Non-racemic amino acids in the Murray and Murchison meteorites

PII S0016-7037(99)00280-X Non-racemic amino acids in the Murray and Murchison meteorites S. PIZZARELLO* and J. R. CRONIN Department of Chemistry and Biochemistry, Arizona State University, Tempe, AZ 85287-1604, USA (Received February 4, 1999; accepted in revised form June 28, 1999) Abstract—Small (1.0–9.2%) L-enantiomer excesses were found in six a-methyl-a-amino alkanoic acids from the Murchison (2.8–9.2%) and Murray (1.0–6.0%) carbonaceous chondrites by gas chromatography-mass spectroscopy of their N-trifluoroacetyl or N-pentafluoropropyl isopropyl esters. These amino acids [2-amino- 2,3-dimethylpentanoic acid (both diastereomers), isovaline, a-methyl norvaline, a-methyl valine, and a-methyl norleucine] are either unknown or rare in the terrestrial biosphere. Enantiomeric excesses were either not observed in the four a-H-a-amino alkanoic acids analyzed (a-amino-n-butyric acid, norvaline, alanine, and valine) or were attributed to terrestrial contamination. The substantial excess of L-alanine reported by others was not found in the alanine in fractionated extracts of either meteorite. The enantiomeric excesses reported for the a-methyl amino acids may be the result of partial photoresolution of racemic mixtures caused by ultraviolet circularly polarized light in the presolar cloud. The a-methyl-a-amino alkanoic acids could have been significant in the origin of terrestrial homochirality given their resistance to racemization and the possibility for amplification of their enantiomeric excesses suggested by the strong tendency of their polymers to form chiral secondary structure. Copyright © 2000 Elsevier Science Ltd 1. INTRODUCTION Molecular dissymmetry has been sought in meteorite organic matter for many years because of its implications for the origin of this material; that is, whether the organic matter was formed by chemical evolution, by extraterrestrial organisms, or is present only as a consequence of terrestrial contamination. Early attempts to make such determinations by measuring optical rotation, a phenomenon associated with molecular dis- symmetry, in meteorite organic compounds gave uniformly negative results. Organic solvent extracts of the Cold Bokkeveld meteorite (Mueller, 1953) and the Mokoia and Haripura meteorites (Briggs and Mamikunian, 1963) failed to show optical rotation, as did both organic solvent and aqueous extracts from eight carbonaceous chondrites of various petro- logic types (Kaplan et al., 1963). Subsequent polarimetric analyses, carried out during a con- troversy over the occurrence of supposed cellular remains in carbonaceous chondrites, gave conflicting results. Whereas op- tical activity was reported in saponified organic extracts of the Orgueil meteorite (Nagy et al., 1964; Nagy, 1965; 1966), it was not observed by others in similar, purified extracts, and the original observation was attributed to impurities and analytical artifacts (Hayatsu, 1965; Hayatsu, 1966; Meinschein et al., 1966). In a comprehensive review of this work, Hayes (1967) concluded that dissymmetry in the optically active compounds in Orgueil had not been proven. By 1969, the development of sensitive gas chromatographic (GC) methods allowed the direct analysis of individual amino acid enantiomers as diastereomeric N-trifluoroacetyl esters of optically active alcohols. Using this approach, Kvenvolden et al. (1970) found that alanine extracted from the recently fallen Murchison meteorite was racemic and that glutamic acid, pro- line, and valine were nearly so. Additional work showed that “all the amino acids that have asymmetric carbon atoms and whose diastereomeric derivatives could be separated by the GC method used appeared to consist of approximately equal amounts of the D and L isomers” (Kvenvolden, et al., 1971). In a separate study, isovaline was reported to exist in Murchison as a racemic mixture (Pollock et al., 1975). These findings, obtained with the presumably pristine Murchison meteorite, clearly established that amino acids were indigenous to the meteorite and were products of an abiotic chemical process. Concurrent analyses of the Murray meteorite using the same methods gave very similar results, i.e., nearly equal amounts of the D- and L-isomers were found for alanine, glutamic acid, proline, valine, a-amino-n-butyric acid, pipecolic acid, and b-aminoisobutyric acid (Lawless et al., 1971). The enantiomeric compositions of several Murchison amino acids were redetermined later by Engel and Nagy (1982). Although they found isovaline to be racemic and a-amino-n- butyric acid apparently so, five protein amino acids (alanine, glutamic acid, proline, aspartic acid, and leucine) showed sub- stantial L-excesses, which they suggested were characteristic of these amino acids as native to the meteorite. This proposal was criticized, largely on the grounds that the sampling procedure did not exclude terrestrial contaminants (Bada et al., 1983). Subsequently, Engel et al. (1990) bolstered the case for a significant enantiomeric excess in Murchison alanine using a chiral GC phase to separate enantiomers in a gas chromatog- raphy-combustion-isotope-ratio mass spectrometer (GC-C- IRMS). They found an L-alanine enantiomer excess of 8% and d13C values for the L- and D enantiomers of 127‰ and 130‰, respectively. On this basis, they argued for an indigenous L-excess on the grounds that terrestrial contamination account- ing for 8% of the alanine would have lowered the L-enantiomer d13C value more than was observed. Recently, Engel and Macko (1997) have extended this approach to 15N and, in this case, argued for indigenous L-excesses in alanine and glutamic acid of 33% and 54%, respectively, based on the similar d15N * Author to whom correspondence should be addressed (spizzarello@ asu.edu). Pergamon Geochimica et Cosmochimica Acta, Vol. 64, No. 2, pp. 329–338, 2000 Copyright © 2000 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/00 $20.00 1 .00 329 values obtained for the D- and L-enantiomers of these amino acids. Recently, we reported small (2.8–9.2%) L-enantiomer ex- cesses in a subset of Murchison amino acids, the a-methyl-a- amino acids (Cronin and Pizzarello, 1997); however, enantio- meric excesses were not observed in a-H-a-amino acids, including alanine (Pizzarello and Cronin, 1998). The latter finding is in agreement with the early results (Kvenvolden et al., 1970), but contrasts with the results of Engel and his coworkers (Engel and Nagy, 1982; Engel et al., 1990; Engel and Macko, 1997). In this article, we report additional Murchi- son amino acids with L-enantiomer excesses and the presence of a similar set of amino acids with L-excesses in the Murray meteorite. 