ISOLEUCINE EPIMERIZATION IN THE HIGH-MOLECULAR-WEIGHT FRACTION OF PLEISTOCENE ARCTICA

The extent of amino acid racemization, as traditionally determined in the entire (total acid hydrolysate) pool of amino acids comprising the organic remains of fossils, is a function of the integrated effects o f a complex diagenetic reaction network. We investigated the possibili ty that some of the complications involved in protein diagenesis might be circumvented by isolating one component of the reaction network an d studying the extent of racemization in that fraction alone. We used gel-filtration to extract the high-molecular-weight (HMW) fraction of proteinaceous matter from fossil and modem molluscan shells. This frac tion contains the largest (ca. > 15,000 MW), most-pristine macromolecu les and has been less affected by diagenesis than the more-degraded, l ower molecular-weight fractions. Variations in the extent of racemizat ion (isoleucine epimerization; aIle/Ile) measured in the HMW fraction of subsamples taken along cross sections of Arctica shells from two in terglacial sites, Bo and Fjosanger, southwestern Norway, are within th e range of analytical uncertainty [coefficient of variation (cv) = 5-8 %], despite the strong gradient (cv = 20-24%) in aIle/Ile of the total amino acid population. Because there is no age difference across a sh ell, this finding supports the idea that the HMW fraction contains mor e geochronologically reliable proteinaceous matter than the total amin o acid pool. Weighted mean aIle/Ile ratios in the HMW fraction of aliq uots of powdered sample from the two shells overlap at +/-1 sigma, des pite significantly different aIle/Ile ratios in the total amino acid p opulation of some shells from the two sites. The difference in aIle/Il e ratios in the total population is attributed to a greater proportion of low-molecular-weight (ca. 300 MW), and hence, extensively epimeriz ed molecules measured in gel-filtered samples from the Fjosanger shell . Because the rate of epimerization in the HMW fraction is much lower than in the total population, the temporal resolution of the HMW techn ique is limited, particularly at these high-latitude sites. Therefore, we cannot use the aIle/Ile HMW data to exclude the possibility that t he two sites are significantly different ages. Analyses of shells rang ing in age from late Pliocene to Holocene indicate that reaction rate in the HMW fraction is about one-fifth the rate in the total amino aci d population, although the difference is expected to decrease with inc reasing aIle/Ile.