Transverse Urea-Gradient Gel Electrophoresis UNIT 7.

Monitoring the cooperative unfolding transition induced when a protein is exposed to BASIC
elevated temperature or a chemical denaturant is an important strategy for characterizing PROTOCOL
the conformational properties of a globular protein. This transition may be analyzed
quantitatively by a variety of spectroscopic techniques, but a simpler alternative is
urea-gradient gel electrophoresis. The pattern produced in the resulting gel can be used
to estimate both the free energy change for unfolding and the rate of the unfolding
transition. In addition, the technique can help identify either covalent or conformational
heterogeneity in a protein sample. Because urea-gradient gel patterns are sensitive to
several parameters, including hydrodynamic volume, net charge, and conformational
stability, the technique can be particularly useful for comparing two forms of a protein,
e.g., a natural form and the product of recombinant bacteria.
Transverse urea-gradient gels are slab gels prepared with a gradient of urea concentration
perpendicular to the direction of electrophoresis (Creighton, 1979). A single sample is
applied to the top of the gel, and as the protein is electrophoresed, molecules at different
positions across the gel are exposed to different urea concentrations. Unfolding is detected
by a decrease in electrophoretic mobility, reflecting the increased hydrodynamic volume
of the unfolded protein.
The gels are prepared with the glass plates rotated 90° from the orientation used for
electrophoresis, which requires minor modifications of most standard electrophoresis
equipment. The major requirements are spacers to fit the sides of the glass plates as
oriented for casting and some arrangement for holding the glass plates in this orientation.
An arrangement devised for preparing gels for the Bio-Rad Mini-Protean II apparatus is
illustrated in Figure 7.4.1. The spacers and a casting box for preparing multiple gels
simultaneously are available from Aquebogue. The gels are prepared using a photoacti-

90

casting electrophoresis

Figure 7.4.1 Spacers for preparing urea-gradient gels for the Bio-Rad Mini-Protean II apparatus.
(A) A gel “sandwich” assembled and oriented for casting. The spacers are designed fit along the
sides of the gel, as oriented for casting, and to keep the glass plates from sliding when placed in a
casting box. (B) After the gel has polymerized, the gel is rotated 90°, the casting spacers are
removed, and short spacers are placed on the sides of the gel to form a single sample well. The
spacers are made to prepare gels 1 mm thick and can be constructed by machining single pieces
of plastic or by glueing layers together. The standard glass plates for the Mini-Protean II apparatus
are used. In the figure, the thicknesses of the spacers and glass plates have been exaggerated for Characteristics of
Recombinant
clarity. Proteins

Contributed by David P. Goldenberg 7.4.1

Current Protocols in Protein Science (1996) 7.4.1-7.4.13
Copyright © 2000 by John Wiley & Sons, Inc. Supplement 3
 vating polymerization catalyst to allow sufficient time to prepare the gradient. After the
gels have polymerized, the spacers are removed to expose the top and bottom of the gel,
as oriented for electrophoresis. The gel is then preelectrophoresed and the sample is
loaded. After electrophoresis, the gel is stained and analyzed.
The following instructions assume that the total volume of the gels is 40 ml for five
10-cm-wide gels in the Aquebogue casting box. The gels are prepared so that 2 cm on the
left and right side (as oriented for electrophoresis) contain 0 and 8 M urea, respectively.
Volumes should be adjusted appropriately for gels of other dimensions.

Materials
Gel overlay solution: 20% (v/v) ethanol/0.002% (w/v) bromophenol blue
Urea/acrylamide gel solutions with photopolymerization catalysts (see recipes)
10× electrophoresis buffer (see recipe)
Protein samples
Acidic or basic protein sample buffer (see recipes)
Coomassie blue R-250 stain solution (see recipe)
Gel rinse solution: 50% (v/v) methanol/7.5% (v/v) acetic acid
Gel destain solution: 5% (v/v) methanol/7.5% (v/v) acetic acid
Gel electrophoresis tank and glass plates (e.g., Bio-Rad)
Spacers for casting gels (Aquebogue)
Gel casting box (Aquebogue)
Three-channel peristaltic pump (e.g., ISCO Tris pump)
Light source to initiate photopolymerization of gels (e.g., two 15-W blue
fluorescent lamps)
Air-displacement pipettor
Gel-loading tip
Plastic boxes

Set up apparatus for casting gradient gels

1. Assemble glass plates and spacers into sandwiches and place them in the casting box.
2. Connect casting box and peristaltic pump as illustrated in Figure 7.4.2 and set up the
reservoir and mixing chamber.
Position the pump so that it is at a slightly higher elevation than the mixing chamber and
slightly below the inlet to the casting box to minimize mixing of the gradient as the solution
of increasing density passes from the mixing chamber to the casting box.
3. Degas the gel overlay solution 5 min with a water aspirator.
4. Fill the bottom 2 to 4 cm of the casting box with overlay solution. Use a 20-ml syringe
to draw the solution through the tubing between the mixing chamber and the casing
box. Clamp the tubing between the peristaltic pump and the casting box. Make sure
that the level of the solution is still above the bottoms of the glass plates.