2. ANALYTICAL PROCEDURES 2.1. Extraction and Isolation of Amino Acids Two interior pieces of the Murray meteorite (Center for Meteorite Studies, Arizona State University) weighing 26 g in total were crushed in a steel press, powdered by using a glass mortar and pestle, and extracted with triple-distilled water for 24 hr at 110°C under vacuum. The powder was sedimented by centrifugation and the extract decanted and combined with two water rinses of the insoluble residue. The extract was then concentrated by rotary evaporation, acidified to pH 2 with 1N sulfuric acid, filtered, and desalted by adsorption on a 50 mL column of cation exchange resin (BioRad AG50W-X4, 200– 400 mesh, H1 form). The resin column was washed with water and the amino acids then eluted with 2N ammonium hydroxide. The Murchison results presented here were obtained by further analysis of data obtained earlier (Cronin and Pizzarello, 1997). 2.2. Amino Acid Fractionation The ammonium hydroxide eluate of the cation-exchange column was dried by rotary evaporation and redissolved in citrate buffer (pH 4.0, 0.2 M Na1), applied to a reversed-phase, semiprep C18 column (Supelcosil-LC 18, 25 cm 3 10 mm) and eluted with the same citrate buffer at a flow rate of 45 mL per hour (Cronin and Pizzarello, 1986). Two-minute fractions were collected, and those containing amino acids of interest were combined and the amino acids reisolated as above on 10–15 mL cation exchange columns. 2.3. Amino Acid Analysis The amino acids were dried by rotary evaporation in a screw-cap vial and esterified with isopropanol/HCl (;2N) at 100°C for 1 hr. The resulting isopropyl esters were dried by rotary evaporation and N-fluoroacylated by adding approxi- mately 50 mL of either trifluoroacetic anhydride or pentaflu- oropropionic anhydride in dichloromethane (1:5, v:v) and heat- ing at 100° for 10 min. The resulting N-trifluoroacetyl- (TAI) or N-pentafluoropropionyl-O-isopropyl (PPI) derivatives of the amino acids were concentrated by evaporation in air and dis- solved in dichloromethane. Aliquots of this solution were care- fully dried on a solids injector for GC-MS analysis. A Hewlett Packard HP 5880/HP 5970 combined gas chromatograph/mass spectrometer was used with a capillary column coated with a chiral phase (Alltech, Chirasil-L-Val, 50m 3 0.25 mm, 0.25 m film thickness). The GC injector was maintained at 230°C, and the oven temperature program was varied to optimize resolu- tion of the derivatized enantiomers of each amino acid as shown in Table 1. 3. RESULTS 3.1. Enantiomeric Analyses of Murchison and Murray Amino Acids The results of enantiomeric analyses of ten amino acids from the Murray and Murchison meteorites are given in Table 2. The corrected enantiomeric excesses (ee) are based on mean values calculated from integrated single ion plots and, when justified (see below), total ion (TI) plots of data obtained from multiple chromatographic runs. Corrections were made in each case for the small mean ee obtained from runs of the corresponding racemic amino acid standards under the same conditions and using about the same amounts of the amino acids. Unless noted otherwise, the differences between the meteorite mean values and those of the standards range in significance from the 90% to 99.9% confidence level using Student’s t test of two inde- pendent means (Ipsen and Feigl, 1970). The significant ee values determined for six a-methyl amino acids from the Murchison meteorite range from 2.8 to 9.2%. Those measured Table 1. Derivatives analyzed, temperature programs for enantiomer separations, and ions integrated. Amino acid Derivative analyzed Initial GC programa M1 (m/z) Ions integrated (m/z) 2-Amino-2,3-dimethyl pentanoic acid 2S,3S/2R,3R PPI 70°, 59, 2° min21 333 246, 217, 83 2S,3R/2R,3S 0 0 0 0 0 0 a-Methyl-norleucine TAI 55°, 59, 1° min21 283 TI, 196, 140, 55 a-Methyl-norvaline 0 50°, 59, 2° min21 269 TI, 184, 182, 166, 69 a-Methyl-valine 0 0 0 182, 167, 166 Isovaline 0 40°, 109, 2° min21 255 TI, 184, 168, 166, 55 Norvaline 0 0 0 168, 126 Valine 0 0 0 168, 153, 55 a-Amino-n-butyric acid 0 70°, 59, 2° min21 241 154, 140, 126 Alanine 0 0 227 141, 140 a In all cases, after completion of the initial program, the temperature was ramped from 120° to 200° at 4° min21 and then held constant for 1 h. TI, Total ion intensity. 330 S. Pizzarello and J. R. Cronin for the same amino acids from the Murray meteorite range from 1.0 to 6.0%, on average about 40% of the Murchison values. All of the amino acids with significant L-excesses have an a-methyl group in addition to the usual a-alkyl (R) group. In contrast, two a-H homologues of these amino acids, a-amino- n-butyric acid and norvaline, were found to be racemic in fractionated extracts of both meteorites. 