Estimation of paleotemperature from racemization of aspartic acid in combination with radiocarbon age

Estimation of paleotemperature from racemization of aspartic acid in combination with radiocarbon age

NIM B Beam Interactions with Materials & Atoms Nuclear Instruments and Methods in Physics Research B 259 (2007) 547–551 www.elsevier.com/locate/nimb ...

148KB Sizes 3 Downloads 73 Views

NIM B Beam Interactions with Materials & Atoms

Nuclear Instruments and Methods in Physics Research B 259 (2007) 547–551 www.elsevier.com/locate/nimb

Estimation of paleotemperature from racemization of aspartic acid in combination with radiocarbon age Masayo Minami

a,*

, Masami Takeyama b, Koichi Mimura c, Toshio Nakamura

a

a Center for Chronological Research, Nagoya University, Nagoya 464-8602, Japan Department of Earth and Planetary Sciences, School of Science, Nagoya 464-8602, Japan c Graduate School of Environmental Studies, Nagoya University, Nagoya 464-8602, Japan

b

Available online 1 February 2007

Abstract We tried to estimate paleotemperatures from two chosen fossils by measuring D/L aspartic acid ratios and radiocarbon ages of the XAD-2-treated hydrolysate fractions in the fossils. The D/L aspartic acid ratio was measured with a gas chromatograph and radiocarbon dating was performed using a Tandetron AMS system at Nagoya University. The radiocarbon age of a fossil mammoth molar collected from Bykovsky Peninsula, eastern Siberia, was found to be 35,170 ± 300 BP as an average value for the XAD-treated hydrolysate fractions. The aspartic acid in the mammoth molar showed a little evidence of racemization, which might be due to in vivo racemization during the lifetime and then suggests negligible or no postmortem racemization during burial in permafrost. From four animal bone fossils collected from a shell mound excavated at the Awazu submarine archeological site in Lake Biwa, Shiga, Japan, the racemization-based effective mean temperature was calculated to be 15–16 C using the D/L aspartic acid ratio of about 0.11 and the 14C age of 4500 BP for the XAD-2-treated hydrolysate fractions in the fossils. The average annual temperature was estimated to be 11– 12 C, which approximates to the temperature that the fossils experienced during burial at the site. Although the application of racemization ratios in fossils as paleotemperature indicators is surrounded with many difficulties, the results obtained in this study suggest its feasibility.  2007 Elsevier B.V. All rights reserved. PACS: 91.65.Dt; 91.65.Hy; 91/70.Bf Keywords: Fossil bone; Aspartic acid; Racemization;

14

C age; XAD-2 resin

1. Introduction Amino acids in proteins of living organisms consist only of L-enantiomers. Over long periods, these L-amino acids undergo slow racemization, producing the corresponding D-amino acids. Fossil materials, such as fossil bones and teeth, contain both L- and D-amino acids and the D/L amino acid ratio increases with the age of the fossil [1]. For example, aspartic acid, which has the fastest racemization rates of stable amino acids, has a half-life of 3500 years at 25 C [2]. The age of fossils, therefore, can be estimated by measuring the D/L amino acid ratio. This method has *

Corresponding author. Tel.: +81 52 789 3091; fax: +81 52 789 3092. E-mail address: [email protected] (M. Minami).

0168-583X/$ - see front matter  2007 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2007.01.201

been applied to various fossil materials, including deepsea sediments, shells, teeth, bones and wood (e.g. [3–7]). Amino acids racemize in accordance with the following equation: k1

l-amino acid

!

d-amino acid

ð1Þ

k2

Because racemization is a reversible first-order reaction, the speed of disappearance of L-amino acids in Eq. (1) can be expressed as: 

dL ¼ k1L  k2D dt

ð2Þ

where L and D are mole concentrations of L- and D-amino acids in the fossil sample, respectively; k1 and k2 are the

548

M. Minami et al. / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 547–551

first-order rate constants for the interconversion of the Land D-enantiomers for a particular amino acid; and t is the age of the sample. Because Dt=0  Lt=0 when t = 0 Lt¼0 ffi L þ D:

ð3Þ

Thus: 

dL ¼ ðk 1 þ k 2 ÞL  k 2 Lt¼0 : dt

By integration of this equation:   1 þ ðD=LÞ ln ¼ ð1 þ KÞk 2 t þ C 1  KðD=LÞ

ð4Þ

ð5Þ

where hK = k2/ki1, and C is the integral constant 1þðD=LÞ C ¼ ln 1KðD=LÞ . Because k1 = k2 = k in racemizat¼0 :

tion, Eq. (5) can be expressed:   1 þ ðD=LÞ ln ¼ 2kt þ C: 1  ðD=LÞ

ð6Þ

The (D/L)t=0 value in Eq. (6) is the racemization rate during hydrolysis of collagen. Little racemization of amino acids takes place during hydrolysis of collagen. The (D/ L)t=0 value is estimated to be 0.07 for aspartic acid during hydrolysis of collagen with 6 M HCl for 24 h [8]. Therefore, Eq. (6) is expressed:   1 þ ðD=LÞ ln  0:14 ¼ 2 kt: ð7Þ 1  ðD=LÞ If the rate of the reaction in Eqs. (5) and (6) is constant, the relationship between time t and the D/L amino acid ratio would follow the predicted reversible first-order expression. However, racemization is a chemical reaction and its rate depends, for example, on temperature, pH, presence of mineralization and amino acid composition (e.g. [9–12]). In particular, the amino acid racemization rate depends mostly on temperature [13]. The value of k, according to the Arrhenius equation, is expressed as follows: k ¼ A expðEa =RT Þ

ð8Þ

where A is the Arrhenius constant, Ea is activation energy, R is the gas constant and T is the temperature in degrees Kelvin. The equation can be evaluated by using a rate constant previously determined on bone or teeth from an environment of known and relatively constant temperature and a measured value of Ea. It is assumed, therefore, that the exponential or effective mean temperature to which a sample has been exposed can be estimated by measuring the D/L amino acid ratio in a sample that has previously been dated by an alternative method, such as radiocarbon dating. The Arrhenius Eq. (8) can be expressed:     k asp ðtoothÞ Ea 1 1 ln  ð9Þ ¼ k asp ðfossilÞ R T calc T tooth where Ea = 33.4 kcal/mol, kasp(tooth) = 8 · 104 yr1, kasp(fossil) = the racemization rate of a sample, Ttooth = 310 K, Tcalc = exponential or effective mean temperature,

R = 1.987 cal/(K mol) [14]. This equation is derived from measurements of aspartic acid racemization in the teeth of living human beings. Estimation of paleotemperature is of primary importance in studies of palaeoenvironnment and the racemization reaction of amino acids in fossils has been used as a paleotemperature indicator (e.g. [15–18]). However, racemization is dependent on the chemical state of amino acids and thus the rate of amino acid racemization in fossils varies with its chemical composition. It is reported, for example, that the amino acid racemization is greatly retarded in anhydrous environments [19] and that free amino acids in bone or teeth are more highly racemized than are bound (acid-insoluble) amino acids [5,20]. Due to the effect of the chemical state of amino acids on amino acid racemization rate, the use of amino acid racemization as a paleotemperature indicator is surrounded with difficulty. In this study, we tested the feasibility of using amino acid racemization of two chosen fossils to estimate paleotemperature by measuring their D/L aspartic acid ratios and radiocarbon ages. Gelatin and XAD-treated hydrolysate fractions in a fossil were both investigated to assess fluctuation of the rate of aspartic acid racemization in the two fractions, which are generally used for measuring accurate radiocarbon ages of fossil bones. We employed the same fossil bone fractions for the measurement of 14C ages and aspartic acid racemization rates. 2. Samples Fossil samples used in this study were a mammoth elephant molar and animal bone fragments. The mammoth elephant molar was collected from the Bykovsky Peninsula, which is in the delta region of the Lena River in eastern Siberia. In this area, there is thick permafrost and the frozen soil, known as ‘‘edoma’’. The molar was recovered from a peat bed at 21.3 m depth and its 14C age had previously been measured to be 34,280 ± 820 BP [21]. The animal bone fossils were collected from a shell mound excavated at the Awazu submarine archeological site in the southern basin of Lake Biwa, Shiga, Japan. The Awazu archeological site is buried by 2–3 m depth of bottom sediments. We previously obtained 14C ages of about 4500 BP for the samples using XAD-2 chromatography [22]. We also used modern collagen standards of bovine Achilles tendon (Sigma and Nacalai) and L-aspartic acid (Kishida) for comparison of racemization rates on hydrolysis. 3. Experimental method 3.1. Sample pre-treatment and

