Global kinetic mechanism describing single nucleotide incorporation for RNA polymerase I reveals fast UMP incorporation

Abstract

RNA polymerase I (Pol I) is responsible for synthesizing ribosomal RNA, which is the rate limiting step in ribosome biogenesis. We have reported wide variability in the magnitude of the rate constants defining the rate limiting step in sequential nucleotide additions catalyzed by Pol I. in this study we sought to determine if base identity impacts the rate limiting step of nucleotide addition catalyzed by Pol I. To this end, we report a transient state kinetic interrogation of AMP, CMP, GMP, and UMP incorporations catalyzed by Pol I. We found that Pol I uses one kinetic mechanism to incorporate all nucleotides. However, we found that UMP incorporation is faster than AMP, CMP, and GMP additions. Further, we found that endonucleolytic removal of a dimer from the 3’ end was fastest when the 3’ terminal base is a UMP. It has been previously shown that both downstream and upstream template sequence identity impacts the kinetics of nucleotide addition. The results reported here show that the incoming base identity also impacts the magnitude of the observed rate limiting step.

Graphical Abstract

Introduction

RNA polymerase I is essential for ribosome biogenesis and is exclusively responsible for synthesis of the 5.8 S, 18 S, and 25 S ribosomal (r)RNAs. However, eukaryotes express at least two additional RNA polymerases (Pols) responsible for transcription of alternative classes of genes. RNA Pol II transcribes protein-coding genes and most regulatory, non-coding RNAs, and Pol III synthesizes primarily transfer (t) RNA and the 5 S rRNA [1-4].

We have reported the development and application of a single-turnover rapid mixing kinetics approach to examine the mechanism of a single incorporation of AMP catalyzed by Pol I [5, 6]. We found that the rate limiting step for this reaction was ~200 s⁻¹. In contrast, the average rate of incorporation over kilobases, both in vivo [7] and in vitro [8, 9], has been shown to be ~60 nt s⁻¹.

The same single-turnover approach was used to interrogate the mechanisms of nine sequential nucleotide additions of AMP and GMP [10-12]. Interestingly, both the first and second additions, corresponding to A followed by G, were observed to be 190 and 256 s⁻¹, respectively [10]. The remaining seven converged on the average of ~ 60 s⁻¹ but this average varied between 22 – 116 s⁻¹. These observations left us asking why the first and second additions are so much faster than the subsequent additions.

Here we sought to determine if the identity of the incoming nucleotide impacts the rate of a single nucleotide addition catalyzed by Pol I. We found that single nucleotide addition time courses for the four canonical nucleotides could be described by the same number of elementary kinetic steps, i.e. the same mechanism. Subtle differences in the binding affinity are detected but, to our surprise, U is incorporated nearly two-fold faster (~630 s⁻¹) than A, C, and G (~320 s⁻¹). Curiously, removal of a dinucleotide from the 3’ end was also found to be nearly an order of magnitude faster when the cognate nucleotide was a U. This dependence of the rate of incorporation on nucleotide identity is consistent with an induced fit mechanism.

Materials and Methods

Buffers

All buffers and reagents were filtered using Millipore-Sigma Stericup Vacuum Bottle Top Filtration System (0.22 μm filters) (MilliporeSigma Billerica, MD) unless stated otherwise.

The reaction conditions (buffer A) for quench flow time courses was 40 mM KCl, 20 mM Tris-Acetate (OAc) pH 8.23 at 25 °C, 2 mM dithiothreitol, and 0.2 mg mL⁻¹ bovine serum albumin (BSA).

Proteins

Pol I was purified from Saccharomyces cerevisiae as previously described [6, 13].

Nucleotides, nucleic acids, heparin, and BSA

The preparation of nucleotides, nucleic acids, heparin, and BSA has been described previously [5, 6]. The sequences of the template DNA, non-template DNA, and RNA primers are shown in Figure 1 A.

Figure 1.

Redesign of DNA template, radiolabeled elongation complex (EC) formation, and chemical quench flow setup. (A) Several different template DNA strands were used to test for the extension of different nucleotides. The sequences of the template-DNA strand as well as the RNA product for UMP, AMP, GMP, and CMP incorporation are shown, as well as a previously published AMP incorporation used for comparison. (B) Elongation complexes (ECs) are built by allowing WT Pol I (gray) to incubate with a pre-annealed DNA_t:RNA. The 9-mer RNA is shown in green and the template DNA is shown in red. A non-template DNA strand, shown in black, is added to form an EC. The Pol is allowed to incorporate an α-³²P-NTP for 10 minutes to produce a radiolabeled EC. The schematic is not to scale. (C) The radiolabeled EC is loaded into syringe 1 of the chemical quench flow, and syringe 2 contains varying concentrations of NTP, magnesium, and heparin. Syringe 1 and 2 are rapidly mixed and the radiolabeled EC extends from a 10-mer to 11-mer. The 11-mer can be cleaved to form a 9-mer and radiolabeled dimer.