3.2. Potential Errors Several alternative explanations for an apparent ee must be ruled out before it can be viewed as truly characteristic of the indigenous amino acid (Cronin and Pizzarello, 1997). The most important of these are (a) systematic bias in the analytical system; (b) terrestrial contamination; and (c) coelution of other meteoritic amino acids. As noted above, analytical bias was eliminated by running racemic standards along with the mete- orite samples and correcting the ee determined for each mete- orite amino acid for any deviation from 0% in the ee deter- mined for the corresponding standard. The probability of terrestrial contamination is assumed to differ among amino acids depending on their abundance in the biosphere. Table 3 summarizes the results of literature searches for reported natural occurrences of the amino acids discussed here. In four cases, to the best of our knowledge, the amino acids have never been reported and the contribution of L- enantiomers by contamination is considered to be unlikely. Furthermore, the observation that alanine and valine, ubiqui- tous protein amino acids, are racemic (Murray) or show only small ee (Murchison) in our samples indicates that contamina- tion of these specimens is minimal, at worst, lowering even further our estimate of the likelihood of a terrestrial contribu- tion to the biologically rare or unknown amino acids. We believe the most likely source of error to be the unrec- ognized coelution of another meteoritic amino acid with one of the amino acid enantiomers of interest. To fully appreciate this possibility, it must be remembered that the Murchison amino acids show complete structural diversity within the general types present (Cronin et al., 1985; Cronin and Pizzarello, 1986). As carbon number increases, the number of isomeric possibilities increases markedly, whereas the concentrations within homologous series decrease exponentially. Taking the seven-carbon amino alkanoic acids as an example, there are (including diastereomers) 18 a-amino isomers, 90 amino posi- tion isomers, i.e., b-z amino acids, and 53 five- and six- membered ring cycloalkanoic amino acids, excluding N-con- taining rings. Even more numerous higher homologues are present in amounts about 30% (eight-carbon homologues) and Table 2. Enantiomeric excesses in amino acids from the Murchison and Murray meteorites. Amino acid Sample Standard Confidence (%) Corrected ee (%)L (%) s n ee (%) L (%) s n ee (%) Murray meteorite 2-Amino-2,3-dimethyl-pentanoic acid 2S,3S/2R,3R 50.6 0.7 12 1.2 50.1 0.6 16 0.2 .95 1.0 2S,3R/2R,3S 51.0 0.6 12 2.0 49.9 0.6 17 20.2 .99 2.2 Isovaline 53.6 0.4 20 7.2 50.6 0.3 24 1.2 .99.9 6.0 a-Methylnorvaline 51.1 0.3 23 2.2 50.4 0.3 30 0.8 .99.9 1.4 a-Methylvaline 51.1 0.8 21 2.2 50.6 0.7 17 1.2 .90 1.0 a-Methylnorleucine 51.1 0.5 18 2.2 50.2 0.6 19 0.4 .99.9 1.8 a-Amino-n-butyric acid 50.8 0.5 14 1.6 51.0 0.4 20 2.0 Not sig. 20.4 Norvaline 50.9 0.5 10 1.8 50.5 0.9 17 1.0 Not sig. 0.8 Alanine 51.0 0.3 21 2.0 50.8 0.3 27 1.6 .95 0.4 Valine 50.6 0.6 14 1.2 50.8 0.4 13 1.6 Not sig. 20.4 Murchison meteorite * 2-Amino-2,3-dimethyl-pentanoic acid 2S,3S/2R,3R 52.6 0.5 5 5.2 48.8 1.9 14 22.4 .99 7.6 2S,3R/2R,3S 54.7 0.6 5 9.4 50.1 0.8 18 0.2 .99.9 9.2 * Isovaline 54.6 0.6 8 9.2 50.4 0.6 15 0.8 .99.9 8.4 * a-Methylnorvaline 51.4 0.4 10 2.8 50.0 0.2 10 0 .99.9 2.8 a-Methylvaline 51.6 0.3 7 3.2 50.2 0.6 10 0.4 .99.9 2.8 a-Methylnorleucine 52.5 1.3 10 5.0 50.3 0.4 13 0.6 .99.9 4.4 *a-Amino-n-butyric acid 50.4 0.2 3 0.8 50.2 0.2 12 0.4 Not sig. 0.4 *Norvaline 50.2 0 3 0.4 50.0 0.2 10 0 Not sig. 0.4 Alanine 50.8 0.4 15 1.6 50.2 0.3 16 0.4 .99.9 1.2 Valine 51.3 0.4 12 2.6 50.2 0.2 13 0.4 .99.9 2.2 * Reported previously (Cronin and Pizzarello, 1997) Table 3. Terrestrial occurrence of the meteoritic amino acids ana- lyzed. Amino acid Occurrence 2-amino-2,3-dimethylpentanoic acid diastereomers (a-Methyl-isoleucine/a-Methyl-alloisleucine) Unknown a-Methyl-norleucine Unknown a-Methyl-norvaline Unknown a-Methyl-valine Unknown Isovaline Rare Norvaline Rare Valine Ubiquitous a-Amino-n-butyric acid Common Alanine Ubiquitous 331Non-racemic amino acids in meteorites 10% (nine-carbon homologues) of the concentrations of the seven-carbon amino acids. Clearly, there is a significant possi- bility of coelution in a chromatographic analysis of these amino acids and, even with the prior fractionation step used in this work, careful mass spectral analysis of each enantiomeric peak is necessary in order to be confident that other amino acids are not contributing to its integrated ion intensity. In attempting to recognize coelution, we have found it useful to sum the mass spectra taken over each of the two enantio- meric peaks and subtract the summed mass spectra of one peak, usually the less intense peak, from the other peak. The fragment ions from extraneous compounds coeluting with either enan- tiomer are abundant in such different spectra (in principle, they are the only ions if the amino acid is racemic) and clearly indicate the presence of a coeluting compound(s). An example is given in Figure 1 based on analysis of the valine-containing fraction from the Murray meteorite. Figure 1a is a total ion chromatogram showing the D-valine and L-valine peaks which, after integration, indicate an L-excess of 1.2%. Figure 1b shows the difference mass spectrum obtained by subtracting the summed mass spectra of the D-enantiomer peak from those of the L-enantiomer peak. Comparison of this difference spectrum with that of a TAI-valine standard (Fig. 1c) shows that, along with the principal ions of the TAI-valine mass spectrum, there are additional ions at m/z 85, 138, and 180 indicating that a calculation based on the total ion chromatogram would over- estimate the valine ee. Figure 1d shows a single mass spectrum taken near the maximum of the L-valine peak and illustrates the importance of using enantiomer difference spectra in evaluat- ing the “purity” of enantiomer peaks. In this spectrum the extraneous ions are not intense and are easily overlooked in the mass spectrum, which in general looks very much like that of TAI-valine. Once the presence of an extraneous coelutant is noted it is also necessary to consider whether its fragmentation in the mass spectrometer might produce ions with masses in