 |
Description  |
|
|
BACKGROUND OF THE INVENTION
This invention was made with U.S. Government support under Contract No. GM
37277 awarded by the National Institute of General Medical Sciences of the
United States Department of Health and Human Services.
This invention relates a method for characterizing polymer molecules or the
like, for example, observing and determining the size of individual
particles and determining the weight distribution of a sample containing
particles of varying size. More particularly, this invention involves the
use of microscopy and/or microscopy in combination with spectroscopic
methods to characterize particles, such as by measuring their positional
and conformational changes when they are subjected to an external force,
and by measuring their length and diameter or radius. Such measurements
include rates of relaxation, reorientation and rotation of particles
subject to an external force, and measurements of length and diameter of
particles before, during or after they are subjected to an external force.
Among other applications, these measurements are used to determine particle
size. The invention is especially well suited to size polymer molecules in
a polydisperse sample (e.g., a sample containing particles of varying
size), when the particles have been placed in some type of medium, and is
useful to measure very large molecules, such as large nucleic acid
molecules which are subject to breakage when placed on a microscope slide
using conventional methods.
Methods for determining the molecular weight distribution of a polydisperse
sample of particles are useful in a variety of different fields. For
example, in polymer chemistry, the properties of an oligomer are often
dependent upon its molecular weight distribution. When a particular
substance is found to exhibit favorable properties and the exact
composition of the oligomer is not known, an analysis of the molecular
weight distribution of the polymer is used for purposes of identification.
In molecular biology, the molecular weight distribution of a polydisperse
sample, such as a sample of DNA restriction enzyme digests, provides
valuable information about the organization of the DNA. This information
may be used to produce chromosome maps and extensive molecular genetics
characterizations.
Traditionally, the molecular weight distribution of a sample of particles
has been determined by measuring the rate at which particles which are
subjected to a perturbing force move through an appropriate medium, e.g.,
a medium which causes the particles to separate according to size. A
mathematical relationship is calculated which relates the size of
particles and their migration rate through a medium when a specified force
is applied. For example, in gel permeation chromatography, a well-known
technique, a polymer to be characterized is dissolved in a solvent and the
resulting solution is then passed through a column which has a
cross-linked or porous gel polymer in the stationary phase. Large
molecules will pass quickly through the gel, while the movement of smaller
molecules will be hindered by their entry into the pores of the substance
comprising the stationary phase. The molecular weight distribution of the
sample is determined by measuring the content of the effluent from the
column, e.g., by measuring the refractive index of the effluent over a
period of time. Several limitations of gel permeation chromatography are
that it cannot be used to separate DNA molecules larger than about 5 kb,
and it can only be used for samples which are soluble in (at least one of)
a limited number of suitable solvents.
Sedimentation is a well-known technique for measuring particle size, but,
when applied to polymers, this method is limited to molecules with a
maximum size of about 50-100 kilobases (kb). Attempting to measure larger
molecules by this technique would probably result in underestimation of
molecular size, mainly because the sedimentation coefficient is sensitive
to centrifuge speed. (see Kavenoff et al., Cold Spring Harbor Symp.
Quantit. Biol., 38, 1 (1974).
Another popular method of separating polymer particles by size is by gel
electrophoresis (see, e.g., Freifelder, Physical Biochemistry, W. H.
Freeman (1976), which is particularly useful for separating restriction
digests. In brief, application of an electric field to an agarose or
polyacrylamide gel in which polymer particles are dissolved causes the
smaller particles to migrate through the gel at a faster rate than the
larger particles. The molecular weight of the polymer in each band is
calibrated by a comparison of the migration rate of an unknown substance
with the mobility of polymer fragments of known length. The amount of
polymer in each band can be estimated based upon the width and/or color
intensity (optical density) of the stained band, however, this type of
estimate is usually not very accurate.
Pulsed field electrophoresis, developed by the present inventor and
described in U.S. Pat. No. 4,473,452, the disclosure of which is hereby
incorporated by reference and relied upon, is an electrophoretic technique
in which the separation of large DNA molecules in a gel is improved
relative to separation using conventional electrophoresis. According to
this technique, deliberately alternated electric fields are used to
separate particles, rather than the continuous fields used in previously
known electrophoretic methods. More particularly, particles are separated
using electric fields of equal strength which are transverse to each
other, which alternate between high and low intensities out of phase with
each other at a frequency related to the mass of the particles. The forces
move the particles in an overall direction transverse to the respective
directions of the fields. It should be noted here that the term
"transverse" as used herein is not limited to an angle of, or close to,
90.degree., but includes other substantial angles of intersection.
