The choice of solvent systems for multidimensional chromatography of postlabeled DNA adducts is dependent on the type of adducts to be chromatographed and the removal of interfering 32P-labeled compounds (i.e., normal nucleotides) that can contribute to the radioactive background on the chromatograms. Urea (used to reduce the interaction of hydrophobic molecules with the PEI-cellulose sheets), pH, and salt concentrations all influence the rate at which adducts migrate. Increasing the solvent strengths of the salts or urea, or both, will generally increase migration. However, as the solubility limits of the solvent components are approached, the risk of these components precipitating out of solution onto the PEI-cellulose sheets increases rapidly. A consequence of this precipitation is the tailing of radioactivity on the chromatograms which can hamper quantitation. Concentrations of urea greater than 8.5 M should be avoided because chromatography times increase significantly without any appreciable improvement in adduct migration. Generally, we dip the edge of the PEI-cellulose sheet in water or the buffer component of the urea/buffer solvents before placing the sheet in the solvent to reduce the problem of supersaturation and precipitation of salts in the solvent front. For large hydrophobic structures derived from PAHs a 1 M sodium phosphate, pH 6.0, solvent is generally used for D1 to migrate normal nucleotides and nonspecific radioactivity off the PEI-cellulose sheets and onto paper wicks. However, for faster migrating and less hydrophobic structures, the phosphate concentration may have to be raised to 2.3 M sodium phosphate, pH 5.5, to get the DNA adducts to salt out of solution and onto the sheets. The lower pH (5.5) is necessary to minimize the problem of sodium phosphate crystallizing on the PEI-cellulose sheets when using phosphate solutions that are 1.7 M or greater. Also, raising the laboratory temperature and sealing the tanks will help to reduce this problem. However, increasing the ionic strength of the phosphate buffer will also cause some interfering 32P-labeled structures to salt out onto the sheets, and the resulting autoradiograms may not be as clean (Beach and Gupta 1992). An optional clean-up chromatography step can be done in the D2 direction (Fig. 5) using 2.5 M ammonium formate, pH 3.5, although this step is not necessary when chromatographing bulky, hydrophobic DNA adducts like those derived from PAHs (>2 rings). The solvent systems for D3 and D4 are used to separate the large, hydrophobic, xenobiotic DNA adducts. The most commonly used D3 solvent system for large, hydrophobic adducts on commercial sheets is 3.5 to 4.5 M lithium formate, 8.5 M urea, pH 3.5 (LFU). If the DNA adducts in D3 are extremely slow migrating then consider 3.5 to 4.5 M pyridinium formate, 8 M urea, and pH 3.4 (made by neutralizing formic acid with pyridine to desired pH). The traditional solvents for D4 are 1.6 M lithium chloride, 0.5 M Tris, 8.5 M urea, pH 8.0 (LTU) or 0.8 M sodium phosphate, 0.5 M Tris, 8.5 M urea, pH 8.2 (PTU) for commercial PEI-cellulose sheets. When using laboratory prepared sheets, dilute the solvents to 80 to 90% of the strength used for commercial sheets.
The PAH-DNA adducts are generally found within a diagonally shaped region when LFU is used for D3 and LTU is used for D4. If the diagonal radioactive zone (DRZ) formed on the chromatogram (as visualized in the autoradiogram) is strong with many intense overlapping spots, then the isopropanol:4N ammonia system is an excellent alternative for D4, because it will disperse the adducts over a larger area of the chromatogram (Fig. 6).
Another promising D4 solvent system is 0.08 to 0.4 M NH4OH that gives a zone of adducts more disperse than LTU but less than the isopropanol/ammonia system (Spencer et al. 1993). In addition, this system develops rapidly, and chromatography for D4 is complete in approximately 0.5 hours. However, if the DRZ is weak, then the LTU system would be superior because the DNA adducts are compressed into a zone of overlapping spots that can be easily quantitated. Sometimes when autoradiograms are hazy, a final development (D5) in 1 M sodium phosphate, pH 6.8, can improve autoradiogram quality by removing some of the nonspecific radioactivity from the chromatogram.
Some of the solvents used for chromatography are based on formic acid or acetic acid. The concentration of these solutions is based on anion concentration, for example 4 M lithium or ammonium formate, pH 3.5 would have 4 moles per liter of formic acid and sufficient lithium hydroxide (or ammonia) would be added to bring the pH to 3.5. Another important consideration is the type of pH electrode used. With Tris solutions, a calomel electrode is essential because Tris interferes with the standard silver/silver chloride electrode in alkaline media. Also, pH must be set at the temperature of chromatography because some buffers, such as Tris, have strong temperature dependent pH profiles. Small changes in pH above 8.0 can have a substantial effect on the migration of some DNA adducts.
