Metabolic degradation, inducing potency, and metabolites of fluorinated and chlorinated-fluorinated dibenzodioxins and dibenzofurans

PII: soo45-6535(%)003fi5.7

Chemosphere,Vol. 34, No. I, pp. 29-40, 1997 copyright0 1996 Ehwia science Lad Printedin Great Britain. All rightsxrservcd 004%6535/97 S17.OOtO.oO

Metabolic Degradation, Inducing Potency, and Metabolites of Fluorinated and Chlorinated-Fluorinated

Dibenzodioxins and Dibenzofurans

R Weber’, H.-J. Schmib?, D. Schrend, H. Hagenmaier’ ‘Institute of Organic Chemistry and ‘Institute of Toxicology, University of Tiibingen, D-72076 Tiibingen (Receivedin Germany 2 September1996; accepted17 October 19%)

ABSTRACT The metabolic degradation of fluorinated, chlorinated-fluorinated and chlorinated congeners was measured in liver homogenate of NMRI mice. While in the time period between 0 and 240 min no degradation of the 2,3,7,8_TCDD/TCDF could be detected, for all fluorinated congeners a perceptible degradation was found, even for the 2,3,7,8-TFDD. Stepwise chlorination of the 2,3,7,8-fluorinated congeners leads to a decrease of the degradation rate. In the EROD test, the exchange of chloro- with fluorosubstituents in the 2,3,7,8-TCDF leads to a decrease of induction potency. 3,7-Dichloro-2,8-difluorodibenzotiran

was about l/lOOOth as potent as 2,3,7,8-

TCDF, while 2,3,7,8-TFDF was complete inactive. Comparison of the metabolic rates of different TCDD with those of the analogous TFDD demonstrates that the order of enzymatic degradation of different TCDD and the analogous TFDD is identical. The TFDD are degraded slightly faster than the corresponding TCDD. Surprisingly 1,4,6,9-TXDD showed the second slowest metabolic rate of the fluorinated and chlorinated TXDD after 2,3,7,8-TXDD although none of the 2,3,7,8_positions were substituted. Judging from 2,3,7,8-TFDD

and’ 1,7-dichloro-2,8-difluorodibenzofiran

the metabolic pathway of

fluorinated and chlorinated-fluorinated congeners seem to be comparable to the chlorinated congeners. Copyright 0 1996 Elsevier Science Ltd

INTRODUCTION Activation of the aryl hydrocarbon receptor probably leads to most of the adverse effects of 2,3,7,8-TCDD (teratogenicity, body weight loss, thymic atrophy, immunosuppression, carcinogenicity, and perturbation of reproductive finctions [ 1, 21). Among the criteria suggested for dioxin-like compounds, persistence in the environment and agonistic biological effects such as dioxin receptor-mediated induction of cytochrome P450 (CUP) 1Al are key determinants [3]. Based on these and other properties, a variety of polychlorinated dibenzodioxins, dibenzofirans, and biphenyls have been allocated Toxicity Equivalency Factors (TEF) describing their relative potency compared to 2,3,7,8-TCDD. In particular, TEF have been established or suggested for all 2,3,7,8-chlorosubstituted dibenzodioxins and dibenzotirans [4], a number of polychlorinated biphenyls [5,6], some chlorinated naphthalines [7], and brominated hydrocarbons [8]. Because fluorine has a smaller Van der Waals radius (1.35 A) than chlorine (1.8 A), fluorinated dibenzodioxins/dibenofirans

are somewhat smaller than their chlorinated analogues. Both substituents 29

30 exhibit similar effects on the aromatic core i.e. a -1 and +M effect, making further electrophilic substitution less probable, and directing additional substituents to the ortho- and puru-positions.

