- Email: [email protected]
protein interactions
FreeRadicalBiology & Medicine,Vol. 7, pp. 231-236, 1989 Printed in the USA.All rightsreserved.
0891-5849/89 $3.00+ .00 © 1989PergamonPressplc
Original Contribution RADICALS
INVOLVED
IN PHOTOALLERGEN
/ PROTEIN
INTERACTIONS
JONATHAN N. DELAHANTY, JEFFREY C. EVANS* and CHRISTOPHER C. ROWLANDS Department of Chemistry, University College Cardiff, P.O. Box 78, Cardiff CFI 1XL, U.K. MARTIN D. BARRATT and RUTH U. PENDLINGTON Environmental Safety Laboratory, Unilever Research, Colworth House, Sharnbrook, Bedford MK44 1LQ, U.K. (Received 24 February 1988; Revised 2 June 1988; Accepted 5 April 1989)
Abstract--Aqueous solutions (pH = 8) of both 3,3'-dimethyl and 4,4'-dimethyl substituted analogues of the photoallergen fentichlor (bis(2-hydroxy-5-chlorophenyl)sulphide) produced stable semiquinone radicals when irradiated with u.v. light (> 310 nm). These radicals have been characterised using electron spin resonance techniques: the results confirm the assignment of hyperfine coupling constants for the parent fentichlor radical. The binding of fentichlor to HSA was found to be partly oxygen dependent demonstrating a role for semiquinone type radicals in the binding mechanism. The stoichiometry and specificity of the binding of the dimethyl analogues to soluble proteins were found to be similar to that of fentichlor itself. Keywords--Photoallergens, ESR, Methyl derivatives of fentichlor, Proteins, Free radicals
bromide were purchased from Sigma Chemical Co. (Poole, U.K.). 4-Chlorophenol, 2-methyl-4-chlorophenol, 3methyl-4-chlorophenol were used as supplied from Aldrich Chemical Co. (Gillingham, U.K.). Elemental sulphur-35 (0.1 ml solution in toluene, activity 2 mCi) was obtained from Amersham International PLC (Amersham, U.K.).
INTRODUCTION
Wilkinson found that the germicide tetrachlorosalicylanilide (TaCS) caused abnormal sensititivy of the skin.~ This condition resulted from the interaction of the germicide with human skin under the influence of u.v. light and was described as photodermatitis. Similar photoallergic reactions have been described for a variety of compounds 2 and in particular for the germicide fentichlor. 3-s The formation of radicals on uv irradiation of fentichlor has been described in a previous paper. 6 This work has now been extended to study the radicals produced by uv irradiation of the 3,3'- and 4,4'-dimethyl analogues of fentichlor together with their binding properties to soluble proteins, using radiotracer methods.
Synthesis o f [35S]-labelled fentichlor, 3,3'-dimethylfentichlor and 4,4'-dimethylfentichlor
These compounds were synthesised by the reaction of sulphur dichloride, prepared from chlorine and sulphur (containing 35S) with the corresponding chlorophenol. 7'8 The yields of fentichlor, 3,3'-dimethylfentichlor and 4,4'-dimethylfentichlor were 62, 13, and 87% respectively. All three compounds were recrystallised from toluene and were found to be 96% pure by isotopic dilution analysis and thin layer chromatography (CHC13:CH3OH, 19: 1). Nonlabelled compounds were also prepared by this method. The melting points were 192-194°C (3,3'dimethylfentichlor) and 193-195°C 4,4'-dimethylfentichlor).
EXPERIMENTAL
Materials
Human serum albumin (HSA fraction V, fatty acid free), bovine insulin, human ~-globulin and cyanogen
*Author to whom correspondence should be addressed. 231
J.N. DELAHANTYet al.
232
Irradiation of protein solutions with fentichlor and dimethyl derivatives Stock solutions of the compounds in methanol were prepared (5 mg/ml) and added to solutions of protein in 0.1 M tris-HCl buffer (3 ml) using a microsyringe. Samples were irradiated in quartz vials (20 ml) with stirring at a distance of 4 cms from the light source for 30 min. A 450 watt Hanovia medium pressure Hg-arc lamp, fitted with a Schott WG-310 filter (1 mm) and a UG-5 filter (2 mm, uv transmitting) giving 58% transmission at 313 nm was used as a light source. After irradiation, samples were passed through a column of Sephadex G-10 equilibrated with 0.05 M ammonium bicarbonate to separate the photoadduct from unbound photoallergen: the photoadduct was eluted in the void volume whilst the unbound photoallergen was eluted later.