Mixing in the quench flow is 1:1, and thus all concentrations stated below are two-fold lower after mixing.

Radiolabeled elongation complex syringe

Elongation complexes (ECs) were built using our previously published method[5, 6, 11, 14, 15]. Figure 1B illustrates the construction of the radiolabeled EC. A 9-mer RNA primer is pre-annealed to the template DNA (DNA_T), followed by the addition of Pol I. After the Pol binds for a fixed amount of time to the RNA:DNA_T hybrid, the non-template DNA (DNA_NT) is added to form an EC. Finally, α-³²P-NTP and magnesium (Mg²⁺) is added, and the bound Pol is allowed to incubate for 10 minutes. During this time, Pol I incorporates the radiolabel at the 3’ end resulting in a 10-mer RNA. After ten minutes, EDTA is added to chelate excess Mg²⁺ and stop the labeling reaction. The radiolabeled EC is composed of ≈16 nM Pol I, 162.75 nM RNA, 54.26 nM template DNA, 162.75 nM non-template DNA, ≈5 nM α-³²P-CTP, 100 μM Mg(OAc)₂, and 1.1 mM ethylenediamine tetraacetic acid (EDTA)-K₃.

NTP syringe

The NTP Syringe contains either adenosine triphosphate (ATP; Sigma Aldrich, A26209-5G), cytidine triphosphate (CTP; Sigma Aldrich C1506-250MG), uridine triphosphate (UTP; Sigma Aldrich U6875-100MG), or guanosine triphosphate (GTP; Sigma Aldrich, G8877-1G) at twice the concentrations indicated in the text, 18 mM Mg(OAc)₂, and 0.05 mg mL⁻¹ heparin.

Chemical Quench Flow time courses

The radiolabeled EC was loaded into one syringe of the chemical quench flow. The contents of the EC syringe and the NTP syringe are rapidly mixed and Pol I extends the radiolabeled 10-mer to a radiolabeled 11-mer for fixed amounts of time before rapid mixing with 1 M HCl to instantaneously stop the reaction. Time points were collected from 0.005 s – 100 s. Time points were separated on sequencing gels to resolve the 10-mer, 11-mer, other products, and quantified using Equation 1 as previously described [6].

Model Independent Analysis

Time courses were fit to sums of exponentials (11-mer: equation 2 and dimer: Equation 3) and fits of the ${\mathsf{k}}_{\text{obs}}$ secondary plots (Equation 4) were performed using nonlinear least squares (NLLS) in KaleidaGraph (Synergy Software, PA).

Model Dependent Analysis

Time courses were fit using a previously published MATLAB toolbox called MENOTR[16]. MENOTR (https://github.com/ZachIngram/2021-MENOTR) uses a combination of genetic algorithm and NLLS to find the best fit parameters for the time courses. Each set of time courses are globally fit as a function of both time and [NTP]. Each concentration contains three experimental replicates.

Time courses were globally analyzed one replicate at a time, and the average and standard deviation of each parameter is reported.

HPLC Analysis

Nucleotide Purity Analysis via High-Performance Liquid Chromatography.

ATP, CTP, GTP, and UTP purity was assessed by subjecting samples of each nucleotide to analysis via high-performance liquid chromatography (HPLC). Nucleotide solutions were prepared from their respective stocks to a final concentration of 0.1, 0.2, 0.3, 0.5, and 1 mM using buffer A. This was done to assess contamination levels over the range of concentrations used in chemical quench-flow assays. Additionally, 1 mM solutions of each nucleotide stock were prepared and spiked with a final concentration of 1 mM ATP as a false positive and 100% buffer A was injected as a blank measurement.

