Recent advances in biomedical applications of accelerator mass spectrometry

The development of biomedical AMS

The use of radioisotopes has a long history in biomedical science. Isotopic enrichment of xenobiotics with ¹⁴C is routinely used as a method of following their metabolic fate in both animals and humans, and a drug is typically synthesized such that the natural abundance of ¹⁴C is increased from the background level of 1.2 × 10^-10% to 20% or even higher depending upon the compound. The low energy β-radioactivity is then used to track the radiolabeled compound and its metabolites in biological samples derived from laboratory animal or human studies. LSC has been generally used for a long time to detect, follow and quantitate levels of radiotracer in such studies. There are occasions, however, when the low sensitivity of LSC becomes experimentally limiting, while the technique of AMS has now changed the experimental paradigm because its extremely sensitive detection limit virtually removed the previous experimental barriers.
The high sensitivity of AMS indeed affects experimental designs in several ways. First, the radioisotopic dose can be reduced to inconsequential levels of radiolysis, hazardous waste streams, and human subject exposure. Secondly, the chemical dose to a biological system, including humans of all ages and health status, is minimized to sub-physiological and sub-toxic doses. This allows a realistic analysis of the effects arising from low chemical doses. For example, children and women of child-bearing ages, who are important targets of increased health-related research, are suitable subjects at the low doses afforded by AMS [10,11], since the administration of such low levels of ¹⁴C are considered non-radioactive from a regulatory point of view. Finally, even if the sampled material needs fractionation to specific biomolecules prior to quantitation, the sample sizes are reduced to amounts that can be obtained from well-defined, and often non-invasive procedures.
For a practical AMS measurement, biological samples containing 0.2–5 mg of carbon must be converted to solid carbon (graphite or fullerene) using a two-step process [12]. In a quartz tube, and using excess copper oxide (CuO), the sample's biological carbon is oxidized to CO₂. The CO₂ is then reduced to solid carbon by both reduction with titanium hydride and zinc powder and catalyzation with either iron or cobalt. Because this process is independent of the chemical nature of the sample, it eliminates interference or suppression from other sample components. Therefore, AMS provides one piece of information about the sample of carbon measured: the precise ¹²C:¹⁴C ratio. In AMS, one measures the isotope ratio with respect to that of a well-known (external) standard in order to produce an absolute isotope concentration for the combusted sample [13,14]. With AMS, experimenters only need the fractional elemental abundance of the sample and the specific activity of the tracer compound in order to obtain, in the units most useful for interpretation, the concentration of the tracer in the sampled material. The mechanics of an AMS instrument, the mathematical conversions of the measured values to meaningful "Modern" values, and the comparisons with LSC are well reviewed in the literature [3,11,15-17].
For the first time in 1990, sensitive and precise quantitation of ¹⁴C was applied to the analysis of biological samples containing enriched ¹⁴C-labeled carcinogens for toxicology and cancer studies by Turteltaub et al. [18]. Their research quantified chemical binding of the ¹⁴C-labeled carcinogens to DNA at the level of 1 binding in 10¹¹ bases. The benefits of using AMS for the analysis of samples derived from radiotracer studies with humans soon became apparent, since AMS produces very specific quantitation with simple analysis [19]. Any isotope concentration greater than the known stable natural ¹⁴C background must arise from an introduced isotope label ("introduced" includes contamination, which must be carefully controlled and avoided). In the simplest experimental design, there is only one external radioactive source, perhaps a radiolabeled compound introduced into the biological system at a specific time. The isotope ratio of the isolated sample is then easily converted to the concentration of the labeled compound and its metabolites per g or ml of the analyte.
Not surprisingly, AMS has soon become a tool of choice for pharmacokinetic analyses [10,11,16]. All the metabolites of the compound that contain the labeled moiety can be directly quantified in chromatographic separations without resorting either to secondary standards or to prior knowledge of metabolic pathways. Although some fluorescent methods quantitate into the amol levels [20,21], they require derivatization procedures that are not suitable for in vivo tracing, create tracers that are not chemically equivalent, and are less general in applicability across many biological systems. Conversely, AMS is specific only to the labeled compound in any chemical or biological medium. Such specificity requires neither prior speciation nor the introduction of either molecular modifications or internal standards. With AMS, it is possible to conduct radiotracer studies in human with the administration of such low levels of ¹⁴C [10,11].
The most recent innovation using AMS technology is the so-called "microdosing" concept [10]. Choosing a drug for clinical trials from numerous candidates is very much a hit-and-miss business. Data are gathered from in vitro, in vivo, and in silico models in order to predict the drug's behavior in humans but such methods are probably only about 60% predictive. Presented with a choice of good candidates, it would be better to take them all into human subjects. This would, however, be prohibitively expensive, as each compound would require a significant package of toxicological safety testing. Alternatively, each candidate drug could be given to human volunteers at very low levels of a few tens, or at most a hundred μg. At these levels, only a limited toxicology package is required and in vivo human data can be acquired for candidate selection [22]. Only AMS has the required sensitivity to conduct such studies at the low μg level.
In this review, the recent development of AMS methods to the present day in biomedical/bioanalytical research where it is being strategically used with high precision (see Figure 2 for the major applications of AMS discussed here) will be followed.