From our current

knowledge it is not possible, however, to predict the consequences of replacement of chlorine by fluorine in a dibenzodioxin/dibenzohnan

molecule upon metabolism, CYPI Al induction and other biological

properties. For 2,3,7,8-TFDD, rapid elimination from mouse blood and liver after i. p. injection was previously reported [9]. It was unclear however whether this finding is a result of rapid metabolism, an alternative means of elimination or transfer into a deep compartment such as the adipose tissue. Metabolism of PCDDNDF

has been investigated in a number of studies showing that despite small differences [lo], the

in viva and in vitro pathways of metabolism are basically identical [l I-131. In the present study, we investigated the influence of fluorination of dibenzodioxins/dibenzofurans on metabolic degradation using activated liver homogenate from NMRI mice. Induction of CYPl Al-catalyzed ‘I-ethoxyresorufin 0-deethylase (EROD) activity e.g. in mammalian cell culture is widely established as a parameter for the biological potency of PCDDNDF

and related

compounds [14, 151. For the relative potency of a number of inducers, and their toxic potential in viva a good correlation was reported [ 16, 171. MATERIALS

and METHODS

Isolation of the liver homogenate Liver homogenates were obtained from male NMRI mice weighing 20-23 g. The animals were pretreated with corn oil or with a total amount of 900 ug B-naphthoflavone (i. p.; dissolved in corn oil) at three consecutive days prior to sacrifice. Livers were pertbsed via the portal vein with sterile saline, removed, and minced in 4 ml TrYsucrose buffer (5 mM Tris, 250 mM sucrose, pH 7.4). The tissue was then homogenized in an ice-cooled Dounce homogenizer, centrifuged at 500 x g, and stored at -7OOC. Metabolism in the liver homogenate Incubations were performed at 37°C by mixing 200 ul 1M Tris-HCI (pH 7.4), 200 ul 1 mM MgCl2, 200 ul 5 mM NADP, 200 1.1150 mM sodium isocitrate, and 40 ul isocitrate dehydrogenase (Boehringer, Mannheim) with liver homogenate containing I .2 mg protein, and with a substrate mixture dissolved in 10 ul DMSO. At various timepoints, the incubations were stopped by adding 1 ml toluene supplemented with analytical standards. Polyhalogenated dibenzodioxins and dibenzotirans were extracted twice for 15 min with 1 ml toluene under sonification. The organic phase was removed, the solvent was evaporated to a final volume of 40 ul, and the samples were analyzed by HRGULRMS.

31 EROD-Induction For analysis of CYPlAl

induction, Hepa-l cells were incubated for 48 h with 2,3,7,8-TCDD, 2,3,7,8-

TCDF, or FzCIZDF (2,8-dithroro3,7-dichlorodibenzo&an),

and for 9 h with 2,3,7,8-TFDD or 2,3,7,8-

TFDF, and EKOD activity was determined as described [ 111. Synthesis of the congeners The chlorinated [ 181and tluorinated [9] diioxins

were synthesked as reported.

HRGCZVZMS-Detection The analysis was carried out with a HP 5890 gas chromatograph coupled directly to a HP 5970 mass selective detector. For separation of the isomers a SIL 88 column (50 m, 0.25 mm i.d., 0.2 urn film thickness, CHROMPACK Frankfurt, FKG) was used. Sample ahquots (1~1) were injected splitless (injector temperature 260 “C). Carrier gas was helium at a head pressure of 160 kPa. Temp. program for CP-SIL 88: A) fluorinated congeners: 90 “C, 2,7 “C/min. to 150 “C, 4 min. isothermal; 8 ‘C/min. to 170 “C, 3 min. isothermal; 20 “C/min. to 245 “C, 10 min. isothermal B) mixed chlorinated-fluorinated

congeners: 120 “C, 3 min. isothermal; 6 “CXnin. to 180 “C, 4 min.

isothermal; 10 “C/min. to 220 “C, 5 min. isothermal; 20 “C/min. to 245 “C, 15 min. isothermal. For the analysis of the metabolites a DB-5 tked

silica column (30 m, 0.32 mm i. d., 0.25 pm film

thickness, J&W Scientific, Folsom, USA) was used. Temp. Program for DB-5 (injector temp. 250 “C): 80 “C, 4 min isothermal; 10 “C/mm to 200 ‘C; 5 min isothermal, 5 oClmin to 300 “C.

