. K . ll. . _ V Lava; ¥§~ufi§rfiu¥lliilfinfl§i£aid§ . .rt'.‘ ~.fv.)kz.¢.€.f?i¥! V tolvggtvl. ..‘ , . "amt... {Keiufin.vitvk15v§§flvti$oc¥!duh 43.1, :5 . .. {1:7 .55 1 13.: ; .. _ . . . . lfihk‘..¥3§§4i¥$v¥1unto¥» . I... . : .QVbhxgfibfi: .135...» {9.4 .: V _. ,, r . . V... fink.“ Ira! 9:39"..- t't‘v-‘Ilogfi , . (1 «3.30153 9....» 13“ . via tr .3. fi 3 $31! axial .1 V . s . . m V. V1.17 .. V V km“. 1.‘ 314.....15“: 2.1 .0 I . I Ii v! o. . .. \s. 51.5}... 10.... ”010.....3 x...) 1 Va“ 30. Is. HM “5. . . .,.. .. . . \ i V...) (V V ‘‘‘‘ 1 . t1. 3. 3. . at .Ii . . 3.21... 135.5. 2. x5 32.} a... s . b ‘ \V I". x... V . 5... 1.22.”, t . a . . X). .r I. .Vn. I t; .724... . .. . . VV . . a s a.) .816. 5...... lhxhzfifiilaw « an... .1... a. z. . .53 V 31:25.. r1..unuu..¢u.t.hnfiV 35$... . . . V . . .. . .V. 6...}. m2? .. V5 .Iéufififimfioa .3. 23.2%.. 3% 5?. £9.28 a 2.; .fi.r..V....s..:2.V.rzuu.. 11:32- . V . a 3. . . V .1 3 VV ..\ .. ... \ . . .J u, x. t u 1. . . . . . 1 V3 V I. ,0 :23}... It. . . .. V .. . . . n 4.52.5.3,» h... . \:._...H x. .f....z. .03.“.1erfiuax... 2,. In... . . . . 1 . it I... it 1 a V 4.. x .V .. .. _. )AHHW\: \Trxux WW. . . I .. . . . .V V. H~vamnkngh$flcx,1(.a .3.er :0 gag)“. . . 1‘ .\ . \ . . iv... A .\ a 1. V . . it, . r . 1 . )1}: a) QR. . ~ . ; km”... 3%!- T J)... 3‘ . 11“ . 1 V V. 04,5.mm»..,siuuu.3a¥n . . . 13...... .52 XX?“ 5. y: . .. . . (I: .MVWXV\\ r 1: V . a . :11. A51).u-\1\kanow. Y x 0.. . a. .. .u. b : . 1.1.x». Rhafibxfimmt. . 2.. . h... 135%:ufiimwofivrfi x s. V} . ., . .xmnfifixsn .. .s 93$: . t u... . . , t hi... . .V.mnhphba.me.un%) , . 24 S.-. .. 5.4 . . .1}....i§..fii§§pxh u. . a 25153 K)... 3:9. wt . I. 2 6» i. )5 If 3x3 a hmwfllrsbiuxxzfil V 1. .1. 0: I V . IV. 1 i. . 31...... .. ... .: V. .. 1 kw. V... ..V. , V. A... .4. .8 is. V x . . . . 2 . A. . . infiflaé \ . . . 3315...} .5». . .. . x ....| . .9312... iii...) , . . . V .1 k\.1..... \>\I\\7... .. V . - 32:}... 311.13 3H1:H111k.fi}.1\a\Vq../}\\\\.\\p . . V .. V k 9\\V\NNK5.A.I .i1x\I..!uw\~V.\.1$1.Vu:sz x8:3~.v€57~.1\|2.. Vfifi. )7" V 333 . 153‘ 3‘ \\)}.\l. .v‘fiMflpz‘suq‘ a ism-kl. V at )3 “VA-I6: (its avvsvxalx . . .l .. t Oasis! 00 I . i, . II J. V V . . ‘Aitgflfpi v! a . . , . . 1 . r :. FOLK V “in... V V . VV . \FI . q .. .31. 2m“: s . A $1....1‘57flvx111A\1\?€N.15\“1.2.P.fl\..12.x3?&u{11§. \. . 3.5!...» 53.1\~\r.\1\£.. ... )1: 1532, 21‘... \AS\\ 3.53.51 ._ 11\\\d..\.‘SV: 1.1!}. «an... I i V 11“ . \figkh 5.35)); 1% 3) an.“ )383‘1. .58“: IV. thLYFVHVfiUHHHmhoitsaqddW‘mkfiui Lhfikfi I run”. 3|}: . V V . V. \ ,lii 2))! .4 V. . - k . Y I .71 . -111! . ..; €094.15 ‘t If- it! v: m} 1”,“ u‘. 'l. ' ' ‘. built I 7. '0? 1L E‘Ev‘fi‘g‘ in Iltbflurq-..alnutfi¢li :O‘S‘Digfifl I a )3! 3. . . t 305 TVrAA. I its: nivtv .- .!bfii 0.90.. \szkunta t.3333\.u\23.«.fl1\. 5.1VV..\V.} ‘72}... SitQflfix ‘ i. “32:. get-oiling... iv» in . . alluu. \)~.x\>\5\1-14.fi7? \S_»;3>\l\..w)1\51 ‘31.}... \I-x\\\1n;‘.,\.r.u.)auwv)? 3.3. V x 1...... Mn” 4|»..vrf. :C 2.775.511 . . . L) Ii . . \ a. . «cthkflw‘xi‘xkx\1\x\1rllvxyx§ i513. i)..%.fl\3fin(}u.2fi \k. 2.11.. is.\). . ,Klwtuuuuwxs. I. h!‘)»iv-§t‘( HI- lvéfiri‘ MV’V . . .. V..}\\Aadfln‘ \l\.\~\«.\.~(. 5.12.9.1“: NSZI1NE55EI‘35. «3.5557! . udthktvfiw. ax. k. ,73‘. . 31:... 3 z...r..rn\ 1%.".- .d 1 {81. .‘tlni‘i. ratti. . Lhrfvvvgrég .V\)3‘.\S\$i\ \x. )1}; 3.. \V 51...}.3... . x.\.\ 151.7\é%\fl{ ‘nxfiifi .3...\Qh.\l)x...13315|$3‘ 3:. ill... 31 V IV Infill.{.gr0‘ -. . .2 5‘ . .. \).¢h‘\!owl..nu (9.1.7.111... A») 1\§\V\. 08$" 31.. \t... HRS) §§abfi\g.\fix\. \Sguhlflfilifiz s. 3%. \- )zxg 1| .il “135"...9-‘tli: 1r yer-ti. I?! x. . . ‘1Juu. 3:1. 1x73. ..x\)2. )3)..ng .. 11 131‘“ Ax . . ..\~ 3.3.1.... 1 1 13.1.1; 410:. . )1} 3 1 Qt1‘711|'.vefil: \ ‘L w. R. .111 «t1\..l.. . )8!s,....31.1:33i\\.2r V. NEH}... 1.1»). fly-‘3‘. 1 ....)\‘(.\uv\llLfi\3fisnzn)1.u£ :7 )3}. r 173.. :i‘i} 0‘ i.¢9vv.i.-.l.vivta .lstvvoln § . “a". gii ...V \filx)..vs\)5\ a... \\.\\.\11~.\\§.~.\.v5v\l)|531\3 “733571....53 121|111¥X|§Si 1 VV\ ‘ V ‘5‘): «a. . H 73K...) 3 a. 0.15 IT! ‘36,? Inn-L. tuigsi" girl-3.5.1.35 5.431 lin<§g~.1fi\.rl7 .\\1\V\\.31§..1\: ..‘ftpV\§§.ri».\¥IA . . .. .tltizl. iii... \ . V I: .. : .u . 2.. O... . zigzt.-. 3LV,;..|‘.... {lit .L.r.{.$ .. .I 1 ..I~.’th.-.‘I4§-4v§t,’ "\ Y E.\.§v,..1. unit“. 329...}..1.‘ 1.2.1.1131)... Elli; . 1M». {5?}..‘V V { iii}: CL .A‘..5!93~l..£c. Prf........n: . . V...v.v...V“..1:-... \?Trv. THESIS We? rm: ;~;-'. ~ rt?" 3;. ‘ :fr. . ‘9 i, / \— (o .._::—- W“:\r—"- -~---~-«- - - This is to certify that the dissertation entitled STUDIES ON THE RELATIONSHIPS BETWEEN THE CHEMICAL AND PHARMACOTOXICOLOGICAL PROPERTIES OF POLYBROMINATED BIPHENYLS presented by Ghazi Abdo Dannan has been accepted towards fulfillment of the requirements for Ph.D. degree in Biochemistry .M Major professor Date May 17, 1982 MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES .525... \’ RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. STUDIES ON THE RELATIONSHIPS BETWEEN THE CHEMICAL AND PHARMACOTOXICOLOGICAL PROPERTIES OF POLYBROMINATED BIPHENYLS By Ghazi Abdo Dannan A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Biochemistry 1981 ABSTRACT STUDIES ON THE RELATIONSHIPS BETWEEN THE CHEMICAL AND PHARMACOTOXICOLOGICAL PROPERTIES OF POLYBROMINATED BIPHENYLS By Ghazi Abdo Dannan Separation techniques have been developed to purify sufficient quantities of most of the P88 congeners of Firemaster. These included the preferential solubilization and crystallization of crude fractions, alumina adsorption column chromatography, and reversed-phase chromatography on a Lipidex-SOOO column and on Kieselguhr-G thin layer plates. The structures of several congeners were determined using mass spectrometry and lH-NMR spectroscopy. According to their sequence of elution from the gas chromatograph the major PBB congeners of Firemaster, including seven previously identified congeners, are 2,4,5,2',S'-penta-, 2,4,5,3',4'-penta-, 2,3,6,2',4',5'-hexa-, 2,4,5,2',4',5'-hexa-, 2,3,4,2',4',5'-hexa-, 2,4,5,3‘,4',5'-hexa-, 2,3,4,5,3',4'—hexa-, 2,3,4,5,2',4',5'-hepta-, 2,3,4,5,Z',3',4'-hepta-, an octa-, an octa-, and 2,3,4,S,2',3',4',S'-octa-bromobiphenyl (congeners I-Q, respectively). The degree of bromination at carbons 95339 to the biphenyl bridge seemed to determine their chromatographic and UV-absorption spectral properties, as well as their three-dimensional crystal structure. The extent of planarity in the biphenyl system appeared inversely related to the number of M bromines. Three mono-ortho brominated congeners, _2_, g, and _7_, seemed to be have intermediate planarity between that of the 9m unsubstituted 3,4,5,3',4',S'- hexabromobiphenyl (H88) and the di-o_r_t£g brominated congeners exemplified by congener fl. Congeners _l_, _2_, 2, Q, _7_, 2, and _l_2_ as well as HBB were evaluated for their acute effects on rat liver microsomal drug metabolizing enzymes and for certain toxic effects that are typically caused by tetrachlorodibenzo-g-dioxin (TCDD). Congeners Z, _5_, g and Z caused a variable 3-methylcholanthrene (MC)- and phenobarbital (PB)-type induction of microsomal enzymes and caused at least one TCDD-like toxic response. The grth_o unsubstituted HBB, however, was at least 10-20 times more similar to TCDD than any of these congeners. Congeners I, 2, and l_2_, like the two major congeners g and g, were strict PB-type inducers and relatively non-toxic. The biological effects of P88 congeners seemed to be determined by their chemical properties, particularly by the extent of planarity. Firemaster BP-6 (lot number 6224-A) and Firemaster FF-l (lot number FH7042) containing nearly 25% and 15%, respectively, of congeners _2_, _5_, _6_ and _7_ caused a similar PB-type induction but the first was a significantly more potent MC-type inducer. A reconstituted PBB mixture formulated to resemble Firemaster BP-6 evoked nearly an identical PB- and MCI-induction effects to that of Firemaster BP-6. Congeners _2_, _5_, g, and Z are, therefore, responsible for most or all of the TCDD-like effects of crude Firemaster. TO ALL THE PEOPLE WHO ASPIRE FOR KNOWLEDGE BUT HAVE NO MEANS TO AFFORD IT. ACKNOWLEDGEMENTS Iconsider myself extremely fortunate to have worked so closely with so many people both in Dr. Aust's laboratory and elsewhere in Michigan State University. I wish to thank Professor Steven D. Aust for his guidance and support throughout the course of my graduate studies. Drs. S. Ferguson-Miller, Pamela J. Fraker, David McConnell, Stuart D. Sleight, and John E. Wilson are also gratefully acknowledged for serving as members of my guidance committee. I also consider myself fortunate to have collaborated with Drs. Stuart Sleight, Janver Krehbiel and Pamela Fraker on several aspects of this research. A more detailed account of their contributions may be found in the Appendix. Several other people including Esther Roege (electron microscopy studies), Betty Baltzer (mass spectrometry studies), Frank Blenis (lH-NMR spectral studies), Rick Jensen (parts of the histopathology studies), and Jennie Kung and Don Ward (crystal structure studies) have provided data and results that have become part of this research. I would like to also acknowledge the members of this laboratory for their friendship, advice, and assistance. I am particularly indebted to Joe Buck, Gerald Mileski, Cynthia Millis, Ming Tien and most of all Debra Metcalf for providing valuable assistance and suggestions. iii TABLE OF CONTENTS LIST OF TABLES . LIST OF FIGURES LIST OF ABBREVIATIONS . INTRODUCTION . LITERATURE REVIEW Hepatic Microsomal Drug Metabolizing Enzymes . Induction of Hepatic Microsomal Drug Metabolizing Enzymes Consequences of Microsomal Drug Metabolizing Enzymes Induction . . . A Model for the Coordinate Enzyme Induction and Toxicity in Response to Halogenated Aromatic Hydrocarbons and Polycyclic Aromatic Compounds-"The TCDD-Receptor" Polychlorinated Biphenyls and Structure-Function Relationships for Induction of Microsomal Drug Metabolizing Enzymes . Polychlorinated Biphenyls and Structure—Activity Relationships for Toxicity and Binding to the Cytosolic Receptor . Induction of Drug Metabolizing Enzymes by Polybrominated Biphenyls . Attempts to Characterize the Inducing Agents in the Crude PBB Mixture of Firemaster. . . . PBB Metabolism: Structure- Function Relationships. Toxicity Due to Polybrominated Biphenyls. CHAPTER 1: STUDIES ON SOME CHEMICAL CHARACTERISTICS OF POLYBROMINATED BIPHENYL CONGENERS . Abstract Introduction. Materials and Methods Materials Gas Chromatography (GC). . . as Chromatography- Mass Spectrometry (GC— MS). H- NMR Spectrometry . . Thin Layer Chromatography (TLC) Acetonitrile Partitioning. UV-Absorption Spectroscopy . Purification of PBB Congeners 15 18 26 30 31 32 34 I. A General Scheme for Purifying Congeners_ l, 2, 4, 5, 6, 7, and 9.. Fractionation of Firemaster BP- 6 by Extraction and Recrystallization in Acetone . . . . . . Neutral Alumina Chromatography . Recrystallization in Hexane Lipidex- 5000 Reversed- Phase Chromatography II. Purification of Congeners_ 3, 6___a, 6b, and l_0 Results. . . . . . . Purification of PBB Congeners Purification of Congeners 4 and 6 . Purification of Congeners 5 and 9. Purification of Congeners_ 1, 2 and 7. Characterization of Several PBB Congeners Some Chemical and Chromatographic Properties of PBB Congeners . Discussion CHAPTER 2: LIVER MICROSOMAL ENZYME INDUCTION AND TOXICITY STUDIES WITH BROMINATED BIPHENYL CONGENERS: RELATIONSHIP BETWEEN TOXICITY AND O__RTHO BROMINATION Abstract Introduction. Materials and Methods Materials Preparation of PBB Congeners Animals . . Pathological and Ultrastructural Evaluations . Clinical Evaluations. . . . . . . Jerne Hemolytic Plaque Assay Preparation of Microsomes. Enzyme Assays . . Preparation of Ethyl Isocyanide. . SDS- Polyacrylamide Gel Electrophoresis. Purification of Epoxide Hydrolase. Results. Effects of 2, 4,5, 3', 4', 5'— Hexabromobiphenyl (Congener 6) and 2, 3, 4, 5, 2', 3', 4', 5'— Octabromobiphenyl (Congener 1__2) . Effects of 2,4,5,3',4'-Pentabromobiphenyl (Congener 2) . Effects of 2,4,5,2',5'—PentabromobiphenyI (Congener 1), 2,3,4,2',4’,5'-Hexabromobiphenyl (Congener 2), 2, 3, 4, 5, 3' ,-4' Hexabromobiphenyl (Congener Z), and 2, 3, 4 ,,5 2', 3', 4'— Heptabromobiphenyl (Congener 9). Discussion . CHAPTER 3: RECONSTITUTION OF SOME BIOCHEMICAL AND TOXICOLOGICAL EFFECTS OF COMMERCIAL MIXTURES OF POLYBROMINATED BIPHENYLS. Abstract Introduction . 111 112 113 116 116 117 117 120 121 121 122 122 126 126 127 128 128 132 151 169 194 195 196 Materials and Methods . Materials Animals . Preparation of Microsomes . Enzyme Assay Results. . . . . Discussion GENERAL DISCUSSION . REFERENCES APPENDIX: LIST OF PUBLICATIONS . vi 198 198 201 201 201 202 207 211 225 236 Table 10 ll 12 13 14 15 LIST OF TABLES Distribution of PBB Congeners in Various Fractions Chemical Shifts and Possible Structures of Partially Characterized PBB Congeners . . . Relative GC Retention Times ”(t R) and ECD- -Responses to Pure PBB Congeners Relative Abundance of PBB Congeners in Firemaster (FF-l and BP—6) and in a Reconstituted Mixture Made from Purified Congeners . . . . . . . . . . . . . Reversed-Phase TLC of PBB Congeners UV-Spectral Data of PBB Congeners . X-Ray Crystallography Studies of PBB Congeners . Some Liver Microsomal Drug Metabolizing Parameters of Rats Administered Congeners 6 or _1_2_ or Other Treatments . Some Liver Microsomal Drug Metabolizing Parameters of Rats Administered Congener 2 or Other Treatments . . . Some Physiological and Clinical Parameters from Rats Administered Congener Z or Firemaster FF-l . . . . . . . . . . Antibody Mediated Response of Rats to Sheep Red Blood Cells Two Weeks after Exposure to Firemaster FF-l or Congener 2 Some Liver Microsomal Drug Metabolizing Parameters of Rats Administered Congener_ 2 or HBB. . . . . . . . Body and Organ Weights of Rats Administered Various PBB Congeners . Some Liver Microsomal Drug Metabolizing Parameters of Rats Administered Various PBB Congeners . . . . . . . . Competition by Brominated Biphenyl Congeners for the Specific Binding of[ [H]- TCDD to Hepatic Cytosol . . . . . vii Page 66 84 86 89 91 98 . 108 . 129 . 135 . 138 . 145 . 147 . 152 . 167 . 190 16 Some Liver Microsomal Drug Metabolizing Parameters of Rats Administered Various Mixtures of PBB . . 17 Body and Organ Weights of Rats Administered Various Mixtures ofPBB. . . . viii . 203 . 205 LIST OF FIGURES Figure Page 1 Gas Chromatographic Elution Profile and Structures of Polybrominated Biphenyls . . . . . . . . . . . . . . . . 56 2 General Outline for PBB Congeners Purification from Firemaster BP-659 3 (a) Gas Chromatographic Elution Profile of F-ML3 before Neutral Alumina Chromatography . . . 67 (b) _ Neutral Alumina Chromatography Elution Profile of Certain PBB Congeners of F— M—L3. . . . . . . . . . . . . . 67 4 (a) Gas Chromatographic Elution Profile of Congener l- Enriched Alumina Column Pool (Fractions 25- 30) . . . 70 (b) Separation of Congener 1 from Congeners_ 2, 4, and 6 by Lipidex- 5000 Chromatography. . . . . . . . . . . 70 5 (a) Gas Chromatographic Elution Profile of Congener_ 2 before Lipidex- -5000 Chromatography. . . . . . 73 (b) Separation of Congener_ 2 from Congener_ 6 by Lipidex- 5000 Chromatography . . . . . . . . . . . . . . 73 6 1H- NMR Spectrum and Structure of 2, 3, 6, 2' 4,', 5'— Hexabromo— biphenyl (Congener 3).. . . . . . 77 7 1H- NMR Spectrum and Structure of 2, 3, 4, 5, 3', 4’— Hexabromo— biphenyl (Congener Z). . . . . . . 80 8 1H- NMR Spectrum and Structure of 2,3, 4, 5, 2' ,3', 4'- -Heptabromo— biphenyl (Congener 9). . . . . 82 9 Thin Layer Chromatographic Separation of Penta—(PBB) and Hexa-(HBB) Bromobiphenyls . . . . . . . . 92 10 UV- Absorption Spectra of 3, 4, 5, 3', 4', 5'- -Hexabflromobimphenyl (HBB) and Congeners_ 7, 6, 2, and 5 . . . . 95 11 UV-Absorption Spectra of Congeners Z, 4, 6, and E. . . . . . . . 99 12 Gas Chromatographic Elution Profiles of All Purified PBB Congeners . . . . . . . . . . . . . . . . . . . 118 13 Effects of Congener 6 on Rat Liver Ultrastructure . . . . . . . . 133 ix 14 15 16 17 18 19 20 21 22 23 Effects of Congener 2 and Firemaster FF-l on Rat Liver Structure . Effects of Congener 2 and Firemaster FF- 1 on Rat Liver Ultra- structure . . . . . . . SDS- -Polyacrylamide Gel Electrophoresis of Liver Microsomal Proteins From Male Rats Administered Congener_ 2 or Other Pretreatments . . Effects of Congenersl, 5, Z, or9on Rat Liver Structure Seven DaysAfterTreatment. . . . . . . . . . . . . . . Effects of Congeners 1, 5, Z, or _9_ on Rat Liver Ultrastructure . An Electron Micrograph of a Portion of a Hepatocyte from a Rat Administered Congener Z. Effect of Congener Z on Rat Thymus Structure. SDS-‘Polyacrylamide Gel Electrophoresis Profile of Liver Microsomal Proteins from Rats Administered Congeners l, 2! Z, or 2 Effects of Various PBB Congeners on the Profile of Epoxide Hydrolase and Other Proteins of Microsomes Subjected to SDS- Polyacrylamide Gel Electrophoresis. Gas Chromatographic Elution Profiles and Structures of Polybrominated Biphenyls in Firemaster FF-l and a Mixture Reconstituted with Pure Congeners 140 143 149 153 158 162 165 170 178 199 AHH butylated hydroxytoluene DNA 8 ECD EtNC FAD FMN GC H and E HBB Hz iP MC NADH NADP+ NADPH NMR PB PBB LIST OF ABBREVIATIONS aryl hydrocarbon hydroxylase activity 2,6-ditert-butyl-p—cresol deoxyribonucleic acid extinction coefficient electron capture detector ethyl isocyanide flavin adenine dinucleotide flavin mononucleotide gas chromatography hematoxylin and eosin 3,4,5,3',4',5'-hexabromobiphenyl Hertz (cycles per second) intraperitoneal coupling constant 3-methylcholanthrene B-nicotinamide adenine dinucleotide, reduced form B-nicotinamide adenine dinucleotide phosphate B-nicotinamide adenine dinucleotide phosphate, reduced form nuclear magnetic resonance phenobarbital polybrominated biphenyls xi PCB PEG Ppm RER RNA SD SDS SEM SER TCDD TLC Tris UDP UV polychlorinated biphenyls polyethylene glycol parts per million relative thin layer chromatographic mobility rough endoplasmic reticulum ribonucleic acid standard deviation sodium dodecyl sulfate standard error of the mean smooth endoplasmic reticulum retention time 2,3,7,8—tetrachlorodibenzop-dioxin thin layer chromatography tris (hydroxymethyl) aminomethane uridine-5'-diphosphate ultraviolet INTRODUCTION In mid 1973, an error was made in the process of preparing a pelleted form of cattle feed at a grinding feed mill operated by the Farm Bureau Services, Inc. at Battle Creek in Michigan. Between 500 and 1000 pounds of polybrominated biphenyl (PBB), a fire retardant mixture closely related to polychlorinated biphenyl (PCB), was mistaken for a feed supplement (magnesium oxide) and as a result was mixed with cattle feed. This mix-up, however, may have originated at the Michigan Chemical Corporation at St. Louis, Michigan, where both PBB and magnesium oxide were manufactured and commercially marketed as Firemaster and Nutrimaster, respectively. Both products had a similar physical appearance but were normally packaged in pre-printed bags carrying their trade mark names. Due to a temporary shortage in the pre-printed bags the two products were packated in plain brown bags with their trade names stenciled at the top. The details of the mix-up at the plant are not completely understood, but a few bags of Firemaster made their way to the feed mill operated by the Farm Bureau Services. The discovery of the mix-up was not made until the end of April 1974 when, through the perseverance of a Michigan dairyman by the name of Fredric Halbert, mysterious cattle ailments were linked to PBB poisoning. Thousands of cattle, hogs 30d sheep in addition to over 1.5 million chickens and large amounts of feed, and various dairy products had to be removed from the market and destroyed. The economic losses obviously ran into the millions of dollars. Several thousand farm families and consumers, who purchased dairy products directly from contaminated 2 farms, consumed relatively large amounts of PBB. By some estimates nearly 90% of Michigan's residents have measurable levels of PBB in their bodies. Health problems associated with PBB consumption are somewhat variable and far from being completely understood. Some reports have cited some neurological disorders, lethargy, stomach discomfort, and musculoskeletal symptons among other ailments felt by the afflicted farm families. However, these and other human related illnesses could not be unambiguously linked to PBB ingestion since many of those people might have been emotionally and psychologically distressed over their farm losses or their fear of the unknown effects of the chemicals in their body. Therefore, no cause and effect relationship have thus far been established between PBB and any damage to human health. Any long-term PBB related toxic and health effects may require decades before they may become apparent. Of immediate concern to the public of Michigan, however, is the possible transfer of PBB to babies nursing from contaminated mothers. Normal growth and development could seriously be affected by PBB. Biological effects due to the crude mixture of PBB have been widely examined in the rat. Among these effects is the induction of hepatic microsomal drug metabolizing enzymes in a manner similar to the combined administration of the two prototype inducers, phenobarbital (PB) and 3—methylcholanthrene (MC), a phenomenon commonly referred to as "mixed-type induction." These effects are also similar to those caused by certain PCB mixtures commonly known as Aroclors. Among various halogenated aromatic hydrocarbons, however, 2,3,7,8-tetrachloro- dibenzo-p—dioxin (TCDD) is believed to be the most potent MC-type inducer of the microsomal enzymes. However, TCDD does not cause any PB-type induction of mic rosomal enzymes. TCDD also evokes a number of characteristic toxic effects that include atrophy of the lymphatic organs, immunosuppression, body weight loss, loss of appetite, unique Ultrastructural changes in the liver, porphyria and 3 eventually death. The extent of most of these effects are species and sex dependent. All of these effects, including the MC-type induction of microsomal enzymes, are believed to be mediated by a liver cytosolic protein that possesses a high binding affinity for TCDD, and as a result is commonly referred to as the "TCDD-receptor." Like TCDD, the Firemaster mixture can also cause most of these effects. On a molar basis, however, TCDD may be at least 104 times more effective than the PBB mixtures. This should still call for concern since large amounts of PBB have made their way to the environment, and since some people may have consumed relatively large quantities of PBB in their diets. Environmental and medicinal agents that stimulate the microsomal drug metabolizing enzymes are commonly believed to cause serious pharmacological and toxicological changes as well. Among these are alterations in the i_n vi_v0 metabolism of endogenous compounds such as steroid hormones, fatty acids, vitamins, bilirubin, and thyroxin where, as a result, serious physiological and biochemical complications may arise. Other complications due to microsomal enzyme induction may include i_n M modifications of the toxicity and carcinogenicity of environmental chemicals. Similar changes and complications are likely to be caused by PBB. Polybrominated biphenyls are highly stable and resistant to microbial or sunlight degradation, and therefore environmental contaminations due to their manufacture, industrial handling or disposal are likely to be a threat for future accidents involving human exposure. Firemaster, as such, is a powder or granular material that is water insoluble but is reasonably soluble in organic solvents and fat. The various isomers and congeners of the mixture are expected to have widely different chemical properties. Most of these differences, as will be shown later (see Chapter 1), seem to be mainly influenced by the number of bromines and their arrangement on the biphenyl nucleus. 