Congener specific polychlorinated biphenyl metabolism by human intestinal microbe Clostridium species: Comparison with human liver cell line-HepG2

Abstract

Polychlorobiphenyls (PCBs), which adversely affect human fetal and infant development, are endocrine disrupter and cause neurological disorders. They may also be carcinogenic. It is not known whether these effects are due to whole PCBs or to its metabolites, produced by the human gastrointestinal system primarily the liver and/or by intestinal microbes such as Clostridium sp. The available data show that Clostridium perfringens, the most prominent species of Clostridium occurs in the human gut. C. beijerinckii is a special type of Clostridium present in the gut of autistic children with late onset autism. Since mixed cultures are better PCB metabolizers than single cultures, mixed cultures of Clostridium were used in this work. The first step in PCB degradation is the removal of the chlorine atoms and then the breaking open of the phenyl ring leading to the final degradation product: CO₂. In this study, GC-MS analyses were done to examine the effect of Clostridium sp. on PCB-153 and PCB-77 and the metabolites obtained with Clostridium sp. therein. In this paper, we report that the unlike human liver cells which cannot produce any PCB metabolites. Mixed Clostridium spp. can degrade these PCBs. Clostridium spp. and were able to dechlorinate PCB 153 (hexachlorobiphenyl) to pentachlorobiphenyl and PCB 77 (tetrachlorobiphenyl) to trichlorobiphenyl. Despite considerable absorption of PCB 153 (40%) and PCB 77 (50%) in 30 minutes and 1.5 hours respectively by human liver (HepG2) cells, they can not dechlorinate PCBs. It has been observed that slight differences in chemical structures of PCBs such as coplanar (PCB-77) vs. non-coplanar (PCB-153) has significant metabolic effects.

Polychlorobiphenyls (PCBs) are used in the manufacture of capacitors, transformers, cooling liquids, hydraulic fluid and lubricants, because of their dielectric properties and chemical stability. The physical and chemical nature of the PCBs including lipophilicity, heat resistance and inertness are reasons for the reported health problems¹. There are 209 congeners of PCBs, each differing in the position of chlorine atoms. However, due to steric hindrance, only approximately half of this number actually occurs in nature. The higher chlorinated PCBs are considered more hazardous than the lower chlorinated ones. However, solubility of the PCBs depends on the number of chlorine atoms with lower chlorinated forms, which are more soluble. Hence, they may lead to more wide spread negative health effects².

Depending upon the presence of chlorine atoms at the ortho positions, PCBs are divided into non-coplanar and coplanar. The non-coplanar PCBs are always ortho substituted (2 or more ortho chlorine) and have the two phenol rings in separate planes. PCB-153 (2,2′,4,4′,5,5′ hexachloro 1-1′ biphenyl) is one such example (Fig 1a). In contrast, a coplanar PCB has no ortho chlorines or one ortho chlorine. PCB-77 (3,3′,4,4′ tetra-chloro 1-1′ biphenyl) is an example of coplanar PCB (Fig 1b). It has been reported³ that even a slight difference in chemical structures of PCBs, such as coplanar (PCB-77) vs. non-coplanar (PCB-153) may result in a profound difference in effects on health.

Fig 1.

Structure of coplanar PCB-153 (2,2′,4,4′,5,5′ hexachloro 1,1′ biphenyl) and non-coplaner PCB-77 (3,3′,4,4′ tetrachloro 1,1′ biphenyl). The presence of more than one chlorine at Ortho position (2 or 2′) made PCB-153 a non-coplanar PCB.

PCBs, which adversely affect the development of human-fetus and infant, are endocrine disrupters and cause neurological disorders⁴. They may also be carcinogenic⁵. It is not yet known whether these negative effects of PCBs on health are due to whole PCBs or its metabolites. PCBs can enter the human body though the gastrointestinal tract, lungs or the skin. Since commercial production of PCBs was banned from USA in 1977, the current source of PCB- exposure in human is mainly through the consumption of contaminated foods. Leachates from PCBs dumps or PCB-contaminated ground fills have contaminated many rivers and lakes; and these compounds now appear to be bio-accumulating in human foods causing detrimental effects in humans⁶.