common with the more abundant fragment ions of the amino acid of interest on which an ee calculation might be based. In this case, the extraneous ions suggest that the coelutant is a TAI-methyl proline (m/z 180 is the expected base peak) and, if this is correct, contributions to the intensity of TAI-valine ions at m/z 55, 153, and 168 would not be expected. Consequently, we felt confident in basing a calculation of ee on these integrated fragment ions, which gave an insignificant ee of 0.4% (Table 1). In contrast with valine, in the case of isovaline there is no indication from the difference mass spectrum of any coeluting substance. The difference mass spectrum (Fig. 2c) matches almost perfectly the mass spectrum of the amino acid standard (Fig. 2b), thus establishing the chromatographic “purity” of the enantiomer peaks, validating the use of the total ion chromato- gram in calculations, and fully supporting the attribution of the difference in area to a true enantiomer excess. In order to minimize the coelution problem, we fractionated the total amino acids of the whole extract by reverse-phase chromatography before GC analysis, a procedure that separates amino acids mainly on the basis of polarity (Cronin and Piz- zarello, 1986). Because the meteoritic amino acids are mainly amino alkanoic acids, the most likely coelutants carried through this fractionation along with a particular amino acid are those of the same or similar molecular mass, i.e., carbon chain isomers, amino position isomers (e.g., b-, g-, d-isomers), N- alkyl isomers, cyclic or unsaturated amino acid analogs (see valine example, above), and homologues with one more or one less carbon atom. Consequently, we gave particular attention to the possible presence of these species in attempting to identify coelutants and determine which fragment ions could legiti- mately be used for calculation of ee in each case. 4. DISCUSSION 4.1. Comparison with Previous Results The significant enantiomeric excesses in the a-amino acids of the Murchison and Murray carbonaceous chondrites appear to be confined to the a-methyl amino alkanoic acids, a subset of the amino acids that has not been extensively analyzed heretofore. Our finding of a significant L-enantiomer excess in isovaline contrasts with Pollock et al. (1975) who reported this amino acid to be racemic in Murchison; however, it should be noted that the latter result was obtained by GC analysis alone of a complex fraction containing numerous other components. It cannot be considered definitive due to the possibility of unrec- ognized coelution. Chiral analyses of the a-methyl amino acids from the Murray meteorite have not been carried out previ- ously. The finding of no or very small ee in alanine, a-amino-n- butyric acid, and valine agrees with earlier findings for both the Murchison (Kvenvolden et al., 1970, 1971) and Murray (Law- less et al., 1971) meteorites. We have not observed the large (33%) L-alanine excess in the Murchison meteorite reported by Engel and Macko (1997). It is important to note that Engel and coworkers carried out their enantiomer separations without either prior fractionation of the amino acids or alteration of the standard gas chromatographic conditions to better resolve the complex mixture present in a meteorite extract (Engel et al., 1990; Engel and Macko, 1997). We have identified five amino acids in the whole meteorite extract with GC retention times on Chirasil-L-Val similar to those of the alanine enantiomers and believe that coelution of one or more of these may have compromised both their measured enantiomer ratio and isoto- pic values (Pizzarello and Cronin, 1998). Fractionation of the meteorite extracts by reverse-phase chromatography, as done in this work, removes most of these amino acids from the alanine fraction. 4.2. Possible Mechanisms for the Formation of Enantiomeric Excesses 4.2.1. Neutron star produced UVCPL We have suggested (Cronin and Pizzarello, 1997) that the enantiomeric excesses observed in meteoritic amino acids could have originated in the presolar cloud as a result of its irradiation by ultraviolet circularly polarized light (UVCPL) from a neutron star, a concept proposed by others (Rubenstein et al., 1983; Bonner and Rubenstein, 1987) and recently dis- cussed in the context of interstellar grains and cometary organic matter (Greenberg, 1996). Substantial deuterium enrichments in the meteoritic amino acids suggest that they or their precur- sors are of interstellar cloud origin (Epstein et al., 1987). The 332 S. Pizzarello and J. R. Cronin Fig. 1. Analysis of valine enantiomers from the Murray meteorite. (a) Chirasil-L-Val total ion chromatogram showing TAI-valine enantiomers (large peaks at ca. 31.3 and 32.4 min.) in reverse-phase fraction of Murray extract. (b) Difference mass spectrum of summed L-peak minus summed D-peak shown in (a). (c) GC-mass spectrum of standard TAI-valine. (d) Single GC-mass spectrum from TAI-L-valine peak shown in (a). Fig. 2. Analysis of isovaline (2-amino-2-methyl butanoic acid) enantiomers from the Murray meteorite. (a) Chirasil-L-Val total ion chromatogram showing TAI-isovaline enantiomers in reverse-phase fraction of Murray extract. (b) GC mass spectrum of standard TAI-isovaline. (c) Difference mass spectrum of summed L-peak minus summed D-peak shown in (a). 