One of the most significant problems with determining the weight of
molecules by indirect measurement techniques, such as those described
above, is that the parameters which are directly measured, e.g., migration
rate, are relatively insensitive to small differences in molecular size.
Thus, a precise determination of particle size distribution is difficult
to obtain. The lack of precision may particularly be a problem when
biological polymer samples, which tend to be unstable and contain single
molecules inches in length, are involved.
While some of the known methods of determining particle size distribution
in a polydisperse sample provide better resolution than others, few, if
any, of the previously known techniques provide resolution as high as is
needed to distinguish between particles of nearly identical size. Gel
permeation chromatography and sedimentation provide resolution of only
about M.sup.1/2 (M=molecular weight). Standard agarose gel
electrophoresis and polyacrylamide gel electrophoresis provide resolution
varying as -logM. Pulsed electrophoretic techniques are effective for
separating extraordinarily large molecules, but do not provide much better
resolution than standard electrophoresis. Thus, the ability to distinguish
between particles of similar size, for example, particles differing in
length by a fraction of percent, is quite limited when the above-described
measurement techniques are used.
Under special experimental circumstances, DNA gel electrophoresis resolves
a polymer mixture to a resolution of M.sup.1. However, this degree of
accuracy is only achieved when variables such as gel concentration and
field strength are carefully controlled.
Particles of higher mass (i.e., up to approximately 600 kb) can be resolved
using conventional gel electrophoresis by reducing the gel concentration
to as low as 0.035% and reducing field strength, however, there are
drawbacks to this method. Most notably, the dramatic reduction in gel
concentration results in a gel which is mechanically unstable, and less
sample can be loaded. An electrophoretic run to resolve very large DNA
molecules using a reduced gel concentration and field strength may take a
week or more to complete. Furthermore, a reduced gel concentration is not
useful to separate molecules in a sample having a wide range of particle
sizes, because separation of small molecules is not achieved. Thus, if a
sample containing molecules having a wide range of sizes is to be
separated, several electrophoretic runs may be needed, e.g., first, a
separation of the larger molecules and then further separation of the
smaller molecules.
Other particle measurement techniques known in the art are useful for
sizing certain molecules which are present in a bulk sample, (e.g., the
largest molecules in the sample, or the average molecular size) but are
impractical for measuring many polymers of varying length in a given
sample. The viscoelastic recoil technique, (see Kavenoff et al,
"Chromosome-sized DNA molecules from Drosophila," Chromosoma 41, 1 (1973))
which is well known in the art, involves stretching out coiled molecules
in a solvent flow field (e.g., a field which is created when fluid is
perturbed between two moving plates) and determining the time required for
the largest molecule to return to a relaxed state. Relaxation time is
measured by watching the rotation of a concentric rotor which moves during
the time of relaxation. While this technique is quite precise in that
sample determinations vary as M.sup.1.66 when applied to large DNA
molecules, it is not useful for sizing molecules other than the largest
molecule in the sample.
Using light scattering techniques, which are known in the art, (e.g.,
quasi-elastic light scattering), the size and shape of particles are
determined by a Zimm plot, a data analysis method which is known in the
art. With these techniques, size dependence varies as M.sup.1. Light
scattering requires that the solution in which the molecules to be
measured are placed is pure, that is, without dust or other contamination,
and it is therefore unsuitable for sizing a DNA sample. Furthermore, it is
not useful for sizing molecules as large as many DNA molecules, and is
useful only for determining the average weight of particles in a sample,
not the weight distribution of a sample with particles of various sizes.
Yet another particle measuring technique which is known in the art for
measuring individual molecules provides measurements of particle size
having limited accuracy. The average size and shape of individual, relaxed
DNA molecules has been determined by observing the molecules under a
fluorescence microscope, and measuring the major and minor axes of
molecules having a spherical or ellipsoid shape (see Yanagida et al, Cold
Spring Harbor Symp. Quantit. Biol. 47, 177, (1983)). The technique
described in the above-cited reference is performed in a free solution,
without perturbation of the molecules.
The movement of small DNA molecules during electrophoresis has been
observed (see Smith et al. Science, 243, 203 (1989)). The methods
disclosed in this publication are not suitable for observation of very
large DNA molecules, and techniques for measuring molecules are not
discussed.
It is noted that practical weight determinations of particles such as
polymer molecules depend not only upon maximizing the size dependencies of
the directly measured parameters, but also upon factors such as the amount
of sample needed, the time required to complete an analysis, and the
accuracy of measurements. Gel permeation chromatography can be
time-consuming and requires a large amount of sample. Methods such as
conventional gel electrophoresis can be relatively time-consuming, require
moderate amounts of sample, and cannot size very large DNA molecules.
SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention to provide a method
for effectively characterizing polymer molecules and the like, for
example, determining the size of individual particles or the weight
distribution of a polydisperse sample of particles.
Another object of this invention is to determine particle size using a
method which provides better resolution than methods known in the prior
art.
Yet another object of this invention is to provide a method for observing
large molecules or other large particles, (including molecules too large
to size using any other methodology), determining their size, and
determining the molecular weight distribution of a sample containing large
molecules.
A further object of this invention is to provide a faster and more
efficient method for determining the size of individual particles and the
weight distribution of a sample of particles.
Another object of this invention is to size one or more particles using an
extremely sensitive method, e.g., one which can use an amount of sample as
small as a single particle.
Yet another object of this invention to provide accurate size information
for a polydisperse sample containing particles having a wide range of
sizes, and to provide this information more quickly than by using
previously known techniques.
These and other objects of the invention will be apparent from the
discussion which follows.
Broadly described, the method of the invention involves characterizing
individual particles, including deformable and non-deformable particles in
a polydisperse sample by placing the particles in a medium, applying an
external force to the particles, thereby causing physical changes,
particularly conformational and/or positional changes, and then observing
and measuring these changes. This method is useful for characterizing
polymer molecules of a variety of sizes, including the smallest molecules
which are detected by a suitable microscope (the microscope optionally may
be attached to a spectroscopic apparatus and thus molecules too small to
be visualized may still be detected), and large polymers, which may be up
to several inches in length when stretched to a linear conformation. Shear
sensitive molecules (e.g., large molecules), which cannot be placed on a
microscope slide without breaking when conventional techniques are used,
are measured according to this invention by collapsing (condensing) the
molecules before they are placed in the medium and then uncollapsing them
after placement in the medium. This invention is useful for characterizing
many types of particles which can be visualized or detected under a light
microscope. Several non-limiting examples include polysaccharides,
polypeptides and proteins.
Deformable particles are particles which have a tendency to change
conformation (shape) as well as position when they are subjected to an
external force. Nondeformable particles tend to have a substantially
stable conformation even when subjected to an external force, but may
undergo changes in position. Deformable particles are usually reversibly
deformable, e.g., they change conformation when an external force is
applied, and then return to a configuration comparable to their original
shape when application of the force is terminated.
This invention is particularly useful for measuring polymer molecules which
are folded, coiled or possibly even supercoiled when in a relaxed state,
and are subject to conformational changes such as stretching, bending,
twisting, contracting, etc., and positional changes such as rotating,
translating etc. This invention is particularly useful when an external
force is applied to molecules which are in some type of medium. However,
if a free solution is used, application of an external force may not be
needed to cause the molecules to change conformation or position.
Particles which are large enough to be seen using a microscope are measured
by visualization, e.g., by direct observation of a microscopic image.
Particles may, alternatively, be measured using microscopy combined with
any suitable spectroscopic technique, particularly if the particles are
too small to be imaged (viewed with acceptable resolution).
Several non-limiting examples of useful spectroscopic methods include using
polarized radiation as generated by a laser combined with measurement of
refractive index or fluorescence dichroism, or using sensitive video
cameras such as cooled charged coupled devices, silicon intensified target
devices, and micro-channel plate detectors.
Samples containing a mixture of both small and large particles, for
example, small and large DNA molecules are sized rapidly, with each
particle in the sample being measured simultaneously. The method of this
invention involves measuring conformational and positional changes of
individual, discrete molecules (or other particles), as contrasted to
method known in the art, which characterize a sample in bulk. The method
of this invention is applied to measure any number of particles, ranging
from a single particle to a large number of particles. If a sample
containing a large number of particles is measured, the number of
particles which are observed at one time will depend in part upon the
field of view of the microscope and the extent to which the particles are
separated from each other. Viewing discrete, individual particles, or
measuring their role of relaxation after applying an external force
permits complete deconvolution or separation of measured parameters.
The medium used in this invention is any suitable material. Preferably the
medium will hold relaxed particles in a relatively stationary position and
yet permit movement of particles which are subjected to an external force.
However, a free solution also may be used. For measurements of molecular
movement, a suitable medium is any medium which will permit different
particles to change conformation and position at different rates,
depending upon their size, and perhaps upon their chemical composition.
For many uses of this invention, the preferred medium is a gel or a liquid.
