Tracing the S-Curve: Molecular Applications of ICP-MS

"Mass Spectrometry Forum" columns focus on the basics of mass spectrometry (MS), and therefore often delve into the tools of the trade: MS instruments and hardware. The columns often begin with a restatement of elementary facts and assumptions so that we can see how those tools fit with analytical needs. We do so again here, by reminding ourselves that MS is one analytical tool, among many tools, that can provide information that we use within "an overall measurement perspective" to solve problems. An exclusive focus on hardware associated with a particular choice within MS — be it a choice of ionization method, of mass analyzer, of a combination known as a hyphenated method, or even of a particular data processing approach — is sometimes an overly artificial construct that focuses on the tool rather than on the creation and completion of an analytical approach that provides data that solve the problem. A properly constructed analytical approach begins with the careful definition of the problem, configures a sampling methodology, collects the samples, prepares the sample, acquires data, and culminates with the integration of the MS data with other data, acquired with other analytical tools. The overall protocol confronts the discrepancies and inconsistencies, and, assessing the data in perspective, finally reaches a reasoned conclusion. The acquisition of MS data is only a part of the measurement suite, and only part of an overall analytical approach that must first determine if MS analysis is even possible; and if so, whether the data will be pertinent. One joy of teaching introductory MS courses is bearing witness to the expansion of MS into new areas of application, and a growing recognition of the synthetic skills that underlie the creation of an insightful analytical approach. The ultimate value of an approach is based upon its unique utility — the rightness of the method for the job — and not on hardware or tools, no matter what commercial acronym describes the instrument, and no matter how much it costs. When MS data is the linchpin, as it often is, mass spectrometrists earn their rightful commendation as analysts. It requires only funds to acquire MS hardware; it takes an insightful analyst to prepare samples, design and validate a method, obtain results of both accuracy and precision, and place the results in proper perspective.

Kenneth L. Busch

The subject of this column is molecular applications of inductively coupled plasma (ICP) linked with MS, and how those applications have developed. If we were to pull our analytical chemistry textbook from the shelf, we might read that ICP-MS is used to provide qualitative and quantitative information about elemental (mostly metal and metalloid) species, with samples reduced to atomic ions within the plasma ionization source, the mass analyzer separating the ions by mass (with higher mass resolution used to avoid isobaric interferences), and optimized detectors providing highly accurate information even at very low limits of quantitation. We would read in the textbook that ICP-MS competes with other spectroscopic methods for elemental analysis but has become a method of choice because of unmatched specificity and sensitivity that routinely achieves parts-per-billion levels. Early reviews and books (1–3) in the field describe tools and instrumentation, and the relevant applications almost entirely within fields of elemental analysis (inorganic MS), including geochemical and environmental analyses. We would read, perhaps, that isotopic dilution methodology has been used extensively in ICP-MS analyses (4,5), and that careful sampling is an intrinsic part of a high-sensitivity analytical method. Therefore, as an example of insightful sampling as part of an overall analytical scheme, we could consider spatially resolved sampling and high-sensitivity ICP-MS applied to the analysis of tree-ring samples to assess environmental exposure to heavy metal toxicants (6). However, the concept that ICP-MS might have molecular applications is usually missing from the textbooks. If our imaginations stopped at the atomic level, satisfied with these extraordinary accomplishments now become ordinary, we would be shortchanging ourselves in our role as analysts.

Analytical chemists tend to be innovative. As ICP-MS has evolved into a mature analytical method, with instruments available from several commercial vendors, a long-term view might place ICP-MS on the plateau of the S-curve of innovation (Figure 1). Analytical chemists have long been familiar with this curve, which traces its conceptual origins to Tarde (7) in 1890, and is now popular among business strategists who lecture about innovation. A common metric in assessing the breadth of a given field of chemistry is publication statistics. Accordingly, Figure 2 plots publications in ICP-MS for each publication year for about the past 25 years (8). The similarity of the trend to the general shape of the S-curve is apparent. According to the authors of the study, the growth rate peaked in about 1997, representing the inflection point in the curve. Additionally, the authors note that "speciation" by ICP-MS represents about one third of the publications tabulated in recent years, and is a rapidly growing application area in ICP-MS. Referring back to Figure 1, speciation could be represented by the second curve labeled "emerging technology." In the context of this column, speciation represents an early, but now established application of ICP-MS in the area of molecular analysis.

