Article pubs.acs.org/est
Chlorine Isotope Effects and Composition of Naturally Produced Organochlorines from Chloroperoxidases, Flavin-Dependent Halogenases, and in Forest Soil Christoph Aeppli,†,⊥,* David Bastviken,‡ Per Andersson,§ and Ö rjan Gustafsson†,∥ †
Department of Applied Environmental Science (ITM), Stockholm University, Sweden Department of Thematic Studies, Water and Environmental Studies, Linköping University, Sweden § Laboratory for Isotope Geology (LIG), Swedish Museum of Natural History, Stockholm, Sweden ∥ The Bert Bolin Centre for Climate Research, Stockholm University, Sweden ‡
S Supporting Information *
ABSTRACT: The use of stable chlorine isotopic signatures (δ37Cl) of organochlorine compounds has been suggested as a tool to determine both their origins and transformations in the environment. Here we investigated the δ37Cl fractionation of two important pathways for enzymatic natural halogenation: chlorination by chloroperoxidase (CPO) and flavin-dependent halogenases (FDH). Phenolic products of CPO were highly 37Cl depleted (δ37Cl = −12.6 ± 0.9‰); significantly more depleted than all known industrially produced organochlorine compounds (δ37Cl = −7 to +6‰). In contrast, four FDH products did not exhibit any observable isotopic shifts (δ37Cl = −0.3 ± 0.6‰). We attributed the different isotopic effect to the distinctly different chlorination mechanisms employed by the two enzymes. Furthermore, the δ37Cl in bulk organochlorines extracted from boreal forest soils were only slightly depleted in 37Cl relative to inorganic Cl. In contrast to previous suggestions that CPO plays a key role in production of soil organochlorines, this observation points to the additional involvement of either other chlorination pathways, or that dechlorination of naturally produced organochlorines can neutralize δ37Cl shifts caused by CPO chlorination. Overall, this study demonstrates that chlorine isotopic signatures are highly useful to understand sources and cycling of organochlorines in nature. Furthermore, this study presents δ37Cl values of FDH products as well of bulk organochlorines extracted from pristine forest soil for the first time.
■
INTRODUCTION Chlorinated organic compounds (organochlorines) of both industrial and natural origin are prevalent in the environment. Organochlorines tend to display high toxicity, persistence, and bioaccumulation potential. Besides the many well-known industrial sources of organochlorines, there is an enormous pool of naturally produced compounds in the environment, with over 2200 different organochlorines identified so far.1,2 A large number of studies have reported their environmental distribution. These compounds, for example, were detected in soils,3,4 seawater and marine sediments,5,6 marine troposphere air,7,8 and along the marine food chain up to marine mammals.9,10 Soil systems are important for production and accumulation of naturally produced organochlorines.11,12 Although abiotic formation ofmostly volatilecompounds is possible,4,13 microbial sources are thought to be the most relevant contributor to nonvolatile natural organochlorines in soils.3 For example, >20% of inorganic chlorine (Cli) added to boreal forest soil samples were later recovered as organically bound chlorine (Clorg) in incubation experiments, while a sterile © 2013 American Chemical Society
control did not lead to significant amounts of organochlorines.14−16 This result suggests substantial ongoing chlorination of organic matter and turnover of chlorine in soils. Although the exact mechanism of such microbial Clorg formation is not yet fully understood, it is hypothesized to be mainly catalyzed by chloroperoxidases (CPOs) or flavindependent halogenases (FDHs).17,18 These enzymes catalyze the oxidation of chloride which forms a reactive species to chlorinate electron-rich moieties of soil organic matter. The resulting products can closely resembleor even have the same chemical structuresas industrially produced organochlorines such as chlorinated phenols, phenoxy phenols, or dibenzo-p-dioxins. Special Issue: Rene Schwarzenbach Tribute Received: Revised: Accepted: Published: 6864
September 17, 2012 January 2, 2013 January 15, 2013 January 15, 2013 dx.doi.org/10.1021/es3037669 | Environ. Sci. Technol. 2013, 47, 6864−6871
Environmental Science & Technology
Article
The aim of this study was, therefore, to investigate chlorine isotope effects and composition of naturally produced organochlorines. Specifically, we aimed to assess chlorine isotopic fractionation associated with two main classes of chlorinating enzymes, CPOs and FDHs. These are thought to be responsible for most natural organochlorine production in soil systems. We then measured δ37Cl of organochlorines extracted from pristine boreal forest soils to explore the use of δ37Cl for source apportionment between different natural and industrial sources.
Many organochlorines detected in the environment affect human or environmental health. In order to assess the potential to curb their industrial emissions, it is important to know natural and anthropogenic organochlorine sources. However, there are often large uncertainties in natural sources and sink terms of these compounds. Therefore, tools to differentiate between compounds of natural and anthropogenic origin are necessary. Isotopes are an excellent tool to apportion natural and anthropogenic sources. Using the radiocarbon (14C) content of compounds has proven a very successful method to this end.9,19 This approach is based on the fact that the feedstock of industrially produced compounds is largely petroleum-derived, and therefore radiocarbon-free, whereas naturally produced organochlorines are made from carbon with contemporary radiocarbon levels. However, the radiocarbon method is limited by the sample amount needed for analysis (>25 μg carbon). As environmental samples mostly contain only trace amount of natural organochlorines, the radiocarbon method is often not applicable. As an alternative, stable isotope analysis recently emerged as a promising tool for source apportionment of organochlorines in the environment. For example, approaches using stable carbon isotopes have been used to apportion trichloroacetic acid or chloroform formed in forest soils.20,21 For larger compounds containing many carbon atoms, however, the small δ13C signals introduced during chlorination can be masked due to isotopic dilution. To overcome this limitation, the use of chlorine isotopes has been pioneered. This isotopic system is more suitable than carbon isotopes since usually only a few chlorine atoms are present in a naturally produced organochlorine molecule, therefore limiting the effect of isotopic masking. Furthermore, large chlorine isotope effects are expected to be associated with natural chlorination. This prediction is based on one laboratory study that found CPO catalyzed chlorinated products to be highly depleted in δ37Cl, with δ37Cl values of −12‰ vs standard mean ocean chloride (SMOC).22 This value is different from all known industrially produced end-members such as chlorinated pesticides and persistent organic pollutants, which can span a range of δ37Cl values of −7 to +6‰ vs SMOC.23−25 However, the outcome of this CPO experiment has not yet been confirmed using different substrates. Furthermore, products from the other important chlorinating enzyme class, FDHs, have not yet been investigated, thus currently limiting the application of δ37Cl for organochlorine source apportionment in the environment. There are also only few studies analyzing δ37Cl of Clorg in the environment, and they focused on measurement of organochlorines in animal fat tissue26−28 and in preindustrial sediments.29 Lastly, all of these studies were complicated and restricted to small sample numbers due to analytical challenges, involving labor-intensive preparative gas chromatography (GC) and bulk measurement including off-line conversion prior to δ37Cl analysis.27,29 Recent progresses in δ37Cl measurement using quadrupole MS (qMS) achieved detection limits of well below 1 nmol (10,000 units mL−1). KCl, H2O2, (30% aqueous solution), NaOH, KNO3, AgNO3, CH3I, phenol, 2chlorophenol (2CP), 4-chlorophenol (4CP), 2,4-dichlorophenol (DCP), and 2,4,6-trichlorophenol (TCP) were obtained from Sigma-Aldrich in the highest available quality. High purity of KNO3 (99.999%) was essential due to high chlorine background in lower-purity products. Using the high-purity products, the typical blank Cli concentrations were
