Gel filtration Chromatography
Gel filtration chromatography, or molecular sieve chromatography,
separates macromolecules on the bases of Stokes radius. The
chromatography material is a composed of beads which contain pores.
There is a distribution of pore sizes, with the average size depending on
the resin. About 1/3 of the volume of liquid in a column containing gel
filtration medium (Sephadex, Sepharose, and others) is contained within
the beads. Proteins up to a certain size can partition into the beads. The
inside of the beads is a stationary phase. Buffer flows past the beads,
providing the applied force on the macromolecules to move down the
column. The liquid outside the beads is a mobile phase. Proteins that do
not partition into the beads because they are larger than the pores, move
down the column most rapidly. Proteins that do partition into the beads
are retarded in their elution because they are effectively in a larger
volume.
Gel filtration chromatography
Fap = solvent flow
Fret = partitioning into stationary phase particles
small molecules partition into beads
- This retards their progress down the column
Large particles are excluded from the bead interior
beads
-stationary phase
matrix
-elute faster
solvent flow
Vo
Ve
VI
Measure elution volume, Ve, or 
Vo = elution volume of totally excluded molecules
VI = elution volume of totally included molecules
 =
Ve - Vo
VI - Vo
elution volume
measure particle velocity
solvent velocity
Gel Filtration Chromatography
 depends on size and shape-
  0.9
small molecule
  0.1
large molecule
accessible volume within the matrix
 =
Ve - Vo
VI - Vo
Proteins that are excluded from the interior of the beads elute at the excluded volume
(Vo), whereas small molecules define the opposite extreme, the totally included volume
(VI). The difference, (VI-Vo) is the volume inside the beads. The elution volume of an
molecule will be in between the Vo and VI. The degree to which the molecule
of interest has access to the interior of the beads is measured as σ, defined above.
The elution volume accurately measures
the Stokes radius of a macromolecule.
Experimentally, the value of σ or the elution value, correlates with the
Stokes radius as measured by other hydrodynamics techniques. The same
factors of size and shape that result in altered frictional drag through
solution also determine the ease of penetration of a molecule into the
bead interior in gel filtration media.
Often gel filtration is calibrated using standards of known molecular
weight and calibrated in terms of molecular weight rather than Stokes
radius. This will be accurate only if all the standards as well as the
unknown sample have the same shape and specific volume, as we
described for the interpretation of sedimentation data. Hence, globular,
spherical proteins can be used as a set of standards, but if the unknown
protein has an elongated shape it will not penetrate the beads, and will
have a large Stokes radius. If the gel filtration data are calibrated in terms
of molecular weight, the answer obtained by gel filtration will be much
too large, since the elongated molecule behaves as a much larger sphere.
Example:
Elution of native globular protein using Sephadex G-200
Experimentally, the retardation of macromolecules in
gel filtration chromatography
correlates very well with the Stokes radius measured by Diffusion
If standards and the unknown have the same shape
then- and only then- does gel filtration chromatography
give a good estimate of molecular weight
sucrose
VI
1
3.0
cyt c
2.5
BSA

2
globular (spherical) proteins
blue dextran
Vo
totally included
catalase
0
1.5
Ve
Vo
1.0 totally excluded
104
105
Log(Molecular weight)
106
The previous graph shows a plot of the elution volume of a
series of globular proteins over a wide range of molecular
weights for the medium Sephadex G-100.
Often such graphs are used to calibrate a column and then
used to obtain M for an unknown protein.
This is incorrect, however, because of the unknown shape
factor. Since any deviation from spherical shape will result in
an increase in the Stokes radius, it can be concluded that any
molecular weight obtained by such a method is a maximum
value.