RESULTS AND DISCUSSION
Irradiation of 3,3'-dimethyl and 4,4 '-dimethylfentichlor The ESR spectra obtained on irradiation of 3,3'dimethyl- and 4,4'-dimethylfentichlor in alkaline solution (pH = 8) are shown in Figure 1A and Figure 2A respectively together with computer simulations (Figs. I B and 2B respectively) based on the hyperfine coupling constants shown in Table 1. It can be seen
METHODS
The protein concentration was determined by the method of Bradford 9 using the Bio-rad protein assay. Liquid scintillation counting was performed using a Packard Prias PLD liquid scintillation spectrometer. The scintillant used was Liquiscint (National Diagnostics) (3 ml added to 0.02 ml of sample). Anaerobic solutions of the protein were prepared by passing a stream of dinitrogen through the solutions for one hour prior to the addition of the [35S]-fentichlor stock solution. These solutions were then treated as previously described.
0 "1 mT
g =2.006
B
Electron spin resonance (ESR ) spectroscopy of irradiated photoallergens Aqueous alkaline solutions of photoallergens contained in a glass flat cell, were irradiated in the cavity of a Varian E3 ESR spectrometer with a medium pressure Hg-arc lamp.
Specificity in photobinding of 4,4'-dimethylfentichlor to monomer HSA Monomer HSA was prepared from the commercial sample by the method of Pedersen.l° The monomerHSA-photoallergen photoadduct was treated with cyanogen bromide according to the method of Meloun et al. 11 The two major fractions obtained by Sephadex G-100 chromatography were assigned according to the method of Meloun and Kusnir. 12
Fig. 1. (A) ESR spectrumobtainedon irradiating aqueous solutions of 3,3'-dimethylfentichlor(pH 8.0); (B) Computersimulationof (A) obtained using parameters in Table 1.
Radicals in photoallergens
j
&
0.linT
g:2.006
I
'
t
Fig. 2. (A) ESR spectrumobtained on irradiating aqueous solutions of 4,4'-dimethylfentichlor (pH 8.0); (B) Computersimulation of (A) obtained using parameters in Table 1.
from Figure 1A that another species is present 2 °, this was not observed in the ESR spectrum of 4 , 4 ' dimethylfentichlor (Fig. 2A). The appearance of a second species was also seen on irradiating solutions of fentichlor. 6 It has been shown earlier by Grabowski ~3 that photolysis of halogenated phenols in aqueous alkaline solutions gives rise to the corresponding dihydroxybenzene by replacing the photolabile halogen with an OH group. Further irradiation in the presence of oxygen would result in the benzosemiquinone radical being produced (1 ° radical Fig. 3). This radical undergoes further oxidation and nucleophilic attack by O H - to give rise to a secondary radical of the 4-hydroxybenzosemiquinone type, which is consistent with the work of Pedersen. ~4 The scheme for 3,3'-di-
233
methylfentichlor (Fig. 3) shows how such radicals are produced.* The initial step of this process involves homolytic fission of the photolabile carbon--chlorine bond leading to the formation of a carbon-centred radical, too short-lived to be directly detectable by ESR. Evidence to support this is obtained in an experiment in which solutions of fentichlor in methanol (containing 1% of 0.1 M sodium hydroxide) were irradiated in the presence of the spin trap 2-methyl-2-nitrosopropane (MNP). Spectra were obtained of the CH2OH MNP adduct (aN = 1.54, azn = 0.61 mT). (Lit. values aN = 1.54, a2H = 0.625 mT). .5 (These spectra also contains a contribution from the di-tertiary butyl nitroxide radical (aN = 1.69 mT) formed by breakdown of the spin trap). In the absence of fentichlor no such spectra were obtained except for that of the di-tertiary butyl nitroxide radical. In a similar experiment, where alkaline aqueous solutions of MNP and fentichlor were irradiated, the di-tertiary butyl nitroxide radical spectrum was again obtained as well as the I°C fentichlor radical; no spin adduct spectrum was seen. Li and Chignell, 16however, have detected a phenyl adduct for both fentichlor and bithionol using the spin traps phenyl-Ntert-butyl-nitrone (PBN) and 5,5'-dimethylpyrrolineN-oxide (DMPO) on anaerobic irradiation. In subsequent experiments using these spin traps similar spin adduct spectra have been detected. It would appear that MNP is not a suitable trap for such phenyl radicals. The production of the CH2OH radical arises as a consequence of the primary carbon-centred radical abstracting an H. from the methanol. In aqueous solution however this radical is converted to a stable semiquinone radical initially by addition of an OH group followed by aerobic oxidation. (In the absence of oxygen, the semiquinone radical is not observed. 6) It is apparent in the case of the 4,4'-dimethylfentichlor that substitution blocks the ability to react further and hence form the 2 ° species.