4 Analysis of Firemaster by gas chromatography reveals twelve major components each of which has a concentration of nearly 0.5% or more of the total mixture. Most scientific research has been performed with the original mixture of Firemaster, and as such very little or no information is available on the biochemical and toxicological effects due to the individual isomers and congeners of the mixture. In this laboratory attention has been mainly devoted to studying the chemical and biological properties of the individual PBB components of Firemaster. First of all, most of these components were purified and their structures were determined. A large portion of this task was accomplished by Robert Moore who also characterized the microsomal enzyme induction due to the two most abundant PBB congeners in the mixture. The available techniques, however, were inadequate for purifying relatively large quantities of the remaining congeners for performing similar studies. As a result those studies had to be put off. The first objective of the research described in this thesis was to devise methods that can be used to purify sufficient quantities of highly pure congeners. Then, the structures of the yet unidentified PBB congeners were determined and certain chemical and chromatographic properties of most congeners were examined and compared. Most of these properties were shown to be mainly influenced by the total number of bromines and the extent of bromination M to the biphenyl bridge. These studies are all presented in Chapter 1 of this thesis. In Chapter 2, seven Firemaster PBB congeners were evaluated for their effects on rat liver microsomal drug metabolizing enzymes, and on the size and structure of certain organs particularly liver and thymus. The main purpose for these studies was to identify the congeners of the PBB mixtures responsible for toxicity and MC-type induction. Previous studies on Firemaster have shown no traces of brominated dibenzofurans or dibenzo-p-dioxins, two classes of compounds _k— 5 that are expected to be highly toxic and potent MC-type inducers. Few PBB congeners were capable of causing some of these effects. Relationships and correlations between PBB structure, TCDD—like toxicity, and microsomal enzyme induction are also considered. In Chapter 3, nine purified PBB congeners totalling nearly 96% of Firemaster were reconstituted to formulate a Firemaster-like mixture, which after being administered to rats caused biochemical and toxicological effects very similar to crude Firemaster. Thus, most biological effects of Firemaster can be accounted for by the PBB congeners in the mixture. LITERATURE REVIEW A comprehensive literature review will be presented on the nature of the microsomal drug metabolizing enzymes and their induction by various chemicals with special reference to polyhalogenated aromatic hydrocarbons. This review will also cover the most widely accepted mechanism by which certain polyhalogenated aromatic hydrocarbons cause a characteristic toxic response and an MC-type induction. - In this regard, the role of a liver cytosolic binding protein, commonly known as the "TCDD-receptor," will be emphasized. Hepatic Microsomal Drug MetabolizinlEnzymes In the process of tissue homogenization, membranes of the endoplasmic reticulum break up into fragments which form vesicles of closed membranes. These spherical sacs are commonly referred to as microsomes. Microsomes can be isolated by differential centrifugation of the homogenized tissue (Claude, 1969). Microsomal drug metabolizing enzymes are present in numerous mammalian tissues, even though they are most abundant in liver. Hepatic microsomes will therefore, be the main subject of the following review. Microsomal drug metabolizing enzymes are organized into an NADPH- dependent electron transport chain which requires the cofactor NADPH and molecular O2 to metabolize a wide range of substrates of exogenous or endogenous origin. At least three components are required for the proper function of the microsomal electron transport chain (Strobel _e_t _e_I_l_., 1970). The first component is a flavoprotein containing an FMN and an FAD per subunit of enzyme (Iyanagi and 7 Mason, 1973), and is commonly referred to as NADPH-cytochrome P-450 reductase. This enzyme accepts electrons from NADPH, then passes them to a terminal group of oxidases (the second component), collectively known as cytochrome P-450 (Lu e_ta_l., 1969). Since the reductase can also pass on electrons to exogenous cytochrome c it is often referred to as NADPH-cytochrome c reductase (Gillette e; a_l., 1972). Based on substrate specificity studies, and electrophoretic and immunochemical evidence, it is commonly believed that multiple forms of cytochrome P-450 are present in liver microsomes from experimental animals (Walton and Aust, 1974; Thomas e_t_ 61., 1976; Warner e_t_ 91., 1977; Guengerich, 1977; Guengerich, 1979; Lu, 1979). These forms of related enzymes are occasionally referred to as microsomal hemoproteins (or cytochrome P-450 hemoproteins) since each subunit contains a single heme prosthetic group. (The term cytochrome P-450 is usually used to collectively refer to the various hemoprotein forms). The association of NADPH-cytochrome P-450 reductase with cytochrome P-450 hemoproteins is commonly referred to as the microsomal mixed- function oxidase system (Mason, 1957; Conney, 1967; Gillette e_t_ a_l., 1972). The proper function of this system is thought to be dependent on a third component which was found to be phosphatidyl choline (Strobel E §ln 1970). Certain non-ionic detergents could substitute for this lipid fraction, and either of these two is thought to function by enhancing the interaction between cytochrome P-450 and the reductase (Lu 6t 61., 1974). In its reduced state (sodium hydrosulfite, N825204’ may be used as an artificial electron donor), cytochrome P-450 binds carbon monoxide to form a complex which shifts its absorption spectrum maximally at 450 nm. The amplitude of the reduced carbon monoxide difference spectrum is the basis for quantitating this enzyme (Omura and Sato, 1964). _—————: 8 A substrate is normally bound to oxidized cytochrome P—450 whose iron is then reduced before oxygen binds to the reduced enzyme-substrate complex. This activated oxygen complex is reduced further before it finally dissociates into hydroxylated product, water and oxidized cytochrome. These events are summarized in the scheme shown below. ROH RH H20 Fe+3 2H+ (RH) Fet3— 0;] (RH) Fe+3 e (RH) [petiog] (RH)Fe+2 (R H)[Fe2-02] (Modified from White, R.E. and Coon, M.J. (1980). In Annual Review of Biochemistry, Vol. 47, pp. 315-356.) In this overall process one atom of oxygen (from Oz) is incorporated into the Substrate as a hydroxyl group or an epoxide, and the second oxygen atom is combined with two hydrogen atoms to form water (Mason, 1957; Posner g a_l., 1961; Gillette g a_l., 1972). ¥ L, 9 Various body constituents (endogenous compounds) and xenobiotics (exogenous foreign compounds) are metabolized by the same hepatic enzymes of the endoplasmic reticulum. Among the endogenous substrates are heme, fatty acids, steroids, vitamins, and bilirubin (Gillette, 1966; Wade g a_l., 1968; Tephly and Mannering, 1968). Several hundred xenobiotics including halogenated hydrocarbons, insecticides, urea herbicides, polycyclic aromatic hydrocarbons, dyes, volatile oils, nicotine and other alkaloids, food preservatives, besides compounds such as safrole, xanthines, flavones, and organic peroxides that occur in food are metabolized by the same enzymes (Kappas and Alvares, 1975; Conney and Burns, 1972; Kuntzman, 1969). The multiplicity of the hemoproteins and their differing substrate specificities (Guengerich, 1977; Coon g 61., 1977) may, at least in part, account for the ability of the drug metabolizing enzymes to metabolize so many diverse substrates. Metabolites of most xenobiotics and endogenous compounds, produced by the enzymes of the mixed-function oxidase, may in turn serve as substrates for a number of additional enzymes. Epoxides, for instance, can be hydrated by epoxide hydrolase to grins-dihydrodiols (Oesch, 1973; Lu and Miwa, 1980) which become less reactive than their parent compounds. Most epoxides are highly unstable, and since they are electrophilic in nature they can react irreversibly with nucleophilic groups of cellular macromolecules (protein, DNA, RNA) to generate adducts. It is generally thought that such modifications may play an important role in chemical carcinogenesis (Heidelberger, 1975), even though the exact nature of this role remains unclear (Grunberger and Weinstein, 1979). It is commonly believed that epoxide hydrolase plays a protective role by converting the highly reactive epoxides to the appreciably more inert dihydrodiols (El-Tantawy and Hammock, 1980). Contrary evidence however, has been reported, in at least one instance, Where epoxide hydrolase seemed to enhance the deleterious effects of 10 benzo(a)pyrene-7,8-oxide. This compound is hydrolyzed by epoxide hydrolase to the corresponding dihydrodiol, which is further metabolized by the mixed-function oxidase system into the highly reactive, mutagenic and carcinogenic 7,8-dihydroxy- 7,8-dihydrobenzo(a)pyrene-9,10-oxide (Wood SE a_l., 1976b; Newbold and Brookes, 1976; Buening _ei a_l., 1978; Osborne, 1979). While epoxide hydrolase had a toxication role in the formation of this compound it may also be detoxifying since in reconstituted enzyme systems and in cell cultures, it reduced the level of DNA binding by the same diol epoxide (Gozukara it. Sin 1981). Therefore, depending on the particular metabolic pathway, epoxide hydrolase may hydrolyze a particular epoxide chemical, such as an arene oxide, into a more or less noxious diol derivativer Epoxides may also be metabolized into glutathione conjugated products by a family of cytosolic enzymes commonly known as glutathione-S-epoxide trans- ferases (Sims and Grover, 1974; Jerina and Daly, 1974). These enzymes are commonly thought to play a protective (detoxifying) role by trapping a reactive intermediate, such as epoxides or other electrophiles, by the -SH group of glutathione. In many instances, therefore, the depletion of glutathione dramatically increases the hepatotoxicity of electrophilic intermediates that are formed by the action of the mixed-function oxidase enzymes (Mitchell 61; 21:: 1975). Not all the glutathione transferase reactions are detoxifying in nature, however, since 1,2-dichloro-, or 1,2-dibromo-ethane is chemically activated by glutathione conjugation (Rannug e_t a_l., 1978; Hill e_t_ 61., 1978). In these instances, the sulphur atom of glutathione replaces a chlorine or a bromine atom by nucleophilic substitution and gives rise to a mustard-like glutathione conjugate which is more mutagenic than the parent compound. Glucuronidation is another means of conjugating certain metabolites of the mixed-function oxidase enzymes. Glucuronic acid (from UDP-glucuronate) may be 11 transferred to an appropriate acceptor group, usually a hydroxyl, by the action of a family of microsomal enzymes called UDP-glucuronyltransferases (Dutton, 1975; Aitio, 1978). Other conjugation reactions include sulphation, acetylation, methylation, or glycine conjugation (Aitio, 1978). In general, the combined actions of the mixed-function oxidase system (Phase 1 reactions) and the conjugating enzymes (Phase 2 reactions) convert hydrophobic compounds into more hydrophilic metabolites which can then be more readily excreted in the bile or in the urine. It should also be remembered that the combined activities of these enzymes may also result in toxifying many other compounds into carcinogenic, mutagenic, or cytotoxic derivatives. Other xenobiotics, however, including most of the polyhalogenated aromatic hydrocarbons may be highly resistant to the actions of these enzymes, a fact that may explain their relatively long half lives and their deposition in certain body compartments particularly the adipose tissue (Tuey and Matthews, 1980; Lutz e6 a_l., 1977). Induction of Hepatic Microsomal Drug Metabolizing Enzymes An important aspect of the drug metabolizing enzymes is their susceptibility to induction by many of the xenobiotics that may or may not be actual substrates for these enzymes. Induction usually increases the ability of these enzymes to metabolize xenobiotics as well as normal body constituents (Conney and Burns, 1972; Conney, 1967; Sher, 1971). Other substances including lead and other heavy metals, organophosphorous insecticides, carbon tetrachloride, as well as certain gases such as ozone and carbon monoxide, and the volatile anesthetic agent fluroxene (2,2,2-trifluoroethyl vinyl ether) may inhibit or depress certain of these microsomal enzymes (Kappas and Alvares, 1975; Conney and Burns, 1972; Ivanetich and Bradshaw, 1977). Different mechanisms may underlie these inhibitory effects, but none is thought to be mediated through interference with the gene expression. 12 Inducers of the hepatic microsomal drug-metabolizing enzymes may be classified into at least two groups; one being represented by the barbiturate phenobarbital (PB), and the other is exemplified by the hepatocarcinogen, 3-methylcholanthrene (MC). By itself, each of these two agents result in a characteristic pattern of induction and distinct changes in the physiology and function of the liver. Phenobarbital, for instance, gives rise to proliferation of the smooth endoplasmic reticulum and hypertrophy of the liver. Stimulation of hepatic blood and bile flow as well as biliary excretion has also been observed in rats administered PB but not MC (Klaassen, 1970). In addition, distinct changes in the content and activities of certain microsomal drug metabolizing enzymes are associated with PB-treatment. The two major components of the mixed-function oxidase, namely NADPH-cytochrome P—450 reductase and one or more forms of cytochrome P-450 are greatly induced. In the latter case, the absorption maximum of the reduced carbon monoxide difference spectrum is unchanged (A m x = 450 a nm), and the microsomal ethyl isocyanide difference spectrum 455/430 nm absorbence ratio is not appreciably changed from its normal value of %. Induction of the microsomal monooxygenase activity is directed towards increased metabolism of a wide variety of substrates including the N-demethylation of aminopyrine, benzphetamine, and ethylmorphine, and the hydroxylation of testosterone at the 1601 position, in addition to increasing the aliphatic hydroxylation of hexobarbital and pentobarbital (Conney, 1967; Parke, 1975). In addition, PB induces a number of other enzyme activities including glutathione-S- transferase (Kaplowitz _e_t 61., 1975), d) aldehyde dehydrogenase (Deitrich gt a_l., 1978), and microsomal epoxide hydrolase (Bresnick 6t 61., 1977), as well as chloramphenicol UDP-glucuronyltransferase (Bock e; §_I_., 1973). The major cytochrome P-450 form(s) in microsomes from PB-induced rats appear to have a subunit molecular weight of 50,000-51,000 as judged by SDS-polyacrylamide gel 13 electrophoresis of purified enzymes (Guengerich, 1977; Guengerich and Martin, 1980). In contrast, MC administration causes only a slight increase in liver weight or proliferation of the endoplasmic reticulum and has no effect on NADPH- cytochrome P-450 reductase. MC induces different cytochrome P-450 species whose activities are directed towards a narrower range of substrates including the 2-biphenyl hydroxylation (Parke, 1975), ethoxyresorufin-O-deethylation (Burke e_t a_l., 1977), benzo(a)pyrene hydroxylation and epoxidation (Parke, 1975), and acetanilide-4-hydroxylation (Atlas and Nebert, 1976). In addition, MC causes a hypsochromic spectral shift to nearly 448 nm in the Soret peak of the difference spectrum of the reduced-CO saturated microsomal hemoproteins, and increases the ethyl isocyanide difference spectral 455/430 nm ratio by nearly 3-fold (Sladek and Mannering, 1966). A number of enzymes or activities including DT-diaphorase (Lind and Ernster, 1974), p-nitrophenol-UDP-glucuronyltransferase (Bock g a_l., 1973), and T aldehyde dehydrogenase (Deitrich e_t a_l., 1978) are also preferentially induced by MC. In the rat, mouse, and rabbit at least two forms of cytochrome P-450 with 54,000 and 56,000-57,000 apparent molecular weights seem to be induced by MC or other polycyclic aromatic hydrocarbons (Atlas 21 61., 1977; Guenthner and Nebert, 1978). The first of these hemoproteins seems to be associated with acetanilide 4-hydroxylase activity and is named cytochrome P-448 since the reduced-CO complex has the Soret peak at 448 nm (Guenthner and Nebert, 1978). The second hemoprotein is named cytochrome Pl-450 and is linked to aryl hydrocarbon hydroxylase (AHH), an activity that is fluorometrically measured by the rate of benzo(a)pyrene hydroxylation (Guenthner and Nebert, 1978; Nebert and Gielen, 1972). Inducers are generally classified as being PB- or MC-type after assaying a number of the microsomal drug metabolizing enzymes in microsomes isolated from 14 animals treated with the compound in question. Also, the microsomal protein profiles are examined by SDS-polyacrylamide gel electrophoresis. It should be pointed out, however, that overlapping substrate specificity may occur whereby an inducer of one type or another (MC or PB) may slightly induce one or more enzymatic activities that are characteristically associated with the opposite type of induction. For example, biphenyl is metabolized by microsomes from P8- and MC-treated rats into mainly the 4-OH and Z-OH derivative, respectively even though either type of induced microsomes also increases the rate of formation of the other metabolite (Creaven and Parke, 1966; Billings and McMahon, 1978). Furthermore, while both PB and MC seem to induce the benzo(a)pyrene hydroxylase activity the second inducer is appreciably more effective than the first one (Parke, 1975). These results may, at least partly, be explained by the existence of a certain degree of overlapping between different purified forms of cytochrome P-450 towards metabolizing certain substrates (Guengerich, 1977). This overlapping and the limited extent of resolution of microsomal hemoproteins by SDS-polyacrylamide gel electrophoresis should call for attention when attempts are made at characterizing the type of inductions in whole microsomes (Guengerich, 1977). While most inducers resemble either of the two prototypes, PB or MC, certain mixtures of polyhalogenated biphenyls can cause both induction responses simultaneously. A mixed-type induction response is also observed when both PB and MC are administered together. An agent or a mixture of different agents that simulates the combined effects of PB + MC may, therefore, be referred to as a mixed-type inducer. The mixed-type induction effects of certain mixtures of polyhalogenated biphenyls could be due to certain congeners in the mixture being PB-type inducers and others being MC-type inducers. However, the possibility exists that one or more of the congeners may each by itself cause these effects 15 simultaneously. More will be discussed later on the subject of mixed-type inducers, and the structure-function relationships for induction of the drug metabolizing enzymes by polyhalogenated biphenyl congeners. It is commonly accepted that the molecular mechanism for the induction of microsomal hemoproteins is mainly dependent on 66 n_o_\Q_ protein synthesis rather than activation of preexisting polypeptides or protein stabilization against degradative processes (Dehlinger and Schimke, 1972; Haugen e_t a_l., 1976). The effects of inducers on d_e rm protein synthesis are thought to be directed primarily at the level of DNA transcription, even though the translational processes of m-RNA may also be affected (Parke, 1975). Consequences of Microsomal Drug Metabolizing Enzymes Induction The ability of various xenobiotics to induce liver microsomal drug metabolizing enzymes may result in various physiological, clinical as well as toxicological and carcinogenic consequences. Since PBB dramatically enhance the activities of the drug metabolizing enzymes they may also cause pharmacotoxico- logical effects similar to other agents that induce these enzymes. Some of the immediate physiological effects of inducers are due to altered rates of biosynthesis or breakdown of certain vital body constituents particularly steroid hormones, lipids, and lipid soluble vitamins. For instance, PB and certain other drugs including halogenated hydrocarbon insecticides enhance the bio- synthesis of cholesterol and increase the rate of hydroxylation of cholic acids and several steroidal hormones (Conney and Burns, 1972; Parke, 1975). It may be expected that alterations in the relative proportions of steroidal sex hormones could have serious physiological implications on reproduction, fertility or libido. The metabolism and function of lipid-soluble vitamins may also be altered by microsomal enzyme induction. Epileptics given prolonged barbiturate treatments k 16 may devleop osteomalacia and hypocalcemia probably due to depletion of vitamin D body pools through enhanced conversion of 25-hydroxycholecalciferol, an essential metabolite of vitamin D, into more polar metabolites that can be excreted into the bile (Breckenridge, 1975). Vitamin K metabolism may also be altered by PB treatment since hemorrhagic episodes, that can be prevented by vitamin K treatment, occur in some babies born to mothers taking PB for epilepsy (Mountain _e_t a_l., 1970). Bilirubin conjugation and, therefore, elimination may also be enhanced by certain inducing agents such as PB, antipyrine, or DDT (Breckenridge, 1975). Increased bile and hepatic blood flow may also be caused by PB (Breckenridge, 1975). Another important aspect of microsomal enzyme induction is the clinical and pharmacotoxicological interactions between certain drugs that are administered simultaneously. Enzyme induction, for instance, is believed to result in shortening the half life of oral anticoagulants (Breckenridge, 1975). The rate of metabolism ‘ of anticoagulants and other drugs including sedatives is also enhanced in individuals who habitually consume excessive amounts of alcohol (Conney and Burns, 1972; Breckenridge, 1975). Certain drugs, including alcohol, may also enhance their own metabolism through enzyme induction, a fact that may explain tolerance to such drugs (Conney and Burns, 1972; Breckenridge, 1975). Microsomal enzyme induction may also alter the toxicity and carcinogenicity of certain xenobiotics by forming metabolites of greater or lesser virulence than the parent compounds. Phenobarbital pretreatment, for instance, potentiates the hepatotoxic effects of carbon tetrachloride, probably by enhancing the formation of the free radical metabolite, 'CC13, which may be involved in the peroxidative damage of membrane lipids (Parke, 1975). Prior i_n_ y_I_\ZC_) induction with MC, on the other hand, offers some protection against CCl4 toxicity (Parke, 1975). Protection by MC may be due to the preferential induction of one or more forms of 17 cytochrome P-450 that are not involved in metabolizing this chemical, while at the same time the content of the form(s) that is (are) normally responsible for producing the free radical metabolite may be reduced. Hepatic necrosis due to bromobenzene may also be potentiated by PB induction while induction by MC offers some protection. The major metabolite of bromobenzene is the 3,4-epoxide derivative which after reacting with and depleting hepatic reduced glutathione can covalently bind to cellular macromolecules, and thereby cause necrosis. Pretreatment with MC, however, does not significantly enhance the metabolism of bromobenzene whose major metabolite in this case is the relatively less reactive (and therefore less toxic) 2,3-epoxide derivative (Gillette _e£ a_l., 1974; Parke, 1975). The carcinogenic activity of numerous chemicals can also be appreciably modified by the type of induced microsomal enzymes. Therefore, the hepatocarcinogenic effects of 3'-methyl-4-dimethyl aminoazobenzene are largely attenuated by prior MC-induction, while PB-pretreatment tends to amplify the bladder carcinogenic effects of 2-naphthylamine through enhanced N—hydroxyla— tion of the parent compound (Parke, 1975). It should be emphasized that polycyclic aI‘Omatic hydrocarbons and similar compounds which do not contain highly reactive eleCtrophilic sites are not, themselves, carcinogenic. Among the most extensively sIIUdied of these compounds is benzo(a)pyrene which is mainly metabolized, by the mixe(LI-function oxidase system and epoxide hydrolase, into various phenols, quODES, dihydrodiols, triols, tetrols, epoxides, diol epoxides and water-soluble COHjugates (Yang £84., 1976, 1977, 1978). Some of these metabolites, particularly CErtain dihydrodiol epoxides, are highly reactive, and therefore they can covalently bind '30 macromolecules and cause mutation and cancer (Yang E1 a_l., 1978; Wood g 3‘1" 1976a,b). It is, therefore, conceivable that prior MC- or PB—type induction of t . he mlCrosomal drug metabolizing enzymes will preferentially enhance or slow ¥ — -—~ ~~ — — 1.- 11"? 18 down certain benzo(a)pyrene metabolic pathways that are ultimately involved in causing mutation and cancer. A Model for the Coordinate Enzyme Induction and Toxicity in Response to Halogenated Aromatic Hydrocarbons and Polycyclic Aromatic Compounds-"The TCDD-Receptor" During the past few years ample evidence has been gathered suggesting the involvement of a hepatic cytosolic protein in the mediation of the coordinate gene induction by TCDD, MC, or other similar halogenated or non-halogenated structurally planar polycyclic aromatic compounds. The many compounds that induce cytochrome Pl—450 and AHH activity can be divided into two groups, the polycyclic aromatic hydrocarbons such as MC, and the halogenated aromatic compounds including dibenzo-p-dioxins, dibenzofurans, PCB, and PBB. A regulatory genetic model has been described for explaining the mechanism by which these planar compounds evoke a coordinate gene expression leading to induction of several enzymes and activities (Guenthner and Nebert, 1978; Poland 61 a_l., 1979). Evidence supporting this model has been mainly gathered by two research groups headed by Daniel W. Nebert in the National Institute of Health, Bethesda, Maryland; and by Alan Poland at the McArdle Laboratory for Cancer Research in the University of Wisconsin. Before presenting a summary of the mOdel I will attempt to discuss some of the results and observations that are in faVOI‘ of this model. In crosses and backcrosses between two inbred strains of mice, C57BL/6 (B6), and DEA/2 (D2), 3 single gene difference between these two groups appeared to be I‘ESPPI‘ISible for the difference in their responsiveness to induction by MC. The first strain (86) responds to induction by MC of cytochrome Pl-450 and associated onPOXygenase actIVItIes IncludIng AHH, while mice of the D2 strain are ¥ 19 non-responsive. This trait for aromatic hydrocarbon responsiveness is inherited as a simple autosomal dominant trait (Nebert and Gielen, 1972; Thomas _e_t a_l., 1972), and is thought to be controlled by a genetic locus named the fl locus (Thorgeirsson and Nebert, 1977). At least ten other monooxygenase activities besides the AHH are also associated with the _A_h_ locus (Thorgeirsson and Nebert, 1977). Other non- monooxygenase activities including p-nitrophenol—UDP-glucuronyltransferase (Owens and Nebert, 1975) and DT-diaphorase (Lind and Ernster, 1974) may also be associated with this locus. Not associated with the locus, however, are the cytochrome P-450 mediated monooxygenase activities, cytochrome P-450 reductase, and epoxide hydrolase all of which are inducible by PB. Studies in rats have shown' that TCDD is nearly 30,000 times more potent than MC in inducing the hepatic AHH activity (Poland and Glover, 1974). TCDD also invariably induced this activity in various strains of mice which are responsive and non-responsive to MC (Poland g a_l., 1974). It, therefore, was suggested that inbred mice which were genetically non-responsive to MC must have the structural and regulatory genes needed for expressing cytochrome Pl-450 and its associated AHH and other activities since those mice do respond to the highly more potent TCDD. Non- responsiveness was attributed to a defective recognizing receptor, probably due to a mutation, which has a diminished affinity for MC or other similar polycyclic aromatic hydrocarbons but which still recognizes and binds TCDD (Poland g a_l., 1974). Other studies have shown that the non-responsive mice were at least ten times less sensitive to TCDD than responsive mice; ED50 for AHH induction was :10 nmole TCDD/kg and 2 1 nmole TCDD/kg, respectively (Poland and Glover, 1975). 12m studies have shown that TCDD can reversibly be bound, with a high affinti)’ (KD= 0.27 nM), by a liver cytosolic macromolecular species in the 86 responsive mice, but the cytosol of non-responsive mice (D2) had a much lower binding affinity (Poland e_t a_l., 1976a). The receptor binding affinity for various ‘ 20 polycyclic aromatic hydrocarbons and halogenated aromatic congeners correlated well with their potency as inducers of hepatic AHH activity in rats and mice (Poland _e_ta_l., 1976a,b; Poland and Glover, 1977). Since a common receptor is allegedly involved in binding these chemicals and eventually expressing their induction effects, it would seem plausible that the agonists of this receptor should have certain common structural features. Among the polycyclic aromatic hydrocarbons no obvious common features that may be necessary for receptor binding could be discerned, and no structure-activity relationship could be established (Poland and Glover, 1977). The halogenated aromatic compounds, on the other hand, seemed to have a well defined structure- activity relationship, and at least one common structural feature was present. The most active agonists in this group are approximate isostereomers whose molecular structures fit an approximately 3 x 10 A rectangle with halogen atoms occupying all four corners (Poland and Glover, 1977). Among the halogenated aromatic compounds that were tested, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) had the highest binding affinity to the receptor and was the most potent inducer of AHH (Poland and Glover, 1977). Dibenzofuran with halogens occupying the same four positions was similar but slightly less potent than TCDD (Poland and Glover, 1977; Poland e_t_ 11." 