Moss Tom. DNA-protein interactions: principles and protocols
Подождите немного. Документ загружается.
98 Papavassiliou
bands corresponding to free and bound species are cut out, and DNA is eluted.
The remaining steps in the procedure, beginning with organic extractions of
the eluates, are identical to those described in Subheading 3.4.1., steps 15–21.
1. Cut a piece of NA-45 membrane and four pieces of filter paper to the exact size
of the gel; cut the membrane between liner sheets wearing gloves.
2. To increase binding capacity, wash the membrane for 10 min in 10 mM EDTA,
pH 7.6, and for 5 min in 0.5 M NaOH, followed by several rapid washes in dis-
tilled, deionized water; let the membrane soak in 1X TBE buffer.
3. Remove the plate with gel from the Pyrex dish and place it on a flat surface.
Carefully lay two pieces of prewet (in 1X TBE buffer) filter paper onto the sur-
face of the gel, making sure no air bubbles are trapped between the filter paper
and gel.
4. Slowly and with extreme care, lift the gel (adhered to the filter paper) and place it
(with the gel facing up) on a clean glass plate. Wet the gel with a thin layer of 1X
TBE buffer.
5. Wearing gloves, lay the wet membrane sheet over the gel, again being careful not
to trap air bubbles beneath the membrane.
6. Complete the “sandwich” by placing the two remaining pieces of prewet (in 1X
TBE buffer) filter paper on top of the membrane.
7. Insert the “sandwich” of filter paper/gel/membrane/filter paper into a gel-holder
cassette, and load the assembly into one of the center slots in a (wet) transfer
apparatus (any of the commercially available electroblot units are suitable), with
the NA-45 membrane positioned between the gel and the anode (positive electrode).
8. Fill the transfer apparatus with 1X TBE buffer (precooled at 4°C) and transfer the
chemically cleaved double-stranded DNA fragments electrophoretically from
the gel to the NA-45 membrane. Electroblotting is performed at 4°C for 3 h, at either
20 V (approx 1 V/cm) if small DNA fragments (40–90 bp) are being transferred or at
35 V (approx 2 V/cm) if fragments >100 bp have been employed in the EMSA.
9. When transfer is completed, turn off the power, remove the gel “sandwich,” lift
the membrane sheet away from gel while wearing gloves, and rinse it in 20 mM
Tris–HCl, pH 8.0, and 0.1 mM EDTA to remove residual polyacrylamide. Do not
let the membrane dry!
10. Place the wet membrane (with transferred DNA face up) on the surface of a used
piece of X-ray film wrapped in plastic wrap and cover it with a tightly drawn
layer of plastic wrap. With a pad of Kimwipes, push out any trapped air bubbles
under the plastic wrap. Efficient transfer can be monitored by checking with a
Geiger counter.
11. Follow steps 2 and 3 in Subheading 3.4.1.; use a metal cassette instead of a
cardboard film holder to expose the membrane to X-ray film.
12. Autoradiograph the membrane at 4°C for 15–45 min (about one-fourth the time
required for the wet gel).
13. Follow steps 5–8 in Subheading 3.4.1. Using sterile forceps, transfer the wet
NA-45 membrane strips into siliconized 1.5-mL Eppendorf tubes.
In Gel
OP–Cu Footprinting 99
14. Add to each NA-45 membrane strip 0.6 mL of NA-45 membrane elution buffer
and spin for a few seconds in a microcentrifuge to submerge the whole strip.
15. Incubate at 55°C for 2–3 h in a water bath with agitation.
16. Vortex the tubes vigorously and pellet the NA-45 membrane strips by centrifug-
ing at room temperature for 10 s in a microcentrifuge.
17. Remove the buffer, and place it in a fresh siliconized 1.5-mL Eppendorf tube.
Monitor paper for loss of radioactivity; typically, approx 90% of the membrane-
bound DNA is released with this technique.
18. Follow steps 15–21 in Subheading 3.4.1.
3.5. Preparation of the G + A Sequencing Ladder
Provided the sequence of the DNA probe is known, you may perform at this
stage a Maxam–Gilbert guanine- and adenine-specific modification/cleavage
reaction (G + A sequencing ladder) of the end-labeled DNA fragment used in
the gel retardation assay. This reaction will be coelectrophoresed with the DNA
samples eluted from the free and bound fractions, to identify nucleotides pro-
tected from chemical cleavage (protein-contact sites) in the final stage of the
footprinting analysis (Subheading 3.6.). Below is a fast version (requiring only
1 h) of this otherwise time-consuming reaction.