Active Transport in E. coli Mediated by
outer membrane proteins and TonB
Example: the FhuA transport
Protein from E. coli
-active transport system coupled
To TonB in the cytoplasmic membrane
-required for ferric ion transport
via ferrichrome complex
Science (1998)282,2202-
Express and characterize TonB without the N-terminal
membrane anchor
put a His-tag on the N-terminus
to facilitate one-step purification
H6-’TonB
Protein is soluble
Calculated molecular weight: 24,880
Size Exclusion Chromatography of H6-TonB
use Superose 12 column
Total volume: 21 ml
measured by the elution of
NaNO3
Void Volume: 7.3 ml
measured by the elution of
Dextran blue 2000
Calibrate vs Rs
thyroglobulin Rs = 8.6 nm; Ve = 8.8 ml
ferritin Rs = 6.3 nm; Ve = 10.7 ml
catalase Rs = 5.2 nm; Ve = 11.7 ml
aldolase Rs = 4.6 nm; Ve = 12.0 ml
bovine serum albumin Rs = 3.5 nm; Ve = 12.5 ml
ovalbumin Rs = 2.8 nm; Ve = 13.4 ml
chymotrypsinogen Rs = 2.1 nm; Ve = 14.9 ml
RNase Rs = 1.75 nm; Ve = 15.5 ml
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Size Exclusion Chromatography of H6-TonB
Rs = 4.1 nm for H6-TonB
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Sedimentation Velocity of H6-TonB
Measure S20,w = 1.4 S
Calculate M
 = 1.007 g/ml
 = 1.04 Poise
V = 0.7376 ml/g (from sequence)
Rs = 4.1 nm (from gel filtration)
Rs = M(1 - V)/6NS20,w
M = 28,000
Conclude that H6-TonB is a monomer in solution
Sedimentation Equilibrium of H6-TonB
0.8 mg/ml protein
18000 rpm
overnight 20oC
M = 30,000
Slope = M(1-V)
 = 1.007 g/ml
V = 0.7376 ml/g
Conclusion: H6-TonB is a monomer in solution
consistent with gel filtration/Sed. velocity
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Size Exclusion Chromatography of H6-TonB
Rs = 4.1 nm for H6-TonB
Calculate Rmin = (3MV/4N)1/3 = 1.9 nm
for anhydrous sphere with
M = 24,900 from sequence
Conclude: molecule is highly asymmetric
Rs/Rmin = 2.1
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
One can re-calculate Rmin assuming hydration of 0.3 g H2O/g protein
Rmin = 2.0 nm
so Rs/Rmin = 2
Consistent with an ellipsoid with an axial ratio of 15:1
240 Å x 16 Å
TonB goes from
the inner to the
outer bacterial
membrane
Size Exclusion Chromatography of FhuA
Rs = 4.8 nm for FhuA
TLN buffer contains detergent,some of which
is bound to the membrane protein to maintain
the protein in solution
consistent with protein plus bound detergent: Mr = 185,000
Mixture of FhuA and TonB Elutes at a Smaller
Volume in the presence of the FhuA Fe ligand
plus ferricrocin: Rs = 6.2 nm
no ferricrocin
Assuming a 1:1 Complex, Rmin can be estimated as 4.4 nm
so Rs/Rmin = 1.4 for the presumed complex. This is more
typical of globular proteins and much less asymmetric
than TonB alone
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Electrophoresis
The next (third) mass transport technique we will discuss is
electrophoresis. In this case the applied force is due to the electric field.
This will be dependent on the net charge on the macromolecule. In
solution, without any retarding matrix such as polyacrylamide or
agarose, the force of retardation would be proportional to the frictional
coefficient or Stokes radius.
The electrophoretic mobility is the velocity per unit of electric field,
related to the (Q/f) or (Q/Rs). Since the electrophoretic mobility depends
on both charge and size, it cannot be interpreted in any simple manner. A
folded protein, for example, might have a net positive or negative charge
and, therefore, might travel towards either the anode or cathode.
Furthermore, the effective charge is going to be dependent on the ionic
composition in solution, making this an even more complex problem.