14

C dating

Samples were ultrasonicated repeatedly in distilled water and then in 0.2 M NaOH and rinsed with distilled water. The lyophilized bone powder was decalcified with 0.6 M HCl at 4 C. A part of the acid-insoluble fraction (decalci-

M. Minami et al. / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 547–551

The gelatin was hydrolyzed with 6 M HCl at 110 C and its gelatin hydrolysate and the XAD-treated hydrolysate fractions were derivatized to trifluoroacetic isopropyl ester in a two-step procedure. The dried sample was esterified in isopropanol and acetyl chloride for 1 h at 100 C. The reaction was then quenched and the amino acid isopropyl ester evaporated to dryness under a stream of N2 at 0 C. The ester was acylated for 10 min at 100 C by addition of trifluoroacetic anhydride and dichloromethane. After acylation, the reagent was removed by evaporation under a stream of N2 at 0 C. The amino acid derivative was dissolved in dichloromethane prior to undergoing gas chromatography.

Gelatin (This study) XAD-treated hydrolysate (This study) Nagaoka et al. (1994) Nagaoka et al. (1994)

42000

40000

38000

C age (BP)

3.2. Measurement of D/L aspartic acid ratio

21&, which is consistent with the 23& to 21& ranges for the d 13C value in collagen of herbivorous animals [24]. The gelatin was dated 35,500–36,500 BP for dentin and 37,000 BP for cementum and root parts, while the XADtreated hydrolysate fraction was dated to about 35,000 BP. The age of the XAD-treated hydrolysate is similar to the 34,250 ± 850 BP age previously obtained for the gelatin for the total parts of the same molar sample and the 35,800 BP age determined on another mammoth molar sample collected from the same edoma layer [21] (Fig. 1). Generally, the XAD-2 chromatography method is more effective in obtaining reliable 14C ages than the gelatin extraction method for badly preserved fossil bones with a collagen yield of less than 1%. There is no difference in the 14C ages obtained by either method for well-preserved fossils [22]. The molar sample used in this study is well preserved and the 14C age of the XAD-treated hydrolysate could be reliable. If so, the apparent age of the gelatin extracted from the cementum and the root of the molar could be older than the actual age of the tooth. This is possibly due to contamination by foreign organic carbon of

36000

14

fied bone) was treated with 0.1 M NaOH, repeatedly rinsed with HCl followed by distilled water and then lyophilized. By heating the decalcified bone in H2O (pH3) for 16 h at 90 C in a test tube, the soluble fraction (gelatin) was extracted. The other part of the acid-insoluble fraction was hydrolyzed with 6 M HCl at 110 C to decompose proteins into amino acids. The hydrolysate was passed through the XAD-2 resin and amino acids were extracted with 6 M HCl. The XAD-treated hydrolysate was rotary evaporated and lyophilized. The gelatin and the XAD-treated fractions were then heated to 850 C in a sealed Vycor tube and the extracted gas was purified, retaining only CO2. The CO2 was sealed in a Vycor tube with H2 gas and heated to 650 C. The graphite obtained was pressed into a target and the 14C content measured by Tandetron AMS at the Center for Chronological Research, Nagoya University, Japan.

549

34000

4. Results and discussion 32000

4.1.