Note: Some researchers report that direct exposure to fluorescent light or stray sunlight can cause photodecomposition of benzo[a]pyrene-DNA adducts on PEI-cellulose sheets (R. Roggeband and M. J. Steenwinkel, pers. commun.). These findings indicate the need to use filtered light (400 nm cutoff) in the laboratory during chromatography.
The following protocol is for large hydrophobic DNA adducts.
Please see section on autoradiography for further information.
Isopropanol/4 N ammonia is an excellent alternative D4 solvent
system that will breakup
the strong diagonal radioactive zones that the LTU solvent system
gives and disperse the
individual adduct spots over a larger area of the chromatogram (see
Fig. 6).
NOTE: The isopropanol/4 N ammonia solvent system may cause some
cracking to occur in
Plexiglas boxes and equipment. Use glass tanks for this solvent
system.
The ratio of isopropanol/4 N ammonia (v/v) can vary from 0.8:1 to
2:1; 1.2:1 works well for large
hydrophobic adducts on laboratory prepared sheets and 1:1 works for
commercial sheets. When
using the isopropanol/4 N ammonia solvent system for D4, the sheet is
placed in
10 mM Tris base after the D3 water rinse step to soak for 3 to 5
minutes (this resets the pH of the
sheet). The sheet is then rinsed under the tap and dried. A 2 cm
filter paper wick (Whatman
No.1) is then stapled to the top edge. A longer wick stapled to the
top of the PEI-cellulose sheet
will increase the distance that the adducts migrate and this can be
used as an additional variable
along with the isopropanol/4 N ammonia ratio to control adduct
migration on the chromatogram.
Place the sheet in a TLC tank containing the isopropanol/4N ammonia,
seal it with plastic wrap,
and develop in the D4 direction. Do not move the tank during
development.
Development time is usually 2 to 3 hr. When the solvent has reached
the top of the wick, remove
the sheet from the tank and let the sheet with the attached wick air
dry in a fume hood. Place the
dry chromatogram with the attached wick in 1 M sodium phosphate (pH
6.0) for D5. The D5 is a
necessary step for the isopropanol/4 N ammonia solvent system,
otherwise a strong dark band
may appear in the middle of the autoradiogram. Trim sheet
(see Fig.
7) for autoradiography.
Dilute ammonia (0.08 to 0.4 M) is also an excellent solvent system for D4, and the dispersion of adducts on the chromatograms falls in between the LTU and isopropanol/4 N ammonia solvent systems. The dilute ammonia solvent systems also require the Tris base pretreatment and the attachment of a 2 cm wick as described for the isopropanol/4 N ammonia system. The development time for the dilute ammonia solvent system is 20 to 40 minutes. After D4, develop in the D5 direction in the same manner as was done for the isopropanol/4N ammonia system.
At present, there are several ways to quantitate radioactivity on chromatograms. The traditional "cut and count method" aligns the autoradiogram with the corresponding chromatogram and uses a marking pen to outline the radioactive regions on the chromatogram. Radioactive regions are then cut from the chromatogram, placed in a liquid scintillation vial containing liquid scintillation cocktail, and counted by liquid scintillation spectrometry (LSS). Cerenkov counting, which measures the light emitted when a charged particle passes through water, can also be used; however, a correction for counting efficiency is necessary. In addition, areas near radioactive spots are excised and counted to obtain background corrections. However, background corrections based on this approach can be somewhat subjective and one must exercise careful judgment when the background correction is substantial relative to the spot or region of interest. Another analytical approach is the use of computer-aided imaging systems that will locate and directly measure the radioactivity on the chromatogram.
Autoradiography is used to locate the position of 32P-derived radioactivity on the chromatograms; and it provides information on the relative intensity of each individual adduct. However, there are restrictions on the interpretation of autoradiograms. First, film has a limited range of response (less than 300) and the range of linear response is less than 100. Multiple exposures are usually required for a complete set of autoradiograms to show both strong and weak regions of radioactivity present on the chromatograms. Moreover, the autoradiographic images of faint spots may be only 20 to 30% of their expected densities based on the level of radioactivity present. Because of these limitations, caution is required when making assumptions on the relative intensity of faint spots and the levels of background radioactivity present based on a visual inspection of the autoradiogram. Autoradiography of the chromatograms either can be run at room temperature or film sensitivity can be enhanced by a factor of up to 18 with the use of intensifying screens at -80°C (Swanstrom and Shank 1978).
ProcedureHandle the chromatograms carefully to avoid any flaking or loose particles being generated that are radioactive and could leave confusing images or spots on the autoradiograms.