Electrophoresis
Fap
Fret= f•(velocity)
applied force: electrical field
retarding force: frictional drag
moving through solution
Fap = Q•E
charge x electric field
in steady state: f•(velocity) = Q•E
electrophoretic mobility: U =
velocity
Q
=( )
E
f
depends on charge (Q) and Rs
However, usually electrophoresis is done in the presence of a
retarding matrix such a polyacrylamide
The ability of a macromolecule to move through the retarding
matrix depends on the Stokes radius
A commonly used retarding matrix is cross-linked polyacrylamide
The practical solution to
obtaining useful information
in the most straightforward
manner is to force the
migrating molecules to pass
through a retarding matrix
(agarose, polyacrylamide).
These crosslinked polymeric
substances make a mesh
though which the
macromolecule must
penetrate. The ease with which
a molecule can pass through
this matrix correlates with the
Stokes radius of the molecule.
Electrophoresis in a retarding matrix such as agarose or
Experimentally, the electrophoretic
polyacrylamide
log U = log Uo - KR (%T)
mobility is dependent on the percent of
acrylamide (%T), and the retardation
coefficient is a measure of the extent
of this dependence.
percent acrylamide
measured
mobility
mobility in
absence of gel
( depends on Q and Rs)
-
no matrix
+
The value of the electrophoretic
mobility is Uo. The crucial
element is that the retardation
coefficient, which describes the
extent to which the acrylamide
matrix itself retards the motion of
the macromolecule, depends only
on the Stokes radius and not on
charge.
-
Uo
+
retardation coefficient
Depends only on Rs
(size and shape)
U
matrix
Electrophoresis of Native proteins
:
log U = log Uo - KR (%T)
Uo is different for each protein
since it depends on both Q and Rs
Since the retardation coefficient KR depends only on Stokes radius and not on
charge, one can obtain Rs by determining KR
Measure the electrophoretic mobility vs % acrylamide
the slope gives KR which can be calibrated in terms of Stokes radius
this is called a Ferguson Plot
For native, folded proteins, one can obtain the retardation coefficient by measuring the
electrophoretic mobility (essentially, how far a protein migrates in a given time) in a series of
different polyacrylamide gels of different percentage. The higher amount of acrylamide results in
a finer mesh that the protein must penetrate. This results in a smaller electrophoretic mobility. A
plot of % acrylamide versus U is called a Ferguson plot, and the slope of the line obtained gives
the retardation coefficient for a particular protein.
Ferguson Plot to determine the Stokes radius of non-denatured protein
log U = log Uo - KR (% T)
Ferguson Plot
non-denatured proteins
KR = retardation coefficient
(%T) = % acrylamide
Uo - depends on Q and Rs
KR - depends only on Rs
If standards and the unknown
Protein have the same shape
Then the Rs is related to M
fast
unknown
increasing Rs
slope
log U
0 1
5
% acrylamide
10
slow
0
10,000
30,000
M
50,000
? M if standards and unknown have same shape
slope  Rs 
Example of the Application of the Ferguson Plot
GOAL: To determine the subunit composition of bacterial dioxygenases
of potential use in environmental cleanup of polychlorinated
biphenyls (PCPs)
purified enzymes contain two kinds of subunits, analyzed by
SDS polyacrylamide gel electrophoresis.
 subunit
 subunit
J. Biol Chem (2001)276,29833-38
Two enzymes were examined which have distinct
substrate specificities. In addition, a hybrid
enzyme was prepared and examined.
SDS-PAGE: gives subunit molecular weight and approximate ratio
Relative staining intensities
are consistent with 1:1 ratio of
the two subunits
lane 2. biphenyl dioxygenase
A1 =  subunit
A2 =  subunit
 (~50KDa)
 (~21 KDa)
lane 3. hybrid
C1 =  subunit
A2 =  subunit
lane 4. toluene dioxygenase
C1 =  subunit
C2 =  subunit
J. Biol Chem (2001)276,29833-38
Ferguson Plot Analysis of dioxygenases
1. for each protein (called “ISPxx”),
determine the mobility (Rf) as a
function of percent acryamide (%T)
2. Determine the slope (Kr) of
the plot of logRf vs %T
3. Plot logKr (slope) vs log (mol wt)
4. If standards and dioxygenases are
the same shape, then molecular
weight
can be determined. If the unknowns
are highly asymmetric, the molecular
weight will be incorrect.