14

C age of the fossil Siberian mammoth molar

The results for the fossil Siberian mammoth molar are shown in Table 1. The C/N ratio was in the range 3.2 ± 0.5, which agrees with the C/N ratio of the collagen [23]. The d 13C value for the fossil molar was about Table 1 CO2 yield, C/N atomic ratio, d

13

C value,

3.2

Cementum

GC-1 GC-2 XAD-1 XAD-2

3.3

GC XAD

3.3 3.0

Root

3.2

3.0

Root

Total

Another mammoth

Fig. 1. 14C ages for the mammoth fossil collected from edoma on Bykovsky Peninsula, eastern Siberia.

C age and D/L ratio of a fossil molar of a mammoth, from Bykovsky, eastern Siberia 14

Lab. code# (NUTA2-)

D/L ratio

20.5 20.5 20.5 20.3

35,980 ± 430 35,410 ± 420 36,410 ± 440 35,050 ± 390

6441 6442 6443 6444

0.088

21.0 21.2 20.7 20.9

37,270 ± 490 36,890 ± 470 35,510 ± 410 34,940 ± 380

6433 6435 6434 6436

0.082

21.5 21.1

37,290 ± 480 –

6445 –

0.069 0.057

d

GC-1 GC-2 GC-3 XAD

Dentin Cementum

14

C/N ratio Dentin

30000

13

CPDB (&)

C age (BP)

0.077

0.073

550

M. Minami et al. / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 547–551

older 14C age from the surrounding ice, which was not completely removed through alkali-treatment and gelatin extraction. The dentin is covered with the cement and thus may not have been contaminated enough to result in an older apparent age. Therefore, the age of the molar is estimated to be 35,170 ± 300 BP, the average value for the XAD-treated hydrolysate fractions. 4.2. D/L aspartic acid ratio of the fossil Siberian mammoth molar The D/L aspartic acid ratios of two collagen standards and the L-aspartic acid reagent are shown in Table 2. The racemization ratio on hydrolysis of the collagen standards was equal to the value of 0.07 obtained by Bada et al. [8]. The racemization ratio on hydrolysis of L-aspartic acid reagent was 0.011, which is much lower than Bada’s value. When a peptide bond is formed from a gathering of amino acids, the racemization rate of the amino acids is generally increased [11]. Therefore, the lower D/L ratio of the L-aspartic acid reagent could be due to the lack of a peptide bond. The D/L ratios of gelatin were about 0.011 higher than those of XAD-treated fractions of the fossil mammoth molar (Table 1). The gelatinization is performed under temperature and time conditions similar to those of hydrolysis and therefore the same degree of amino acid racemization might occur as the L-aspartic acid reagent was applied. The D/L ratio for the XAD-treated fraction in dentin was a little higher than the racemization ratio of the modern hydrolyzed collagen standards, 0.070–0.072, while that in cementum was almost identical. Since collagen in dentin is metabolically stable and not resynthesized, the D/L aspartic acid ratio observed in fossil dentin is the sum of in vivo racemization during the lifetime and the postmortem racemization. A little higher D/L ratio in dentin of the mammoth molar might be due to in vivo racemization during the lifetime of the mammoth, suggesting negligible or no postmortem racemization during burial in the permafrost. The D/L ratio of the root part of the molar was less than 0.07. Gillard et al. [25] showed that the D/L aspartic acid variations exist in whole or vertically sectioned dentin and root dentin and that local variations within teeth can be relatively large. Amino acid racemization is dependent on pH, presence of mineralization and amino acid composition, in addition to the environmental temperature and thus this low racemization for the root part of the molar Table 2 D/L aspartic acid ratios of collagen standard and L-aspartic acid reagent D/L