Separations were performed using a Waters 1525 Binary HPLC Pump with a 1500 series column heater, injected using a Waters 2707 Autosampler, and a Waters 2998 Photodiode Array Detector (Waters Corporation, USA) was used for detection. UV absorbance data was collected between 200-400 nm, and chromatograms were directly collected at the ${\lambda }_{\max }$ for GTP, ATP, UTP, and CTP (252, 260, 262, and 272 nm respectively). The column used was a Waters 4.6x150 mm XBridge BEH130 C₁₈ column (3.5 μm particle size) with a 4.6x20 mm guard column of matching chemistry and was heated to 40°C for the duration of the separation. The separation method was adapted from Stocchi, et Al., 1987 [17]. Mobile phases A and B consisted of 0.1 M potassium phosphate (pH 6.5) with 6 mM tetrabutylammonium hydrogen sulfate (TBAHS). Mobile phase A was 100% water, while mobile phase B was made with 60%/40% water/acetonitrile (v/v) (HPLC grade, Fisher Scientific). Mobile phases were vacuum filtered and degassed using 0.22 μm nylon filters (Whatman, GE Healthcare). 5 μL injections of each sample were made for each analysis. Initial solvent conditions were set to 10% B at a flow rate of 0.8 mL min⁻¹. A linear gradient was applied starting at 2.50 minutes to bring the solvent ratio to 50% B at 7.00 minutes. 50% B was maintained from 7.00 to 13.00 minutes. The ratio was then immediately changed to 100% B at 13.00 minutes and held for 3 minutes before reequilibration to initial conditions.

Qualitative Peak Area Analysis.

HPLC chromatograms were analyzed using the Breeze 2 software (Waters Corporation, USA). UV absorbance was directly monitored over time at the ${\lambda }_{\max }$ for GTP, ATP, UTP, and CTP (252, 260, 262, and 272 nm, respectively). Nucleotide analyte identities were determined by extracting the UV spectrum at the respective retention time (${\mathsf{t}}_{\mathrm{R}}$) of the observed peaks and determining the ${\lambda }_{\max }$. Identities were further confirmed by injecting analyte standards to verify peak identities. Nucleotide concentrations were determined at 260 nm, so peaks were integrated to obtain peak areas using the traces collected at 260 nm. Peak areas were then used to determine the relative percent composition of the sample components observed at 260 nm. Percentages are given as the average of three injections.

Results

Single nucleotide addition of A, C, G, and U

To interrogate the impact of nucleotide identity on the mechanisms of a single nucleotide addition we performed experiments for each of the four cognate nucleotides, ATP, GTP, CTP, and UTP. This required a redesign of the template and primer compared to what we have previously reported for a single incorporation of an A [5]. The new template and primer sequences are shown in Figure 1A along with our previous template and primer design, labeled Appling et al.

The redesign of the template is because the original primer terminates at the 3’ end with a radiolabeled C incorporated by Pol I, see Materials and Methods. Thus, we redesigned the template DNA to allow for radiolabeling with ³²P-UTP for the examination of the incorporation of CMP to the 3’-end, see Figure 1A CMP Incorporation. Since it is possible that the kinetics could be impacted by the identity of the base at the 3’-end we designed a template to terminate at the 3’ end with UMP and encode for the addition of AMP, see Figure 1A AMP Incorporation. By doing this we can compare the kinetics of AMP addition to a 3’ terminal C vs. a 3’-terminal U see Figure 1A “Appling et al AMP incorporation” and “AMP incorporation”, respectively.

Assembly of the ECs requires the sequence of steps shown in Figure 1B, as previously reported [5, 6, 8, 10, 11, 14, 15, 18-22]. Briefly, ECs are assembled by allowing Pol I to incubate with the pre-annealed RNA:DNA_T. A non-template strand is added, followed by addition of α-³²P-CTP or ³²P-UTP, depending on the template. This results in RNA primers that terminate at the 3’ end with either C or U thereby yielding kinetics for nucleotide addition to different 3’ terminal ends. Pol I adds the radiolabel to the 3’-end for ten minutes followed by halting the reaction by addition of EDTA (Fig. 1B). This EC containing a 3’ radiolabeled RNA is then loaded into Syringe 1 of a chemical quench flow across from Syringe 2, which includes the NTP of interest, Mg²⁺, and heparin. The contents of the two syringes are rapidly mixed within 2 ms. After mixing, the reaction proceeds for user defined amounts of time before mixing with 1 M HCl to instantaneously halt the reaction (Fig. 1C).

Time courses were collected at final mixing concentrations of 10, 20, 50, 100, 300, 500, and 1000 μM for ATP, GTP, CTP, and UTP. A representative gel image for an experiment performed with a final concentration of 10 μM UTP is shown in Figure 2A and 10 μM ATP, GTP, and CTP are shown in Supplemental Figures 1 - 3. As expected, the starting 10-mer disappeared to form the 11-mer and, in the absence of the next cognate nucleotide, Pol I catalyzed nucleolytic removal of a dimer from the 3’ end, in this case, ³²P-CU (labeled as dimer in Fig. 2A). The time courses for the formation of 11-mer and the dimer products are shown in Figure 2C.

Figure 2.