Prepare the gel solutions

5. Working in subdued lighting, prepare 30 ml each of 0 M and 8 M urea/acrylamide
gel solutions containing photopolymerization catalyst. Do not add TEMED (for
riboflavin polymerization) or methylene blue (for methylene blue polymerization).
Because of the time required to prepare the gradients, it is convenient to use a photoacti-
vated polymerization catalyst, such as riboflavin or methylene blue. The gels can be cast
under subdued lighting (complete darkness is not necessary), and then exposed to direct
Transverse light to initiate polymerization.
Urea-Gradient
Gel Electro- Appropriate pH and acrylamide concentrations of gel solutions must be chosen for each
phoresis protein (see Critical Parameters).
7.4.2
Supplement 3 Current Protocols in Protein Science
 gel sandwiches
in casting box resembles a label

mixing chamber

0M peristaltic pump
urea
8M
urea

magnetic stirrer reservoir

Figure 7.4.2 Preparation of urea-gradient gels using a three-channel peristaltic pump. The gel
sandwiches are held in a clear plastic casting box with an inlet for the gel solution placed below the
bottoms of the glass plates. Gel solutions containing 0 and 8 M urea are placed in the mixing
chamber and reservoir respectively. As solution from the mixing chamber is pumped into the casting
box using two channels of the pump, the 8 M urea solution is pumped into the mixing chamber using
a single channel, leading to a linear increase in urea concentration in the mixing chamber. Before
the gel solution is pumped into the casting box, the bottom of the box and the tubing between it and
the mixing chamber are filled with gel overlay solution. The flow rate from the pump to the casting
box should be ∼2 ml/min.

Urea should be dissolved just prior to preparing the gels, because urea in solution breaks
down to produce cyanate that can react with protein amino groups.
6. Degas the gel solutions 5 min with a water aspirator.
7. For photopolymerization, add 36 µl TEMED (for riboflavin) or 1.5 ml 2 mM
methylene blue.

Cast the urea-gradient gels

8. Place 12 ml of 8 M urea gel solution in the reservoir. Use a 20-ml syringe to fill the
tubing connecting the reservoir to the mixing chamber. Clamp the tubing in the
peristaltic pump and temporarily place the outlet in the 8 M urea reservoir.
9. Place 8 ml of 0 M urea solution in the mixing chamber. Remove the clamp between
the pump and the casting box. Turn on the peristaltic pump to begin pumping 0 M
urea solution into the casting box at a rate of ∼2 ml/min (1 ml/min per channel). When
the last of the 0 M urea solution has just entered the tubing, stop the pump.
Be careful not to allow any air into the tubing.
10. Place 12 ml of 0 M urea solution in the mixing chamber and position the outlet from
the 8 M urea solution in the mixing chamber. Turn on the magnetic stirrer and the
peristaltic pump.
The concentration of urea in the flask will now increase linearly as the solution is pumped
into the casting box. Characteristics of
Recombinant
Proteins

7.4.3
Current Protocols in Protein Science Supplement 3
 11. When the last of the solution has entered the tubing, stop the pump.
12. Place the remaining 8 M urea solution in the mixing chamber and turn on the
peristaltic pump. When the boundary between the ethanol overlay solution and the 0
M urea reaches the top of the gel plates, turn off the pump.
13. Clamp off the inlet to the casting box and disconnect it from the peristaltic pump.
14. Place the gels ∼10 cm from the light source to initiate polymerization, which should
take ∼30 min.
15. Promptly rinse the gel solution out of the tubing before it begins polymerizing.

Set up the gel

16. After the gels have polymerized, remove them from the casting box, remove the
spacers, and mount the gel in the electrophoresis chamber.
Urea-gradient gels can be stored 1 to 2 days at 4°C without significant dissipation of the
gradient. Before storing, wrap the gels in plastic wrap to prevent drying.
17. Dilute electophoresis buffer to 1× and fill the electrophoresis tank.
18. Preelectrophorese the gel 30 min at 100 V in the same direction as will be used for
electrophoresis.
For negatively charged proteins, electrophorese with the positive electrode at the bottom
of the gel; for positively charged proteins, electrophorese with the negative electrode at the
bottom of the gel.
These conditions are usually adequate to remove reactive compounds resulting from the
breakdown of urea or the polymerization process in a gel that is 1 mm thick and 10 cm
wide.

Prepare and electrophorese samples

19. Prepare the sample for electrophoresis by diluting ∼50 µg of protein to a total volume
of 60 µl. Add 15 µl sample buffer appropriate for acidic or basic protein.
Small samples of unfolded proteins in 8 M urea can be prepared by dissolving 48 mg solid
urea in 63 ìl aqueous solution (including the protein) to yield a total of 100 ìl. It is not
necessary to add glycerol to samples that contain 8 M urea.
20. Apply the sample evenly across the top of the gel using an air-displacement pipettor
with a gel-loading tip.
21. Electrophorese the sample.
The voltage and duration of electrophoresis appropriate for a given protein and set of
electrophoresis conditions must be determined empirically, but 50 to 100 V for 3 to 4 hr is
a good starting point.
22. At the end of the run, turn off the power supply and disconnect the electrodes. Remove
the gel from the apparatus and remove the plates. Mark the orientation of the gel by
cutting off one corner.