1. In a siliconized 1.5-mL Eppendorf tube mix successively:
• 20,000 cpm of the end-labeled DNA fragment used in the EMSA
• 1.5 µL of carrier DNA stock solution
• Distilled, deionized water to 10 µL
2. Chill the tube on ice and add 1.5 µL of 88% aqueous formic acid.
3. Incubate at 37°C for 14 min in a water bath.
4. Chill again on ice and add 150 µL of freshly prepared 1.0 M aqueous piperidine.
Close the tube and wrap the cap tightly with a conformable self-sealing tape.
5. Heat at 90°C for 30 min in a thermostatted heating block, with the wells filled
with water. It is necessary to put a lead weight on top of the tube to prevent it
from popping open as pressure builds up inside.
6. Cool the tube on ice. Remove the conformable tape and spin for a few seconds in
a microcentrifuge; transfer to a fresh siliconized Eppendorf tube.
7. Add 1 mL of 1-butanol. Vortex vigorously until only one phase is obtained.
8. Mark the position of the tube in the rotor and spin at 12,000g for 2 min in a
microcentrifuge (room temperature).
9. Carefully remove and discard the supernatant using a drawn-out Pasteur pipet,
taking care not to disturb the tiny pellet or the area of the tube where the pellet
should be located.
10. Add 150 µL of 1% SDS and vortex the tube. Add 1 mL of 1-butanol. Mix well
by repeatedly inverting the tube. This step removes remaining traces of pip-
eridine trapped in the precipitate that interfere with the subsequent electro-
phoretic separation.
11. Resediment the precipitate by centrifuging at 12,000g for 2 min in a micro-
centrifuge and remove the supernatant as in step 9.
100 Papavassiliou
12. Spin for a few seconds in a microcentrifuge to collect any traces of liquid at the
bottom of the tube, and carefully remove it using a micropipet.
13. Dry the pelleted DNA for 5 min in a vacuum centrifuge.
14. Resuspend the samples (as in step 21 of Subheading 3.4.1.) in 5 µL of sequencing-
gel loading buffer, and store at –70°C until ready to load onto the sequencing gel.
3.6. Analysis of the Chemical Cleavage Products
on a DNA Sequencing Gel
Visualization of the length(s) on the DNA affected by protein(s) binding
specifically to it (i.e., the protein-binding site[s], or footprint[s] left by the
protein[s] on the DNA) requires electrophoretic fractionation of the single-
stranded fragments resulting from the chemical nuclease attack in a denaturing
polyacrylamide gel of the type employed in DNA sequencing, followed by
autoradiography. The location of the footprint(s) in the known DNA sequence
is identified by including the sequencing marker G + A track prepared in Sub-
heading 3.5.
1. Assemble and pour a 34 × 40 × 0.04-cm 6–15% (w/v) sequencing polyacryla-
mide gel, containing 1X TBE buffer and 8.3 M urea (20). The percentage of
acrylamide depends on the size of the DNA fragments to be separated as well as
on the size and location (relative to the labeled end) of the suspected protein-
binding site(s). As for the preparative EMSA, you should siliconize the inside
surface of the back glass plate to aid pouring into gel mold and removal at the end
of electrophoresis. To avoid dispersing of radioactivity across lanes (which might
produce significant errors in a subsequent densitometric analysis of the free and
bound DNA chemical cleavage patterns on the autoradiogram; see Note 12), it is
recommended to use a 0.4-mm custom-made sample comb with 5-mm lanes and
5-mm spacing. You may wrap the polymerized gel in plastic wrap and keep it at
room temperature until use (it can be stored as such for up to 36 h).
2. Attach the gel apparatus to the gel electrophoresis tank. Fill both the top and
bottom electrode chambers with 1X TBE buffer and remove the well-forming
comb. Check that wells are free from “tails” of polyacrylamide adhering to sides,
which may lead to uneven loading of samples and consequently band-shape dis-
tortion.