UMP incorporation time courses. (A) Representative gel image from a single nucleotide addition experiment catalyzed by WT Pol I in the presence of 10 μM UTP. The two left most lanes are time zero reactions, and the time course extends from 5 ms to 100 s. Only a 10-mer, 11-mer, and dimer are visualized on the gel. (B) Representative gel image from a single nucleotide addition experiment catalyzed by WT Pol I in the presence of 1 mM UTP. The two left most lanes are time zero reactions, and the time course extends from 5 ms to 100 s. Only a 10-mer, 11-mer, dimer, and 12-mer⁺ bands are visualized on the gel. (C) The density of each band in (A) and (B) is quantified and normalized to the amount of signal in each lane and the time zeros. These values are then used in a plot to show the amount of each product over time.

Surprisingly, at high [UTP], we observed extension products beyond the 11-mer (labeled as 12-mer+ in Figure 2 B and Supp. Fig. 4A), which were unexpected since no ATP was added. In fact, extension products beyond the 11-mer were always observed for time courses collected at [UTP] ≥ 300 μM, (gel images not shown). Interestingly, additional extension products were detected with [ATP] ≥ 300 μM, [GTP] ≥100 μM, and [CTP] = 1 mM. Supplemental Figure 4B - D show example gel images for experiments carried out with 1 mM ATP, GTP, or CTP, respectively.

Based on the template design, there are two possible explanations for Pol I to continue to extend beyond the 11-mer when only one nucleotide is provided. First, low concentrations of contaminating nucleotides in the NTP stocks could cause extension beyond the 11-mer. Second, Pol I could be misincorporating under conditions where high concentrations of NTP are provided.

To distinguish between these two possibilities, we first sought to test for the presence of contaminants in the nucleotide stocks. As a result, we performed HPLC on each nucleotide stock, see Materials and Methods. Supplemental Figure 5A - D show the chromatograms for 1 mM UTP, ATP, GTP, and CTP, respectively. Supplemental Table 1 shows the results of the quantitative analysis of the chromatograms.

Indeed, the UTP stock contains trace amounts of ATP (~0.2 %) but no detectible GTP is present. Both contaminating ATP and GTP would need to be present to reconcile extension to the 19-mer as observed in Supplemental Figure 4A, see template strand in Figure 1A. The GTP stock also contains ~0.06% ATP, which may account for the observation of additional extension products since the template encodes for incorporation of AMP and GMP, see GMP incorporation template in Figure 1A. However, the ATP stock does not exhibit any detectible contaminating nucleotides, see Supplemental Figure 5B and Supplemental Table 1. Yet, extension to 12-mer is observed at high [ATP]. Similarly, the CTP stock does not contain detectible NTP contaminations (Supp. Fig 5D) yet extension to 12-mer is still detected at elevated nucleotide concentrations, see Supplemental Figure 4D.

In summary, for ATP and CTP the emergence of a 12-mer is consistent with misincorporation. Whereas, for GTP and UTP the observed extension beyond 12-mer is consistent with trace amounts of contaminating nucleotide. However, it cannot be ruled out that misincorporation also occurs when contaminating NTPs are present. To further confound the interpretation, misincorporation and extension due to low levels of contaminating NTP are predicted to exhibit the same nucleotide concentration dependence. Thus, from the analysis of the HPLC results, coupled with the quenched-flow experimental results, we conclude that there is a mixture of both misincorporation and trace contaminating NTPs.

Accounting for unexpected extension products

Since misincorporation and extension due to trace contaminating NTPs may both be occurring in the same reaction, we concluded that further purification of the stocks is not likely to completely solve the problem of observing unexpected extension products. We sought to determine if we could account for the additional extension products and extract meaningful kinetic parameters for a single nucleotide addition in the presence of unexpected extension.

Because of the sequential nature of this reaction, all products beyond the 11-mer were first 11-mers. We reasoned by conservation of mass that the sum of all counts of 11-mer and above would represent a time course for the 11-mer in the absence of any additional extension. We have shown, by extensive re-examination of previously published multi-nucleotide addition time courses, that summing the signal from longer products into the signal for the 11-mer reproduces a single nucleotide addition time course [23]. Therefore, in this study, when products longer than an 11-mer were detected, they were added to the counts for the 11-mer.

Determination of the minimum number of steps in the nucleotide addition cycle

With a method to account for unexpected extension products in hand we sought to determine the mechanism that best describes single nucleotide addition for each of the four cognate nucleotides. Time courses collected over a range of UTP from 10 μM to 1 mM were subjected to nonlinear least squares (NLLS) using sums of exponentials, see Figure 3A, 3B, and Materials and Methods. This was done to determine the minimum number of steps required to define a reaction mechanism for the incorporation of a single UMP. A sum of two exponentials, Equation 2, was required to describe the rise and fall of the 11-mer population, see Figure 3A. This analysis yielded two observed rate constants, 11-mer ${\mathsf{k}}_{\text{obs},1}$ and 11-mer ${\mathsf{k}}_{\text{obs},2}$, plotted in Figure 3C and D, respectively. The formation of the dimer species was described by a single exponential function given by Equation 3 yielding a single observed rate constant, Dimer ${\mathsf{k}}_{\text{obs}}$ plotted in Figure 3E.