Conclusions from Ferguson Plot anaylysis
1. biphenyl dioxygenase is a hexamer: 3 3
2. toluene dioxygenase is a tetramer: 2 2
3. hybrid enzyme is a hexamer: 3 3
J. Biol Chem (2001)276,29833-38
Electrophoresis
of double stranded DNA
of denatured RNA (no secondary structure)
of SDS - protein complexes
log U = log Uo - KR (%T)
Within each set:
(1) shape is the same
(2) charge and size vary proportionally
Uo
does not vary with molecular weight
Uo = Q  offsetting effects
f 
Hence, in comparing (for example) DNA samples, U varies in proportion to
Stokes radius
- No need to vary % acrylamide , so one gel is sufficient
Electrophoresis of SDS-Protein Complexes
The procedure of using the Ferguson plot requires running a number of
separate gels since the percent of acrylamide must be varied.
Furthermore, the unknown shape factor makes it risky to interpret the
results directly in terms of molecular weight. The strategy employed to
simplify and improve the usefulness of electrophoresis is to use dodecyl
sulfate (SDS) to denature the proteins, and then determine the
electrophoretic mobility of the SDS-protein complex. In many cases, this
allows one to obtain a reasonably accurate molecular weight of the
polypeptide by running only one gel.
Why does this work? There are two reasons.
1. The SDS complex results in the unfolding of the polypeptide, with
about 1.4 grams of SDS bound per gram of protein. The complex is very
highly asymmetric, like a chain which is covered with the negatively
charged SDS. Larger proteins essentially form longer chains. The longer
chains have both more charge (Q) and a larger frictional coefficient (f or
Rs), and these two have offsetting effects on the value of Uo.
Hence, in the absence of the polyacrylamide retarding matrix, all proteinSDS complexes will run with the same electrophoretic mobility.
2. The SDS-protein complexes all have the same shape. Hence, there is
a uniform relationship between the Stokes radius and the molecular
weight for all (with exceptions) proteins. Hence, KR really does vary
with molecular weight.
Constant value of Uo for SDS-Protein Complexes allows the Rs to be determined by
Determining the electrophoretic mobility on a single gel and comparing to standards
Cannot get Rs from
A single gel
SAME UO
Only one gel is
Needed to get Rs
fast
increasing Rs
0 1
5
% acrylamide
10
NATIVE PROTEINS
increasing Rs
log U
slow
0 1
5
% acrylamide
10
SDS-PROTEIN COMPLEXES
Since the SDS-Protein complexes all have the same shape:
Rs correlates well with molecular weight (M)
Electrophoresis of Protein - SDS Complexes
-
-
-
-
-
-
-highly asymmetric
-as length of protein increases, so does Q
log U = log Uo - Kr (%T)
-
-
-
-
-
Uo is the same for all complexes
relative mobility depends on Rs only
Since the shape is the same for all complexes
Rs  Molecular Weight
Problems with SDS-PAGE one can get incorrect answers!
-Uo might be different for unknown and for the standards
1. Membrane proteins (run fast)
2. Glycoproteins (run slow)
-more or less SDS bound: so (Q/f) is not the same
(usually 1.4 g SDS / g protein for a typical soluble protein)
-glycoproteins can have substantial carbohydrate component that
does not bind SDS
-shape may not be same due to incomplete unfolding (membrane proteins)
There are two major classes of proteins that present problems with this
approach.
1) Membrane proteins are very hydrophobic and may not unfold
completely, even in the presence of SDS. Furthermore, the very
hydrophobic nature of the side chains may result in abnormal (excess)
binding of SDS. Hence, the assumption that Uo is the same for the
unknown as for the standards may not be correct. These effects cause
hydrophobic membrane proteins to run faster in SDS-PAGE than they
should. As a result, the molecular weights are often underestimated very
significantly.