334 S. Pizzarello and J. R. Cronin presence of graphite, SiC, and Si3N4 (supernova grains) in carbonaceous chondrites (Amari and Zinner, 1990) provides evidence for the supernova precursor of a neutron star; conse- quently, exposure of meteoritic organic matter, at least at an early stage in its formation, to electromagnetic radiation from a neutron star does not seem unlikely. In laboratory experiments, UVCPL photolysis (212.8 nm) has been shown to give rise to enantiomeric excesses in race- mic leucine of 1.98% (right handed) and 2.50% (left handed) with 59% and 75% overall decomposition, respectively (Flores et al., 1977). Resolution of racemic mixtures by UV photolysis is a race between overall decomposition and the selective decomposition of a particular enantiomer; thus, the enantio- meric purity achieved depends on the extent of photolysis and the difference in the absorbance of polarized light by the enantiomers. The latter is expressed as the anisotropy factor, g, where g 5 [«S 2 «R]/«), «S and «R are the extinction coeffi- cients of enantiomers for CPL, and « is the extinction coeffi- cient of the racemate(Balavoine et al., 1974). In theory, the enantiomeric excesses expected by photodecomposition of leucine (g 5 0.0244; Flores et al., 1977) are 1.7% at 80% decomposition and 3.2% at 95% decomposition, values in fair agreement with experiment and comparable to most of the meteoritic values. The order of magnitude larger L-excesses reported for alanine (Engel and Macko, 1997) seem to be outside the range achievable by UVCPL photolysis given the g values of a-amino acids (Katzin and Gulyas, 1968). Mason (1988; 1997) has criticized the Rubenstein–Bonner hypothesis as a mechanism for the production of enantiomeric excesses on the grounds that the Kuhn–Condon zero sum rule (cancellation of enantiomer CD spectrum over the full electro- magnetic spectrum) prevents photochemical discrimination by the broad band emission of a neutron star. In response, Bonner (1991) pointed out that even though there may be no net absorption difference across the full spectrum, the particular absorption bands excited are not the same for enantiomers and therefore their tendency to undergo photolysis will not neces- sarily be the same. In addition, Roberts (1984) raised an ob- jection on theoretical grounds to the production of CPL by neutron stars soon after Rubenstein et al. (1983) set out their proposal. 4.2.2. UVCPL from Mie scattering Recently, Bailey et al. (1998) have suggested a modified form of the Rubenstein–Bonner hypothesis that avoids the objections noted above. They have observed substantial circu- lar polarization in the infrared from the Orion molecular cloud and interpreted this as originating not from a neutron star, but from scattering of unpolarized star light by magnetically aligned spheroidal interstellar grains (Mie scattering). They calculate that UV light would be similarly circularly polarized, although difficult to observe. They point out that, in the case of many stars, there is a major drop in UV intensity below 220 nm, such that the selective destruction of amino acid enanti- omers of a particular configuration would be possible as a result of the excitation at only their longest wavelength band; i.e., the objection based on the Kuhn-Condon rule would not apply. They suggest the origin of UVCPL by Mie scattering as an alternative to the neutron star hypothesis, but a hypothesis with the same implications for the chirality of interstellar organic matter. 4.2.3. Enantiomeric excesses and amino acid structure The presence of enantiomeric excesses in the a-methyl amino acids may be due to the fact that racemization of these amino acids is made difficult by methyl substitution of the a-H atom (Pollock et al., 1975) and, as a result, they have retained their original enantiomeric excesses even though exposed to conditions that over time allowed racemization of the a-amino acids with an a-H-atom, for example, during aqueous process- ing of the meteorite parent body (Bunch and Chang, 1980). Alternatively, it is possible that the a-methyl- and a-H- amino acids are unracemized products of different formation processes, with the former having uniquely involved some asymmetric influence. This idea is consistent with the observa- tion that the two types of amino acids vary in relative amount among different Murchison specimens (Cronin and Pizzarello, 1983). For example, a specimen from the Field Museum has been found to have about twice the concentration of isovaline, an a-methyl amino acid, of a specimen from the ASU Nininger Collection, but only about one-half the content of valine, an a-H isomer. Similar differences were seen in the relative amounts of other a-methyl- and a-H-amino acids. We have also noted that the two types of amino acid differ in their ease of extraction from meteorite powders (Pizzarello, unpublished results). Taken together, these observations suggest the associ- ation of the a-methyl- and a-H-amino acids with distinct matrix phases that were mixed in somewhat different ratios throughout the Murchison parent body. If this idea is correct, the possibility of separate origins might be inferred: for exam- ple, presolar formation of the a-methyl amino acids and parent body formation of those with a-H atoms. 4.3. Implications for the Formation of Meteoritic Amino Acids The formation of meteoritic amino acids has come to be viewed as a two-step process in which their precursors were formed in the presolar cloud and, after incorporation into an icy, volatile-rich primitive parent body, underwent Strecker reactions during its aqueous processing (Cronin et al., 1995). The high deuterium content of meteorite amino acids suggests formation of at least their hydrocarbon chains by the low- temperature ion-molecule reactions that occur in interstellar clouds (Epstein et al., 1987). The abundance of amino acids with branched chain structures and the decrease in concentra- tion with increasing chain length are also consistent with in- terstellar chemistry. On the other hand, the presence of an abundant suite of a-hydroxy acids, which essentially match the a-amino acids with respect to carbon chain structures, suggests the formation of both sets of compounds by Strecker reactions under aque- ous-phase conditions that might plausibly have existed in the meteorite parent body (Peltzer and Bada, 1978; Cronin et al., 1993). Amino acids and the other classes of organic compounds found in carbonaceous chondrites are abundant only in the CI and