Binding studies It has been shown previously that there are fundamental differences between the photochemical binding of fentichlor and TnCSto soluble proteins. TaCSreacts photochemically at a major site on albumin 17.18 with which it has a strong noncovalent affinity; ~9 it has a *We would like to thank the referee for pointing out the possibility that the secondaryradical could be formedby an oxygen independent redox reaction between the hydroxy derivative and the unreactive quinone. However, in the e.s.r, studies, oxygen is always present, as spectra are not observed in the absence of oxygen.
234
J. N. DELAHANTYet al. Table 1. Hyperfine Coupling Constants for the Radicals of Fentichlor and its Methylated Analogues Primary Radical 1°
mT 3
O-
OH
O
C1
OC
H
3
0.2
OH ~
S
~
Secondary Radical 2° 4
6
0.29
0.125
3&6
CH3
4
6
0.185
0.24
0.135
O-
OH
O
C1
O-
OH
6
CH~
0.15
0.24
--
--
" 2
O
CI
O-
0.09
CH3
.o- - f O
mT
OH
3
CH3
6
0.14
0.285
0.125
CI
No 2° Radical
c H ~ S ~ c H 3 O
OH
CH3x
C1
OH
O-
S
OH
O-
OH
CI
OH
CI
H3 ao DOH"
CI
CI
O"
HOII 0
I
OH
0
OH
O"
OH
CI
0
CI
O
C1
\
I ° radical
OH
OH
O"
OH
0
CI
oN.
OH
Cl
2 • radical Scheme
Fig. 3. Scheme for the production of free radicals from 3,3'-dimethylfentichlor.
Radicals in photoallergens low p h o t o c h e m i c a l reactivity with other proteins such as 'y-globulin. 2° F e n t i c h l o r shows no specific affinity for any particular site on a l b u m i n and reacts photoc h e m i c a l l y with a n u m b e r of different soluble proteins.2 The results o f b i n d i n g studies using 35S-labelled 3 , 3 ' - and 4 , 4 ' - d i m e t h y l f e n t i c h l o r are shown in Table 2. F r o m these results it can be seen that 3 , 3 ' - and 4 , 4 ' d i m e t h y l f e n t i c h l o r are capable o f bonding to more than one site on human serum a l b u m i n , "y-globulin and bovine insulin. The p h o t o c h e m i c a l binding characteristics o f both d i m e t h y l f e n t i c h l o r analogues are, not surprisingly, similar to those o f fentichlor z~ and unlike those of T4CS.~7"2° In o r d e r to d e t e r m i n e whether 4 , 4 ' - d i m e t h y l f e n tichlor e x h i b i t e d any specificity at all in its photoc h e m i c a l b i n d i n g to H S A , a p h o t o a d d u c t was p r e p a r e d with an initial ratio o f 1:1 ( 4 , 4 ' - d i m e t h y l f e n t i c h l o r : H S A ) . This had a molar ratio of b o u n d 4 , 4 ' d i m e t h y l f e n t i c h l o r : H S A o f 0.91 : 1. The 4 , 4 ' - d i m e t h y l f e n t i c h l o r - H S A p h o t o a d d u c t was treated with c y a n o g e n b r o m i d e . This cleaves the H S A m o l e c u l e at its six m e t h i o n i n e residues. Elution o f the c y a n o g e n b r o m i d e digest through S e p h a d e x G - 1 0 0 yields two main fractions d e s i g n a t e d N (residues 1 123 and 1 2 4 - 2 9 8 ) and C (residues 2 9 9 - 5 8 5 ) . (Note c y a n o g e n b r o m i d e c l e a v a g e of H S A leads initially to three fragments. Two o f these fragments are of similar size and elute together. The e x p e c t e d seven fragments o f H S A are o b t a i n e d when the disulphide bridges are c l e a v e d . See refs. 1 1 and 12.) R a d i o a c t i v e counts from 35S-4,4'-dimethylfentichl o r - H S A treated with c y a n o g e n b r o m i d e showed the 4 , 4 ' - d i m e t h y l f e n t i c h l o r to be distributed equally between the C and N fragments o f H S A showing that, as with fentichlor. 2~ 4 , 4 ' - d i m e t h y l f e n t i c h l o r shows no specificity for any particular part o f the H S A molecule. Role of oxygen
W h e n fentichlor and d i m e t h y l f e n t i c h l o r ( s ) are irradiated in the absence o f o x y g e n , no free radicals can Table 2. Covalent Binding of 4,4'-Diemthylfentichlor (4 MeF) and 3,3'-Dimethylfentichlor (3 MeF) to Soluble Proteins Binding Ratios (m:m) Nonirradiated
Irradiated
Protein
3MeF
4MeF
3MeF
4MeF
HSA Human "y-globulin Bovine insulin
0.8:1 0.4:1 0.6:1
1.4:1 0.6:1 0.2:1
7.7:1 1.4: 1 2.4:1
7.8:1 2.9:1 2.5:1
Irradiationtime--30 min. Initial ratio 10:1 (photoallergen:protein).