2) The second class of proteins that don’t behave similarly as the
standards for calibration, are glycoproteins. These have covalently
attached carbohydrate residues, which interfere with the binding to SDS.
Hence, glycoproteins usually run slower than they would do otherwise in
SDS-PAGE and the molecular weights are overestimated.
SDS-PAGE of Purified H6-TonB Gives an incorrect
value of the molecular weight
Expected molecular weight: 24.9 kDa
Estimated Mr from SDS-PAGE Mobility: 35 kDa
same as wild type TonB
reason is not known
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Chemical Crosslinking of FhuA and His6-TonB
Shows the presence of a 1:1 Complex Enhanced
by ferricrocin and inhibited by 1 M NaCl
Journal of Bacteriology, May 2001, p. 2755-2764, Vol. 183, No. 9
Electrophoresis
of Small Double Strand DNA and of RNA
It is standard procedure to separate linear, double strand DNA species according to the
number of basepairs by polyacrylamide or agarose electrophoresis. Only a single gel is
necessary, and given standards run along side of an unknown, to determine the
molecular weight (or number of base pairs) of DNA. The reason this can be done is the
same as with SDS-protein complexes.
1. Different lengths of DNA all have the same value of Uo since both charge
and size vary proportionately.
2. All the DNA molecules have the same shape if they are linear, so the Stokes
radius varies in proportion to the number of base pairs. This will not work if the DNA is
circular or supercoiled.
These arguments do not hold for RNA, which has considerable tertiary structure.
Hence, the molecular weight of native RNA cannot be determined as for DNA, but
presents the same complications as native, folded proteins. As a result of this it is
necessary to denature RNA during the electrophoresis. Using alkaline conditions, or
solvents like dimethylformamide, which disrupt the tertiary structure and secondary
structure of RNA. Under these conditions, as with double strand DNA, and as with
SDS-protein complexes, the only difference between RNA species that results in
different electrophoretic mobilities is the length of the RNA strand.
Electrophoresis of double-strand DNA
log U = log Uo - KR (%T)
constant
Relative mobility in gel  Rs  # basepairs
Uo constant
Same shape for unknown and
standards
compare with standards
Whenever you have secondary structure this approach fails
1. Uo need not be the same for unknown and standard set
2. Rs will not simply be related to # basepairs (or mol wt) if the
shape of the unknown is not the same as the shape of the standard set
-Hence, one needs to denature RNA to obtain the correct molecular weight from
a single gel.
Stacking of the Sample in Electrophoresis
Electrophoresis is a zonal method, in which the sample is layered on top
of the separating retarding matrix and the mobility is measured by
observing the migrating band of material. Hence, it is important to have a
narrow band of material to optimize the resolution of the technique.
Stacking is used to produce a very sharp band of material prior to
entering the separating gel. The procedure is shown below.
Stacking and Discontinuous Gels
Zonal analysis depends on having sharp, well defined bands
1 create a large electric field (voltage drop) in the sample buffer to concentrate
the protein prior to separation by the gel
sample
buffer
+
want large Electric Field to rapidly
pile up protein at the surface of the
gel.
The logic
1. The charges moving through the gel constitutes the current in the
presence of the applied voltage. The flux of charges across any reference
plane through the gel must be the same.
2. The flux (J) of ions of any particular type is the product of the
concentration of the ionic species multiplied by the net velocity of the
ions.
3. The electrophoretic mobility of each ion is different and is defined as
the ion velocity per unit electric field (U= V/E).
4. This leads to the basic equation that
J1 = C1U1E1 = C2U2E2 = J2
This means that at places in the gel where either the concentration or the mobility of the
major charge carrier is low, the result is a high electric field. This just means that where
the resistance to current flow is high, there will be a large voltage drop (meaning a large
electric field). In the Laemmli buffer system, used routinely for SDS PAGE systems,
there is a discontinuous buffer (hence, disc-gel electrophoresis). Glycine is the major
buffer in the sample, and the pH is such that the average charge is low on the glycine,
reducing its electrophoretic mobility (U1in the next slide). Chloride is the major ionic
carrier of charge in the gel itself. As a result of the electric resistance due to the low
mobility of glycine, the electric field across this part of the system is very high (E1).
