Environ. Sci. Technol. 2005, 39, 2115-2119
Isomer-Selective Adsorption of Amino Acids by Components of Natural Sediments M. WEDYAN AND M. R. PRESTON* Department of Earth and Ocean Sciences, University of Liverpool, 4 Brownlow Street, Liverpool L69 3GP, United Kingdom
We present evidence that under circumstances of low pH and organic-free surfaces an ordinary estuarine sediment can exhibit strong optical isomer selectivity in its absorption of a number of amino acids. This selectivity can also be seen to a lesser degree in the minerals quartz, montmorillonite, and kaolin. Adsorption reactions were performed with racemic amino acid mixtures, and after equilibrium, deviations from a D/L ratio of 1 were measured and in many cases were found to be significant. This was particularly pronounced at pH 4.0, where selective removal of the L isomers by adsorption onto sediment fractions was almost total. Changes in both the nature and degree of selectivity were also observable in different sediment size fractions. While we are at this stage unable to identify the mode of primary selectivity, adsorption experiments with these candidate sediment components, quartz, kaolin, and montmorillonite do exhibit some selective behavior. We believe that the existence of natural chirally selective components in sediment may indicate a new approach to the development of chiral catalysis and synthesis.
Introduction Chirally selective adsorption processes are a major focus of much current research with numerous applications in both laboratory and manufacturing processes (1). In general it is necessary to spend considerable effort on the design of very specific chiral systems, though some recent studies have suggested that even some achiral surfaces can exhibit selectivity, at least at small scales (1-3). For example, the adsorption of organic right-handed (R,R)-tartaric acid onto a copper surface leads to the formation of a chiral template on it (4). Can such systems exist naturally or are they confined to highly specific experimental conditions? Recent research has tended to address the second of these types of system, whereas this paper addresses the issue for a variety of sediment components and assemblages. The most obvious and widespread examples of chiral molecules in nature are perhaps the amino acids, and these have been extensively used for the characterization of organic matter in natural systems (5, 6); however, studies have concentrated on total amino acids, and isomer-selective analysis has been rare. While homochirality is the accepted paradigm for most biological systems, the picture is not that simple with, for example, D enantiomers of aspartic acid, glutamic acid, serine, and alanine being present as significant components of peptidoglycan in bacterial cell walls with D * Corresponding author phone: +44-0-151-794-4093; fax: +440-151-794-5196; e-mail:
[email protected]. 10.1021/es040474o CCC: $30.25 Published on Web 02/12/2005
2005 American Chemical Society
enantiomers being created from L counterparts apparently as a defense against proteolytic degradation (7-9). Abiotic synthesis reactions, however, normally yield racemic mixtures (10, 11). A significant challenge in any studies of the origins of homochirality in nature is therefore to identify natural processes that can lead to the selection, concentration, and polymerization of molecules from an initially chiral mixture, and the physical process that initiated chiral selection in biological systems remains a challenging problem in understanding the origin of life (12). As long ago as 1974 it was recognized that the adsorption of D- and L-alanine from dimethylformamide solution onto D- and L-quartz crystals was selective with D-alanine adsorbing preferentially on D-quartz and vice versa (13). More recently, a number of workers (e.g., 2, 14-17) have suggested that minerals such as calcite and quartz may have some chirally selective properties. Questions remain as to whether natural materials assemblages are net “isomer-blind” at the macro scale or whether inherent selectivity could have led to the homochirality that is evident in biological systems. The work described in this paper examines the adsorption of amino acids onto quartz, clay minerals, and some natural sediments and presents evidence that such systems are not isomer-blind but rather show a variety of chirally selective adsorptive effects. We conclude that this may be derived either from intrinsic mineral selectivity or from the effects of chiral “contamination” of modern systems by fossil chiral influences as suggested by Hazen (personal communication) as a potential criticism of the recent work by Shinitzky et al. (12).