ratio

might be due to a difference in chemical state between the root and other tooth parts. 4.3. Animal bone fragments The D/L aspartic acid ratios of fossil animal bones were 0.108–0.119, which means that the amino acids have racemized since they were buried (Table 3). The kasp(fossil) is calculated from Eq. (7) to be 6.19 · 106  1.14 · 105 yr1 and using Eq. (9), the racemization-based mean temperature of 13.0–13.6 C is calculated. On the above calculation, only one kasp(tooth) value at 37 C of 8 · 104 yr1 is used. Different values, however, have been reported for kasp(tooth) in total amino acid fraction of dentin at 37 C, such as 6.06 · 104 yr1 [26] and 6.05 · 104 yr1 [27]. Ohtani and Yamamoto [20] reported that the rate constant was 6.4 · 104 yr1 in total amino acid fraction in dentin, 5.1 · 104 yr1 in an acid-insoluble collagen fraction and 1.88 · 103 yr1 in a soluble peptide fraction. The racemization rate is dependent on the chemical state of amino acids and the rate is slightly lower in acid-insoluble collagen fractions than in total amino acid fractions. The XAD-treated hydrolysates are the fractions obtained by hydrolysis and then purification with XAD-2 resin of acid-insoluble collagen. Therefore, the racemization rate in the acid-insoluble collagen fraction should be used in this study. When the value of 5.1 · 104 yr1 is used for kasp(tooth) in the Eq. (9), an effective mean temperature of 15.2–15.8 C is calculated. Amino acid racemization responds exponentially to temperature, so a racemization-based effective temperature Te does not correspond to a simple average environmental temperature Ta, but to a higher value than Te. Lee [28] provided a theoretical relationship for calculating Ta as function of Te and the annual temperature range. Using his equation with the annual temperature range of 20 C, Ta is estimated to be 11–12 C. The Awazu submarine archeological site flourished through the Jomon period, but no fossil bones were discovered there after that period. The animal bones could have Table 3 14 C age, D/L aspartic acid ratio, rate constant of aspartic acid enantiomer and racemization-based temperature to which four animal bone fragments were exposed Sample AWA-8 AWA-10 AWA-11 AWA-12 1

Collagen standard Sigma Nacalai

0.070 0.072

L-aspartic acid reagent Non-treated Hydrolysis

0.000 0.011

14 C age (BP)1

D/L

4530 ± 80 4360 ± 70 4570 ± 100 4500 ± 80

0.112 0.119 0.109 0.114

kasp

Te (C)2

Ta (C)3

9.38 · 106 1.14 · 105 8.63 · 106 9.89 · 106

15.64 (13.4)5 16.5 (14.3) 15.2 (13.0) 15.8 (13.6)

12 13 12 12

ratio

Data from Minami and Nakamura [22]. Racemization-based effective mean temperature. 3 Average annual temperature estimated using the equation of Lee [28] and the racemization-based effective mean temperature when 5.1 · 104 yr1is adopted for kasp(tooth). 4 Calculation using the value of 5.1 · 104 yr1 for kasp(tooth) [20]. 5 Calculation using the value of 8 · 104 yr1for kasp(tooth) [14]. 2

M. Minami et al. / Nucl. Instr. and Meth. in Phys. Res. B 259 (2007) 547–551

sunken into Lake Biwa not long after the animals’ death. The Awazu submarine archeological site is at 2–3 m depth under the bottom of Lake Biwa and the temperature of the sediments that buried the site may have been close to or slightly lower than the bottom water temperature. Lake Biwa has not experienced large environmental change during the last 4500 year. Therefore, the average annual temperature that the bones experienced could be close to or slightly lower than that of the water in Lake Biwa. The temperature of the lake bottom water in the southern basin of Lake Biwa ranges between 7 and 28 C during the year and its average value is about 15 C [29], which roughly agrees with the average annual temperature obtained from the animal bone fragments. 5. Conclusion The fossil mammoth molar collected from eastern Siberia showed little evidence of amino acid racemization. The fossil has been preserved at temperatures too low for amino acid racemization to proceed. The D/L ratios of gelatin fractions were higher than those of XAD-treated fractions, because the amino acid racemization might proceed during gelatinization, which is performed at a temperature of 90 C. Therefore, XAD-treated hydrolysate fractions in fossils are useful in estimating paleotemperature when measuring both the D/L aspartic acid ratios and 14C ages. The results from animal bone fossils collected from the Awazu submarine archeological site show that racemization of aspartic acid and radiocarbon age in fossils could be useful in estimation of paleotemperature, though it is surrounded with many difficulties. Acknowledgements We are grateful to Dr. D. Nagaoka (Graduate School of Environmental Earth Science, Hokkaido University) and Mr. I. Iba (the Board of Education of Shiga Prefecture) for providing fossil samples. This work was supported in part by a Grant-in-Aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports,

