1. Single-molecule fluorescence spectroscopy maps the folding landscape of a large protein Menahem Pirchi, Guy Ziv, Inbal Riven, sharona sedghani Cohen, nir Zohar, Yoav Barak & Gilad Haran Gong, Ping Department of Chemistry and Biochemistry University of Delaware

2. Small single-domain protein Two-state folding behavior Smooth Minimize the number of intermediates and kinetic traps Large protein with multiple domains More than 70% of eukaryotic proteome Metastable intermediates Complex folding pathway

3. Single-molecule fluorescence resonance energy transfer spectroscopy (smFRET) 2 Reliable "ruler" for measuring structural changes in proteins Dispelling unknown orientation factor From PubMed: http://www.ncbi.nlm.nih.gov/pubmed/7577238

4. 3 Adenylate kinase Labeled at position 73 and 203 Encapsulated within lipid vesicles Series of GdmCl concentrations Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

5. In lack of a very long single-molecule temporal trajectory that maps the whole landscape Multiple short trajectories collected in experiment Availability of a large number of equilibrium trajectories facilitates reconstruction of the folding landscape Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

6. (a) Three examples of fluorescence trajectories of individual AK molecules Single-molecule FRET trajectories of AK Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

7. (b) Comparison of single-molecule results to the bulk denaturation curve (c) Comparison of the probability distribution of FRET efficiency values obtained from single-molecule trajectories to a free- diffusion single-molecule experiment Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011). GdmCl concentration: 0.65M

8. Change-point analysis of trajectories Transition density map constructed from the 0.65M GdmCl data set Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

9. HMM reveals six states HMM: Hidden Markov Model State is not directly visible Output (dependent on the state) is visible The sequence of tokens generated by an HMM gives some information about the sequence of states From wikipedia: http://en.wikipedia.org/wiki/Hidden_Markov_model

10. The dynamics obey detailed balance —the flux from i to j equals the flux from j to i. Add an extra state presenting the photobleached molecules. Baum-Welch algorithm is used to obtain a maximum likelihood estimate of the HMM parameters.

11. Correlation between the transition density map based on change-point analysis and maps based on the HMM analysis Focusing on the data set taken at 0.65 M GdmCl Repeat the HMM analysis for different values of N, from 2 to 14. Use the HMM parameters to generate a transition map cross-correlated this map with the one obtained from the change- point analysis. optimal N is between 5 to 7 Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

12. State probability distribution histograms, as a function of GdmCl concentration State connectivity changes with denaturant concentration Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

13. Transition maps at three indicated GdmCl concentrations, constructed from the experimental data using Hmm analysis results. As the concentration of denaturant increases ： more transitions tend to occur between states of lower FRET efficiency the fraction of sequential transitions of the type i→i ± 1 increases significantly Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

14. One-dimensional projections of the folding landscape of AK at the three indicated GdmCl concentrations The widths of the lines depict the relative productive flux flowing between each pair of states their colours represent the transition rates. Pirchi, M. et al. Nat. Commun. 2:493 doi:10.1038/ncomms1504 (2011).

15. Discussion Single-molecule FRET spectroscopy can provide a comprehensive description of the folding landscape of a large, multidomain protein The dynamics involve a large set of possible pathways on the landscape Provides the experimental means to characterize folding dynamics-- considerably richer than the simple sequential dynamics Suggest to combine the results from our smFRET experiment with those obtained from a method like native-state hydrogen exchange, also combine computer simulations and measurements