This is where the protein is loaded, so the SDS-protein complex is subject to this very
large field, which causes the complexes to move rapidly down, where all the proteins
pile up on top of the separating acrylamide matrix. This forms a very narrow band,
which then enters the gel, and separation takes place. After this initial formation of the
narrow band, during the rest of the process, the band broadens by diffusion. This is
illustrated in the next slide.
The alternative way to achieve the same high resistance to current flow in the region
where the sample is first loaded onto the gel, is by simply lowering the ion
concentration in this region. Hence, the same ion can be in both the separating gel and
in the sample, but in the sample the buffer is diluted 10-fold.
Stacking
-
current (ion flux) must
be constant throughout
the gel
J1
NH2CH2CO2- (charged)
pH-dependent
mobility
glycine-
NH3+CH2CO2- (neutral)
Cl-
J2
+
J1 = J2
C1V1 = C2V2
C = concentration
V = net velocity
U = Mobility = ( V )
E
glycine
J1 = C1U1E1 = C2U2E2 = J2
Cl-
if (C1U1) << (C2U2)  E1 >> E2
Protein in the glycine buffer experiences a high electric field because of the low
electrophoretic mobility of glycine at the pH used
-the protein moves to the ion boundary if
Ugly < Uprot < UCl-
Stacking
to
_
glycine-
t1
_
t2
_
protein in sample
buffer
Cl-
gel
ion boundary
gel
+
start
+
stack
+
sieve
Uglycine- < Uce
(depends on pH)
Alternative: Dilute running buffer (1/10) to make sample buffer (U1= U2)
J1 = C1U1E1 = C2U1E2 = J2
(low) (high) in sample region
Two variants of gel electrophoresis
1. Gel mobility-shift assay for protein-DNA interactions
2. Pulsed field gel electrophoresis for separating very large DNA (chromosomes)
Electrophoresis of Protein-DNA complexes
An interesting application of electrophoresis technology is the gel
mobility shift assay for DNA-protein complexes.
This is summarized in teh next slide. The main idea is quite simple. A
protein complex with DNA is formed in the sample. The mobility of this
complex is different (less than) that of the DNA alone. As a result, the
complex is visualized as a separate band in the gel after it is run and then
visualized for the presence of the DNA. Small fragments of DNA are
used in these assays.
Gel Mobility - Shift Assay
for quantitative and qualitative characterization of DNA-protein interactions
+
DNA
protein
(eg lac repressor)
DNA-protein complex
1 Equilibrate Reaction Mixture
sample
gel
3 Run gel
+ protein
DNAProtein
complex
2 Load onto gel (e.g., 7.5% Polyacrylamide Tris-BorateEDTA buffer)
- protein
DNA
4 Stain with ethidium bromide for DNA (or detect using
radioactive labels)
Gel Mobility shift assay
Why does this work?
1. Free DNA and Bound DNA species (along with free protein) are stacked
and moved into the gel before there is any time for protein dissociation
Takes ~ 1 min to complete this process
+
koff
1
koff
> 1 min
2 Once in the gel, the DNA-protein complexes are virtually locked
together
WHY?
- low salt
- excluded volume effect of gel
- “cage” effect of the gel
(not fully understood)
The “cage effect” of the gel is such that when the protein and DNA
dissociate they are trapped by the gel and cannot diffuse apart. Hence,
the entropic advantage to dissociation is lost, so the effective dissociation
constant shifts to greatly favor the complex staying together.
The mobility shift assay can be used qualitatively to detect the presence
of complexes, or it can be used quantitatively to determine the affinity
constant in solution between the protein and DNA. This is because the
assay provides information about the amount of DNA that is free in
solution and the amount which is bound to the protein in the solution
prior to getting trapped in the gel. As shown below, enough parameters
can be measured to get a binding isotherm and a dissociation constant.
