Folding-unfolding transitions in single titin molecules characterized with laser tweezers

Science

Volumn 276, Issue 5315, 1997, Pages 1112-1116

  Kellermayer, Miklós S Z ^a,c   Smith, Steven B ^b   Granzier, Henk L ^a   Bustamante, Carlos ^b  

^a WASHINGTON STATE UNIVERSITY   (United States)

^b UNIVERSITY OF OREGON   (United States)

^c UNIVERSITY OF PÉCS   (Hungary)

Author keywords

[No Author keywords available]

Indexed keywords

ARTICLE; ELASTICITY; ENTROPY; FORCE; HYSTERESIS; MUSCLE CELL; PRIORITY JOURNAL; PROTEIN ANALYSIS; PROTEIN FOLDING; PROTEIN STRUCTURE;

AMINO ACID SEQUENCE; ELASTICITY; ENTROPY; IMMUNOGLOBULINS; LASERS; MODELS, CHEMICAL; MUSCLE CONTRACTION; MUSCLE PROTEINS; MUSCLE RELAXATION; MUSCLE, SKELETAL; PROTEIN DENATURATION; PROTEIN FOLDING; PROTEIN KINASES; STRESS, MECHANICAL;

EID: 0031002460     PISSN: 00368075     EISSN: None     Source Type: Journal    
DOI: 10.1126/science.276.5315.1112     Document Type: Article

References (37)

1. R. Horowits and R. J. Podolsky, J. Cell Biol. 105, 2217 (1987); H. L. M. Granzier, H. A. Akster, H. E. D. J. ter Keurs, Am. J. Physiol. 260, C1060 (1991).
2. R. Horowits and R. J. Podolsky, J. Cell Biol. 105, 2217 (1987); H. L. M. Granzier, H. A. Akster, H. E. D. J. ter Keurs, Am. J. Physiol. 260, C1060 (1991).
3. K. Wang, Cell Muscle Motil. 6, 315 (1985); K. Maruyama, Biophys. Chem. 50, 73 (1994).
4. K. Wang, Cell Muscle Motil. 6, 315 (1985); K. Maruyama, Biophys. Chem. 50, 73 (1994).
5. D. O. Fürst, M. Osborn, R. Nave, K. Weber, J. Cell Biol. 106, 1563 (1988).
6. H. L. M. Granzier and T. Irving, Biophys. J. 68, 1027 (1995).
7. R. Horowits, E. S. Kempner, M. E. Bisher, R. J. Podolsky, Nature 323, 160 (1986).
8. A. Soteriou, A. Clarke, S. Martin, J. Trinick, Proc. R. Soc. London Ser. B 254, 83 (1993).
9. H. P. Erickson, Proc. Natl. Acad. Sci. U.S.A. 91, 10114 (1994).
10. S. Labeit, M. Gautel, A. Lackey, J. Trinick, EMBO J. 11, 1711 (1992); S. Labeit et al., Nature 345, 273 (1990); J. D. Fritz, J. A. Wolff, M. L. Greaser, Comp. Biochem. Physiol. B 105, 357 (1993); K. M. Pan, S. Damodaran, M. L. Greaser, Biochemistry 33, 8255 (1994).
11. S. Labeit, M. Gautel, A. Lackey, J. Trinick, EMBO J. 11, 1711 (1992); S. Labeit et al., Nature 345, 273 (1990); J. D. Fritz, J. A. Wolff, M. L. Greaser, Comp. Biochem. Physiol. B 105, 357 (1993); K. M. Pan, S. Damodaran, M. L. Greaser, Biochemistry 33, 8255 (1994).
12. S. Labeit, M. Gautel, A. Lackey, J. Trinick, EMBO J. 11, 1711 (1992); S. Labeit et al., Nature 345, 273 (1990); J. D. Fritz, J. A. Wolff, M. L. Greaser, Comp. Biochem. Physiol. B 105, 357 (1993); K. M. Pan, S. Damodaran, M. L. Greaser, Biochemistry 33, 8255 (1994).
13. S. Labeit, M. Gautel, A. Lackey, J. Trinick, EMBO J. 11, 1711 (1992); S. Labeit et al., Nature 345, 273 (1990); J. D. Fritz, J. A. Wolff, M. L. Greaser, Comp. Biochem. Physiol. B 105, 357 (1993); K. M. Pan, S. Damodaran, M. L. Greaser, Biochemistry 33, 8255 (1994).
14. S. Labeit and B. Kolmerer, Science 270, 293 (1995); W. A. Linke et al., J. Mol. Biol. 261, 62 (1996).
15. S. Labeit and B. Kolmerer, Science 270, 293 (1995); W. A. Linke et al., J. Mol. Biol. 261, 62 (1996).
16. 2, 1 mM EGTA, 1 mM dithiothreitol (DTT), pH 7.4] containing 0.2% Tween-20 and T4 lysozyme (0.2 mg/ml) to prevent the nonspecific binding of parts of the titin molecule to any of the beads. Throughout the protocols, AB contained 42 μM leupeptin (Peptides International) and 10 μM E-64 (Sigma). Although the silica beads did not provide sequence-specific attachment, they significantly increased the success of titin tethering.
17. S. B. Smith, Y. Cui, C. Bustamante, Science 271, 795 (1996).
18. K. Wang, M. R. Ramirez, D. Palter, Proc. Natl. Acad. Sci. U.S.A. 81, 3685 (1984); R. Nave, D. O. Fürst, K. Weber, J. Cell Biol. 109, 2177 (1989).
19. K. Wang, M. R. Ramirez, D. Palter, Proc. Natl. Acad. Sci. U.S.A. 81, 3685 (1984); R. Nave, D. O. Fürst, K. Weber, J. Cell Biol. 109, 2177 (1989).
20. W. Kuhn and F. Grün, Kolloid Z 101, 248 (1942).
21. O. Kratky and G. Porod, Recl. Trav. Chim. Pays-Bas 68, 1106 (1949).
22. S. B. Smith, M. S. Z. Kellermayer, C. Bustamante, H. L. Granzier, in preparation.
23. C. J. Bustamante, J. F. Marko, E. D. Siggia, S. B. Smith, Science 265, 1599 (1994); J. F. Marko and E. D. Siggia, Macromolecules 28, 8759 (1995).
24. C. J. Bustamante, J. F. Marko, E. D. Siggia, S. B. Smith, Science 265, 1599 (1994); J. F. Marko and E. D. Siggia, Macromolecules 28, 8759 (1995).
25. The length of the fully unfolded, denatured skeletal titin molecule extending from the T12 epitope to the M line is ∼10 μm. assuming ∼30,000 amino acids (4) at 3.5 Å per ammo acid.
26. Absence of a uniform distribution of contour lengths out to 10 μm, expected in the case of nonspecific binding by silica beads, could be explained by several mechanisms; (i) the molecules were never stretched beyond half their theoretical length at the forces used; (ii) a fraction of the tethers were loops attached at both ends to one bead and snagged in the middle by the other (this would in turn be a double tether); and (iii) there was specificity in the interaction between titin and silica beads due to the uneven charge distribution along the titin molecule (25). The possibility that the transitions in the force versus extension curves could be explained by the presence of multiple molecules of different contour lengths between the beads was excluded by a simulation based on the behavior of WLCs of unequal lengths. This simulation, however, was not used to exclude any multimolecule tether from the data analysis shown in Fig. 3B. Thus, tethers of nonintegral effective persistence lengths in Fig. 3B may reflect such uneven yoking of molecules between the beads.
27. stretch in Fig. 3A) varied between 1.5 to 3.0 μm. Assuming this length also included ∼1 μm of native titin, the size of the pre-unfolded fraction is 0.5 to 2.0 μm, representing 5 to 20% of the entire ∼10-μm-long primary structure. If some of these molecules were tethered in their middles, the pre-unfolded fraction could comprise 10 to 40% of the tethered half-molecule. The unfolded PEVK domain of titin in skeletal muscle, with 2174 residues, may contribute up to 0.8 μm to the contour length of the whole molecule (4). Early electron micrographs of titin, which showed the molecule as strings of beads connected with thin strands, have already implied the presence of regions that easily extend or unfold under stress [(12); J. Trinick et al., J. Mol. Biol. 180, 331 (1984)]. Recent electron microscopic evidence further points to the presence of an easily unfolding or pre-unfolded region in titin, which may correspond to the PEVK domain (25). The relatively small unfolded fraction seen in the electron micrographs could be due to the short time (a few milliseconds per molecule) allowed for denaturation by the preparation procedure.
28. H. Higuchi, Y. Nakauchi, K. Maruyama, S. Fujime, Biophys. J. 65, 1906 (1993).
29. The relaxation time for the force-denaturation of titin was estimated by rapidly stretching the molecule to high force (∼60 pN) and holding its length constant while watching the force decay. This relaxation was best fit with a fast decay time of ∼4 s and a slower phase of ∼70 s. Thus, these processes are slow compared with the stretch and release rates and very slow compared with the ∼1 s required to denature reduced-disulfide Ig domains in 4 M GuCl [Y. Goto and K. Hamaguchi, J. Mol. Biol. 156, 911 (1982)].
30. -1 for both processes. A refolding intermediate that creates a new fold in the β barrel of an Ig or FNIII domain must shorten the denatured molecule's extension by Δx ≈ 8 nm before the proper set of interactions can be made, but an unfolding intermediate with a high (rate-limiting) free energy could be one in which many bonds or hydrophobic interactions are disrupted by a relatively small shift (Δx ≈ 0.3 nm) away from the ideal shape of the native domain. Source code and simulation results are available at http://alice.uoregon. edu/∼cjblab on the Web.
31. The rate of refolding of the FNIII domains could strongly depend on whether or not the unfolding leads to trans-cis isomerization of the eight prolines at the corners of the β folds of these domains [A. L. Main et al., Cell 71, 671 (1992)]. Protein refolding is often limited by the rate of proline isomerization [T. Kiefhaber, H. H. Kohler, F. Schmid, J. Mol. Biol. 224, 217 (1992)]. Because the state of all eight prolines should randomize in a domain that is unfolded for 10 to 100 s, correct isomerization of all eight prolines might take thousands of seconds to occur randomly. Contrary to this view, K. Plaxco et al. [Proc. Natl. Acad. Sci. U.S.A. 93, 10703, (1996)] found that an isolated FNIII domain refolded rapidly (<1 s) after chemical denaturation. Titin force-refolding may be slower than FNIII chemical-refolding for several reasons: (i) no residual chemical denaturant helps prevent kinetic traps during refolding, (ii) pulled prolines may preferentially adopt the incorrect cis configuration, (iii) a variety of proline-rich FNIII and Ig domains are present in titin, and (iv) titin domains are connected axially, limiting their accessible conformations during refolding. Thus, "wearing-out" could still reflect the systematic increase in the number of prolines in the cis state and the consequent increased fraction of slow-to-renature domains in the molecule. Such wearing out may explain why the pre-unfolded fraction of titin was longer than that attributable to the PEVK region. Because all tethers were fully stretched one or more times before data was taken, the pre-unfolded titin could also contain unfolded Ig or FNIII domains that needed more time to renature. Indeed, single molecules recover if left relaxed for several minutes and regain some of their hysteresis behavior. Notably, wearing out and recovery has also been seen in muscle fibers (24). 24. H. L. M. Granzier and K. Wang, Biophys. J. 65, 2141 (1993).
32. The rate of refolding of the FNIII domains could strongly depend on whether or not the unfolding leads to trans-cis isomerization of the eight prolines at the corners of the β folds of these domains [A. L. Main et al., Cell 71, 671 (1992)]. Protein refolding is often limited by the rate of proline isomerization [T. Kiefhaber, H. H. Kohler, F. Schmid, J. Mol. Biol. 224, 217 (1992)]. Because the state of all eight prolines should randomize in a domain that is unfolded for 10 to 100 s, correct isomerization of all eight prolines might take thousands of seconds to occur randomly. Contrary to this view, K. Plaxco et al. [Proc. Natl. Acad. Sci. U.S.A. 93, 10703, (1996)] found that an isolated FNIII domain refolded rapidly (<1 s) after chemical denaturation. Titin force-refolding may be slower than FNIII chemical-refolding for several reasons: (i) no residual chemical denaturant helps prevent kinetic traps during refolding, (ii) pulled prolines may preferentially adopt the incorrect cis configuration, (iii) a variety of proline-rich FNIII and Ig domains are present in titin, and (iv) titin domains are connected axially, limiting their accessible conformations during refolding. Thus, "wearing-out" could still reflect the systematic increase in the number of prolines in the cis state and the consequent increased fraction of slow-to-renature domains in the molecule. Such wearing out may explain why the pre-unfolded fraction of titin was longer than that attributable to the PEVK region. Because all tethers were fully stretched one or more times before data was taken, the pre-unfolded titin could also contain unfolded Ig or FNIII domains that needed more time to renature. Indeed, single molecules recover if left relaxed for several minutes and regain some of their hysteresis behavior. Notably, wearing out and recovery has also been seen in muscle fibers (24). 24. H. L. M. Granzier and K. Wang, Biophys. J. 65, 2141 (1993).