When PCB contaminated foods or water are ingested by people, it is partly absorbed in its original form into the liver, where it may undergo some chemical modifications. It then goes to the adipose tissue where it remains deposited for years according to Agency for Toxic Substances and Disease Registry (ATSDR). However, some portions of it may undergo chemical change in the human intestine by the commensal microbes. One such intestinal anaerobic microbe is Clostridium sp. Clostridium perfringens is one of the most prominent Clostridium species in human gut⁷. Clostridium beijerinckii is a special type of Clostridium present in the gut of autistic children with late onset autism^{7, 8}. Since PCBs are believed to be in part responsible for late onset autism, any disturbance of the microflora in human intestine can lead to changes in PCBs metabolism. Earlier it was shown in our laboratory that mixed cultures of Clostridium can better metabolize PCBs than single culture⁹. Hence mixed cultures of Clostridium spp. were used in this research work.

The first step in PCB degradation is the removal of the chlorine atoms and then there is breaking open of the phenyl ring^10-12. The final degradation product is CO₂¹³. In this present investigation, metabolites formed by Clostridium spp. were examined from both the PCB-153 and PCB-77. We also report here that whereas human liver cells cannot produce any PCB metabolites, Clostridium spp. are able to do so.

Materials and Methods

Chemicals and reagents

The human liver cell line HepG2 (Cat#HB-8065) and Clostridium beijerinckhii (Cat#25752) cultures were obtained from the ATCC (Manassas, VA). Dulbecco's modified Eagle's medium (DMEM, Cat# 11995-065), fetal bovine serum (FBS, Cat# 12318-028) and penicillin-streptomycin (PS) were from Invitrogen. Fluid thioglycollate medium (FTM, Cat#269720) was from Difco. Non-coplanar PCB-153 (2,2′,4,4′,5,5′-hexachlorobiphenyl, Cat# PRC-047, CAS#035065) and coplanar PCB-77 (3,3′,4,4′-tetrachlorobiphenyl, Cat# PRC-036, CAS#032598) were obtained from Ultra Scientific. Stock solutions (50mM) of each type of PCB were prepared in dimethyl sulfoxide (DMSO from Sigma-Aldrich). Sterile phosphate buffer saline (PBS) was obtained from Quality Biologicals Inc. (Gaithersburg, MD).

Bacterial and cell culture

In 1000mL of distilled water 29.5 g FTM was dissolved, dispensed 20 mL in screw capped vials and autoclaved at 121°C for 30 min. Screw caps were tightened immediately after sterilization to prevent entry of air. Before inoculation, the tubes were heated to drive away oxygen; and checked for the contamination with oxygen by recording the color of the medium. Medium contaminated with oxygen turns pink. There was little pink coloration in the top layer of the medium, but the total relative volume of that was less than 20%. The tubes were tightly capped and cooled to <50°C. A mixed culture was prepared by mixing equal volumes (500μL) of Clostridium acetobutylicum/beijerinckii culture and C. perfringens (from our laboratory) in 3 - 4 stock culture tubes containing 5mL of FTM medium and incubating for 48 h at 37°C. This mixed culture was used in experiments for PCB metabolism. Henceforth these mixed cultures are referred as “Clostridium spp”. The bacterial cultures were stored in FTM medium at 4° C for future use.

Human liver cells (HepG2) were grown in DMEM medium, supplemented with heat activated 10% FBS and 1X penicillin-streptomycin in 25cm² tissue culture flasks. The complete medium was filtered through membrane filter of 0.22μ, pore size. The cells were grown at 37°C in an incubator containing 5% CO₂ and stored in liquid nitrogen.

Experimental design

All experiments were done in triplicates. Equal amount of human cells and bacterial cells were inoculated. For the cell cultures, medium was refreshed before experiments and checked under microscope for confluence. For bacterial metabolite studies, absorbance were re-checked. The exposure times for the liver cells were 0, 0.5, 1, 3, 6, 12, 18, 24, and 48 h and for bacterial cells it was 72 h, the almost similar time to which these bacteria get exposure to the ingested material in human intestine. The 0 h served as a control.

Growth curve of Clostridium sp.

Growth curves of the Clostridium spp. - cultures were prepared to evaluate the growth of the bacteria and to compare with growth in the presence of 70μM of PCB-153 or PCB-77. A concentration of 70μM was chosen for both PCB congeners. This is the LD₅₀ dose for the liver cell with both types of PCB. LD₅₀s were not calculated for the bacteria since they were apparently toxic to the bacteria. The bacteria were grown in FTM medium and the growth was measured spectrophotometrically at 600nm at after 0, 0.5, 1, 2, 3, 6, 12, 24, 48 and 72 h of inoculation.