CM meteorites, which are types that have been extensively altered by liquid water. The icy, volatile-rich planetesimals 335Non-racemic amino acids in meteorites from which the parent bodies of these meteorites were formed might reasonably be expected to have contained aldehydes, ketones, ammonia, and hydrogen cyanide, the reactants in Strecker reactions. All of these compounds have been shown to be present in interstellar clouds (Irvine, 1998). Thus, the parent body of the organic-rich carbonaceous chondrites might well have contained all the reactants, as well as providing the reaction conditions necessary for a Strecker synthesis. The presence of enantiomeric excesses in the Murchison and Murray amino acids is difficult to reconcile with this formation hypothesis if UVCPL provided the necessary asymmetric in- fluence as suggested above. If their asymmetry was achieved by preferential photolysis, it must have been manifest after (or during) formation of the amino acids; however, if this process occurred in the early solar system it seems likely that the amino acids would have been shielded by the parent body. If the exposure to UVCPL occurred in the presolar cloud, only the precursor ketones of the a-methyl amino acids would have been affected; consequently, there would have been only a secondary effect on the chiral a-carbons of the amino acids (formed later) and then, only in the case of those formed from chiral aldehydes/ketones. The C7 amino acid, 2-amino-2,3- dimethyl pentanoic acid, is interesting in this regard. The precursor of the carbon chain of this amino acid in a Strecker synthesis is the chiral ketone, 3-methyl-2-pentanone (asymmet- ric C3). In this case, if the presolar cloud had been exposed to UVCPL, the ketone would have carried an enantiomeric excess when incorporated into the primitive parent body. Formation of the amino acid from this ketone by Strecker reactions would then have given four stereoisomers, but because of the preex- isting enantiomeric excess in the ketone, an L-excess would have been observed in one pair of enantiomers, and a D-excess in the other pair (Cronin and Pizzarello, 1997).1 To the con- trary, as seen in the data presented in Table 2, an L-excess was observed in both enantiomeric pairs. Consequently, if the asymmetric influence was, in fact, UVCPL, formation of the nonracemic a-methyl-a-amino acids seems likely to have been entirely a presolar process. 4.4. Origin of Life Implications As Pasteur realized almost 150 years ago (Geison, 1995), understanding the origin of homochirality may be key to un- derstanding the origin of life. Since that time, numerous biotic and abiotic theories have been proposed (Bonner, 1991); ac- cording to the former, life initially was based on achiral mol- ecules and/or racemates and the use of specific enantiomers came about through evolution, whereas the latter theories pro- pose that a tendency toward homochirality was inherent in chemical evolution. The finding of L-enantiomeric excesses in the amino acids of meteorites clearly supports the latter pro- posal. Since the accretion of dust, meteorites and, possibly, cometary material provided the early earth with a significant fraction of its organic inventory (Chyba and Sagan, 1992), the amino acids so delivered may have provided the initial enan- tiomeric excesses necessary for amplification by further chem- ical evolution. The a-methyl amino acids, which have not been viewed heretofore as important for the origin of life (they are generally unimportant in terrestrial biochemistry), are abundant in carbonaceous chondrites and may be well-suited for such a role. Polymerization accompanied by formation of regular sec- ondary structure, for example, a-helices and b-sheets, has been shown to be an effective way to amplify modest initial enan- tiomeric excesses (Bonner et al., 1981; Brack and Spach, 1981) and a-methyl amino acids are known to have strong helix inducing and stabilizing effects (Altman et al., 1988; Formag- gio et al., 1995). Furthermore, these amino acids would not be subject to the “catastrophe of racemization” (Bonner, 1991). Any hard won gains in their enantiomeric excesses would be stable to chemical racemization (Pollock et al., 1975), whereas this would not be the case for amino acids lacking an a-methyl substituent (Bada and Miller, 1987). A transition to a-H-amino acids obviously must have occurred later, perhaps to allow amino acid formation by transamination of a-keto analogues, a process that is very important in the metabolism of protein amino acids but impossible for the a-methyl amino acids. Be that as it may, a-methyl amino acids could conceivably have played a significant early role in the chemical evolution of homochirality and perhaps have been significant in the bio- chemistry of a pre-RNA world. 4.5. Molecular Dissymmetry and the Search for Extraterrestrial Life In the years after Pasteur’s discovery of the molecular basis for the asymmetry of biomolecules, work in organic chemistry showed them to be in general largely chiral compounds and, moreover, almost exclusively homochiral. For example, of the 20 protein amino acids, 19 are chiral and exclusively of the L-configuration (Greenstein and Winitz, 1961). Biochemical studies of the relationship between biological activity and mac- romolecular structure have made it clear why this is the case. Taking proteins as an example, X-ray crystallography has shown them to have a unique three-dimensional structure, and denaturation studies have demonstrated repeatedly that even small alterations in this characteristic structure commonly lead to loss of activity. Interchanging the a-H and the a-R-group of a single amino acid residue, the structural change equivalent to substituting the opposite enantiomer, would lead to an inactive structural variant of a protein in many cases and several such inversions of configuration would almost assuredly cause