Figure 3.

Model independent analysis of Single nucleotide addition time courses of UMP. (A) Time courses were collected at seven concentrations of UTP indicated in the legend. Time courses were fit to a sum of two exponentials, Eq. 2, and the best fit lines are shown. (B) The dimer was fit to one exponential, Eq. 3, and the best fit lines for each concentration are shown. (C) Secondary plot of the faster of the two ${\mathsf{k}}_{\text{obs}}$ values from fitting the 11-mer was plotted against [UTP]. This ${\mathsf{k}}_{\text{obs}}$ showed nucleotide dependence, and was described using a rectangular hyperbola, Eq. 4. (D) The second, slower ${\mathsf{k}}_{\text{obs}}$ from 11-mer fits was plot against [UTP] and well described by a flat-line. (E) The ${\mathsf{k}}_{\text{obs}}$ from dimer fits was plot against [UTP] and well described by a flat line.

The observed rate constant for the formation of the 11-mer depends, hyperbolically on the [UTP] and was subjected to NLLS using Equation 4, see Figure 3C. The hyperbolic dependence indicates that nucleotide binding is in rapid equilibrium with the next step and the step is kinetically coupled to nucleotide binding. From this analysis the maximum observed rate constant extrapolated to infinite [UTP], ${\mathsf{k}}_{\text{obs},\max }=(800\pm 200){\mathrm{s}}^{\text{-}1}$, and the [UTP] to achieve half ${\mathsf{k}}_{\text{obs},\max }$, ${\mathsf{K}}_{1∕2}=(200\pm 100)\mu \mathsf{M}$.

UMP incorporation is surprisingly fast. The maximum observed rate constant of ~800 s⁻¹ predicts a half-life of, ${\mathsf{t}}_{1∕2}$ ~0.9 ms, which is outside of the millisecond temporal resolution of the chemical quench flow. Moreover, it is faster than what we previously detected for AMP incorporation (~ 260 s⁻¹, ${\mathsf{t}}_{1∕2}$ ~3 ms) [5]. Consistently, for experiments collected in the presence of [UTP] ≥ 300 μM most, ~65%, of the reaction has occurred before collection of the first time point of 5 ms and likely within the mixing time of 2 ms.

Figure 3D shows the second observed rate constant, 11-mer ${\mathsf{k}}_{\text{obs},2}$, describing the disappearance of the 11-mer, was independent of [UTP], which is expected for a unimolecular reaction. The average value, $<{\mathsf{k}}_{\text{obs}}>=(1.7\pm 0.1){\mathrm{s}}^{\text{-}1}$ was determined over all [UTP]. Consistently, from the analysis of the time courses describing the dimer formation, Figure 3B, the observed rate constant is independent of [UTP] with an average value of $<{\mathsf{k}}_{\text{obs},\text{dimer}}>=(2.0\pm 0.1){\mathrm{s}}^{\text{-}1}$, see Figure 3E. The agreement between the rate constant describing the disappearance of 11-mer and formation of the dimer indicates that these are the same processes and represent the rate constant for nucleolytic cleavage. Interestingly, for UTP, the disappearance of the 11-mer is ~5-fold faster than what we previously reported for the cleavage step in the presence of ATP.

Experiments were also performed to examine A, G, and C incorporations, and example gels and plots can be seen in Supplemental Figures 1-3. The same model independent analysis was carried out for A, G, and C. Supplemental Figures 6-8 show the time courses, results of NLLS analysis, and observed rate constants vs. [NTP].

The resultant kinetic parameters for all four nucleotides are shown in Table 1. Except for UMP incorporation, the other three nucleotides incorporate with a similar observed rate constant (~320 s⁻¹), have a similar binding affinity (~70 μM), and perform cleavage within the same order of magnitude (~0.2 s⁻¹). In contrast, UMP incorporation has significantly different kinetic parameters of ~800 s⁻¹, ~200 μM, and ~1.8 s⁻¹ for nucleotide incorporation, binding affinity, and cleavage, respectively.

Table 1:

Model independent parameters from each nucleotide incorporation.