235
Table 3. Covalent Binding of [35S]-Fentichlorto HSA: Effect of Oxygen Nonirradiated m:m
Irradiated m:m
0.18:1 0.18:1
0.80: 1" 0.60:1"
Oxygen not excluded Deoxygenated
(n = 2) (n = 4)
Initial ratio 1:1 (m: m) (fentichlor/HSA). *Statisticallysignificantdifference(p < 0.0005). be detected 6 by E S R s p e c t r o s c o p y . The p h o t o c h e m i c a l binding o f 35S-fentichlor to H S A was d e t e r m i n e d in the presence and absence o f o x y g e n . The results (Table 3) show a small but significant reduction in the photoc h e m i c a l binding o f fentichlor to H S A in the absence o f o x y g e n . This suggests that whilst the semiquinone radicals f o r m e d after irradiation o f fentichlor are capable of reacting with the protein to form adducts, the m a j o r reaction is i n d e p e n d e n t o f o x y g e n and m a y involve direct reaction b e t w e e n the p r i m a r y carboncentred radical and the protein. Acknowledgement--One of us (JND) wishes to thank SERC and
Unilever for a CASE award. REFERENCES
1. Wilkinson, D. S. Photodermatitis due to tetrachlorosalicyalnilide. Br. J. Dermatology 73:213; 1961. 2. Harber, L. C.; Harris, H.; Baer, R. L. Photoallergic contact dermatitis. Arch Dermatol. 94:255; 1966. 3. Bury, J. N. Photoallergies to fentichlor and multifungin. Arch. Dermatol. 95:287; 1967. 4. Clayton, R. From fentichlor sensitivity to actinic reticuloid? Proc. Roy. Soc. Med. 69:379; 1976. 5. Ramsay, C. A. Skin responses to ultraviolet radiation in contact photodermatitis due to fentichlor. J. Invest. Dermatol. 72:99; 1979. 6. Delahanty, J. N.; Evans, J. C.; Rowlands, C. C.; Barratt, M. D. J. Chem. Soc. Faraday Trans 1:135; 1987. 7. Dunning, F.; Dunning, B.; Drake, W. E. Preparation and bacteriological study of some symmetrical sulphides. J. Am. Chem. Soc. 53:3466; 1931. 8. Tueda, T.; Tsubui, K. Dihydroxydiphenyl sulphide derivatives. Jap. Patent 776; 1950. 9. Bradford, M. A rapid and sensitive method for the quantitization of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248; 1976. 10. Pedersen, K. O. Exclusion chromatography. Arch. Biochem. Biophys. Suppl.:157; 1962. 11. Meloun, B.; Saber, M. A.; Kusnir, J. Structure of N terminal cyanogen bromide fragment of human plasma albumin. Biochim. Biophys. Acta. 393:505; 1975. 12. Meloun, B.; Kusnir, J. Cleavage of human plasma albumin by cyanogen bromide and characterization of the fragments. Collect Czech. Chem. Comm. 37:2812; 1972. 13. Grabowski, Z. R. Photochemical reactions of some aromatic halogen compounds. Z. Physik. Chem. (Frankfurt) 27:239; 1961. 14. Pedersen, J. A. Electron spin resonance studies of oxidative processes of quinones and hydroquinones in alkaline solution. Formation of primary and secondary semiquinone radicals. J. Chem. Soc. Perkin Trans 2:424; 1973. 15. Buettner, G. R. Spin trapping: ESR parameters of spin adducts. Free Radicals Biol. Med. 3:259; 1987.
236
J.N. DELAHANTYet al.
16. Li, A. S. W.; Chignell, C. E Spectroscopic studies of cutaneous photosensitizing agents. Photochem and Photobiol. 46:445; 1987. 17. Rickwood, D. M.; Barratt, M. D. Evidence for a major strong binding site for tetrachlorosalicylanilide on human serum albumin. Photochem. Photobiol. 35:643; 1982. 18. Rickwood, D. M.; Barratt, M. D. Identification of the major covalent binding site for tetrachlorosalicylanilide on human serum albumin. Photobiochem. Photobiophys. 5:365; 1983.
19. Barratt, M. D.; Rickwood, D. M. Binding of a spin-labelled photoallergen to human serum albumin. Biophys. Chem. 19:69; 1984. 20. Kochevar, I. E.; Harber, L. C. Photoreactions of 3,3',4',5tetrachlorosalicylanilide with proteins. J. Invest. Dermatol. 68:1512; 1977. 21. Rickwood, D. M.; Barratt, M. D. Photochemical reactions of fentichlor with soluble proteins. Chem. Biol. Interactions 52:213; 1984.
