Anal. Chem. 1997, 69, 926-929
Stable Nitrogen Isotope Analysis of Amino Acid Enantiomers by Gas Chromatography/Combustion/ Isotope Ratio Mass Spectrometry Stephen A. Macko* and Maria E. Uhle
Department of Environmental Sciences, The University of Virginia, Charlottesville, Virginia 22903 Michael H. Engel and Vladimir Andrusevich
School of Geology & Geophysics, The University of Oklahoma, Norman, Oklahoma 73019
The analysis of the stable nitrogen isotope compositions of individual amino acid stereoisomers through the use of gas chromatography/combustion/isotope ratio mass spectrometry (GC/C/IRMS) is presented. Nitrogen isotopic compositions of single amino acids or of their enantiomers is possible without the labor-intensive and time-consuming preparative-scale chromatographic procedures required for conventional stable isotope analysis. Following hydrolysis and derivatization, single-component isotope analysis is accomplished on nanomole quantities of each of the stereoisomers of an amino acid, utilizing the effluent stream of gas chromatographic separation. Nitrogen isotope fractionation is minimal during acylation of the amino acid, with no additional nitrogen being added stoichiometrically to the derivative. Thus, the isotopic composition of the nitrogen in the derivative is that of the original compound. Replicate stable nitrogen isotope analyses of 11 amino acids, and their trifluoroacetyl (TFA)/isopropyl (IP) ester derivatives, determined by both conventional isotope ratio mass spectrometry (IRMS) and GC/C/IRMS, indicate that the GC procedure is highly reproducible (standard deviations typically 0.3-0.4‰) and that isotopic differences between the amino acid and its TFA/IP derivative are, in general, less than 0.5‰. Stable nitrogen isotope analysis of bulk organic materials is a well-established method for tracing biosynthesis1 as well as the sources and history of organic matter in the geosphere.2 For example, nitrogen isotopes have been used to assess trends in early diagenesis,3-5 to elucidate conditions on the early Earth,6 and to assess the origins of organic nitrogen in extraterrestrial materials7 as well as to establish trophic orders in modern and fossil food chains.8,9 Often complicating the interpretation of bulk (1) Fogel, M. L.; Cifuentes, L. A. In Organic Geochemistry, Principles and Applications; Engel, M. H., Macko, S. A., Eds.; Plenum: New York, 1993; p 73. (2) Macko, S. A.; Engel, M. H.; Parker, P. L. In Organic Geochemistry, Principles and Applications; Engel, M. H., Macko, S. A., Eds.; Plenum: New York, 1993; p 211. (3) Wada, E. In Isotope Marine Chemistry; Goldberg, E. D., Horibe, Y., Sarahashi, J. J., Eds.; Uchida Rokakuho Publ.: Tokyo, 1980; p 375. (4) Altabet, M. A. Deep-Sea Res. 1988, 35, 535-554. (5) Qian, Y.; Engel, M. H.; Macko, S. A. Isotope Geosci. 1992, 15, 201-210. (6) Schidlowski, M.; Hayes, J. M.; Kaplan, I. R. In Earth’s Earliest Biosphere: Its Origin and Evolution; Schopf, J. W., Ed.; Princeton Univ. Press: Princeton, NJ, 1983; p 149.