3. Pre-electrophorese the gel for 45–60 min before loading the samples. This
removes persulfate ions and heats the gel. Prerunning of the gel is performed at
constant temperature (approx 55°C), which is most easily achieved by applica-
tion of constant power (approx 50–70 W). If the surface temperature becomes
too high (>65°C), the glass plates will crack.
4. Thaw the DNA samples (if frozen), heat-denature them (including the G + A
sequencing ladder) at 90°C in a dry-block heater for 5 min, and quick-chill in
wet ice.
5. Disconnect the power supply, and immediately prior to applying the samples,
thoroughly rinse out (using a 10-mL syringe with a 22-gage needle) the wells of
In Gel
OP–Cu Footprinting 101
the gel with the upper reservoir TBE buffer; this prevents streaking of the DNA
samples caused by urea that has diffused into the wells.
6. Using calibrated glass capillaries with finely drawn tips or, preferably, dispos-
able flat-capillary pipet tips, load (as quickly as possible) 5 µL of each sample
(plus the G + A sequencing ladder) onto the wells of the sequencing gel, sweep-
ing the sample evenly from side to side. An untreated naked DNA sample (i.e.,
not subjected to the gel retardation assay) should always be diluted in sequenc-
ing-gel loading buffer, heat-denatured, and coelectrophoresed with the treated
samples to verify the integrity of the DNA, as single-strand nicks can mask pro-
tein binding sites or even create artificial ones.
7. Remove all bubbles from the bottom of the gel (they may prevent even migration
of the samples) using a 30-mL syringe with a bent 20-gage needle.
8. Run the gel under pre-electrophoresing conditions (constant power, approx
50–70 W), taking care not to overheat the glass plates. Uneven migration of frag-
ments (“smiling”) caused by an uneven gel temperature can be avoided by clamping
an aluminum plate to the front glass plate. It is customary to electrophorese the
samples until the bromophenol blue marker dye is about 3–5 cm from the bottom of
the gel, but longer electrophoresis times may be required to obtain single-band
resolution in the area of the footprint(s). The location of this region(s) depends on
the distance between the radioactive label and the suspected protein-binding site(s)
as well as on the length of the DNA fragment. Make use of available tables in the
literature (referring to the migration of oligodeoxynucleotides in sequencing gels
in relation to marker dyes) to determine how long to run your gel in order to
achieve the desired electrophoretic resolution in the region of the expected
footprint(s) (19); this will allow you to discern differences between the chemical
cleavage patterns of the free DNA and that derived from the complexed fraction(s).
9. After completion of electrophoresis, remove the gel from the apparatus, and with
the aid of a razor blade, slowly lift the siliconized plate. The thin polyacrylamide
sheet will stick to the unsiliconized plate. Fix the gel for 15–20 min by gently
immersing it (still attached to the lower plate) in a tank containing enough fixing
solution; this removes excess urea that would otherwise crystallize out.
10. Carefully remove the plate, bringing the gel on it from the tank, and lay it on a
flat surface. Place a prewet (in fixing solution) sheet of backing paper (cut slightly
bigger than gel dimensions) over the gel, press it gently down on the gel, roll out
any air pockets (using a 10-mL glass pipet), and peel it off patiently and with
extreme care together with the gel attached.
11. Cover the gel surface (but not the back of the filter paper) with a tightly drawn layer
of plastic wrap. With a Kimwipe, push out any trapped air bubbles under the plastic
wrap that might interfere with good uniform contact among the film, gel, and
screen. Add two sheets of filter paper next to the backing paper as a support pad and
put the “sandwich” into a gel dryer (paper pad closest to vacuum source).
12. Dry the gel under vacuum at 80°C for 45–60 min. Do not release the vacuum
before the gel is completely dried (sequencing gels with acrylamide concentra-
tions >10% are susceptible to fracturing).
102 Papavassiliou
13. In the dark room, place a sheet of Kodak X-Omat AR film against the plastic-
covered face of the gel. It is advisable to preflash the X-ray film, that is, exposing
the film to a 1-ms flash of light prior to placing it in contact with the sample to an
optical density of 0.15 (A
540
) above the absorbance of the unexposed film. This
increases sensitivity (all of the time of exposure to the radioactivity-generated
light produces blackening) and linearity of the film response (the degree of black-
ening above the background is proportional to the amount of radioactivity), which
are both essential if a densitometric analysis of the chemical cleavage products is
to be performed (see Note 11). Preflashing requires a photographic flash unit
appropriately fitted with filters and adjusted as described (21).