Preferably, the medium is anticonvective, but this is not absolutely
necessary. The medium may or may not be inert. The choice of an
appropriate medium will depend in part upon the size of the particles
which are measured, the tendency for the particles to change position and
shape, and the desired precision of the measurements. For example, when
large molecules (or other particles of similar size) are measured, a gel
with a large pore size is preferably used.
The external force applied to the particles is any force which causes the
non-deformable or deformable particles to undergo changes in conformation
or position. For example, the force may be an electric field, solvent flow
field, or a magnetic field, but is not limited to these types. The force
may vary in direction, duration and intensity. A particularly useful way
to perturb the particles is by using electrophoresis.
The types of changes which are measured in this invention primarily include
changes in conformation or shape, including stretching and relaxation
rates, as well as length and diameter (or radius) measurements, and
changes in position, including changes in orientation and rotation as well
as translation within the medium. Particles may undergo changes in
conformation or position, or both. Different types of changes are measured
according to various embodiments of the invention.
The techniques for measuring conformational and positional changes include,
but are not necessarily limited to, microscopy (alone), and microscopy
combined with spectroscopy. Several non-limiting examples of useful
spectroscopic techniques include birefringence, linear or circular
dichroism, and detection of fluorescence intensity.
Particles which are large enough to be seen under a microscope can be
measured by visualizing (imaging) the particles. As non-limiting examples,
a light microscope or a scanning/tunneling microscope may be used. While
particles may be viewed directly, it is useful to link the microscope to a
low light sensitive video camera, connected to a computerized image
processor (described in detail below) which records a series of
photographs, even a motion picture, by digitizing the images which are
received. The image processor may itself comprise a computer, or may be
linked to a computer which processes data based upon the images. Use of a
computerized apparatus enables the movement of each individual molecule to
be measured simultaneously. Furthermore, the relationship of molecules to
one another may be detected, and several different parameters of a single
particle can be measured simultaneously.
Optionally, the microscope and image processor are connected to a
spectroscopic apparatus. This technique is particularly useful for
particles which are too small to be visualized, but is useful for sizing
larger particles as well.
In order to transform measurements of change in conformation and position
into size measurements, it is generally necessary to generate (or
otherwise obtain) data relating to physical changes of particles of known
size when the particles are subject to external forces. "Markers" are
developed by measuring the parameters of molecules with known values of
molecular weight. This information may be input into the computer in order
to establish a relation between molecular weight and particular
conformational and positional changes which are measured. Preferably, the
markers are particles of similar structure to the particles of unknown
size (e.g., both particles contain the similar chemical components),
because rates of relaxation, reorientation and rotation may be dependent
upon particle composition. However, this may depend upon several
variables, e.g., polymer size, composition, etc., and thus it may not
always be necessary for the "markers" to have a composition similar to
that of the particles of unknown size.
Shear sensitive particles are particles which are subject to breaking when
they are placed on a microscope slide using conventional methods.
According to another aspect of this invention, such particles are
collapsed into a higher density conformation before they are placed in a
medium, in order to prevent breakage when the particles are mounted on a
microscope slide. Once they have been placed in the medium, they can be
uncollapsed and measured by the same methods as the smaller molecules.
In one embodiment of the invention, fluorescently stained, deformable
molecules which are coiled, folded or otherwise configured in a relaxed,
native conformation are placed in a medium and are temporarily deformed,
or stretched by applying an external force. When application of the force
is stopped, the relaxation time of the molecules (e.g., the time required
for the molecules to return to their original, relaxed state) is
determined by direct microscopic observation of molecular movement, or by
a combination of microscopy and spectroscopy. Alternatively, the kinetics
of stretching are measured by following the stretching of the molecule
after initiation of the external force. Rate measurements are calculated
in various ways, for example, by determining an amount of change per unit
time. Rates of change for molecules of unknown size are determined based
upon rates of molecules of known size, such as by interpolation or
extrapolation.
As with the viscoelastic measurement technique known in the art, the
relaxation time of particles in a liquid according to this embodiment
varies as about M.sup.1.66. In a gel, it is believed that resolution may
be as high as M.sup.2-4. This is based upon theoretical principles which
show that molecules rotate in gels or confining matrices, and their
relaxation time is much greater in a gel than in a solution (DeGennes, P.
E., Scaling Concepts in Polymer Physics, Cornell University Press, N.Y.
(1979).
