Biophys J, June 1998, p. 2743-2744, Vol. 74, No. 6
NEW AND NOTABLE
Making Movies of Molecular Motions
David
Keller
Department of Chemistry, University of New Mexico, Albuquerque, New
Mexico 87131 USA
 |
ARTICLE |
Traditionally, biomolecules have been viewed as
chemical entities, to be characterized and understood in terms of their
thermodynamic properties or by the kinetics of their chemical
reactions. But in recent years, with the growing body of
high-resolution structural information from crystallography, and new,
detailed mechanisms of large "mechanochemical" molecules such as
the molecular motors, F1Fo ATP synthases, type
II topoisomerases, RNA polymerases, and others, it has become
fashionable to think of living things as collections of tiny machines,
each carrying out a specific task in the overall process of growth and
cell division. Most standard biophysical methods require large numbers
of molecules and measure ensemble-averaged properties. This has the
distinct advantage that signals that depend on the number of molecules
are greatly amplified, making the experimentalist's job easier. But
many of the most important properties of any machine (small or large) do not scale up with numbers, and many of the most interesting puzzles
cannot be easily solved by ensemble-averaged measurements. It seems
reasonable to expect that small machines are best studied the same way
we would naturally study big machines: by observing what they do in
well-defined situations one at a time. Considerable effort has been
devoted in recent years toward developing methods for doing exactly
that. Examples include direct measurements of forces and stepping
distances on individual molecular motors (Coppin et al., 1996
),
measurements of stretching and protein unfolding forces in titin (Rief
et al., 1997; Kellermayer et al., 1997), and the direct observation of
transcription by RNA polymerase (Kasas et al., 1997), among many
others.
Now, in the paper by van Noort et al. on page 2840 of this issue of
Biophysical Journal, this "single molecule manipulation and measurement" approach is taken to a new level of sophistication. The authors show that by careful attention to detail, and by careful optimization of the parameters that control imaging forces and speed in
an atomic force microscopy (AFM) experiment, it is possible to record
the detailed movements of DNA molecules and proteins. This includes
such difficult-to-detect processes as sliding of nonspecifically bound
protein along DNA molecules and transient annealing of sticky ends of
DNA molecules previously cleaved by an endonuclease. Perhaps most
importantly, the authors have been able to distinguish clearly which
regions on any given DNA molecule are firmly pinned to the substrate
surface, and which regions are free to move and interact with other
molecules. This kind of detailed molecule-by-molecule information goes
a long way toward making real-time AFM investigations of
protein-nucleic acid interactions both interpretable and routine.
Although it has been clear from almost the beginning that AFM has the
potential for this kind of powerful, single-molecule cinematography,
several technical problems have hindered it. The essence of
movie-making is the ability to quickly take many images of the same
field of view. AFM images are formed by measuring the force of
interaction between the sample molecules and a sharp probe tip as the
tip is scanned over the sample surface. This method of creating an
image makes the two requirements for a molecular movie (images must be
collected quickly and many images must be collected from the same spot)
difficult to satisfy at the same time. On the one hand, interaction
forces must be kept small, or the sample will be damaged by repeated
scans. On the other hand, each force measurement must be made quickly,
or the time resolution of the experiment will suffer. This is an
example of the trade-off between sensitivity and bandwidth that is
encountered in many high-sensitivity measurements, and is a
nearly universal problem in single-molecule experiments of all kinds,
no matter what the instrumental setup. In AFM, the main source of noise is the ever-present, irreducible Brownian forces acting on the force
sensor. These "thermal fluctuations" are caused purely by the
presence of the buffer solution surrounding the sample, and are
impossible to distinguish from sample-induced forces without extensive,
relatively slow, time averaging.