Materials and Methods The analytical method used was modified from one developed recently for the determination of amino acid enantiomers in fossils (18). This employs precolumn derivatization of amino acids with o-phthaldialdehyde (OPA) and N-isobutyryl-Lcysteine (IBLC) to yield fluorescent diastereomeric derivatives of primary amino acids. IBLC was obtained from Fluka. All amino acid standards and OPA were obtained from Sigma Chemical Co. HPLC-grade water and buffers were obtained from BDH. Kaolin and montmorillonite (Aldrich) and quartz (Sigma) were used in the adsorption experiments. The amino acids analyzed were aspartic acid (Asp), glutamic acid (Glu), serine (Ser), arginine (Arg), valine (Val), methionine (Met), phenylalanine (Phe), leucine (Leu), and isoleucine (Ile). An Agilent 1100 Series high-performance liquid chromatography (HPLC) system equipped with a quaternary pump, vacuum degasser, autoinjector/autosampler, and fluorescence detector (λex ) 330 nm and λem ) 445 nm) was used for all analyses. HP Chemostation computer software was employed to perform the integration and calibration of the fluorescence signals and also to control solvent gradients and the automation of the sample derivatization and injection program. The derivatization of DL-amino acids was performed online with the automated injector (18). OPA/IBLC derivatizing reagent (2 µL; 260 mM IBLC and 170 mM OPA in 1 M potassium borate buffer, pH 10.4, adjusted with potassium hydroxide pellets), 1M potassium borate buffer (2µL; pH 10.4) and sample (1 µL) were mixed, during which the reaction takes place at room temperature, and injected into the HPLC system after a reaction time with a complete derivatization cycle of 5 min. HPLC separations were performed on a reversed-phase column (Hypersil BDS, 5 µm, 250 × 4 mm Thermoelectronic). VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2115
TABLE 1. HPLC Solvent Elution Program time (min)
%A
%B
%C
0 2 31 85 95 105 120
95 95 76.6 0 0 95 95
5 5 23.0 95 95 5 5
0 0 0.4 5 5 0 0
The mobile phase consisted of three solvents: acetate buffer (solvent A, 23 mM sodium acetate adjusted to pH 6.00 with 10% acetic acid and sodium hydroxide, freshly prepared weekly), methanol (solvent B), and acetonitrile (solvent C). The modified solvent program from ref 18 is given in Table 1. The column was maintained at 25 °C. Amino acids were quantified by reference to an internal standard (L-homoarginine). The system was found to be linear in calibration and stable in response factors over the concentration range 1-20 pmol‚L-1 for each amino acid analyzed. The analytical repeatability of D/L ratios was obtained by multiple injections of amino acid standards and ranged from 0.16% for alanine to 0.75% for serine. To maintain a high repeatability of D/L ratios, standard solutions were run alongside samples and response factors calculated for each batch of samples (normally 1 standard/10 samples). Sediments were obtained from the intertidal zone in the River Mersey Estuary with a hand-driven core tube. Sediments in this region are predominantly fine-grained muds and silts with a small proportion of coarser sand. The top centimeter of sediments is oxygenated, but deeper sediments are suboxic or anoxic. One core was sectioned into 1 cm slices and each slice was subsectioned. Pore waters were extracted by centrifugation and analyzed for their amino acid content as described below. Samples (unfiltered pore water) were prepared for total hydrolyzable amino acid (THAA) analysis by a modified version of the method of Cowie et al. (19). With an automatic pipet, 900 µL of sample was pipetted into a clean Pyrex culture tube; to this was added 100 µL of internal standard (LHomarg), followed by 1 mL of 50% hydrochloric acid solution (50 mL of acid and 50 mL of water). Nitrogen was then bubbled through the sample for 1 min prior to sealing with a Teflon-faced screw-top lid to prevent oxidation of the amino acids. The sealed tubes were then hydrolyzed at 110 °C for 6 h and subsequently freeze-dried. The freeze-dried hydrolyzed amino acid samples were then redissolved in a minimum amount of Milli-Q water divided into equal aliquots and were either analyzed immediately or stored frozen (-20 °C). Adsorption experiments were performed on wet sediment, freeze-dried sediment, and ashed (450 °C, 15 h) sediment. Kaolin and montmorillonite were similarly ashed before use. Ashed sediment was also size-fractionated by dry sieving into four fractions: >500, 250-500, 125-250, and
