Beam me up. A view of the accelerator mass spectrometer; beams run from left to right in the foreground.
For many environmental chemicals, scientists must extrapolate data from high-dose animal experiments to estimate human risk from low exposures. Because there is often great disparity between doses, intermediate markers of exposure that respond to lower doses must be used. One common marker is the DNA adduct: DNA covalently bound to an active chemical intermediate. Accelerator mass spectometry (AMS) is a new method that significantly increases the accuracy and range of measuring DNA adducts.
DNA binding has traditionally been measured by tagging the chemical of interest with carbon-14 or hydrogen-3 isotopes and, after extensive procedures to eliminate any unbound isotope, counting the isotope label still covalently bound to the DNA. More recently, monoclonal antibodies have been developed to recognize an intermediate of the chemical covalently bound to a DNA base and have been used in immunoassays to quantify adduct binding. Another sensitive method to analyze DNA adducts is breaking down DNA into its constituent nucleotides by enzymatic digestion, followed by labeling the nucleotides with phosphorous-32 and separating adducted from normal nucleotides by multidimensional thin-layer chromatography. Both monoclonal antibodies and enzymatic digestion are useful for assessing DNA damage, but they lack either sensitivity at low doses or the capability to give reliable quantitative information. Both methods do, however, have the advantage of allowing assessment of human exposures because they are noninvasive (people do not have to be exposed to radioactivity or analytical chemicals). In contrast, AMS allows for an ultrasensitive measurement of radiolabeled compounds and thus is ideally suited for measuring carcinogen-DNA adducts following an extremely low-dose exposure. AMS allows the investigator to analyze small samples, measure with extreme precision, and detect extremely small amounts of labeled compound.
How AMS Works
Accelerator mass spectrometry uses a tandem electrostatic (Van de Graaff) accelerator to dissociate and accelerate negative ions generated from graphite by a cesium-sputter ion source. The ions are sorted by a low-energy mass spectrometer, which selects the proper mass and injects the ions into the accelerator beam tube. The increasing positive potential in the beam tube accelerates the negative ions toward the 2-15 megavolt terminal at the center of the accelerator. Quadrapole lenses select and focus the correct ion and charge state for a second high-energy mass spectrometer, which separates the rare isotope from the ion beam of an abundant elemental isotope. A third magnet selects by momentum. The ion of interest is further selected based on its velocity, and finally the ions of interest are counted in a multianode gas-ionization detector. All these steps are necessary to obtain 1 part in 1015 sensitivity in detecting rare isotopes such as 14C.
For counting samples with 14C, approximately 1 milligram of carbon is optimal. This material is generated by combusting the sample to carbon dioxide and reducing it to graphite on a cobalt catalyst This is the rate-limiting part of the AMS measurement: one technician can batch process approximately 250 samples per week. Although AMS is a super-sensitive scintillation counter, it does have the short-comings of isotope counting, including contamination problems.
Measuring Carcinogens and DNA
Traditional animal bioassays use either the maximum tolerated dose (MTD) or doses close to this level. Yet most environmental exposures are orders of magnitude lower. Therefore, methods like DNA binding, cytogenetics, and mutation analysis are used to assess intermediate genetic and molecular damage leading to carcinogenesis. These methods allow exposures to be measured at lower levels than the traditional cancer bioassays, but in most cases do not approach actual levels in food, the workplace, or the environment. Measurement of DNA adducts is probably the most sensitive intermediate method of assessing events leading to cancer. Immunoassay analysis can detect 1 adduct per 108 nucleotides, and 32P-postlabeling can detect 1 adduct per 1010 nucleotides. In practice, quantitative measurements are one to two orders of magnitude less sensitive than these limits. In contrast, AMS can directly measure 14C nuclei in as few as 1000 atoms, two orders of magnitude below the natural abundance of 14C in living organisms. Therefore, AMS can accurately measure 1 adduct per 1011 bases, or 1 adduct per 100 mammalian cells.
Kenneth W. Turteltaub and colleagues at the Lawrence Livermore National Laboratory made the first measurements of DNA adducts using AMS. They discovered a linear relationship of binding with dose of the carcinogen 2-amino-3,8-dimethylimidazo[4,5-f]quinoxaline (MeIQx) from 1 adduct per 5 106 nucleotides to 1 per 1011 nucleotides. 32P-postlabeling data for MeIQx also fit on the same linear dose-response curve but in the higher dose region.
These linear AMS adduct measurements suggest that repair of MeIQx adducts is not concentration dependent at even 1 adduct per 100 cells; in fact, it appears that repair is no more efficient at these low doses than it is at 1 million times the damage. The same can be said for metabolism and any other physiological activity affecting the steady-state level of these adducts. Thus, at least for acute exposures, extrapolation of damage is linear over an extremely large range of doses. If this relationship holds for other genotoxic carcinogens, risk analysis for these types of compounds can be done with fair certainty that a linear relationship exists with respect to dose.
Scientists at Lawrence Livermore have developed methods that allow separation of the nucleotide DNA adducts by high-performance liquid chromatography and then counting of fractions by AMS. This procedure allows measurement of specific adduct levels at extremely low exposures. The retention times for each individual adduct, which are derived from high-dose experiments, are related to peaks seen from AMS counted fractions.
One drawback of AMS is that analysis of human samples such as blood, urine, or tissue would require the administration of both carcinogens and labeled material before analysis. Although less than 0.1 millirad of radiation and 100-500 nanogram/kilogram dose of chemical would be an insignificant exposure for a human, it is difficult to get approval for such studies. An attractive alternative for analyzing low-level human exposures is an AMS-modified radioimmunoassay. This method would allow detection of specific adducts by antibodies and then the use of competition radioimmunoassay coupled to AMS to quantify adducts.