The central result of the experiments by van Noort et al. is that with
existing instrumentation it is possible to take high-quality movies
involving many images (25 or more) of the same field of view with
reasonable time resolution (~1 min/frame, or 5000 image pixels/s) and
minimal disturbance to the sample molecules. Furthermore, by aligning
and averaging many frames together, areas of loosely attached, rapidly
moving DNA could be distinguished from areas strongly bound to the
substrate. When molecules of photolyase, a DNA repair enzyme, are
introduced, they bind and diffuse in the rapidly moving regions of the
DNA, but stop at the bound regions. This is the most systematic
demonstration so far that detailed, almost real-time dynamical
information can be obtained by in situ AFM imaging. Moreover, as the
authors themselves point out, AFM is still a relatively young
technique, and breakthrough improvements are likely in the near future.
This can only make the basic capability demonstrated here even more
powerful.
One of the main questions still to be addressed is the effect of the
substrate and the imaging tip on the movements of target molecules.
This is especially crucial when the process of interest involves
loosely bound or mobile molecules, as in the diffusion of a protein
along a DNA molecule. In the experiments of van Noort et al.,
photolyase was in some cases observed to diffuse in the direction
opposite that of the of the AFM tip, suggesting that the effects of
imaging forces may not be large. But this is a point that must be
carefully tested with well-understood systems before quantitative
conclusions can be drawn.
Real-time, single-molecule movies have been demonstrated only rarely in
the past, mostly with nucleic acids. For example, Guthold showed that
individual cuts of surface-bound DNA by the Bal 31 restriction nuclease
could be followed by AFM (Bustamante et al., 1994). And in perhaps the
most spectacular demonstration of real-time imaging, Kasas et al. were
able to directly observe transcription by Escherichia coli
RNA polymerase on a DNA template by AFM (Kasas et al., 1997).
Collectively, these experiments and others point the way to a general
AFM-based approach to studying protein-nucleic acid interactions of
many kinds. One particularly intriguing long-term possibility is the
use of AFM to perform "in vitro motility assays" for molecules like
the RNA and DNA polymerases. In the molecular motor field, "gliding
filament" assays, combined with site-specific mutagenesis, have
played a crucial role in sorting out the important structural features of myosins and other motors, and in testing hypothetical mechanisms of
force generation. If it can be demonstrated that neither the substrate
nor the imaging tip interferes significantly with movement, the AFM may
make similar experiments possible for polymerases.
| |
FOOTNOTES |
Received for publication 1 April 1998 and in final form 2 April 1998.
Address reprint requests to Dr. David J. Keller, Department of
Chemistry, University of New Mexico, Clark 103, Albuquerque, NM 87131. Tel.: 505-277-3621, 505-277-2060; Fax: 505-277-2609; E-mail:
dkeller{at}triton.unm.edu.
| |
REFERENCES |
-
Bustamante, C.,
D. A. Erie, and D. Keller.
1994.
Biochemical and structural applications of scanning force microscopy.
Curr. Opin. Struct. Biol.
4:750-760.
-
Coppin, C. M.,
J. T. Finer,
J. A. Spudich, and R. D. Vale.
1996.
The detection of sub-8 nm movements of kinesin by high resolution optical trap microscopy.
Proc. Natl. Acad. Sci. USA.
93:1913-1917[Abstract].
-
Kasas, S.,
N. H. Thompson,
B. L. Smith,
H. G. Hansma,
X. S. Zhu,
M. Guthold,
C. Bustamante,
E. T. Kool,
M. Kashlev, and P. K. Hansma.
1997.
Escherichia coli RNA polymerase activity observed using atomic force microscopy.
Biochemistry.
36:461-468[Medline].
-
Kellermayer, M. S. Z.,
S. B. Smith,
H. L. Granzier, and C. Bustamante.
1997.
Folding-unfolding transitions in single titin molecules characterized with laser tweezers.
Science.
276:1112-1116[Abstract/Full Text].
-
Rief, M.,
M. Gautel,
F. Oesterhelf,
J. M. Fernandez, and H. E. Gaub.
1997.
Reversible unfolding of individual titin immunoglobulin domains by AFM.
Science.
276:1109-1112[Abstract/Full Text].
Biophys J, June 1998, p. 2743-2744, Vol. 74, No. 6
© 1998 by the Biophysical Society 0006-3495/98/06/2743/02 $2.00
Top
Article
References