926 Analytical Chemistry, Vol. 69, No. 5, March 1, 1997
stable nitrogen isotope measurements are the many modes of fractionation of nitrogen.1 Furthermore, important information at the molecular level is obscured in bulk analyses, a consequence of differences in the biosynthetic pathways of individual components10,11 and diagenesis.12,13 Attempts have been made to isolate individual components of complex mixtures for off-line combustion and stable nitrogen isotope analysis by conventional IRMS.14,15 However, the preparative chromatographic steps required to isolate sufficient milligram quantities of a component for conventional 15N analysis are labor intensive and often are complicated by column bleed and isotopic fractionation resulting from incomplete recovery of the entire component. Thus, numerous samples of interest remain unanalyzed owing to the complexities of this type of approach, and available data sets of δ15N values of individual compounds are quite limited. We previously reported a GC/C/IRMS method for the on-line stable carbon isotope analysis of complex mixtures of individual amino acid enantiomers at nanomole levels.16 More recently, the feasibility of determining δ15N values of amino acids by GC/C/ IRMS has been demonstrated.17-20 However, these initial GC/ C/IRMS methods are not amenable for determining the δ15N values of individual D- and L-enantiomers of amino acids in complex mixtures at natural abundance levels with a sufficiently high degree of precision. The ability to make such measurements is important for documenting the origin(s) and diagenesis of amino (7) Epstein, S.; Krishnamurthy, R. V.; Cronin, J. R.; Pizzarello, S.; Yuen, G. U. Nature 1987, 326, 477-479. (8) Macko, S. A.; Engel, M. H. Phil. Trans. R. Soc. London B 1991, 333, 367374. (9) Ostrom, P. H.; Macko, S. A.; Engel, M. H.; Russell, D. A. Geology 1993, 21, 491-494. (10) Macko, S. A.; Estep, M. L. F.; Hare, P. E.; Hoering, T. C. Isotope Geosci. 1987, 65, 79-92. (11) Hare, P. E.; Fogel, M. L.; Stafford, T. W., Jr.; Mitchell, A. D.; Hoering, T. C. J. Archaeol. Sci. 1991, 18, 277-292. (12) Silfer, J. A.; Engel, M. H.; Macko, S. A. Isotope Geosci. 1992, 15, 211-221. (13) Macko, S. A.; Engel, M. H.; Qian, Y. Chem. Geol. 1994, 114, 365-379. (14) Bidigare, R. R.; Kennicutt, M. C.; Keeney-Kennicutt, W. L.; Macko, S. A. Anal. Chem. 1991, 63, 130-133. (15) Engel, M. H.; Goodfriend, G. A.; Qian, Y.; Macko, S. A. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 10475-10478. (16) Silfer, J. A.; Engel, M. H.; Macko, S. A.; Jumeau, E. J. Anal. Chem. 1991, 63, 370-374. (17) Merritt, D. A.; Hayes, J. M. J. Am. Soc. Mass Spectrom. 1994, 5, 387-397. (18) Brand, W. A.; Tegtmeyer, A. R.; Hilkert, A. Org. Geochem. 1994, 21, 585594. (19) Preston, T.; Slater, C. Proc. Nutrit. Soc. 1994, 53, 363-372. (20) Metges, C. C.; Petzke, K. J.; Hennig, U. J. Mass Spectrom. 1996, 31, 367376. S0003-2700(96)00956-0 CCC: $14.00
© 1997 American Chemical Society
acid stereoisomers in fossil systems (e.g., refs 8 and 15), as well as for establishing the origin of chirality in our solar system prior to life’s origin (e.g., ref 21) and, perhaps, for confirming the existence of life on other planetary bodies such as Mars.22 Here we report a new GC/C/IRMS method for the stable nitrogen isotope analysis of individual amino acid enantiomeric components of complex mixtures.