In a second embodiment, the reorientation time of a deformable or
non-deformable particle is measured. When particles are first subject to a
perturbing force in one direction, and the direction of the perturbing
force is then changed, for example, by 90.degree., small particles quickly
reorient themselves and start a new migration along the new path. Larger
particles, on the other hand, remain substantially immobile until they are
reoriented in the direction of the electric field. Then, they too begin to
move in the new direction. By that time, the smaller particles will have
moved ahead. Measurements of the rate at which the position of a molecule
changes with respect to an external force may be measured, for example, by
measuring changes in position (e.g., lateral and/or rotational movement)
per unit time.
In a third embodiment, the rate at which a particle rotates is determined
when a series of external forces are applied. This method is particularly
applicable to rod-shaped molecules, such as small DNA molecules, and
elongated molecules which are maintained in a relatively uniform
conformation. "Rotation time" according to this invention is the amount of
time required for a molecule to undergo a positional rotation of a
particular angular increment, for example, 360.degree., when a particular
set of external forces are applied.
By periodically switching pulse direction, intensity and length, molecules
are caused to move slightly back and forth as they are rotated. This
facilitates rotation, and is analogous to the way in which an automobile
is manipulated into or out of a parallel parking space by alternating
backward and forward motion. However, unlike an automobile, a rod-shaped
or coil molecule may bend somewhat as it rotates. A pulsing routine may
also function to keep a deformable particle in a generally consistent
conformation, in order to provide useful measurements, e.g., measurements
which relate rotation time to molecular size.
Data for reorientation and/or rotation rates for particles of known size
may be used to develop a relationship between reorientation and/or
rotation rate and molecular size, which then may be used to determine the
size of various polymer molecules of similar composition and unknown size,
such as those which are present in a polydisperse sample. Reorientation
and rotation rate may be determined using microscopy (preferably combined
with image processing) to directly observe positional changes, or by
combining microscopy with spectroscopic measurements. Thus, these
embodiments are useful not only for mid-sized and large molecules, but
also for molecules that are too small to be imaged with acceptable
resolution.
In yet another embodiment of this invention, the length of a molecule or
other particle which has been placed in a medium is directly measured
using microscopy. This technique provides direct measurement of the
molecular size of any number of molecules. This method generally involves
observing the curvilinear length of deformed molecules which are in a
stretched state, e.g., during the application of an external force, or
soon after termination of a force which has stretched a molecule. However,
this method also may be applied to non-deformable molecules having an
elongated shape, and measurement of such molecules does not require
application of an external force before measurements are made. Preferably
this embodiment uses the same microscopy and imaging equipment as is
described above.
In a fifth embodiment, the diameter (or radius) of molecules or other
particles suspended in a medium is measured. Application of a perturbing
force is optional, because the diameter of a deformable molecule is
preferably measured when the molecule is in a relaxed state, and the
molecule is spherical, ellipsoidal or globular in shape. This embodiment
may be used to measure particles which are deformable or non-deformable,
and involves the use of a light microscope attached to a computerized
imaging device.
These five embodiments may be combined such that some or all of the
above-mentioned parameters are measured simultaneously for one or more
molecules.
A sixth embodiment of the invention is directed particularly to sizing very
large particles which tend to break if they are mounted on a microscope
slide using conventional methods. In brief, this new technique involves
collapsing the particles before they are placed in the medium, using an
agent which causes them to condense, and then uncollapsing the particles
after they have been placed in the medium. The molecules are then sized
according to the method of embodiments one to five. The method for
chemically collapsing molecules also may be used when it is desirable to
place a large number of molecules in a small area, such as in
microinjection, even if the molecules are not large or shear sensitive.
This invention provides a novel technique for mapping nucleic acid
molecules. For example, when a nucleic acid is placed in a matrix and
digested, the fragments are ordered by the computerized apparatus, and are
sized by the methods described above. Thus, the order of the digests is
quickly and accurately determined.
A further aspect of this invention provides for sequencing nucleic acid
molecules by hybridizing probes to portions of a molecule. A nucleic acid
is placed in a medium, to which suitable, desired probes are added. A
recombinational enzyme may be added if necessary. Reaction is initiated by
an appropriate means, for example, the addition of ATP (adenosine
triphosphate) and magnesium ions. After the probes have hybridized they
are detected by the methods described above, namely, microscopy (alone) or
microscopy in combination with spectroscopy.
Thus, the present invention provides an accurate method of determining the
size of individual particles and the weight distribution of a polydisperse
sample of particles. Another important advantage of this invention over
the techniques of the prior art is that the measurable parameters for each
of the particles in a polydisperse sample, not just the largest particle,
are determined. Additional advantages are that (1) only one molecule is
needed, and the sample may be very small, e.g., may consist of only one,
or only a few molecules (2) measurements may be based on one
representative particle for each size in the sample, (3) the technique can
be use | | |