1976a). The isostereomeric 3,4,3',4'-tetrachlorobiphenyl was nearly 100-fold less potent than either TCDD or 2,3,7,8-tetrachlorodibenzofuran (Poland and Glover, 1977). Two other compounds, 3,4,5,3',4',5'-hexachloro(bromo)biphenyl had similar potency to 3,4,3',4'-tetrachlorobiphenyl in inducing the AHH activity in chick embryo livers (Poland and Glover, 1977). More will be discussed later on the structure-activity relationships in regard to the pharmacotoxicological effects of polyhalogenated biphenyls. _Ifl m9 and i_n £1319 studies have recently shown that the cytosolic receptor mediates the nuclear uptake and binding of TCDD (and possibly other similar 21 agonists), an event that would be essential for the subsequent gene-expression of monooxygenase activity (Greenlee and Poland, 1979). Following the administration of 3H-TCDD, hepatic nuclei of B6 responsive mice showed 5 times greater binding of the radio-ligand than the nuclei of the less responsive D2 mice. This binding was drastically diminished if unlabelled TCDD or 2,3,7,8-tetrachlorodibenzofuran, but not biologically inactive congeners, were coadministered i_n_v_iy6 with 3H-TCDD. .12 yigg experiments have also shown a much greater specific binding of 3H-TCDD by hepatic nuclei of responsive mice if the radioligand was preincubated first with cytosol from the same strain of mice than if it were preincubated with the cytosol of the less responsive D2 mice. No similar experiments were reported on the i_n_ £612 uptake of 3H-TCDD by nuclei of the less responsive D2 mice. If the diminished nuclear uptake of 3H-TCDD by this strain of mice were mainly due to a defective cytosolic receptor then the application of 3H-TCDD charged cytosol from B6 mice should restore their ability for radioligand uptake. The capacity of unlabelled polycyclic aromatic hydrocarbons and halogenated aromatic congeners to compete against 3H--TCDD for specific nuclear binding correlated well with their ability to compete with the same radioligand for specific cytosol binding (Greenlee and Poland, 1979). The degree of binding to the cytosol and nucleus also correlated with their potency as i_n yifl inducers of hepatic AHH activity (Greenlee and Poland, 1979). Somewhat similar studies were recently reported on rat liver nuclear uptake of 3H-MC, a process that apparently was also mediated by a cytosolic binding protein (Tierney _e_t a_l., 1980). An 3H-MC binding protein peak (termed Peak B) was obtained by fractionation of the cytosol by Sephadex chromatography. Incubation of this 3H-MC bound protein with rat liver nuclei resulted in translocation of radioactivity to the nucleus in a process that was time and temperature dependent. MC binding to the cytosolic proteins had an apparent equilibrium dissociation constant of nearly 2.8 nM, an order of magnitude higher 22 than the binding of TCDD to its respective receptor (Tierney e_t a_l., 1980). A number of other polycyclic aromatic hydrocarbons were also bound by the cytosolic Peak B. This MC-binding protein may be different from the TCDD-receptor for they both had significantly different apparent molecular weights and sedimentation coefficients (Tierney 6t a_l., 1980; Greenlee and Poland, 1979). The possibility that this MC-binding protein may be the same as the cytosolic estradiol receptor was also excluded (Tierney 6t a_l., 1980). All of these studies are in support of a model which does not seem unrelated to the molecular mechanism of steroid hormones action. As mentioned earlier, this model has been mainly advanced by the efforts of Nebert and his coworkers (Guenthner and Nebert, 1978) and Poland and his coworkers (Poland fl a_l., 1979). Briefly summarized, the effects of the polycyclic aromatic hydrocarbons and the halogenated aromatic inducers are mediated through hepatic cytosolic protein receptors (two kinds may exist, each being highly specific to one of the two groups of ligands) to which these compounds are bound avidly with a high degree of stereospecificity. The ligand—receptor protein complex translocates to the nucleus where in some unknown manner it evokes a pleiotrophic response involving the activation of numerous structural genes including cytochrome Pl-450 and its associated AHH activity. The cytosolic receptor protein(s) is genetically determined by the regulatory Ag gene which may be inherited as a single autosomal dominant trait. Mutation in MC-nonresponsive mice may result in an altered receptor of a diminished affinity for the inducing compound and/or a decreased number of binding sites. These mice, however, should have all the structural genes for induction of AHH and other associated activities by very potent agonists such as TCDD. A schematic representation of the proposed model is also shown on the following page. 23 CYTOPLASM NUCLEUS TCDD\ ‘TCDD + BP;>TCDD -BP;:TCDD -BP ll AHHACTIVITY A . TOXICITY Besides the biochemical responses evoked by TCDD and isostereomeric planar halogenated aromatic hydrocarbons, a variety of toxic effects in laboratory animals and human beings are also produced. In most animal species, toxicity due to these compounds is characterized by a slow wasting syndrome with ultimate death, involution of lymphoid tissues particularly the thymus, and embryotoxicity and/or teratogenesis. In a more limited number of species, these compounds also cause hepatotoxicity (hepatocellular necrosis), chloracne and hyperkeratosis (symptoms that are reported in exposed humans and monkey among other species), and chick edema syndrome (McConnell and Moore, 1979; McConnell, 1980; Kimbrough, 1974). The actual mechanism(s) for toxicity and the cause(s) of death are still an enigma. Among all known polyhalogenated aromatic hydrocarbons TCDD appears to be the most toxic and therefore it has been studied the most. Even though data on the toxic effects of specific halogenated aromatic congeners 24 are limited it is suggested that congeners that are isosteric with TCDD will produce the same TCDD—toxic syndrome if administered at a sufficient dose (Poland e_ta_l., 1979). Numerous physiological and biochemical studies have been conducted in order to explain the mechanism of toxicity and lethality due to TCDD. These studies are well summarized and reviewed by Neal g a_l. (1979). It was concluded that TCDD does not act through effects on DNA synthesis, cell growth or replication, or metabolism to reactive metabolites that can bind covalently, nor does it appear to interfere with absorption or utilization of nutrients, the synthesis of ATP, the cellular oxidation state or other metabolic processes (Neal E a_l., 1979). However, some of the effects due to TCDD including toxicity in the lymphoid tissues, muscular atrophy (due to protein catabolism), mobilization of fatty acids and edema can be mimicked by overproduction of glucocorticoids (Neal E a_l., 1979 and cited references). It was, therefore, considered possible that TCDD may bind to the specific glucocorticoid receptors and thereby cause glucocorticoid-like effects, or it may cause an adrenal hyperfunction by inducing the cytochrome P-450 mediated hydroxylation pathways that eventually lead to overproduction of glucocorticoids (Neal e_t_ a_l., 1979). These and other related possibilities were investigated, but the conclusion was reached that glucocorticoids are unlikely to be involved in TCDD toxicity (Neal 53$ 61., 1979). A model for the mechanism of toxicity by TCDD and similar congeners has been recently hypothesized by Poland and coworkers (Poland e_t g, 1979; Poland and Glover, 1980). It should be emphasized that this model was based on a rather limited number of toxicity studies that at best were done on one or just a few congeners of each of the several classes of halogenated aromatic compounds. Furthermore, most of the toxicity studies were conducted in only one animal species or another (except for TCDD which has been tested in almost all 25 experimental animals), and therefore extrapolation to other species may have to be made. Hence, the structure—toxicity relationships are generally less thoroughly studied than the structure-activity relationships for AHH induction and binding to the TCDD-receptor. With these limitations, it generally appeared that the potency of these compounds to induce AHH activity and their receptor-binding affinity correlated with their toxic potency. It was, therefore, suggested that toxicity due to chlorinated aromatic hydrocarbons is mediated through the cytosolic receptor, and that the initial event in their toxic action is the stereospecific recognition and binding by the receptor species (Poland e_t a_l., 1979). In support of this model, it was found that thymic involution and fetal cleft-palate formation (two measures of toxicity) in response to TCDD seemed to segregate with the fl-locus, which presumably determines the receptor protein in inbred strains of mice (Poland g a_l., 1979; Poland and Glover, 1980). According to this model, binding of a halogenated aromatic congener by the cytosolic protein species is required for exerting its toxic action; however, the actual mechanism(s) by which toxicity and lethality are triggered remains unexplained. Further support for this model was obtained when the binding capacity to the cytosolic receptor of a few congeners including 3,4,5,3',4',5'-hexabromobiphenyl and TCDD correlated with their ability to cause thymic atrophy (Poland and Glover, 1980). Some other nonhalogenated polycyclic aromatic compounds, including MC and B-naphthoflavone, which are relatively weak agonists for the receptor, caused thymic atrophy but were less potent than TCDD (Poland and Glover, 1980). In conclusion, during the past few years correlations have been suggested to exist between the structures of polyhalogenated aromatic congeners, the binding affinity to a liver cytosolic species, the induction of AHH and several other activities, and the degree of toxicity. It is becoming increasingly accepted that the correlations between toxicity and enzyme induction effects are not fortuitous. 26 The recognition and binding of certain polyhalogenated aromatic congeners by the cytosolic receptor presumably causes the sustained expression (or repression) of one or more genes that are controlled by the receptor(s) (Poland and Glover, 1980). It is feasible that these sustained effects on the genome may be involved in the mechanism of toxicity which remains to be elucidated. Polychlorinated Biphenyls and Structure-Function Relationships for Induction of Microsomal Drug Metabolizing Enzymes Commercial PCB are complex mixtures of chlorinated biphenyl isomers and congeners which have had a widespread industrial application due to their heat resistance properties (Sissons and Welti, 1971). PCB mixtures, particularly those with a high chlorine content are potent mixed-type inducers of hepatic microsomal drug metabolizing enzymes (Alvares g a_l., 1973; Alvares and Kappas, 1977). The spectral properties of microsomal cytochrome P—450 from rats pretreated with PCB are intermediate between those of PB and MC-induced microsomes (Alvares fig” 1973). Furthermore, AHH and ethylmorphine-N-demethylase, two activities that are preferentially induced by MC and PB, respectively, are induced by a commercial mixture containing 54% chlorine known as Aroclor 1254. Three forms of cytochrome P-450 with approximate molecular weights of 48,000 (P-4508), 52,000 (P-450b), and 56,000 (P-450C), were isolated from rats treated with this mixture (Ryan _e_t a_l., 1979). These hemoproteins were structurally distinct since they differed in their spectral, catalytic, immunologic, and electrophoretic prOperties; and upon limited proteolytic treatment they had different peptide maps. Moreover, using the same criteria, the three hemoproteins from Aroclor 1254-treated rats were indistinguishable from a mixture of hemoproteins from PB and MC-treated rats (Ryan e_t_ a_l., 1979). All these results are in support of the 27 contention that this PCB mixture induces similar cytochrome P-450 forms to those inducible by PB as well as those that MC or TCDD induces. It would appear that the duality of this biochemical response could arise from certain individual congeners evoking a PB-type induction and others causing an MC-type induction. Subsequent research with various pure PCB congeners has shown that the position and degree of halogenation govern the type and extent of the biochemical response to these chemicals (Ecobichon and Comeau, 1975; Goldstein _e_t a_l., 1976, 1977). It should be emphasized, however, that these studies were mostly performed with symmetrical PCB congeners totalling nearly thirty or so compounds, approximately 15% of the 209 possible congeners. In these studies, the tri or 'less substituted congeners were either weak or inactive inducers of microsomal enzymes. For the higher chlorinated congeners there seemed to be two distinct groups of inducers, a PB-type group represented by 2,4,5,2',4',5'- hexachlorobiphenyl, and an MC-type group which may be represented by 3,4,5,3',4',5'-hexachlorobiphenyl (Goldstein it. a_l., 1977, Goldstein, 1979). Based on a rather limited number of studies it was concluded that all PCB inducers can fit into either of these two groups (Goldstein e_t_ 61., 1977). The following rather generalized structure-activity relationship rules were suggested: (1) PCB congeners that are chlorinated in both the 9E9. and gag positions of both phenyl rings are PB-type inducers which upon further chlorination may have the quantitative but not the qualitative effects altered; (2) PCB congeners chlorinated in only the m_efii and p_a_13 positions of both phenyl rings are MC-type inducers which upon addition of even a single 22.1539 chlorine are considered to lose their MC-like induction activity (Goldstein fl a_l., 1977). At the same time when these rules were formulated, however, there seemed to be some contradiction since certain congeners, 2,3,4,2',4',5'- and 2,3,4,2',3',4'-hexachlorobiphenyl, were reported to have mixed-type induction properties (Stonard and Greig, 1976). In a 28 later study (Goldstein it. a_l., 1978), however, 2,3,7,8-tetrachlorodibenzofuran, a potent MC-type inducer, was proven to be a trace impurity that accounted for the partial MC-type induction effects of another congener, 2,4,5,2',4',5'-hexachloro- biphenyl, claimed (Alvares and Kappas, 1977) to be a mixed-type inducer, but in actuality was a strict PB—type. These later findings (Goldstein e_t .511" 1978) were used to defend the rules (particularly rule 2) mentioned above against any results that apparently disagreed with those rules (Goldstein e_t a_l., 1978, 1979; Goldstein, 1979). More recently, however, Parkinson and coworkers demonstrated that several highly purified di-9_r_t_i_i_o_ chlorinated PCB congeners, including 2,3,4,2',4',5'- and 2,3,4,2',3',4'-hexachlorobiphenyl, displayed the effects of PB and MC, though they appeared more like PB than MC (Parkinson e_t_ a_l., 1980a, 1981). These congeners, which were collectively called mixed-type inducers, had 3 2,3,4- trichloro substitution pattern on at least one ring. In a series of other recent studies several highly pure mono-66mg halogenated biphenyls were shown to also have intermediate induction effects between PB and MC (or TCDD). 2,4,5,3',4',5'-Hexabromobiphenyl was the first mono-M halogenated biphenyl to be reported with mixed-type induction properties (Dannan gt a_l., 1978a). This finding was followed by a series of reports showing similar effects by several PCB congeners having in common, besides other similar molecular features, a single halogen at a carbon o_rt_lE to the biphenyl bridge (Y oshihara SE a_l., 1979; Parkinson g 61., 1979, 1980b,c). Most of these congeners, however, seemed to have more MC-like than PB-like induction effects. A few of them were also shown to cause some of the typical TCDD-like toxicity effects (Yoshihara _ei _a_l_., 1979). Some appeared to be present in commercial PCB mixtures, and others could be detected in human tissues (Yoshihara e; 61., 1979; Parkinson e_t_ 61., 1980c). Besides having a single M halogen, each of these congeners had halogen substituents at both para carbons and at least at two of the 29 four meg carbons of the biphenyl nucleus. Most of these congeners, therefore, could be considered as mono-61113 chloro derivatives of 3,4,3',4'-tetra-, 3,4,5,3',4'- penta, and 3,4,5,3',4',5'-hexa-chlorobiphenyl, all three are known to have biological effects similar to, but nearly two orders of magnitude less potent than, TCDD. In general, the introduction of one halogen M to the biphenyl bridge of the latter compounds appears to weaken but not abolish their TCDD-like biological effects. In addition, the derived compounds seem to acquire (at variable degrees) some PB- type induction activity, hence the naming mixed-type inducers (Parkinson e; a_l., 1979, 1980b,c). However, in contrast to the di-6r_th_6 chlorinated PCB congeners, which have a 2,3,4-trichloro substitution pattern, the mono-61316 halogenated congeners are more similar to MC than to PB. Other studies have also demonstrated that w chloro substitution is not an absolute requirement, as suggested by Goldstein e_t_ a_l. (1977) (also refer to the above mentioned rule (1)), for a PCB congener to cause a PB-type induction. A strong PB-type induction was caused by the w unsubstituted 2,3,5,2',3',5'- and 2,3,5,6,3',5'-hexachlorobiphenyl (Kohli g a_l., 1979; Parkinson e_t_ a_l., 1980c). However, two other pa_r_a unsubstituted congeners, 2,3,6,2',3',6'- and 2,3,5,6,2',5'- hexachlorobiphenyl, had no or very slight PB-type induction activity (Goldstein fl a_l., 1977; Parkinson _e_t:__ayl_., 1980c). All these results should caution against making unwarranted generalized conclusions with regard to the structure-function relationships for microsomal enzymes induction by polyhalogenated biphenyls. While the type and extent of induction may be influenced mainly by the number of halogens 92312 to the biphenyl bridge, the extent of induction may also be determined by other factors including substitution at the m_et_a and pa_ra_ carbons. Induction by these compounds is most likely mediated by specific receptors that recognize and bind the two types of inducers of polyhalogenated biphenyls. (It is commonly accepted that the 30 TCDD-receptor(s) is responsible for binding certain polyhalogenated biphenyls and mediating their MC-type induction effects; however, no similar receptor has yet been demonstrated for binding PB and other similar agonists including polyhalogenated biphenyls (Poland gt _a_lt, 1980).) The persistence of polyhalogenated biphenyls is another rather important factor that should be considered when various congeners are assessed and compared for their induction potency. Certain PCB or PBB congeners can be metabolized at considerable rates, and, therefore, are appreciably less persistent than others (Matthews and Kate, 1979; Dannan e_t a_l., 1978b; Moore e_t gt, 1980). Differences in rates of metabolism and excretion should determine their effective Ifl M concentrations and, henceforth, their availability for binding by the induction receptors. Polychlorinated Biphenyls and Structure-Activity Relationship: for Toxicity and Binding to the Cytosolic Receptor Toxicity studies seem to indicate that, like other classes of halogenated aromatic compounds, correlations do exist between the ability of PCB congeners to cause an MC (TCDD)-like induction response and their TCDD-like toxic action. Therefore, PCB congeners which resemble TCDD as inducers (3,4,3',4'-tetra-, 3,4,5,3',4'-penta-, and 3,4,5,3',4',5'-hexa-chlorobiphenyl) are much more toxic than the congeners which belong to the class of PB-type inducers such as 2,4,5,2',4',5'- hexachlorobiphenyl (McKinney i gt, 1976; Yoshimura gt gl_., 1979: Biocca gt gt, 1981). Mono-gt'_t_hg chlorinated congeners which were proven as mixed-type inducers seem to also have intermediate TCDD-like toxic effects (Yoshihara gt g_l., 1979; Parkinson e_t_ g_l_., 1980b,c). Biochemical studies have also shown that the relative binding of certain PCB congeners to the TCDD-receptor correlates well with their TCDD-like activity (Poland and Glover, 1977; Poland i gt, 1979). 31 Future studies should show if the binding affinity of the mono-ortho substituted congeners, like their biological effects, is intermediate between the non-ortho substituted congeners and the di—ortho substituted ones. Induction of Drug Metabolizing Enzymes by Polybrominated Biphenyls Initial research on the effects of brominated biphenyls on microsomal drug metabolizing enzymes have been mainly performed with crude mixtures of PBB, particularly with the mixture of Firemaster which allegedly was involved in contaminating Michigan. This mixture, however, contains at least twelve congeners and isomers of PBB in addition to trace amounts of at least twenty other non PBB contaminants, including roughly 220 ppm of penta- and hexa-brominated naphthalenes (Moore a a_l., 1980; Hass _e_t gl_., 1978; and Chapter 1 of this thesis). Preliminary studies have shown that Firemaster induced cytochrome P-450 and certain associated activities in rat, mouse, quail, and dog (Cecil gt a_l., 1975; Babish _e_t a” 1975, 1976; Farber gt gt, 1976; Corbett gt g_l_., 1975; Sleight and Sanger, 1976). None of these studies, however, accurately defined the exact type of microsomal enzyme induction arising from the mixture. In this laboratory, the mixture was first shown to cause the combined induction effects of PB and MC, and therefore was considered a mixed-type inducer of rat liver microsomal enzymes (Troisi, 1975). The induction effects of Firemaster were still persistent even ten days after administering a single ip injection of 90 mg Firemaster/kg body weight. These findings were confirmed by the more elaborate studies of Dent and coworkers who investigated the effects of the mixture in rat liver, kidney and mammary gland. In some of these studies both the chronic and acute administration of the PBB mixture caused induction effects in the liver that were almost indistinguishable from those arising from the combined administration of PB 32 and MC (Dent gt 536., 1976a,b). The mammary tissue of lactating rats was also susceptible to PBB which seemed to induce benzo(a)pyrene hydroxylase activity; while in the same tissue there was a 50% reduction in epoxide hydrolase activity (Dent gta_l., 1977). The kidney seems to be another tissue where epoxide hydrolase seemed preferentially depressed to only 14% of the normal activity after rats were fed diets containing 100 ppm PBB for three months (McCormack gt 536., 1977). It was postulated that induction of kidney benzo(a)pyrene hydroxylase and depression of epoxide hydrolase in the same tissue may increase the susceptibility of the kidney to toxicological damage by a second agent, such as benzo(a)pyrene, that can be metabolized into reactive epoxides (McCormack gt gt, 1977). The same conclusion ~was reached regarding the similar effects of the mixture on rat mammary tissue (Dent g a_l., 1977). Effects of PBB on rat neonatal drug metabolizing enzymes were also investigated. In one study, lactating rats were fed 0, 0.1, 1.0, or 10 ppm PBB for eighteen days after parturition when the dams and their pups were sacrificed (Dannan, 1978). The livers of the dams fed the 10 ppm PBB and their nursing pups showed the full spectrum of induction characteristics of both types of inducers, PB and MC. Pups were nearly 10—fold more sensitive than their dams since mothers fed the 1.0 ppm PBB diet were not affected while their pups had a mixed-type induction of liver microsomal enzymes. Therefore, components responsible for both the MC- and PB-like aspects of induction in the PBB mixture must be transmitted via the milk. This was further verified when the milk of rats was shown to contain virtually all the major PBB components of the mixture, whose composition was slightly altered (Dannan, 1978; Dannan i gt, 1978b). A later study has also shown that nursing rats fed diets containing 50 ppm PCB or PBB had nearly 293 ppm and 180 ppm, respectively, of the two mixtures in their milk at parturition (McCormack _e_t gt, 1979). The apparent concentration of these 33 chemicals in rat's milk may, at least in part, explain the previous results on the higher sensitivity of the pups relative to their dams (Dannan, 1978). PBB may also be transmitted transplacentally since pups that were exposed to PBB i_n u_tefl, but were later nursed by untreated rats showed mixed-type induction of liver microsomal enzymes (Dent e_t_ a_l., 1977). Dent gt gt (1978) have also attempted to characterize the nature of the PBB- inducible cytochrome P-450 hemoproteins. Female rats were administered a single ip injection of 150 mg Firemaster/kg and the time-course of changes in ethylisocyanide spectral properties, reaction kinetics, inhibitor effects, and microsomal SDS-polyacrylamide gel electrophoretic profiles were examined. Some anomalies were observed between the kinetic data and electrophoretic data, and it was suggested based on the electrophoretic results that the later stages (following the first 24 hr) of induction by PBB resembled an "induced control" situationi(Dent gt gt, 1978). It was concluded that PBB may represent a new class of inducing agents since they seemed to have some but not all of the mixed-type induction properties of both PB and MC. With the exception of the results of Dent e_t gt (1978) all data seem to agree that the crude mixture of Firemaster evokes a mixed-type induction response in rat (and mouse) liver microsomal drug metabolizing enzymes in a manner very similar if not identical to that normally evoked by the combined administration of PB + MC. Several attempts have been made to characterize the PB-like inducers in the mixture and more importantly to account for the agents that cause the TCDD-like toxic consequences, so that the potential hazards of PBB-related environmental contaminations could be determined. The following will be a brief summary of what has been accomplished at the time the present research was being pursued and completed. 