Bulletin of Glaciological Research ,, (,**/) 03ῌ1. 69 ῍Japanese Society of Snow and Ice Determination of amino acids in ice samples Fumio NAKAZAWA+, Keiichi OHTA,, Naomi HARADA-, Koji FUJITA. and Masayoshi NAKAWO/ + Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya .0.ῌ20*+, Japan , Department of Ecosystem Studies, School of Environmental Science, University of Shiga Prefecture, ,/** Hassaka-cho, Hikone, Shiga Pref. /,,ῌ2/--, Japan - Japan Agency for Marine-Earth Science and Technology, ,ῌ+/ Natsushima-Cho, Yokosuka-City, Kanagawa ,-0ῌ**0+, Japan . Graduate School of Environmental Studies, Nagoya University, Furo-cho, Chikusa-ku, Nagoya .0.ῌ20*+, Japan / Research Institute for Humanity and Nature, --/ Takashima-cho, Marutamachi-dori Kawaramachi nishi-iru, Kamigyo-ku, Kyoto 0*,ῌ*212, Japan (Received October ., ,**.; Revised manuscript received November -*, ,**.) Abstract This study investigates amino acid concentrations and compositions in ice as a preliminary approach for applying amino acid analysis to ice core study. For the amino acid analysis, *..,/ to *.1/* kg ice samples are used. The ices from Antarctica have +0.2 and +2.- nmol kgῌ+ of amino acids and these concentrations are close to laboratory blank levels, while the ice from the Chongce Ice Cap in China has +0* nmol kgῌ+ of amino acids. Therefore, the analysis for Antarctic ice is estimated to need more than several kg of ice. In contrast, the analysis from the Chongce Ice Cap should be reducible to only several hundred grams of ice because the concentrations in this cap are higher than those in Antarctica. Further, amino acid sources are present in glacier surroundings of temperate regions, whereas polar regions are far from sources. Serine, aspartic acid and alanine are the major components in the samples. However, the amino acid compositions show variation with each sample. seem to originate mainly from microorganisms, such +.Introduction as snow algae and bacteria, that breed on glacier surfaces in summer season (Kohshima, +321, +323; Ionic species, such as inorganic ions and light Ling and Seppelt, +33*, +33-; Yoshimura et al., +331; carboxylic acids, in ice cores have provided important Takeuchi et al., ,**+), plant products, such as pollen, in information on past climatic and environmental con- glacier surroundings (Koerner et al., +322; Bourgeois, ditions. The chemical species yield data on historical +33*, ,***;Liuet al., +332; Reese et al., ,**-; Nakazawa events including biogenic emission, biomass burning, et al., ,**.) and the bacteria in precipitation (Herlihy et anthropogenic emission and volcanic eruption (e.g., al., +321; Sattler et al., ,**+; Mace et al., ,**-). Thus, the Hammer, +311; Hammer et al., +32*; Legrand and amino acids may have good environmental informa- Angelis, +330; Olivier et al., ,**-). However, there are tion in ice core study. Since these biological activities some other candidates for the source tracers in cases increase in summer season and decrease in winter, when interpretation is di$cult. For example, natural amino acid-rich layers in an ice core may indicate events, such as volcanic eruptions and large forest summer layers. In addition, amino acid concentrations fires, increase concentrations of chloride, nitrate and in each summer layer may be useful as a good marker sulfate ions in the atmosphere. In addition, anthropo- of meteorological conditions such as temperature and genic e#ects such as biomass burning and fossil fuel solar radiation that a#ect these activities. However, combustion can also increase the atmospheric concen- no previous study has reported the amino acids in trations of these constituents. glacier ice. Therefore, as a preliminary approach to ice In contrast, amino acids that build proteins appar- core research, this study aims to survey the level of ently originate from and are abundant in living organ- amino acid concentrations in ice from the Antarctic isms, for example, approximately /*ῌ of the human ice sheet and Chongce Ice Cap in central Asia, and to body’s dry weight is protein (McMurry and Castellion, compare their compositions. +333). Moreover, amino acids contained in glacier ices 70 Bulletin of Glaciological Research Table +. Detection system for high-performance liquid chromatography (HPLC). HPLC Shimadzu LC-+*AD Detector Fluorescence detector RF-+*AXL (-/*ῌ.0* nm) Integrator Shimadzu chromatopac GR1Ae plus Column GL Science Intersil ODS-- (+/* mm L, ..0 mm ID, / mm particle size) Guard column GL Science Intersil ODS-- (/* mm L, ..0 mm ID, / mm particle size) Column temperature .*῎ Eluent E+: / mM Na, HPO.ῌ+,H, O// mM NaH, PO.ῌ,H, O(pH1.*/) E,:E+/CH- CN( ,: +, v/v) E-:CH- CN/Distilled water (HPLC grade) (.: +, v/v) Flow-rate +.* mL minῌ+ Table ,. Working gradient conditions used for amino ,. Sampling sites and sample preparation acid determination. Time (mim) E+ E, E- Ice core sampling was conducted at two distinct Program + Program , (῍) sites; Asuka camp in East Antarctica (1+῍-+῎S, ,.