sample. Details of the EA method have been previously reported.23 Differences between the two methods for the assessment of the 15N were minimal and within the precisions of the measurements reported previously (∼0.1‰, one standard deviation). Additionally, the results of the 13C EA analyses were in excellent agreement with previously reported values for the same amino acid standards.16 Stable nitrogen isotope values are reported using the standard convention
EXPERIMENTAL SECTION Standards and Reagents. Standard solutions of individual amino acid stereoisomers and of racemic mixtures of amino acids were prepared following the methodology of Silfer et al.16 Briefly, crystalline amino acids (Sigma Chemical, St. Louis, MO) were dissolved in distilled 0.1 N HCl (final concentration of 0.05 M) and stored at 4 °C until further use. 2-Propanol (commonly isopropanol, IP; HPLC grade, Fisher Scientific, Fairlawn, NJ) was acidified to 2.8 M by the addition of 250 µL (per milliliter of IP) of acetyl chloride (99+%, Aldrich, Milwaukee, WI). Trifluoroacetic anhydride (TFAA, 99+%, Pierce Chemical Co., Rockford, IL) was used without further purification. All reagents were of the same lot numbers for the derivatization procedures described below. Derivatization Procedures. Two-hundred microliter aliquots of a 0.05 M solution of each amino acid (50 µmol) were placed in an ashed screw-cap vial and evaporated to dryness at 40 °C under a stream of N2. Approximately 0.5 mL of acidified IP was added to each sample vial. The vials were sealed with Teflon-lined caps and heated at 100 °C for 1 h. The reaction was quenched by placing the vials in a freezer at -5 °C. The residual IP was removed under a gentle stream of N2. To remove excess water and IP, 0.25 mL of distilled CH2Cl2 was added to each vial in two successive aliquots, each of which was evaporated under a gentle stream of N2 at room temperature. The resultant amino acid esters were acylated by the addition of 0.5 mL of TFAA and 0.5 mL of distilled CH2Cl2. The vials were sealed and heated at 100 °C for 10 min. The vials were then placed in an ice bath, where the excess TFAA and CH2Cl2 was removed under a gentle stream of N2. To remove residual TFAA and CH2Cl2, 0.25 mL of distilled CH2Cl2 was added to each vial, and this was again removed at 0 °C under a gentle nitrogen stream. Conventional IRMS. The individual amino acid enantiomers and aliquots of their respective TFA/IP derivatives were converted to gases of suitable purity for isotope analysis using one of two procedures. In some instances, following the procedure of Macko et al.,10 crystalline amino acids or the derivatives were added to purified granular copper oxide (850 °C; Mallinckrodt Baker, Phillipsburg, NJ) and copper wire (Alpha Resources, Racine, WI) inside an ashed (550 °C, 1 h) quartz tube. Following evacuation, sealing, and combustion at 850 °C, the resulting gases were cryogenically purified, and the isolated N2 was isotopically analyzed on a PRISM (Micromass Instruments, Manchester, UK) stable isotope ratio mass spectrometer. Alternatively, certain of the bulk isotope measurements were made on a Micromass Optima instrument which had been interfaced to a Carlo Erba elemental analyzer (EA). In this case, the resulting gases were introduced directly into the source of the mass spectrometer, following chromatographic separation and chemical trapping of water, with both the 13C and the 15N being determined on a single
where R is the 15N/14N of the standard or sample. The standard for 15N analysis is atmospheric N2 and is defined to be 0.0‰. For routine analysis, the samples are analyzed against a laboratory reference tank of high-purity molecular nitrogen (99.999%, Linde Specialty Gases, Union Carbide, Danbury, CT) which has been calibrated against atmospheric nitrogen as well as other international standards. GC/C/IRMS. The individual stereoisomers, as well as several racemic mixtures of amino acids, were analyzed directly for their stable nitrogen isotope compositions using a Micromass Instruments Optima GC/C/IRMS system. This system is a modification of that which has been previously reported for the analysis of molecular carbon isotope abundances.16 The GC/C/IRMS system for 15N analysis of individual