Stain and destain the gel

23. Place the gel in a plastic box containing ∼50 ml Coomassie blue R-250 stain solution.
Agitate gently ≥2 hr at room temperature.

Transverse
The high acid concentration of the recommended stain makes this solution particularly
Urea-Gradient effective for fixing smaller proteins that are not easily precipitated by milder staining
Gel Electro- conditions. For some applications, staining solutions commonly used for SDS gel
phoresis

7.4.4
Supplement 3 Current Protocols in Protein Science
 electrophoresis—e.g., 0.1% (w/v) Coomassie blue R-250/50% (v/v) methanol/7.5% (v/v)
acetic acid—may also be suitable.
24. Transfer the gel to a clean plastic box containing gel rinse solution and rinse ≤5 min.
25. Transfer the gel to a plastic box containing gel destain solution. Agitate gently and
change the gel destain solution as necessary to obtain a nearly clear background.

REAGENTS AND SOLUTIONS

Use Milli-Q-purified water or equivalent in all recipes and protocols steps. For common stock solutions,
see APPENDIX 2E; for suppliers, see SUPPLIERS APPENDIX.
Acidic protein sample buffer
50% (w/v) glycerol
0.01% (w/v) bromophenol blue
Store at room temperature
Acrylamide/bisacrylamide stock solution
30 g acrylamide (30% final)
0.8 g N,N′-methylene bisacrylamide (0.8% final)
H2O to 100 ml
Filter through paper or an 0.45-µm filter
Store at 4°C in the dark
CAUTION: Acrylamide and bisacrylamide are toxic. The powders should be handled in a
fume hood, and both the powder and solutions should be handled wearing gloves.
Basic protein sample buffer
50% (w/v) glycerol
0.2% (w/v) methyl green
Store at room temperature
Coomassie blue R-250 stain solution
0.5 g Coomassie blue R-250 dissolved in ∼400 ml H2O (1 mg/ml final)
50 g trichloroacetic acid (0.61 M final)
50 g 5-sulfosalicylic acid (0.39 M final)
H2O to 500 ml
Store at room temperature
CAUTION: Trichloroacetic acid and 5-sulfosalicylic acid are very caustic. Wear gloves
when handling the solids or stain solution.
Electrophoresis buffer, 10×
Buffer solutions that have been used successfully for urea-gradient gels include:
0.5 M acetate-Tris, pH 4.0: 0.5 M acetic acid adjusted to pH 4.0 with Tris base
0.5 M imidazole/0.5 M MOPS, pH ∼7.0
0.5 M Tris-acetate, pH 8.0: 0.5 M Tris base adjusted to pH 8.0 with acetic acid
0.5 M Tris/0.5 M Bicine, pH ∼8.4
Store all buffer solutions at room temperature
Urea/acrylamide gel solutions containing photopolymerization catalysts
0 M urea/15% acrylamide gel solution for riboflavin-polymerized gels
15 ml acrylamide/bisacrylamide stock solution (see recipe)
3 ml 10× electrophoresis buffer (see recipe; 1× final)
3.75 ml 0.04 mg/ml riboflavin (5 µg/ml final)
8.25 ml H2O
Immediately before casting gels, degas gel solution, then add 36 µl TEMED. Characteristics of
Recombinant
continued Proteins

7.4.5
Current Protocols in Protein Science Supplement 3
 8 M urea/11% acrylamide gel solution for riboflavin-polymerized gels
14.4 g solid urea
11 ml acrylamide/bisacrylamide stock solution (see recipe)
3 ml 10× electrophoresis buffer (see recipe; 1× final)
3.75 ml 0.04 mg/ml riboflavin (5 µg/ml final)
1.2 ml H2O
Immediately before casting gels, degas gel solutio, then add 36 µl TEMED.
0 M urea/11% acrylamide gel solution for methylene blue–polymerized gels
11 ml acrylamide/bisacrylamide stock solution (see recipe)
3 ml 10× electrophoresis buffer (see recipe; 1× final)
1.5 ml 20 mM sodium toluene sulfinate (1 mM final)
1.5 ml 1 mM diphenyliodonium chloride (50 µM final)
11.5 ml H2O
Immediately before casting gels, degas solution and add 1.5 ml 2 mM methylene
blue (100 µM final).
8 M urea/7% acrylamide gel solution for methylene blue–polymerized gels
14.4 g solid urea
7 ml acrylamide/bisacrylamide stock solution (see recipe)
3 ml 10× electrophoresis buffer (see recipe; 1× final)
1.5 ml 20 mM sodium toluene sulfinate (1 mM final)
1.5 ml 1 mM diphenyliodonium chloride (50 µM final)
4.4 ml H2O
Immediately before casting gels, degas solution and add 1.5 ml 2 mM methylene
blue (100 µM final).