0891-5849/89 $3.00+ .00 © 1989PergamonPressplc
Original Contribution RADICALS
INVOLVED
IN PHOTOALLERGEN
/ PROTEIN
INTERACTIONS
JONATHAN N. DELAHANTY, JEFFREY C. EVANS* and CHRISTOPHER C. ROWLANDS Department of Chemistry, University College Cardiff, P.O. Box 78, Cardiff CFI 1XL, U.K. MARTIN D. BARRATT and RUTH U. PENDLINGTON Environmental Safety Laboratory, Unilever Research, Colworth House, Sharnbrook, Bedford MK44 1LQ, U.K. (Received 24 February 1988; Revised 2 June 1988; Accepted 5 April 1989)
Abstract--Aqueous solutions (pH = 8) of both 3,3'-dimethyl and 4,4'-dimethyl substituted analogues of the photoallergen fentichlor (bis(2-hydroxy-5-chlorophenyl)sulphide) produced stable semiquinone radicals when irradiated with u.v. light (> 310 nm). These radicals have been characterised using electron spin resonance techniques: the results confirm the assignment of hyperfine coupling constants for the parent fentichlor radical. The binding of fentichlor to HSA was found to be partly oxygen dependent demonstrating a role for semiquinone type radicals in the binding mechanism. The stoichiometry and specificity of the binding of the dimethyl analogues to soluble proteins were found to be similar to that of fentichlor itself. Keywords--Photoallergens, ESR, Methyl derivatives of fentichlor, Proteins, Free radicals
bromide were purchased from Sigma Chemical Co. (Poole, U.K.). 4-Chlorophenol, 2-methyl-4-chlorophenol, 3methyl-4-chlorophenol were used as supplied from Aldrich Chemical Co. (Gillingham, U.K.). Elemental sulphur-35 (0.1 ml solution in toluene, activity 2 mCi) was obtained from Amersham International PLC (Amersham, U.K.).
INTRODUCTION
Wilkinson found that the germicide tetrachlorosalicylanilide (TaCS) caused abnormal sensititivy of the skin.~ This condition resulted from the interaction of the germicide with human skin under the influence of u.v. light and was described as photodermatitis. Similar photoallergic reactions have been described for a variety of compounds 2 and in particular for the germicide fentichlor. 3-s The formation of radicals on uv irradiation of fentichlor has been described in a previous paper. 6 This work has now been extended to study the radicals produced by uv irradiation of the 3,3'- and 4,4'-dimethyl analogues of fentichlor together with their binding properties to soluble proteins, using radiotracer methods.
Synthesis o f [35S]-labelled fentichlor, 3,3'-dimethylfentichlor and 4,4'-dimethylfentichlor
These compounds were synthesised by the reaction of sulphur dichloride, prepared from chlorine and sulphur (containing 35S) with the corresponding chlorophenol. 7'8 The yields of fentichlor, 3,3'-dimethylfentichlor and 4,4'-dimethylfentichlor were 62, 13, and 87% respectively. All three compounds were recrystallised from toluene and were found to be 96% pure by isotopic dilution analysis and thin layer chromatography (CHC13:CH3OH, 19: 1). Nonlabelled compounds were also prepared by this method. The melting points were 192-194°C (3,3'dimethylfentichlor) and 193-195°C 4,4'-dimethylfentichlor).
EXPERIMENTAL
Materials
Human serum albumin (HSA fraction V, fatty acid free), bovine insulin, human ~-globulin and cyanogen
*Author to whom correspondence should be addressed. 231
J.N. DELAHANTYet al.
232
Irradiation of protein solutions with fentichlor and dimethyl derivatives Stock solutions of the compounds in methanol were prepared (5 mg/ml) and added to solutions of protein in 0.1 M tris-HCl buffer (3 ml) using a microsyringe. Samples were irradiated in quartz vials (20 ml) with stirring at a distance of 4 cms from the light source for 30 min. A 450 watt Hanovia medium pressure Hg-arc lamp, fitted with a Schott WG-310 filter (1 mm) and a UG-5 filter (2 mm, uv transmitting) giving 58% transmission at 313 nm was used as a light source. After irradiation, samples were passed through a column of Sephadex G-10 equilibrated with 0.05 M ammonium bicarbonate to separate the photoadduct from unbound photoallergen: the photoadduct was eluted in the void volume whilst the unbound photoallergen was eluted later.