LD₅₀ studies with HepG2

To study cell death, liver cells were grown in 25cm² tissue culture flasks for 45-50% confluency in DMEM medium with 10% FBS. The cells were washed with PBS and the medium was refreshed before treating with graded concentrations (0 to 100μM) of 2′2′,4,4′,5,5′ hexachloro-PCB or 3,3′,4,4′ tetrachloro-PCB (in triplicates) and incubated for 18 and 24 hours, respectively. Medium was removed from the plates by pipetting and centrifuged to get rid of any floating dead cells. 1ml of 1X Trypsin-EDTA was added to the plates for 5 minutes, and the cells dislodged. The trypsinized cells were placed into the tubes and then resuspended with 3ml of 4% Trypan Blue stain (1:1 with PBS) and allowed to stand at room temperature for 5 minutes (total volume is 4ml). A hemocytometer was filled with 10μl of the suspended cells and observed under microscope using low power. The four corner squares and the center squares in the central zone were counted. The lethal dose-50 (LD₅₀) for PCB-153 and PCB-77 was calculated by counting the number of viable cells after treatment with different PCB concentrations. The viable cells remained unstained with Trypan blue staining while the dead cells look blue whereas the viable cells exclude the Trypan blue stain. The appropriate percentage of living cells was established and a non-linear regression analysis performed (dose-response curve). This made the curve hyperbolic rather than sigmoidal¹⁴.

Sample preparation for GC-MS

The cell population of Clostridium spp. in each tube (1:5 dilutions) was spectrophotometrically checked at 600nm wavelength and made equal by diluting with sterile FTM medium, wherever required. In separate sets of tubes, 70μM of PCBs were added and incubated for 0, 0.5, 1, 3, 6, 12, 18, 24, 48 and 72 h. After the stipulated time, the tubes were rolled from side-to-side to mix the contents without disturbing the top layer of FTM contaminated with oxygen. Centrifuged 1.5mL of the culture at 13000g for 5 min to pellet the cells. The medium was removed to a new tube and extacted the PCB by adding an equal volume of hexane:acetone (1:1) with 10ng mL^-1 biphenyl as internal control in a 1.5mL eppendorf tubes. The tubes were vortexed for 1 min and the phases were separated by centrifuging at 5000g for 5 min. The upper phase (250μL) was transferred to an amber colored glass vial. For cellular extractions of PCB / metabolites from the cells, the cells were washed thrice, re-suspended in 50μL of distilled water, allowed to rupture by freeze and thawing (freezing for 30 min at -20°C for 30min followed by quick heating to 50°C for 10 min), followed by grinding in a hand-held motorized pestle inside the eppendorf tube (Knoles) with 500μL hexane: acetone (1:1).

The liver cells (10⁶ cells) were grown in DMEM medium with 10% FBS for 48 h in 25 cm² tissue culture flasks. For experimental purpose, 70μM of PCBs were added in separate sets of flasks (in triplicates) and incubated for 0, 0.5, 1, 3, 6, 12, 18, 24 and 48 hs. After the stipulated time, the medium was removed and the cells were first washed with 1X PBS and then harvested with 1mL of 1X Trypsin-EDTA and incubated for 5 min in 37°C. The cells were re-pelleted and washed three times with PBS. PCB was extracted from the medium and the cells as mentioned above.

GC-MS analysis of PCBs

Gas chromatography/MS spectrum (GC/MS) was done using Agilent 6890N interfaced with an Agilent 5973 inert Mass Spectra. The oven had an initial temperature of 250°C and a final temperature of 300°C, with a ramp rate of 30. The inlet helium pressure was 9.8 psi and the pulse pressure was 25psi. The ionizer temperature was 230°C with electron ionization (EI) energy of 70eV. High resolution gas chromatogram capillary column SPB-5 (30m length, 0.25mm i.d. and 0.25um film) was used (Sigma-Aldrich). One μL of the prepared sample was injected in the GC-MS for analytical purposes. The data was analyzed by use of HP Chemstation Productivity Software (vD1.00) and using the NIST 2002 library.

Quality assurance

Standard curves were drawn with pure PCBs dissolved in DMSO with concentrations of 1, 5, 10, 25, 50, 75 and 100μm of PCB-153 and PCB-77, respectively. A solution of 10ng mL^-1 of biphenyl was used as internal control for all extraction and the level of e biphenyl peak was monitored for consistent loading of the samples.

Calculation of total amount of PCBs

The total amounts of PCBs were calculated by first calculating the PCB:Biphenyl ratio followed by conversion in to μM from the standard curves.

Statistical analysis

All the results were statistically analyzed and expressed as mean±SEM using Graphpad Prism. In all cases p<0.05 were considered significant change.

Results and Discussion

Clostridium sp. growth curve

The growth curves of Clostridium spp were prepared with both types of PCBs and without PCB to determine whether the bacteria could grow in presence of 70μM of either type of the PCBs. Strikingly, the better growth rate of bacteria was observed in presence of PCBs (Fig. 2). This may be due to the fact these bacteria are using PCBs as carbon source. This phenomenon has also been observed in the PCB degrading fungus (Phanerochaete chrysosporium)¹⁵. The continuous increase in the bacterial number in exponential pattern proves that these bacteria were not killed by PCBs. The LD₅₀ dose was not calculated for the bacteria.

Fig 2.

Growth curve of mixed Clostridium spp., grown either in absence or presence of PCB-77 or PCB-153 taken at 600 nm.

LD₅₀ for liver HepG2 cells

The results of the LD₅₀ studies show that about 50% of the liver cells die at a PCB-153 concentration of 67μM after treatment for 18 hours. Therefore, for all subsequent experiments, the cells were treated with 70μM or 25 ppm of PCB-153. The results of the LD₅₀ studies also show that 50% of the cells die at a PCB-77 concentration of 63μM after treatment for 24 hours. Therefore, for all subsequent experiments, the cells were treated with 70μM i.e. 20 ppm of PCB-77 (Fig 3).

Fig 3.

Cellular viability of HepG2-cells after the treatment of PCB-153 and PCB-77 during 24 h. Non-linear regression curve (dose-response curve) to determine LD₅₀ was drawn. LD₅₀ was found about 67μM for PCB-153 and 63μM for PCB-77.

GC-MS analysis

The peak for biphenyl, PCB-153 and PCB-77 were observed after 6.47, 9.48 and 9.77min, respectively (Fig 4A and 5A). The figures show that the PCB-153 entered the cell slowly for the first 6 h since un-accounted for PCBs may be due to fixation of PCB-153 to the cell membrane and the capsule of the bacteria. Further analysis showed that Clostridium dechlorinated PCB-153 (hexachlorobiphenyl) to pentachlorobiphenyl (Fig 5). The dechlorinated product was detectable around 12 h and their concentration continued to increase up to 24 h. During this time the maximum height and area of peak for pentachlorobiphenyl were detected. There after, decrease in height and area of peak were observed and were probably due to the formation of lower chlorinated forms or other metabolites. A new peak appeared at 9.235 min. As the peak was very small, the extracted ion chromatogram showed the new peak of molecular wt of 324 (Fig 5B1) and the peak of PCB-153 with molecular wt 360 (Fig 5B2) as obtained from the GC-MS spectrum library. This new peak at 9.235 min was identified as 2,2′,4,5,5′-pentachlorobiphenyl. This was likely due to meta- dechlorination by Clostridia as was previously suggested¹⁴. As the Ortho position remains unchanged, there was no change in the planarity. It had been formed by loosing one chlorine at position 4′. This dechlorination process started around 12 h and continued until 72 h. The maximum height and area of peak were found at 24 h, as after that probably minute quantities of lower chlorinated forms, which were however not detected, were forming.

Fig 4.

Concentration of PCB 153 (A) and PCB 77 (B) at different time exposures on Clostridium spp. It depicts that PCBs depletion from the medium correlated with an increase in the intracellular concentration. PCB 153 (70μM) (C) and PCB 77 (70 μM) (D) measured in HepG2-cells and medium at different time measured by GC-MS. The rate of absorption was 0-30 min, when PCB -153 started decreasing in the medium and cytosolic concentration increases in caparison with PCB – 77, where it takes about 1-3 h.

Fig 5.

Total ion chromatogram of PCB-153 treated Clostridium spp. at 24 h with a new peak at 9.235 min (A). The extracted ion chromatogram (EIC) shows a new peak of molecular wt 324 at the time (arrow) (B1) and the peak of PCB-153 with molecular wt of 360 (B2) as per GC-MS spectrum library. The peak at 9.235 min (C1) was matched with the NIST 2002 database (C2) with a 98% concurrence of 2,2′,4,5,5′-pentachlorobiphenyl..

Similar study with PCB-77 showed slow entry of the PCB-77 into the cell concomitant with the gradual depletion of PCB from the medium (Fig 4B). Fig 5B shows the composite graph of the rate of PCB entry into the cell. Moreover, after 12 h, the PCB peak started falling suggesting degradation. Increased dechlorination after 12 h onwards might be due to time required to induce different enzyme.

Further analysis showed that Clostridium spp. dechlorinated PCB-77 (tetrachlorobiphenyl) to trichlorobiphenyl (Fig 6). This dechlorination started around 1.5 h and continued until 72 h. The maximum peak height and area were found at 12 h. The detected new peak of the pentachlorobiphenyl is shown in Fig 6A. The peak appeared at 8.771 min. As the peak was very small, the extracted ion chromatogram showed the new peak of molecular wt 256 (Fig 6B1) and the peak of PCB-77 with molecular wt 292 (Fig 6B2). The new peak at 8.771 min was of 3,4,4′-tricholobiphenyl and was due to para-dechlorination by Clostridia, as also suggested earlier. The comparative analysis is given in Table 1.

Fig 6.

Total ion chromatogram (TIC) of PCB-77 treated Clostridium spp. at 12 h with a new peak at 8.771 min (A). The extracted ion chromatogram (EIC) shows a new peak of molecular wt of 256 at the same time (arrow) (B1) and the peak of PCB-77 with molecular wt of 292 (B2) as per GC-MS spectrum library. The obtained peak at 8.771 min (C1) was matched roughly with NIST 2002 database (C2) as 3,4,4′-trichlorobiphenyl.

Table 1. Summary of the comparative data of metabolism.

	Clostridium sp.		Liver cells (HepG2)
PCB Congeners	PCB-153	PCB-77	PCB-153	PCB-77
Original nomenclature	(2,2′,4,4′, 5,5′-hexachlorobiphenyl)	(3,3′,4,4′-tetrachlorobiphenyl)	(2,2′,4,4′, 5,5′-hexachlorobipheny)	(3,3′,4,4′-tetrachlorobiphenyl)
Time to enter 50% of PCB (h)	9	2.5	0.75	2
Time to deplete 50% PCB from medium (h)	1	1	0.75	2
Time lag between depletion and entry of PCBs from the medium (h)	8	1.5	0	0
Biphenyl Peak (Internal Standard) (min)	6.47	6.47	8.67	8.02
PCB peak (min)	9.77	9.48	13.64	11.20
New peak (min)	9.235	8.773	None	None
Metabolite formed	2,2′,4,5,5′ hexachloro-1,1′ biphenyl (PCB-37)	3,4,4′trichloro-1,1′biphenyl (PCB-101)	None	None
Identification of metabolite from NIST database	98% similarity to CAS#037680-73-2	94% similarity to CAS#038444-90-5	None	None
Maximum metabolite generated after h:	24	12	None	None

The GC-MS study for PCB-153 with liver cells showed that the 40% of 2,2′,4,4′,5,5′-hexachlorobiphenyl was removed from the medium by the liver cells within 30 min concomitant with an increase in the intracellular PCB concentration (Fig 5C). The decrease in the PCB concentration in the medium was due to entry of PCB in the cell. No PCB metabolites were identified. Similarly, experiments with PCB-77 showed that 50% of 3,3′,4,4′-tetrachlorobiphenyl was removed from the medium by the liver cells within 1.5 h along with increase in the intracellular PCB concentration (Fig 5D). Hence, these studies with human liver cells showed that PCB-153 rapidly entered the liver cell within 1.5 h, whereas the entry of the PCB-77 took up to 6 h. This may be attributable to the enhanced membrane destabilizing property of PCB-153. In contrast, in Clostridium, both PCB-153 and PCB-77 took longer time to enter the cell. This was attributable to the presence of the cell wall in bacterial cells which are lacking in animal cells.