RESULTS Metnbolism of polyfluorinated dibenzodkzins4ibenzofurans For the study of their metabolism, defined PFDD mixtures were incubated for 0 - 240 min in the presence of mouse liver homogenate and an NADPH-regenerating system. After incubation for 120 min a decrease was found for all congeners [Tab. 1] including the 2,3,7,8-substituted congeners which however were found to be the most stable. In particular, 2,3,7,8-TFDD was degraded to a measurable extent, which was not the case for its chlorinated counterpart 2,3,7,8-TCDD. In a separate experiment, using only 10 pmol OFDDIOFDF a significant degradation was also observed for these congeners, indicating however, a very low rate of metabolism (data not shown).

32 Table 1: Metabolism of PFDD in liver homogenates of mice after two hours of incubation

(C PFDD = 2 nmol/mg protein).

1,4,0-

l&D 6,7,9-l 1,2,4,6,7,9-WD 1,2,4,6,9-P#DD 1,4,6,9-T,FDD 1,2,3,4,6,7,9_H,FDD 1,2,4,7,8P#DD 1,4,7,8-T,FDD 1,2,4,6,7-T,FDD 12489-P$DD >. , .

78 69 67 63 61 55 32 32

PFDD-ccngener 1,2,6,9-/1,3,7,8-/1,3,6,9-/1,2,8,9-TAFDD 2,3,7-/1,2,4-/1,3,6-T,FDD 1,6DzFDD 1,3,9-/1,4,7-11.2.6~/1,2,9- T,FDD 1,7-/1,8-/1,9-/1,3-/2,3/1,2-D:FDD l-/2-M,FDD

1

1,2,7,9-TJT)D 1,2,3,4,6,8-HJ-1,2 ,3 ,4 .7-PJ 1,2,7-T,FDD 16 X2,8-/2,7-D#DD 16 X1,2,4.6-/1,2,4,71,2,4,7,9-TJDD 15 14 X1,3,7-/1,3,8-TFDD 1,2,3,4-T$DD 12 1,2,7,8-TdFDD 11 11 I-MFDF non metabolized [%]

I between five and ten

smaller than five

To exclude non specific degradation of PFDD, incubations with heat-inactivated homogenate were performed in parallel. Even after 240 min, no decrease of individually analyzed PFDD congeners was found under these conditions. Thus, fluorinated dibenzodioxins and dibenzomrans (especially the 2,3,7,8_substituted congeners) can be degraded enzymatically. This also demonstrates that the rapid elimination of 2,3,7,8-TFDD in our kinetic studies of mice [9] was due to metabolic degradation. To address the question of the catalyzing enzyme(s), liver homogenates from D-naphthoflavone-pretreated mice were used. This type of pretreatment inducing hepatic CYPlAlIIA2

led to a ca. fivefold increase in

metabolic degradation of the PFDD (Fig. 1). These results are comparable with the five- to six-fold increase of the rate of 2,3,7,8-TCDD metabolism by rat hepatocytes pretreated with 3-methylcholantren [201.

Figure 1: Relative concentration of some PFDD/PFDF a&r 60 min in liver homogenate induced with R-naphthoflavone (Ind), without addition of a NADPH-regenerating system (Ind (no NADPH)) and non induced homogenate (Nor) Omission of the NADPH-regenerating system, which provided NADPH, the cofactor of CUP-catalyzed monooxygenase reactions, almost completely abrogated PFDD metabolism (Fig. 1). When the CYPlA-inhibitor a-naphthoflavone was added to the incubation at a concentration of 10 pM, no significant PFDD metabolism was detectable within the first 30 min (Fig. 2). After 1 h, when the inhibitor was almost completely metaboliied, PFDD metabolism became detectable. 100

i

Figure 2: Degradation of TFDDII’FDF in liver homogenate (afkr 30,60 and 120 min) with and without the addition of a-naphthoflavone Our findings suggest that (an) NADPH-requiring

enzyme(s), presumably O-naphthoflavone-inducible

CYPIA isozymes play a major role in PFDD metabolism, similar to what is reported for PCDD [21].