34 Attempts to Characterize the Inducing Agents in the Crude PBB Mixture of Firemaster Attempts at identifying the potentially toxic (toxicity due to PBB will be discussed later) and TCDD (MC)-like inducing agents in the crude Firemaster mixture have been frustrating and largely unproductive. The structures of seven PBB congeners totalling nearly 90% by weight of the mixture have been identified, and each of these congeners has been shown to contain at least one bromine 9.1m to the biphenyl bridge (Sundstrom gt a_l., 1976a; Jacobs gt gt, 1976; Moore gt a_l., 19783; Moore and Aust, 1978). During that time, however, the "dogma" that PCB congeners with one or more halogens M to the biphenyl bridge could not have any MC-like induction effects was still unchallenged (Goldstein e_tgt, 1977; Poland and Glover, 1977). For this reason, and since the purification of large quantities of homogeneously pure PBB congeners was a rather formidable task, most views have turned away searching for possible non-PBB contaminants that may account for the TCDD-like effects of the mixture. Several studies have involved the fractionation of Firemaster by column chromatography and testing the biological effects of the resulting crude fractions. These studies had certain features in common, namely the mixture was applied to an alumina (or a Florisil) column which was eluted with organic solvents of increasing polarity. Several fractions (usually two to five), each corresponding to a particular eluting solvent, would be collected and tested for their effectiveness in inducing the microsomal drug metabolizing enzymes or causing a TCDD-like toxicity response. Aside from the PBB congeners, any TCDD-like agents, such as brominated-dibenzo-g-dioxins or dibenzofurans, that may be present in the mixture would be expected to elute in the most polar fraction(s) since chemically they are the most polar, and therefore chromatographically speaking, would be retained the most. 35 In one such study, a polar fraction constituting nearly 0.17% of the original Firemaster sample was found to contain twenty or more non-PBB contaminants, but no brominated dibenzofurans or dibenzo-g-dioxins could be seen at a detection limit of 0.5 ppm of Firemaster (Hass i gt, 1978). This polar fraction was administered to guinea pigs at 1 mg/kg, a dosage equivalent to 500 mg of the original mixture, and also to mice at 15 mg/kg and all animals were watched closely and weighed regularly for the duration of the study (30 days). Most body organs including the spleen and thymus were weighed and examined histologically and analysis of hepatic porphyria was also performed. In the same study, the acnegenic properties of the polar and non-polar fractions of Firemaster were examined by painting each on rabbit ears (the rabbit ear test). The results of all these studies agreed that the polar fraction was inert toxicologically, while Firemaster or the non-polar fraction, both of which containing in addition to the PBB congeners nearly 220 ppm brominated naphthalene, were somewhat toxic by the rabbit ear test. It was concluded that the low concentrations of brominated naphthalenes may not be sufficient to account for the toxic effects of Firemaster, and therefore the observed biological effects are most probably due to the brominated biphenyl congeners themselves (Hass gt gt, 1978). In another study, three fractions, F1’ F2 and F3 (in order of increasing polarity), were collected by chromatographing Firemaster on an alumina column; and each or Firemaster was administered twice into rats at probably 100 mg/kg body weight (the exact dosage was not given but the samples were dissolved at 100 mg/2 ml corn oil) (Safe g gt, 1978). A toxicity index (T/D, which means Toxification/Detoxification) was defined and determined for each type of induced microsomes based on the results of the following i_g v_i_t_tg assay. Microsomes were incubated with 3H-4-chlorobiphenyl in the presence of NADPH, and radiolabelled metabolites were analyzed and separated into soluble metabolites and low 36 molecular weight conjugates (detoxification products), and high molecular weight adducts (toxication products). The toxicity index was 23, 30, 29, 55 and 49 for control microsomes and microsomes from F1’ F2, F3, or Firemaster-pretreated rats, respectively. These results were inadequate for making any conclusions regarding the identity of the potentially toxic agents in Firemaster (Safe gt a_l., 1978). Using the rabbit ear test Kimbrough e_t a_l. (1977) have also evaluated the hyperkeratotic activity of a polar and non-polar PBB fraction that were resolved by alumina chromatography. The polar fraction produced a pronounced hyperkeratosis of the rabbit ear, while the non-polar fraction caused only a very mild reaction. Firemaster: also caused hyperkeratosis at a total dose of 60 mg/rabbit ear while similar reactions were elicited by 20.5 or 17.5 pg of a tetrachloro- or a tetrabromodibenzofuran, respectively. It was concluded that the PBB mixture contains chloracne-producing chemicals whose nature was undetermined (Kimbrough gt gl_., 1977). Four fractions were also resolved by alumina column chromatography of Firemaster, and these fractions were tested for their microsomal enzyme induction effects (Moore gt gl_., 1980). Relative to the other tested fractions the fourth one (constituting 4.5% by weight of the starting material) was found to be a potent MC-type inducer. However, no information was given on the actual composition of this polar fraction or the relative amounts of PBB congeners in this fraction (Moore gt gl_., 1980). In a rather recent study, Firemaster BP-6 was fractionated into two and three fractions by chromatography on alumina (fractions AA and AB) and Florisil (fractions FA, FB’ and FC), respectively (Robertson gt _a_l., 1981b). These fractions were administered (ip) twice in rats at 100, 6.7, 100, 0.6, and 0.4 mg/kg, respectively and an attempt was made to correlate the extent of MC-type 37 induction with the PBB composition of these fractions. From this study, even though four PBB congeners (_2_, _5_, 6, and Z whose actual structures are given in Chapter 1) were prematurely implicated as the agents that were responsible for the MC-type induction effects, it was concluded that other yet unidentified components or synergistic effects may also be important (Robertson _e_t 53., 1981b). Like the previous studies, therefore, no definite conclusions could be reached in terms of identifying the agents that are totally responsible for the TCDD-like effects of Firemaster. A somewhat different approach was taken by Goldstein e_t a_l. (1979) who studied the effects of 2,3,6,7-tetrabromonaphthalene, which was predicted as the most potent MC-like inducer of all brominated naphthalenes. At a level equivalent to the concentration of brominated naphthalenes in Firemaster, this substance had no effect either by itself or when administered with 1 mmole/kg of the PB-like inducer 2,4,5,2',4',5'-hexabromobiphenyl. The tetrabromonaphthalene was found to be an MC-type inducer, but it was estimated that nearly 80 g Firemaster/kg body weight would have to be administered in order for the naphthalenes in the mixture to cause an appreciable MC—like effect. It, therefore, was concluded that the nearly 200 ppm brominated naphthalenes in Firemaster are unlikely to account for the MC-type induction effects of the mixture (Goldstein Qgt, 1979). In summary, some of these studies have indicated that the brominated naphthalenes are unlikely to account for the TCDD-like effects of Firemaster, and that these effects could not be due to brominated-dibenzofurans or dibenzo-g- dioxins since the latter compounds could not be detected in the mixture. Other investigations have provided evidence for the presence of the TCDD-like agents in the relatively polar fractions, but have not been able to rule out the possibility that certain PBB congeners are themselves responsible for most or all of these effects. 38 In spite of these and other efforts the question as to which of the PBB congeners, if any, would be responsible for these effects has therefore remained unanswered. Efforts to characterize the PB—type inducers in Firemaster have been relatively simpler and more successful. The two major components, 2,4,5,2',4',5'- hexa- and 2,3,4,5,2',4',5'-hepta-bromobiphenyl, were purified from PBB mixtures by alumina chromatography and recrystallization. Each was tested in rats and found to evoke a strict PB-type induction response (Moore, 1978; Moore g a_l., 1978c, 1979). Both congeners combined make up nearly 70% of Firemaster, and therefore the majority of the similar effects of the mixture can readily be accounted for by these two congeners. Based on the structure-activity relationships for PCB a number ofother yet untested congeners were predicted to contribute to the PB- like induction in Firemaster (Moore gt gt, 1980). As will be shown in Chapter 2 of this thesis most of these predictions were found true. PBB Metabolism: Structure-Function Relationships 12 M metabolism of a certain chemical may, in general, contribute to its overall toxicity. As discussed earlier, some metabolites of benzo(a)pyrene such as certain dihydrodiol epoxides are extremely reactive, and therefore they can covalently bind to macromolecules and ultimately cause mutation and cancer (Yang e_t a_l., 1978; Wood e_t a_l., 1976a,b). Compounds such as PCB or PBB congeners are unlikely to be as carcinogenic as benzo(a)pyrene and similar polycyclic aromatic hydrocarbons. Evidence, however, still indicates that certain PCB or PBB mixtures, including Firemaster, can cause liver tumors or neoplastic nodules in mice or rats (Ito g a_l., 1973; Kimbrough gt a_l., 1975, 1978). While the cause of such effects is still largely unknown, it Is possible that these effects are mediated through metabolism and covalent binding to DNA. PCB congeners or mixtures of low chlorine content are known to undergo metabolism via the microsomal 39 mixed-function oxidase enzymes into products or intermediates that can covalently bind to cellular macromolecules (Wyndham gt a_l., 1976; Shimada and Sato, 1978). An extensive amount of research on the i_n yiyg and D My; metabolism of chlorobiphenyls, including metabolite identification studies, has been performed (for reviews refer to Sundstrom _e_t a_l., 1976b; and Safe, 1980). With the exception of one or two recent studies (Kennedy _e_t a_l., 1980, 1981) little or no research has explored the structural requirements for chlorinated biphenyls metabolism. Ten dichlorobiphenyls were recently studied for their it v_ittg metabolism by microsomes from rats given different pretreatments (Kennedy g gl_., 1981). Among the £112 (mono- or di-) substituted dichlorobiphenyls the greatest rates of metabolism were observed when microsomes from PB-treated rats were used. In contrast, M unsubstituted dichlorobiphenyls were most efficiently metabolized by microsomes from B-naphthoflavone (an MC-type inducer) treated rats. Enzymes of the mixed-function oxidase and epoxide hydrolase were postulated to be involved in metabolizing the various dichlorobiphenyls into monohydroxy and dihydrodiol derivatives through arene oxide intermediates. Several jg v_iyg studies have recently attempted to search for and characterize the metabolites of certain lower brominated biphenyls, and the PBB metabolites of Firemaster. Some of the results were summarized and reviewed by Moore gt gt (1980). A rather different approach was recently undertaken for studying the lflflfl structure-function relationships for PBB microsomal metabolism (Dannan, 1978; Dannan gt gt, 1978b; Moore gt a_l., 1980). tgyitrt experiments were designed to resemble common drug metabolizing enzyme assays where hepatic microsomes were incubated aerobically with the substrate (PBB) in the presence of an NADPH- generating system. Incubation mixtures would then be extracted in ethyl acetate, and extracts would be chromatographed on a Florisil column to remove any 40 contaminating lipids. The rate of PBB metabolism would be determined after measuring the amount of substrate disappearance by gas chromatography. Initial experiments have shown that only two PBB components, 2,4,5,2',5'—penta~ and 2,3,6,2',4',5'-hexa-bromobiphenyl were metabolized (preferentially disappeared) when Firemaster was incubated with microsomes from PB- or PBB-treated rats, but no metabolism could be observed with control microsomes or microsomes from MC-treated rats. Of all PBB congeners in the mixture only these two are known to have two adjacent unsubstituted carbons one of which is ggtg to the biphenyl bridge. Three other congeners, 2,4,5,3',4'-penta-, 2,3,4,2',4',5’-hexa-, and 2,3,4,5,2',3',4'-hepta-bromobiphenyl have two adjacent unsubstituted carbons, but were not susceptible to metabolism. Based on these findings it was concluded that the presence of adjacent unsubstituted carbons was not by itself a sufficient criterion for rendering a PBB congener susceptible to metabolism (Dannan, 1978; Dannan gt a” 1978b; Moore gt a, 1980). It was hypothesized that two ~unsubstituted carbons one of which was at a £12 position would be required in order for a PBB congener to be metabolized. The structure—function relationships were further tested by employing seven model PBB compounds that were not believed to be present in Firemaster. Among the most striking findings was the extremely high rate of metabolism of 2,2'- dibromobiphenyl (estimated to exceed 2100 pmoles/min~mg protein), and the lack of metabolism of its fig disubstituted isomer, 4,4'-dibromobiphenyl, despite the fact that both compounds possessed four pairs of adjacent unsubstituted carbons. Each of the remaining five model compounds (each containing four or five bromines) had at least one pa_1a_ unsubstituted carbon (not necessarily adjacent to another unsubstituted carbon), and all five could be metabolized. It was, therefore, concluded that a w unsubstituted carbon seemed to be absolutely required for the i_n vitro metabolism of PBB congeners. Significant rates of metabolism will not 41 be observed if both pfl carbons were halogenated (Dannan fl a_l., 1978b; Moore gt a_l., 1980). The apparent rates of metabolism for a number of PBB congeners including several congeners of Firemaster were also compared to see if other structural features may influence the relative rates of metabolism (Moore e_t gt, 1980). At least two apparent trends seemed important. Rates of metabolism seemed directly proportional to the number of gt_h_o_ bromines, but appeared inversely related to the total number of bromines. The presence of two adjacent unsubstituted carbons including the one at a fig position may also enhance the rate of PBB metabolism. It should be emphasized that these studies were only performed on a rather small number of PBB congeners using mainly one type of microsomes (PB-induced). More research will be needed on other untested congeners to determine the validity of the apparent trends. Toxicity Due to Polybrominated Biphenyls According to Dorland's Medical Dictionary toxicity is "the quality of being poisonous" (Friel, 1974). With such a broad definition it would be extremely difficult to attempt to explore all the various aspects of PBB toxicity. Such tasks may be further complicated since different PBB congeners may also have quite distinct toxicological and pharmacological properties. For the purpose of this thesis, therefore, I will attempt to briefly review some of the TCDD-like toxic effects of PBB including their immunotoxicity and the lesions they cause, particularly in the liver. The actual mechanisms by which most of the toxic effects of PBB, and other similar halogenated aromatic hydrocarbons, are expressed remain largely unknown. In all animal species that have been studied to date, when sufficient amounts of PBB are administered the initial sign of PBB toxicity is weight loss or reduced 42 weight gain which is sometimes accompanied by a decrease in food intake (Sleight and Sanger, 1976; Corbett g a_l., 1978; Farber gt a_l., 1978; Lambrecht at _a_l., 1978; Ringer, 1978; Gupta and Moore, 1979; Fraker, 1980). Other characteristic changes include atrophy of the lymphatic tissues (particularly the thymus and to a lesser degree the spleen), and in certain cases death is imminent (Ringer, 1978; Farber gt gl_., 1978; Gupta and Moore, 1979; Fraker, 1980). Most of the atrophy in the thymus can histologically be accounted for by a marked lymphatic depletion in the cortical area (Sleight and Sanger, 1976; Gupta and Moore, 1979; Fraker, 1980). All these effects are typical of TCDD and other similarly toxic halogenated aromatic hydrocarbons (McConnell and Moore, 1979). A number of immunological tests have been conducted to evaluate the effects of PBB on certain aspects of the immune response. Kateley and Bazzell (1978) examined fifty-two contaminated cattle that had up to 30 ppm of PBB in the fat. There were no noticeable effects on the number of lymphocytes, concentrations of serum immunoglobulins, antibody-mediated responses, or mitogenic responses (Kateley and Bazzell, 1978). In rodents, immunological alterations were seen when mice and rats were administered 22 daily oral doses of Firemaster at levels ranging from 0.03 to 30 mg/kg body weight (Luster gt a_l., 1978). Cell mediated immunity was severely depressed in both rats and mice given the largest dose and to a lesser extent in animals given the next largest dose (3.0 mg/kg). The humoral immune function, as determined by measuring the antibody production, serum immunoglobulin levels, and splenic lymphocytes response to mitogens, was depressed only in mice at the 30 mg/kg dosage level (Luster e_t gt, 1978). In more chronic studies in which 0.1 to 10 mg Firemaster/kg body weight was administered daily (five days a week for six months) several immunological responses were depressed in mice and rats that were given the higher dosages, but immunity against bacterial challenge was not significantly affected (Luster gt a_l., 43 1980). It was concluded that although PBB could induce immune alterations in rodents, immunological effects become most apparent at exposure levels that approximate the doses needed to cause other significant toxic responses such as liver lesions and decreased body weight gain (Luster gt a_l., 1980). In another study, a dose-dependent reduction in B- and T—helper cells was observed in female mice that were fed diets containing 0, 1, 10, or 100 ppm Firemaster for 30 days (Fraker, 1980). This indicated that the humoral immune system may be dramatically affected by PBB. In contrast, no effects were seen on T-cell dependent delayed type hypersensitivity in any PBB—treated animals during the same time period (Fraker, 1980). Certain aspects of the immune function were also evaluated after feeding 0, .100 or 200 ppm PBB-containing diets to pregnant swine during the last half of gestation and the following four weeks of lactation (Howard E a_l., 1980). Significant reductions were seen in mitogenic response of peripheral blood lymphocytes of PBB fed sows and their four week-old offspring. Conversely, the bactericidal activity of whole blood in sows fed PBB or in their offspring was normal. The mechanisms for the PBB-caused immunosuppression were not readily known (Howard gt a_l., 1980). In summary, while PBB appear to be immunosuppressive in most laboratory animals, dosages that can cause other apparent toxicological changes may be required before immune alterations can be seen. Variable effects could also be observed where in certain experiments PBB primarily seemed to affect the cell mediated immunity (Luster gt a_l., 1980), while in others the humoral immune function was the primary target (Fraker, 1980). These differential effects could be time-dependent (chronic vs. acute exposure) since the normal proliferative processes of B- and T-lymphocytes are not identical (Vos fl a_l., 1980). Finally, in PBB-contaminated humans (farmers and PBB-manufacturing workers) no abnor- malities in lymphocyte number or function could be detected even in subjects with 44 > 50 ppm PBB in their fat (Stross fl gt, 1981). No other human health effects could be directly correlated with exposure to PBB (Stross g a_l., 1981). Besides the immune system the liver is another sensitive target for PBB. Liver enlargement (hepatomegaly) mainly due to hepatocellular hypertrophy is reported in all species studied including rat, mouse, chick, cattle, and monkey (Sleight and Sanger, 1976; Gupta and Moore, 1979; Corbett g gt, 1978; Fraker, 1980; Ringer and Polin, 1977; Moorhead gt a_l., 1978; Allen e_t _a_l., 1978). Ultrastructurally, hepatocyte enlargement is mainly due to extensive proliferation of endoplasmic reticulum (ER) particularly the smooth type (Sleight and Sanger, 1976; Corbett gt a_l., 1978; Gupta and Moore, 1979). Other characteristic hepatic Ultrastructural changes include increased numbers of cytoplasmic lipid droplets, disorganized rough ER and formation of concentric arrays of laminated ER membranes (also called myelin figures or whorls) around lipid droplets (Sleight and Sanger, 1976; Gupta and Moore, 1979). Formation of neoplastic nodules can also be induced in rat liver only six months after administering a single oral dose of 1 g Firemaster/kg body weight (Kimbrough g a_l., 1978). These neoplastic nodules appeared to develop into hepatocellular carcinomas when rats were killed nearly two years after a similar dose of PBB was administered (Kimbrough _et gt, 1981). The mechanism(s) by which PBB can induce hepatic nodules or carcinomas remains to be established. Research that will be described in this thesis is concerned with studying the relationships between the chemical and pharmacotoxicological properties of polybrominated biphenyls. The objectives are to account for the TCDD-like toxic chemical(s) that are present in Firemaster, and to examine the structure-activity relationships for enzyme induction and toxicity by polyhalogenated biphenyls. The original rules (Goldstein e_t gt, 1977; Goldstein, 1979) did not allow an @- halogenated biphenyl to have any TCDD-like toxic effects. However, those rules 45 were rather generalized since they were based on studying a rather limited number of PCB congeners (30-40 compounds), none of which was singly gthg substituted. Firemaster, contains three mono-gift brominated congeners, 2,4,5,3',4'-penta-, 2,4,5,3',4',5'-hexa-, and 2,3,4,5,3',4'-hexa—bromobiphenyl, which may be considered the derivatives of 3,4,3',4'-tetra- and 3,4,5,3',4'-penta-brominated biphenyl. Even though the last two PBB congeners have not been tested they are expected, like 3,4,5,3',4',5'-hexabromobiphenyl (HBB), to be the most toxic PBB congeners since their chlorinated analogues are also the most toxic of the PCB congeners. Preliminary chemical studies (refer to Chapter 1) have indicated that the singly gr_th_o brominated congeners possess intermediate chemical properties between the di-g‘tfg brominated compounds such as 2,4,5,2',4',5'-hexabromobiphenyl and the m unsubstituted HBB. It would, therefore, seem likely that they also have intermediate biological effects in terms of binding to the TCDD-receptor and eliciting the biological responses characteristic of TCDD or HBB. If these expectations were true then most of the similar biochemical effects of Firemaster may be accounted for by these congeners since they represent a significant portion of these mixtures (mg i_ntig). These hypotheses and the relationships between the structures of PBB congeners and their biological and toxicological effects have been examined by a number of different approaches. Results of i_rt M TCDD- receptor binding affinity to various PBB congeners, including those with a single pr_thg bromine, have been compared to i_n_ M pharmacotoxicological studies. However, since PBB congeners have a limited solubility in aqueous media results of the IE fittg binding studies may not, by themselves, be reliable for estimating the extent of tr; gig; toxicity and induction by these congeners. Furthermore, these compounds seem to have unusual pharmacokinetic properties since they are predominantly deposited in adipose tissue and, except for two congeners, are extremely resistant to metabolism and elimination (Dannan g gl_., 1978b; Matthews 46 and Kate, 1979). lg vivo studies, are, therefore, absolutely essential for understanding the pharmacotoxicological properties of PBB congeners. CHAPTER 1 STUDIES ON SOME CHEMICAL CHARACTERISTICS OF POLYBROMINATED BIPHENYL CONGENERS 47 48 ABSTRACT A commercial preparation of polybrominated biphyenyls (PBB), marketed as Firemaster, contains twelve major PBB congeners each of which is nearly 0.5% or more of Firemaster. The structures of seven of these congeners, 2,4,5,2',5'-penta—, 2,4,5,3',4'penta-, 2,4,5,2',4',5'-hexa-, 2,3,4,2‘,4',5'—hexa-, 2,4,5,3',4',5'-hexa-, 2,3,4,5,2',4',5'-hepta, and 2,3,4,5,2',3',4',5'—octa—bromobiphenyl, have previously been assigned. The strucutres of three additional congeners, 2,3,6,2',4',5'—hexa-, 2,3,4,5,3',4'-hexa-, and 2,3,4,5,2',3',4'-hepta-bromobiphenyl, have been determined by mass spectrometry and lH-NMR spectroscopy. Several separation techniques have been. used to purify these congeners including selective solubilization, repeated recrystallization, reversed—phase thin layer chromatography, and alumina adsorption column chromatography, as well as reversed-phase Lipidex-5000 column chromatography. The relative concentrations of all twelve congeners in Firemaster were determined from GC—detector response curves for pure PBB. UV- absorption spectra for some PBB congeners have also been recorded. Certain chemical and chromatographic properties have been compared in relation to the structures of the congeners. Most of these studies seem to indicate that the physico—chemical properties of PBB congernes are mainly determined by their total bromine content and more importantly by the number of bromines at carbons Lthg to the biphenyl bridge. INTRODUCTION The term "polybrominated biphenyls" refers to a group of chemicals (209 possible isomers and congeners) which collectively share the empirical formula C12H10-nBrn’ where n ranges from 1 to 10. Nearly 50 of these congeners have 49 been synthesized, and 42 of these are mono- through tetra-bromo (Sundstrom gt gl_., 1976) while a number of other congeners are commercially available in rather small samples from Ultra Scientific, Inc., formerly known as RFR Corp. Several preparations of PBB mixtures with an average bromine content of six, eight, nine or ten bromines per molecule have been produced for industrial application as flame retardants in USA and Europe (Brinkman and DeKok, 1980). The process of manufacturing PBB mixtures involves a Friedel-Crafts reaction of biphenyl with bromine in an organic solvent (e.g. dichloro- or dibromoethane) in the presence of chloride and a catalyst, such as aluminum chloride, under anhydrous conditions (Brinkman and DeKok, 1980). The reaction may be completed in few hours to give an 80% or'better yield. The reaction temperature does not seem to be critical since from as low as 100 to as high as 1500C have been reported in various synthetic processes (Brinkman and DeKok, 1980). After the desired degree of bromination is attained, excess bromine may be removed by adding ethylene to form ethylene dibromide, and the solid reaction products can be physically separated from the organic solvent by filtration or centrifugation. The solid products can be washed, dried, and pulverized to yield commercial preparations of PBB. Firemaster BP-6, with an average bromine content of six bromines per molecule, is one such preparation from which Firemaster FF-l can be derived by adding nearly 2% of calcium silicate as an anticaking agent (DiCarlo gt gt, 1978). Nearly 500-1000 pounds of Firemaster FF-l (believed to be of lot no. FH7042) was inadvertently introduced into Michigan's food chain in 1973 (Carter, 1976). Various attempts have been made to analyze and determine the composition of Firemaster. At least twelve major PBB components can be visualized when Firemaster is analyzed by gas chromatography, and the structures of seven of these congeners have been identified (Moore, 1978; Moore and Aust, 1978; Moore gt gt, 1978a; Sundstrom e_t a_l., 1976a; Jacobs gt a_l., 1976). Nearly twenty other non-PBB 50 trace impurities, including approximately 200 ppm of penta- and hexa-brominated naphthalenes, were detected by GC-MS analysis of a highly polar fraction of Firemaster (Hass gt gt, 1978). However, no brominated-dibenzofurans or dibenzo- g-dioxins could be detected at < 0.5 ppm (Hass fl a_l., 1978). While numerous biological studies have been performed with the crude mixture of Firemaster, there is an urgent need to understand the biological properties of the individual purified congeners. Such studies would obviously require rather large quantities of absolutely pure compounds. However, except for the major component in Firemaster, namely 2,4,5,2',4',5'-hexabromobiphenyl (congener 4), only small quantities of certain congeners could be obtained pure by the available techniques (Moore, 1978). It was necessary to devise methods for purifying sufficient amounts of as many PBB congeners as possible so that relationships between their chemical properties and pharmacotoxicological effects can be investigated. The structures of some purified congeners were identified, and some chemical and chromato- graphic properties of the individual congeners were examined. The results of these chemical characterization studies are the subject of this chapter. MATERIALS AND METHODS Materials Firemaster BP-6 (lot no. 6224-A) was a gift from Michigan Chemical Corporation (St. Louis, Michigan). A sample of Firemaster FF-l (lot no. FH7042, the lot which was allegedly involved in the Michigan incident) was a gift from Farm Bureau Services (Lansing, Michigan). Neutral and basic alumina (activity grade I) were purchased from Sigma Chemical Company (St. Louis, Missouri). 