FRET efficiency distributions derived from HMM

The FRET efficiencies in the different conformations of the domain sensor in the presence of ADP and ATP resolved by the HMM analysis are compared to FRET efficiencies of substrate-bound Ssc1 in Figure 6.13. For the spFRET measurement of substrate-bound Ssc1(341,448), the protein was encapsulated together with ADP and P5, a short peptide acting as a model substrate for Ssc1 (Section 6.3.2). It was shown before using single-molecule burst analysis (Section 6.3.1) that Ssc1 assumes a defined, narrow distance distribution with undocked domains and a closed lid in the substrate-bound state. The spFRET TIRF assay using the encapsulated Ssc1 domain sensor yielded a FRET efficiency of 26% with bound substrate (Figure 6.13A and Tab. 6.4). This can be related to the distance between the dyes using a Förster radius of 6.1 nm (Mapaet al., 2010), resulting in a 7.3 nm separation in the substrate-bound state in excellent agreement with the previous measurements in solution (Table 6.4). This is a good indication that there are no artifacts introduced by possible interactions with the lipids or other encapsulation-related effects.

3

The increased width of theE≈50% FRET state in the measurement of Ssc1(341,448) with ATP is probably due to the short dwell time of this state (refer to Figures 6.10 and 6.17) and the resulting low statistics for this state. The correlation beteen the initial and final FRET values (lower initial FRET (∼90%) transitions to lower final FRET (∼40%) and vice versa) may be due to erroreous determination of theγ factor. Theγfactor was determined for most traces by the dynamic method (refer to Section 4.3.5). The performance of the dynamicγdetermination may suffer from the low percentage of time in the∼50% FRET state.

A B

Domain sensor + ADP

0 10 20 30 Intensity [a.u.] 1 1.5 2 0 0.5 1 FRET Time [s] C D

Domain sensor + ATP Lid sensor + ADP

Figure 6.12:Hidden markov modeling and transition density plots of smTIRF time traces of Ssc1. (A) FRET efficiency trace (blue) with corresponding idealized trace (red) from the HMM analysis to identify the FRET efficiencies and transition rates underlying the experimental data. Donor and acceptor intensity (green and red, upper panel) and FRET efficiency (lower panel) are shown over time. The experimentally measured FRET efficiency (blue) is overlaid with the result of the HMM analysis (red). (B-D) Transition density plots for Ssc1(341,448)/ADP (B), Ssc1(448,590)/ADP (C), and for Ssc1(341,448)/ATP (D) constructed from the transitions found by the HMM analysis.

The two conformations observed for the domain sensor in the presence of ADP as well as the two conformations observed in the presence of ATP are clearly different from the substrate-bound confor- mation. While the respective high FRET populations have clearly different mean FRET efficiencies, the low FRET subpopulations observed in the presence of ATP and ADP have surprisingly similar average FRET values. The greater width of the low-FRET state with ATP than with ADP is due to the increased statistical uncertainty because of the low number of frames the molecule spends in this conformation. At this point, it could be speculated that it is indeed the same conformation which is assumed both in the presence of ADP and ATP, and Ssc1 cycles between this conformation and another nucleotide dependent conformation.

For the Ssc1 lid sensor, we measured a FRET efficiency of 13% in the ATP state and 73% in the substrate-bound state, corresponding to 8.4 nm and 5.2 nm, respectively (Figure 6.13B). This is also in good agreement with our previous results (Table 6.4). In the experiments with bound substrate, we observed a subpopulation (19%) of molecules showing identical switching behavior as in the presence of ADP alone. Although the bulk concentration of substrate peptide was high enough to ensure the presence of several substrates in each vesicle, the local concentration inside the vesicles may have been lower due to incorporation of the hydrophobic substrate peptide into the lipid bilayer. We thus attribute this subpopulation to vesicles containing no available substrate and excluded them from the histogram for the substrate-bound state. Apart from these molecules, we did not observe any dynamics in these measurements as expected from the narrow distance distributions in the ATP and ADP/P5 states (Mapaet al., 2010).

The two conformations of the SBD observed by the lid sensor resolved by the HMM analysis in the presence of ADP are clearly distinct from the conformations observed in the presence of ATP and when substrate is bound (Figure 6.13B). While the presence of ATP results in a widely open lid (E ≈ 13%) and substrate binding triggers lid closure (E ≈ 74%), in the presence of ADP the SBD cycles between an intermediate state with∼ 37%FRET efficiency and a tightly closed state displaying a FRET efficiency even higher than in the substrate-bound state (E≈90%).

Table 6.4:FRET efficiencies and distances compared for TIRF and MFD-PIE experiments.

FRET efficiency [%] Distance [nm]

TIRF1 TIRF2 MFD-PIE

Ssc1(341, 488) ADP High-FRET 81.1±0.2 4.8 (4.2;5.6) 4.4 ADP Low-FRET 50.2±0.9 6.1 (5.3;7.1) 6.2 ATP High-FRET 90.6±0.8 4.2 (3.6;4.8) 4.3 ATP Low-FRET 47.8±3.1 6.2 (5.3;7.3) 6.1 ADP/P5 26.5±0.6 7.2 (6.3;8.4) 7.5 Ssc1(488, 590) ADP High-FRET 88.7±0.9 4.3 (4.1;4.8) 4.3 ADP Low-FRET 37.1±1.3 6.7 (6.2;7.3) 6.8 ATP 12.8±0.9 8.4 (7.8;9.4) 7.73 ADP/P5 73.5±0.8 5.1 (4.8;5.7) 5.2 1

Error estimated from covariance matrix of the Fit)

2 _{Maximum error fromκ}2_andEerror) 3

The relatively large deviation between the MFD-PIE data and the TIRF data for Ssc1(448,590)/ATP is possibly due to the high FRET efficiency in this state. The highE results in large errors in the distance determination due to the_R16 dependence.

0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 0.5 FRET efficiency [%]

Normalized histogram count

ADP HF state (Mean= 81 %, SD=3 %) ATP LF state (Mean= 48 %, SD=15 %) ATP HF state (Mean= 90 %, SD=4 %) Peptide bound state (Mean= 26 %, SD=6 %)

0 10 20 30 40 50 60 70 80 90 100 0 0.1 0.2 0.3 0.4 FRET efficiency [%]

Normalized histogram count

ADP LF state (Mean= 37 %, SD=8 %) ADP HF state (Mean= 89 %, SD=7 %) ATP state (Mean= 13 %, SD=8 %) Peptide bound state (Mean= 73 %, SD=5 %)

Figure 6.13:SpFRET histograms derived from the HMM analysis of Ssc1. (A) FRET efficiency histograms of the two inter-domain conformations identified by the HMM analysis in the presence of ADP (dark/light red) and in the presence of ATP (dark/light blue), together with the molecule-wise FRET histogram of the domain sensor in the presence of ADP and P5 (green). (B) FRET effi- ciency histograms of the two SBD conformations identified by the HMM analysis in the presence of ADP (dark/light red) and molecule-wise FRET histograms of the SBD conformation in the presence of ATP (blue) and in the presence of ADP and P5 (green). (A,B) Single Gaussian fits to the data are shown by dashed lines. The average FRET value and the standard deviation are given in the figure legend.