molecular components consists of a Hewlett Packard 5890 gas chromatograph interfaced to an Optima stable isotope ratio mass spectrometer, through a combustion system consisting of an oxidation furnace composed of braided nichrome/copper oxide wire and a reduction furnace of granulated copper, followed by a H2O/CO2 trap at liquid N2 temperature (Figure 1). Other details of the system hardware and software are similar to those reported previously by Freedman et al.24 The separation of the amino acid stereoisomers was accomplished using a 50 m × 0.25 mm i.d. fused silica capillary column bonded with an optically active stationary phase (Chirasil-Val, Alltech Associates, Deerfield, IL) which is capable of baseline resolution of TFA/IP esters of amino acid enantiomers. The GC conditions are as follows: splitless injection (∼1-2 nmol of each enantiomer derivative was injected, combusted, and subsequently introduced directly into the source of the MS); the carrier gas was ultrapure He (99.9999%) at a head pressure of 80 kPa; the injector temperature was 200 °C, and the temperature of the interface between the GC and the oxidation furnace was 350 °C; the GC temperature program was 45 °C for 3 min, 45-90 °C at 45°/min, 90 °C isothermal for 15 min, 90-190 °C at 3 °C/min, and then 190 °C isothermal for 30 min. The solvent (ethyl acetate) peaks were removed from the effluent of the GC through a heart split valve which is open to the FID at the time of injection. The valve is programmed to close at 1500 s to allow the column effluent to be directed to the oxidation furnace. Calibration of the stable nitrogen isotope composition of each component is accomplished by comparison to three reference gas pulses (each of 30 s duration) introduced at the start of the run and following the opening of the heart split valve at the end of each run, i.e., after 4500 s. Unlike conventional IRMS, which requires the combustion of ∼1 mg of an amino acid for isotope analysis, the N2 effluent
(21) Engel, M. H.; Macko, S. A.; Silfer, J. A. Nature 1990, 348, 47-49. (22) McKay, D. S.; Gibson, E. K., Jr.; Thomas-Keprta, K. L.; Vali, H.; Romanek, C. S.; Clemett, S. J.; Chillier, X. D. F.; Maechling, C. R.; Zare, R. N. Science 1996, 273, 924-930.
(23) Fry, B.; Brand, W.; Mersch, F. J.; Tholke, K.; Garritt, R. Anal. Chem. 1992, 64, 288-291. (24) Freedman, P. A.; Gillyon, E. C. P.; Jumeau, E. J. Am. Lab. (Fairfield, Conn.) 1988 (June), 114-119.
δ15N(‰) ) [Rsample/Rstandard - 1] × 103
Analytical Chemistry, Vol. 69, No. 5, March 1, 1997
927
Figure 1. Schematic of the GC/C/IRMS system.
Figure 2. Chromatogram of the TFA/IP esters of the racemic mixtures of amino acids. The trace at the top is the instantaneous ratio of mass 29 to mass 28 ion currents. The trace at the bottom is of the mass 29 and mass 28 ion currents as a function of time. The y-axis is the amplifier response for the minor (mass 29 ion).
from the reduction furnace of the GC/C/IRMS system is introduced directly into the ion source (5 × 10-6 mbar) of the Optima, allowing for analysis of each component at the low nanomole concentration of injection. RESULTS AND DISCUSSION Conventional IRMS. We previously reported16 a stable carbon isotope fractionation during amino acid acylation that likely resulted from the preferential incorporation of 12C during bond rupture and formation. In general, however, the δ15N values of the TFA/IP amino acid derivatives determined by conventional IRMS were within 0.5‰ of the δ15N values of their respective 928 Analytical Chemistry, Vol. 69, No. 5, March 1, 1997
underivatized values (Table 1). As in the case of carbon,16 small discrepancies may, in part, reflect impurities of unknown stable isotope compositions either in the reagents used for derivatization or in the bulk amino acid standards. GC/C/IRMS. It is apparent that the chromatographic separation employed in GC/C/IRMS eliminates impurites from the bulk sample derivatives. The δ15N values of the amino acid derivatives determined by GC/C/IRMS were, within the analytical error of the methods, almost identical to the δ15N values of their respective, underivatized amino acids (Table 1). In fact, average replicate values for seven of the amino acids determined by GC/C/IRMS were within 0.1‰ of their underivatized values. These results