4,4'-Di-, 2,2'-di-, 3,5,3',5'-tetra-, 2,4,2',5’-tetra-, 2,5,2',5'-tetra-, 2,5,2',6'-tetra-, 51 2,4,5,2',6'-penta-, and 3,4,5,3',4',5'-hexa-brominated biphenyl were purchased from Ultra Scientific, Inc. (formerly RFR Corp.) Hope, Rhode Island. Each of these congeners was at least 95% pure. Hexane (n-hexane, b.p. 68-690) and all other solvents were of glass-distilled grade for pesticide analysis, from Burdick and Jackson Laboratories, Inc., Muskegon, Michigan. Hexane was purified by passing over basic alumina. Three percent OV-l on Gas Chrom O, 100/120 mesh was purchased from Applied Science Laboratories, Inc., State College, Pennsylvania. Lipidex-5000 was obtained from Packard Instruments Company, Inc. (Downers Grove, Illinois). Kieselguhr-G-TLC plates (0.25 mm and 0.5 mm thickness) were purchased from Altech, Inc., Newark, Delaware. Gas Chromatography (GC) PBB mixtures and fractions were routinely analyzed on a Varian 3700 gas chromatograph equipped with a pulsed 63Ni electron capture detector (ECD), and a Varian COS-111 digital integrator. A 6 ft, 2 mm ID, glass column packed with 3% OV—l on 100/120 mesh Gas Chrom O was maintained at 2450C with N2 (40 ml/min) as carrier and purge gas. Standard curves for the GC-ECD response (log area counts) to various amounts of each of the purified congeners of Firemaster (log gram PBB) were plotted. Linear regressions on the data points for each curve were obtained, and the slope, intercept, and correlation coefficient were calculated from each linear curve. These detector response data were used to determine the percentage of each congener in Firemaster FF-l and Firemaster BP-6, and to calculate the absolute concentration of each congener from the GC chromatograms of the various purified fractions. I thank Skip Mileski for constructing the GC-ECD response standard curves and determining the percentage of the congeners in the mixtures of Fire master. 52 Gas Chromatography-Mass Spectrometry (GC-MS) Molecular weights were determined by GC-MS (Hewlett-Packard 5840A/5985) at 70 eV and a source temperature of 200°C, using a 6 ft column packed with 3% OV-l or 5% SE-30. Selective Ion Monitoring (SIM), by GC-MS or by direct probe mass spectrometry, was occasionally used to specifically search for ions pertaining to congeners of brominated dibenzo—g-dioxins and brominated dibenzofurans in purified PBB samples. I thank Betty Baltzer of the Michigan State University Mass Spectrometry Facility for performing most of these analyses. 1l-l-hlviR Spectrometry lH--NMR spectra of purified congeners were recorded on a Bruker WP 180 spectrometer at ambient temperature. Each sample was dissolved in CDCI3 at nearly 1% w/v, and MeQSi was used as the internal standard. I thank the Chemistry Department of Michigan State University and the operators of the instrument for obtaining the spectra. Thin Layer Chromatography (TLC) Kieselguhr-G-TLC plates were prepared for reversed-phase chromatography and brominated compounds were detected as first described by DeVos and Past (1971), when the same technique was used for resolving certain PCB mixtures. Briefly the plates were impregnated with paraffin oil (8% in petroleum ether) and air dried before the PBB compounds were spotted. Between 3 and 5 pg of each congener in hexane (at 1 mg/ml concentration) was spotted with a 10 ul Hamilton syringe. The spotted plates were developed in a paraffin oil-saturated solvent mixture of acetonitrile:methanol:acetone:water (20:20:9:1). Plates that were spotted with congeners of five or more congeners were redeveloped two more times after they were allowed to air-dry between successive developments. 53 Completely dry plates were then sprayed with an ethanolic solution of AgNO3 (0.85 g AgNO was dissolved in 100 ml ethanol (95%) and 0.5 ml NH OH was added 3 4 immediately before spraying). This spraying solution is useful only for few minutes since silver ions start to quickly come out of solution upon addition of NHQOH. The plates were exposed to steam for a few seconds, and were finally irradiated with an ultraviolet (UV) lamp for approximately half an hour. Brominated biphenyl congeners appeared as dark brown spots on a brownish background. Using preparative plates (0.5 mm thick) the same procedure was used to purify milligram amounts of certain PBB congeners (see later). The PBB sample to be purifigd was streaked with a commercial apparatus across as few plates as required. 'Each plate was developed three times under the same conditions described above, then covered with aluminum foil except for a narrow section (0.5 cm wide) along the middle of the plate. Through this exposed section each plate was visualized by spraying with the AgNO3 solution and irradiation with the UV- lamp. The PBB composition of each of the visualized spots was assayed by the GC. The section of each plate across the spot containing the PBB of interest was scraped off. Scrapings of similar composition were combined from all plates and extracted (in a Soxhlet apparatus) twice in hexane. Extracts, however, contained in addition to PBB a fair amount (as much as 1 ml in certain cases) of paraffin oil that was impregnated into the coating of the plates. PBB can successfully be separated from paraffin oil by acetonitrile partitioning. Acetonitrile Partitioning This procedure is commonly used for extracting halogenated pesticides from fat-containing biological samples (McMahon and Sawyer, 1971). A detailed description of the method will not be given since it was presented elsewhere (McMahon and Sawyer, 1971). The procedure involves a preferential extraction of 54 a fat or an oil sample into petroleum ether in the presence of acetonitrile, which preferentially extracts the polyhalogenated hydrocarbons. PBB in the acetonitrile phase would then be back extracted into fresh petroleum ether in the presence of excess water. Under these conditions, acetonitrile would be miscible with water, and both would be removed and discarded; PBB would be retained into the petroleum ether phase. This procedure, if done carefully, can allow better than 95% recovery of oil-free PBB. UV-Absorption Spectroscopy The UV-absorption spectra of pure PBB congeners were recorded on a Cary 219 double'beam spectrophotometer between 370 and 190 nm (UV-grade hexane was the solvent). The extinction coefficients (8) of the major UV-absorption bands were calculated from the slope of the plots of Amax vs. PBB concentration. Purification of PBB Congeners As will become apparent, several techniques were concurrently required for resolving the PBB congeners of Firemaster. In general, none of the purification techniques was, by itself, adequate for purifying any congener. Furthermore, certain techniques were practical for purifying some of the congeners but not others. Following numerous preliminary experiments, the most suitable methods were selected for purifying each of the congeners. Procedures will be presented in two sections each of which was written to include common methods used for purifying several PBB congeners. 55 I. A General Scheme for Purifying Congeners t, 2, l_i, 2, 6, Z, and 2* In order to purify these congeners Firemaster BP-6 was selected over Firemaster FF-l since it contains roughly 10% more of the relatively minor congeners (particularly 2, 2, 2, and 6) than FF-l (Figure 1), and is free of calcium silicate which is present in FF-l. One or more of the following procedures were Fractionation of Firemaster BP-6 by Extraction and Recrystallization in Acetone Practically, 10-200 9 of Firemaster BP-6 can be employed to start each purification. The procedure followed in one experiment using 100 g of Firemaster BP-6 will be cited as an example (Figure 2). The initial fractionation step was essentially that of Bairstow e_t gt (1978). One hundred grams of BP-6 was stirred in 200 ml acetone at room temperature for 2 hr. The mixture was filtered on a Buchner funnel and, under these conditions, the insoluble fraction (IF) was rinsed with 25-50 ml of acetone. The IF, 65 g, was completely dissolved in hot acetone and allowed to crystallize in the dark at 4°C. After several days the resulting crystals (IF-Crystals) were harvested and successively recrystallized (3 times) in acetone at room temperature until crystals of over 99% pure congener 6 were obtained. Purified material can further be treated (batchwise) with activated charcoal and chromatographed on neutral alumina (in hexane) to remove any polar contaminants that may be hard to detect by GC analysis. * The major congeners which are present in Firemaster BP-6 and Firemaster FF-l will be assigned underlined numbers (t through t2) according to their sequential elution from the GC as shown in Figure 1. These are: 2,4,,5 2', 5'- penta- -(1), 2,4, 5, 3', 4'- -penta- -(2), 2, 3, 6, 2', 4', 5'- hexa— (3), 2, 4, 5, 2', 4', 5'- hexa- (4), 2, 3, 4, 2', 4', 5'-hexa- (5), 2,4, 5, 3‘, 4', 5'-hexa- (6), 2,3,4, 5, 3',4'—hexa- (Z), 2, 3, 4, 5, 2', 4' ,5'- hepta- (2), 2, 3, 4, 5, 2', 3', 4'- -hepta- -(2), Br8(1_0), Br (11), and 2, 3, 4, 5, 2', 3', 4', 5'- octa-(Q) bromobiphenyl. Two other congeners (g_a and 6_b), with intermediate retention times between congeners 6 and Z, are also present in Firemaster FF-l. 56 FIGURE 1. GAS CHROMATOGRAPHIC ELUTION PROFILE AND STRUCTURES OF POLYBROMINATED BIPHENYLS Firemaster BP-6 and Firemaster FF-l, 3.1 and 3.7 ng, respectively, were injected into an electron capture (63Ni) GC equipped with a 3% OV-1 column. Injector port, column, and detector temperatures were 2600, 2450, and 330°C, respectively; attenuation = 10 x 16 (Amp/mV). Twelve major congeners are assigned numbers according to their elution order from the GC column. The structures of ten congeners and the number of bromines in two others are shown. Two additional congeners, 6g and 612 whose retention times are intermediate between those of congeners 6 and Z, are also believed to be present in Firemaster (see later). FBr Q)r rBr Br Br ”a: a: Br Br I 8'5' 9' © © 3' 1 2: BP-6 59. © \ 8r 3! Br 3' Br Br Br 9' 3' BrHBr Br Br 1...; a o 3' 3' Br a: 3: Bi/ 8: / 3' 3' a: Br Br Br © © Br M Br Br Br 1 L I I I I I I IOOFS "4 F8 g FF-l 5 - 5 IL g I 50~I ' E 2 < -J 1 it I l 6 ' 23 7 9 6" Io II I2 0.1 I I I I I I L 0 4 8 l2 16 20 24 RETENTION TIME (min) FIGURE 1 58 The mother liquor (F-MLl) resulting from allowing the filtrate fraction (F) to crystallize was enriched in some of the relatively minor congeners of Firemaster. This fraction was concentrated and again allowed to crystallize in acetone at room temperature. The resulting mother liquor (F-MLZ) was concentrated and another batch of crystals (F-Crystals3) harvested. These crystals and those (F-Crystalsz) of the preceeding step were predominantly composed of congeners 4 and 2 and were not of particular use for further purification of any other congeners. The last mother liquor (F-MLB) was chromatographed on alumina. Neutral Alumina Chromatography Acetone was evaporated off the F-ML3 fraction, which when dry was an oil. About 1 g of this material was chromatographed on neutral alumina. The column (40 x 2.5 cm), fitted with a Teflon stopcock and a glass fritted disk, was filled with hexane and the stopcock opened while powdered alumina (110 g) was added. After packing, nearly 300 ml of fresh hexane was passed through the column. Nealry 1 g of F-ML3 was dissolved in 25 ml hexane and applied to the column which was eluted with the same solvent at 2 ml/min. Fractions (200) were collected at 10 min intervals, and every fifth fraction (or more as needed) was analyzed by GC. GC detector response standard curves (log area counts vs. log gram congener) were used to determine the amount of each congener in the fractions. Similar fractions were pooled into five separate fractions from which congeners _1_, 2, 2, Z, and 2 were purified by recrystallization and reversed-phase Lipidex-5000 chroma- tography. Recrystallization in Hexane Final purification of congeners t, 2, 2, Z, and 2 was accomplished either by successive recrystallization or by both recrystallization and reversed-phase 59 6:326 988 c3 260522 ocm 22.552 mom wuam awhm<2wm~u _ZOME ZO_Hfiocoo8820 @000me8_:_._ 3.8.3 SP. 3. P. 3.8. 68.78. m 36.5 o 36.5 326.5 «36.5 .3660 >88026EE8 65:56 .2662 B 29:62. .8220 AN humhnwuwflwmuw 6:268 :_ 8:35.338? «268830: mafia? I... T... 32:8 5 6.3.668 :66: t 3.2-“. 68.80 c. 6.3.668 2068.“: ..s.-.... .2668-.. .42-... 6:286 6:236 :_ :ozouzoaboom :_ 8_BN___2»>L0 C: 8:02“. 0328:. F: 225.". 8:8: 05235 go 056:: 68 88:33 0868 5.3 8826.2. 045 3.862.“. @ N umDuE 61 Lipidex-5000 column chromatography. Recrystallization proved to be a useful technique when the congener of interest constituted at least 50% of the total sample and when other structurally similar congeners (see later) were not present. Between 15 and 100 mg of each of congener 2, 2, or 2 enriched fractions was transferred into a 50 ml screw-capped culture tube and dissolved in 1-10 ml of hot (50-600C) hexane. Since the congeners have vastly different solubility in hexane, heating (600C in a water bath) and frequent vortexing of the tightly capped tubes were necessary to achieve a saturated solution. The completely solubilized samples were then allowed to crystallize at room temperature in the dark for a day or more. The crystals were dissolved and recrystallized further (two to four more times) until pure congener 2 (>99%) and congener 2 (> 97%) were obtained. Congener 2 could only be recrystallized to 80-85% purity. Further recrystallization or column chromatography on alumina could not separate this congener from the structurally very similar congener 6. Similar attempts to purify congener 2 from the structurally very similar congener g were not successful either. Lipidex-5000 Reversed-Phase Chromatography Between 150-200 g Lipidex-5000 (suspended in methanol) was centrifuged at 1000 x g for 10 min to remove methanol. The gel was suspended and stirred in two volumes of acetonezmethanolzheptane (3:1:1). This slurry was centrifuged another time and the solvent was removed and discarded. The gel was then suspended in three volumes of the same solvent and stirred in an ultrasonic bath to break up adhering gel beads. The slurry was then left 2-3 hr to equilibrate before packing. The gel suspension was poured into a 55 x 4.5 cm glass column, equipped with a Teflon stopcock and a glass fritted disk, which was partially filled with the solvent mixture, and packed by gravity as the column drained. About a liter of the same 62 solvent was passed through the column. The same column was used repeatedly to chromatograph a large number of different PBB samples with no significant cross contamination or deterioration in performance. Before each application, however, about half a liter of the solvent mixture described above was passed through the column. Usually 50-200 mg of a particular sample may be dissolved and applied to the column in a minimum volume of the same solvent mixture. The column was developed at room temperature at a flow rate of nearly 2 ml/min. Fractions were collected at 10 min intervals and dilutions were analyzed by GC. Recycling the less pure fractions was also practiced. Impure fractions that eluted from one or more columns were pooled and rechromatographed on the Lipidex column. 11. Purification of Congeners _3_, 6_a, 6p, and _1_0_ Each of congeners _3_, 6_a, 6p, and _12 represents less than 1.5% of either Firemaster mixture, and therefore, purification of relatively large quantities of these congeners is rather difficult. The major objective, therefore, was to obtain sufficient amounts of these congeners mainly for structural characterization studies. Congener _12 was purified from a crude fraction containing 22%, 53%, and 25% of congeners t2, Q, and _1_2, respectively. This fraction was produced during the process of purifying congeners 2, tt, and _1_2, from an octabromobiphenyl mixture (manufactured by Aldrich Chemical Co.) by repetitive alumina chromatography and recrystallization (Moore, 1978). The fraction containing 22% congener Q was a donation by Robert W. Moore. Reversed-phase preparative TLC chromatography, described earlier, was the only technique that was used to obtain pure congener 22. Using this technique congener _1_(2 had a larger R value (greater f mobility) than either of congeners t1_ and 1_2_. It was extracted and freed of paraffin oil as described earlier in this section. Nearly 5 mg of over 98% pure 63 congener _1_2 was recovered, and was used for structure characterization by GC-MS and lH-NMR spectroscopy. Congeners 2, 6g, and 6b_ were purified from Firemaster FF-l since this mixture had relatively more of these congeners than Firemaster BP-6 (the last two congeners are believed to be exclusive to Firemaster FF-l). Firemaster FF-l (2 g) was fractionated by neutral alumina chromatography (2 columns) in hexane as previously described (Moore, 1978), and the fractions that were most enriched in congeners _3_, 6_a and 6_b were pooled separately. The combined fractions that were most enriched in congener 2 contained nearly 10% of congener 2 and almost 85% of congener 2, in addition to minor amounts of congeners 2 and 2. This sample was rechromatographed on a similar alumina column, and fractions enriched in congener _3_ were combined and rechromatographed a third time. The most concentrated fractions of congener 2, which were mainly contaminated with congener 2, were pooled to give slightly less than 20% pure congener 2. At this point it seemed that further alumina column chromatography would not significantly improve the separation of congeners 2 and 2. Reversed—phase TLC (as describe earlier) can, however, separate the two congeners. After extraction and oil partitioning, nearly 6 mg of over 98% pure congener 2 was recovered and used for structure elucidation by lH-NMR. The two alumina columns on which 2 g of Firemaster FF-l was chromatographed to obtain pure congener 2 were also used to purify congeners 6_a and @. Fractions containing both congeners, which eluted together, and were mainly contaminated with congeners 2, 6, 4_, and 2 (both 6_a and 6_b were 15% of the total congeners), were combined. Reversed—phase preparative TLC was also used to purify congeners 6_a and 6_b from this rather crude mixture. Each plate was visualized as described earlier, and samples corresponding to different spots were analyzed by GC. Areas corresponding to congeners 6_a_ and 62 were scraped off the 64 plates, extracted with hexane, and separated from the paraffin oil as described earlier. Congener _6_a was about 80% pure and was contaminated with nearly equal amounts of congeners fl and g, but was mostly free of congener Q. The sample containing congener 6_b_ was apparently free of congener 2a, but was only 54% pure since it contained 41% and 5% of congeners g and 2, respectively. Another attempt was made to improve the purity of congeners _6_a_ and 6_b, where each of the two samples was rechromatographed on three TLC plates. The plates were treated as described earlier to obtain congeners 2a and _6_t_J_ at concentrations of 85% and 60%, respectively. The two samples, containing congener g as the major contaminant, were used to characterize the structures of congeners 2a and 6_b. RESULTS Purification of PBB Congeners There were at least two objectives for obtaining purified PBB congeners. The first was to characterize those PBB congeners whose structures were partially or completely unknown. For most of these studies the main priority was assigned to acquiring sufficiently pure samples for structure assignment studies. No further details on the results of purifying these congeners will be provided beyond what has been presented in the Materials and Methods section. The second objective was to test the chemical and biological properties of as many PBB congeners as possible. Such studies would require absolutely pure compounds in quantities larger than normally needed for structural studies. Results of this research will be presented in some detail in the following sections. rz. .‘ ' 65 Purification of Congeners {L and 6 An outline of the scheme used to purify congeners fl and 2 as well as five other congeners is shown in Figure 2. The method for purifying congener fl is similar to that of Bairstow gt; a_l. (1978). Nearly 65% (by weight) of the Firemaster BP-6 starting material was recovered in the insoluble fraction (1F) which was relatively enriched in congeners fl and 2 (Table 1). Nearly 14 g of congener 2 (>99% pure) was recovered after this fraction was completely solubilized in a sufficient amount of acetone and allowed to crystallize. This represents almost 28% recovery of congener 3 since it constituted nearly 50% of Firemaster BP-6. The mother liquor fraction (IF-ML) was enriched in congener 2 and was therefore combined ‘with the similar F-Crystalsl fractions for purifying congener 2. Successive recrystallization (4 times) yielded nearly 700 mg of pure congener 2 crystals (> 99% pure). This is about 13.5% recovery from Firemaster BP-6 which contains 5.2% congener 2. Purification of Congeners 5 and 2 Congeners 2 and 2 could be purified from F-ML3 by a combination of recrystallization and alumina chromatography. By allowing the original filtrate (F) fraction to crystallize, the resulting mother liquor (F-MLl) became enriched in some minor components, including congeners 2 and 2. Concentration and crystallization of F-ML twice yielded nearly 5 g of material (F—MLB) in which the 1 two major congeners _4 and 2 decreased from 47.8% and 15.1% in Firemaster BP-6 to 12.6% and 6.3%, respectively (Table l, and Figure 38). At the same time congeners 2 and 2 increased from 11.6% and 0.8% in Firemaster BP-6 to 29% and 45%: respectively in F'Ml—3 (Table 1). Figure 3b shows the elution profile when l g of F-ML was chromatographed on alumina (not shown are the elutions of 3 congeners 2 and 2 which eluted mainly in fractions 39—70 and 39-80, respectively). 66 6326866 so: u 0.2 n 3205523 0cm 95525500 .50 N 0.59:.— mmm m .O.Z .D.Z .O.Z 0% m6 0.5 0.0 0.0m 0.2 0.0 3.3 0.0a m.:>_-u._ .D.Z .D.Z .O.Z .G.Z 0.3 04 0.m 0.5. 0.3. m.0 m.N 0.0 23:20-11: .O.Z .O.Z 04 .D.Z 0.0 .O.Z 0.9V m.N 0.5m .D.Z 0.2 0.m aflmumiuiu T0 «.0 m.0 0.0 0.3 0.N 0.0 0.0 0.00 0.0 in m.N 1.: .O.Z n.U.Z m4 0.q 0.0a mic m.m 0.NN «.2 0.m 0.0a m0 1.. m.0 70 N.0 0.0 HRH 0.N Nd 04H 0.mm 04 m.m N.m wuam amummEmaC II. III I m M a m s N H H H 0 r—l oxl MI NI HI 233 00 £0 mocwocand. amcoocoo mcosomi .QcmEmSmmoE 009:3 03203.: oEobomB E9; mam 333m A 2:91.. .50 DmnCOmmU $9: 3 25.8950 $2309: 50:: DO .3 85:95 295 mcofiomau mcofiomau waotm> E mamcmmcoo 00a .5 55:07:20 H 0.569 67 FIGURE 3. (a) GAS CHROMATOGRAPHIC ELUTION PROFILE OF F-ML3 BEFORE NEUTRAL ALUMINA CHROMATOGRAPHY The GC conditions were similar to those in Figure 1 except that the amount of sample injected was not determined. Congener numbers are indicated over the GC peaks. (b) NEUTRAL ALUMINA CHROMATOGRAPHY ELUTION PROFILE OF CERTAIN PBB CONGENERS OF F-ML3 One gram of F-ML3 was applied to a 40 x 2.5 cm column packed with 110 g of neutral alumina. The column was eluted (2 ml/min) with hexane, and 20 ml fractions were collected and analyzed by GC. The amount of each PBB congener in the fractions was determined from standard detector response curves. 68 (a) IOOr 2 5 FIGURE 3 u 0) 2 O a. m In] “7' l m 50- Z ‘5 4 u C 0 L I L L I l l 4 0 4 8 I2 I6 20 24 28 RETENTION TIME (min) (b) 42 I T l T I Br Br I Congener —.—.— Br © C) I Br Br 36' B, . Congener —-.— Br 6) (9 Br Br J Br Br 30? Congener —H— Br (9 © Br 5 1 Br Br , Br Br Br éZ‘I- t . Congener —o-o— tar—{69m . L r Br 3 I 8 Congener —-*— Br {9 © : 3 gr Br ': Congener —o<>— avg-tar Br r lab : -r T Congener 9 —-- 5:38-83: Br Br J 6' ‘ &' 4 0 Kb 3 \ ° 5: I ' 'L' ’-‘_‘?.‘-_ - 20 4O 60 80 I00 l20 I40 I60 I80 Fraction Number 69 Fractions 55-78 yielded nearly 200 mg of material of which 75% was congener 2. At least 100 mg of 99% pure congener 2 can be recovered from this fraction by repeated recrystallization in hexane. At least an additional 400 mg of congener 2 may be recovered from the remaining 4 g of F-ML3 by similar alumina chromatography and repeated recrystallization. Therefore, based on its abundance in Firemaster BP-6 the yield of pure congener 2 could be at least 4.5%. Nearly 30 mg of material was recovered in fractions 151-200 and 80% of the material was congener 2. This was repeatedly recrystallized in hexane until at least 97% pure congener 2 was obtained. Less pure fractions containing from 55 to 75% congener 2 may also be pooled and recrystallized. A total of at least 25 mg of 97% pure congener 2 may be recovered from 1 g of F-ML3 material. An additional 100 mg of this congener can be obtained from the remaining 4 g of F-ML3. The yield of congener 2 from Firemaster BP-6 may be as high as 15%. Purification of Congeners 1: 22 and Z Final purification of congeners 2, 2, and 2 was achieved primarily by reversed-phase Lipidex-5000 column chromatography (Nystrom and Sjovall, 1975). The fractions from the alumina column that were most enriched in congener 2 (fractions 25-30), congener 2 (fractions 31-38), and congener _7_ (fractions 105-125) were pooled separately (Figure 3b). Compared to their abundance in F-ML3 there was a 4-, 3-, and ll-fold purification of congeners 2, 2, and 2 respectively (Table l and Figure 2). Lipidex-5000 column chromatography of the pool enriched in congener 2 resulted in the elution profile shown in Figure 4b. Most of congener 2 (65 mg) eluted in fractions 23 and 24 and was 99% pure. Nearly 35 mg of the same congener was distributed over fractions 25-27 which also contained most of congeners 2 and 2 (Figure 4b). Nearly 60 mg of congener 2 were obtained upon crystallizing fractions 23 and 24. An additional 240 mg could be recovered if the 70 FIGURE 4. (a) GAS CHROMATOGRAPHIC ELUTION PROFILE OF CONGENER 2-ENRICHED ALUMINA COLUMN POOL (FRACTIONS 25-30) The GC conditions were similar to those in Figure 1 except that the amount of sample injected was not determined. (b) SEPARATION OF CONGENER 2 FROM CONGENERS 2, 3, AND 2 BY LIPIDEX-5000 CHROMATOGRAPHY The mixture (nearly 200 mg) was applied to a 55 x4.5 cm column packed with nearly 200 g Lipidex-5000 gel. The column was eluted with acetone:methanol:heptane (3:1:1) at 2 ml/min, and each fraction (20-23 ml) was analyzed by GC. The amount of each PBB congener in the fractions. was determined from the GC standard detector response curves. RELATIVE RESPONSE (M AMGJNT (Mg) 71 IOO- FIGURE 4 I G C I I 50- 4 2 OJ L 6 . l I I i l l I J O 4 8 I2 I6 20 24 28 RETENTION TIME (MIR) LIPIDEX‘ 5000 60f Br r —rr—I— = 37,—? (I) w‘ : .r 8r ml '**H§H%ym Br Br 30‘I -O—-O— : BrHBr (4) Br Br 20- Br Br -¢—+—: 3',—*3, (5) Br -r IO‘ éiaaamammmawauamwifik FRACTION NUMBER 72 remaining 4 g F-ML3 was also sequentially chromatographed on alumina and Lipidex—5000. Therefore, relative to its original concentration in Firemaster BP-6 the yield of congener 2 could be at least 9.5%. The fractions (31-38) from the alumina column that were enriched in congener 2 weighed 388 mg of which 20% was comprised of congeners 2, 2, and 2 (Figure 3b). When crystallized in hexane nearly 170 mg of crystals containing 85% congener 2 were obtained (Figure 5a). Several attempts to remove the contaminant, congener 2, by alumina chromatography or recrystallization were unsuccessful. However, separation by Lipidex-5000 column chromatography was successful (Figure 5b). Approximately 120 mg of congener 2 (> 99% pure) was recovered in fractions 25-27 (Figure 5b). An additional 20 mg eluted with 30 mg of congener 2 in fractions 28-32. These later fractions can be pooled and rechromatographed on the same column to increase the recovery of congener 2. Chromatography of a relatively less pure sample (from 20 to 80% congener 2) on the Lipidex column did not result in any significant loss in the resolution of congener 2 from congener 2. Furthermore, the elution volume did not seem to be affected by the amount of sample applied to the column (compare, for example, the elution volumes of congeners 2 and 2 in Figures 4b and 5b). The same results were obtained for Lipidex column chromatography of congeners 2 and 2. After crystallization in hexane nearly 120 mg of congener 2 (> 99% pure) was obtained. An additional 480 mg can be recovered from the remaining 4 g of F-ML3 by alumina and Lipidex chromatography. The yield of congener 2 relative to its abundance in Firemaster BP—6 is nearly 11.3%. Fractions 105-125 from the alumina column (Figure 3b) contained 45 mg of material, 75% of which was congener Z. This material was chromatographed on the same Lipidex column (elution profile is not shown). Most of the contaminants, including congeners 2, 2, 2, 2, and 2, eluted before congener _7_, and congeners fl, 73 FIGURE 5. (a) GAS CHROMATOGRAPHIC ELUTION PROFILE OF CONGENER 2 BEFORE LIPIDEX-5000 CHROMATOGRAPHY The GC conditions were similar to those in Figure 1 except that the amount of injected sample was not determined. (b) SEPARATION OF CONGENER 2 FROM CONGENER 2 BY LIPIDEX—5000 CHROMATOGRAPHY The mixture (167 mg) was applied to the same Lipidex column that was used in Figure 4b under identical conditions. Congeners 2 and 2 were analyzed and quantitated as outlined in Figure 4b. (a) RELATIVE RESPONSE (b) AMOUNT UWQ) 74 IOOF- FIGURE 5 2 G C SOI- 6 .I L L l I L l l o 4 8 I2 IS 20 24 28 RETENTION TIME (min) LIPIDEX- 5000 60 50‘ 40‘ w. 20 d 8r H. a. I 2) r Br H: BrfiBde) Br :I 1 A..