Photography of Autoradiograms
We use a 4 by 5 inch format Polaroid camera mounted on a stand
and Polaroid type 55
positive/negative film for making prints of individual
autoradiograms. Satisfactory prints can be
made using a 35 mm camera and Kodak TX200 film. However, the larger
format of the 4 by 5
inch Polaroid type 55 film yields a sharper print.
Quantitation of Radioactivity by the "Cut and Count"
Method
Place the autoradiogram on the chromatogram and align it with
the fluorescent ink marks.
If you do not have fluorescent ink marks to align the chromatogram,
then use the solvent front
edges on 2D autoradiograms to align the autoradiogram with
chromatogram. Carefully outline
the regions of interest including background areas on the
chromatogram with a marking pen.
A map is drawn of the radioactive regions that are to be excised and each region is numbered. This will give you a permanent record of areas quantitated and their respective positions on the autoradiograms. The regions are then excised using a sharp razor blade or a scalpel. Each of the excised chromatogram pieces is weighed (this allows a background correction to be made per unit weight of chromatogram) and placed in a 20 mL scintillation vial. Add 10 mL of scintillation cocktail to each vial and count.
Liquid Scintillation Spectrometry
Liquid scintillation counting has some restrictions. Samples
containing low levels of
radioactivity (i.e., faint spots and samples for estimation of
background corrections) must be
counted for longer time periods for adequate counting statistics.
Counting statistics are based on
total counts measured and not the counting rate (i.e., cpm or dpm).
The following numbers give a
comparison between total counts and % error (i.e., total counts / %
error): 200 / ± 14%, 500 /
± 9%, 1000 / ± 6%. Also, the liquid scintillation spectrometer has
an inherent background of 20 to
25 cpm. A small variation of 3 to 10 cpm in the estimation of
background per cm2 of surface
(easily achievable because of poor counting statistics for samples
with low radioactivity levels)
can greatly affect adduct level computations. For single,
well-resolved spots, one can use a clean,
adjacent area to define background. However, for diagonal
radioactive zones, which often have a
faint hazy area around them (the haze may be due to adducts), the
question of where to sample
for defining background corrections becomes highly subjective. The
subjective aspect of defining
background from a remote part of the chromatogram points to the need
to do the
chromatography and measurement of radioactivity carefully. Moreover,
proper handling and
storage of tissue and purified samples is important, because DNA
breakdown products will raise
the overall background.
The liquid scintillation counting efficiency for 32P in weakly quenched solutions is 98 to 99% which means cpm is essentially the same as dpm.
Correction for background is as follows:
corrected cpm (region of interest) = cpm (region of interest) - [wt
(mg) of region of
interest x (background cpm/wt of background region in mg)].
Cerenkov Counting
An alternative to counting samples by LSS is Cerenkov counting
which is cheaper because
it does not require the use of scintillation cocktail and eliminates
disposal problems associated
with some organic-based cocktails. However, the counting efficiency
for radioactivity can be
substantially lower. A correction factor for counting efficiency can
be made by counting aliquots
of 32P in both liquid scintillation cocktail and water.
Since the counting efficiency in
the cocktail is approximately 99%, the ratio of cpm-cocktail to
cpm-water will give a correction
factor for converting Cerenkov counts to dpm.
We are currently using storage phosphor imaging technology to locate and measure radioactivity associated with the DNA adducts present on the chromatograms. This technology offers high sensitivity (approximately 0.2 nmol adduct/mol nucleotides), extremely low counting background, a large linear range of response to radioactivity (105 to 1), and sufficient data points to allow accurate mapping and quantitation of radioactivity using computer image analysis methods. (For an in-depth discussion on storage phosphor imaging see Reichert et al. 1992). An additional benefit of this imaging system is that, unlike radioactivity measurements by liquid scintillation spectrometry, all raw data for the distribution and intensity of radioactivity on the chromatograms is available for review and reprocessing at any time in the future.
Scanning the Chromatograms
The imaging screens are first erased using an Image Eraser
(Molecular Dynamics,
Sunnyvale, CA). Storage phosphor imaging is performed at ambient
temperature by placing
storage phosphor screens (Molecular Dynamics, Sunnyvale, CA) over the
radioactive
chromatograms and exposing for a specified time period. A factor to
convert the screen signal to
dpm is generated by imaging a test strip containing a serial dilution
of a 32P solution
concurrently with the samples. The signal response from the screen
is then divided by the
32P dpm in the test spots to generate a factor for
converting the imaging screen
signal response directly to 32P dpm. The length of screen
exposure to a
chromatogram is dependent on the maximum levels of radioactivity
present and is usually about
one-tenth the time required for autoradiography. Generally, for a
10 µg DNA sample a 6 to 24 hr exposure is sufficient. The
chromatograms for DNA bases
and the specific activity analyses require only a 10-minute exposure.
