sarisky, C.C. Kao, Terminal nucleotidyl transferase activity of recombinant Flaviviridae RNA-dependent RNA polymerases: implication for viral RNA synthesis, J. Virol. 75 (2001) 8615– 8623. [33] J.H. Sun, C.C. Kao, Characterization of RNA products associated with or aborted by a viral RNA-dependent RNA polymerase, Virology 236 (1997) 348–353. [34] J.H. Sun, S. Adkins, G. Faurote, C.C. Kao, Initiation of (-)-strand RNA synthesis catalyzed by the BMV RNA-dependent RNA polymerase: synthesis of oligonucleotides, Virology 226 (1996) 1– 12. [35] A. Lundin, A. Thore, M. Baltscheffsky, Sensitive measurement of flash induced photophosphorylation in bacterial chromatophores by firefly luciferase, FEBS Lett. 79 (1977) 73–76. [36] M. DeLuca, W.D. McElroy, Kinetics of the firefly luciferase catalyzed reactions, Biochemistry 13 (1974) 921–925. [37] M. Zuker, Mfold web server for nucleic acid folding and hybridization prediction, Nucleic. Acids Res. 31 (2003) 1–10.