- l2! 22824252627282.9303 323334353637 FRACTION NUMBER o 75 _l_._l_, and £2. eluted after congener Z. Fractions 30—32 were pooled to give 25 mg of 87% pure congener Z. Another 30 mg of material of similar composition was collected from an identical experiment and after combining both samples they were rechromatographed on the same Lipidex column. Nearly 45 mg of 92% pure congener 1 was recovered. About 20 mg of highly pure (99%) congener 1 could be obtained by crystallization in hexane. The mother liquor of these crystals can be combined with a number of other Lipidex fractions of similar purity (BO-85%), then repeatedly recrystallized in hexane. An additional 20 mg of pure congener Z (98%) can thus be recovered. Since nearly 20 mg of pure congener Z may be recovered from 1 g of F-Ml.3 it is estimated that the yield may total 100 mg. This is a recovery of nearly 3.5% relative to congener _7_ in Firemaster BP-6. Characterization of Several PBB Congeners Structure assignments have been made after considering the following general assumptions derived from 1l-I-NMR spectral studies on a number of PBB congeners (Moore and Aust, 1978; Moore, 1978). Chemical shifts of 7.8-8.0 ppm, 7.3-7.7 ppm, and 6.9-7.2 ppm are indicative of protons adjacent to 2, l, and 0 bromines, respectively. The chemical shifts for protons adjacent to both a brominated carbon and a bridge carbon fall into three categories, depending on the total number of 911239 bromines. In the presence of only one M bromine the m protons have chemical shifts between 7.54-7.59 ppm. When two m bromines are present the chemical shifts of the M protons range between 7.45-7.49 ppm. With three M bromines chemical shifts of 7.37 ppm and 7.38 ppm were reported for the single gm proton of two congeners, }_1_ and g, respectively. Coupling constants also fall into three categories. Two protons that split each other with coupling constants of 7-9 Hz, and 2-3 Hz should be m and meta to each other, respectively. Unsplit signals (3 <0.5Hz) are due to protons 76 that are either pa_r§ to each other or are on opposite rings. Furthermore, for biphenyls containing 3-7 bromines, two singlets (J <0.5 Hz) necessarily mean that one ring has a 2,4,5-tribromo substitution pattern, since this is the only possible w proton arrangement. With these considerations the following assignments were made. Congener 3 is a hexabromobiphenyl, of which 42 possible isomers can exist. A highly purified sample (> 98% pure) of congener _3_, whose GC profile is shown in the insert of Figure 6, gave rise to four proton NMR signals, two of which are singlets, and have chemical shifts at 7.960 ppm and 7.405 ppm. These two signals must be due to the two protons of a 2,4,5-tribromo substituted ring. Protons on the second-ring gave rise to two _o_r_th_o-split signals (J = 8.6 Hz), and, therefore, the two protons must occupy either the 4 and 5, or the 5 and 6 positions. However, the chemical shifts at 7.572 ppm and 7.504 ppm indicate that each proton is adjacent to a bromine, which rules out the second possibility. The second ring, therefore, has bromines at the 2, 3, and 6 positions. The structure of congener 3 is, therefore, 2,3,6,2',4',5'-hexabromobiphenyl (Figure 6). CBC-MS analysis of congener _7_ indicated six bromines. The NMR splitting pattern and coupling constants (Figure 7) indicated two ELIE-coupled protons one of which was also {pita-coupled. Therefore, one ring must have three protons Which are split ortho, meta, and ortho + meta. Three structures can be postulated for this ring where the two bromines are at carbons 2 and 4, 2 and 5, or 3 and 4. But the ortho + meta-split proton at 7.134 ppm could not be adjacent to a bromine since its chemical shift is not nearly far enough downfield. Therefore, the first tWC) possibilities can be ruled out, and a 3,4-dibromo substitution is the only POSSible structure for this ring. The single proton on the second ring, with a shift Of 7.538 ppm, should be adjacent to only one bromine, and therefore it must be fig to the bridge carbon. The only structure that agrees with these chemical ¥ 77 FIGURE 6. 1H-NMR SPECTRUM AND STRUCTURE OF 2,3,6,2',4',5'-HEXA- BROMoeIPI—IENvL (CONGENER 2) The gas chromatographic profile of purified congener 2 sample is shown in the insert. 78 CHCI3 RESPONSE CD "" 7.44 l l I 7.50 l LJ ’ i O 6 l2 RETENTION TIME (min) l l L l l l L l l l L L l l 8.00 7.00 6.00 5.00 ppm FIGURE 6 79 shifts, splitting pattern, and coupling constants for the four protons of congener Z is 2,3,4,5,3',4'-hexabromobiphenyl (Figure 7). The NMR spectrum of congener 2 (Figure 8) indicates that two _grt_h_o-coupled protons are present. Since the signal for one of these is at 6.998 ppm it cannot be adjacent to any bromines, so these protons occupy the 5- and 6-positions on one ring. The unsplit proton must be on the other ring, and from the magnitude of its chemical shift (7.460 ppm) it has to be adjacent to one bromine. This is only possible if this proton is at a Carbon o_r_th_o to the biphenyl bridge. Therefore, the only possible structure for congener _9_ is 2,3,4,5,2',3',4'—heptabromobiphenyl (Figure 8). GC-MS analyses showed that congeners 6_b (60% pure) and E (98% pure) are a hepta- and an octa-bromobiphenyl, respectively. Congener 6_b was mainly contaminated by congeners 4 and, to a lesser extent, _8, but both were adequately resolved from congener 6_b to permit an accurate mass spectral analysis. The NMR spectral analysis of congener 6_b sample revealed five major unsplit signals at 7.97, 7.96, 7.93, 7.47, and 7.38 ppm. The signals at 7.93 and 7.47 ppm should arise mainly from congener 4 since they seem to exactly match the chemical shifts of pure congener 4 (Moore, 1978). The remaining three signals are all singlets, and have an integral ratio of unity to one another (Table 2). The presence of these singlets indicates that one ring is 2,4,5-tribrominated, where the proton at position 6 should give rise to the signal at 7.38 ppm, while either of the remaining two signals could be due to the proton on carbon number 3. The remaining proton, with a chemical shift of 7.96 or 7.97 ppm, should be adjacent to two bromines, and, therefore, it should occupy a position w or meta to the biphenyl bridge. The structure of congener 6_b, should, therefore, be 2,3,5,6,2',4',5'- or 2,3,4,6,2',4',5'- heptabromobiphenyl. 80 FIGURE 7. lI--l-NMR SPECTRUM AND STRUCTURE OF 2,3,4,5,3',4'-HEXA- BROMOBIPHENYL (CONGENER Z) The gas chromatographic profile of purified congener Z is shown in the insert. RESPONSE 81 HGURE7 Br Br CHCI3 Br Q Br Br Br _J\ A' IK\L_ 0. 6 E R RETENTION TIME (min) 7, l34 ZSBI 7687 7. so 7, 60 7.40 7.20 7.00 ppm 82 .fimmE 05 E 550% 2 29:3 cotton >22“me 05 Co 0503 oanEmsmEoEo mom 9:. Am muZmozoov u>ZmIa50.20%?EmlénfiafiiRN uo umoeoompm oz< EDEBEm £22-} .m mmDOE 83 m UZDOC 88 com, 8... con 8.... 08. 08 pun.r___‘____________..m_.._4_d___. I 5% is} E}... .m cm 1 I m .m cm I I y mmN «EEC m2; ZOFZMHm—m N. m 0 u d NOV ESNOdSBH _n 0 I 0 more 84 s m u m 5 ©©s ...I II Soc .5 22: 3 mmmN cm 55 cm . onto 3 ONmN 5 5 cm 50*: © no 5cm 5 . Sam 3 22: 3 000m 55 a .073. g... A... s... cm .m cm .55 cm 5: <2 E some .m 5 5 <2 3 wmmN .55 5:6 .0 3 mth .m a .m :205 E2: m mmmzhonmhm wumfimon. mkn=Im 440.295 220350805209 eéocafioEoEofig O. 222930805285 3 mam—232 mwzwozoo mm...m<_2mm_n_ mmmzmozoo mma .QMN_mm._.oI_u_<_km<& mo mmmahoamhm mumfimoa oz< mhmfm I_85% pure) gave rise to a molecular ion cluster characteristic of a heptabromobiphenyl. However, the sample was too small to obtain an 1H-NMR spectrum with clear signals (signal to noise ratio was too low), and, therefore, the structure of _63 remains unknown. Congeners. 63 and _6_b are believed to be present in Firemaster FF-l but not in Firemaster BP-6. In the GC chromatogram of Firemaster FF-l peak 7 is believed to be composed of at least three PBB components which include congeners _6_a_, 60 and Z (Figure l). Congener 6a appears as a shoulder peak between the peaks of congeners 6 and Z. Congener _6_b_, however, has a very similar retention time to that of congener Z, and, therefore, both give rise to a single unresolved peak. The GC retention time of purified congeners _6_a_, 6_b, and Z relative to that of congener _6_ is 1.059, 1.105, and 1.158, respectively. Some Chemical and Chromatographic Properties of P88 Congeners Some gas chromatographic properties of PBB congeners of Firemaster are summarized in Table 3. First of all, it can be concluded that the retention time (tR) for the elution of the PBB congeners from the gas chromatograph (GC) is directly proportional to the total number of bromines. Two pentabrominated congeners (_1_ and g) have smaller retention times than any of the five hexabrominated congeners (2, fl, 2, Q, and Z) whose retention times are yet smaller than those of the heptabrominated congeners 8 and 9. And congeners 12, Q, and 86 22 30.0 R . N on .m a: -.m..¢..n..~.m.e.m.m --- --- --- 0.: 3|: 020 3% 8|: Pm. $0 804 BA :0 Amy -.e..m..~.m.e.m.w 03- 000.0 80 an .N a. -.m.e..N.m.q.m.N and 30.0 $4 34 3 ginninN ohm «00.0 00.. RA .8 Jeannfiiw 0.3 03.0 30 MN; 9 -.m..e..~.e.m.m .22. 30.0 80 84 av -.m..e..N.n.e.N o: 30.0 30 5.0 AU -.m.e..N.0.m.N 3.0 80.0 on; 2.6 9 -.s..m.m.e.~ men 30.0 84 N90 3 0.0.03 mg x 3:3000 ANEV min. 0.0EE M 0:09:00 0P 0:0:0m:o0 mmn. A000000u£u> 20.0.3000 33 x 3:300 my 0>30_0m. :0320200 00:00.03”. 00.00000 20:00:00 mmn. 0.50. on. 008801-000 0:0 Am: 00E: :ozc0u0m 0.0 0>SE0m M 015.09 87 _1_2, which are the most substituted (each has 8 bromines) are the last to elute from the GC column. Relative to the retention time of congener 4 all PBB congeners have retention times that range between 0.52 and 5.30 (Table 3). The presence of bromines at carbons o_rtfl)_ to the biphenyl bridge also seems to determine the GC retention times of PBB congeners. This is best illustrated by comparing the GC retention times of isomeric PBB congeners where there is an inverse relationship between tR and the number of M bromines (Table 3). Of the two pentabrominated compounds congener 2 with a single o_rtm bromine has a longer tR than the di-gth_o brominated congener Z. And among five hexabrominated congeners, the smallest t is observed for the only congener with three ortho R bromines (congener _3_). The mono-93mg brominated congeners _6_ and Z have longer retention times than the di-gtjlg brominated congeners 4 and 5. Furthermore, of two isomers containing equal numbers of M bromines the one with a 2,3,4- substitution pattern in one or both rings seems to have a longer tR than its isomer which has none or only one 2,3,4-tribrominated ring, respectively (a 2,3,4- tribromination may also be included in a 2,3,4,5-tetrabrominated ring). Compare, for instance, the retention times of congeners 4 and 2, congeners _6_ and Z, or congeners 8 and 2 (Table 3). In each pair, the congener with one more 2,3,4- tribromination pattern is the one with a higher t Finally, the non ortho R' brominated 3,4,5,3',4',5'-hexabromobiphenyl, which is not present in Firemaster (DeKok _e_t _a_l_., 1977), has a nearly identical tR to congener _8_; and therefore, it has a larger tR than any hexabrominated congener of Firemaster (data not shown). For each PBB congener there was a linear relationship between the amount of injected PBB (10-2000 pg) and the GC-ECD response with a correlation coefficient > 0.98 (Table 3). The slopes of these standard curves, expressed as integrator counts per mmole PBB were of the same order (1016) even though their actual Values varied within a range of nearly 2-fold. No apparent relationship seemed to 88 exist between molecular structure, including total bromine content, and the GC-ECD response for the PBB congeners of Firemaster. In Firemaster FF-l and Firemaster BP-6 the abundance of each congener was determined from its corresponding standard curve after an exact amount of either mixture was injected into the (3C (Table 4). Congener 4 is the most abundant PBB where it constitutes nearly 50% of either mixture. Congener 8 is the second most abundant congener and comprises nearly 15% and 25% of Firemaster BP-6 and Firemaster FF-l respectively. Therefore, BP-6 may be expected to contain nearly 10% more of the remaining congeners than Firemaster FF-l. This seems to be true since the total concentration of congeners _1_, _2_, 5, _6_, and Z is nearly 10% higher in Firemaster BP-6 than in Firemaster FF-l. Congeners 9 and _1_2_, however, are slightly more abundant in the second of the two mixtures. Nearly equal concentrations of congeners 3, 19, and Q are present in the two mixtures. Table 4 also lists the percent PBB composition of the two Firemaster mixtures as determined directly by the digital CDS electronic integrator. This method assumes an identical detector response to all PBB congeners. Values from this method can thus be compared to those that were obtained by using the standard curves of the individual PBB congeners (Table 4). With only one exception, the two methods yielded results that were within 25% of each other. Using the standard curve response to pure congener Z nearly 5% of this congener was estimated to be present in Firemaster BP-6. In contrast, only 2.9% of the same congener was directly measured in the same mixture by the electronic integrator. A reconstituted mixture was formulated by mixing nine pure PBB congeners in pr0portions similar to their concentrations in Firemaster BP-6. The percent abundance of each congener, which was directly measured by the electronic integrator, is also presented in Table 4. The amounts of most congeners in the reconstituted mixture seem to closely match their respective concentrations in Firemaster BP-6 (Table 4). The 89 mdv m5 mA 0.0 0.N a To 0.0 H v a v 0 N6 0.0 H v a v 9.0 0.0 0.0 H4 04 RNA 0.2 CNN 13 5.3 0.0 Rm 0.m Rm m0 m0 Nd N.m 0.0 Rm 0.: 0.: mfi RNA 0.0 0.0m 700 n.0m 0.3 79 0 m4 NA 04 m4 Rm Om 0.0 N0 Co 0.0 NR RN 0.0 Rm 0:352 0-0m. :. o\o HIRE :. o\o 0-0m :. oxo 70.1.. :. axe 0003:2000m :20 @A V 340 Am. Amv .9 AU 00. V .9 3 AU 9 3 0.3.0.0 .m . a. fim m:m. m 0m. -.e..m..~.m.e.m.w -.m..¢..~.m.e.m.m -e..m.m.q.m.w 0.3.0ng -.m.e..N.q.m.N 10.4.0.3} -.m..¢..~.0.m.~ I.¢..M.W.¢.N 0.0.0.0..“ 3:0E003000S. “00:00 20:00:00 900. 0.5m. 3 00:0000N. 030.5005 0.:05020 E00“. 000-00 “.0 00>.50 0:00:05 E00... 20:20:00 man. 20:00:00 0032?. E0: 000.2 003x22 000330:oo0m 0 :. 0:0 Guam 0:0 Tun: 00000600.“. :. 0:0:0o:o0 mmn. .8 00:00:30.0. 0300.00. .0 0:50.... 90 biological effects of the reconstituted mixture were evaluated and compared to the same effects of the two crude mixtures of Firemaster (see Chapter 3). Pure PBB congeners were also compared for their thin layer chromatographic relative mobility (Rf) under reversed-phase conditions (Table 5 and Figure 9). Since the lower molecular weight congeners (not believed to be present in Firemaster) had relatively high mobility the plate containing the di- and tetra- brominated congeners was developed only once. The plates containing congeners with five or more bromines, however, were developed three times so that resolution could be improved. The chromatographic mobility of PBB congeners under these conditions seems to be determined by the total number of bromines, the number of bromines M to the biphenyl bridge, and by the presence of a 2,3,4-tribromo substitution pattern (Table 5). Isomeric PBB congeners have Rf values that are directly proportional to the number of bromines gym to the biphenyl bridge. 2,2'-Dibromobiphenyl, for instance, has a larger Rf value than the non-m brominated 4,4'—dibromobiphenyl. Similarly, in each group of tetra-, penta-, or hexa-brominated compounds, the congener with three _o_rt_hg_ bromines has the largest Rf value, while the least M substituted congener has the smallest Rf value. As compared to other substitution patterns, particularly the one which is the most common (2,4,5-tribromo), a 2,3,4-tribromo substitution also seems to increase the mobility of a PBB congener. Congeners 4 and _5_, are hexa- and congeners 8 and 9 are hepta-brominated with two 9349 bromines each. Each pair shares an identical ring whereas congeners 4 and _5_ have a 2,4,5-tribromo substituted ring and congeners 9 and 9 have a 2,3,4,5-tetrabrominated ring. In each pair, however, the second congener has a larger Rf value, and it is the one which contains a 2,3,4-tribromination pattern on the second ring. Also, each of congeners Q and Z has a single ortho bromine, but only congener Z has the 2,3,4-substitution pattern, which is included in the 2,3,4,5-tetrabrominated ring, and it has the larger 91 TABLE 5 Reversed-Phase TLC of PBB Congeners Total #Ortho Amount #Times Compound #Br Br Applied (ug) Developed Rf 4,4'- 0 5 1 0.68 2 2,2'- 2 5 l 0.84 3,5,3',5'- 0 5 1 0.37 2,4,2',5'- ' 4 2 5 1 0.60 2,5,2',5'. 2 5 1 0.64 2,5,2',6'- 3 5 1 0.76 2,4,5,3',4'- (2) 1 4 3 0.78 2,4,5,2',5'- (y 5 2 4 3 0.84 2,4,5,2',6'- 3 4 3 0.88 2,3,4,5,3',4'- (Z) 1 4 3 0.75 2,4,5,3',4',5'- (g) 1 4 3 0.67 2,4,5,2',4',5'- (4) 6 2 4 3 0.73 2,3,4,2',4',5'- (g) 2 4 3 0.82 2,3,6 , 2',4',5'- (3) 3 4 3 0.89 2,3,4,5,2',4',5'- (£3) 2 4 3 0.70 2,3,4,5,2',3',4'- (_9_) 7 2 4 3 0.80 2,3,4,5,2',3',4',5'- (_1_2_) 8 2 4 3 0.68 92 FIGURE 9. THIN LAYER CHROMATOGRAPHIC SEPARATION OF PENTA-(PBB) AND HEXA-(HBB) BROMOBIPHENYLS Each congener (3-5 pg) was spotted on a paraffin oil-coated Kieselguhr-G-TLC plate which was developed (3 times) in acetonitrile:methanol:acetone:water (20:20:9:1). Spots of PBB were visualized under UV-light after spraying with an ethanolic solution of AgNO3 containing NHAOH. Among the penta- and hexa-brominated congeners notice that the R values are directly proportional to the f number of ortho bromines. 93 Front 5! n e W o S _ -Origin 00:-.0..n..~.n.¢.~ 00:-.0..¢..~.n.¢.~ 00:-.0..n..~.0.¢.~ mm:-.n..¢..n.n.¢.~ .00: .6 :2 80.. 3 :2 000400.060 0000.00.00 000-3360 FIGURE 9 94 Rf value. Besides the number of 2529 bromines and the 2,3,4-tribromination pattern, the separation by TLC seemed to also be affected by the total number of bromines. In general, the mobility of PBB congeners seems to be retarded by increased bromination, and, therefore, an inverse relationship seems to exist between Rf values and the total number of bromines. However, the retarding effects due to increased bromination are counteracted by M bromination and by the presence of a 2,3,4-tribromo substitution pattern. The separation of PBB congeners by reversed-phase Lipidex-5000 column chromatography seemed to also be governed by the total number of bromines and by the number of bromines £332 to the biphenyl bridge. A direct correlation seems to exist between VR of a PBB congener and its degree of substitution. On the other hand, congeners with equal number of bromines have VR inversely related to the number of 21213119. bromines. Furthermore, the effects of these two factors on VR are equal and additive. For example, adding a bromine at the gar_a carbon of congener _1_ yields congener 4 whose V is larger (Figure 4b). A similar increase in R VR is also evident when an extra bromine (meta) is added to congener _2_ to yield congener 4 (Figure 5b). Adding an ortho bromine to congener _2_, however, caused no change in the V of the resulting congener 4. An increase in VR due to the R extra bromine is neutralized by an equivalent decrease in VR due to ortho bromination (Figure 4b). On the other hand, V increases when an ortho bromine R of a di-9r_tflc_i substituted PBB, such as congener i or 4, is moved to a non-grim position, as in congener g or _6_, respectively (Figure 4b). The best resolution is obtained when two compounds, such as congeners _1_ and g, differ in their total bromine content and the less brominated compound of the two has more o_rtm bromines. The degree of ortho bromination appears to also play a major role in determining the UV-absorption spectral characteristics of PBB congeners. In an 95 FIGURE 10. UV-ABSORPTION SPECTRA OF 3,4,5,3',4',5'-HEXABROMO- BIPHENYL (HBB) AND CONGENERS Z, 4, g, AND 9 Spectra were recorded in hexane at 5.0 ppm concentration except for congener _5_ whose concentration was 3.4 ppm. Notice in comparison to the spectrum of the m unsubstituted HBB the reduction in the intensity of the conjugation (k) band and its shift to a lower wavelength in the spectra of congeners with a single ortho bromine. 96 Absorbance l I I J I 1 I90 220 250 280 3K) 340 370 Wavelength (nm) FIGURE 10 97 ailing unsubstituted congener, such as 3,4,5,3',4',5'-hexabromobiphenyl, two major absorption bands with Amax at about 227 nm and 272 nm can be observed (Figure 10 and Table 6). The main band at 227 nm is caused by (TI-90*) electron transitions, while the band at 272 nm, commonly referred to as the k-band, is generally believed to arise from conjugation of the biphenyl system (DeKok e_t g}, 1977). The intensity of the k-band, therefore, should be proportional to the extent of interaction between the two phenyl rings, which in turn should be proportional to the degree of planarity of the biphenyl system. Each of three mono-M brominated congeners, _2_, _6_, and Z, has a less intense but well resolved k-band which displays a slight blue shift (1 max = 255-260 nm) (Table 6 and Figure 10). Di-M brominated compounds have the least distinct k-band which gets shifted further towards the UV-region, and, therefore, becomes hard to resolve from the main absorption band. The main UV-absorption band appeared at 214 to 229 nm. This band seems to undergo a slight bathochromic shift as additional bromines are added at 032911 and pa_ra_ positions to the biphenyl bridge. The main UV-absorption band of congener _1_ (K max = 215 nm) undergoes a slight progressive bathochromic shift in the order of increasing bromine content in congeners 4, _8_, and g, all of which, like congener l, have two thhg bromines (Figure 11 and Table 6). Similar bathochromic shifts are observed when an additional _mgtg bromine is added to each of congeners g and 9 to give congeners Q and 9, respectively (Table 6). _Qr_t_h_g bromination, however, seems to shift this band in the opposite direction. Congener _2_ with a single Elm—o bromine for instance, displays a )‘max at 222 nm, while its di-gglq brominated isomer, congener 9, has a Amax of 215 nm. Furthermore, among five hexabrominated congeners, the tri-ortho brominated congener 9 has the shortest A max (214 nm), while congeners _6_ and Z, each with a single ortho bromine, display the longest Amax (229 and 225 nm, respectively). Congeners 4 and 9 give rise to a A max at a 98 TABLE 6 UV-Spectral Data of PBB Congeners Main Band k-band PBB )‘max Loge Amax Congeners (nm) (M'l cm-l) (nm) 2,4,5,2',5'- (l) 215 4.53 229-231 2,4,5,3',4'- (9) 222 4.74 254-257 2,3,6,2',4',5'- (9) 214 5.01 229-231 2,4,5,2',4',5'- (4) 217 4.81 229-231 2,3,4,2',4',5'- (9) 216 4.81 228-230 2,4,5,3',4',5'- (4) 229 4.76 254-258 2,3,4,5,3',4'- (Z) 225 4.76 255-260 2,3,4,5,2',4',5'- (9) 224 4.86 228-232 2,3,4,5,2',3',4'- (9) 223 4.85 228-232 2,3,4,5,2',3',4',5'- (92) 227 4.93 229-231 3,4,5,3',4',5'- (H88) 227 4.86 272 99 FIGURE 11. UV-ABSORPTION SPECTRA OF CONGENERS 9, 4, 9, AND E Spectra were recorded in hexane at 5 ppm concentration. Notice the absence of k—band in the 250-270 nm region in all spectra of the di- ortho brominated congeners. 