34 Metabolism of chlorinated-fluorinated

dibenzodioxitw’dibenzofurans

In mouse liver homogenate, no metabolic degradation of 2,3,7,8-TCDD and 2,3,7,8-TCDF was detectable over an incubation period of 180 min (Fig. 3). In contrast, stepwise fluorination of the lateral positions of the molecules led to a considerable rate of degradation. A number of non-2,3,7,8-substituted PFDD were exceptionally labile towards metabolic degradation (Fig. 3), and, in most instances, were completely metabolized after 3 h of incubation with homogenate from pretreated animals. In general, 2,3,7,8substituted congeners were more resistant towards degradation than non-2,3,7,8_substituted congeners. The extent of metabolism of 2,3,7,8-TFDF, which showed the highest rate of all 2,3,7,8_fluorinated congeners, is comparable to that of the non-2,3,7,8_substituted chlorinated congeners (Fig. 3).

Figure 3: Comparison of metabolic degradation of the chlorinated, chlorinated-fluorinated and fluorinated

congeners in mouse liver homogenate (after 180 min) Metabolism of tetrajluorinated

vs. tetrachlorinated

dibenzodioxins

To investigate the effects of fluorination vs. chlorination on metabolic degradation, a defined mixture of tetrahalogenated dibenzodioxins was incubated with activated mouse liver homogenate. It was found that the rank order of metabolic stability is nearly identical for tetrachlorinated

and tetrafluorinated

dibenzodioxins (Fig. 4). In both cases, the 2,3,7,8_substituted congeners show the lowest rate of degradation. The fluorinated congeners consistently showed a higher rate of metabolic degradation than the chlorinated analogues. Thus, fluorine and chlorine substitution apparently had a parallel effect on the aromatic core regarding metabolic attack. Interestingly, the 1,4,6,9-substituted dibenzodioxins are surprisingly stable. Although not substituted at the lateral positions, they showed the second highest metabolic stability, being only surpassed by the 2,3,7,8substituted congeners. This is in good agreement with the presence of considerable amounts of 1,4,6,9TCDD besides the 2,3,7,8_substituted congeners in the liver, 14 and 24 h after application of a defined

35 PCDD mixture to rats [22,23]. It is interesting to note that the polar TXDD are metabolized faster than the non-polar. In this sense, 1,3,7,9-TXDD is degraded more rapidly than the 1,3,6,8-TXDD congeners and the 1,2,8,9-TXDD faster than the 1,2,6,7-TXDD congeners.

100

I

z 20

0

Figure 4: Metabolic degradation of TCDD and TFDD in mouse liver homogenate

Induction of CYPIA activity (EROD) with 2,3,7,&TFDF, 3,7-dichloro_2,&dijluorodibenzofrcran, and 2,3,7,8-TCDF Log-probit concentration-response curves from EROD induction data in Hepa-l mouse hepatoma cells are shown in Fig. 5. The curves allowed the calculation of ECJo values as parameters for the inducing potency of each congener. Substitution of two chlorine atoms in the 2,3,7,8-TCDF molecule by fluorine led to a dramatic decrease in potency while the completely 2,3,7,8-fluorinated congener TFDF was inactive. These effects of chlorine replacement by fluorine may result both from increased metabolic instability and possible yet undefined perturbations of receptor binding/activation or from effects on the DNA binding of the activated receptor. 2,3,7,8-TFDD however was a highly potent inducer of CYPlAl

mRNA and protein and a potent

transcriptional activator of CYP 1Al promoter-containing reporter gene construct, showing approximately ten-fold lower potency than 2,3,7,8-TCDD [24]. In contrast to OCDD which exhibits a moderate potency as inducer of EROD activity in H4IIE rat hepatoma cells and rat hepatocytes [ 141, the octafluorinated analogue OFDD was inactive as inducer.

36

-z IO0 \ 0 80E ,a ,h

60-

2

40-

: 0

20-

.----.----~

0 Kl

o-

-12

-13

-I

I

-10

-8

-9

lg concentration Figure 5: EROD-activity OFDD

of 2,3,7,84’FDF,

These results allow the conclusion dibenzofurans

that the stepwise fluorination

-5

-6

(M)

3,7-dichlor-2,8-diiuordibenzoIirran,

can be viewed as a kind of detoxification

decreasing EROD-activity.