Figure 1

In ICP-MS (9), speciation refers to identification of molecular entities that contain an element of interest, using a separation method (often liquid chromatography [LC] or capillary electrophoresis [CE]) to separate various molecular species containing the metallic element before their elution into the ionization source. ICP-MS provides an "atomic" detector that can be tuned to the presence of specific elements or can provide a total elemental fingerprint (vide infra). The retention time on the chromatographic column, by comparison with the retention times of standards, provides a confirmation of the molecular identity. As an example, tin (Sn) appears in the environment in many different molecular forms, not all of which are of equal toxicity or of equal regulatory concern (9). The ICP-MS provides a signal based in the intensity of the Sn⁺ ion (and its various isotopes) with high sensitivity, and the data exhibit superb accuracy and precision. The chromatographic separation, LC in this case, must provide sufficient resolution to separate the molecular species clearly in terms of elution time, and standards must be available for direct comparison of elution times. In this example of the tin speciation, four molecular entities, each containing tin, were of interest. With sufficient sample preparation and cleanup (and assuming that the recoveries of each molecular species in that process are known), chromatographic separation can be completed in short runs of just a few minutes. For tin speciation, the LC run was 12 min in length and could probably be shortened still. It might be argued that the LC–ICP-MS combination is not inherently as powerful in establishing a molecular identity as is GC–MS or LC–MS, in which the distribution (mass and relative abundance) of ionic species in the mass spectrum as well as the retention time match (should standards be available) confirms the molecular identity. However, the high sensitivity of the ICP-MS coupled with the enhanced resolution of modern LC or CE (and the ability to use columns with different retention characteristics) combine to provide a powerful analytical tool, and ICP-MS becomes a means of molecular analysis. As noted, the relative simplicity of the detected signal as elemental ions, and the concordant application of isotope dilution methods, means that the accuracy of the quantitation can be very high. Relative standard deviations of just a few percent are reported routinely in the literature. The October 2007 supplement to Spectroscopy titled Applications of ICP and ICP-MS Techniques for Today's Spectroscopist highlights some current capabilities of these speciation applications. To reiterate, the unique capabilities of the analytical method determine its value.

Figure 2

Figure 3 represents the S-curve growth curve specifically in "speciation papers" abstracted as such in "Analytical Abstracts" (10); this curve represents results obtained with all analytical tools, not just ICP-MS. Speciation using ICP-MS is probably still in the exponential growth part of its own particular S-curve. Whether or not the inflection point has been reached depends more so upon the pressures placed on various governments and regulatory agencies rather than on the established capabilities of the analysts and their instruments. Scholars of the S-curve often present it as an amalgamation of many disparate individual curves. We could in fact deconvolute a curve such as Figure 3 into element-by-element curves, or method-by-method curves, or curves that purport to show any particular dependent variable. Historians and regulators have access to modern bibliometric tools that would allow them to do so. Such researchers might also comment on the timeline of publications, and the influence of topical conferences, funding initiatives, and (more recently) resources such as web sites for both commercial instrument manufacturers and academic researchers. However, returning again to the basics, we emphasize that innovation often takes leaps into new dimensions, and coplanar S-curves as a model for innovation might be insufficient. Instead, consider an aphorism that underlies a great deal of modern chemical metrology: "New instrumentation begets new chemistry." A simple two-dimensional representation such as Figure 1 cannot fully represent the broader, shifting multidimensional world, driven by utility and innovation, and by funding and regulation. Any emerging demand for answers, any set of newly identified problems, any new funding source for research, and any substantial leap in analytical capabilities will draw the performance S-curve into a new dimension. Here then we have a possible explanation for the fact that ICP-MS analysis supplanted decades-old elemental analysis protocols in a few short years. We then might explain how ICP-MS elemental analysis protocols now address brand new problems in expanded environmental and geochemical areas of research. We might then understand how ICP-MS has not simply expanded into the area of "speciation," but now finds applications in metallomics, metabolomics, proteomics, and other burgeoning "molecular" areas of MS application. The predictive answer resides in the fundamental axiom that the value of an analysis is based upon its unique utility. Application of ICP-MS in these recent research venues might not have been foreseen based upon the descriptions provided in the analytical chemistry textbooks, but performance needs, regulatory needs, and the demands for answers have evolved into new innovative dimensions.