Saturation of the screens
occurs as the pixel values approach 80,000 and is evident when the
radioactivity profiles of lines
drawn through radioactive areas on the chromatogram images have a
truncated appearance.
However, this is a problem only with extremely "hot" samples where
the radioactivity is
concentrated in a small area. After exposure of the chromatograms to
the screens, the latent
radioactivity image is read by scanning the screen with a
PhosphorImager (Molecular Dynamics,
Sunnyvale, CA).
Data Storage and Computer Requirements
Image data from the PhosphorImager is stored on a 620 MB optical
disk (Pioneer
DEC-702 rewritable optical disk, Upper Saddle River, NJ) using a
Pioneer rewritable optical disk
drive (Upper Saddle River, NJ). Data are then processed on a 486-33
MHz computer equipped
with a
330 MB hard drive, 32 bit local bus and 48 MB of RAM. A Diamond
Stealth VRAM video
accelerator card is used to increase the rate of screen repainting.
Because the file sizes generated
by the imaging process are up to 41 MB, it is desirable to place the
files in a virtual RAM drive
where processing is 3 to 6 times faster. This computer, as
configured, processes large data files
quickly and efficiently. The files can be processed on a 386-33 MHz;
however, there is a
substantial increase in data processing time. Software used to
process the data is ImageQuant
(version 3.15) from Molecular Dynamics.
Processing of Image Analysis Data
The 32P-postlabeling results are usually presented as a ratio of the number of adducts detected divided by the amount of DNA used in the assay (e.g., nmol DNA adducts/mol DNA, amol DNA adducts/µg DNA, fmol DNA adducts/µg DNA). Presenting the data in this fashion avoids the problem of not being able to extract DNA quantitatively from tissues.
It is necessary to keep track of all numbers used in the final calculations. It is also important that your laboratory protocols assure that the necessary information regarding aliquot sizes and dates when samples are counted is recorded. See Table 1 for sample calculations that include the raw data from the imaging system (in Microsoft Excel 4.0 format).
Sample calculations
Calculation for total DNA adducts measured:
fmol adduct(s) | = dpm for DNA adduct spot (or zone) x
labeling aliquot
correction factor x
(1/specific activity of (gamma-32P)ATP). |
dpm for DNA adduct spot (or zone) | =2130 |
aliquot correction factor | = 2 (15 µL spotted on chromatography sheet of a 30 µL 32P-labeled DNA sample) |
specific activity of (gamma-32P)ATP | = 2000 |
Ci/mmol | = 4400 dpm/fmol |
fmol adduct(s) | = 2130 x 2/4400 = 0.968 fmol |
Calculation for total DNA analyzed for adducts:
nmol DNA | = dpm for DNA base spot (dG) x factor to convert from nmol dG to nmol DNA x spotting aliquot factor x DNA enzyme hydrolysate dilution correction factor x (1/specific activity of DNA bases labeling (gamma-32P)ATP). |
Where: | |
dpm for DNA base spot (dG) | = 24600 dpm |
factor to convert from nmol dG to nmol DNA | = 100% / % of dG in DNA |
= 4.76 | |
spotting aliquot factor | = 2.4 (10 µL spotted on chromatography sheet of a 24 µL 32P-labeled DNA bases sample ) |
DNA enzyme hydrolysate dilution correction factor | = (500 µL dilution volume/10 µL taken for base labeling) x 2 (to change from 5 µL of enzyme hydrolysate used for measuring DNA content to 10 µL of hydrolysate used in the DNA adduct labeling part of the assay) = 100 |
specific activity of DNA bases labeling (gamma-32P)ATP | = 968055 dpm/nmol |
nmol DNA | = 24600 dpm x 4.76 x 2.4 x 100 x 1/(968055 dpm/nmol) |
nmol DNA | = 29 nmol DNA used in the DNA
adduct assay
|
Calculation of DNA damage level:
nmol DNA adducts/mol DNA | = amount of DNA adducts measured in a sample/amount of DNA used. |
= 0.968 fmol DNA adducts/29 nmol DNA | |
= 968 amol DNA adducts/29 nmol DNA x
(109/109)
| |
nmol DNA adducts/mol DNA | = 33 nmol DNA adducts/mol DNA |
The PPL method is an involved procedure that requires numerous determinations to be made that are dependent on enzyme efficiencies and radioactivity measurements. The method should be considered semiquantitative when working with tissue samples from animals exposed to complex mixtures, because the identities of the adducts formed from complex mixtures are unknown, and it is not possible to simultaneously optimize the conditions for all adducts present. However, if sets of samples are run under uniform conditions, then meaningful comparisons between samples collected from different sites can be made. To have confidence in the results, appropriate controls must be used to signal the presence of significant errors. The following are quality assurance procedures that should be included in the 32P-postlabeling assay: