Limits...
Transient misfolding dominates multidomain protein folding.

Borgia A, Kemplen KR, Borgia MB, Soranno A, Shammas S, Wunderlich B, Nettels D, Best RB, Clarke J, Schuler B - Nat Commun (2015)

Bottom Line: Simulations suggest that a large fraction of these misfolds resemble an intramolecular amyloid-like state reported in computational studies.However, for naturally occurring neighbours with low sequence identity, these transient misfolds disappear much more rapidly than for identical neighbours.We thus propose that evolutionary sequence divergence between domains is required to suppress the population of long-lived, potentially harmful misfolded states, whereas large populations of transient misfolded states appear to be tolerated.

View Article: PubMed Central - PubMed

Affiliation: Department of Biochemistry, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland.

ABSTRACT
Neighbouring domains of multidomain proteins with homologous tandem repeats have divergent sequences, probably as a result of evolutionary pressure to avoid misfolding and aggregation, particularly at the high cellular protein concentrations. Here we combine microfluidic-mixing single-molecule kinetics, ensemble experiments and molecular simulations to investigate how misfolding between the immunoglobulin-like domains of titin is prevented. Surprisingly, we find that during refolding of tandem repeats, independent of sequence identity, more than half of all molecules transiently form a wide range of misfolded conformations. Simulations suggest that a large fraction of these misfolds resemble an intramolecular amyloid-like state reported in computational studies. However, for naturally occurring neighbours with low sequence identity, these transient misfolds disappear much more rapidly than for identical neighbours. We thus propose that evolutionary sequence divergence between domains is required to suppress the population of long-lived, potentially harmful misfolded states, whereas large populations of transient misfolded states appear to be tolerated.

Show MeSH

Related in: MedlinePlus

Dependence of the transfer efficiencies of individual species on GdmCl concentration.(a,b) Transfer efficiency values obtained for the population of unfolded molecules by fitting FRET efficiency histograms recorded in the microfluidic-mixing device at different final GdmCl concentrations 3.8±0.3 ms after mixing. (c) Same experiments as in b, but histograms here were recorded after 4.6±0.5 s to allow for complete refolding of I27 in the presence of unfolded I28 (see main text). (d) I27 monomer transfer efficiencies measured 3.8±0.3 ms after mixing at different concentrations of GdmCl. The transfer efficiency of this species is taken to be representative of a misfolded state with the central domain formed and the terminal domain unfolded (, see main text). In all plots, a solid vertical bar at 0.23 M GdmCl, that is, the GdmCl concentration in which we measured refolding, represents the uncertainties obtained from the 90% confidence interval of an extrapolated fit (equation 10 in Methods) to all data points excluding the one at 0.23 M GdmCl (error bars indicated for individual data points). For the plots in c and d, it was possible to measure the transfer efficiency directly at 0.23 M GdmCl (red and purple circle, respectively), but these points were not included in the fits: the excellent agreement of the extrapolated and measured transfer efficiency values confirms the suitability of the extrapolation approach. A representative structure of the corresponding molecular species is given in each panel; the residues labelled with fluorescent dyes are indicated as orange spheres. All plots are colour coded according to the transfer efficiency populations in Fig. 3.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4660218&req=5

f4: Dependence of the transfer efficiencies of individual species on GdmCl concentration.(a,b) Transfer efficiency values obtained for the population of unfolded molecules by fitting FRET efficiency histograms recorded in the microfluidic-mixing device at different final GdmCl concentrations 3.8±0.3 ms after mixing. (c) Same experiments as in b, but histograms here were recorded after 4.6±0.5 s to allow for complete refolding of I27 in the presence of unfolded I28 (see main text). (d) I27 monomer transfer efficiencies measured 3.8±0.3 ms after mixing at different concentrations of GdmCl. The transfer efficiency of this species is taken to be representative of a misfolded state with the central domain formed and the terminal domain unfolded (, see main text). In all plots, a solid vertical bar at 0.23 M GdmCl, that is, the GdmCl concentration in which we measured refolding, represents the uncertainties obtained from the 90% confidence interval of an extrapolated fit (equation 10 in Methods) to all data points excluding the one at 0.23 M GdmCl (error bars indicated for individual data points). For the plots in c and d, it was possible to measure the transfer efficiency directly at 0.23 M GdmCl (red and purple circle, respectively), but these points were not included in the fits: the excellent agreement of the extrapolated and measured transfer efficiency values confirms the suitability of the extrapolation approach. A representative structure of the corresponding molecular species is given in each panel; the residues labelled with fluorescent dyes are indicated as orange spheres. All plots are colour coded according to the transfer efficiency populations in Fig. 3.

