OUP user menu

Microbial reductive dehalogenation of polychlorinated biphenyls

Juergen Wiegel , Qingzhong Wu
DOI: http://dx.doi.org/10.1111/j.1574-6941.2000.tb00693.x 1-15 First published online: 1 April 2000

Abstract

Under anaerobic conditions, microbial reductive dechlorination of polychlorinated biphenyls (PCBs) occurs in soils and aquatic sediments. In contrast to dechlorination of supplemented single congeners for which frequently ortho dechlorination has been observed, reductive dechlorination mainly attacks meta and/or para chlorines of PCB mixtures in contaminated sediments, although in a few instances ortho dechlorination of PCBs has been observed. Different microorganisms appear to be responsible for different dechlorination activities and the occurrence of various dehalogenation routes. No axenic cultures of an anaerobic microorganism have been obtained so far. Most probable number determinations indicate that the addition of PCB congeners, as potential electron acceptors, stimulates the growth of PCB-dechlorinating microorganisms. A few PCB-dechlorinating enrichment cultures have been obtained and partially characterized. Temperature, pH, availability of naturally occurring or of supplemented carbon sources, and the presence or absence of H2 or other electron donors and competing electron acceptors influence the dechlorination rate, extent and route of PCB dechlorination. We conclude from the sum of the experimental data that these factors influence apparently the composition of the active microbial community and thus the routes, the rates and the extent of the dehalogenation. The observed effects are due to the specificity of the dehalogenating bacteria which become active as well as changing interactions between the dehalogenating and non-dehalogenating bacteria. Important interactions include the induced changes in the formation and utilization of H2 by non-dechlorinating and dechlorinating bacteria, competition for substrates and other electron donors and acceptors, and changes in the formation of acidic fermentation products by heterotrophic and autotrophic acidogenic bacteria leading to changes in the pH of the sediments.

Keywords
  • Reductive dehalogenation
  • Polychlorinated biphenyl
  • Microbial
  • Anaerobic
  • Environmental factor
  • Aquatic sediment

1 Introduction

Polychlorinated biphenyls (PCBs) are composed of two benzene rings linked at the C-1 carbon. They are substituted in a great variety of combinations, with 1–10 chlorines in the ortho, meta and/or para positions (see Fig. 1 for nomenclature) leading to 209 possible compounds, referred to as PCB congeners. When PCBs are categorized by the degree of chlorination, the term homolog is used, e.g. the trichlorobiphenyl homologs. PCBs of a given homolog with different chlorine substitution positions are called isomers. The individual chlorine substitutions are further classified into singly or doubly flanked and unflanked chlorines, referring to whether the neighboring positions carry chlorine substitutions or not, respectively. Until banned in the USA in 1978, PCBs were commercially manufactured as complex mixtures, each of which contains 60–90 PCB congeners [1]. Trade names for these mixtures include Aroclor and Pyroclor (Monsanto, USA), Clophen (Bayer, Germany), Fenclor (Caffaro, Italy), Fenoclor (S.A. Cros, Spain), Kanechlor (Kanegafuchi, Japan) and Phenoclor and Pyralène (Prodelec, France) [2]. PCBs were widely used in industry because of their excellent chemical and physical properties including low vapor pressures, low water solubility, excellent dielectric properties, stability to oxidation, flame resistance and relative inertness [2]. The extensive use of PCBs has resulted in widespread contamination of air, water, soil and sediments [3]. For example, one-third of the total USA production of PCBs (approximately 1.4×109 lb) has been released into the environment through various venues by deliberate or accidental discharges [4].

1

Nomenclature for PCBs. The 12 carbon positions in the biphenyl are numbered using carbon 1 for the phenyl–phenyl bond and then numbers 2–6 for the first ring and 2′–6′ for the second ring. Thus, there are four ortho (i.e. 2-, 2′-, 6- and 6′-), four meta (i.e. 3-, 3′-, 5- and 5′-) and two para (i.e. 4- and 4′-) positions. Short notations used for halogen-substituted biphenyls use either the notation pictured above, giving the substituted position in order of increasing numbers and using the prime notations for the second ring (e.g. for a given hexa-chlorobiphenyl=2,2′,3,3′,5,6′-CP), or give first the positions of chlorines on the most chlorinated ring followed by the chlorinated positions on the second ring, separated by a hyphen without using the primes (e.g. 2,3,5-2,3,6-CP or 235-236-CP). Bromo-substituted biphenyls (BB) are numbered the same way.

Although both the production for industrial use and the discharge of PCBs have been banned in the USA since 1978, contamination with PCBs still occurs and is of great public concern due to bioaccumulation and potential toxicity to humans and wildlife [3, 5]. New contamination can still occur through damage to old PCB-containing transformers by lightning and by disturbance of PCB-containing sludge and sediments. PCBs are ubiquitous pollutants. The recent use of more effective extraction and quantification methods (e.g. [68]) led to finding more and more organisms and environments being contaminated with PCBs, e.g. seaweeds and ice from Antarctica [9]. Because of their hydrophobicity, major portions of PCBs released into the aquatic environment are eventually expected to end up either adsorbed onto sediment (humic acids) or resting as sludge (oil–water mixtures) on the bottom of rivers, lakes and oceans. PCBs have been found adsorbed on marine microorganisms, the ratio of adsorbed and taken up PCB congeners depended on the structure and lipids of the cells and the hydrophobicity of the congeners [10]. Through mechanical disturbances including sediment dwelling organisms, wind and rain water [11, 12], PCBs subsequently enter into food chains ([3, 6] and literature cited therein) and finally end up in the tissue of nearly all terrestrial and marine plants and animals, fish, mammals, fish- and non-fish-eating birds (e.g. barn owls; [13]), and humans [3], where it, through the effect of bioaccumulation/biomagnification, can reach harmful concentrations (e.g. [14]). The uptake of and physiological reactions against PCBs by animals have even led to the proposal to use animals as bioindicators for detecting PCBs in aqueous environments (e.g. eels; [15]). PCBs are generally regarded as stable compounds which are not degraded easily. Biodegradation in the environment, although slow, has been demonstrated. Although recently plant-associated dehalogenation has been proposed (Meager et al., personal communication), PCBs are mainly biodegraded by two distinct microbial processes: (i) aerobic oxidative degradation and (ii) anaerobic reductive dechlorination.

The aerobic degradation process has been studied extensively ([1620] and literature cited therein). Aerobic degradation involves the oxidative destruction of the PCB molecule through a series of intermediates and can commonly be carried out by a single or by two, possibly synergistically acting, aerobes. However, the aerobic degradation of PCBs usually attacks only lightly chlorinated congeners, i.e. biphenyls with five or fewer chlorines. Furthermore, in many instances, only the top few millimeters of sediments are aerobic. Therefore, the largest reservoirs of PCBs in rivers and lakes are the anaerobic sediments, which are not suitable for the growth of aerobic microorganisms. Effective microbial transformation of PCBs in these contaminated sediments can only occur through anaerobic processes, i.e. reductive dehalogenation.

Anaerobic reductive dechlorination involves the removal of the chlorine substituents as halogen ions and their replacement by hydrogen in the form of electrons and protons. Reductive dechlorination is an important process in biodegradation of various halogenated aliphatic and aromatic compounds (for reviews of reductive dechlorination, see [2123]). Anaerobic dechlorination can attack a large array of highly chlorinated PCBs including decachlorobiphenyl ([24]; unpublished results). This reductive dechlorination of highly chlorinated PCBs decreases their toxicity and thus increases their degradability, e.g. by converting the ‘co-planar’, dioxin-like congeners into congeners with fewer chlorines [2528]. Based on the published literature, PCB dechlorination is widespread in the environment although the rate, extent and specificity of microbial dechlorination vary widely, even within the same sediment. Experimental evidence indicates that environmental factors such as temperature, pH, partial pressure of H2, and the presence or absence of utilizable carbon sources, electron donors and electron acceptors affect the dechlorination. After a brief review of anaerobic dechlorination in a variety of sediments and an overview of observed dechlorination processes, this mini-review will primarily explore the importance of these environmental factors as elucidated by studies involving microcosms (also called sediment cultures) and PCB-chlorinating enrichment cultures. Finally, field studies and environmental applications for anaerobic PCB dechlorination will be briefly discussed. For a more extensive listing, see the reviews of Abramowicz [16], Abramowicz and Olson [29], and Bedard et al. [30].

2 Anaerobic dechlorination of PCBs

Anaerobic dechlorination of PCBs was first reported by Brown et al [3133]. They observed in situ dechlorination of PCBs based on the altered distribution of PCB congener patterns in the sediments compared with those of the commercial PCB mixtures discharged at the investigated site. Subsequently, Quensen et al. [34] confirmed microbial PCB dechlorination in the laboratory using anaerobic sediment slurries from the Hudson River. Since then, anaerobic dechlorination of PCBs has been reported from other laboratories ([30] and literature cited therein, [3548]). Evidence to date demonstrates that microbial PCB dechlorination is widespread in many anaerobic environments, including freshwater (pond, lake and river), estuarine and marine sediments ([29, 30]Table 1). In the laboratory, dechlorination activity of PCBs in uncontaminated sediments was also detected after PCBs were added to the sediments (Table 1). Presumably due to environmental selections, microbial communities in PCB-contaminated sites are better adapted for the PCB dechlorination than those from environments containing no or only traces of PCBs [42, 49]. For example, a higher dechlorination rate (46 versus 16 μmol l−1 day−1) and a higher extent of dechlorination (1.6+0.1 versus 2.1 total chlorines per biphenyl remaining after 1 year of incubation) were observed for the dechlorination of the supplemented tetrachlorobiphenyl 2,3,4,6-CB to trichlorobiphenyls in microcosms from a contaminated Woods Pond (MA, USA) grab sample versus from a non-contaminated site (a small pond in a wooded area of the Sandy Creek Nature Park, Athens, GA, USA) [42]. PCB-dechlorinating organisms are apparently only present in low numbers in PCB-non-contaminated sites, as is also suggested by the observed longer lag times before dechlorination was observed [42, 49].

View this table:
1

Example of sites containing microorganisms capable of anaerobic PCB dehalogenationa

Location
PCB-contaminated sites
Otanabee River/Rice LakeCanada
Salt MarshGA, USA
Rhine RiverGermany
Lake ShinjiiJapan
Lake KetelmeerThe Netherlands
Rhine RiverThe Netherlands
Escambia BayFL, USA
Waukegan HarborIL, USA
Chesapeake BayMD, USA
Hoosic RiverMA, USA
Housatonic RiverMA, USA
New Bedford HarborMA, USA
Silver Lake and Woods PondMA, USA
Kalamazoo RiverMI, USA
Grasse RiverNY, USA
Hudson RiverNY, USA
St. Lawrence RiverNY, USA
Lake HartwellSC, USA
Fox River/Green BayWI, USA
Sheboygan RiverWI, USA
PCB-uncontaminated sites
Sandy Creek Nature CenterGA, USA
Center PondMA, USA
Red Cedar RiverMI, USA
Saline RiverMI, USA
Hudson River (Spiers Falls)NY, USA
  • aModified table from cited references in [29, 30].

2.1 Dechlorination patterns and processes

Commercial PCBs such as Aroclor 1242, 1248, 1254, 1260 and 1268 were produced by catalytic chlorination of biphenyl to a specified wt% chlorine during the manufacturing process [50]. The individual PCB congeners vary greatly in their susceptibility to evaporation, adsorption, solubilization, microbial dechlorination and other alteration processes that may occur in specific environments or small niches [3, 30]. As a result, the distribution pattern of PCB congeners in an environmental sample will exhibit a record of all alteration processes to which that sample has been exposed [51]. To investigate and characterize the potential of given sites for PCB dehalogenation, relatively large (>100 ml) microcosms prepared in triplicate from pooled samples need to be examined. Grab samples from one site can be very heterogeneous in respect to both the amount and congener distribution of the hydrophobic PCB adhering to the sediment and the microbial dehalogenation activity.

Brown and co-workers [3133, 51] described various patterns of PCB dechlorination for both environmental and laboratory samples and gave each a letter designation. There are at least eight distinct microbial dechlorination processes (Processes M, Q, H′, H, P, N, LP and T) which can be identified through careful comparison of the patterns of congener loss and product formation ([30, 44, 5254], Table 2, Fig. 2). Other processes can be explained as combinations of these eight (for detailed discussion, see [30]). Microbiologically mediated dechlorination of PCBs typically removes meta and/or para chlorines to generate primarily ortho-substituted mono- through tetrachlorobiphenyls, though biphenyl has been observed in dechlorination of 4-4-dichlorobiphenyl (4-4-CB) [55], 345-CB [56], 34-34-CB [57] and 23456-CB [58]. Also ortho dechlorination of several individual PCB congeners (e.g. 24-CB, 246-CB and 2356-CB) has been reported [37, 43, 56, 59, 60]. As an example of various dehalogenation routes of a PCB congener under anaerobic conditions, Fig. 2 depicts the different observed routes for 2,3,4,6-CB in sediments samples from Woods Pond of the Housatonic River (MA, USA). It should be stressed that not all the depicted dehalogenations were observed throughout the investigated temperature range (between 4 and 60°C) or at all pH values between 5.0 and 7.5. Thus, temperature is one of the important environmental factors determining what kind of dechlorination occurs in a given environment (see below). Fig. 3 shows as an example the Process N dechlorination followed by the Process LP (also called Process UP) in Woods Pond sediment samples. Process N mainly removes chlorines in flanked meta positions (i.e. with a neighboring chlorine in ortho and (=doubly flanked)/or (=single flanked) in para position). Recently, microbial ortho dechlorination of Aroclor 1260 was demonstrated in anaerobic slurries of estuarine sediments [44]. After 6 months of incubation, extensive meta dechlorination (45–65% decrease in meta chlorines per biphenyl) and moderate ortho dechlorination (9–12% decrease in ortho chlorines per biphenyl) of Aroclor 1260 were observed in samples amended with 800 parts per million (ppm) Aroclor 1260 with and without the addition of 350 μM 2345-CB or 2356-CB. The ortho dechlorination activity mainly removed flanked ortho chlorines, i.e. from PCBs with 235- and 2356-chlorophenyl groups. The presence of relatively smaller amounts (14%) of 235- and 2356-chloro-substituted congeners in Aroclor 1260 may explain why only moderate ortho dechlorination of Aroclor 1260 was observed.

View this table:
2

Positions of chlorines removed by each dechlorination processa

Dechlorination processSusceptible chlorines
MFlanked and unflanked meta
QFlanked and unflanked para, meta of 23-group
H′Flanked para, meta of 23- and 234-groups
HFlanked para, doubly flanked meta
PFlanked para
NFlanked meta
LPFlanked and unflanked para
TFlanked meta of 2345-group, in hepta- and octachlorobiphenyls
  • aModified table from [30].

2

Temperature-dependent routes of microbial reductive dechlorination of spiked 2346-CB in Woods Pond sediment samples (Lenox, MA, USA) incubated between 4 and 66°C.

3

Major dechlorination reactions for Process N and LP as observed, e.g., in incubations with Woods Pond sediment at 25°C and pH 6.0–7.5. At temperatures below 18°C, Process LP is only a minor reaction.

2.2 Factors influencing PCB dechlorination

It has been hypothesized [30, 35, 43, 5254, 61] that a variety of microorganisms with distinct dehalogenating enzymes, each exhibiting a unique pattern of congener selectivity, are responsible for the various patterns of PCB dechlorination observed both in the laboratory and in the environment. Environmental factors and conditions affect the growth and the variety of metabolic activities of different microorganisms differently and hence influence divergently the extent and rate of the various PCB-dechlorinating activities. Consequently, a better knowledge of whether and to what extent individual environmental factors can influence PCB dechlorination is important for obtaining an understanding of the diversity of PCB dehalogenation and the conditions under which a particular PCB dehalogenation pattern can or cannot occur. This knowledge will help in predicting the potential for PCB dechlorination in a given environment and will aid in developing bioremediation schemes. Below, environmental factors are discussed which have been shown to influence PCB dechlorination and include especially temperature, pH, available carbon source, H2 as electron donor and the presence or absence of electron acceptors other than PCBs. So far, no extensive and systematic studies have been done on the influence of different types of soil/sediment (including with different PCB adsorption properties) on the reductive dehalogenation. Also yet to be undertaken are extensive studies of (expected) changes in community structures during enrichments and a comparison of them in PCB-contaminated and non-contaminated sediments showing dehalogenation activity from the same site or from different sites but containing the same soil/sediment type.

A major factor in the biotransformation and biodegradation plays the bioavailability of PCBs, i.e. the adsorption and desorption processes. The rates of desorption are different for the different PCB congeners; the most hydrophobic congeners (with high Kow values) behaved roughly as predicted from models, whereas the less hydrophobic ones with low Kow values desorbed up to four magnitudes slower than models predicted [62]. Unfortunately, this subject of bioavailability including the so called process of aging of PCB contaminations, which leads over time to stronger adsorption and lower recovery rates of PCBs, cannot be discussed here in detail due to its complexity and the space limitation of the mini-review.

2.2.1 Temperature

Besides the effect of temperature on the bioavailability as well as on transport between sites (due to the influence of temperature on the heats of surface–air exchanges; [63, 120]), temperature has a significant effect on the growth and the physiological activity including uptake and enzymatic dehalogenation of PCB congeners. However, most laboratory studies of microbial PCB dechlorination reported to date have been conducted at room temperature, i.e. around 25°C. PCB-contaminated sediments in the environment typically experience a different and much wider range of temperatures. The range of temperatures depends on the climate and on the depth of the water and the sediment itself. The influence of temperature is multifacetted. Effects include changes in the adsorption and desorption kinetics of PCBs from soil particles [64] and thus both the hydrolytic (abiotic) dehalogenation and the availability of PCBs for microbial transformations. However, these effects are probably minor in comparison with the effect of temperature on the growth of microorganisms and the catalytic activity of enzymes. Wu et al. investigated the dechlorination of added 2346-CB and residual Aroclor 1260 in Woods Pond sediment (Housatonic River, Lenox, MA, USA, collected in May 1991 and stored at around 7°C) incubated at 18 temperature points from 4 to 66°C [4244]. The incubations were maintained at the desired temperatures with a variation of less than ±1°C. Whereas the residual Aroclor 1260 had been only marginally dehalogenated in the environment with no indications that a significant dechlorination was still occurring, in the microcosms incubated with 350 μM 2346-CB as primer, dechlorination occurred at 8–34°C and at thermophilic temperatures of 50–60°C. The optimal temperatures for overall chlorine removal from 2346-CB and from residual PCBs were 18–30°C and 20–27°C, respectively. Between 8 and 34°C and 50–60°C, flanked meta dechlorination occurred whereas flanked para dechlorination was observed only between 18 and 34°C. However, dechlorination of doubly unflanked para chlorines occurred only in the temperature range of 18–30°C, whereas unflanked ortho dechlorination of 246-CB and 24-CB was observed at 8–30°C. These data indicate that temperatures have a profound influence on the rate, extent and products of PCB dechlorination. Taking into account the typical average summer temperatures for Woods Pond, which range from about 15°C at 45-cm depth to between 18 and 20°C at a depth of 10–15 cm, and the winter temperatures for these depths being between 1 and 4°C, one can conclude that during about half of the year no measurable dechlorination should occur. However, in the summer months, the temperature at a 10–15-cm depth is in the range at which the modified N-dechlorination pattern changed to the simple N-pattern, i.e. unflanked (lonely) para chlorine substituents will not be removed by this dechlorination pattern. This is in agreement with the distribution of PCB congeners in Woods Pond sediment. It is interesting that at the prevalent summer temperature of ∼18°C, the highest variability between repeated experiments and among the triplicate incubations was observed. Furthermore, the highest dechlorination rates were observed at much higher temperatures than the normal summer temperatures in the upper 45 cm of the sediment. The latter observation is not total surprisingly since similar observations have been made with other systems. Thus, incubations at room temperatures (around 20–25°C) would not provide a correct and probably therefore not an applicable estimation of the dehalogenation potentials in such sediments. Importantly, without priming with a congener [44] or Aroclor 1260 [65] at elevated concentrations (above 500 ppm), no dehalogenation of the residual PCBs in Woods Pond sediment occurred at any of the incubation temperatures. This repeatedly observed effect is consistent with the lack of continuing in situ dehalogenation at measurable rates in most areas of Woods Pond. Apparently elevated concentrations are required to ‘activate’ the dehalogenation system in the bacteria involved. This study was the first detailed one on the effect of temperature on the occurrence of PCB dehalogenation patterns. Since two different systems were analyzed (uncontaminated pond at Sandy Creek Nature Park in Athens, GA, USA, and the contaminated Woods Pond in Lenox, MA, USA) and similar effects were observed for chlorophenol dehalogenation, we speculate that such effects may also be observed for PCB dehalogenation in sediments from other locations [42, 43]. Fish's [66] observations on temperature effects on the dehalogenation of Aroclor 1242 in Hudson River sediments are in agreement with this statement. Tiedje et al. [67] reported that in an experiment with Hudson River sediment, samples incubated at 12, 25, 37, 45 and 60°C exhibited reductive dechlorination of Aroclor 1242 at 12°C and (roughly twice as fast) at 25°C but not at 37°C or above.

In contrast to temperature-dependence studies in the laboratory where samples were incubated at constant temperatures, sediments in nature are subjected to temperature shifts. In many environments, the temperature fluctuates not only on a seasonal basis, but also from day to night and as a result of strong rain or hot spells, although diurnal fluctuations are minimal in most water sediments covered with several feet of water. Presently, it is unknown how much or whether at all seasonal temperature changes the PCB dechlorination pathways after they have been induced in a given sediment. Wu and Wiegel [68] simulated seasonal temperature changes to some extent by introducing temperature downshifts and upshifts and investigated the effects of these temperature shifts on the dechlorination of 2346-CB added as a primer and of residual PCBs in Woods Pond sediment. These experiments showed that the temperature at the time of priming (initiation) of PCB dechlorination and not the subsequent incubation temperature was the primary determinant of which of the dechlorination processes occurred. Furthermore, these dechlorination processes were more or less retained during subsequent up- and downshifts of the incubation temperature. Changes in the incubation temperature, however, strongly influenced the rate and extent of dehalogenation. These observations explain the differences observed between the dechlorination in the field study in Woods Pond [69] and the above described microcosms studies at various temperatures. The observations are congruent with the hypothesis that the observed dechlorination patterns are due to different populations which are established in response to the different temperatures at the time of priming and which change only slowly with subsequent changes in the incubation temperatures [35, 43, 44, 5254, 61].

2.2.2 pH

Sediments are frequently well-buffered systems, but in contrast to strictly aerobic processes, anaerobic microbial processes may lead to an increase in acidic fermentation products and thus cause local changes in the pH. Like the effects of temperature and carbon sources, the effect of pH on PCB dechlorination in sediments is complex because of various possible interactions between the different dehalogenating and non-dehalogenating microbial populations. Furthermore, pH affects the equilibrium between PCBs that are dissolved and those that are adsorbed to organic matter and thus influences the bioavailability of PCBs in soil [64]. Dechlorination of 2346-CB added as a primer and of residual PCBs in Woods Pond sediment was studied at pH values between 5.0 and 8.0 and at incubation temperatures where changes in the dechlorination processes have been observed, i.e. 15, 18, 25 and 34°C [65]. The pH of each slurry was adjusted and maintained by periodically (2∼10 days) adding sterile anaerobic 2 N NaOH or HCl. Except for 34°C at pH 5.0, at all temperatures, some PCB dechlorination was observed at all pH values examined. The optimal pH for overall removal of chlorines was around 7.0–7.5. However, the stereospecificity of the dechlorination varied: e.g. for 2346-CB, flanked meta dechlorination occurred at pH 5.0–8.0, unflanked para dechlorination at pH 6.0–8.0 and ortho dechlorination at pH 6.0–7.5. However, at pH 7.0 and 15°C, ortho dechlorination dominated whereas at 18 (Fig. 4) and 25°C, unflanked para dehalogenation outpaced the other dehalogenation reactions. These results indicate that the pH of the incubation also strongly influences not only the rate and extent of dechlorination but also the route of dechlorination of the 2346-CB and the residual Aroclor 1260.

4

Example how the incubation of pH influences the prevalence of specific dehalogenation reactions in Woods Pond sediment samples as shown for the environmentally important temperature of 18°C. The relative proportions of the untransformed 246-CB (◻), which is the first dehalogenation product of 2346-CB, and the products of ortho dehalogenation (24-CB and 4-CB; ○) and of unflanked para dehalogenation (26-CB and 2-CB; △) were calculated. The 2-CB was only formed by unflanked para dehalogenation from 24-CB but not from 26-CB because the time courses revealed that 26-CB, 2-CB and 4-CB were not further dehalogenated at a detectable rate within 1 year of incubation (modified from [65]).

2.2.3 Supplementation of carbon sources

Anaerobic PCB dechlorination is a reductive process that presumably uses PCBs as electron acceptors, but does not cleave the rings. Thus, the PCB dechlorinators should need other compounds as sources for carbon and electrons for growth. Until very recently, all research on anaerobic PCB dechlorination was conducted in the presence of sediment which could provide a variety of organic matter. In addition, all experiments were performed with primary microcosms or enrichment cultures containing both PCB dechlorinators and non-PCB dechlorinators. Thus, the addition of a specific carbon source to a culture could enhance PCB dechlorination by providing a desirable carbon source and electron source to the PCB dechlorinators or to non-dechlorinating bacteria which might provide the PCB-dechlorinating microorganisms with growth stimulating factors such as more suitable electron donors, vitamins or carbon sources. Also, the addition could stimulate the utilization of substances that inhibit the PCB dechlorination. On the other hand, such additions could also inhibit PCB dechlorination by supplying a carbon source to non-PCB dechlorinators. These could then out-compete the PCB dechlorinators for electron donors or whose products would be more preferred electron acceptors than PCBs to the dechlorinators and thus inhibit the dehalogenation. Obviously, both effects can occur simultaneously to different dehalogenation populations. All of these possibilities, which are difficult to differentiate in a mixed culture, complicate the interpretation of results obtained when carbon sources have been supplemented.

Alder et al. [70] demonstrated that repeated addition (500 mg l−1 initially and 250 mg l−1 monthly) of fatty acids (acetate, propionate, butyrate and hexanoic acid) stimulated dechlorination of added PCBs in carbon-limited sediment slurries, but not in sediment slurries which had higher organic carbon contents. The addition of 0.1% (v/v) thioglycolate medium with beef extract or acetate to sediment slurries enhanced PCB dechlorination by shortening or eliminating the lag time for the dechlorination of PCBs [71] and increasing the observed overall dechlorination rate [72]. The addition of 0.06% pyruvate and malate substantially increased the extent of PCB dechlorination in Aroclor 1248-contaminated soils [38], whereas in Aroclor 1260-contaminated Woods Pond sediment, malate added together with a primer shortened only the lag time before the onset of the reductive dehalogenation of PCBs [65, 73, 74]. The stimulation of the onset of dechlorination was highly dependent on the incubation temperature and sediment pH; e.g. at pH 7.5, the addition of 10 mM malate to the sediment led to a significantly shortened lag and t50 time for the removal of the first chlorine (mainly 2346-CB to 246-CB) at 15, 18 and 25°C and to an increase in the maximal observed dechlorination rate at 15°C but not at 18–34°C. At pH 6.0, however, malate had no effect on the t50 value nor on the dechlorination rate [65]. Based on later enumeration experiments [75] for priming with 26-bromobiphenyl (26-BB) and the influence of malate on hydrogen utilization [68], we hypothesize that the observed effects are due to stimulation of growth of specific microorganisms by malate only under the specified conditions.

Compared to sediments, enrichment cultures provide more defined conditions. Nies and Vogel [76] found some dependence on the carbon used for the enrichment. Addition of glucose, methanol, acetate or acetone enhanced both the rate and extent of the PCB dechlorination. The addition of 20 mM pyruvate and 10 mM malate enhanced meta dechlorination of 2346-CB by 2346-CB enrichment cultures, while the addition of 20 mM pyruvate increased the rate of para dechlorination of 246-CB and decreased the lag time of dechlorination in 246-CB enrichment cultures [65], indicating again that different types of PCB-dehalogenating microorganisms are present in the sediment and that a definite characterization and understanding of this process will require pure cultures. Two PCB-dechlorinating enrichment consortia were derived from estuarine sediment (Charleston Harbor, SC): a 2356-CB meta dechlorinating enrichment and a 2345-CB para dechlorinating enrichment culture [77]. After the cultures had been transferred eight times into estuarine medium containing only 0.1% (wet w/v) of sediment and 173 μM 2356-CB or 2345-CB, the addition of the potential e-donor formate (10 mM) resulted in an increase in the extent of the 2356-CB dehalogenation to 236-CB from 17% to 75% chlorine substituents removed. Surprisingly, the addition of 10 mM fumarate, an electron acceptor potentially competing with PCB, also increased the extent of dehalogenation. These findings are somewhat contradictory demonstrating that the current understanding of reductive PCB dehalogenation and the interactions with non-dehalogenating microorganisms is indeed poor. This makes it difficult to apply specific laboratory-derived results to in situ processes as long as no pure cultures of reductive PCB dehalogenators are available for control experiments and to study individual responses under well controlled conditions.

2.2.4 Supplementation of H2 as electron donor

Reductive dechlorination is a two electron transfer reaction in which H2 is assumed to be directly or indirectly the electron donor [7880] and water the proton source [81, 82]. Lake sediments contain H2 producers and usually a variety of competing H2 utilizers with different affinities for H2. The successful competition for H2 by a microorganism depends not only on the partial pressure (i.e. availability) of H2 and the affinity of the microorganism for H2 but also on the presence of utilizable carbon sources and electron acceptors. H2, depending on the partial pressure, can stimulate or inhibit this microbial dechlorination process [79, 80, 83, 84]. Wu [85] incubated Aroclor 1260-contaminated sediment from Woods Pond (amended with 2346-CB) at 15, 25 or 34°C, and at pH 6.2 or 7.2 under a frequently replenished gas atmosphere of 0, 1 or 10% H2 (v/v) in O2-free N2 gas and under shaking conditions. Generally, no large differences in the dechlorination rate or extent were observed between samples incubated in a gas atmosphere with or without 1% H2 gas. Assumingly, this amount did not change drastically the available amount of H2 compared to the microbially produced H2. Lower dechlorination rates and a lower extent of dechlorination of 2346-CB and sediment PCBs were found in samples incubated under 10% H2. Higher inhibition was observed at pH 7.2 than at 6.2. 10% H2 in the head gas inhibited or changed the dechlorination reactions of 2346-CB and residual PCBs depending on the incubation temperature and pH. Sokol et al. [86] found that H2 altered the pathway and products of reductive dechlorination of supplemented 234-CB by Hudson River sediment microorganisms. Under H2/CO2, 234-CB was dechlorinated to 24-CB and 23-CB and then 2-CB. Under N2 or N2/CO2, 234-CB was converted to 24-CB only.

2.2.5 Electron acceptor

Reductive dehalogenation of haloaromatic compounds can lead to energy conservation ([21] and literature cited therein, [87]). Some results support the proposal [31, 34, 51] that PCBs are used by the dechlorinating microorganisms as electron acceptors. Kim and Rhee reported [88] that addition of 300 ppm of Aroclor 1248 to PCB enrichment cultures in anaerobic sediment resulted in a 188-fold increase in the number of PCB dechlorinators (from 2.5×105 to 4.6×107 cells per g of sediment). Conversely, the number decreased by 93% from 4.6×105 to 3.1×104 initial value in samples without addition of Aroclor 1248. They concluded that the growth of PCB dechlorinators requires the presence of PCBs. Recently, we observed [75] that the number of microorganisms in Woods Pond sediment capable of dehalogenating 26-BB and PCBs increased nearly 1000-fold (from 3–4.9×105 to 2–5.8×108 cells per g sediment (dry weight)) after priming with 26-BB (1050 μM) plus 10 mM malate. These results demonstrate that halogenated biphenyls prime PCB dechlorination primarily by stimulating the growth of PCB-dechlorinating microorganisms. An additional stimulation of the dehalogenation through the induction of dehalogenation enzymes is possible, but presently regarded as less important.

Several investigations of PCB dechlorination were conducted in the presence of common electron acceptors for anaerobic microorganisms. PCB dechlorination was usually observed under methanogenic conditions [70, 8991], and the addition of bromoethane sulfonic acid (BESA), an inhibitor of methanogenesis, inhibited dechlorination of certain PCB congeners [91, 92] or dechlorination processes [90]. However, ethanol-treated, pasteurized cultures obtained from Hudson River exhibited meta dechlorination of Aroclor 1242 [93] and the addition of BESA did not inhibit meta dechlorination of 2346-CB by a 2346-CB enrichment culture [65], indicating the methanogens may not carry out the dechlorination but influence the availability of electron donors in these cultures. The addition of sulfate (10–30 mM), an electron acceptor used by sulfate reducing bacteria, completely inhibited dechlorination or favored one dechlorination process over others [60, 8991]. Nitrate (10–16 mM) had either no effect or an inhibitory effect on dechlorination [90, 91]. Ferric oxyhydroxide (50 mM) decreased the extent of PCB dechlorination [90]. Dechlorination activities of 2346-CB and 246-CB enrichment cultures were inhibited in the presence of 5 mM sulfate, thiosulfate, sulfite or nitrate [65], suggesting that under the test conditions, either these electron acceptors were preferred over PCB or that non-dehalogenating bacteria using these acceptors were out-competing the PCB dehalogenator(s) for the available electron donor(s).

Zwiernik et al. [94] found FeSO4-enhanced anaerobic reductive dechlorination of Aroclor 1242 by microorganisms from Hudson River sediments. Addition of FeSO4 or both Na2SO4 and PbCl2 stimulated para dechlorination. However, addition of Na2SO4, FeSO4 and molybdate, or Fe(OH)3, did not have the stimulatory effect on para dechlorination. These authors propose that the addition of FeSO4 or Na2SO4 (plus PbCl2) provides sulfate ions as electron acceptors which stimulated the growth of para dechlorinating bacteria. The increased concentrations (supplementation) of Fe2+ or Pb2+ ions removed through precipitation sulfide ions formed during sulfate reduction and thus reduced the bioavailability and hence toxicity of the sulfide ions and/or H2S. These results suggest that if the sulfide/H2S concentrations can be kept low enough, PCB dehalogenation and sulfate reduction can coexist in a sediment.

In summary, the different results suggest that reductive PCB dechlorination can occur under methanogenic, sulfidogenic, iron(III) reducing and denitrifying conditions and that which process(es) occur or predominate depends on the physiology of the indigenous microflora and the nutritional conditions of a given environment. The data also support further the hypothesis that PCB congeners are used as respiratory e-acceptors.

2.3 PCB-dechlorinating anaerobic enrichment cultures

The isolation of microorganisms capable of dechlorinating PCBs is necessary to fully elucidate the process. As mentioned above, PCB dechlorination is prevalent in anoxic habitats; however, to date, all attempts to isolate pure cultures of anaerobes reductively dehalogenating PCBs have been unsuccessful. Only a few enrichments have been obtained ([24, 46, 52, 58, 61] and literature cited therein, [95, 96]). In this mini-review, the term ‘enrichment culture’ refers to a culture obtained through several sequential transfers into sterile sediment-containing medium and exhibiting increased dechlorination rates or an increased specificity. The terms microcosms and sediment culture refer to incubations of sediment slurries with less than two transfers into fresh sediment. Dechlorination activities could only be maintained in enrichments in the presence of sediment or sediment substitutes. Furthermore, anaerobic microbial communities frequently exhibit strong interactions among their members including dependences between non-dechlorinating and dechlorinating species, and this is apparently also true for PCB-dechlorinating communities. All the enrichment cultures reported were obtained by using either a PCB mixture (e.g. Aroclor 1242) or an individual PCB congener. Presumably, the PCBs act as selective electron acceptors. Boyle et al. [97] reported that periodic supplementation of sterile sediment and 236-CB resulted in a 100-fold increase in the dechlorination rate of 236-CB in a sediment slurry. This may prove to be one method to enrich for and isolate a desirable PCB-dechlorinating microorganism. Similarly, various methods have been used to separate two different dechlorination activities which existed in primary enrichment cultures. For example, Ye et al. [47, 93] reported that meta and para dechlorination activities of Aroclor 1242 were separated by using heat or ethanol and antibiotics. The effects of temperature and pH on the different dehalogenation activities reported above have been used to maintain para dechlorination activity of 2346-CB and to eliminate the meta dechlorination of 2346-CB in the 2346-CB enrichment culture [61]. Some of the enrichment cultures showed a narrow substrate spectrum [61, 65]. A 2346-CB enrichment culture para-dechlorinated only PCBs with flanked para chlorines whereas a 246-CB enrichment culture para-dechlorinated PCBs with flanked or unflanked para chlorines. Both of the enrichment cultures did not meta dechlorinate PCBs. Natarajan et al. [58] reported that a methanogenic consortium developed in a granular form exhibited complete dechlorination of 23456-CB to biphenyl via 2346-CB, 246-CB, 24-CB and 2-CB as intermediates by removal of meta, para and ortho chlorines. To the knowledge of the authors, this is the first report of complete dechlorination of a PCB congener to biphenyl by an anaerobic consortium.

Recently, two laboratories demonstrated that the presence of sediment was not necessary for PCB dechlorination activity in their enrichments. Hartcamp-Commandeur et al. [24] reported ortho, meta and para dechlorination of 234-234-CB and 236-236-CB by anaerobes derived from Dutch sediment. Enrichment cultures were grown in a rich anaerobic growth medium containing 2.5% sediment. After visible growth, the organisms were transferred twice (1:10) into the medium amended with 1 μg l−1 234-234-CB and 1 μg l−1 236-236-CB. After incubation, dehalogenation products (234-34-CB, 234-23-CB, 236-23-CB) were detected. No dechlorination products were observed in sterile controls. However, the authors did not mention whether the culture growing in medium was sustainable when further serially transferred in sediment-free medium. Cutter et al. [95] clearly demonstrated a sustainable PCB dechlorination activity in medium without the addition of sediment. Primary culture from estuarine sediments grew in minimal medium containing 5% of dry weight sediment, 173 μM 2356-CB, and a mixture of sodium acetate, butyrate and propionate at a final concentration of 2.5 mM each. After incubation, ortho and meta dechlorination products (236-CB, 235-CB, 26-CB, 25-CB, 35-CB and 3-CB) of 2356-CB were found. After the first transfer, only 235-CB and 35-CB were detected, indicating that ortho dechlorination activity was maintained and meta dechlorination activity was eliminated. The dechlorination activity was sustainable and the lag time decreased up to the fifth transfers in sediment-free culture. This is the first report of sustained anaerobic PCB dechlorination in the complete absence of soil or sediment.

3 Dehalogenation of hydroxylated PCBs, an aerobic PCB degradation product, by a pure anaerobic culture

Aerobic dehalogenation and transformation by aerobic bacteria, birds, mammals and humans leads to various hydroxylated PCBs ([16, 98] and literature cited therein). The major route of PCB transformation in these organisms is via monohydroxylation which involves the mixed-function oxidases in the microsomal cytochrome P-450 system [99]. Since the hydroxylated PCB products are relatively stable in these organisms, they are excreted via urine and droppings [88, 100, 101] and they can be detected in the environment, including drinking water [102]. Hydroxylated PCBs can exhibit estrogenic or anti-estrogenic activities in humans, depending on in which organ or tissue they are active and on their substitutions ([98, 103, 104] and literature cited therein). Although the amendments to the Safe Drinking Water and Food Quality Protection Act require the monitoring of estrogenic substances in drinking water, there is very little known about anaerobic transformations of hydroxylated PCBs in the environment. The 4-hydroxylated PCB congeners are the main PCB metabolites excreted by birds and mammals. Wiegel et al. [105] demonstrated that Desulfitobacterium dehalogenans, a Gram-type positive bacterium isolated as an ortho chlorophenol dehalogenating anaerobe, can reductively dehalogenate the meta chlorines in para-hydroxylated PCBs, e.g. all four chlorines from 3,3′,5,5′-tetrachloro-4,4′-dihydroxybiphenyl. The dechlorination rates are equivalent to the dechlorination rates of, e.g. 2,4-dichlorophenol, the electron acceptor on which the organism was isolated and other 4-substituted 2(6)-chlorophenyl derivatives [106]. It should be stressed, however, that this dehalogenation occurs as a reductive dehalogenation process of chlorophenol derivatives which is assumed to follow a different mechanism than that affecting the more lipophilic PCBs. PCBs are not dehalogenated by D. dehalogenans.

4 Anaerobic debromination of polybrominated biphenyls (PBBs)

In contrast to PCBs, PBBs are still used, e.g. as the flame retardant Firemaster BP6. Probably due to the fact that the brominated compounds are not regarded as highly recalcitrant and that no accumulation of PBBs in sediments or bioaccumulation in vertebrates has been reported, relatively few studies on brominated biphenyls have been done. The microbial debromination of PBBs under anaerobic conditions has been demonstrated in a few instances [73, 75, 107, 108]. Morris et al. [107] demonstrated that the PBB-contaminated sediments and the PCB-contaminated sediments (i.e. their microbial communities) partially debrominated the commercial PBB mixture Firemaster BP6 by removal of meta and para bromide. Recently, Bedard and Van Dort [73] demonstrated complete microbial reductive debromination of 16 different mono- to tetrabrominated biphenyls in PCB-contaminated sediment via loss of meta, para and ortho bromide. Wu [61] and Chang [65] demonstrated that PCB-dehalogenating enrichments could dehalogenate the corresponding di- and tribromo biphenyls, but that some of the parallel obtained PBB debrominating enrichment cultures could not dechlorinate the corresponding PCB congeners. These results indicate the existence of a microbial population which can only dehalogenate bromo-substituted biphenyl congeners but not the chloro-substituted one. Similar observations have been made with chlorinated and brominated phenols ([109, 110] and literature cited therein). On the other hand, it needs to be stressed that the enrichment cultures obtained with PCB congeners kept the PCB-dehalogenating activity when subcultured in the presence of only the brominated congeners (unpublished results, [65]). The addition of BB congeners (e.g. 2,4-BB) also stimulated the PCB dechlorination [65, 73, 75] and it is proposed that this effect is due to that they stimulated growth of the dechlorinating microorganism by being used as respiratory electron acceptors.

5 Outlook on environmental applications

Microbially mediated reductive dechlorination can remove chlorines from higher chlorinated PCBs, producing mono- through tetrachlorobiphenyls which subsequently can be aerobically dehalogenated and mineralized. Therefore, for active or facilitated bioremediation processes, it is currently assumed that an anaerobic reductive dechlorination process phase (which transforms higher chlorinated PCBs to mono- to tetrachlorinated congeners) needs to be followed by an aerobic phase to mineralize the less chlorinated products and the biphenyl. Fish and Principe [111] demonstrated that Aroclor 1242 was biologically transformed by both aerobic degradation and anaerobic dechlorination in test tube microcosms. Their microcosms, which consisted of sediments with overlying water as the only source of oxygen, developed aerobic and anaerobic compartments. In the surface sediment of these microcosms, the added Aroclor 1242 was biotransformed by both aerobic biodegradation and anaerobic dechlorination and total PCB concentration decreased from 64.8 to 18 μmol kg−1 sediment. One suggestion is to add suitable analogs to PCB-containing sediments to initiate and stimulate the growth of PCB-dechlorinating microorganisms. Suitable analogs means compounds which effectively induce the dehalogenation but are fast and completely degraded and thus do not cause additional ecological or health risks as increasing the PCB concentrations in the environments would [74, 112]. The research group of Bedard reported [113, 114] that addition of 26-BB stimulated in situ extensive anaerobic meta dechlorination of residual Aroclor 1260 in caissons installed in Woods Pond (Lenox, MA, USA). PCB dechlorination was first observed at 3 weeks. During the summer, the dechlorination was slow in the bottom sediments because they were colder, but dechlorination in the bottom sediment continued during the winter because these sediments were not in direct contact with the colder water. By 312 days, the PCBs in the lower sediments had been dechlorinated almost as extensively as those in the upper sediments. Recently, DeWeerd and Bedard [115] reported using halogenated benzoates and other halogenated aromatic compounds (e.g. 4-bromobenzoate, 4-iodobenzoate and 2,5-dibromobenzoate) to stimulate the microbial dechlorination of Aroclor 1260 by a decrease of up to ∼35%meta and para chlorines per biphenyl after 163 days of incubation. The mineralization of halobenzoates under anaerobic conditions is well documented [21]. Therefore, halogenated benzoate could be considered as a potential chemical for the stimulation of PCB dehalogenation in bioremediation attempts in the environment.