100 FIGURE 11 Absorbonce o 0 © 0 0 H l L L l J I90 220 250 280 3|O 340 370 wavelength (nm) 101 somewhat intermediate position (217 and 216 nm, respectively). Finally, the extinction coefficient of the main UV-absorption band seems to be similar for all the PBB congeners (Table 6). DISCUSSION ’ The gas chromatographic profiles of Firemaster FF-l (lot no. FH7042) and Firemaster BP-6 (lot no. 6224-A) are shown in Figure 1. At least twelve PBB congeners appear to be common to both mixtures. The structures of seven of these congeners have previously been characterized by GC-MS and lH-NMR spectroscopy. In this laboratory, congeners _1_, 9, 9, 9, 9 and 99 were assigned the structures 2,4,5,2',5'-penta-, 2,4,5,3',4'-penta-, 2,3,4,2',4',5'—hexa-, 2,4,5,3',4',5'- hexa-, 2,3,4,5,2',4’,5'-hepta-, and 2,3,4,5,2',3',4',5'-octa-bromobiphenyl (Moore, 1978; Moore and Aust, 1978). The major component of Firemaster, congener 4, has also been previously identified as 2,4,5,2',4',5'-hexabromobiphenyl (Sundstrom g; §_l_., 1976a; Jacobs g _a_l_., 1976). Congener l_l_ was partially characterized as 2,3,4,5,2',3',4',6'-, 2,3,4,5,2',3',5',6'-, or 2,3,4,5,6,2',4',5'-octabromobiphenyl (Moore 95 a_l., 1980). Two different empirical formulas, an octa- (DeKok _e_t_ 159., 1977) and a hepta-bromobiphenyl (Moore, 1978), have been reported for congener 1_0_. Peak number 7 of the gas chromatogram of Firemaster FF-l was believed to be composed of at least two different components, one or both of which being due to a heptabromobiphenyl, but a hexabromobiphenyl was also considered likely (Moore, 1978). In Firemaster BP-6 this peak was considered to be due to a heptabromobiphenyl (DeKok _ei a_l., 1977). In this research, besides recrystallization, several chromatographic methods were used to fractionate Firemaster FF-l and Firemaster BP-6 into individual PBB congeners. Using GC-MS and lH-NMR spectroscopy the structures of congeners 9, 102 Z, and 9 were determined as 2,3,6,2',4',5'-hexa-, 2,3,4,5,3',4'-hexa-, and 2,3,4,5,2',3',4'-hepta-bromobiphenyl, respectively. Congener 6_b was partially characterized as 2,3,4,6,2',4',5'- or 2,3,5,6,2',4',5'-heptabromobiphenyl, while congener .19 can have any of three structures, 2,3,4,5,6,2',4',5'-, 2,3,4,5,2',3',4',6'-, or 2,3,4,5,2',3',5',6'-octabromobiphenyl, one of which was also proposed to be the structure of congener _1_; (Moore fi a_l., 1980). Congener 99 should contain seven bromines, but its structure remains unresolved. In the gas chromatogram of Firemaster FF-1 peak 7 is believed to be composed of at least 3 congeners, 6_a, _6_b_, and Z with retention times relative to that of congener _6_ of 1.059, 1.105, and 1.158, respectively. Congener Z is believed to be the predominant, if not the only, component of peak 7 in the gas chromatogram of Firemaster BP-6 (Figure 1). The patterns of lH-NMR chemical shifts and coupling constants are similar to those seen with other brominated or chlorinated biphenyls (Moore and Aust, 1978; Moore, 1978; Welti and Sissons, 1972). Chemical shifts of protons 91m to the biphenyl bridge tended to decrease with increasing bromination at carbons oLho to the biphenyl bridge. On the other hand, deshielding effects were felt by protons that were adjacent (M) to one or two bromines. All protons had chemical shifts that ranged between 7.00 and 8.00 ppm. Signals with chemical shifts near either 7 or 8 ppm were due to protons that had no or two adjacent bromines, respectively. Intermediate chemical shifts (around 7.5 ppm) were due to protons with one adjacent bromine. The significant reduction in the GC retention time with increased _o_r_t_h_g bromination is presumably due to a corresponding decrease in the degree of interaction with the stationary phase. Retention times of alkyl substituted biphenyls were similarly decreased by increasing the number and size of the 93119 alkyl groups (Beaven e_t_ a_l., 1957). These effects were attributed to steric hindrance explained by a decrease in the planarity of the biphenyl rings and a 103 reduction in the overall degree of biphenyl conjugation, which, in effect, resulted in less interaction with the non-polar stationary phase (Beaven g a_l., 1957). Likewise, with increasing number of bromines 92313. to the biphenyl bridge, the dihedral angle between the two phenyl rings is expected to increase, with a concomitant reduction in the overall degree of biphenyl conjugation. This could result in less interaction between these non-planar PBB congeners and the stationary phase (methyl silicone, OV-l). However, interaction with the stationary phase seemed to increase as the total number of bromines is increased. Similar results were reported for PCB congeners (Hutzinger e_t_ _a_l., 1974). The gas chromatographic separation of polyhalogenated biphenyls may also be dependent on other factors including their volatility (boiling point), which may be inversely proportional to the degree of halogenation. The slight degree of variation (2-fold) in the GC-electron capture detector response to penta- through octa-brominated PBB congeners is consistent with previous results on PCB congeners (Hutzinger e_t_ _a__l_., 1974). The response to decachlorobiphenyl is over 500 times stronger than the response to 4-chlorobiphenyl. However, most of this difference occurs between the mono- to tri-chlorobiphenyls, while the response to tetra- through deca-chlorobiphenyls varies only by a factor of 2 to 3 (Hutzinger g9 91., 1974). Knowing the GC-electron capture detector response to pure PBB congeners, it was possible to determine their concentrations in Firemaster. These determinations are generally in close agreement with direct measurements by electronic integration of the GC eluting peaks. Few attempts have been made to resolve PCB mixtures by reversed-phase TLC, where the mobile phase is relatively more polar than the stationary phase (Kieselguhr coated with paraffin oil) (DeVos and Peet, 1971; Stalling and Huckins, 1973). No studies, however, have attempted to explore the relationships between 104 PCB structures and their chromatographic mobility. Using similar TLC conditions the results of this chapter indicate that as the number of 9rt_hg_ bromines is increased there is less interaction between isomeric PBB congeners and the non- polar stationary phase. These results fully agree with those concerning the effects of Mbromination on PBB separation by the GC. The retardation effects due to increasing total bromination are also consistent with the similar GC results. However, the apparent increase in the mobility of congeners with a 2,3,4- tribrominated ring contrasts with the GC retardation effects due to a 2,3,4- substitution pattern. Alumina column chromatography and recrystallization have until recently been the only two techniques for purifying sufficient quantities of a limited number of PBB congeners for biological studies (Bairstow _e_t a_l_., 1978; Moore _e_t a_l., 1978b, 1979). However, certain congeners including _1_, _2_, 9, 9, Z, and 9 could not be purified by these techniques (Moore and Aust, 1978). To purify such relatively minor congeners most of congeners 4 and 9, which together constitute nearly 65% of Firemaster BP-6, were removed by acetone extraction of the mixture and repeated crystallization of the filtrate (F) fraction (Figure 2). The resulting mother liquor (F-MLB) which contained 12.6% and 6.3%, respectively, of congeners 4 and 9 (Table l and Figure 3a) was then fractionated by alumina chromatography. Repeated recrystallization in hexane and reversed-phase Lipidex-5000 chroma- tography were then used to purify relatively large quantities of congeners _1_, Z, 9, Z, and 9. The lipophilic Sephadex derivatives such as LH-20 and Lipidex have been successfully employed, under normal or reversed-phase conditions, to separate various lipids and other hydrophobic substances (Nystrom and Sjovall, 1975). The purification of congeners _1_ and 9 by Lipidex-5000 (Figures 4b and 5b) marks the first application of this technique for separating structurally similar halogenated 105 aromatic hydrocarbons. Nearly complete recoveries were attained when the impure fractions were pooled and rechromatographed on the same column. Other characteristics include linear elution (near symmetrical or Gaussian elution profiles) with little or no tailing or fronting. The retention volume (VR) of the eluting congeners was insensitive to the initial purity, amount, or number of congeners in the sample (for example, compare V of congeners 9 and 9 in Figure R 4b to V of the same congeners in Figure 5b). However, not all congeners can be R separated by this technique since congeners 9 and 4, for example, have identical retention volumes (Figure 4b). Reversed-phase Lipidex chromatography was a particularly useful technique for resolving structurally similar brominated biphenyl congeners-that could not be separated by crystallization or alumina column chromatography. Congeners eluted in the order of increasing bromine content (molecular weight) and decreasing number of 99919 halogens. Similar trends were reported when chlorinated biphenyls were separated by carbon or carbon/foam chromatography (Huckins _e_t_ 99., 1980). Other practical aspects of Lipidex-5000 chromatography include the stability and inertness of the gel towards repeated use, reproducibility of elution profiles over a wide range of solute concentrations, preparative scale separations, low retention volumes, rapid flow rates, and simplicity. A number of factors have been considered of general importance for separations on the lipophilic Lipidex gel (Nystrom and Sjovall, 1975). Some of these factors include gel partitioning effects, adsorption effects, and gel filtration (exclusion) effects, all of which may be important for PBB separation as well. Numerous studies and reviews are available on the steric effects arising from introducing various types of substituents into one or more of the four 99919 positions in biphenyl (Eliel, 1962; Finar, 1975). These effects have been frequently studied by UV-absorption spectroscopy. The high intensity band of unsubstituted biphenyl, at approximately 250 nm, commonly termed the conjugation band 106 (k-band), is highly sensitive to M substitution. This band is progressively diminished (or is completely lost), and is shifted to a lower wavelength (hypsochromic shift) upon increasing the size and number of 9% substituents. In general, this reflects a progressive loss in molecular planarity and conjugation in the biphenyl system (Beaven, 1958). The UV-absorption characteristics of PBB congeners, particularly with regard to the conjugation band also indicate that the degree of planarity is mainly determined by the extent of M bromination. It is of particular interest to note that mono-gl_thg brominated PBB congeners (g, _6_ or Z) still display a distinctive k-band whose intensity is at least 25% of that of the same band in the o_rt_llg unsubstituted 3,4,5,3',4',5'-hexabromobiphenyl (Figure 10). Therefore, a fair degree of planarity and conjugation should be retained in these mono-M brominated congeners. The various chromatographic and spectral studies all seem to agree that the extent of planarity and conjugation of the biphenyl system is mainly determined by the degree of bromination g_r_thg_ to the biphenyl bridge. It is commonly believed that steric repulsion between the M substituents, and therefore the energy barrier to rotation around the biphenyl axis, is proportional to the size and number of 933mg substituents (Eliel, 1962; Finar, 1975). Of particular interest is that the mono-m brominated biphenlys (congeners ;, Q and Z) seemed to display somewhat intermediate chemical characteristics between the grit; unsubstituted 3,4,5,3',4',5'-hexabromobiphenyl and the di-g‘_thp_ brominated congeners exemplified by congener 3. While certain studies have indicated that biphenyl or certain M unsubstituted derivatives tend to be planar or nearly planar in their crystal form (Robertson, 1961; Charbonneau and Delugeard, 1976) other studies have shown that various 92112 unsubstituted derivatives of biphenyl have a dihedral angle of 42-540 between the two planes of the benzene rings (Bastiansen e_t_ gl_., 1971). It was, 107 therefore, of interest to study the crystal structure of an gr_th_o unsubstituted toxic PBB congener such as 3,4,5,3',4',5'-hexabromobiphenyl (H88) and compare it to that of a non toxic congener such as 2,4,5,2‘,4',5'-hexabromobiphenyl (congener 3). The purpose was to see if toxicity may be related to the dihedral angles. Since, as will be seen in Chapter 2, the mono M substituted congeners g, _6_, or Z seemed to have intermediate biological effects between H88 and congener 3 it would have been desirable to see if the dihedral angle of any of these congeners may be intermediate between those of H88 and congener 5. However, crystals suitable for x-ray crystallography studies could only be grown for H88 and congener 3. Even though various solvents and conditions were used, all efforts to obtain suitable crystals of any of the mono-M substituted congeners failed. It was therefore decided to study HBB, congener fl, and a tri-9_rt_h£ brominated compound, 2,5,2',6'- tetrabromobiphenyl (TBB), to see if a relationship exists between the dihedral angle and the degree of grth_o bromination. Results of these structural studies are not yet completely concluded; however, the dihedral angles for all three congeners have been obtained and are shown in Table 7. All three congeners appear to prefer a non-planar conformation in the solid state. However, while the di and tri-gyghg substituted congeners have a similar dihedral angle of 85-860, the M unsubstituted HBB has an angle of twist of only 50-520. This seems to be in excellent agreement with the values of 42-540 reported for a number of 9333 unsubstituted biphenyl derivatives that were studied by gas phase electron diffraction and x-ray crystallography (Bastiansen _e_t at, 1971). Therefore, in the solid state, and possibly in solution, planarity does not seem to be the most stable conformation of H88. This should not mean that HBB cannot assume a planar conformation since it is widely believed that the two phenyl rings of an M unsubstituted biphenyl have a relatively low energy barrier to rotation (Eliel, 1962; Finar, 1975). Therefore, the molecules of H88 are likely to assume various 108 $8.3 n 3 0%-? 2: NS.“ $3 Roi N momma: u E AmmIV o Jeanne} Am cmcmocoov o3 $.N 892 RE 20.2 s N -.m..¢..~.m.s.~ Aommfi: u 3 am: can 3N Sam 2%: an; s m seaflm 293. AmEoBv o n m ANV zoO boa cm 05.5% 952550 _maomca Aonc bison. 39 222085 :mO $30222 o .o_c:oo:oE mam mncsanoo mob: 9: ho 239:0 .Smb 55:35 93 boozoo 3 tom: 2m; c2362 2 mmmofio n ,3 fox £2 pcm 53830236 0 TmOcm£a_noEo.fimxchm..q..m.m.q.m mfizmumio 3 com: mm; mcmfimoS—zofiumJ BE; .mcmcmmcoo 025 ”.2: on... be #538 or: am? mcmxmc @353 95:38 c3233 E 232an3 502 am 550.5 203 233.6 mcmcmmcoo mmn. ho mofinum xcamamozmumfo xmmux H mumdfi. 109 conformations including completely planar, which seems to be required for binding to the TCDD-receptor and causing the typical TCDD-like toxicity (Poland e_t al., 1979). It should be pointed out that the fact that HBB is not absolutely planar should agree with the fact that it is at least two orders of magnitude less active than TCDD in binding to the cytosolic protein receptor(s) and inducing the AHH activity (Table 15; and Poland e_t_ a_l., 1979). While the dihedral angle of any of the mono-o_r_t_h_o_ brominated congeners could not be determined it may be expected that each will have an angle that is intermediate between that of H88 and congener fl. Such expectations are supported by the UV-absorption spectral characteristics and chromatographic properties that were discussed earlier. On the other hand, since a di- and a t‘ri-gflhg brominated congeners have virtually the same dihedral angle it may be concluded that only two gr_t_h_g bromines (each on a ring) will have sufficient bulkiness to almost completely restrict the free rotation of the phenyl rings around their major axis. The steric interaction created by two or more gm bromines forces the two phenyl rings to become nearly perpendicular to one another. Other biphenyls with two or more bulky Lthq substituents are also thought to have a dihedral angle in the vicinity of 90°, even though no mono-grim substituted biphenyl is known to be in a completely twisted conformation (Finar, 1975). Conformation can also be studied by measuring the optical activity and racemisation rates of biphenyl compounds containing different number and size of M'substituents (Eliel, 1962). Two conditions are required for a biphenyl compound to exhibit optical activity. First, neither ring should have a vertical plane of symmetry. Secondly, free rotation around the biphenyl axis should be restricted by the steric effects of sufficiently large gm; substituents. In biphenyl, itself, rotation around the biphenyl axis is considered facile where planarity is assumed to be the most favorable conformation. In planar form, biphenyl can be assumed to exist in a resonance hybrid with its conjugated form, 110 which imparts a partial double bond character across the two rings (Finar, 1975). Any substituted biphenyl (except those that have symmetry in one or both rings) can exhibit optical activity if, due to steric hindrance, the molecule assumes a twisted conformation. Conversely, when the energy barrier to rotation is not sufficiently high the enantiomeric forms of biphenyl can interconvert with one another to yield a racemic mixture that has no optical activity. Biphenyls containing four, three, or two sufficiently large substituents have been resolved into optically active enantiomers (Eliel, 1962). Among these was the di-flg substituted 2,2'-dibromobiphenyl-4, 4'-dicarboxylic acid. However, no mono-91th:; substituted biphenyl could be obtained in an optically active form. Even a biphenyl with the bulky As(CH3);I- w substituent was found to have a low energy barrier through which the enantiomeric forms pass as they readily interconvert (Eliel, 1962). As will be discussed later, the degree of 91mg bromination and, hence, the conformation of PBB congeners plays a crucial role in determining their biological effects. CHAPTER 2 LIVER MICROSOMAL ENZYME INDUCTION AND TOXICITY STUDIES WITH BROMINATED BIPHENYL CONGENERS: RELATIONSHIP BETWEEN TOXICITY AND ORTHO BROMINATION lll 112 ABSTRACT Seven PBB congeners comprising between 20 and 30% of two Firemaster mixtures (FF-1 and BP-6) were evaluated for some of their short-term, pharmacotoxicological effects. These are 2,4,5,2',5'-penta-, 2,4,5,3',4'pentafi 2,3,4,2',4',5'—hexa-, 2,4,5,3',4',5'-hexa-, 2,3,4,5,3',4'-hexa-, 2,3,4,5,2',3',4'-hepta-, and 2,3,4,5,2',3',4‘,5'-octa-bromobiphenyl (congeners 1, 2, _5_, _6_, Z, _9_, and _12, respectively). Each was administered (ip) at 90 mg/kg into rats one week before sacrifice, except for congener _2_ which was administered two weeks before sacrifice. Results of liver microsomal drug metabolizing enzymes induction, the histopathological and Ultrastructural studies, and changes in body weight gain and organ weights all agree that correlations do exist between the type of microsomal enzyme induction and the extent of toxicity. Congeners _2_, 2, _6_, and Z caused a variable degree of 3-methylcholanthrene (MO-type induction as evidenced by the shifts in the cytochrome P-450 difference spectra absorption maxima and the relative increase in the 455/430 nm ratios of ethyl isocyanide (EtNC) difference spectra. These congeners also induced benzo(a)pyrene hydroxylase and p-nitrophenol-UDP-glucuronyltransferase to extents that were generally comparable to those caused by MC. They also evoked one or more of the 2,3,7,8- tetrachlorodibenzo-p—dioxin (TCDD)-like toxicity responses including thymus atrophy, impairment of the humoral immune response, body weight loss, and hepatic Ultrastructural changes that included increased formation of cytoplasmic lipid droplets and extensive proliferation and disorganization of the rough endoplasmic reticulum. In addition, a phenobarbital (PB)-type induction, as eVIdenced by the increased aminopyrine-N-demethylase and NADPH-cytochrome P-450 reductase activities, was evoked by congeners _5_ and 6 and to a lesser extent by Congeners 2 and Z. Congeners 1, _9_, and _1_2, on the other hand, caused only 113 PB-type induction and were relatively non-toxic. Microsomal epoxide hydrolase activity was invariably induced by all congeners in a manner that seemed to suggest that this enzyme may not necessarily be coordinately expressed with the mixed-function oxidase activities that are normally induced by PB. Congeners 2 and 2 displaced 3H-TCDD binding by the TCDD-receptor at nearly 700-fold higher concentration than TCDD. Results of this assay and those of the MC-type microsomal enzyme induction seemed to agree that 3,4,5,3',4',5'-hexabromo- biphenyl (HBB) is at least 10- to 20-fold more potent than any of the mono-93g}; brominated congeners 2, 6 or Z or the di-ortho brominated congener 5. INTRODUCTION Polyhalogenated biphenyls have experienced a widespread industrial applica- tion mainly due to their heat resistance and chemical stability. As a result, these chemicals, particularly mixtures of polychlorinated biphenyls (PCB), have entered the environment and resulted in contaminating the populations of many industrial nations (Kimbrough, 1974). Mixtures of polybrominated biphenyls (PBB) have by far been less widely applied and commercialized; and therefore PBB are not known to be as ubiquitous in the environment as PCB. However, a mixture of PBB, commercially known as Firemaster, has inadvertently contaminated Michigan's environment between Fall of 1973 until early Spring of'1974 (Carter, 1976). Nearly 90% of Michigan's population are estimated to have measurable levels of PBB in their bodies. As was discussed in Chapter 1, each of twelve PBB congeners (2112), in addition to congeners 99 and 6_b, are present in Firemaster at a concentration of at least 0.5%. The structures of ten of these congeners totalling over 98% of the mixture are known, and the other PBB congeners are partially characterized. At 114 least twenty other minor chemicals including nearly 200 ppm of penta- and hexa- brominated naphthalenes have been detected in a mixture of Firemaster (Hass gt al., 1978). However, no traces of brominated-dibenzo—p-dioxins or dibenzofurans could be found at < 0.5 ppm (Hass gt a_l., 1978). Rats administered Firemaster have their hepatic microsomal drug metabolizing enzymes induced in a manner similar to that when both phenobarbital (PB) and 3-methylcholanthrene (MC) are coadministered, a phenomenon commonly termed mixed-type induction (Troisi, 1975; Dent g a_l., 1976a,b; McCormack e_t_al., 1977; Moore at a_l., 1978b, 1979). Firemaster can also cause a number of toxic effects that are typical of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), except the dose required to cause the TCDD-like toxicity is relatively much higher. In addition to liver enlargement and porphyria, neoplastic nodules were seen in rats administered high doses of Firemaster (Strik, 1978; Kimbrough e_t_ a_l., 1978, 1981). The PBB mixture also suppresses the immune response and causes thymus atrophy in mice and rats (Fraker, 1980; Gupta and Moore, 1979). Edema in chicks and reduced egg hatchability have also been reported (Polin £11., 1979). The polyhalogenated biphenyls, particularly the polychlorinated biphenyls (PCB), are generally classified as to the type of microsomal enzyme induction they produce because of the correlation between MC-type induction and toxicity (Goldstein e_t a” 1977; Poland and Glover, 1977; Goldstein, 1979). Congeners which give rise to an MC-like induction of microsomal enzymes also evoke a toxic response similar to that usually caused by TCDD, including body weight loss, involution of the thymus, porphyria, and edema (Goldstein et _a_l_., 1977; Goldstein, 1979; Poland and Glover, 1977; Poland e_t a_l., 1979). Congeners of this group also bind to a liver cytosolic protein commonly referred to as the "TCDD-receptor", which is believed to be involved in the mechanism of enzyme induction and perhaps toxicity (Poland and Glover, 1977; Poland gt a_l., 1979; Greenlee and Poland, 1979). 115 There appears to be a very good correlation between the binding to the TCDD- receptor, the induction of enzymes usually induced by MC, and the TCDD-like toxic response. There are at least two reasons for characterizing the biological effects of the individual congeners in Firemaster. First, most human exposure to PBB occurred indirectly through contaminated meat and dairy products; therefore, it is possible that one or more of the toxic PBB components may have been concentrated in the consumed food products. In Firemaster-exposed rats the composition of extracted PBB from liver and milk was considerably different than that of the original mixture (Dannan _e_t a_l., 1978b). Therefore, studies with purified congeners will help in predicting the toxicity and pharmacological effects related to human or environmental contaminations with PBB. Secondly, studies with a series of congeners will develop the structure-activity relationships for toxicity and the type of microsomal enzyme induction. The two major congeners of Firemaster, namely 2,4,5,2',4',5'-hexa- and 2,3,4,5,2',4',5'—heptabromobiphenyl (congeners 3 and 2, respectively), are strict PB-type inducers and are not believed to cause any TCDD- like toxicity (Moore g a_l., 1978b, 1979; Render, 1980). These two congeners constitute nearly 65-75% of the two mixtures Firemaster BP-6 and Firemaster FF-l (Figure 1, and Table 4). Research in this Chapter was intended to characterize the pharmacotoxicological effects of seven other congeners totalling nearly 20-30% of either of the two Firemaster mixtures. These congeners are 2,4,5,2',5'-penta-, 2,4,5,3',4'-penta-, 2,3,4,2',4',5'-hexa—, 2,4,5,3',4',5'-hexa-, 2,3,4,5,3',4'-hexa-, 2,3,4,5,2',3',4'-hepta-, and 2,3,4,5,2',3',4',5'-octa-bromobiphenyl which, according to their order of elution from the GC, will be referred to as congeners 2, 2, _5_, 6, Z, 2 and _12, respectively. 116 MATERIALS AND METHODS Materials Benzo(a)pyrene, polyethylene glycol (PEG, molecular weight approximately 400), butylated hydroxytoluene (BHT), sodium dodecyl sulfate (SDS), bovine serum albumin (BSA, fraction V), horse heart cytochrome c (Type VI), highly purified pig heart isocitric acid dehydrogenase (Type IV), trisodium DL-isocitrate (Type I), NADP”, tetrasodium NADPH (Type III), disodium NADH (Grade 111), glycine, ammonium UDP-glucuronate, Tween 80, MC, and PB were all purchased from Sigma Chemical Co., St. Louis, Missouri. Aminopyrine, and Coomassie Brilliant Blue G-250 (xylene brilliant cyanin G) were obtained from K and K Rare and Fine Chemicals, Plainview, New York. Acrylamide, N,N'-methylenebisacrylamide, N,N,N',N'-tetra-methylenebisacrylamide (TEMED), and ammonium persulfate were purchased from Canalco, Rockville, Maryland. Agar was obtained from Difco Laboratories, Detroit, Michigan, and pyronin B was from Hartman-Leddon Co., Philadelphia, Pennsylvania. Styrene oxide, and p-nitrophenol were purchased from Aldrich Chemical Co., Inc., Milwaukee, Wisconsin. [7 3H] -Styrene oxide (specific activity = 29.2 mCi/mmole) was purchased from New England Nuclear, Boston, Massachusetts. Dimethyl-POPOP (1,4-bis[2-(4-methyl-S-phenyloxazolyl)-J- benzene), PPO (2,5-diphenyloxazole), and Triton X-100 were all 03 scintillation grade, and were purchased from Research Products International Corp., Elk Grove Village, Illinois. Sheep red blood cells (SRBC) were purchased from GIBCO Diagnostic, Madison, Wisconsin, while minimum essential culture medium (MEM) and rabbit anti-rat IgG were purchased from Grand Island Biologicals, Grand Island, New York. A sample of Firemaster FF-l (lot no. FH7042), that allegedly was involved in contaminating Michigan, was a gift from Farm Bureau Services, Lansing, Michigan. 117 Preparation of P88 Congeners Congeners _1_, 2, _5_, 6, Z, and 2 were purified from Firemaster BP-6 (lot no. 6224-A), which was manufactured by Michigan Chemical Corp., St. Louis, Michigan. The details of the purification methods which involved repeated recrystallization and chromatography on both Alumina and Lipidex-5000 were presented in Chapter 1. The first four congeners had a purity of 99% or better, while congeners Z and 2 had a 98% and 97% purity, respectively. However, all impurities can be accounted for by one or more of the PBB congeners. Congener 12, which was purified by Robert A. Moore by chromatography on a series of Alumina colums and by repeated recrystallization, was nearly 98% pure. Figure 12 shows the GC elution profile of all purified congeners including some others (2, fl, and 2) that were not used in these studies. Congeners 2, _5_, 6, and 2 were also analyzed by GC/MS (Hewlett-Packard 5840A/5985) selective ion monitoring (SIM) for. tri-through octa-brominated dibenzodioxins and dibenzofurans. At nearly 1/1000 detection limit (2 ug PBB injected) none of these ions of concern could be identified. Crude 3,4,5,3',4',5'-hexabromobiphenyl (HBB) was purchased from Ultra Scientific, Inc. (formerly RFR Corp.), Hope, Rhode Island. It was purified by repeated alumina chromatography in hexane to a purity of over 98% and its structure confirmed by NMR. Animals Outbred male Sprague-Dawley rats were purchased from Spartan Research Animals, Haslett, Michigan, and were allowed an acclimation period of two days prior to any treatment. Unless indicated otherwise, all PBB treatments were administered as a single ip injection at 90 mg/kg body weight one week before killing. The research was accomplished on stages where each experiment had its own set of control rats. In the first experiment congeners 6 and _12 were 118 FIGURE 12. GAS CHROMATOGRAPHIC ELUTION PROFILES OF ALL PURIFIED PBB CONGENERS 119 Br Br \‘J/PH Br Br Br Br Br .. © (9 er /8( 3" Pk? \11“ Br Br Br ., © © er /Bf Bf \ PkB \ .A‘2_ -2“ _~__.____ """"" _.._ Br Br Br Br 8' / Br Br \ Pk9 N Br I f\/ FIGURE 12 120 administered into rats weighing between 150 and 180 g. Control rats received the same volume of vehicle (4 ml PEG/kg body weight). In the same experiment, four rats per treatment were also administered Firemaster FF-l, PB, or MC. PB was administered in five daily ip injections at 50 mg/kg in 1 ml water, and rats were killed on the sixth day. MC was administered (ip) at 20 mg/kg in 2 ml corn oil 36 and 24 hr before killing. In the second experiment, six rats weighing between 150 and 180 g were administered a single ip injection of congener 2 or Firemaster FF-l at 90 mg/kg in 4 or 3 ml PEG, respectively (control rats were administered PEG at 3 ml/kg), and all rats were killed two weeks later. Rats (five per treatment) were also treated with PB, MC, or the combination of these two chemicals at the same dose and schedule used for each as described above. In another part of this study, twelve rats were divided into four groups. The first and second groups received an ip injection of 2 or 30 mg 3,4,5,3',4',5'-hexabromobiphenyl/kg body weight, while the last group was injected with the vehicle (PEG) only. All rats in these four groups were killed one week after treatment. In the third experiment four rats weighing between 100 and 125 g were administered a single ip injection of congener l, 2, Z, or 2 one week before sacrifice. PEG was also injected into four control rats at 3 ml/kg. All rats were housed in stainless steel wire cages (two rats per cage) and were given free access to feed and water except for the night before sacrifice when feed was removed. Pathological and Ultrastructural Evaluations At necropsy weights of the liver, spleen and thymus were recorded for each rat. Samples from these tissues and from kidney, thyroid, trachea, lung, heart, adrenal, pancreas, and small intestine were fixed in 10% buffered formalin, embedded in paraffin and sectioned at 5 pm. All tissues were stained with ~~‘.