Chemistry Reference
In-Depth Information
9.4.1.3 Equilibrium Unfolding Studies on Simple Model Two-state Folders
In order to properly study protein folding, it is important to have access to informa-
tion on many time scales, from nanoseconds to seconds or longer. Since confocal-
based single-molecule measurements currently have detectors with much better
time resolution, most, if not all, single-molecule protein folding studies have used
this optical isolation technique. In the
first protein-folding measurements described
in this section and the next, a focused, continuous-wave (cw; non-pulsed) laser was
used to excite the donor
uorophore. The
fluorescence emission from single
molecules was collected and refocused onto a detection pinhole used to select the
small observation volume. The
fluorescence emission is split using a dichroic mirror
and two bandpass
filters into two detection channels, one for the donor and one for
the acceptor.
For the studies on freely diffusing molecules, time traces of donor and acceptor
emission were formed with 100-
s
-
1-ms time resolution, and intense, single-
molecule signals (or bursts) above the background were selected. For each selected
single molecule, the ratio E
m
¼
I
A
/(I
A
þ g
I
D
) was calculated, where
g
is the ratio of the
quantum ef
fluorophores. In the initial
studies, this ratio was not precisely calibrated, although it was typically near 1.
Histograms are formed from these E ratios, showing how many proteins have a
speci
c value of E. In most of these studies, there was a large subpopulation of
proteins with a photobleached or inactive acceptor. For these proteins, only donor
emission is observed, and E is 0 (assuming that leakage of donor emission into the
acceptor channel is subtracted). Since this donor-only subpopulation is indistin-
guishable from a donor- and acceptor-labeled protein with dimensions too large for
FRET to occur, this limits the usable range of E to be well above 0 (generally 0.3 and
above).
In the
first paper to demonstrate single-molecule measurements of proteins
undergoing folding and unfolding, the Hochstrasser group studied immobilized
GCN4 using FRET [79]; they constructed similar E histograms, where molecules
were searched in an image (rather than a burst in a time trace). Extended time
trajectories lasting several seconds were recorded for single molecules, allowing for
dynamical information to be accessed. Correlations calculated on these time trajec-
tories were used to provide information on the temporal dynamics. Adding denatur-
ant to the protein increases the fraction of proteins that are unfolded. Upon addition
of 7M of denaturant, all GCN4 molecules were unfolded, increasing on average the
distance between D and A, and lowering E. In this completely denatured state, a
variety of distances was seen, covering a large range in E. This was attributed to the
sticking of the unfolded protein to the surface. This prevented the authors from
making a strong correspondence between the results of the study with the results
from previous, ensemble level studies. Even so, the general pattern of increasing the
number of unfolded proteins with denaturant was seen at the single-molecule level.
In order to study proteins undergoing folding and unfolding in a less perturbative
environment, it was necessary to keep the protein (especially unfolded proteins) away
from glass surfaces. In [80], we introduced a method of studying protein folding on
freely diffusing molecules, and demonstrated it using the extensively studied two
ciencies and detection ef
ciencies of the
