OUP user menu

Characterization of 3-chlorobenzoate degrading aerobic bacteria isolated under various environmental conditions

Janneke Krooneman, Arne O. Sliekers, Teresa M. Pedro Gomes, Larry J. Forney, Jan C. Gottschal
DOI: http://dx.doi.org/10.1111/j.1574-6941.2000.tb00698.x 53-59 First published online: 1 April 2000

Abstract

The rates of bacterial growth in nature are often restricted by low concentrations of oxygen or carbon substrates. In the present study the metabolic properties of 24 isolates that had been isolated using various concentrations of 3-chlorobenzoate, benzoate and oxygen as well as using continuous culture at high and low growth rates were determined to investigate the effects of these parameters on the metabolism of monoaromatic compounds. Bacteria were enriched from different sampling sites and subsequently isolated. In batch culture this was done both under low oxygen (2% O2) and air-saturated concentrations. Chemostat enrichments were performed under either oxygen or 3-chlorobenzoate limiting conditions. Bacteria metabolizing aromatics with gentisate or protocatechuate as intermediates (gp bacteria) as well as bacteria metabolizing aromatic compounds via catechols (cat bacteria) were isolated from batch cultures when either benzoate or 3CBA were used as C sources, regardless of the enrichment conditions applied. In contrast, enrichments performed in chemostats at low dilution rates resulted in gp-type organisms only, whereas at high dilution rates cat-type organisms were enriched, irrespective of the oxygen and 3-chlorobenzoate concentration used during enrichment. It is noteworthy that the gp-type of bacteria possessed relatively low μmax values on 3CBA and benzoate along with relatively high substrate and oxygen affinities for these compounds. This is in contrast with cat-type of bacteria, which seemed to be characterized by high maximum specific growth rates on the aromatic substrates and relatively high apparent half saturation constants. In contrast, bacteria degrading chlorobenzoate via gentisate or protocatechuate may possibly be better adapted to conditions leading to growth at reduced rates such as low oxygen and low substrate concentrations.

Keywords
  • 3-Chlorobenzoate
  • Substrate limitation
  • Hypoxic condition
  • Catechol pathway
  • Gentisate
  • Protocatechuate pathway

1 Introduction

Many studies have been performed to investigate the biodegradation of xenobiotic haloaromatic compounds in order to predict the fate of such chemicals in the environment and to develop effective waste treatment techniques. It is known that halogenated monoaromatic compounds can be degraded rather easily in oxic habitats. The degradation of monoaromatic compounds catalyzed by initial dioxygenases often proceeds via (chloro)catechol, with ring-fission by ring-cleavage dioxygenases and subsequent release of halogens takes place following this initial dioxygenation [17]. A common feature of such initial dioxygenases is that they possess relatively poor affinities for oxygen, and under low partial pressures of oxygen the activities of these enzymes are low [810]. As a result, intermediates such as (chloro)catechols accumulate, growth rates are reduced, and toxic polymeric structures consisting of polymerized (chloro)catechols may be formed [1113].

More recent studies have shown that bacteria isolated from hypoxic habitats appear to have modified metabolic pathways for the degradation of aromatic compounds so that accumulation of toxic intermediates can be avoided. For example, Kukor and Olsen [14] isolated toluene degrading Pseudomonas spp. which possessed catechol dioxygenases with improved affinities for oxygen. Moreover, the ability to denitrify allowed these bacteria to enhance their toluene degradation rate under oxygen limiting conditions in the presence of nitrate [15]. We recently described a 3-chlorobenzoate (3CBA) degrading Alcaligenes sp. which was also isolated under hypoxic conditions. This bacterium possessed an increased affinity for oxygen and a relatively high affinity for 3CBA. Dioxygenation reactions leading to the formation of chlorocatechols did not occur and 3CBA was degraded via a pathway involving hydroxylated benzoates [16]. Under both oxygen and 3CBA limiting conditions it was competitively superior to a Pseudomonas sp. which degraded 3-chlorobenzoate via chlorocatechol [13]. Based on these observations we speculated that enrichment cultures with either a reduced partial pressure of oxygen or under 3CBA limiting conditions will be dominated by bacteria using the gentisate or protocatechuate pathway (gp-type of bacteria) and not by bacteria using the catechol pathway (cat-type bacteria). To explore this possibility, batch cultures, with either air saturating conditions or with low oxygen concentrations in the gas phase were used to enrich and isolate 3CBA or benzoate degrading bacteria from various sampling sites, thus selecting for bacteria with highest μmax values. In addition, most probable number (MPN) counts of both gp bacteria and cat bacteria were performed to determine the abundance of bacteria with these two metabolic pathways in the samples used. To selectively enrich for bacteria on the basis of maximum substrate affinities at limiting substrate concentrations (3CBA or O2), chemostats were used which were operated at various dilution rates. The isolates obtained were characterized with respect to the metabolic pathway used for the degradation of 3CBA or benzoate. These results are discussed in relation to the presumed functional roles of these isolates in natural environments.

2 Materials and methods

2.1 Inocula, media and growth conditions

Inocula were obtained from the following sources: peat soil (Appèlbergen, Haren, The Netherlands), humus soil (Hortus, Haren, The Netherlands), sandy soil polluted with oil (polluted area, The Netherlands), agricultural soil (sandy potato field, The Netherlands), contaminated fresh water sediment (Biesbosch, The Netherlands), and sediment from an uncontaminated ditch (Hortus, Haren, The Netherlands). Fifty gram of each sample was suspended in 100 ml low chloride minimal medium (LMM) [17] and shaken for 2 h in a rotary incubator at room temperature. After allowing for sedimentation, 25 ml of the above liquid was filtered through a 5 μm filter to free the inoculum from protozoa. The filtrate was then used as a 20% inoculum for enrichment cultures wherein LMM was used as the medium and either 3-chlorobenzoate or benzoate were added as growth substrates at a final concentration of 2 mM. Vitamins (1 ml l−1) were added after autoclaving the medium [18]. Batch cultures were incubated in closed flasks under air saturating (20% O2) or hypoxic (2% O2+98% N2) conditions. The oxygen concentration in the liquid phase was kept in equilibrium with the gas phase by incubation in a rotary incubator at 150 rpm. The ratio between the gas phase volume and the volume of the liquid phase was 6.5, to ensure that consumption of oxygen by aerobic metabolism did not significantly affect the oxygen concentration. Oxygen concentrations were monitored in time with a gas chromatograph (Pye Unicam 104, equipped with a katharometer and a Poropack Q (Water Associates Inc.) 100 to 120 mesh column). At least three transfers (1%) were performed in these batch culture enrichments before selection and subsequent isolation of the dominant species were performed. Microtiter plates which were used for MPN counts, and agar plates (LMM+BA or 3CBA) were incubated in a jar with a gas phase that contained either 98% nitrogen plus 2% oxygen (MPN counts) or air (approximately 20% O2). Growth in continuous culture, under oxygen limited conditions or substrate limiting conditions, were performed in oxygen regulated chemostats as described previously [16].

2.2 Isolation, identification and characterization of bacteria

Benzoate and chlorobenzoate degrading bacteria were isolated from primary enrichments by serially diluting the culture in LMM plus either 3CBA or BA, and spreading of 0.1 ml of various dilutions on agar plates. All strains obtained were screened for growth on various substrates by determining increases in turbidity (OD433). Growth on the following aromatic compounds, 2 mM final concentration added from 100 mM sterile stock solutions, was tested: benzoate, 3-chlorobenzoate, 3-hydroxybenzoate, 4-hydroxybenzoate; and the following sugars (5 mM final concentrations added from 100 mM sterile stock solutions): glucose, fructose, arabinose, maltose, mannitol, and mannose.

The Gram-type of the bacteria was determined by adding 40% KOH to colony material [19]. API tests 20 NE and 20 E and supplementary tests described by API were performed to identify the isolates.

To obtain preliminary information on which metabolic pathway is used for the degradation of benzoate or 3-chlorobenzoate, the rates of O2 consumption by washed cell suspensions of exponentially grown cells were determined using key intermediates from the catechol pathway, protocatechuate pathway and gentisate pathway as substrates [16]. Isolates obtained from chemostat enrichments were cultivated in batch culture prior to the determination of the respiration rates with catechol, gentisate, or protocatechuate as substrates. In addition, the activities of key enzymes of these pathways, ortho- and meta-(chloro)catechol dioxygenase [20], 2,3-, 4,5- and 3,4-protocatechuate dioxygenase [21] and gentisate dioxygenase [22] were determined in cell free extracts from cells grown in batch culture with (chloro)catechol, protocatechuate and gentisate as substrates, respectively [16].

The number of viable heterotrophic 3CBA and BA degrading aerobes was estimated using the most probable number (MPN) technique. Serial dilutions (dilutions 1 in 10) were made (n=5) in microtiter plates from the same ‘Biesbosch filtrate’ as described in the previous section, in LMM medium containing either 2 mM 3CBA or BA. The last dilution that was positive for growth (ΔOD433>0.1) was re-inoculated into fresh LMM medium. To demonstrate the presence of gp-type and/or cat-type bacteria in these enrichments, cells were subsequently harvested (mid-log phase), washed twice and the rates of O2 consumption were determined using key intermediates from the catechol pathway, the protocatechuate pathway and the gentisate pathway.

3 Results

3.1 Batch culture enrichments

To investigate the effects of the enrichment conditions employed on the pathway used by various strains for the metabolism of 3CBA or BA, batch culture enrichments under air saturating conditions or low pO2 (2% O2) with either 3-chlorobenzoate or benzoate as the substrate were carried out. The pathway used by the isolates obtained was subsequently determined by measuring (i) the activity of the key enzymes (chloro)catechol dioxygenase, gentisate dioxygenase and protocatechuate dioxygenase and (ii) respiration rates using the key intermediates catechol, gentisate, and protocatechuate.

To get an impression of the number of viable heterotrophic 3CBA and BA degrading bacteria in a natural sample and the relative abundance of gp- and cat-type of bacteria, most probable number counts were performed with one sediment sample. Results obtained from such MPN counts with 3CBA or BA as the substrate, both at low oxygen concentrations and air saturating conditions, indicated that gp- as well as cat-type bacteria were present in comparable numbers (Table 1).

View this table:
1

Oxygen uptake rates in washed cell suspensions of exponentially grown enrichments inoculated with the last growth-positive dilution of dilution series on 3CBA or BA

SubstrateLast dilution positive for growthOxygen uptake rates (μmol O2 min−1 OD−1) with
CatecholProtocatechuateGentisate
3CBA106015.827.3
BA10886.9033.4

Nine isolates (strains BBH1–BBH9) were obtained from enrichments with benzoate as the substrate under air saturating conditions, and five strains (strains BBL5–BBL9) were isolated under hypoxic (2%) conditions (Table 2). Characterization of the isolates showed that both gp- and cat-type of bacteria were isolated under hypoxic as well as under air saturating conditions. Bacteria using the gentisate or protocatechuate pathway possessed gentisate and protocatechuate dioxygenase activities and no catechol dioxygenase activity. The opposite was true for cat-type bacteria which possessed catechol dioxygenase activity and no gentisate dioxygenase activity (Table 2). These results corresponded with respiration activities measured in whole cells of these isolates: gentisate and protocatechuate were respired by gp-type bacteria whereas catechol was respired by cat-type bacteria (data not shown).

View this table:
2

Isolates obtained from enrichments in batch and continuous culture on both benzoate (BA) and 3-chlorobenzoate (3CBA) and their main characteristics

Enrichment conditionSourceIsolateGrowth onμmax (h−1) onEnzyme activitiesfIdentified as
Glucose or fructose3-HydroxybenzoateBenzoate3CBACatechol dioxygenaseGentisate dioxygenaseProtocatechuate dioxygenaseSpeciesTypeg
Batch culture
BA, high pO2asedimentBBH1+e0.55e+Pseudomonas sp.cat
BA, high pO2sedimentBBH2+0.46+Pseudomonas sp.cat
BA, high pO2peat soilBBH4+0.52+Pseudomonas putidacat
BA, high pO2humusBBH6+0.82+Aeromonas sp.cat
BA, high pO2oil soilBBH7+0.52+Pseudomonas sp.cat
BA, high pO2oil soilBBH8+0.52+Pseudomonas sp.cat
BA, high pO2humusBBH9+0.36+Aeromonas sp.cat
BA, high pO2sedimentBBH3+0.12++Bordetella sp.gp
BA, high pO2peat soilBBH5+0.12++Bordetella sp.gp
BA, low pO2bsedimentBBL5+0.52+Pseudomonas putidacat
BA, low pO2sedimentBBL6+0.46+Pseudomonas alcaligenescat
BA, low pO2humusBBL7+0.35+Aeromonas sp.cat
BA, low pO2peat soilBBL8+0.32+n.d.cat
BA, low pO2oil soilBBL9+0.20++Bordetella sp.gp
3CBA, high pO2ditchCBH2+0.370.08+Comamonas sp.cat
3CBA, high pO2ditchCBH6+0.330.09+Comamonas sp.cat
3CBA, high pO2sedimentCBH3+0.130.07++Bordetella sp.gp
3CBA, high pO2oil soilCBH4+0.150.17++n.d.gp
3CBA, high pO2humusCBH5+0.110.16++Xanthomonas maltophiliagp
3CBA, low pO2ditchCBL3+0.490.10+Pseudomonas aeruginosacat
Continuous culture (3CBA)
low Dc, O2-limsedimentCH1+0.160.17++Alcaligenes sp.gp
low D, 3CBA-limsedimentCH2+0.180.17++Alcaligenes sp.gp
high Dd, O2-limsedimentCH3+0.50.21+Pseudomonas aeruginosacat
high D, 3CBA-limsedimentCH4+-0.50.21+--Pseudomonas aeruginosacat
  • aAir saturation.

  • b2% O2 plus 98% N2 in the gas phase.

  • cD=0.025 h−1.

  • dD=0.15 h−1.

  • eNo growth visible ΔOD430 nm≤0.05.

  • fEnzyme activity ‘+’, ≥20 nmol min−1 mg protein−1; no enzyme activity ‘−’ <20 nmol min−1 mg protein−1.

  • gCat-type represents bacteria using the catechol pathway for degradation of (3C)BA; gp-type represents bacteria expressing the gentisate or protocatechuate pathway for growth on (3C)BA.

Enrichments done in media that contained 3-chlorobenzoate as a substrate and different levels of oxygen yielded six isolates in total. Five of these strains were isolated under air saturating conditions (strains CBH2–CBH6), of which two used the catechol pathway for 3CBA degradation, while the others expressed gentisate and protocatechuate dioxygenase but not catechol dioxygenase. Under reduced oxygen tensions only one isolate was obtained (strain CBL3) and it possessed catechol dioxygenase activity and no gentisate and protocatechuate dioxygenase activities. As was also observed with enrichments and isolations on benzoate, it appears that the metabolic pathway used is independent of the oxygen concentrations used during the enrichment procedures.

All isolates were identified using API tests. This revealed that all of the isolates were β-proteobacteria. Bacteria expressing the gentisate or protocatechuate pathway belonged to the family Alcaligenaceae, whereas the cat-type of bacteria were members of the genera Pseudomonas or Aeromonas. Isolates that metabolized (3-chloro)benzoate via catechol were all able to grow on glucose and fructose, whereas those that metabolized (3-chloro)benzoate via gentisate or protocatechuate could not. In addition, none of the cat-type strains could use 3-hydroxybenzoate as a carbon source, whereas all the gp-type bacteria could. These differences may be characteristic for these organisms and coupled to the capacity to express either the gentisate or the catechol pathway or this may be due to the fact that both groups of bacteria belong to different taxa [23].

All isolates able to degrade 3-chlorobenzoate could also metabolize benzoate. The growth rates of all isolates on 3-chlorobenzoate were in the same range for all strains. Remarkably, the gp-type bacteria had comparable growth rates on benzoate and 3-chlorobenzoate, whereas the cat-type of bacteria showed significantly reduced growth rates on the chlorinated compound relative to the growth rates achieved on benzoate (Table 1).

3.2 Continuous culture enrichments

To selectively enrich for bacteria under air saturating plus 3CBA limiting conditions or oxygen limiting conditions in the presence of excess 3CBA, chemostats were used and operated at different dilution rates. Four bacterial strains, CH1–CH4, were isolated from such chemostat enrichment cultures (Table 2). Strains CH1 and CH2 were isolated at a low dilution rate (D=0.025 h−1). Strain CH1 was isolated under oxygen limiting conditions in the presence of 3CBA excess and was shown to degrade 3CBA via gentisate or protocatechuate. Strain CH2, isolated under air saturating and 3-chlorobenzoate limiting conditions, also used the gentisate pathway. In contrast, strains CH3 and CH4 were isolated at a higher dilution rate (D=0.15 h−1) and both strains were characterized as cat-type bacteria. As with the isolates obtained from batch culture enrichments, the cat-type bacteria belonged to the genus Pseudomonas, whereas the gp-type of bacteria were Alcaligenes spp. Within this group of chemostat isolates, the same differences with respect to the use of substrates and differences in growth rates on benzoate and 3-chlorobenzoate were observed, as with the isolates obtained in batch culture (strains CBH2–CBH6 and CBL3). Subsequently, the apparent half saturation constants for oxygen were determined. These apparent half saturation constants, as determined by respiration of 3CBA by intact washed whole cells, reflect the kinetic properties of recombined influence of a series of biochemical reactions, including the uptake of the substrate, intercellular conversions and the subsequent respiration of the organism. It revealed that strains CH3 and CH4 possessed relatively high apparent half saturation constants for oxygen (24.4–26.1 μM), whereas strains CH1 and CH2 possessed lower apparent half saturation constants for oxygen (∼16 μM). In addition, apparent half saturation constants for 3CBA of the gentisate oxidizing bacteria were approximately ten times lower (20–30 μM) than the apparent half saturation constants for 3CBA of the catechol oxidizing bacteria (100–200 μM).

4 Discussion

An important finding of this study is that bacteria degrading (chloro)benzoate via the catechol pathway are not only characterized by high apparent half saturation constants for both 3-chlorobenzoate and oxygen, but also in general by high maximum specific growth rates on the aromatic substrate. In contrast, gp-type bacteria, using the gentisate or protocatechuate pathway for 3-chlorobenzoate and benzoate degradation, possess lower maximum specific growth rates in combination with relatively low apparent half saturation constants for oxygen and (3-chloro)benzoate. This suggests that gp-type bacteria may be well adapted to substrate and/or oxygen limiting growth conditions, whereas cat-type bacteria may be better tuned to environmental conditions with substrate and oxygen excess, allowing them to grow at relatively high rates.

Cat- and gp-type bacteria were isolated from batch enrichments, under both air saturating conditions and reduced partial pressures of oxygen. The organisms isolated under reduced levels of oxygen were all capable of good growth under atmospheric levels of oxygen. Apparently, the oxygen concentration used in the ‘low oxygen’ batch enrichments did not favor growth of a single particular metabolic type of bacteria. This may be due to the experimental set up that was chosen in such a way that the oxygen concentrations applied remained present in the cultures in levels comparable to those initially present (2%). It could be that these oxygen levels were not sufficiently low to establish an environment that strongly selected for microaerophilic bacteria that are usually defined as organisms requiring oxygen for respiratory purposes but growing best at levels significantly below atmospheric concentrations [24].

Six of the isolates (30%) used the gentisate pathway for growth on either benzoate or 3-chlorobenzoate, indicating that there is a chance that this metabolic type is also common in nature. MPN counts in one of the soil samples showed indeed the presence of high numbers of gp-type bacteria, comparable to those of cat-type bacteria. Such relatively high initial numbers of gentisate oxidizing bacteria present in the inocula may, at least partially, be responsible for the outcome of the batch enrichments.

Enrichments in continuous culture resulted in dominance of gp-type bacteria at low dilution rates (0.025 h−1). At high oxygen concentrations plus 3-chlorobenzoate limitation, these bacteria appeared to have a competitive advantage over the cat-type bacteria, probably due to lower 3CBA affinities of the latter, as shown previously [13, 16]. In addition, the higher affinities for oxygen that are characteristic of these newly isolated gp-type bacteria probably resulted in their dominance under oxygen limiting conditions. Applying high dilution rates and oxygen limiting conditions resulted in dominance of cat-type bacteria which is predicted on the basis of the outcome of competition experiments performed between a gentisate oxidizing Alcaligenes sp. and a catechol oxidizing Pseudomonas sp. [13]. This Pseudomonas species outcompeted the Alcaligenes only at high growth rates. Surprisingly, no growth inhibiting chlorocatechols accumulated as might have been expected for cat-type bacteria that grow under oxygen limiting conditions. In general, a distinction can be made between low μmax-low Ks and high μmax-high Ks types of bacteria, based on numerous competition and enrichment studies with heterotrophic and autotrophic bacteria [2528]. It now seems that this general concept also applies to growth on (chlorinated) benzoates.

Most probable number counts showed that the abundance of the gp-type bacteria in a soil sample may be comparable to the abundance of benzoate and chlorobenzoate degrading bacteria that use the catechol pathway. Although these were only enumerated in one sample, it implies that the gp-type bacteria may play a role in the decomposition of chlorobenzoate in nature. Moreover, the gp-type bacteria were also enriched and isolated from other sampling sites, indicating that they are widespread. Fulthorpe et al. [31] showed that 3-chlorobenzoate degrading bacteria are common members of soil microbial communities in many different natural environments. In addition, the fact that Fulthorpe and co-workers could not detect the presence of 3CBA catabolic genes encoding for the ortho-pathway via chlorocatechol (1,2-catechol dioxygenase) or the meta-pathway via protocatechuate (3,4-protocatechuate dioxygenase) among the 16 strains they tested, may suggest that bacteria using another pathway, possibly via gentisate, may indeed be pandemic. To explore this possibility a statistically significant number of bacteria should be isolated, preferably from different (globally dispersed) sampling sites and characterized to determine the presence, abundance and distribution of both types of bacteria and their specific metabolic activities. Molecular detection techniques could be valuable tools to study the presence, abundance and distribution of both types of bacteria and their specific metabolic activities.

In the catechol pathway, for degradation of chlorinated benzoates, chloride release generally takes place after ring-fission and cycloisomerization of the chlorosubstituted muconates [7]. Chlorinated intermediates formed in this pathway, such as chlorinated catechols, may lead to reduced growth rates [8, 11, 13]. In contrast, in the gentisate and protocatechuate pathway, chloride release probably takes place during initial hydroxylation of the chlorinated benzoate [29, 30]. If this is true, no chlorinated intermediates will be formed and inhibitory effects by such intermediates are prevented, thereby allowing undisturbed growth on the chlorinated compound. This corresponds with the clear difference between the two metabolic types of bacteria isolated on 3-chlorobenzoate. Those using the gentisate pathway, possessing relatively low maximum specific growth rates, did not show any significant difference in their μmax on chlorinated benzoate in comparison with the μmax on benzoate. This is in contrast with the bacteria using the catechol pathway, which possessed relatively high growth rates on benzoate and showed severely reduced growth rates on 3-chlorobenzoate.

Low growth rates due to substrate limitation and/or oxygen limitation might be expected in natural (hypoxic) environments such as oxic–anoxic interfaces in soil, and top layers of aquatic sediment surfaces. Moreover, in such environments organic substrates are often adsorbed to soil or sediment particles, resulting in even more severely substrate limiting conditions in the liquid phase. This emphasizes the potential ecological significance of the newly isolated gp-type bacteria, that are able to grow at substantial rates under severe substrate and oxygen limitations.

Another feature concerns the simultaneous use of multiple substrates by bacteria. This aspect has not been investigated within the context of the present study but it will undoubtedly add significantly to the competitiveness of 3CBA metabolizing heterotrophs [32], but whether this holds for gp- and cat-type of bacteria to the same extent can at present only be speculated.

Paradoxically, the cat-type bacteria are usually the ones that have been studied extensively with respect to aerobic degradation of halogenated monoaromatic compounds. The reason why the gp-type bacteria have remained unnoticed so far is not entirely clear. But probably the combination of high oxygen tensions and excess of (chloro)benzoate concentrations in previous enrichment and isolation procedures has always strongly selected for bacteria with the highest μmax and hence precluded isolation of gp-type of organisms with much lower μmax values.

When making predictions on the involvement of gp-type of bacteria in the degradation of haloaromatics in nature, one may speculate, on the basis of our results, that their role could be substantial. To substantiate such predictions quantitative data on their actual abundance are needed.

Acknowledgements

This investigation received financial support from the National Institute of Public Health and Environmental Protection, The Netherlands.

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].
View Abstract