RESULTS AND DISCUSSION
Irradiation of 3,3'-dimethyl and 4,4 '-dimethylfentichlor The ESR spectra obtained on irradiation of 3,3'dimethyl- and 4,4'-dimethylfentichlor in alkaline solution (pH = 8) are shown in Figure 1A and Figure 2A respectively together with computer simulations (Figs. I B and 2B respectively) based on the hyperfine coupling constants shown in Table 1. It can be seen
METHODS
The protein concentration was determined by the method of Bradford 9 using the Bio-rad protein assay. Liquid scintillation counting was performed using a Packard Prias PLD liquid scintillation spectrometer. The scintillant used was Liquiscint (National Diagnostics) (3 ml added to 0.02 ml of sample). Anaerobic solutions of the protein were prepared by passing a stream of dinitrogen through the solutions for one hour prior to the addition of the [35S]-fentichlor stock solution. These solutions were then treated as previously described.
0 "1 mT
g =2.006
B
Electron spin resonance (ESR ) spectroscopy of irradiated photoallergens Aqueous alkaline solutions of photoallergens contained in a glass flat cell, were irradiated in the cavity of a Varian E3 ESR spectrometer with a medium pressure Hg-arc lamp.
Specificity in photobinding of 4,4'-dimethylfentichlor to monomer HSA Monomer HSA was prepared from the commercial sample by the method of Pedersen.l° The monomerHSA-photoallergen photoadduct was treated with cyanogen bromide according to the method of Meloun et al. 11 The two major fractions obtained by Sephadex G-100 chromatography were assigned according to the method of Meloun and Kusnir. 12
Fig. 1. (A) ESR spectrumobtainedon irradiating aqueous solutions of 3,3'-dimethylfentichlor(pH 8.0); (B) Computersimulationof (A) obtained using parameters in Table 1.
Radicals in photoallergens
j
&
0.linT
g:2.006
I
'
t
Fig. 2. (A) ESR spectrumobtained on irradiating aqueous solutions of 4,4'-dimethylfentichlor (pH 8.0); (B) Computersimulation of (A) obtained using parameters in Table 1.
from Figure 1A that another species is present 2 °, this was not observed in the ESR spectrum of 4 , 4 ' dimethylfentichlor (Fig. 2A). The appearance of a second species was also seen on irradiating solutions of fentichlor. 6 It has been shown earlier by Grabowski ~3 that photolysis of halogenated phenols in aqueous alkaline solutions gives rise to the corresponding dihydroxybenzene by replacing the photolabile halogen with an OH group. Further irradiation in the presence of oxygen would result in the benzosemiquinone radical being produced (1 ° radical Fig. 3). This radical undergoes further oxidation and nucleophilic attack by O H - to give rise to a secondary radical of the 4-hydroxybenzosemiquinone type, which is consistent with the work of Pedersen. ~4 The scheme for 3,3'-di-
233
methylfentichlor (Fig. 3) shows how such radicals are produced.* The initial step of this process involves homolytic fission of the photolabile carbon--chlorine bond leading to the formation of a carbon-centred radical, too short-lived to be directly detectable by ESR. Evidence to support this is obtained in an experiment in which solutions of fentichlor in methanol (containing 1% of 0.1 M sodium hydroxide) were irradiated in the presence of the spin trap 2-methyl-2-nitrosopropane (MNP). Spectra were obtained of the CH2OH MNP adduct (aN = 1.54, azn = 0.61 mT). (Lit. values aN = 1.54, a2H = 0.625 mT). .5 (These spectra also contains a contribution from the di-tertiary butyl nitroxide radical (aN = 1.69 mT) formed by breakdown of the spin trap). In the absence of fentichlor no such spectra were obtained except for that of the di-tertiary butyl nitroxide radical. In a similar experiment, where alkaline aqueous solutions of MNP and fentichlor were irradiated, the di-tertiary butyl nitroxide radical spectrum was again obtained as well as the I°C fentichlor radical; no spin adduct spectrum was seen. Li and Chignell, 16however, have detected a phenyl adduct for both fentichlor and bithionol using the spin traps phenyl-Ntert-butyl-nitrone (PBN) and 5,5'-dimethylpyrrolineN-oxide (DMPO) on anaerobic irradiation. In subsequent experiments using these spin traps similar spin adduct spectra have been detected. It would appear that MNP is not a suitable trap for such phenyl radicals. The production of the CH2OH radical arises as a consequence of the primary carbon-centred radical abstracting an H. from the methanol. In aqueous solution however this radical is converted to a stable semiquinone radical initially by addition of an OH group followed by aerobic oxidation. (In the absence of oxygen, the semiquinone radical is not observed. 6) It is apparent in the case of the 4,4'-dimethylfentichlor that substitution blocks the ability to react further and hence form the 2 ° species.