loss of activity, a distressing fact well known to synthetic peptide chemists. Taking a small 100 residue protein as an example and assuming that the correct amino acid is incorporated sequen- tially at each position but the D- or L-enantiomer choice left to chance, 1.27 3 1030 isomeric forms of the protein would be possible, of which only a relatively small number would be likely to show the biological activity of the homochiral isomer, i.e., of the all L-amino acid protein. Thus the chiral specificity of each amino acid residue must be a sine qua non for protein- based life wherever it may occur, and overall amino acid homochirality may be a more efficient way of achieving this than specifying both the amino acid and its configuration at each chain elongation step. 1The situation here is somewhat analogous to the epimerization of L-isoleucine in which the configuration change at the asymmetric b-carbon is slow relative to that at the a-carbon. The effect is the conversion of L-isoleucine to an equilibrium mixture of L-isoleucine and D-alloisoleucine (D-alloile/L-ile ’ 1.3) 336 S. Pizzarello and J. R. Cronin As others have pointed out (MacDermott et al., 1996), searching for extraterrestrial homochirality may be a relatively unambiguous and cost-effective strategy for seeking extant life elsewhere in the solar system. A sensitive polarimetric instru- ment might be used for this purpose, which could also be useful for the recognition of extinct life as well as prebiotic organic matter. Fossil proteins undergo diagenesis accompanied by temperature-dependent racemization of their constituent amino acids with measurable enantiomeric excesses surviving for hundreds of thousands of years or longer (Bada, 1991). It has been estimated that the amino acid enantiomeric excesses of an extinct martian biota could still survive in cold, dry environ- ments (Bada and McDonald, 1995). As described here, some prebiotic amino acids also carry a chiral signature and, as a result, an observation of net optical rotation cannot be taken as a completely unambiguous indication of life. However, remem- bering the diversity of molecular structure apparent in meteor- itic organic compounds, that is, in the products of organic chemical evolution, measurements of optical rotation combined with amino acid analysis could, in principle, differentiate pre- biotic chemistry from protein-based life and/or its diagenetic products, in extraterrestrial bodies. 5. CONCLUSIONS 1) Six a-methyl-a-amino acids have been extracted from both the Murchison and Murray meteorites and found to have small excesses of the L-enantiomers. These amino acids are either unknown or of rare occurrence in the biosphere. 2) The enantiomeric excesses range from 2.8% to 9.2% in the amino acids from the Murchison meteorite and from 1.0% to 6.0% in the corresponding amino acids from the Murray meteorite. 3) Thus far, amino acids that lack an a-methyl group in addition to the usual a-alkyl group have been found to be racemic or nearly so as extracted from these two meteor- ites. The amino acids analyzed include the protein amino acids alanine and valine. 4) Exposure of meteoritic amino acids in the presolar cloud to UVCPL originating from either Mie scattering of starlight or possibly from a nearby neutron star are reasonable hypotheses for the origin of their enantiomeric excesses. 5) The occurrence of enantiomeric excesses in a-methyl-a- amino acids may have interesting origin of life implica- tions assuming meteoritic organic matter to be representa- tive of organic matter on the early earth. When polymerized, these amino acids are avid helix formers, suggesting a mechanism for the amplification of their initial small enantiomeric excesses, and their resistance to racemization would prevent the loss of gains in enantio- meric excess thus achieved. Acknowledgments—The authors gratefully acknowledge research sup- port from the NASA Exobiology Program (NAG5-4131), meteorite samples provided by Carleton Moore from the collections of the Center for Meteorite Studies, Arizona State University, and constructive re- views by Keith Kvenvolden and Jeffrey Bada. REFERENCES Altman E., Altman K. H., Nebel K., and Mutter M. (1988) Conforma- tional studies on host-guest peptides containing chiral a-methyl-a- amino acids. Int. J. Pept. Protein Res. 32, 344–351. Amari S. and Zinner E. (1990) Supernova grains from meteorites. Nature 345, 238–240. Bada J. L. (1991) Amino acid cosmogeochemistry. Phil. Trans. R. Soc. Lond. B. 333, 349–358. Bada J. L., Cronin J. R., Ho M.-S., Kvenvolden K. A., Lawless J. G., Miller S. L., Oro´ J., and Steinberg S. (1983) On the reported optical activity of amino acids in the Murchison meteorite. Nature 301, 494–497. Bada J. L. and McDonald G. D. (1995) Amino acid racemization on Mars: Implications for the preservation of biomolecules from an extinct martian biota. Icarus 114, 139–143. Bada J. L. and Miller S. L. (1987) Racemization and the origin of optically active organic compounds in living organisms. BioSystems 20, 21–26. Bailey J., Chrysostomou A., Hough J. H., Gledhill T. M., McCall A., Clark S., Me´nard F., and Tamura M. (1998) Circular polarization in star formation regions: Implications for biomolecular homochirality. Science 281, 672–674. Balavoine G., Moradpour A., and Kagan H. B. (1974) Preparation of chiral compounds with high optical purity by irradiation with circu- larly polarized light, a model reaction for the prebiotic generation of optical activity. J. Am Chem. Soc. 96, 5152–5158. Bonner W. A. (1991) The origin and amplification of biomolecular chirality. Origins Life Evol. Biosphere 21, 59–111. Bonner W. A., Blair N. E., and Dirbas F. M. (1981) Experiments on the abiotic amplification of optical activity. Origins Life 11, 119–134. Bonner W. A. and Rubenstein E. (1987) Supernovae, neutron stars and biomolecular chirality. BioSystems 20, 99–111. Brack A. and Spach G. (1981) Enantiomer