Parameter	AMP (Appling et al.) add to C	UMP Add to C	AMP Add to U	GMP Add to C	CMP Add to U
11-mer
${\mathsf{k}}_{\text{obs},\max }({\mathrm{s}}^{\text{-}1})$	270 ± 30	800 ± 200	290 ± 20	340 ± 50	340 ± 10
${\mathsf{K}}_{1∕2}(\mu \mathsf{M})$	170 ± 30	200 ± 100	60 ± 20	80 ± 40	69 ± 9
$<{\mathsf{k}}_{\text{obs},2}>({\mathrm{s}}^{\text{-}1})$	0.27 ± 0.02	1.7 ± 0.1	0.088 ± 0.001	0.4 ± 0.2	0.25 ± 0.02
Dimer
$<{\mathsf{k}}_{\text{obs}}>({\mathrm{s}}^{\text{-}1})$	0.40 ± 0.02	2.0 ± 0.1	0.094 ± 0.002	0.15 ± 0.02	0.26 ± 0.02

The maximum observed rate constant for AMP incorporation on the template used here is in good agreement with the value determined with the template previously reported in Appling et. al, see Table 1 [5]. However, the rate constant defining the cleavage step is ~4-fold faster and binding is ~3-fold weaker using the Appling et al. template compared to the newly designed template. Recall, the previous template encodes for the addition of A to a C, whereas this template encodes for the addition of A to a U. Thus, the differences in the kinetics of cleavage suggests that the rate of cleavage may depend on the sequence of the last two nucleotides at the 3’ end, i.e. ³²P-CA vs ³²P-UA. Further, the strength of the interaction between the template and the next cognate nucleotide may depend on the identity of the nucleotide at the 3’ end.

Elucidation of mechanism describing single nucleotide addition

To identify the elementary kinetic steps and reveal the presence of additional reaction intermediates, we sought to determine a mechanism that can simultaneously describe all time courses for 11-mer and dimer as a function of a given [NTP]. The model independent analysis (exponential fitting) reveals the minimum number of steps in a reaction and places bounds on a global kinetic mechanism. However, each observed rate constant is a convolution of elementary steps in a complete cycle of nucleotide addition and thus there is additional information in the time courses to be extracted.

Here we have constructed normalized fraction of RNA as a function of time by adding all the counts on the gel image for the 11-mer and 12-mer+, see Figure 3A and 3B. Although we have validated this approach [23], cleavage of an extension product beyond 11-mer will impact the time courses for both the 11-mer and the dimer. This is because when an extension product longer than the 11-mer is cleaved, the radiolabel remains in the product and does not produce a detectable radiolabeled dimer. For example, a 14-mer is cleaved to a radiolabeled 12-mer and an undetectable dimer. Upon adding the 12-mer counts to the 11-mer counts, this phenomenon is represented in the time courses as a steady-state population of the 11-mer. This is best illustrated for 1 mM UTP in Figure 3A where the time course plateaus at ~0.18 after about 1 s instead of approaching zero like the other time courses in that panel. This is also the explanation for a decrease in total amplitude of the dimer as a function of increasing [UTP], see Figure 3B, which is phenomenon that we did not observe in our initial study on A incorporation [5].

Logically, we started the analysis of the time courses collected with UTP using the reaction mechanism previously reported to describe a single A incorporation [5]. However, we included a step describing irreversible formation of 11-mer governed by rate constant, ${\mathsf{k}}_6$, in Supplemental Scheme 1, Supplemental Figure 9. This is because of the steady state population of 11-mer described in the preceding paragraph. However, this step is a local parameter since it does not become significant until 300 μM UTP.

Supplemental Scheme 1 did not adequately describe all time courses as a function of [UTP], see Supplemental Figure 9. In particular, the best fit line deviates from the time points describing 10 μM UTP. Equally important, the rate constant, ${\mathsf{k}}_5$, floated to large values indicating that it is not contributing to the fit. In this analysis, when a rate constant floats to a value of ~1x10⁵ s⁻¹ (${\mathsf{t}}_{1∕2}$ ~7 μs) then this is undetectably fast and is not contributing to the fit. Thus, Supplemental Scheme 1 was simplified to Scheme 1 by removing the conformational change denoted as (EC₁₁)₃.

Scheme 1.

The best fit of the UMP incorporation time courses to Scheme 1 is shown in Figure 4A. As shown in the 11-mer panel, the 10 μM UTP time course is better described by Scheme 1 than Supplemental Scheme 1 and, in addition, all dimer time courses are better described by Scheme 1 than Supplemental Scheme 1. It’s important to note that in Supplemental Scheme 1, ${\mathsf{k}}_{\text{np},\mathsf{r}}$ and ${\mathsf{k}}_4=0$. However, during the fitting of the time courses using Scheme 1 the productive to non-productive conformational change and, more importantly, the bond formation step exhibited reversibility, see ${\mathsf{k}}_{\text{np},\mathsf{r}}$ and ${\mathsf{k}}_4$, respectively in Scheme 1 Figure 4A.