Figure 3

Expansion of ICP-MS into speciation and more generally "-omics" analysis (and generation of a new S-curve of innovation) does not change the basic fact that the ICP still atomizes the sample introduced into the plasma and still ionizes the elements as atomic ions, that a mass analyzer still separates these ions by mass, that the detection system still provides high-accuracy data for the distribution of isotopes, and that the overall sensitivity of the instrument extends routinely into the parts-per-billion range for many elements. Instead, developments in chromatography, specifically LC and CE, and in the interfaces to the ICP, and in the preparation of samples, were instead all "instrumental" in expansion of ICP-MS into areas of molecular analysis. Redesigns of sample introduction systems (nebulizers and skimmer cones) and the introduction of collision cells and regions, were modifications incorporated into a new generation of instruments (11). The new instruments provided new analytical abilities and engendered new "chemistries." As a simple example of what instrument development makes possible, consider that the use of array detectors, and data systems with newer abilities to collect, visualize, and interpret all data for all elements in a sample, has led to intriguing new applications grouped under the term of "elemental fingerprinting." For instance, the elemental profiles of agricultural crop products vary with the soil in which they were grown, and elemental fingerprinting therefore becomes a tool to establish the origin of a food product. Counterfeit drugs can be differentiated from real drugs based upon the elemental profile. Forensic applications based upon elemental fingerprinting are growing rapidly, centered in the middle of their own steep performance S-curve.

Molecular applications in metallomics, metabolomics, and proteomics might still seem distant, given that elemental ions constitute the detected species. However, remember that innovation occurs in the unexpected new dimensions. We begin with a description of ICP-MS in metallomics, defined as the distribution and speciation of a specific element within an organism. The "distribution" aspect of the research extends the simple delineation of molecular species containing the element of interest in a particular sample, and extends the overall study to an organ, or an organism, and sometimes to an evolutionary aspect of time as well. A similar conceptual framework can apply to environmental research, in which an environmental sink might be considered as part of a larger environment, or even a geographical area. How is distribution studied? Consider a recent metallomic study that takes advantage of the high sensitivity of ICP-MS (12). Laser ablation is used to sample proteins in gel electrophoresis bands, a small fraction of which is desorbed and analyzed by ICP-MS to provide elemental information. Once the location of metal-containing bands of interest is established, a matrix-assisted laser desorption ionization (MALDI) or electrospray ionization (ESI) mass spectrum is acquired to identify the protein in that band. The ICP-MS analysis requires so small a fraction of the sample that sufficient protein remains for the second-step MS analysis. The distribution of the element of interest in the various separated proteins on the gel is established in this conjoined analysis.

Metabolomics is, of course, more complex, studying the metabolic pathways of metals in a biological system, and how the molecular species and their distributions might change with time (or with subcultures of organisms, such as gut flora and fauna, subsumed within the host organism). In some recent studies, the very-high resolution capabilities of CE are coupled with an interface design that includes a sheath flow supplemented with stable isotopes of selected isotopes, which provided a calibration point for the analysis of metallothionein complexes from liver and brain samples (13,14). As analyses move into these more complex fields, sample preparation and derivatization strategies that take advantage of the high elemental specificity of ICP-MS provide the means to solve the analytical problem at the molecular level. The derivatization can be relatively straightforward, as in the use of a copper- or iron-containing reagent that reacts with amines or carboxylic acids, respectively, in a mixture, allowing speciation with ICP-MS to be used (15). But the derivatization can be more innovative as well. Molecular applications of ICP-MS usually involve heteroatom and heteroisotope tagging, a form of derivatization in which the heteroatom itself is the label, formed as the elemental ion in the ICP and monitored by the mass spectrometer. As the terms suggest, in heteroatom tagging, the label introduced contains a distinctive element not usually expected to be found in the sample before derivatization (following the concept as in the copper and iron derivatizations described above). In heteroisotope tagging, the label introduced contains an enriched isotopic abundance of an endemic element, and will thus alter the signature isotopic pattern observed in the mass spectrum. Szpunar has provided a comprehensive recent review (16) and defines heterotagging-tagged proteomics as "the use of a heteroelement, naturally present in a protein or introduced in a tag added by means of derivatization, for the spotting and quantification of proteins." The ICP-MS detector in such a project is coupled not only to LC or CE, but also to gel electrophoresis (as described earlier), and often used in conjunction with MALDI and ESI-MS. The ICP-MS results lead the analysis, serving as the targeting sensor that focuses other analytical methodologies that provide molecular spectroscopic information. But more than that, the ICP-MS data, through the use of isotope dilution techniques, provide the quantitative data with levels of accuracy and precision more difficult to attain with molecular MS methods. The new instrumentation has begotten new chemistry. In work with gel electrophoresis, the importance of maintaining all of the measured MS data in a correlated (x,y) set of spatial coordinates is evident. These abilities were already in place from past developments in planar chromatography coupled with MS, and they represent the translation of expertise and innovation into a new dimension.