33. The rate of refolding of the FNIII domains could strongly depend on whether or not the unfolding leads to trans-cis isomerization of the eight prolines at the corners of the β folds of these domains [A. L. Main et al., Cell 71, 671 (1992)]. Protein refolding is often limited by the rate of proline isomerization [T. Kiefhaber, H. H. Kohler, F. Schmid, J. Mol. Biol. 224, 217 (1992)]. Because the state of all eight prolines should randomize in a domain that is unfolded for 10 to 100 s, correct isomerization of all eight prolines might take thousands of seconds to occur randomly. Contrary to this view, K. Plaxco et al. [Proc. Natl. Acad. Sci. U.S.A. 93, 10703, (1996)] found that an isolated FNIII domain refolded rapidly (<1 s) after chemical denaturation. Titin force-refolding may be slower than FNIII chemical-refolding for several reasons: (i) no residual chemical denaturant helps prevent kinetic traps during refolding, (ii) pulled prolines may preferentially adopt the incorrect cis configuration, (iii) a variety of proline-rich FNIII and Ig domains are present in titin, and (iv) titin domains are connected axially, limiting their accessible conformations during refolding. Thus, "wearing-out" could still reflect the systematic increase in the number of prolines in the cis state and the consequent increased fraction of slow-to-renature domains in the molecule. Such wearing out may explain why the pre-unfolded fraction of titin was longer than that attributable to the PEVK region. Because all tethers were fully stretched one or more times before data was taken, the pre-unfolded titin could also contain unfolded Ig or FNIII domains that needed more time to renature. Indeed, single molecules recover if left relaxed for several minutes and regain some of their hysteresis behavior. Notably, wearing out and recovery has also been seen in muscle fibers (24). 24. H. L. M. Granzier and K. Wang, Biophys. J. 65, 2141 (1993).
34. The rate of refolding of the FNIII domains could strongly depend on whether or not the unfolding leads to trans-cis isomerization of the eight prolines at the corners of the β folds of these domains [A. L. Main et al., Cell 71, 671 (1992)]. Protein refolding is often limited by the rate of proline isomerization [T. Kiefhaber, H. H. Kohler, F. Schmid, J. Mol. Biol. 224, 217 (1992)]. Because the state of all eight prolines should randomize in a domain that is unfolded for 10 to 100 s, correct isomerization of all eight prolines might take thousands of seconds to occur randomly. Contrary to this view, K. Plaxco et al. [Proc. Natl. Acad. Sci. U.S.A. 93, 10703, (1996)] found that an isolated FNIII domain refolded rapidly (<1 s) after chemical denaturation. Titin force-refolding may be slower than FNIII chemical-refolding for several reasons: (i) no residual chemical denaturant helps prevent kinetic traps during refolding, (ii) pulled prolines may preferentially adopt the incorrect cis configuration, (iii) a variety of proline-rich FNIII and Ig domains are present in titin, and (iv) titin domains are connected axially, limiting their accessible conformations during refolding. Thus, "wearing-out" could still reflect the systematic increase in the number of prolines in the cis state and the consequent increased fraction of slow-to-renature domains in the molecule. Such wearing out may explain why the pre-unfolded fraction of titin was longer than that attributable to the PEVK region. Because all tethers were fully stretched one or more times before data was taken, the pre-unfolded titin could also contain unfolded Ig or FNIII domains that needed more time to renature. Indeed, single molecules recover if left relaxed for several minutes and regain some of their hysteresis behavior. Notably, wearing out and recovery has also been seen in muscle fibers (24). 24. H. L. M. Granzier and K. Wang, Biophys. J. 65, 2141 (1993).
35. L. Tskhovrebova and J. Trinick. J. Mol. Biol. 265, 100 (1997).
36. K. Trombitás and G. H. Pollack, J. Muscle Res. Cell Motil. 14, 416 (1993).
37. Supported by grants from the Whitaker Foundation and the National Institute of Arthritis and Musculo-skeletal and Skin Disease (AR-42652) to H.L.G., and by grants from NIH (GM-32543) and NSF (MBC 9118482) to C.B. H.L.G. is an Established Investigator of the American Heart Association. The T12 and T51 antibodies were donated by D. O. Fürst. We thank M. Hegner for his help with atomic force microscopy, and G. Flynn, J. Schellman, E. Reisler, G. Yang, K. Campbell, C. Cremo, G. H. Pollack, and B. Slinker for their insightful comments on the manuscript.