Figure 4.

Model Dependent fits of UMP incorporation for Supplemental Scheme 1 and Scheme 1. (A) Both the 11-mer and dimer were globally fit to Supplemental Scheme 1 using MENOTR. The chi² value was (3.2 ± 0.5) x 10⁵. The kinetic parameters of (B) AMP, (C) GMP, and (D) CMP incorporation catalyzed by Pol I when fit to Scheme 1.

The simpler model, Scheme 1, was found to be an improvement over Supplemental Scheme 1 for not only U incorporation, but also for A, G, and C incorporations (Supplemental Figures 10 - 12). Thus, we tested if Scheme 1 could better describe the Appling et al. AMP incorporation results (Supp. Fig. 13) [5]. Upon inspection, Scheme 1 describes the data well, and the majority of the parameters are the same as previously reported [5]. This re-fitting of the AMP incorporation set of time courses shows that Scheme 1, the simpler model, can describe all single nucleotide addition time courses that we have collected using Pol I and shows Scheme 1 is a global model across all five sets of time courses.

Overall, these results show that when using the same DNA non-template strand and, as similar as possible, template DNA strand and RNA primers, the four nucleotides are incorporated into the growing RNA chain using the same kinetic mechanism. However, U is incorporated much faster than the other three nucleotides. Further, both U and G exhibit comparable but weaker binding than C and A. These observations are consistent with our previous multi nucleotide addition studies [10-12] that revealed large variability in the rate constants when using only A and G.

Discussion

One objective of our work has been to determine the elementary kinetic mechanisms describing single and multi-nucleotide additions catalyzed by Pol I. To date, our evidence supports a mechanism where cognate nucleotide binding is in rapid equilibrium and is immediately followed by rate limiting bond formation [5, 8, 10, 24]. This bond formation step occurs through an ${\mathrm{S}}_{\mathrm{N}}2$ substitution reaction where the 3’ hydroxyl group of the ribose sugar attacks the α-phosphate of an incoming nucleotide. Consequently, the identity of the base would not be predicted to impact the elementary rate constant defining bond formation. Nevertheless, we observe that U was incorporated substantially faster than the other three nucleotides.

Regardless of data analysis approach, model fitting, etc., a simple visual inspection of the time course for U incorporation, shown in Figure 4, reveals that nearly 70% of the reaction is complete within the first 5 ms for the highest concentrations of UTP. This revealed that the reaction for addition of a U is much faster than what we previously reported for the incorporation of an A. If bond formation is rate limiting and the base identity is not involved in catalysis, then what is the reason for the variability in the rate constants?

We observe that the trend for fastest to slowest is U > G > C > A. Perhaps a more stably bound cognate nucleotide allows the enzyme to catalyze bond formation faster. If this were the case, then one would expect G and C to be faster than A and U. However, base stacking interactions also play a role in stability. Here, since we use the polymerase for radio labeling, we are not examining single nucleotide addition to the same nucleotide at the 3’ end of the RNA. So, listing the trend of the last two nucleotides may be the better way to think about the problem and, from this study, this would be (CU) > (CG) > (UC) > (UA) > (CA). However, no clear trend currently emerges, and substantially more nucleotide pairs may be required to fully reveal the rules.

Accelerated incorporation of UMP may play an important role in termination. Although Pol I exhibits protein dependent pausing, we have also shown that Pol I favors pausing when incorporating multiple UMPs [25, 26]. Transcription through A-rich template DNA tracts may result in destabilization of the transcription elongation complex, which may favor pause and collapse.

Interestingly, the rate of removal of a dinucleotide from the 3’ end is also influenced by the base identity. Again, removal of a dinucleotide with a 3’ terminal U is the fastest and the trend for endonuclease activity listed fastest to slowest is (CU) > (CA) > (UC) > (CG) > (UA). Although the reaction is different, Hein et al reported a similar trend for pyrophosphorolysis catalyzed by prokaryotic RNA Polymerase (RNAP). They showed that U at the 3’ end was removed by pyrophosphorolysis faster than any other nucleotides [27]. Further, they showed this phenomena to be evolutionarily conserved by comparing Thermus thermophilus RNAP, E. coli RNAP, S. cerevisiae Pol II, and Bos taurus Pol II [27].