Table 1. δ15N Values of Amino Acids Determined by IRMS and GC/C/IRMS amino acid D-Ala L-Ala
β-Ala D-Val L-Val Ivald Gly D-Leu D-Asp D-Phe R-Aibae
Table 2. δ15N Values for Racemic Mixtures (GC/C/IRMS)
δ15Na
δ15N derivativeb
δ15N GC/C/IRMSc
amino acid
δ15Na
amino acid
12.5 -6.3 -3.6 4.6 7.8 9.7 1.2 2.8 -5.3 -1.7 5.1
11.8 -6.3 -4.0 4.4 7.2 9.3 1.3 2.8 -4.8 -2.0 4.7
12.4 -5.9 -3.7 4.2 7.7 10.0 1.3 2.7 -5.2 -1.5 5.0
D-Ala
4.29 4.27 18.39 18.44 7.87 7.63
D-Asp
The δ15N values are for underivatized amino acids determined by conventional IRMS; standard deviation (0.1‰. b The δ15N derivative values were determined by conventional IRMS; standard deviation (0.1‰. c The GC/C/IRMS values are averages of nine runs, with an average standard deviation for each amino acid of e(0.4‰. d Ival, isovaline. e R-Aiba, R-aminoisobutyric acid. a
clearly indicate that stable nitrogen isotope fractionation during derivatization is minimal. As a further test of the method, replicate GC/C/IRMS analyses were performed on a mixture of racemic amino acids. Excellent resolution of the D- and L-enantiomers of the respective amino acids was achieved with the chromatographic column employed (Figure 2). Unlike carbon,25 the instantaneous 29/28 ratio for N2 clearly indicates a normal chromatographic discrimination (Figure 2), in which molecules containing the lighter nitrogen isotope (m/z 28) elute prior to molecules containing the heavier isotope (m/z 29). This isotope effect is the reverse of that observed for 13CGC/C/IRMS25 and is indicative of the lack of influence of the chromatographic separation on the 15N component of the derivative. As previously documented for carbon,26 the lack of shoulders on the peak inflections of the 29/28 ratios can be used as an independent check of compound purity. Coeluting compounds are likely to have at least slight differences in isotope compositions, which would be observed as minor abberations (shoulders) on the 29/28 ratios. The δ15N values for individual amino acid enantiomer derivatives in this racemic mixture had an average standard deviation of e(0.4‰. Thus, with the exception of glutamic acid, the δ15N values for the respective stereoisomers of the individual amino (25) Hayes, J. M.; Freeman, K. H.; Popp, B. N.; Hoham, C. H. Org. Geochem. 1990, 16, 1115-1128. (26) Lichtfouse, E.; Freeman, K. H.; Collister, J. W.; Merritt, D. A. J. Chromatogr. 1991, 585, 177-180.
L-Ala D-Val L-Val D-Leu L-Leu
L-Asp D-Glu L-Glu
δ15Na -2.58 -2.77 17.86 18.80
a Values are averages of nine GC/C/IRMS runs. The standard deviation for each analysis is e(0.4‰.
acids were, within the current limits of the method, identical (Table 2). The larger discrepancy observed for D- and L-glutamic acid may result from a minor contribution of column bleed, as these two components elute at a higher temperature at the end of the run. CONCLUSIONS It is now possible to determine δ15N values of amino acid enantiomers by GC/C/IRMS. Within experimental error, δ15N values of amino acid derivatives obtained by GC/C/IRMS are indistinguishable from δ15N values of the underivatized amino acids determined by conventional IRMS. This new GC/C/IRMS approach is superior in that (1) it permits the analysis of much smaller samples (nanomoles as opposed to milligram quantities required for IRMS) and (2) it will enable researchers to monitor stable nitrogen isotope variation of individual components within bulk samples, thus providing a better understanding of isotope fractionations associated with biosynthetic pathways and the diagenesis of organic matter in the geosphere. Furthermore, stable nitrogen isotope analysis of amino acid enantiomers by GC/ C/IRMS will provide an independent verification of the molecular stable carbon isotope method currently used to establish the indigeneity of these compounds in fossil15 and extraterrestrial21 materials. ACKNOWLEDGMENT This work was supported by the National Science Foundation (Grant EAR-9504618). We thank T. Brockwell (Micromass, Inc.) for his assistance with the GC/C/IRMS analyses. Received for review September 18, 1996. December 3, 1996.X
Accepted
AC960956L X
Abstract published in Advance ACS Abstracts, January 15, 1997.
Analytical Chemistry, Vol. 69, No. 5, March 1, 1997
929