Binding studies It has been shown previously that there are fundamental differences between the photochemical binding of fentichlor and TnCSto soluble proteins. TaCSreacts photochemically at a major site on albumin 17.18 with which it has a strong noncovalent affinity; ~9 it has a *We would like to thank the referee for pointing out the possibility that the secondaryradical could be formedby an oxygen independent redox reaction between the hydroxy derivative and the unreactive quinone. However, in the e.s.r, studies, oxygen is always present, as spectra are not observed in the absence of oxygen.
234
J. N. DELAHANTYet al. Table 1. Hyperfine Coupling Constants for the Radicals of Fentichlor and its Methylated Analogues Primary Radical 1°
mT 3
O-
OH
O
C1
OC
H
3
0.2
OH ~
S
~
Secondary Radical 2° 4
6
0.29
0.125
3&6
CH3
4
6
0.185
0.24
0.135
O-
OH
O
C1
O-
OH
6
CH~
0.15
0.24
--
--
" 2
O
CI
O-
0.09
CH3
.o- - f O
mT
OH
3
CH3
6
0.14
0.285
0.125
CI
No 2° Radical
c H ~ S ~ c H 3 O
OH
CH3x
C1
OH
O-
S
OH
O-
OH
CI
OH
CI
H3 ao DOH"
CI
CI
O"
HOII 0
I
OH
0
OH
O"
OH
CI
0
CI
O
C1
\
I ° radical
OH
OH
O"
OH
0
CI
oN.
OH
Cl
2 • radical Scheme
Fig. 3. Scheme for the production of free radicals from 3,3'-dimethylfentichlor.
Radicals in photoallergens low p h o t o c h e m i c a l reactivity with other proteins such as 'y-globulin. 2° F e n t i c h l o r shows no specific affinity for any particular site on a l b u m i n and reacts photoc h e m i c a l l y with a n u m b e r of different soluble proteins.2 The results o f b i n d i n g studies using 35S-labelled 3 , 3 ' - and 4 , 4 ' - d i m e t h y l f e n t i c h l o r are shown in Table 2. F r o m these results it can be seen that 3 , 3 ' - and 4 , 4 ' d i m e t h y l f e n t i c h l o r are capable o f bonding to more than one site on human serum a l b u m i n , "y-globulin and bovine insulin. The p h o t o c h e m i c a l binding characteristics o f both d i m e t h y l f e n t i c h l o r analogues are, not surprisingly, similar to those o f fentichlor z~ and unlike those of T4CS.~7"2° In o r d e r to d e t e r m i n e whether 4 , 4 ' - d i m e t h y l f e n tichlor e x h i b i t e d any specificity at all in its photoc h e m i c a l b i n d i n g to H S A , a p h o t o a d d u c t was p r e p a r e d with an initial ratio o f 1:1 ( 4 , 4 ' - d i m e t h y l f e n t i c h l o r : H S A ) . This had a molar ratio of b o u n d 4 , 4 ' d i m e t h y l f e n t i c h l o r : H S A o f 0.91 : 1. The 4 , 4 ' - d i m e t h y l f e n t i c h l o r - H S A p h o t o a d d u c t was treated with c y a n o g e n b r o m i d e . This cleaves the H S A m o l e c u l e at its six m e t h i o n i n e residues. Elution o f the c y a n o g e n b r o m i d e digest through S e p h a d e x G - 1 0 0 yields two main fractions d e s i g n a t e d N (residues 1 123 and 1 2 4 - 2 9 8 ) and C (residues 2 9 9 - 5 8 5 ) . (Note c y a n o g e n b r o m i d e c l e a v a g e of H S A leads initially to three fragments. Two o f these fragments are of similar size and elute together. The e x p e c t e d seven fragments o f H S A are o b t a i n e d when the disulphide bridges are c l e a v e d . See refs. 1 1 and 12.) R a d i o a c t i v e counts from 35S-4,4'-dimethylfentichl o r - H S A treated with c y a n o g e n b r o m i d e showed the 4 , 4 ' - d i m e t h y l f e n t i c h l o r to be distributed equally between the C and N fragments o f H S A showing that, as with fentichlor. 2~ 4 , 4 ' - d i m e t h y l f e n t i c h l o r shows no specificity for any particular part o f the H S A molecule. Role of oxygen
W h e n fentichlor and d i m e t h y l f e n t i c h l o r ( s ) are irradiated in the absence o f o x y g e n , no free radicals can Table 2. Covalent Binding of 4,4'-Diemthylfentichlor (4 MeF) and 3,3'-Dimethylfentichlor (3 MeF) to Soluble Proteins Binding Ratios (m:m) Nonirradiated
Irradiated
Protein
3MeF
4MeF
3MeF
4MeF
HSA Human "y-globulin Bovine insulin
0.8:1 0.4:1 0.6:1
1.4:1 0.6:1 0.2:1
7.7:1 1.4: 1 2.4:1
7.8:1 2.9:1 2.5:1
Irradiationtime--30 min. Initial ratio 10:1 (photoallergen:protein).
