Main

In agreement with earlier analyses2, we find the alanine from the interior of Murchison to be racemic and believe that incomplete chromatographic resolution of l-alanine and other meteoritic amino acids may have affected both the enantiomer ratio and the δ15N determinations of Engel and Macko.

The Murchison meteorite contains a complex suite of amino acids. Whereas terrestrial samples are dominated by the 20 protein amino acids, over 70 amino acids have been positively identified in this meteorite, many of which appear to be uniquely extraterrestrial. Others are present that have not been positively identified. This complexity challenges methods of amino-acid analysis. For example, in our study of enantiomeric ratios of meteoritic amino acids3 we found that we needed a fractionation step before gas-chromatographic analysis in order to avoid resolution problems.

We find that several meteoritic amino acids are racemic, including alanine. An l-alanine excess was seen only in an extract of a sample that included exterior fragments. Figure 1 shows the chromatographic resolution of alanine from this extract, in the same chromatographic phase (Chirasil-l-valine) as that used by Engel and Macko1. There are five amino acids that elute over a period of about a minute, from just before l-alanine to sarcosine. Although only one of these amino acids overlaps l-alanine in our chromatogram, in the separation shown in ref. 1 the chromatogram seems to be compressed and baseline separation is not quite achieved between l-alanine and sarcosine. Thus it seems probable that more than one of these five amino acids eluted or overlapped with l-alanine, contributing to the appearance of an l-alanine excess. It is unlikely that these amino acids were absent in Engel and Macko's sample as different Murchison stones are qualitatively similar with respect to amino-acid content4.

Figure 1: Total ion chromatogram (Chirasil-l-Val, 50 m×025 mm, Alltech) of amino acids (N -trifluoroacetyl isopropyl esters) from a desalted Murchison water extract.
figure 1

The dotted line is a single ion (m/z 180) trace. Amino acids were identified from mass spectra. 1, D-alanine; 2, 2-amino-2-ethyl butanoic acid; 3, l-alanine; 4, methyl proline isomer or unsaturated acyclic C6 amino acid; 5, DL-2-methyl norvaline; 6, DL- N -methyl alanine; 7, unsaturated C4 amino acid; 8, sarcosine.

Full size image

We do not know the δ15N values of the amino acids that we suspect may be interfering but, if their δ15N values were close to that of indigenous l-alanine, their presenceould go undetected isotopically. On the other hand, if their δ15N values were greater than that of l-alanine, their presence could mask the presence of terrestrial l-alanine. Given Murchison amino-acid values of up to +184‰, it is clear that their chromatographic overlap, even if they were present in relatively small amounts, would increase the 15N content of the l-alanine peak and compensate, wholly or partially, for the lower 15N content of terrestrial l-alanine that might have been present.

In conclusion, Engel and Macko's argument that their chromatographic peak for l-alanine is composed almost entirely of meteoritic l-alanine depends on the peak containing only l-alanine. Their argument for a large excess of the l-enantiomer in the alanine indigenous to their Murchison sample would be greatly strengthened if they were to provide evidence that l-alanine and the closely eluting amino acids that we have identified were completely resolved in their gas chromatography/isotope-ratio mass spectrometry analyses. Such analyses of appropriate amino-acid standards might also help to establish whether or not incomplete resolution affected their result.