Mentions: Even without considering misfolding, other populations are expected to contribute at early refolding times: molecules with both domains unfolded (U27–U27; Supplementary Fig. 3a) and with only one domain folded, (F27–U27/U27–F27; Supplementary Fig. 3b). To obtain the transfer efficiency of the completely unfolded state in 0.23 M GdmCl, we measured U27–U27 and U27–U28 between 4.6 and 0.5 M GdmCl ∼4 ms after mixing (when the totally unfolded population is >95%) and extrapolated their transfer efficiency to 0.23 M GdmCl (Fig. 4a,b), yielding E=0.27±0.02 and E=0.32±0.03, respectively. A similar procedure was employed to identify the transfer efficiency of F27–U27 and F27–U28. In this case, we took advantage of the slow refolding of I28 compared with I27 (∼4 × 10−3 s−1 versus ∼15 s−1, respectively, in 0.23 M GdmCl (refs 31, 32)), allowing us to transiently populate F27–U28 in the I27–I28 tandem and obtain its transfer efficiency of 0.28±0.02 in 0.23 M GdmCl (Fig. 4c). In view of the positions of the fluorophores in the tandem repeats, both F27–U27 and U27–F27 are expected to exhibit this transfer efficiency, within experimental uncertainty.


Transient misfolding dominates multidomain protein folding.

Borgia A, Kemplen KR, Borgia MB, Soranno A, Shammas S, Wunderlich B, Nettels D, Best RB, Clarke J, Schuler B - Nat Commun (2015)

Dependence of the transfer efficiencies of individual species on GdmCl concentration.(a,b) Transfer efficiency values obtained for the population of unfolded molecules by fitting FRET efficiency histograms recorded in the microfluidic-mixing device at different final GdmCl concentrations 3.8±0.3 ms after mixing. (c) Same experiments as in b, but histograms here were recorded after 4.6±0.5 s to allow for complete refolding of I27 in the presence of unfolded I28 (see main text). (d) I27 monomer transfer efficiencies measured 3.8±0.3 ms after mixing at different concentrations of GdmCl. The transfer efficiency of this species is taken to be representative of a misfolded state with the central domain formed and the terminal domain unfolded (, see main text). In all plots, a solid vertical bar at 0.23 M GdmCl, that is, the GdmCl concentration in which we measured refolding, represents the uncertainties obtained from the 90% confidence interval of an extrapolated fit (equation 10 in Methods) to all data points excluding the one at 0.23 M GdmCl (error bars indicated for individual data points). For the plots in c and d, it was possible to measure the transfer efficiency directly at 0.23 M GdmCl (red and purple circle, respectively), but these points were not included in the fits: the excellent agreement of the extrapolated and measured transfer efficiency values confirms the suitability of the extrapolation approach. A representative structure of the corresponding molecular species is given in each panel; the residues labelled with fluorescent dyes are indicated as orange spheres. All plots are colour coded according to the transfer efficiency populations in Fig. 3.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4660218&req=5