235
Table 3. Covalent Binding of [35S]-Fentichlorto HSA: Effect of Oxygen Nonirradiated m:m
Irradiated m:m
0.18:1 0.18:1
0.80: 1" 0.60:1"
Oxygen not excluded Deoxygenated
(n = 2) (n = 4)
Initial ratio 1:1 (m: m) (fentichlor/HSA). *Statisticallysignificantdifference(p < 0.0005). be detected 6 by E S R s p e c t r o s c o p y . The p h o t o c h e m i c a l binding o f 35S-fentichlor to H S A was d e t e r m i n e d in the presence and absence o f o x y g e n . The results (Table 3) show a small but significant reduction in the photoc h e m i c a l binding o f fentichlor to H S A in the absence o f o x y g e n . This suggests that whilst the semiquinone radicals f o r m e d after irradiation o f fentichlor are capable of reacting with the protein to form adducts, the m a j o r reaction is i n d e p e n d e n t o f o x y g e n and m a y involve direct reaction b e t w e e n the p r i m a r y carboncentred radical and the protein. Acknowledgement--One of us (JND) wishes to thank SERC and
Unilever for a CASE award. REFERENCES
1. Wilkinson, D. S. Photodermatitis due to tetrachlorosalicyalnilide. Br. J. Dermatology 73:213; 1961. 2. Harber, L. C.; Harris, H.; Baer, R. L. Photoallergic contact dermatitis. Arch Dermatol. 94:255; 1966. 3. Bury, J. N. Photoallergies to fentichlor and multifungin. Arch. Dermatol. 95:287; 1967. 4. Clayton, R. From fentichlor sensitivity to actinic reticuloid? Proc. Roy. Soc. Med. 69:379; 1976. 5. Ramsay, C. A. Skin responses to ultraviolet radiation in contact photodermatitis due to fentichlor. J. Invest. Dermatol. 72:99; 1979. 6. Delahanty, J. N.; Evans, J. C.; Rowlands, C. C.; Barratt, M. D. J. Chem. Soc. Faraday Trans 1:135; 1987. 7. Dunning, F.; Dunning, B.; Drake, W. E. Preparation and bacteriological study of some symmetrical sulphides. J. Am. Chem. Soc. 53:3466; 1931. 8. Tueda, T.; Tsubui, K. Dihydroxydiphenyl sulphide derivatives. Jap. Patent 776; 1950. 9. Bradford, M. A rapid and sensitive method for the quantitization of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248; 1976. 10. Pedersen, K. O. Exclusion chromatography. Arch. Biochem. Biophys. Suppl.:157; 1962. 11. Meloun, B.; Saber, M. A.; Kusnir, J. Structure of N terminal cyanogen bromide fragment of human plasma albumin. Biochim. Biophys. Acta. 393:505; 1975. 12. Meloun, B.; Kusnir, J. Cleavage of human plasma albumin by cyanogen bromide and characterization of the fragments. Collect Czech. Chem. Comm. 37:2812; 1972. 13. Grabowski, Z. R. Photochemical reactions of some aromatic halogen compounds. Z. Physik. Chem. (Frankfurt) 27:239; 1961. 14. Pedersen, J. A. Electron spin resonance studies of oxidative processes of quinones and hydroquinones in alkaline solution. Formation of primary and secondary semiquinone radicals. J. Chem. Soc. Perkin Trans 2:424; 1973. 15. Buettner, G. R. Spin trapping: ESR parameters of spin adducts. Free Radicals Biol. Med. 3:259; 1987.
236
J.N. DELAHANTYet al.
16. Li, A. S. W.; Chignell, C. E Spectroscopic studies of cutaneous photosensitizing agents. Photochem and Photobiol. 46:445; 1987. 17. Rickwood, D. M.; Barratt, M. D. Evidence for a major strong binding site for tetrachlorosalicylanilide on human serum albumin. Photochem. Photobiol. 35:643; 1982. 18. Rickwood, D. M.; Barratt, M. D. Identification of the major covalent binding site for tetrachlorosalicylanilide on human serum albumin. Photobiochem. Photobiophys. 5:365; 1983.
19. Barratt, M. D.; Rickwood, D. M. Binding of a spin-labelled photoallergen to human serum albumin. Biophys. Chem. 19:69; 1984. 20. Kochevar, I. E.; Harber, L. C. Photoreactions of 3,3',4',5tetrachlorosalicylanilide with proteins. J. Invest. Dermatol. 68:1512; 1977. 21. Rickwood, D. M.; Barratt, M. D. Photochemical reactions of fentichlor with soluble proteins. Chem. Biol. Interactions 52:213; 1984.