Harkness et al. [116] reported a field study of in situ aerobic biodegradation of PCBs in the Hudson River. The results showed that indigenous aerobic microorganisms can degrade the lightly chlorinated PCBs present in these sediments. Addition of non-analogous compounds such as inorganic nutrients (ammonia-nitrogen and phosphate), biphenyl and oxygen enhanced PCB biodegradation.

The use of PCB-dechlorinating enrichment cultures (bioaugmentation), another approach to improve bioremediation of PCBs, showed promise for stimulation of dechlorination of PCBs in PCB-contaminated sediments. Bedard et al. [52] demonstrated their culture, which had been enriched on 23456-CB, enhanced Process N and Process LP dechlorination (meta and unflanked para dechlorination) of Aroclor 1260 residue in Woods Pond sediment. Wu and Wiegel [63] reported that the addition of one PCB-dechlorinating culture and a PCB primer (246-CB) into Aroclor 1260-contaminated Woods Pond sediment enhanced dechlorination of the residual PCBs, but also that other enrichments were less effective. However, both reports also stated that bioaugmentation without the addition of PCB primers resulted in little or no dechlorination, indicating that primers are apparently required to sustain the dechlorination activity. These results suggest again that chemicals which are quickly degraded in the environment and which can serve as economical substrates for growth of PCB dechlorinators and prime PCB dehalogenation need urgently to be identified to successfully accelerate PCB dechlorination in contaminated environments.

The use of anaerobic microbial consortia in the form of granules shows promise for extended anaerobic degradation. Natarajan et al. reported reductive dechlorination of spiked 23456-CB [58], Aroclor 1254 [117] and residual PCBs in PCB-contaminated sediment [118] by anaerobic microbial granules. Extensive dechlorination of PCBs occurred at low temperature [21]. Oxygen exposure inhibited the dechlorination, but the dechlorination activity was recoverable with incubation in the absence of oxygen [119]. Granules stored at room temperature (20–25°C) appeared to be more active compared to those stored at 4 or −20°C. These properties may make the granules suitable for in situ bioremediation of PCBs. However, from the reported data, the amount of granules that would be required to be added to an environment raised doubts on whether such an approach is practical and cost effective.

6 Conclusions

The potential for microbially mediated, anaerobic PCB dechlorination is apparently widespread in a variety of soils and sediments, i.e., the capability to induce anaerobic PCB dehalogenation has been observed in microcosms and sediment cultures obtained from soil and sediments without detectable PCB contamination (Table 1). The environment can be predominant methanogenic, sulfidogenic (sulfate reducing) or iron(III) reducing. Different dechlorination activities (flanked and unflanked meta, para and ortho dechlorination) have been documented. The influence of environmental factors such as temperature, pH, carbon content, the concentration of H2 and/or other electron donors, and the presence or absence of electron acceptors have been studied in more detail in only a few cases. The studies revealed that these various factors strongly influence the rate, extent and route of dechlorination. Other important environmental factors such salinity or co-contaminants (metals and polycyclic aromatic hydrocarbons) need to be much more intensively investigated. Pure cultures of reductively PCB-dehalogenating microorganisms especially need to be isolated in order to study the influence of microbial interactions on the dehalogenation processes in re-associated cultures. PCB-dechlorinating enrichment cultures growing in defined medium without the addition of sediment could provide a means for isolation of PCB-dechlorinating microorganism. The isolation of axenic PCB-dehalogenating anaerobic cultures remains a top priority in this field.

Furthermore, to enable development of effective bioremediation processes that are superior to mechanical removal of the contaminated soil and sediments, appropriate chemicals to induce and stimulate PCB dechlorination in contaminated environments with no or very slow ongoing PCB dehalogenation need to be identified and tested in field studies. Based on the present results, simple inoculation of a contaminated environment with enrichment cultures appears to be neither suitable nor practicable for successful stimulations of the significant dechlorination rates required for a practical and economical process. Thus, both further laboratory studies with pure cultures (or at least defined, sediment-free co-cultures) and in situ studies are required to obtain a better understanding of the importance of PCB as a potential electron acceptor in contaminated environments. Understanding the microbial interactions and the influences of environmental factors on them are paramount for evaluating the potential of reductive dehalogenation for possible commercial bioremediation processes removing aged PCB from environments but also to evaluate how contamination with PCBs and similar compounds disturbs the ecological balance of different microbial communities in various environments. Because the microbial actions are the underlying basis of healthy and thriving eukaryotic communities, major disturbances in the balanced function of the microbial community will have negative impacts on the life of animal and plants in the contaminated environments.

Acknowledgements

The authors are indebted to Donna L. Bedard for her helpful advice and General Electric Company for financial support in the past for work done in the laboratory of J.W., who also acknowledges the present support form the Office of Naval research, US Department of Defense (Grant # N00014-97-1-0955 given to J. Kostka, K. Maruya and M. Frischer). Q.W. acknowledges present support from the Office of Naval Research, US Department of Defense (Grant # N00014-99-1-0978 given to Harold D. May). We also thank Cara Runsick-Mitchell for help in editing the manuscript.

References

  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
  35. [35].
  36. [36].
  37. [37].
  38. [38].
  39. [39].
  40. [40].
  41. [41].
  42. [42].
  43. [43].
  44. [44].
  45. [45].
  46. [46].
  47. [47].
  48. [48].
  49. [49].
  50. [50].
  51. [51].
  52. [52].
  53. [53].
  54. [54].
  55. [55].
  56. [56].
  57. [57].
  58. [58].
  59. [59].
  60. [60].
  61. [61].
  62. [62].
  63. [63].
  64. [64].
  65. [65].
  66. [66].
  67. [67].
  68. [68].
  69. [69].
  70. [70].
  71. [71].
  72. [72].
  73. [73].
  74. [74].
  75. [75].
  76. [76].
  77. [77].
  78. [78].
  79. [79].
  80. [80].
  81. [81].
  82. [82].
  83. [83].
  84. [84].
  85. [85].
  86. [86].
  87. [87].
  88. [88].
  89. [89].
  90. [90].
  91. [91].
  92. [92].
  93. [93].
  94. [94].
  95. [95].
  96. [96].
  97. [97].
  98. [98].
  99. [99].
  100. [100].
  101. [101].
  102. [102].
  103. [103].
  104. [104].
  105. [105].
  106. [106].
  107. [107].
  108. [108].
  109. [109].
  110. [110].
  111. [111].
  112. [112].
  113. [113].
  114. [114].
  115. [115].
  116. [116].
  117. [117].
  118. [118].
  119. [119].
  120. [120].
View Abstract