Because of the ability to detect very low concentrations of DNA, very
high affinity binding can be measured.
Gel Mobility Shift Assay
can provide all the information for a binding isotherm
After running the gel: visualize the DNA: protein-DNA complex is retarded
[DNA-Prot]
[DNA]
[DNA]total = [DNA]free + [DNA-protein]
know this measure relative amounts from gel
[protein]total = [protein]free + [DNA-protein]
know this calculate this measure from gel
[DNA
-prot]
can measure Kd ~10-8 to 10-12 M
[protein]free
(see JBC (1991) 266 13661-)
Gel Mobility Shift Assay of a Protein-RNA Complex
Goal is to measure the Kd of the complex formed between a t-RNAgln
mutant and the glutaminyl-tRNA synthetase
Residues altered in high affinity mutant
Gel shift assay
1. Use radioactive label on tRNA to detect on the gel
2. Incubate tRNA/protein mixture for 15 min
200 pM tRNA
6.6 nM to 66 pM Gln-tRNA Synthase (GlnRS)
3. Load onto 20% polyacrylamide gel and run for 5 h at 4o C
4. Autoradiography to determine bound and free tRNA
5. Fit to binding isotherm for 1:1 complex formation
Increased [protein]
Binding isotherm for tRNA/protein Complex
Kd = 0.27 nM
Pulsed Field Gel Electrophoresis
In normal electrophoresis - electrophoretic mobility is independent of
molecular weight for large DNA (> 50 kbp)
because it becomes elongated in the electric field
elongate
relax
Pulsed field gel electrophoresis is designed
for separating very large DNA fragments (>100 kbp)
and takes advantage of this relaxation/elongation property
(PNAS (1987) 84, 8011 - 8015;PNAS (1991) 88, 11071 - 11075)
In pulsed field gel electrophoresis, the DNA is allowed to “relax”
after a brief pulse of electric field and then the direction
of the electric field is changed
- this results in a strong length-dependence of electrophoretic behavior
E
t1
t2
t3
t4
Several variations of technique using different field direction
changes and pulse times
critical parameters
- DNA relaxation time (TR) vs electrophoresis pulse time (TP)
elongate
relax (TR)
Field Inversion Gel Electrophoresis
(FIGE).
Another interesting and useful application of electrophoresis technology
is field inversion gel electrophoresis (FIGE), which is used to separate
very large DNA. The usual agarose electrophoresis is limited in the
ability to separate DNA of sizes larger than 50 kbp (50,000 basepairs).
This is because the DNA becomes elongated in the electric field and the
gel no longer separates. FIGE takes advantage of this elongation
produced by the electric field to separate DNA fragments larger than 100
kbp. The main idea is to change the direction and of the electric field at a
certain pulse rate (TP, pulse time). The large DNA is elongated by the
field in the initial direction, and this elongation makes it much easier for
the DNA to move in this direction, and very difficult for it to penetrate
the gel in the perpendicular direction. The amount of time it takes the
DNA to relax back to a spherical shape from the elongation is dependent
on size.
If the electric field direction is changed before the DNA has had a chance
to relax, then the DNA will not be able to respond to the altered direction
and move in the new direction since it is elongated in the “wrong” way.
Smaller DNA will relax more rapidly, and will move along with the
altered field direction. Large DNA will not respond.
Hence, the path of migration in the gel will vary with DNA size. The
critical parameters are the pulse time (TP) for the electric field and the
relaxation time for the DNA elongation (TR).
The next slide shows three conditions, where the relaxation time is much
shorter than, equal to, or much larger than the pulse time. The mobility of
the DNA actually is minimized when the pulse time is about equal to the
DNA relaxation time, as pictured.
Alternate field direction by 90o
pulse time: Tp
small DNA
responds rapidly
to changing field
larger DNA
TR  Tp
slow migration: field changes just
as DNA changes shape
(minimum mobility)
Very large DNA
After continuous
field along ^x, no
^
adjustment along y,
so it resumes motion
in ^x direction
-faster mobility
PNAS 88 11071