Here we have shown that the nucleotide concentration dependence for all four nucleotides can be described by the same number of kinetic steps. Because the binding step is in rapid equilibrium, we only acquire constraints on the ratio of the binding and dissociation rate constants, i.e. ${\mathrm{K}}_{\mathrm{D}}$. The ${\mathrm{K}}_{\mathrm{D}}$ for all four nucleotides varies between ~80 to 170 μM. Although this is as much as a factor of two, an obvious trend does not emerge. The weakest binding is for U and G, followed by C, when one may expect G and C to exhibit the tightest binding. It will be of interest to determine if these subtle differences emerge from either the association or dissociation rate constants. Such an examination can speak to base selectivity.

Ribosome biogenesis is coupled to the rate of transcription catalyzed by Pol I and implicated in several human diseases [22, 28-33]. Thus, it is critical to gain a complete understanding of the mechanisms that control transcription by Pol I. Previous work has shown that the upstream DNA sequence impacts both prokaryotic RNAP and Pol II [27, 34]. We have recently shown that downstream sequence directly impacts the kinetics of transcription elongation prior to the sequence entering the active site of Pol I both in vivo and in vitro [20]. Here we have shown that the identity of the cognate nucleotide and the dinucleotide pair at the 3’ end also impacts the rate of transcription. Going forward there is a need to examine a sufficiently large number of incorporations to deconvolute the rules governing these effects.

$${\mathit{Frac}}_{\mathit{i}}(\mathit{t})=\frac{\frac{[{\mathit{RNA}}_{\mathit{i}}](\mathit{t})}{{\sum }_i[{\mathit{RNA}}_{\mathit{i}}](\mathit{t})}-\frac{[{\mathit{RNA}}_{\mathit{i}}](0)}{{\sum }_i[{\mathit{RNA}}_{\mathit{i}}](0)}}{1-\frac{[{\mathit{RNA}}_{\mathit{i}}](0)}{{\sum }_i[{\mathit{RNA}}_{\mathit{i}}](0)}}$$ Equation 1

$$Frac={\alpha }_1(1-e^{-k_{obs,1}})+{\alpha }_2(1-e^{-k_{obs,2}})$$ Equation 2

$$Frac={\alpha }_1(1-e^{-k_{obs,1}})$$ Equation 3

$$k_{obs}=\frac{(k_{obs,max})\times [NTP]}{K_{1∕2}+[NTP]}$$ Equation 4

Table 2:

Kinetic Parameters from Global Analysis using Scheme 1

Parameter	AMP (Appling et al.) Add to C	UMP Add to C	AMP Add to U	GMP Add to C	CMP Add to U
${\mathsf{k}}_{\text{np},\mathsf{f}}({\mathrm{s}}^{\text{-}1})$	0.075 ± 0.006	0.16 ± 0.06	20 ± 30		0.07 ± 0.01
${\mathsf{k}}_{\text{np},\mathsf{r}}({\mathrm{s}}^{\text{-}1})$		0.02 ± 0.01	30 ± 50		0.002 ± 0.001
${\mathsf{K}}_{\mathrm{D}}(\mu \mathsf{M})$	96 ± 6	170 ± 20	80 ± 30	150 ± 30	106 ± 6
${\mathsf{k}}_3({\mathrm{s}}^{\text{-}1})$	182 ± 4	630 ± 50	319 ± 8	400 ± 60	370 ± 10
${\mathsf{k}}_4({\mathrm{s}}^{\text{-}1})$		9 ± 1	2 ± 1	3 ± 2	7 ± 2
${\mathsf{k}}_5({\mathrm{s}}^{\text{-}1})$	0.35 ± 0.02	2.2 ± 0.1	0.092 ± 0.004	0.110 ± 0.002	0.301 ± 0.006
${\mathsf{k}}_6({\mathrm{s}}^{\text{-}1})\text{-}100$				0.008 ± 0.006
${\mathsf{k}}_6({\mathrm{s}}^{\text{-}1})\text{-}300$		0.13 ± 0.07	0.015 ± 0.005	0.031 ± 0.004
${\mathsf{k}}_6({\mathrm{s}}^{\text{-}1})\text{-}500$		0.2 ± 0.1	0.026 ± 0.006	0.07 ± 0.02
${\mathsf{k}}_6({\mathrm{s}}^{\text{-}1})\text{-}1000$		0.51 ± 0.07	0.046 ± 0.003	0.15 ± 0.02	0.03 ± 0.01
$[{\mathsf{EC}}_{10}]∕[{\mathsf{EC}}_{10,\text{total}}]$	0.77 ± 0.03	0.89 ± 0.01	0.91 ± 0.02		0.82 ± 0.02

Highlights.

• Pol I employs one universal kinetic mechanism independent of base identity

• UMP is incorporated substantially faster than GMP, CMP, and AMP

• Fast incorporation of UMP may be important for termination

Abbreviations

EC

Elongation complex

Pol

RNA polymerase

rRNA

Ribosomal RNA

DNA_t

Template DNA