enrichment in early pep- tides. Origins Life 11, 135–142. Briggs M. H. and Mamakunian G. (1963) Organic constituents of the carbonaceous chondrites. Space Sci. Rev. 1, 647–682. Bunch T. E. and Chang S. (1980) Carbonaceous chondrites. II. Carbo- naceous chondrite phyllosilicates and light element geochemistry as indicators of parent body processes and surface conditions. Geochim. Cosmochim. Acta 44, 1543–1577. Chyba C. F. and Sagan C. (1992) Endogenous production, exogenous delivery, and impact-shock synthesis of organic molecules: An in- ventory for the origins of life. Nature 355, 125–132. Cronin J. R., Cooper G. W., and Pizzarello S. (1995) Characteristics and formation of amino acids and hydroxy acids of the Murchison meteorite. Adv. Space Res. 15, 91–97. Cronin J. R. and Pizzarello S. (1983) Amino acids in meteorites. Adv. Space Res 3, 5–18. Cronin J. R. and Pizzarello S. (1986) Amino acids of the Murchison meteorite. III. Seven-carbon acyclic primary a-amino alkanoic acids. Geochim. Cosmochim. Acta 50, 2419–2427. Cronin J. R. and Pizzarello S. (1997) Enantiomeric excesses in mete- oritic amino acids. Science 275, 951–955. Cronin J. R., Pizzarello S., and Yuen G. U. (1985) Amino acids of the Murchison meteorite. II. Five-carbon acyclic primary b-, g-, and d-amino alkanoic acids. Geochim. Cosmochim. Acta 49, 2259–2265. Cronin J. R., Pizzarello S., Epstein S., and Krishnamurthy R. V. (1993) Molecular and isotopic analyses of the hydroxy acids, dicarboxylic acids, and hydroxydicarboxylic acids of the Murchison meteorite. Geochim. Cosmochim. Acta 57, 4745–4752. Engel M. H. and Macko S. A. (1997) Isotopic evidence for extrater- restrial non-racemic amino acids in the Murchison meteorite. Nature 389, 265–268. Engel M. H., Macko S. A., and Silfer J. A. (1990) Carbon isotope composition of individual amino acids in the Murchison meteorite. Nature 348, 47–49. Engel M. H. and Nagy B. (1982) Distribution and enantiomeric com- position of amino acids in the Murchison meteorite. Nature 296, 837–840. Epstein S., Krishnamurthy R. V., Cronin J. R., Pizzarello S., and Yuen G. U. (1987) Unusual stable isotope ratios in amino acid and car- 337Non-racemic amino acids in meteorites boxylic acid extracts from the Murchison meteorite. Nature 326, 477–479. Formaggio F., Crisma M., Bonora G. M., Pantano M., Valle G., Toniolo C., Aubry A., Bayeul D., and Kamphuis J. (1995) (R)- Isovaline homopeptides adopt the left-handed 310-helical structure Peptide Res. 8, 6–14. Flores J. J., Bonner W. A., and Massey G. A. (1977) Asymmetric photolysis of (R,S)-leucine with circularly polarized light, J. Am Chem. Soc. 99, 3622–3625. Geison G. L. (1995) The Private Science of Louis Pasteur. Princeton Univ. Press. Greenberg J. M. (1996). Chirality in interstellar dust and in comets: life from dead stars. In Physical Origin of Homochirality in Life (ed. D. B. Cline) pp. 185–210. American Institute of Physics. Greenstein J. P. and Winitz M. (1961) Chemistry of the Amino Acids. Wiley. Hamilton P. B. (1965) Amino acids on hands. Nature 205, 284–285. Hayatsu R. (1965) Optical activity in the Orgueil meteorite. Science 149, 443–447. Hayatsu R. (1966) Artifacts in polarimetry and optical activity in meteorites. Science 153, 859–861. Hayes J. M. (1967) Organic constituents of meteorites—a review. Geochim. Cosmochim. Acta 31, 1395–1440. Ipsen J. and Feigl P. (1970) Barcroft’s Introduction to Biostatistics, 2nd ed., Harper & Row. Irvine W. (1997) Extraterrestrial organic matter: A review. Origins Life Evol. Biosphere 28, 365–383. Kaplan I. R., Degens E. T., and Reuter J. H. (1963) Organic compounds in stony meteorites. Geochim. Cosmochim. Acta 27, 805–834. Katzin L. I. and Gulyas E. (1968) Absorption, rotatory dispersion, and circular dichroism studies on some hydroxy and amino acids. J. Am Chem. Soc. 90, 247–251. Kvenvolden K., Lawless J., Pering K., Peterson E., Flores J., Pon- namperuma C., Kaplan I. R., and Moore C. (1970) Evidence for extraterrestrial amino acids and hydrocarbons in the Murchison meteorite. Nature 228, 923–926. Lawless J. G., Kvenvolden K. A., Peterson E., Ponnamperuma C., and Moore C. (1971) Amino acids indigenous to the Murray meteorite. Science 173, 626–627. Kvenvolden K. A., Lawless J. G., and Ponnamperuma C. (1971) Nonprotein amino acids in the Murchison meteorite. Proc. Nat. Acad. Sci. (USA) 68, 486–490. MacDermott A. J. (1996) Looking for the homochiral signature of life: the SETH cigar. In Chemical Evolution: Physics of the Origin and Evolution of Life (ed. J. Chela-Flores and F. Raulin) pp. 373–379. Kluwer. Mason S. F. (1988) Biomolecular homochirality. Chem. Soc. Rev. 17, 347–359. Mason S. F. (1997) Extraterrestrial handedness. Nature 389, 804. Meinschein W. G., Frondel C., Laur P., and Mislow K. (1966) Mete- orites: Optical activity in organic matter. Science 154, 377–380. Mueller G. (1953) The properties and theory of genesis of the carbo- naceous complex within the Cold Bokevelt meteorite. Geochim. Cosmochim. Acta 4, 1–10. Nagy B. (1965) Optical activity in the Orgueil meteorite. Science 150, 1846. Nagy B. (1966) The optical rotation of lipids extracted from soils, sediments, and the Orgueil carbonaceous meteorite. Proc. Nat. Acad. Sci. (USA) 56, 389–398. Nagy B., Murphy M. T. J., Modzeleski V. E., Rouser G., Claus G., Hennessy D. J., Colombo U., and Gazzarini F. (1964) Nature 202, 228–233. Peltzer E. T. and Bada J. L. (1978) a-Hydroxycarboxylic acids in the Murchison meteorite. Nature 272, 443–444. Pizzarello S. and Cronin J. R. (1998) Alanine enantiomers in the Murchison meteorite. Nature 394, 236. Pollock G. E., Chang C.-N., Cronin S. E., and Kvenvolden K. A. (1975) Stereoisomers of isovaline in the Murchison meteorite. Geochim. Cosmochim. Acta 39, 1571–1573. Roberts J. A. (1984) Supernovae and life. Nature 308, 318. Rubenstein E., Bonner W. A., Noyes H. P., and Brown G. S. (1983) Supernovae and life. Nature 306, 118. 338 S. Pizzarello and J. R. Cronin