COMMENTARY
Background Information
One of the most general properties of folded
globular proteins is the tendency to undergo a
cooperative unfolding transition when exposed
to elevated temperature or to chaotropic agents
such as urea or guanidinium chloride (Tanford,
1968). The unfolding transition for a particular
protein under defined conditions is charac-
terized by several thermodynamic properties,
including the free energy change for unfolding,
the enthalpy change, the entropy change, and,
even more fundamentally, whether or not the
transition is sufficiently cooperative that only
fully folded and fully unfolded molecules are
present at equilibrium. The free energy change
for unfolding, ∆Gu, is a particularly important
parameter that represents the net stability of the
folded conformation relative to the unfolded
state and is widely used for comparing the
stabilities of related proteins. Unfolding-re-
folding transitions are also characterized by
their kinetic properties, including their rates
and the accumulation of any transient interme-
Transverse diates.
Urea-Gradient
Gel Electro- The thermodynamics and kinetics of protein
phoresis unfolding and refolding reactions have been
studied extensively using a variety of different
biophysical methods, including spectroscopic
and calorimetric techniques. These studies have
led to important advances in understanding the
physical interactions that stabilize folded pro-
teins and the mechanisms by which these struc-
tures form (Schellman, 1987; Privalov, 1989;
Kim and Baldwin, 1990; Creighton, 1992; Mat-
thews, 1993). Although spectroscopic tech-
niques and calorimetry have been mastered in
a growing number of laboratories specializing
in protein stability and folding, these measure-
ments are quite technically demanding and re-
quire equipment and expertise that is not always
found in typical protein biochemistry laborato-
ries.
Urea-gradient gel electrophoresis was de-
veloped by T.E. Creighton (1979) as a simple
alternative to spectroscopic methods for study-
ing protein folding and unfolding. In this pro-
cedure, a single sample is applied across the top
of a polyacrylamide slab gel containing a hori-
zontal gradient of urea concentration. As the
sample is electrophoresed through the gel,
molecules at different positions across the gel
are exposed to increasing urea concentrations.