14. Autoradiograph in a metal cassette containing a single calcium tungstate intensi-
fying screen (place the flashed face of film toward intensifying screen) at –70° to
–80°C (to reduce scattering) for 12–16 h. Shorter or longer periods of time may
be also required, as the clarity of the footprint depends on the intensity of the
bands in the autoradiograph.
15. Immediately remove the film from the cassette and develop it, preferably in an
automatic processing machine for X-Omat films. If re-exposing the gel, it is nec-
essary to let the cassette warm up before inserting a second film (cold cassettes
will quickly collect moisture from the air).
16. Compare the chemical cleavage pattern of the naked DNA to that of the DNA–
protein complex(es). The position of a band in the gel corresponds to the distance
between the label and the point at which the DNA has been cleaved by the chemi-
cal reagent. Accordingly, bands at the bottom of the gel represent the smallest
end-labeled DNA fragments, increasing in size as one reads up the gel until the
pattern terminates abruptly in a strongly labeled band, corresponding to the
uncleaved full-length probe (see Fig. 3). The protected region(s) (indicating
sequence-specific protein binding) appears as an area resistant to cleavage (foot-
print), resulting in an almost complete absence of fragments (gap) arising from
within the protein-binding site(s) in the chemical cleavage pattern of the
complexed DNA. The nucleotides exhibiting protection are identified by align-
ing the bands in the cutting pattern of the free DNA with positions (bonds) in the
sequence of the coelectrophoresed Maxam–Gilbert marker G + A track. In this
comparison, it is necessary to note that, regardless of the end of the DNA frag-
ment labeled in the experiment (5' or 3'), the obtained set of products from the
chemical cleavage reaction matches the mobilities of the G + A sequencing frag-
ments exactly. This is a consequence of the identical 3' and 5' ends generated by
both chemistries at the cleavage points (22).
4. Notes
1. Optimization of the analytical EMSA. This is generally achieved by assessing
binding-reaction parameters and gel electrophoresis conditions. Furthermore,
success in interpretation of results from the coupled gel retardation/in situ
OP–Cu footprinting assay depends critically on some properties of the DNA frag-
ment used in the initial binding reaction.
In Gel
OP–Cu Footprinting 103
a. Properties of the DNA fragment used in the binding reaction. An EMSA
employing crude extracts works best with short DNA fragments, as these
reduce nonspecific interactions of proteins in the extract with sequences flank-
ing the specific binding site(s) and are able to detect binding of large proteins
more readily. Optimal sizes range between 100 and 150 bp, with the putative
protein-binding site(s) located at approximately the center or at least 20–25 bp
away from the radioactive labeled end. If a 20- to 25-bp synthetic oligonucle-
otide is to be used as a probe, it is advisable to design it in a way that it can be
readily subcloned into the polylinker region of a suitable vector, and then
labeled and released as a 40- to 45-bp restriction fragment, in order to obtain
the desired single-base resolution within and around the expected footprint(s).
Although the DNA fragment may be labeled at all ends for EMSAs, the sub-
sequent footprinting analysis requires the DNA fragment to be radioactively
labeled (to a high specific activity) at the 3' or 5' end of one of the two strands.
Klenow enzyme-labeled probes are preferable to kinased probes because some
protein extracts contain substantial phosphatase activities. Finally, the labeled
probe should be unnicked, because the resulting fragments may obscure the
cleavage pattern obtained after the chemical attack in the footprinting analy-
sis. Therefore, sufficient care should be taken to minimize nuclease activities
during all steps of preparation, labeling, and isolation. To this end, we have
found that purification of singly end-labeled probes from native polyacryla-
mide gels by “crush-and-soak” methods (similar to that described in Sub-
heading 3.4.1.) results in less damage to DNA than electroelution.
b. Binding-reaction parameters. These include binding-buffer composition (pH,
ionic strength, metal ion content, and presence or absence of nonionic deter-
gents and/or stabilizer polycations), amount of crude or partly-fractionated
extract or purified protein, concentration of labeled DNA probe, type and
amount of bulk carrier DNA, and temperature and duration of incubation.
