Dissecting the multistep reaction pathway of an RNA enzyme by single-molecule kinetic “fingerprinting”

Liu et al. 10.1073/pnas.0610597104.

Supporting Information

Files in this Data Supplement:

SI Figure 7
SI Figure 8
SI Figure 9
SI Figure 10
SI Text

SI Figure 7

Fig. 7. Schematic secondary structures of the 2WJ and 4WJ hairpin ribozymes. The hairpin ribozyme consists of two helix-loop-helix domains. The two domains may be connected by a 2WJ in the minimal form or a 4WJ in its natural form. The AC₅ loop connecting the 2WJ shown here as the red circular loop may be removed (as in the nicked form of the 2WJ ribozyme) without substantially affecting the conformational dynamics and catalytic activity of the enzyme (1).

1. Bokinsky G, Rueda D, Misra VK, Rhodes MM, Gordus A, Babcock HP, Walter NG, Zhuang X (2003) Proc Natl Acad Sci USA 100:9302-9307.

SI Figure 8

Fig. 8. FRET histograms of the 2WJ hairpin ribozyme. (A) FRET histograms constructed from different 20-s time intervals in the FRET time trace such as the one shown in Fig. 1C, with 0 s as the time point when Mg²⁺ was added. From top to bottom: -20 to 0 s, 0 to 20 s, 100 to 120 s, 300 to 320 s, 600 to 620 s, and 1,200 to 1,220 s. The evolution of FRET histograms clearly links the FRET levels to different physical states: before addition of Mg²⁺, the ribozyme exhibits a FRET level averaged at 0.18; after addition of Mg²⁺, two new FRET levels appears, with the 0.30 FRET level representing the undocked states (U_L and U_C) and the 0.76 FRET level representing the docked states (D_L and D_C); as time further increases, a FRET = 0.18 peak surfaces again, but this time due to cleavage and product release (the P_R state). As expected, when a high concentration of 3′P was added to the buffer to prevent product dissociation or when a ribozyme with a noncleavable substrate was probed, the FRET level of 0.18 does not appear in the presence of Mg²⁺ (Fig. 1E). (B) FRET histograms before and after release of 3′P. For each trace, a single transition event from 0.76 to 0.18 FRET was identified as the 3′P release event. In the before-product-release histogram, two peaks at FRET = 0.30 and FRET = 0.76 exist, representing the undocked (U_L and U_C) and docked (D_L and D_C) states. In the after-product-release histogram, only a single peak at FRET = 0.18 exists, representing the product-released state (P_R). The transitions from 0.76 to 0.18 FRET were identified by visual inspection or by HMM analysis (see SI Fig. 9 for details of this analysis method), both methods yielding the same answer.

SI Figure 9

Fig. 9. Analysis of the binding and dissociation kinetics of the cleavage product 3′P. 3′P with fluorescence quencher dabcyl (3′P-D) was added to the immobilized and cleaved ribozyme in the presence of Mg²⁺. (A) A representative time trace of Cy3 and Cy5 fluorescence and of FRET for a single ribozyme molecule in the presence of 500 nM 3′P-D. The trace was recorded at the frame rate of 33 fps. In the undocked state, 3′P-D stochastically bound to and dissociated from the ribozyme, leading to repetitive quenching and dequenching of the Cy5 signal, whereas the signal from the more distant Cy3 stayed large. The FRET trace fluctuates between a FRET ˜ 0.2 level, as expected for the P_R state, and a FRET ˜ 0 level due to quenching of Cy5 in the 3′P-D bound state. In the docked state, 3′P-D remained bound, quenching signals from Cy3 and Cy5, now both in proximity to the quencher. (B) (Upper) Analysis of the Cy5 fluorescence trace (the undocked portion, black) by a threshold fitting method. We assigned quenched and unquenched states by setting two threshold levels: a transition from quenched to unquenched state is identified as the Cy5 fluorescence increases to larger than threshold 1 and a transition from unquenched to quenched state is identified as the Cy5 fluorescence decreases to smaller than threshold 2. The transition assignment is indicated by the blue line. (Lower) Analysis of the same Cy5 fluorescence trace by the HMM method (1). Based on probability only, this method determines the most likely FRET values of underlying states and the most likely time sequence of those states, which may be originally hidden in the FRET trajectories because of noise. The transition assignment is indicated by the pink line. The two analysis methods yield nearly identical transition assignments. From these two methods we got nearly identical product binding and dissociation rate constants. For 500 nM 3′P, the threshold fitting method reported k_off = 2.6 s^-1 and k_on(obs) = 2.8 s^-1; the HMM method reported k_off = 2.7 s^-1 and k_on(obs) = 2.9 s^-1.

1. Mckinney SA, Joo C, Ha T (2006) Biophys J 91:1941-1951.

SI Figure 10

Fig. 10. Heterogeneous undocking kinetics of the hairpin ribozyme. (A) Representative FRET time traces (1 fps) of three ribozyme molecules at equilibrium. The majority of molecules (60%) display slow undocking kinetics as exemplified by the data in Top, while the other molecules display significantly faster undocking kinetics or undetectable docking, as shown in Middle and Bottom. This is consistent with our previous observation of heterogeneous undocking kinetics, in which a major population (61%) docks most stably and a few minor populations show one or more orders of magnitude faster undocking rates (1). (B) Histogram of docking equilibrium constant (K) constructed from many individual molecules at equilibrium. The value K is defined as the ratio between total docked time and total undocked time for each trace (i.e. each molecule). A total of 989 single-molecule FRET traces that lasted at least 600 s before photobleaching were used to construct the histogram, which is plotted in a semilog scale since ln K scales with the free energy difference between the docked and undocked states. Three peaks are shown in the histogram: the peak centered at 1.2 corresponds to the major population; the peaks centered at -2.5 and at infinity (arising from the traces without detectable docking event) correspond to the minor populations. To isolate the major population, we only picked the FRET traces with K > 0.5 (ln K > -0.7). From the multi-Gaussian fit of the histogram (green line), we calculated the misidentification error of this selection criterion to be ~5%, which includes contributions from both missing traces in the major population and misidentifying traces in the minor populations as the major population due to stochasticity. Based on this criterion, we identified 595 FRET traces as belonging to the major population. The overall equilibrium constant between the docked and undocked states is 3.2, very close to the center of the major peak in this histogram (ln 3.2 = 1.16). (C) After all the rate constants of the major population were determined (Fig. 6A), we simulated 200 FRET traces (600 s long) at equilibrium (with 30 μM 3′P) and constructed a histogram of ln K (black bars). The histogram of ln K for all the molecules obtained experimentally is superimposed as red line. The simulated histogram agrees well with the major peak of the experimental histogram, indicating that our selection criterion correctly identifies the major population.