῍2῎E; *4*+ *4*+ 3-* m a.s.l.) and Chongce Ice Cap in the West Kunlun 3/ / * .4* .4* 2/ +/ * Mountains, China (-/῍+.῎N, 2+῍*1῎E, 0-,1 m a.s.l.). Ice 24* 24* 2* ,* * blocks were taken from the surface of an iceberg o# +04* +04* 1- ,1 * Syowa Station (03῍**῎S, -3῍-/῎E) in the Antarctic +24* +24* 1* -* * Ocean. One section was collected from each ice core; ,/4* ,14* // ./ * +** 2 from the Asuka core, the increment between . and -*4* -,4* /* /* * +*+ - . m depth, while from the Chongce core, the incre- .,4* ..4* -/ 0/ * , 2 - * ment between . and . m depth were used. For the .,4*+ ..4*+ -* 1* * amino acid analysis, *.1/*, *./** and *..,/ kg ice blocks ./4* .14* ,/ 1/ * were used for the Asuka ice, the iceberg ice and the /+4* /-4* ,* 2* * Chongce ice, respectively. /+4*+ /-4*+ * +** * To prevent contamination by laboratory controls, /,4* /.4* * * +** precleaned glasswares, which include filters, were 2.4* 3*4* * +** * used in all laboratory work. The glasswares were 2.4*+ 3*4*+ +** * * prepared by washing in alkaline detergent, placing in 2.4*, (stop) 3*4*, (stop) +** * * +NHCl overnight followed by rinsing with organic- The program + was used for the particulate and free water obtained with a Milli-Q system and then dissolved fractions of the blank sample, the particulate heating in a furnace at ./*῎ for - to . h. Additionally, fraction of the Asuka sample, and the particulate , to eliminate contamination of each sample, about + cm fraction of the Chongce sample. The program was used for the dissolved fraction of the Asuka sample, of the surface was scraped o# with a clean knife in a the particulate and dissolved fractions of the iceberg cold room, and then the sample was melted in a bea- sample, and the dissolved fraction of the Chongce ker. The first melted ice used to wash the surface of sample. the ice samples was discarded. This procedure was repeated - to / times, depending on the sample. The melted ice discarded in this way came to *.+./ to *.++* acid content of the ices. kg. After the ice was completely melted, the melt A laboratory procedural blank were also ana- water was immediately filtrated through .1 mm GF/F lyzed. For particulate amino acids, a new GF/F glass glass fiber filters with a pore size of *.1 mm. fiber filter (diameter: .1 mm; pore size: *.1 mm) was hy- Subsequently, the samples were separated into drolyzed and the following procedure was the same as two portions, a filter sample and filtrate sample. These that for samples. In contrast, for dissolved amino ac- 0 sub-samples (for amino acid analysis) were hydro- ids, .-ml of the organic-free water was concentrated in lyzed with 0NHCl at +*/ ῎ for ,, h. The HCl super- a flask under vacuum, and passed through the cation natant was then transferred to a flask and dried under exchange resin. The following procedure was the vacuum. After drying, the dried residue was desalted same as mentioned above. using cation exchange resin (Dowex /*W-X2), derivat- A detection system for HPLC and working gradi- ized with o-phthalaldehyde and then analyzed by gra- ent conditions used for determination of amino acid dient reversed-phase high-performance liquid chro- are summarized in Table + and , respectively. Twelve matography (HPLC) with UV fluorescence detection amino acids were measured in this study (Table -). in order to determine the total hydrolyzable amino The yield for the purification of amino acids was 3/ to Nakazawa et al. 71 Table -. Concentration of amino acids in the ice (nmol kgῌ+). Abbreviation: Glu (glutamic acid), Ala (alanine), Val (valine), Gly (glycine), Ile (isoleucine), Ser (serine), Thr (threonine), Arg (arginine), Asp (aspartic acid), Met (methionine), Phe (phenylalanine), Tyr (tyrosine). Sample Type sample Glu Ala Val Gly Ile Ser Thr Arg Asp Met Phe Tyr Total Blank Particulate metter +.0 +.- +.+ *.0 -.3 ,.+ +./ *./ +./ +.. n.i. *.3 +0 Dissolved matter *.. *./ n.i. *., n.i. *.. ,.* n.i. +.. *.- n.i. n.i. /., Asuka Particulate metter +.0 +.0 +.2 *.+ n.i. -.* +., +.+ *.3 *./ n.i. n.i. +,.* Dissolved matter *.+ *.0 *.- *., *.0 +.* *.2 *., *.3 n.i. n.i. *.+ ..2 Iceberg Particulate metter *.. +., *.- *.. +., *.2 +.* n.i. *.3 n.i. n.i. *.+ 0.. Dissolved matter +.2 +.+ *.1 *./ +.. +.3 +./ *., +.. +., n.i. *., ++.3 Chongce Particulate metter -* ,. +/ ,.1 1.3 ,* +1 /./ ,* -.- ,.3 -.* +/, Dissolved matter *./ *.3 +.1 *.. n.i. *.3 *.3 *.0 *.3 *./ n.i. *., 1./ The n.i. is “not identified”, meaning that it was impossible to identify the peak of amino acid because amounts were below the detection limit or inseparable from adjacent peaks. +**ῌ (Harada, +33+). The precisions (RSD, n῍,) were acid concentrations in Antarctic ice should be gener- +- to +ῌ except for valine (Val), isoleucine (Ile), me- ally low. thionine (Met), phenylalanine (Phe) and tyrosine (Tyr). The Chongce sample shows that Glu, Ala, Ser and The precisions of the five amino acids were not deter- Asp are the major components of the particulate frac- mined. tion and comprise 0,ῌ of the total. In contrast, Val, Ala, Ser, Thr and Asp are the major components of the -.