7.4.6
Supplement 3 Current Protocols in Protein Science
Those molecules that unfold have a larger hy-
drodynamic volume than the folded protein
molecules and therefore migrate more slowly.
After electrophoresis, the stained gel produces
a graphic representation of the unfolding tran-
sition. As discussed later (see Anticipated Re-
sults), the pattern generated can provide infor-
mation about whether or not the protein under-
goes a urea-induced unfolding transition,
information about the presence of significantly
populated intermediate states in this transition,
and semiquantitative information about the net
stability of the folded protein and the rates of
interconversion between the native and un-
folded states.
Although the data obtained from urea-gra-
dient gels are not generally as quantitative as
those derived from spectroscopic or calorimet-
ric measurements, the electrophoretic method
offers a number of advantages for many appli-
cations. In addition to being relatively simple
to implement, urea-gradient gel electrophoresis
requires only small amounts of protein—a 50-
µg sample is sufficient for detection with
Coomassie blue, and even less can be used with
more sensitive detection methods (UNIT 10.5 &
10.6). Furthermore, heterogeneous samples may
often be used, because different components
are likely to be separated during electrophoresis
(Hollecker and Creighton, 1982). Indeed, urea-
gradient electrophoresis is a particularly sensi-
tive method for detecting heterogeneity, be-
cause molecules that differ in their electro-
phoretic mobilities in either the native or
unfolded form can be distinguished, as can
molecules that differ only with respect to the
urea concentration at which they unfold.
In addition to being convenient, urea-gradi-
ent gels can often provide information about
protein-folding transitions that is complemen-
tary to the information obtained from more
conventional biophysical methods. Most spec-
troscopic techniques—e.g., UV absorbance
(UNIT 7.1), fluorescence, and circular di-
chroism—measure only the averaged proper-
ties of a sample. As a consequence, it is often
difficult to determine if, for instance, the signal
measured at an intermediate urea concentration
arises exclusively from a mixture of fully un-
folded and fully native molecules or if a par-
tially folded species may be present. Separation
techniques such as gel electrophoresis can
sometimes help resolve such ambiguities, par-
ticularly if different conformational states in-
terconvert relatively slowly on the time scale of
the separation.
Current Protocols in Protein Science
The most useful applications of urea-gradi-
ent gels are often those where a qualitative
comparison of different proteins is required, as
in comparing the protein produced in recombi-
nant bacteria with that obtained from a natural
source. If the recombinant-produced protein
has the same covalent structure and has folded
to the same conformation as that of the natural
protein, the two samples should display identi-
cal electrophoresis patterns. If, on the other
hand, the two proteins differ in their covalent
structure, this might be reflected in differences
in the mobilities of either the native or unfolded
forms, particularly if there are differences in net
charge. If the conformations of the folded pro-
teins are different, this may lead to differences
in the mobilities of the native form or in the
urea concentration at which the proteins unfold.
Other applications include screening protein
variants for differences in folding thermody-
namics and kinetics (Klemm et al., 1991;
Creighton and Shortle, 1994) and carrying out
preliminary experiments to define conditions
for more detailed studies using spectroscopic
methods.
Critical Parameters
Because the behavior of a protein during
urea-gradient gel electrophoresis depends on
several properties of the particular protein, in-
cluding its hydrodynamic volume, net charge,
and conformational stability, there are more
parameters that must be considered than for
SDS gel electrophoresis, for instance. These
parameters include pH, direction of electropho-
resis, gel composition, and duration of electro-
phoresis.
pH
Because electrophoresis depends on the in-
trinsic net charge of the protein, the pH must
be sufficiently far from the isoelectric point to
maintain a significant net charge, usually 5 to
10 charge units (the charge of a proton or
electron). For proteins with extremely high or
low isoelectric points, it will probably be nec-
essary to work below or above the isoelectric
point, respectively, but proteins with isoelectric
points near neutrality can be electrophoresed
under either acidic or basic conditions. Al-
though it may be desirable, in general, to use a
pH close to physiological, in some cases it may
be necessary to use more extreme conditions.
For instance, it may be necessary to use a higher
or lower pH to maintain a large enough net
Characteristics of
charge on the protein for electrophoresis or to Recombinant
Proteins
7.4.7
Supplement 3
 keep the protein soluble, because many proteins
are least soluble close to their isoelectric points.
In addition, pH can be used as a means to
manipulate the stability of the folded protein.
Many proteins are too stable to be unfolded in
8 M urea at neutral pH, but can be unfolded by
urea under alkaline or acidic conditions.
Several different buffers have been used
successfully for urea-gradient gels (see Re-
agents and Solutions). Ideally, the ionic
strength of the buffer solution should be as low
as possible to minimize the heat generated dur-
ing electrophoresis. Meeting this condition can
be facilitated by using buffers in which both the
anionic and cationic component provide buff-
ering capacity; several buffer systems of this
type have been described by McLellan (1982).
The direction of electrophoresis is determined
by the pH chosen. If the pH is above the isoelec-
tric point, the protein will have a net negative
charge and will migrate towards the anode in
the electrophoresis tank, as in SDS–polyacry-
lamide gel electrophoresis. If, on the other
hand, a pH below the isoelectric point is used,
the orientation of the electrodes must be re-
versed.
Gel composition
As in other forms of electrophoresis, the gel
composition must be chosen to match the hy-
drodynamic volume of the protein of interest,
with larger proteins requiring lower acrylamide
concentrations. The sieving properties of the
gel and, therefore, the choice of acrylamide
concentration also depend on the presence of
urea in the gel and the method of polymeriza-
tion used. For reasons that have not been fully
characterized, urea in a polyacrylamide gel
decreases the electrophoretic mobility, so that
even in the absence of an unfolding transition
there will be a progressive reduction of mobility
across a urea-gradient gel. This can be largely
compensated for by incorporating an acry-
lamide gradient so that there is a lower acry-
lamide concentration at the high-urea side of
the gel. In his original protocol, Creighton
(1979) used a 15% to 11% (w/v) acrylamide
gradient (photopolymerized with riboflavin)
superimposed on the 0 to 8 M urea gradient,
and these conditions have worked well for a
variety of monomeric proteins, including bo-
vine serum albumin, with Mr = 66,000. Larger
proteins, however, are likely to require lower
acrylamide concentrations.
Transverse Gel sieving is also greatly influenced by the
Urea-Gradient
Gel Electro- conditions used for polymerization. The two
phoresis catalysts that have been most widely used to
7.4.8
Supplement 3
polymerize acrylamide gels are ammonium
persulfate and riboflavin, the latter used for
photoactivation. Both are used in conjunction
with N,N,N′,N′-tetramethylethylenediamine
(TEMED). Photopolymerization is particularly
convenient for preparing gradient gels, because
the timing of polymerization can be easily con-
trolled. A major drawback to the use of ribofla-