-7

2,3,7,8TCDF/TCDD

of the chlorinated

of the congeners

and

dibenzodioxins

and

due to faster metabolism

and

This is confirmed by our own experiments (not shown here) that 2,3,7,8-TFDD

did not exhibit any signs of overt toxicity within six weeks after weekly applications of lo-100

@kg

to

NMRI mice.

Identification

of metabolites offluorinated

For identification was synthesized

of metabolites,

and chlorinated-fluon’nated

1 mg of each 2,3,7,8-TFDD

congeners

and 1,7-dichloro-2,8-difluoro-dibenzofuran

and purified. Because of the lack of cofactors, the liver homogenates used had a very low

capacity for phase II reactions such as glucuronidation

or sulfation. Therefore,

the investigation

was

restricted to phase I metabolites. In the case of 2,3,7,8-TFDD

the two metabolites

difluoro-o-quinone

whereas

(Fig. 6)

hydroxylated

identified were dihydroxytetrafluorobiphenyl fluorinated

dibenzodioxins

ether and

could not be detected.

Obviously stepwise cleavage of central oxygen bridges occurs, first leading to the biphenyl ether which then, in a second cleavage step, may be converted to difluorocatechol

Under the conditions of CC/MS-

analysis the latter is probably oxidized to ditluoro-o-quinone. In case of halogen

1,7-dichloro-2,8-difluorodibenzofman,

substituents

four monohydroxylated

(Fig. 7) and a monohydroxylated

dibenzotirans

monochloro-difluoro-dibenzofuran

with intact could

be

identified. Fluorine substituents were not removed during metabolism, Cleavage of the central oxygen bridges is also found in the metabolism of PCDD, but only in the case of 2,3,7,8_substituted

congeners.

I-hydroxy-2,3,7,8-TCDD,

In contrast to 2,3,7,8-TFDD

however, 2,3,7,8-TCDD

8-hydroxy-2,3,7-trichlorodibenzodioxin

and, via NIH

is also metabolized to shift, to 2-hydroxy-

37

Figure 6: Mass spectra of diiuoro-o-chinon

(metabolite of 2,3,7,&TFDD)

120 ,

15

I

16

17

18

19

20

21

22

23

24

25

Minutes

Figure 7: Reconstructed mass hgmentograms of the metabolites of 1,7-dichioro-2,8-difluorodibenzhran

38 1,3,7,8-TCDD [13, 25, 261. Thus epoxidation at Cl-Cl a seems to occur with both congeners, but Cl-C2 epoxidation is prevented by 2,3,7,8_replacement with fluorine (Fig. 8). Interestingly, the total rate of metabolism was found to be considerably higher with 2,3,7,8-TFDD than with 2,3,7,8-TCDD

Figure 8: Scheme of the metabolism of 2,3,7,8-TCDD and 2,3,7,8-TFDD The nature of the metabolites as well as the relative metabolic rates indicate that fluorine and chlorine exhibit the same influence upon metabolic attack at the dibenzodioxin molecule (regarding metabolism, the electronic aromatic system is the object of concern here). Therefore, results obtained with fluorinated species should also give usetirl indications for the chlorinated analogues. Special emphasis should be placed on the 2,3,7,8_substituted congeners: Several authors have reported about the biotransformation of PCDD/PCDF and the selective enrichment of the 2,3,7,8-subst. congeneres [ 13,27 - 291. Several explanations for this behavior were offered [30], but at present, strong regioselective metabolic attack on either the dibenzodioxin or dibenzofbran molecule can be considered as the most plausible reason. Clearly, the regioselectivity of oxidation by cytochrome P450 must play an important role in this process [ 131. On the other hand, very little information is available on this subject for PCDDKDF. Regioselectivity in biotransformation needs to be treated more quantitatively in the sense of electron density distributions in congeners in order to be able to postulate the location of hydroxylation, as

39

demanded by Van den Berg [ 131. So far, only one study reported such an approach [3 11. Unfortunately the theoretically calculated positions for oxygen insertion were not in accordance with the result from laboratory experiments [13]. The situation is particularly problematic for PCDD/PCDF because calculation times are demanding and results only moderately reliable due to the combination of chlorination (third period of the periodic table of the elements), degree of chlorination and size of the molecule. Fluorine, like oxygen and carbon, belongs to the second period of the periodic table of the elements. Consequently calculations for fluorinated dibenzodioxins and dibenzoturans are easier, faster and give more reliable results. Investigations regarding this aspect are currently carried out. Acknowledgments We gratefilly acknowledge the expert technical assistance of Silvia Vetter REFERENCES [l]

A. Poland, J. C. Knutson, AMU. Rev. Pharmacol. Toxicol., 22, 5 17 (198’2).

[2]

S.. Safe, Annu. Rev. Pharmacol. Toxicol, 26, 371 (1986).