Improvements in AMS
How AMS works. Diagram illustrates the components of an accelerator mass spectrometer and the flow of particles through the equipment.
There are three types of potential improvements in AMS. First, AMS technology has more sensitivity than the biological system allows. This problem can be overcome by using organisms depleted in 14C. Depletion can be accomplished by growing organisms on carbon sources derived from petroleum (all the 14C has decayed in feedstocks derived from these materials). Turteltaub showed that it is possible to gain two orders of sensitivity by using bacteria such as Methylosinus trichosporium, which are grown on petroleum-derived methane as their sole carbon source. The same methodology can be used in rodent diets. Well-balanced diets containing protein and carbohydrate sources depleted in 14C could easily gain one order of sensitivity for DNA binding, allowing detection of more than 1 adduct per 1012 nucleotides. Unlabeled nutrients and biochemicals would act as if they were labeled in rodents fed such a diet (carbon dating of those with 20% residual 14C would make them equivalent to a 13,000-year-old organism) because of their natural 14C levels. Labeled experiments could be done without isotope tracers in these rodents.
Second, further improvements should allow for detection using a second isotope. This would give all the advantages of double-isotope counting such as rates of formation and loss of nutrients, adducts, or biologically important macromolecules. At Lawrence Livermore, 3H has been counted with the same device as 14C using a different detector. This isotope is as sensitive as 14C (approximately 1 part per 1015), but sample preparation methodology needs to be developed futher. Currently, a method is being tested to measure Tritum (TiH2) for the ion source in the form of TiH2. In addition to allowing double-label experiments, tritium capability would allow a much cheaper source of labeled compounds. Although tritium introduces problems associated with isotope exchange, it should be valuable to the biomedical AMS capability.
Third, the instrument currently being used at Lawrence Livermore is designed for research on isotopes up to mass 131 (i.e., 131I). Although this instrument is operational in an automatic mode, where 60 samples can be run in batch mode in less than 2 hours (multiple batches per day), the capital cost and maintenance of a research AMS facility is expensive and is not needed for biomedical samples. An accelerator for 14C and 3H measurements that would cost under $1 million and fit in a typical research laboratory space is a goal of AMS development. Such equipment would allow more of the biomedical research community to conduct these types of high-sensitivity measurements.
Future Directions
The capability to make high-sensitivity measurements has applications throughout biomedicine. Receptor studies where receptor and ligand interactions are limited would benefit from application of AMS. Questions about interactions of chemicals with DNA at any dose could be examined. AMS could also be used to determine if chemicals bind DNA, even at the lowest levels detectable following high exposures.
Another important arena for AMS is the radioimmunoassay field. Turteltaub and colleague John Vogel have shown that a very sensitive, direct radioimmunoassay can be adapted for use with AMS with a sensitivity of less than 100 attomoles (10-18 moles) for the pesticide heptachlor. They also have measured the AIDS drug AZT in a competitive radioimmunoassay at the low femtomole level (10-15 moles). The advantage of this type of assay for humans is that adducts and drugs can be measured without having to dose with radiolabeled chemicals.
AMS is valuable in experiments where high sensitivity, low sample size, or low molecular interactions are problems. DNA adducts appear linear, at least in acute dosing experiments over a million-fold range of dose. The ability to count such low levels of material with precision and speed enables scientists to answer a large spectrum of questions.
James S. Felton
Kenneth W. Turteltaub
James S. Felton is group leader of the Molecular Toxicology Group and Kenneth W. Turteltaub is a senior staff scientist in the Biology and Biotechnology Program at Lawrence Livermore National Laboratory.
Suggested Reading
Felton JS, Turteltaub KW, Gledhill BL, Vogel JS, Buonarati MH, Davis JC. DNA dosimetry following carcinogen exposure using accelerator mass spectrometry and 32P-postlabeling. In: New horizons in biological dosimetry (Gledhill BL, Mauro F, eds). New York:Wiley-Liss, 1991;243-253.
Randerath K, Randerath E. Monitoring carcinogen actions on dna by 32P-postlabeling. Int Symp Prin Tak Cancer Res Fund 21:317-328(1990).
Turteltaub KW, Felton JS, Gledhill BL, Vogel JS, Southon JR, Caffee MW, Finkel RC, Nelson DE, Proctor ID, Davis JC. Accelerator mass spectrometry in biomedical dosimetry: relationship between low-level exposure and covalent binding of heterocyclic amine-carcinogens to DNA. Proc Natl Acad Sci USA 87:5288-5292 (1990).
Turteltaub KW, Frantz CE, Creek MR, Vogel JS, Shen N, Fultz E. DNA adducts in model systems and humans. J Cell Biochem Suppl 17F:138-148(1993).
Turteltaub KW, Vogel JS, Frantz CE, Fultz E. Studies on DNA adduction with heterocyclic amines by accelerator mass spectrometry: a new technique for tracing isotope-labelled DNA adduction. In: Postlabelling methods for detection of DNA adducts (Phillips DH, Castegnaro M, Bartsch H, eds). Lyon:International Agency for Research on Cancer, 1993;293-301.
Vogel JS, Turteltaub KW. Biomolecular tracing through accelerator mass spectrometry. Trends Anal Chem 11:142-149(1992). |
Last Update: August 7, 1998