Specifically, the following considerations should be evaluated: The optimal
ratio of protein to DNA for the assay is best determined by titrating a fixed
concentration of the radioactively labeled DNA fragment with increasing
amounts of crude or partly-fractionated extracts, or purified protein. Fre-
quently, as protein concentration increases, binding passes through a
maximum. Note, however, that increasing amounts of protein to a fixed con-
centration of DNA will not necessarily increase the yield of specific
complex(es) seen. This is because of the fact that whereas a given preparation
of any DNA-binding protein(s) tends to be fully active in nonspecific bind-
ing, it is typically only fractionally active in site-specific binding activity (the
apparent fractional activity varying from 5% to 75%, depending on the par-
ticular protein[s] and the individual sample). Therefore, too much protein,
particularly with crude extract preparations, leads to occlusion of the binding
site(s) by proteins interacting with DNA in a sequence-independent manner.
This problem can be minimized by raising simultaneously the concentration
of bulk carrier (competitor) DNA (typically of the order of 250- to 5000-fold
104 Papavassiliou
[w/w] excess over binding-site DNA); this increases the occupancy of the
binding site(s) by sequestering nonspecifically bound proteins, including non-
specific DNA-binding nucleases that may degrade the end-labeled DNA dur-
ing the binding incubation. Bear in mind, however, that although some
proteins are able to locate their target binding site(s) in the presence of vast
excesses of nonspecific natural DNAs (sonicated salmon sperm or calf thy-
mus DNA), other proteins cannot tolerate natural DNA carriers, but bind
readily in the presence of an excess of synthetic polynucleotides, such as poly
d(I-C) · d(I-C) or poly d(A-T) · d(A-T). It is noteworthy in the latter case that
the efficacy of competition for nonspecific binding can vary among different
batches from the same vendor. On the other hand, if too much carrier DNA is
added, it will compete for the specific factor(s) of interest, and the level of
complex(es) will decrease. Finally, provided an adequate resolution of DNA-
bound species is obtained for a fixed concentration of competitor, increasing
the amount of probe increases the fraction of DNA driven into complex(es),
until the limit set by the binding constant(s) is reached.
c. Gel electrophoresis conditions. Gel parameters, such as percentage of
acrylamide, degree of crosslinking, and pH and type of gel/running-buffer
system (high- or low-ionic-strength) dramatically affect the size, aggregation
state, and stability of DNA–protein complexes, hence their abundance and
quality of separation. Accordingly, it may be necessary to try more than one
gel fractionation/buffer system to obtain sufficient formation of the DNA–
protein complex(es) of interest. The electrophoresis time has to be optimized
for the complex(es) studied and the separation required (if more than one
complex has to be mapped). The most promising conditions can then be
applied in the subsequent preparative EMSA.
2. Binding competition analysis. If competitor DNA is identical to and in relatively
large excess over the labeled DNA, >90% of the radioactive signal should be
eliminated from complexes corresponding to protein(s) that interact with the
binding-site DNA in a sequence-specific manner. Complexes unaffected or only
slightly affected by the addition of competitor are thought to arise from nonspe-
cific, low-affinity binding by abundant proteins that are present in excess to bind-
ing-site DNA; DNA derived from these complexes after the in situ chemical
cleavage reactions can serve as a negative control in the footprinting analysis,
because its cutting pattern will closely resemble that of the free (unbound) probe.
3. Assaying relative dissociation rates of DNA–protein complexes. To follow dis-
sociation kinetics, complexes are allowed to form under optimal reaction condi-
tions and, at time zero, exposed to a large excess of an agent that does not
perturb their stability (commonly 100- to 250-fold mass excess of nonspecific
competitor DNA or, preferably, of the same DNA fragment unlabeled). The
“scavenger” molecules sequester the protein(s) as it dissociates from its specific
binding site(s), hence preventing it from rebinding. Analysis by the EMSA of
aliquots at various times (from a few seconds to 2 h) after quenching the reaction
with the sequestering agent shows the amount of free DNA increasing while the
In Gel
OP–Cu Footprinting 105
level of protein-bound label diminishes. The experiment can be designed in a
way that individual reactions can be started and quenched at different times, such
that all reach the point at which they will be applied to the gel more or less simul-
taneously and electrophoresed for the same period of time. Because dissociation
of typical DNA–protein complexes is a first-order process (i.e., independent of
the concentration of complexes), the results of the analytical study are also
applied to the subsequent preparative assay.
4. Optimizing the time of in situ chemical treatment. The length of incubation
period is determined by several factors, among which the following are of par-
ticular importance:
a. Kinetic stability of the complex(es) (step 3 in Subheading 3.1.). In principle,
the higher the dissociation rate of the complex(es), the lower the time of
exposure to the chemical nuclease. However, because of the gel “caging
effect” and the in situ (on the DNA surface) funneling of the reaction (see
Subheading 1.1.1.), this rule of thumb is applicable only for DNA–protein
complexes with half-lives either <<1 min or >>60 min; provided the
complex(es) is relatively abundant (see below), incubation times <8 min and
>30 min, respectively, should be used in these cases.
b. Relative abundance of the complex(es). The aim of the reaction is to generate,
on average, about one chemical cleavage event per DNA strand. Assuming
that the cleavage process is governed by Poisson statistics, the product oligo-
nucleotides will statistically be the result of a single cleavage (“single-hit
kinetics”) when the concentration of the full-length labeled strand is
approx 50–70% of its original value (i.e., approx 50–70% of the DNA frag-
ments should be left uncleaved). Accordingly, when the abundance of the
target complex(es) is very low (i.e., a bound DNA to free DNA ratio of <<1),
early termination of the reactions will be critical for high-“off”-rate com-
plexes, beneficial for intermediate-stability complexes, and safe for extremely
stable complexes.
c. Temperature of incubation. If optimization of the DNA-binding reaction and,
consequently, of the EMSA requires that both be performed at low tempera-
ture (4°C), you should carry out the chemical nuclease treatment at low tem-
perature as well. However, the amount of dissolved air oxygen in the reaction
mixture under these conditions is considerably higher (increases with decreas-
ing temperature); since molecular oxygen catalyzes a rate-limiting step that
generates in situ hydrogen peroxide (an essential coreactant for the chemical
nuclease activity; see Fig. 2), its presence in more than stoichiometric amounts
will shift the subsequent equilibria toward the right side, leading to increased
rates of DNA-strand scission. To compensate for this accelerated cleavage
kinetics, you should decrease the time of exposure to the chemical nuclease.
Inversely, incubation times should be longer than originally established (or
you may add hydrogen peroxide exogenously) for chemical treatments
performed at bench temperature during hot summer days in non-air-condi-
tioned rooms.
106 Papavassiliou
d. Concentration of reagents. Increasing the concentration of OP while holding
the concentrations of copper(II) sulfate and 3-mercaptopropionic acid con-
stant increases the overall rate of DNA-strand scission, without significantly
affecting the sequence preferences of cleavage and the resulting fragment size
distribution (15). On the other hand, substituting 3-mercaptopropionic acid
by the same concentration of ascorbic acid reduces the overall rate of DNA
cleavage both at low (1:1) and high (>4.5:1) 1,10-phenanthroline/copper
molar ratios (15). These observations are of practical significance to estimat-
ing the time of chemical treatment, particularly when extremely labile or
extraordinary stable complexes are being mapped. A higher concentration of
OP and a short incubation time might be used in the former case, whereas
ascorbate (as the reducing agent) and prolonged treatments are recommended
in the latter.
e. Presence of dithiothreitol (DTT). DTT (often employed in EMSAs to main-
tain the activity and/or stability of DNA-binding proteins) slows the DNA
cleavage rate because this chemical sequesters the copper necessary for the
oxidative cleavage to occur. If excessive amounts of DTT (i.e., >1 mM) have
been used either in the preparative binding reaction or for casting the gel in
which the (OP)
2
Cu
+
cleavage reaction is to be performed, longer incubation
times or increased (OP)
2
Cu
+
concentrations should be used to restore the
cleavage efficiency.
5. Trailing of the bands during electrophoresis. Glycerol-containing binding buff-
ers tend to cause significant trailing at the edges of the bands during electro-
phoresis. If you noticed this trailing under the optimized conditions in the
analytical assay, you may substitute glycerol for Ficoll (2.5% [w/v]; Type 400,
Pharmacia) in the preparative binding buffer. The presence of Ficoll (a copolymer
of sucrose and epichlorohydrin), although not interfering with the thermodynamic
and kinetic parameters of the binding reaction, gives rise to straight bands in the
gel (by tending not to spread so much because of surface-tension effects when
loading the sample), thus minimizing the size of the polyacrylamide strips in the
subsequent DNA-elution steps (Subheading 3.4.1.).
6. Loading the preparative polyacrylamide gel. If you have problems with the sample
not sinking to the bottom of the well (which may be caused by substantial
differences between the binding buffer and the electrophoresis buffer, and/or the
large volume of the preparative reaction), you can preload the well(s) of the gel
with binding buffer or, alternatively, load your sample with the power supply running
at 10–15 mA. (Wear dry plastic gloves and use only plastic tips if you do this!)
7. Preparation of the MPA, OP, and Cu
2+
solutions:
a. Use the recommended suppliers to obtain the liquid and powdered reagents.
The care given to the preparation of reagents is crucial. In particular, the water
used must be of the highest quality. Our laboratory uses only water purified
with a Milli-Q system (Millipore), which removes virtually all organics, ions,
and bacteria. Such precautions help prevent spurious reactions of impurities
with the reagents.
In Gel
OP–Cu Footprinting 107
b. You may try to footprint the complexes using a 1:1 ratio of 1,10-phenanth-
roline to copper, if your suspected protein-binding site(s) or the adjacent
regions are particularly rich in 5'-CG-3' elements (see Subheading 1.2.1.). In
this case, prepare the following solution: 100 mM of 3-mercaptopropionic
acid, 5 mM OP, and 5 mM CuSO
4
.
8. 2,9-Dimethyl–OP is a redox inert analog of OP which acts as a Cu
+
-specific
chelator; it will bind all the metal ion, preventing further oxidative chemistry and
DNA cleavage.
9. Phase inversion during organic extractions of the gel- or membrane-eluted DNA
samples. Because of the high-salt content of the elution buffers, the aqueous phase
(which normally forms the upper layer) may sometimes be dense enough to form
the lower layer. If this is the case, the aqueous phase can be easily identified by
monitoring the eluted radioactivity with a Geiger counter or by following the
strong yellow color of the organic phase (contributed by 8-hydroxyquinoline that
is added to phenol during equilibration as an antioxidant; see Subheading
2.4.1.1.).
10. Excess acrylamide in the gel-eluted DNA samples. If the DNA pellet is highly
contaminated with acrylamide monomers or other impurities from the gel matrix
(a problem sometimes encountered when employing low-percentage mobility-
shift gels and is usually apparent from the formation of an excessive turbidity
during the ethanol-precipitation step), you can further purify it by passing through
a pre-equilibrated Elutip™-d mini-column (an ion-exchange column) according
to the manufacturer’s instructions. Adsorption to and desorption from the col-
umn can be followed with a Geiger counter. Alternatively, you may follow the
experimental strategy developed by Ragnhildstveit et al. (22).
11. Wondering about the footprint: scanning the autoradiogram. If the effects of protein
binding on the cleavage rate of the chemical nuclease are not clearly discernible by
eye, or when partial protection is obtained, you may analyze the ladders quantita-
tively by subtracting the cleavage pattern of the DNA derived from the bound
fraction from that derived from the free fraction. This involves calculating the
probability of cleavage at each bond (which is related to the amount of radioactivity,
or intensity, in the corresponding band in the cutting pattern) and, finally, for each
lane, the average number of cuts in the DNA strand, with the aid of automated laser
densitometers linked to a computer (available from, for instance, Bio-Rad or LKB-
Pharmacia). Because quantification of band intensity is limited within the linear
response of Kodak X-Omat AR films to radioactivity (bands with an optical density
>0.15 and <1.8 absorbance units), it is essential not to use overexposed films for
this type of analysis. Moreover, the film should be free of scratches, fingerprints, or
other blemishes that will appear as optical signals indistinguishable from
32
P, thus
interfering with the calculations. A difference probability plot along the entire
length of the binding site(s) and surrounding DNA can be obtained, revealing the
protein-binding site(s) much more clearly and reliably than can be done by
comparison by eye of the chemical cleavage patterns. If still in doubt, however,
you may analyze the footprint on the other DNA strand.