vin, however, is that polymerization is not as
complete as when ammonium persulfate is
used. Although this is not a serious difficulty
for gels prepared at slightly acidic pH, at neutral
or alkaline pH the gels may be very soft or the
solution may not gel at all. The efficiency of
photopolymerization can be increased by add-
ing a small amount of ammonium persulfate,
but it can be very tricky to find a concentration
that will stimulate photopolymerization with-
out also causing early polymerization. More
recently, a procedure for photopolymerizing
gels using methylene blue has been described,
and it appears that this catalyst leads to much
more efficient polymerization than riboflavin
does (Lyubimova et al., 1993). Urea-gradient
gels containing 11% to 7% (w/v) acrylamide
and polymerized with methylene blue have
been reported to have sieving properties similar
to those of 15% to 11% (w/v) acrylamide gels
polymerized with riboflavin (Creighton and
Shortle, 1994).
Sample size
For urea-gradient gels ∼10 cm wide, a sam-
ple volume of 75 µl is appropriate. This is large
enough to apply evenly across the top of the
gel and small enough to give reasonably nar-
row bands. It should be possible, in principle,
to use discontinuous buffer systems with a
stacking gel, but this does not appear to be
necessary. Even in the absence of a discontinu-
ous buffer system, there is usually some stack-
ing effect when a macromolecular sample en-
ters a gel, because the first molecules to enter
the gel are retarded and accumulate near the
top of the gel while the rest of the molecules
enter the gel. The difference between the mo-
bilities of molecules above and within the gel
can be enhanced by making the ionic strength
of the sample lower than that of the electro-
phoresis buffer.
Duration of electrophoresis
The time required for electrophoretic sepa-
ration depends on the voltage applied to the gel,
and this parameter can be manipulated accord-
ing to the type of information desired. As dis-
cussed later (see Anticipated Results), the elec-
Current Protocols in Protein Science
trophoresis pattern generated depends to a large
degree on whether the interconversion between
native and unfolded states is rapid on the time
scale of the electrophoresis. In particular, if the
two forms interconvert many times during the
electrophoresis, then a smooth, continuous
band of protein should be generated, and the
position of the observed transition can be used
to estimate the thermodynamic stability of the
protein. On the other hand, if the two forms
interconvert slowly, a discontinuous or blurred
band will be generated, and under favorable
conditions it is possible to estimate the rate of
the transition from the appearance of the bands
and the duration of electrophoresis. Significant
separations have been observed after as little as
8 min of electrophoresis (Creighton, 1980), but
times as long as 10 hr have been used to gen-
erate smooth patterns for relatively slow tran-
sitions (Klemm et al., 1991).
An additional parameter that can be manipu-
lated to advantage is the presence of other
molecules or ions in the gel that influence the
unfolding transition by binding selectively to
either the native or unfolded forms. For in-
stance, the influence of a divalent cation can be
examined by comparing the patterns observed
when the ion is present in the electrophoresis
buffers with the patterns obtained when EDTA
is added to the buffers. This approach is limited,
of course, to compounds that do not interfere
with either gel polymerization or electrophore-
sis. Thiol reagents, for instance, react with free
radicals generated during polymerization, and
high salt concentrations lead to very high tem-
peratures during electrophoresis.
Electrophoresis
native protein
0 M
Urea gradient
Figure 7.4.3 Diagram of an unfolding curve produced by urea-gradient gel electrophoresis of a
protein undergoing rapid interconversion between the folded and unfolded states.
Current Protocols in Protein Science
Troubleshooting
The difficulties most likely to arise in using
urea-gradient gel electrophoresis are the forma-
tion of protein aggregates or covalent modifi-
cations either before or during the electropho-
resis. Aggregation is likely to lead to smearing
of the protein bands or a complete failure of the
protein to enter the gel. It may be possible to
minimize aggregation by choosing a different
pH or ionic strength for electrophoresis or by
decreasing the concentration of protein applied
to the gel. Covalent modifications are likely to
cause multiple protein bands, particularly if
they lead to changes in net charge. Such modi-
fications may arise from residual free radicals
from the polymerization reaction or from cy-
anates generated by hydrolysis of urea. To mini-
mize these reactions, the urea solutions should
be prepared immediately before casting the
gels, and the gels should be preelectrophoresed
before applying the sample.
Anticipated Results
The simplest patterns generated by urea-
gradient gel electrophoresis arise when the na-
tive (N) and unfolded (U) forms of the protein
are in rapid equilibrium and there are no sig-
nificant populations of partially folded inter-
mediates. In these cases, a simple sigmoidal
band is produced (see Fig. 7.4.3).
If the two forms are in rapid equilibrium on
the time scale of the electrophoresis, the band
should remain sharp through the transition re-
gion, and the same pattern should be obtained
whether native or urea-unfolded protein is ap-
plied to the gel. Provided that only the native
unfolded protein
8 M
Characteristics of
Recombinant
Proteins
7.4.9
Supplement 3
 and unfolded forms are significantly populated,
the observed mobility at the intermediate urea
concentrations will reflect the relative concen-
trations of N and U ([N], [U]), and the pattern
can be interpreted as a plot of the fraction of
protein unfolded versus urea concentration.
Although urea-gradient gels are used pri-
marily for qualitative analysis of protein un-
folding transitions, they may also be used to
obtain an estimate of the conformational stabil-
ity of a protein, provided that the conditions
described above are met. The free energy
change for unfolding is given by
[U]
∆Gu = −RT ln
[N]
where R is the gas constant and T is the absolute
temperature. The value of ∆Gu will depend on
the urea concentration, becoming more nega-
tive at higher concentrations. Because ∆Gu can
only be measured at urea concentrations where
N and U are both detectable, i.e., in the transi-
tion zone, it is usually necessary to extrapolate
values measured near the middle of the transi-
tion to 0 M urea in order to estimate the stability
under physiological conditions. Numerous
fu = 1
m = –4RT
∆C
0 2
Urea concentration (M)
Transverse
Urea-Gradient
Gel Electro-
phoresis Figure 7.4.4 Estimating the parameter m from a urea-gradient gel. See text for details.
7.4.10
Supplement 3
studies of protein unfolding transitions have
shown that the dependence of ∆Gu on denatur-
ant concentration can usually be well described
by a simple linear relationship,
∆Gu = m ⋅ (C − Cm)
where Cm is the urea concentration at the tran-
sition midpoint (i.e., where ∆Gu = 0, and the
average mobility is midway between that of the
native and unfolded forms), C is any other urea
concentration and m is an empirical parameter
(Schellman, 1978; Pace et al., 1989).
Hollecker and Creighton (1982) have shown
that m is proportional to the derivative of the
fraction unfolded (fu) with respect to urea con-
centration evaluated at the midpoint according
to the relationship:
dfu (Cm)
m = −4RT
dC
This derivative can be estimated from urea-gra-
dient gel patterns by drawing a line tangent to
the gel band at the midpoint and measuring the
change in urea concentration, ∆C, associated
with the extrapolated change in fu from 0 to 1.
This procedure is illustrated in Figure 7.4.4.
fu = 0
∆C
4 6 8
Current Protocols in Protein Science

Electrophoresis
0 M
Urea gradient
Figure 7.4.5 Urea-gradient gel pattern produced by a protein that slowly interconverts between
the native and unfolded states when native protein is applied to the gel.
The derivative is simply 1/∆C, and m is calcu-
lated according to:
−4RT
m=
∆C
From the value of m and Cm, the value of ∆Gu
can be calculated for any other urea concentra-
tion, including 0 M, where ∆Gu = −mCm (Hol-
lecker and Creighton, 1982; Klemm et al.,
1991).
Quantitating urea-gradient gel results in this
manner requires knowing the urea concentra-
tion at different positions across the gel, which
can be estimated from the dimensions of the gel
and from the known volumes of the linear
gradient and the two bordering segments con-
taining 0 and 8 M urea. Although values of Cm,
m, and ∆Gu estimated from urea-gradient gels
are certainly not expected to be as accurate as
those obtained by spectroscopic methods, they
have been found to be surprisingly consistent
with more rigorous measurements (Hollecker
and Creighton, 1982; Pace et al., 1989).
If the native and unfolded forms do not
interconvert rapidly during electrophoresis, a
more diffuse band or a discontinuity will be
generated in the transition zone (see Fig. 7.4.5).
In addition, different patterns are expected
when the native and unfolded forms are applied
to the gel, with the transition appearing slightly
above the equilibrium midpoint (Cm) when
native protein is applied and below Cm when
unfolded protein is applied. These patterns arise
because the rate constants for unfolding and
refolding, as well as the equilibrium constant,
vary with urea concentration, with unfolding
rates increasing and folding rates decreasing at
higher urea concentrations. When native pro-
Current Protocols in Protein Science
Cm
8 M
tein is applied to the gel, it remains in its initial
state at low urea concentrations both because
this state is thermodynamically favored and
because unfolding is very slow. At urea concen-
trations near the equilibrium midpoint, the na-
tive conformation does not predominate ther-
modynamically, but at least a fraction of the
molecules remain in this state simply because
they do not have time to unfold during the time
of electrophoresis. At urea concentrations
somewhat above the midpoint, the unfolding
rate is fast enough that most or all of the mole-
cules unfold during the electrophoresis. When
unfolded protein is applied to the gel, the op-
posite effect is seen, with a fraction of the
molecules remaining unfolded even at urea
concentrations below the equilibrium mid-
point.
When the unfolding and refolding rates are
not fast compared to the time of electrophore-
sis, the patterns cannot be used to estimate
thermodynamic parameters, but it may be pos-
sible to obtain some information about the
kinetics of the transition. If the half-time for
unfolding (or folding) at the midpoint is ap-
proximately equal to the time of electrophore-
sis, a smeared band will be generated in the
transition zone, and slower transitions will give
rise to a discontinuity, as described above. The
rate constants can be estimated by comparing
the observed patterns with those predicted by
numerical simulations, as discussed in greater
detail elsewhere (Creighton, 1979; Goldenberg
and Creighton, 1984; Goldenberg, 1989).
Much more complicated gel patterns can
arise if there are additional species present
Characteristics of
either at equilibrium or accumulating as tran- Recombinant
sient kinetic intermediates. If there is a partially Proteins
7.4.11
Supplement 3
 A

Electrophoresis
unfolded protein

native protein

0 M 8 M
Urea gradient

B Electrophoresis

unfolded protein

native protein

0 M 8 M
Urea gradient

Figure 7.4.6 Urea-gradient gel patterns produced by proteins with additional species. (A) A
protein with a partially unfolded species that predominates at intermediate urea concentrations. (B)
A protein with two unfolded species, one of which refolds very slowly.

unfolded species that predominates at interme-
diate urea concentrations, a two-step transition
may be observed (see Fig. 7.4.6A).
A frequently observed type of conforma-
tional heterogeneity is one in which there are
multiple unfolded forms that refold at different
rates. Multiple unfolded forms can arise from
isomerization of prolyl peptide bonds (Brandts
et al., 1975) and can give rise to the pattern
depicted in Figure 7.4.6B when unfolded pro-
tein is applied to the gel (Creighton, 1980). This
pattern is generated because some of the un-
folded molecules can rapidly refold and estab-
lish an equilibrium with the native protein, but
others do not have time to refold during the
electrophoresis and remain unfolded even at
urea concentrations at which the native protein
is thermodynamically preferred. At the lowest
urea concentrations, some of the slowly refold-
Transverse ing form may have time to fold to the native
Urea-Gradient form, giving rise to a blurred or discontinuous
Gel Electro-
phoresis
transition. Also, the slowly folding form may
equilibrate with partially folded forms, giving
rise to a downward curvature in the upper band.
When native protein is applied to the gel, only
the lower, sigmoidal, band is expected, because
the slowly refolding form is generated rela-
tively slowly after the protein unfolds. Other
complexities in the patterns may arise from
multiple folded forms, intermediates that con-
vert slowly with the native or unfolded protein,
or oligomeric proteins that undergo dissocia-
tion either prior to or in concert with unfolding.
Time Considerations
Most urea-gradient gel electrophoresis ex-
periments can easily be completed in 1 day. A
few hours should be allocated for preparing the
gels, and electrophoresis may take as little as
20 min or as long as several hours. Once pre-
pared, the gels can be kept at least 2 days at 4°C
without apparent changes in their properties.

7.4.12
Supplement 3 Current Protocols in Protein Science
Literature Cited
Brandts, J.F., Halvorson, H.R., and Brennan, M.
1975. Consideration of the possibility that the
slow step in protein denaturation reactions is due
to cis-trans isomerism of proline residues. Bio-
chemistry 14:4953-4963.
Creighton, T.E. 1979. Electrophoretic analysis of
the unfolding of proteins by urea. J. Mol. Biol.
129:235-264.
Creighton, T.E. 1980. Kinetic study of protein un-
folding and refolding using urea-gradient elec-
trophoresis. J. Mol. Biol. 137:61-80.
Creighton, T.E. 1992. Protein Folding. W.H. Free-
man, New York.
Creighton, T.E. and Shortle, D. 1994. Electro-
phoretic characterization of the denatured states
of staphylococcal nuclease. J. Mol. Biol.
242:670-682.
Goldenberg, D.P. 1989. Analysis of protein confor-
mation by gel electrophoresis. In Protein Struc-
ture: A Practical Approach (T.E. Creighton, ed.)
pp. 225-250. IRL Press, Oxford.
Goldenberg, D.P. and Creighton, T.E. 1984. Gel
electrophoresis in studies of protein conforma-
tion and folding. Anal. Biochem. 138:1-18.
Hollecker, M. and Creighton, T.E. 1982. Effect on
protein stability of reversing the charge on amino
groups. Biochim. Biophys. Acta 701:395-404.
Kim, P.S. and Baldwin, R.L. 1990. Intermediates in
the folding reactions of small proteins. Annu.
Rev. Biochem. 59:631-660.
Klemm, J.D., Wozniak, J.A., Alber, T., and Golden-
berg, D.P. 1991. Correlation between mutational
destabilization of phage T4 lysozyme and in-
creased unfolding rates. Biochemistry 30:589-
594.
Lyubimova, T., Caglio, S., Gelfi, C., Righetti, P.G.,
and Rabilloud, T. 1993. Photopolymerization of
polyacrylamide gels with methylene blue. Elec-
trophoresis 14:40-50.
Current Protocols in Protein Science
Matthews, C.R. 1993. Pathways of protein folding.
Annu. Rev. Biochem. 62:653-683.
McLellan, T. 1982. Electrophoresis buffers for
polyacrylamide gels at various pH. Anal. Bio-
chem. 126:94-99.
Pace, C.N., Shirley, B.A., and Thomson, J.A. 1989.
Measuring the conformational stability of a pro-
tein. In Protein Structure: A Practical Approach.
(T. E. Creighton, ed.) pp. 311-330. IRL Press,
Oxford.
Privalov, P.L. 1989. Thermodynamic problems of
protein structure. Annu. Rev. Biophys. Chem.
18:47-69.
Schellman, J.A. 1978. Solvent denaturation. Biopo-
lymers 17:1305-1322.
Schellman, J.A. 1987. The thermodynamic stability
of proteins. Annu. Rev. Biophys. Chem. 16:115-
137.
Tanford, C. 1968. Protein denaturation. Adv. Protein
Chem. 23:121-282.
Key References
Creighton, 1979. See above.
The original description of urea-gradient gels; in-
cludes extensive discussion of most of the important
parameters and considerations and detailed analy-
sis of the relationships between transition rates and
electrophoretic patterns.
Goldenberg and Creighton, 1984. See above.
General review of gel electrophoresis and applica-
tions to studies of protein folding, including urea
gradients.
Contributed by David P. Goldenberg
University of Utah
Salt Lake City, Utah
Characteristics of
Recombinant
Proteins
7.4.13
Supplement 3