Sulfur and phosphorus are common targets for metabolomic molecular analyses with ICP-MS. Heteroisotope tagging with ³⁴S-enriched methionine (Met) provided a handle for green fluorescent protein and JNK stimulatory phosphatase-1 genes. The mixture was then treated and its components separated with LC and then detected in a triple system consisting of a photodiode-array detector, a fluorescence detector, and an ICP-MS system (17). But there is much new chemistry that can be made part of an analytical protocol using ICP-MS, and new performance S-curves to ride upon. The absolute configuration of selenomethionine in Antarctic krill is established using a chiral derivatization reagent, LC, and ICP-MS (18). Transformation reactions can be used to create new information metrics for high-accuracy quantification (19). The S-curve is steep; reviews help to document the extraordinary pace of innovation (20). And in a few years, ICP-MS will be in the midst of new molecular applications. Improved sample collection and preparation will lead this new charge, along with improvements in detection sensitivity, and new chemistry in new applications areas will continue to provide answers to new questions.

Correction: In the last column, one of a series on quantitative MS, I misstated the area under the normal distribution curve that lies within one standard deviation on either side of the mean. The correct percentage of the population of measurements that lies within this interval is of course about 68%, not 95% (which is two standard deviations out on either side). We return to the topic of quantitative MS in the next column, after I go sit in the corner for a standard deviation unit's worth of time out.

Kenneth L. Busch recalls (not personally) the first uses of MS (Aston and Dempster) to map the abundances of the elemental isotopes. It took 30 more years of instrument development for molecular applications of MS to come to the fore. Development of ICP-MS, first with atomic applications and now with a growing suite of molecular applications, may be following a similar chronology. It appears that KLB simply gets older and more spherical, while MS technology "matures" and follows an S-curve. This column represents the views of the author and not those of the National Science Foundation. KLB can be reached at wyvernassoc@yahoo.com

References

(1) A. Montaser, Ed., Inductively Coupled Plasma Mass Spectrometry (Wiley-VCH, New York, 1998).

(2) H.E. Taylor, Inductively Coupled Plasma-Mass Spectrometry: Practices and Techniques (Academic Press, New York, 2000).

(3) J.R. de Later, Mass Spectrom. Rev. 17(2), 97 (1998).

(4) D. Schaumlöffel and R. Lobinski, Int. J. Mass Spectrom. 242, 217–223 (2005).

(5) S.P. Rodriguez-Gonzales, J.M. Marchante-Gayon, J.I. Garcia Alonso, and A. Sanz-Medel, Spectrochim. Acta B 60, 151 (2005).

(6) For an early example using ICP (but not ICP-MS), see C.F. Baes III and S.B. McLaughlin, Science 224, 494–497 (1984).

(7) J. Kimmunen, Acta Sociologica 39(4), 431–442 (1996).

(8) http://www.speciation.net/Public/Document/2007/06/20/2907.html.

(9) B. Fairman and R. Wahlen, Spectroscopy Europe 13(5), 16–22 (2001).

(10) B. Fairman, "VAM and Speciation Analysis," VAM Bulletin, 23, 18 (2000).

(11) R. Thomas, Spectroscopy 17(2), 42–48 (2002).

(12) G. Ballihaut, C. Pecheyran, S. Mounicou, H. Preud'homme, R. Grimsrud, and R. Lobinski, Trends Anal. Chem. 26(3), 183–190 (2007).

(13) K. Polec-Pawlak, D. Schaumlöffel, J. Szpunar, A. Prange, and R. Lobinski, J. Anal. At. Spectrom. 17, 908–912 (2002).

(14) D. Schaumlöffel, A. Prange, G, Marx, K. G. Heumann, and P. Bratter, Anal. Bioanal. Chem. 372, 155–163 (2002).

(15) P.S. Marshall, B. Leavens, O. Heudi, and C. Ramirez-Molina, J. Chromatography A 1056(1–2), 3–12 (2004).

(16) J. Szpunar, Analyst 130(4), 442–465 (2005).

(17) Y. Ogra, T. Kitaguchi, N. Suzuki, and K.T. Suzuki, Anal. Bioanal. Chem., online as DOI 10.1007/s00216-007-1546-y (2007).

(18) J. Bergmann, S. Lassen, and A. Prange, Anal. Bioanal. Chem. 378, 1624–1629 (2004).

(19) M.E. Del Castillo Busto, M. Montes-Bayon, A. Sanz-Medel, Anal. Chem. 82, 8218–8226 (2006).

(20) A. Sanz-Medel, M. Montes-Bayon, M. del Roasario Fernandex de la Campa, J.R. Encinar, and J. Bettmer, Anal. Bioanal. Chem., online as DOI 10.1007/s00216-007-1615-2 (2007).