551

Science and Technology (MEXT) of Japan (subject #15068206). References [1] P.E. Hare, P.H. Abelson, Carnegie Inst. Wash. Yearb. 66 (1968) 526. [2] J.L. Bada, M.-Y. Shou, in: P.E. Hare, T.C. Hoering, K. King Jr., (Eds.), Biogeochemistry of Amino Acids, 1980. [3] J.L. Bada, B.P. Luyendyk, J.B. Maynard, Science 170 (1970) 730. [4] M.-Y. Shou, P.F. Karrow, L.J. Bada, Geology 8 (1980) 200. [5] S. Matsu’ura, N. Ueda, Nature 286 (1980) 883. [6] S. Ritz, A. Turzynski, H.W. Schu¨tz, Forensic Sci. Int. 69 (1994) 149. [7] S. Ohtani, Y. Matsushima, Y. Kobayashi, K. Kishi, J. Forensic Sci. 43 (1998) 949. [8] J.L. Bada, P.M. Masters, E. Hoopes, D. Darling, in: R. Berger, H.E. Suess (Eds.), Radiocarbon Dating, University of California Press, 1979. [9] J.L. Bada, J. Am. Chem. Soc. 94 (1972) 1371. [10] N. Kriausakul, R.M. Mitterer, Science 201 (1978) 1011. [11] S. Steinberg, J.L. Bada, Science 213 (1981) 544. [12] G.C. Smith, K.M. Williams, D.M. Wonnacott, J. Org. Chem. 43 (1978) 1. [13] J.L. Bada, R.A. Schroeder, Naturwissenschaften 62 (1975) 71. [14] J.L. Bada, Earth Planet Sci. Lett. 55 (1981) 292. [15] J.L. Bada, R. Protsch, R.A. Schroeder, Nature 24 (1973) 394. [16] J.L. Bischoff, W.M. Childers, Earth Planet. Sci. Lett. 45 (1979) 172. [17] D.S. Kaufman, Quat. Sci. Rev. 22 (2003) 899. [18] Y. Takano, J. Kudo, T. Kaneko, K. Kobayashi, Y. Ishikawa, K. Marumo, Bull. Chem. Soc. Jpn. 77 (2004) 1029. [19] J.L. Bada, X.S. Wang, H.N. Poinar, S. Pa¨a¨bo, G.O. Poinar, Geochem. Cosmochim. Acta 58 (1994) 3131. [20] S. Ohtani, K. Yamamoto, J. Forensic Sci. 36 (1991) 792. [21] D. Nagaoka, T. Sone, T. Nakamura, M. Fukuda, T. Najayama, V. Kunitsky, Summaries of Researches Using AMS at Nagoya University, V (1994) 197 (in Japanese, with English abstract). [22] M. Minami, T. Nakamura, Nucl. Instr. and Meth. B 172 (2000) 462. [23] P.E. Hare, D. von Endt, Annual Report of Director of the Geophysical Laboratory, Carnegie Inst. Wash, 1989–1990, Geophysical Laboratory, Washington DC, 1990, 115. [24] M. Minagawa, S. Egawa, Y. Kabaya, K. Karasawa-Tsuru, Mass Spectr. 40 (1992) 47. [25] R.D. Gillard, A.M. Pollard, P.A. Sutton, D.K. Whittaker, Archaeometry 32 (1) (1990) 61. [26] A. Shimoyama, K. Harada, Chem. Lett. 13 (1984) 1661. [27] T. Ogino, H. Ogino, J. Dental Res. 67 (1988) 1319. [28] R. Lee, J. Appl. Meteorol. 8 (1969) 423. [29] Shiga prefecture board education, fourth ed., The natural of Lake Biwa, 1997.