[3]

S.. Safe, CRC Crit. Rev. Toxicol., 13, 3 19 (1984).

[4]

NATO/CCMS, Report No. 178 (1988) (North Atlantic Treaty Organization, CCMS, New York).

[S]

S. Safe, Crit Rev. Toxicol., 24, 87 (1994).

[6]

U. G. Ahlborg et al., Chemosphere, 28, 1049 (1994).

[7]

A. Hanberg, F. Waern, L. Asplund, E. Haglund und S. H. Safe, Chemosphere, 20, 1161 (1990).

[8]

T. Zacharevski, M. Harris, S. Sage, H. Thoma, 0. Hutzinger, Toxicology, 51, 177 (1988).

[9]

R. Weber, D. Schrenk, H.-J. Schmitz, A. Hagenmaier, H. Hagenmaier, Chemosphere, 30, 629 (1995).

[lo]

T. Sawahata, J. R. Olson, R. A. Neal, Biochem. Biophys. Commun., 105,341 (1982).

[1 I] J. R. Olson et al., Proc. 1lth Int. Symp. on Dioxins and Related Compounds. Research Tringle Park, NC, 1991, 145. [12] V. J. Wroblewski, J. R. Olson, Toxicol. Appl. Pharmacol., 81, 231 (1985). [13] M. Van den Berg, I. De Jongh, H. Poiger, J. R. Olson, Crit. Riv. Toxicol., 24, 1 (1994). [14] D. Schrenk, H. Hagenmaier, T. Wiesmtiller, H. P. Lipp, K. W. Bock, Arch. Toxicol., 65. 114 (1991). [ 151 D. W. Nebert, F. J. Gonzales, Ann. Rev. Biochem., 56,945 (1987). [ 161 A. Poland, J. C. Knutson, Ann. Rev. Pharmacol. Toxicol., 22, 5 17 (1982). [17] S. Sate, Ann. Rev. Pharmacol. Toxicol., 26, 371 (1986). [18] W. Zoller, K. Ballschmiter, Fresenius Z. Anal. Chem., 323, 19 (1986). [19] P. W. OXeefe et al., Bull. Environ. Contam. Toxicol., 36,452 (1986). [20] J. V. Wroblewski, J. R. Olson,Drug. Metab. Dispos., 16,43 (1988).

40 [21] J. R. Olson, J. H. McReynolds, S. Kumar, B. P. McGarrigle, B. P. Gigliotto, Proc. 11 th Int. Symp. On Dioxins and Related Compounds, Research Triangle Park, NC, 199 1,145. [22] K. Abraham,T. Wiesmtiller, H. Brunner, R. Krowke, H. Hagenmaier, D. Neubert, Arch. Toxicol., 63, 193 (1989). [23] W. Kcimer, Dissertation, University of Ttlbiien,

1995.

[24] H. Schmitz et al., manuscript in preparation. [25] H. Poiger, C. S&latter, Nature, 281,706 (1979). [26] H. Poiger, H.-R. Buser, Biol. Banbury Rep. 18: Mechanisms Dioxin Action, p. 39, 1984. [27] K. Abraham, Arch. Toxicol., 63, 193 (1989). [28] R. J. Norstrom et al., Environ. Pollut., 66, 1 (1990). [29] M. Van den Berg, K. Olie, 0. Hutzinger, Chemosphere, 12,537 (1983). [30] M. Van den Berg, H. Poiger, Chemosphere, 18,677 (1989). [31] W. Veerkamp, P. Seme, 0. Hutzinger, J. Chem. Sot. Perkin Trans., 11,353 (1983).