f4: Dependence of the transfer efficiencies of individual species on GdmCl concentration.(a,b) Transfer efficiency values obtained for the population of unfolded molecules by fitting FRET efficiency histograms recorded in the microfluidic-mixing device at different final GdmCl concentrations 3.8±0.3 ms after mixing. (c) Same experiments as in b, but histograms here were recorded after 4.6±0.5 s to allow for complete refolding of I27 in the presence of unfolded I28 (see main text). (d) I27 monomer transfer efficiencies measured 3.8±0.3 ms after mixing at different concentrations of GdmCl. The transfer efficiency of this species is taken to be representative of a misfolded state with the central domain formed and the terminal domain unfolded (, see main text). In all plots, a solid vertical bar at 0.23 M GdmCl, that is, the GdmCl concentration in which we measured refolding, represents the uncertainties obtained from the 90% confidence interval of an extrapolated fit (equation 10 in Methods) to all data points excluding the one at 0.23 M GdmCl (error bars indicated for individual data points). For the plots in c and d, it was possible to measure the transfer efficiency directly at 0.23 M GdmCl (red and purple circle, respectively), but these points were not included in the fits: the excellent agreement of the extrapolated and measured transfer efficiency values confirms the suitability of the extrapolation approach. A representative structure of the corresponding molecular species is given in each panel; the residues labelled with fluorescent dyes are indicated as orange spheres. All plots are colour coded according to the transfer efficiency populations in Fig. 3.
Mentions: Even without considering misfolding, other populations are expected to contribute at early refolding times: molecules with both domains unfolded (U27–U27; Supplementary Fig. 3a) and with only one domain folded, (F27–U27/U27–F27; Supplementary Fig. 3b). To obtain the transfer efficiency of the completely unfolded state in 0.23 M GdmCl, we measured U27–U27 and U27–U28 between 4.6 and 0.5 M GdmCl ∼4 ms after mixing (when the totally unfolded population is >95%) and extrapolated their transfer efficiency to 0.23 M GdmCl (Fig. 4a,b), yielding E=0.27±0.02 and E=0.32±0.03, respectively. A similar procedure was employed to identify the transfer efficiency of F27–U27 and F27–U28. In this case, we took advantage of the slow refolding of I28 compared with I27 (∼4 × 10−3 s−1 versus ∼15 s−1, respectively, in 0.23 M GdmCl (refs 31, 32)), allowing us to transiently populate F27–U28 in the I27–I28 tandem and obtain its transfer efficiency of 0.28±0.02 in 0.23 M GdmCl (Fig. 4c). In view of the positions of the fluorophores in the tandem repeats, both F27–U27 and U27–F27 are expected to exhibit this transfer efficiency, within experimental uncertainty.

Bottom Line: Simulations suggest that a large fraction of these misfolds resemble an intramolecular amyloid-like state reported in computational studies.However, for naturally occurring neighbours with low sequence identity, these transient misfolds disappear much more rapidly than for identical neighbours.We thus propose that evolutionary sequence divergence between domains is required to suppress the population of long-lived, potentially harmful misfolded states, whereas large populations of transient misfolded states appear to be tolerated.

View Article: PubMed Central - PubMed

Affiliation: Department of Biochemistry, University of Zurich, Winterthurerstrasse 190, 8057 Zurich, Switzerland.

ABSTRACT
Neighbouring domains of multidomain proteins with homologous tandem repeats have divergent sequences, probably as a result of evolutionary pressure to avoid misfolding and aggregation, particularly at the high cellular protein concentrations. Here we combine microfluidic-mixing single-molecule kinetics, ensemble experiments and molecular simulations to investigate how misfolding between the immunoglobulin-like domains of titin is prevented. Surprisingly, we find that during refolding of tandem repeats, independent of sequence identity, more than half of all molecules transiently form a wide range of misfolded conformations. Simulations suggest that a large fraction of these misfolds resemble an intramolecular amyloid-like state reported in computational studies. However, for naturally occurring neighbours with low sequence identity, these transient misfolds disappear much more rapidly than for identical neighbours. We thus propose that evolutionary sequence divergence between domains is required to suppress the population of long-lived, potentially harmful misfolded states, whereas large populations of transient misfolded states appear to be tolerated.

Show MeSH
Related in: MedlinePlus