‘-.._-.4— 121 hematoxylin and eosin (H and E), and the liver sections were also stained with oil red O. Prior to examination by electron microscopy the formalin-fixed liver tissues were postfixed in 1% osmium and stained with uranyl acetate and lead citrate. Clinical Evaluations Rats administered congener 2 or Firemaster FF—l were examined for possible hematologic or other clinical changes. After these rats and their respective control were killed by C02, blood samples were drawn into evacuated glass tubes containing EDTA as an anti—coagulant. Hemoglobin concentration, packed cell volume, total erythrocyte and leukocyte counts were determined on each sample. A specimen of bone marrow from the femur of each rat was smeared onto a glass slide and stained with Wright's stain for evaluation. Jerne Hemolytic Plague Assay Following treatment by congener 2 or Firemaster FF-l, the immune response to SRBC was evaluated by using a previously described modification of the Jerne plaque assay (Fraker e_t a_l., 1977; Luecke e_t_ a_l., 1978). Briefly, nine days after treatment rats were injected ip with 5 x 108 SRBC in 0.4 ml of phosphate buffered saline and were killed on Day 14. The spleens were removed, minced and passed through a 100 mesh stainless steel screen. After washing, the lymphocyte suspension was brought to a final volume of 1 ml in sterile MEM containing Earles balanced salts supplemented with streptomycin and penicillin. A 0.1 ml aliquot of spleen cells at the proper dilution was mixed with 2 x 108 SRBC and added to 1 ml of MEM containing 0.6% agarose. This mixture was overlaid on a 60 mm petri dish containing a 5 ml MEM/agarose base. Following a 1.5 hr incubation at 370C, the IgM-producing anti-SRBC plasmacytes (direct plaques) were visualized by the 122 addition of non-hemolytic guinea pig complement (diluted 1:100). Indirect or IgG- producing plasmacytes were identified by adding rabbit anti-rat IgG (diluted 1:40) prior to the addition of complement. All samples were assayed in duplicate with each rat spleen being processed individually. Indirect plaques were corrected for the small number of direct plaques which developed on the indirect plates. All data are expressed as the average number of plaque forming cells (PFC) produced per rat spleen. Preparation of Microsomes Liver microsomes were isolated from each individual rat according to previously ‘described procedures (Pederson and Aust, 1970). Microsomal pellets were then resuspended in 0.3 M sucrose buffer containing 0.1 M tetrasodium pyrophosphate-HCI, pH 7.5, to remove ribosomes and adsorbed proteins. Microsomes were repelleted by ultracentrifugation at 105,000 x g for 90 min, then resuspended in a 0.05 M Tris-HCl buffer (pH 7.5 at 25°C) containing 50% glycerol and 0.01% (w/v) butylated hydroxytoluene, and stored at -200C under argon until USS. Enzyme Assays Protein concentrations were determined by the method of Lowry _e_t 31. (1951), using BSA as the standard according to Rutter (1967). Cytochrome P-450 was assayed by its reduced carbon monoxide saturated difference spectrum in 10% glycerol (Omura and Sato, 1964) using an extinction coefficient of 91 mM"l-cm'l between Amax and A500. NADPH-cytochrome P-450 reductase was assayed by the rate of cytochrome c reduction at 550 nm using an extinction coefficient of 21 mM'locm-l. 123 Aminopyrine-N-demethylation was assayed as described by Pederson and Aust (1970) and later modified by Moore at al. (1978b). The amount of formaldehyde formed was measured according to the method of Nash (1953). Briefly, the reaction mixture (5 ml total volume) contained 50 mM Tris-HCI, pH 7.5, 5 mM 0.5 mM NADP+, 4 mM sodium D, L-isocitrate, 0.44 units of MgCl 5 uM MnCl 2’ 2’ isocitrate dehydrogenase, 20 mM aminopyrine (recrystallized twice from hexane before use), and 2-3 mg microsomal protein. The complete reaction mixture was prepared on ice in a 20 ml beaker, where aminopyrine was added just prior to transferring the beaker to a 37°C Dubnoff metabolic shaker to start the enzymatic reaction. Formaldehyde production was measured in 1.0 ml aliquots taken from the reaction mixture at 1,4,7, and 10 min and mixed with 1.0 ml of 10% trichloroacetic acid to stop the reaction. After allowing few minutes for the protein to precipitate, 2.0 ml of Nash reagent (2 M ammonium acetate, 0.05 M acetic acid, and 0.02 M 2,4-pentanedione) was added to each tube and the mixtures were heated at 60°C for 10 min to develop the color. After cooling the samples were centrifuged for about 10 min, and the absorbance of the supernatant was measured at 412 nm against a blank consisting of 1.0 ml 10% trichloroacetic acid, 1.0 ml buffer, and 2.0 ml Nash reagent. An extinction coefficient of 7.08 mMnlocm"l and a dilution factor of 4 were used to calculate the formaldehyde content. Benzo(a)pyrene hydroxylation was assayed fluorometrically according to the method of Gielen _e_t _a_l_. (1972). Details of the procedure including some technical advice were also described by Moore (1978). Briefly, a 1 ml volume reaction mixture (in 10 ml Erlenmeyer flask) contained 50 mM potassium phosphate, pH 7.2, 0.36 mM NADPH, 0.39 mM NADH, 3 mM MgC12, 0.08 mM benzo(a)pyrene in methanol, 0.6 mg BSA, and nearly 250 ug of microsomal protein. Benzo(a)pyrene was added last (within 1 min before the incubation) and the reaction mixture was mixed well before being incubated for 10 min in a Dubnoff shaker at 37°C. 124 Reactions were terminated by adding 1 ml of ice cold acetone, and flasks placed on ice to facilitate protein precipitation. After adding 3.25 ml of hexane, flasks were incubated at 37°C for an additional 10 min to extract the metabolites, then 1 ml of the organic phase was transferred to a culture tube. Just before measuring the fluorescence, the hexane extract was vigorously mixed with 3 ml of 1 N sodium hydroxide for approximately 1 min, then centrifuged in a table top centrifuge for nearly 2 min. The organic phase was aspirated off, and the fluorescence of the aqueous phase was measured in an Aminco-Bowman spectrofluorometer using an excitation wavelength of 396 nm and an emission wavelength of 520 nm. The fluorescence of quinine sulfate was used to calibrate the amount of metabolite as described 'by Rickert and Fouts (1970), where 0.036 nmole of 3-hydroxy- benzo(a)pyrene/ml of 1 N NaOH emits a fluorescence equivalent to that due to 0.3 ug quinine sulfate in 1 ml of 0.1 N H2304. UDP-glucuronyltransferase was measured using p-nitrophenol as a substrate according to the method of Lucier e_t a_l. (1977). Reaction mixture (14 ml volume) contained 150 mM Tris-HCl (pH 7.4 at 37°C), 10 mM MgClZ, 0.8 mM p-nitrophenol, 1.4 mM UDP-glucuronic acid, 0.2 ul of Triton X-100 per mg of microsomal protein (0.3-0.6 mg). The microsomes were mixed with the detergent and a small volume of Tris buffer, then preincubated at 370C for 5 min before adding 0.9 ml solution containing p-nitrophenol, Tris, and MgCl Finally, UDP-glucuronic acid was added 2. to initiate the reaction which was allowed to continue for 15 min at 37°C. At the end of the incubation period, 5 ml of 0.2 M sodium glycine (pH 10.4) was added, and the absorbance was measured at 403 nm using cuvettes of 0.5 cm path length. Blanks contained microsomes and all the other components except that Tris buffer was substituted for UDP-glucuronate. Rates were calculated using an extinction coefficient of 18.0 uM‘l.cm-l. 125 Epoxide hydrolase activity was measured radiometrically with styrene oxide as the substrate according to the procedure of Oesch gt a_l. (1971). In 18 x 150 mm test tubes, mixtures containing 0.1 M Tris-HCl (pH 9.0 at 37°C), 0.02% (w/w) Tween 80, and approximately 0.5 mg of microsomal protein were preincubated at 37°C for 20-25 sec. Reactions were initiated by adding 8 mM 3H-styrene oxide (approximately 20,000-30,000 cpm/assay) in acetonitrile (final acetonitrile concentration of 4%), and after 20 min were terminated by adding 10 ml of petroleum ether. After vortexing, the samples were allowed to stand for few minutes until the phases separated, then the tubes were placed in a dry ice-acetone bath for approximately 5 min. The organic phase was decanted and a second petroleum 'ether extraction was repeated. The aqueous phase was then extracted (vortexing) with 2 ml ethyl acetate, and after phase separation, 0.5 ml of the ethyl acetate extract was transferred to a scintillation vial. Radioactivity was determined with a Packard Model 3310 Tri-Carb liquid scintillation spectrometer after adding 15 ml of scintillation fluid containing 2.5 g PPO and 0.1 g dimethyl POPOP dissolved in 1 L of toluene: Triton X-100 (2:1 v/v). Blanks (with no microsomes) were also incubated at 370C for 20 min, before being extracted and analyzed. Ethyl isocyanide (EtNC) difference spectra were recorded using 1 mg microsomal protein/ml in 0.2 M potassium phosphate buffer (pH 7.45) containing 30% (v/v) glycerol. A few crystals of sodium dithionite were added to both the sample and reference cuvettes, then EtNC (final concentration of 3.3 mM) was added to the sample cuvette. Both cuvettes were capped and mixed well before difference spectra were recorded between 500 and 400 nm. (Caution: EtNC is a volatile chemical with an extremely vile odor). The ratio of the absorbance peaks at approximately 455 and 430 nm was measured as A455-500. A430-500 126 Preparation of Ethyl Isocyanide Since ethyl isocyanide has an extremely vile odor, and since it has been known to explode, all steps of synthesis and purification were carried out in a well- ventilated hood behind a shield of safety glass. It was synthesized by the method of Jackson and McKusick (1963) using one mole of each of ethyl iodide and silver cyanide. These two reagents were mixed in a 2L, 3 necked, round bottom flask fitted with a refluxing condenser and a sealed stirrer, and heated on a steam bath for about 2 hr until a viscous, homogeneous, brown liquid was formed. The flask was removed from the steam bath and 100 ml of water was added through the condenser followed by 2.75 moles of potassium cyanide in 85 ml of water. The mixture was stirred for 10 min, then the apparatus was arranged for simple distillation. Distillation was continued until the temperature was 115-1200C. In a separatory funnel sodium chloride (2.5 g) was added to the distillate, and the aqueous layer removed and discarded. Two portions of ice-cold, sodium chloride saturated water were used to wash the crude product of EtNC which was dried overnight over magnesium sulfate. The dry product was distilled through a 15 inch column packed with glass beads. The distillate coming off at 77-79% (second fraction) was collected. Nearly 25 ml of clear ethyl isocyanide was recovered and was stored in a tightly sealed round bottom flask at -20°C. SDS-Polyacrylamide Gel Electrophoresis SDS-polyacrylamide gel electrophoresis was carried out according to O'Farrell (1975) with the following exceptions. The stock acrylamide solution was 22.2% acrylamide and 0.6% N,N'-methylene-bis-acrylamide (bis). The stacking gel Was 3.3% acrylamide and 0.09% bis, and the separating (running) gel was 7.46% Berylamide and 0.2% bis. Buffers for running gel, stacking gel, samples and 127 electrophoresis were identical to those described by O'Farrell except that a lithium salt of dodecyl sulfate was used instead of the sodium dodecyl sulfate. A vertical slab gel apparatus similar to that used by Studier (1973) was employed. Concentrated sulfuric acid was used to clean the glass plates, while the plexiglass spacers were cleaned with detergent. The pieces were assembled using pinch-type paper clamps, then sealed with hot 1.5% agar. The running gel was poured into the assembled mold to a height of 10 cm and overlaid with 2 ml water. After an hour of polymerization time, water was removed and the stacking gel was poured and allowed to stand an additional hour for polymerization. Electrophoresis was started at 20 mA until the dye entered the running gel when the current was turned up to 25 mA. When the tracking dye approached the bottom, the gel was removed and fixed in 10% trichloroacetic acid for 1 hr. The gel was stained in a Brilliant Blue R solution (prepared by dissolving 1.25 g of dye in 454 ml of 50% methanol plus 46 ml of glacial acetic acid) for an overnight, then destained in 25% methanol-10% acetic acid, and stored in 10% acetic acid. Purification of Epoxide Hydrolase Liver microsomal epoxide hydrolase was purified according to the procedure of Guengerich and Martin (1980). Male Sprague-Dawley rats (125-150 g) were administered 0.1% of PB in drinking water for 10 days before killing. Liver microsomes were isolated in 0.1 M Tris-HCI (pH 7.4) buffer containing 0.1 M KCl, 1.0 mM EDTA, and 0.23 11M BHT, and were washed once in 0.1 M potassium pyrophosphate (pH 7.4) containing 1.0 mM EDTA and 0.23 pM BHT. Cholate solubilization of microsomes and column chromatography on n-octylamine- Sepharose 4B, DEAE-cellulose, CM-cellulose, and hydroxylapatite were essentially as described (Guengerich and Martin, 1980). Fractions were assayed according to the method of Oesch e_t a_l. (1971) as described above except that Tween 80 was 128 omitted since detergents were present in the purification buffers. Fraction II (Guengerich and Martin, 1980) contained most of the epoxide hydrolase activity and had a specific activity of nearly 350 nmoles-mg protein'l-min'l. Several fractions of highly purified epoxide hydrolase have been reported to have a specific activity of 465-885 nmoles-mg protein-1cmin.l (Guengerich and Martin, 1980). SDS-polyacrylamide gel electrophoresis showed this preparation to contain a major band of nearly 48,000 molecular weight and a minor one with a larger molecular weight (approximately 80,000). RESULTS Results of this Chapter will be presented in sections since the studies were performed in stages as the purified PBB congeners became available. In each section, the pharmacotoxicological effects of one or few congeners in rats were compared to the same parameters in control rats, and in certain sections the effects of Firemaster or the prototype inducers, PB, MC or both of PB and MC were also evaluated for reference. It should be mentioned that even though these studies were all performed with male Sprague-Dawley rats, seasonal and age variations were inevitable and may have affected certain forms of cytochrome P-450 and associated activities (Gielen e_t_a_l., 1972; Parke, 1975). Effects of 2,4,5,3',4',5'-Hexabromobiphenyl (Congener 6) and 2,3,4,5,2',3',4',5'- Octabromobighenyl (Congener 1_2) Seven days after the administration of congeners 6 and g most of the hepatic drug metabolizing parameters were substantially altered (Table 8). Both treatments increased the liver weight to body weight ratios to nearly 150% of Control. Liver weight to body weight ratios were also substantially increased by PB 129 .Om H 558 mm ummwmaxm 5.5 550 62:53 5.5.55 .3 SN cam. on 5 5223 EC cmamuflEEvm 53 3x38 oNv DZ :65 5x5 9.: co 55:3 5.53 32 5.5 9.3:. on .5 52535 3 >255 m>C E nmamuflEEcm mm; mm .552 $553 025 55:3. 95 Tum 555852.; 5 .Mia. emcmmcoo .w aocmmcoo 5 335 am .5 20.558 minim. _E a Co c2535 2 532m 5 505525855 5.52, 83... .chEumm: .5ch .5 Mia. .5 .0. 25:50:00 nocmuflEEUd. 3mm “.0 mamumEmamQ GEE—05322 GEO _mEOmSBE 53'. mEom .m mumx; 130 Qood v 8 55.53% 25:55:95 5.5 53m> 525 :5 ”god A 3 58h. 955533 9.6: 5.550 on. 52:5 3525355 a 0 £555 mE.EE\mm_oEc a 5555 08529:: w 813 3 n 5 n2 H 95 c: H 5 2 H as dz H S c: H EN 25 H GA No H E o: 83 3 u 5 cm; H can 5 H NS so H 5 so H 92 an H a3 35 H SH so H an me $533 3 n 5 75 3 H soc 2 H 5: no H 3: .1 H N: 36 H :4 93 H 3m 36 H man 3 H ca 288827.. 8%: R u 5 one H a: 2 H n2 no H 3 a; H n2 :3 H a; NH H see Rd H RH so H 3 NIH. 5:880 8.33 3 u 5 3: H 98 3 H 0.2 an H 5.2 no H 3; 86 H as; 2 H RM 85 H Bk so H 5 m $898 83 3 .i. 5 e: H NR 3 H ex Nd H 3 3 H em 55 H Rd R H EN 86 H 84 no H .3 6.580 55— 55535.— D a me 555.85: 5552 $5.5»: 55155955 A055,. :5 Dwain. ACE . .3 Ac\cv 59:55.; 4.30.3033 -05.»: 5553 2.5553 ©3233 mEoEooiO momain. is 25m iQOD mExoam AwVONcmm ioEE< UZum iIQO‘max to 448.5 nm. The 455/430 nm absorbance ratio in the EtNC difference spectra was increased almost 4-fold. Such an increase was slightly higher than that caused by MC which also shifted the A max of the CO-difference spectra to 448.5 nm. Aryl hydrocarbon hydroxylase (AHH) activity, measured as benzo(a)pyrene hydroxyla— tion, was induced about 11-fold by both congener 2 and MC. This increase in AHH activity was significantly higher than that caused by Firemaster FF-l or the 133 FIGURE 13. EFFECTS OF CONGENER 6 ON RAT LIVER ULTRASTRUCTURE (A) Portion of a hepatocyte from a control rat administered (ip) 3 ml PEG/kg body weight a week earlier. (B) Portion of a hepatocyte from a rat administered (ip) 90 mg/kg of congener 6 a week earlier. Notice the formation of concentric arrays of ER membranes around lipid droplets and mitochondria. (Uranyl acetate-lead citrate stain, 15, 150X). 135 .05 H cmmE mm ummmmhaxm 2m memo .chm :35 comm coup cum: m_sumcom ncm omen mem 9: am m_mo_Emco mmmfi cmZm 953 0.2 ncm ma fion 5:2, @385 33m .8583 2an .E 3 Eu mm 33E om am @225 Es nmgmuflEEom mm; 02 .xmc 5me 9: co 8:3 2m; 32 new 3995 €02 9.3.: on ”.0 953035 E xzmv m>E E UmhmuflEEnm mm; mm. £32 mxmmz, 03... vax ncm TIE hmummEmeC go .N. gmcmmcoo be 9.3:. cm .5 :ohcoov ov<0ma _E M US co_uom_c_ & 2me m 3232:..Evm 2m; 3mm .ScmEumm; .550 .8 M emcmmcoo umgmuflEEUd. 3mm U8 EmumEmEn. mENzonmum—z 9:0 85.8222 .621. mEom .m Him}: 136 20.0 v 3 223:5 25cmoEc9m 2m 83:; .550 :m “God A 3 ammuum mLchBm 9.5: 35:00 3 227:5 0 530.5 9: .c_E\mm_oEc a 530.5 958.08: m 3.33 R u 5 as H nfi 3 H 98 no H 2 3 H #2 Ed H m: 2 H Re Ed H 3H 02 + mm. 3.33 a u 5 g H 98 as H no 3 H 3: one H 3: deH 34 NH H SH 036 H 34 o: 8.81: a u 5 and H new m: H QR 3 H 3 3 H MEN 85 H Rd 3. H 5 Ed H 8H mm. 3qu G n 5 Th: 1} H NB 3 H nan E H 3 m: H EN 26 H 84 R H :3 36 H SH 1885:; 3.33 G n 5 24 H no: as H QR so H 3: 3 H mi and H SH. 2 H an 86 H 3N M 5898 8.23 G n 5 in H «.2 A: H 2: do H o; E H 3: Ed H and R H «8 25 H a: 6526 www— meuODUmh D D me nmmmgmhmcmb comm. £68.63 now—3.338% A052 E: Dwain. AEc . .3 Emcbmm; 420.5030 no.6»; ocmta 2-9.2.8 ©9133 @5550in momqua do: .2265 -AmVONcmm 65:3 026 -IanZz 05.26in mucmEumo: .850 .5 IN. gmcmocoo 52335EU< 3mm u_o mgmumEmgwa GEE—03322 020 38895:). .524 mEom m mind.» 137 combined administration of PB+MC. Even though Firemaster FF-l and the combination of PB +MC caused nearly 2.5-fold increase in cytochrome P-450 content, these treatments caused a blue shift of only 0.5 nm. In addition, microsomes from rats pretreated with Firemaster or with PB + MC had only a 2- fold increase in their EtNC difference spectra ASS/430 nm ratio. PB alone caused the smallest increase in this ratio and it failed to shift the absorption maximum of the hemoproteins, even though it increased AHH activity by nearly S-fold. E—Nitrophenol-UDP-glucuronyltransferase activity was increased nearly three times by congener _2_. This activity was increased by 70-90% after Firemaster F'F-l, PB + MC, or MC treatment, while PB alone had no significant effect. The ‘N-demethylation of aminopyrine was slightly increased by congener ; (136% of control), doubled by Firemaster FF-l, PB, or PB + MC, but was unchanged by MC alone. NADPH-cytochrome P450 reductase activity was increased 36% by congener _2_ and 60-70% by Firemaster FF-l, PB, or PB + MC, while MC alone had no effect. Epoxide hydrolase activity increased to 270% of control after administering congener _2_. In agreement with previous results (Table 8), Firemaster was the most effective inducer of this activity (about 350% of control) and PB was the next most effective (300% induction). The combination of PB + MC elevated this enzyme to twice its control level, while this activity was not affected by MC (Table 9). There was no differences in hematologic parameters between control and treated rats (Table 10). Total serum protein values were similar in all groups. Examination of the bone marrow did not reveal any changes in cellularity. No alteration could be detected in the myeloid, erythroid, or lymphoid components. Liver weight to body weight ratios were increased in rats given Firemaster FF-l and congener g. Spleen weight to body weight ratio and thymus weight to body weight ratio were slightly but significantly decreased in rats given congener g. 138 God v e 683. 88.2% .3 .228 52: 220:6 2:52:35 m 6.3qu 33.3 SBHRK 3112.2 «NodHono mBBHmNd moaoHBa mmmHHm: mamcmocoo m6 + 55 .I. .I. .I. .I. .I. .I. .I. .I. .I. 7:: mH+Rm 0:..qu mo+m3 NNo+mHR com+mw ma moo+mmo Bo+mmo mNmo+mwm omo+amm 5556521.. 26H? 33.? 3H5: 563:: R.mHmN.D .5st3 BdHano NdoHRa SBHR.“ 8:80 233 36V 233 $88 AmEE 555m 5382: 52:. A9 5955.; 8:8 9:29, Ease $2 x 055 #2 x 0523 a; zoom co as $25.6 8:55 :50 655.1 5:09.525 5:09:51. .52, >502 55C. umxoma 55% 55C. mom5>< .05 H G n 5 55E 55 5.5 523 :< .5053 505 miOwn. GE m 35 52502 65.. 5550 .552 5553 525 5:3 55 5953 255 ox 5n. 7...“. 55585:“. 5 M 5:50:00 0E om 35 555552555 553 35m THE 555F527; 5 .N. 55950 555£EEU< 35m E9: m55Em5d 52:20 55 56565625 5Eom OH mjmdc. 139 Also these rats gained less weight per day than their corresponding control (Table 10). Rats given Firemaster FF-l or congener g had swollen hepatocytes many of which contained small vacuoles (Figure 14). Although all portions of the liver lobules were affected, it appeared that the peripheral areas were more severely affected in rats given congener _2_. Results of the red O stain indicated only a slight increase in fat in hepatocytes of rats given Firemaster FF-l, but fat accumulation was more apparent in hepatic cells of rats given congener g. No treatment related histological changes could be seen in the remaining tissues. Striking Ultrastructural changes were also noticed in electron micrographs of hepatocytes of rats administered Firemaster FF-l or congener _2_. Firemaster FF-l caused a marked proliferation and dilatation of the SER. Congener ; caused similar but less dramatic changes (Figure 15). Mitochondria from Firemaster FF-l or congener _2_-treated hepatocytes were apparently normal. Lipid droplets of different size were common features to both treatments. Some of these lipid droplets were occasionally surrounded by concentric arrays of laminated ER membranes (Figure 15). Two weeks after exposure to either Firemaster FF-l or congener g, the antibody mediated response to a primary immunization with SRBC was substantially reduced (Table 11). Compared to controls, those rats exposed to the P88 mixture produced 43% as many direct and 52% as many indirect anti-SRBC plasmacytes per spleen. Rats exposed to congener _2_ gave even poorer responses producing only 27% as many IgM and 31% as many IgG PFC per spleen as normal rats. The equivalent reduction in the percentage of both direct and indirect responding plasmacytes indicates that there was no preferential impairment of either 19M or IgG producing lymphocytes. Another experiment was carried out to compare congener g to 3,4,5,3',4',5'- hexabromobiphenyl (H88), 3 previously reported strictly MC-like inducer (Poland 140 FIGURE 14. EFFECTS OF CONGENER _2_ AND FIREMASTER FF-l ON RAT LIVER STRUCTURE (A) (B) (C) Liver section of a control rat two weeks after an ip injection of 3 ml PEG/kg body weight. Notice normal appearance of liver cells and sinusoids. Liver section of a rat two weeks after congener ; treatment (90 mg/kg). Hepatocytes are swollen, somewhat vacuolated and have a slightly individualized appearance. Liver section of a rat liver two weeks after treatment with Firemaster FF-l (90 mg/kg). Notice swelling of hepatocytes and mild vacuolation. (H and E stain, 160x.) 141 -;. , , A ,‘u (> “-7 ‘ 5i" ‘ .4; ". .. ‘ y... '..' er, ?: _..::.e°:-. 4:35; a gilmvcfiubz‘b. “A FIGURE 14C 143 FIGURE 15. EFFECTS OF CONGENER 2 AND FIREMASTER FF-l ON RAT LIVER ULTRASTRUCTURE (A) Hepatocyte of a rat two weeks after treatment (ip) with Firemaster FF-l (90 mg/kg). (8) Hepatocyte of a rat two weeks after treatment (ip) with congener g (90 mg/kg). Notice in both (A) and (B) the proliferation of SER and extensive formation of lipid droplets which are surrounded by concentric arrays of laminated ER membranes (myelin figures). (Uranyl acetate-lead citrate stain, 15, 150x). 144 145 TABLE 11 Antibody Mediated Response of Rats to Sheep Red Blood Cells Two Weeks after Exposure to Firemaster FF-l or Congener 2 Six rats per group were administered (ip) 3 ml PEG/kg (control) or 90 mg/kg of Firemaster FF-l or Congener 2. Direct and Indirect Plaques were determined five days after primary immunization with SRBC. All values are the mean 1'. SEM. Direct Indirect Plaques (IgM) Plaques (IgG) Treatment (PFC/Spleen) PFC/Spleen) Control 158,000 1 5,850 176,000 : 5,950 Firemaster FF-l 68,100 1 7,3308 92,500 : 13,320a Congener ; 42,600 : 5,6403’b 55,100 : 6,1803’b Significantly different than mean of control by Student's t-test (p< 0.001) Significantly smaller than mean of Firemaster (p < 0.05) 146 and Glover, 1977). Each compound was administered at 30 mg/kg body weight but because of its extreme toxicity the second congener was also given at 2 mg/kg. These rats were sacrificed one week after treatment. Table 12 shows that even though cytochrome P-450 was only slightly induced (125% of control) by 30 mg of congener 2/kg its absorption maximum was shifted to 448.5 nm. Likewise 30 or 2 mg HBB shifted the Amax to 448.5 nm, but resulted in a significantly larger induction of cytochrome P-450 hemoproteins (230 and 190% of control, respectively). Both doses of the latter congener induced benzo(a)pyrene hydroxylase activity by 7.8-fold, a level that slightly exceeds that caused by 30 mg congener glkg (6.25-fold). The ability to conjugate p-nitrophenol was similarly induced by'all three treatments by about 3-fold. The 455/430 nm absorbance ratio of ethyl isocyanide difference spectra increased over 7-fold in microsomes from rats given the higher dose (30 mg/kg) of HBB. An equivalent injection of congener g increased this parameter by 3.3-fold, a level slightly less than that produced by a much lower dose (2 mg/kg) of HBB. Other enzymes that are typically induced by PB were also assayed in microsomes from rats given HBB. Interestingly, aminopyrine-N-demethylase activity was only 28% of control in microsomes from rats given the larger dose (30 mg/kg). However, at 2 mg/kg this congener caused only a slight depression (18%) in this activity. No depression resulted from 30 mg/kg congener 2. This treatment, however, slightly induced epoxide hydrolase (138% of control) which was not significantly changed by either dose of HBB. NADPH-cytochrome P-450 reductase was depressed 40% by the larger dose of HBB. The profile of microsomal protein induction was also monitored by polyacrylamide gel electrophoresis (Figure 16). Band II with an apparent molecular weight of 51,000 daltons seems to be the predominant polypeptide(s) in control microsomes and appears to specifically respond to PB-induction but not 147 050205 553 : 5.56555 0:5 .:0_005_:_ 3.05.. 5 00 5:0 20500.5 6050005 .6 5:06 0: 30502550 053055 55: 0555.5-M 5:50:00 50.: 55505053. *0 53:55 5:0 .05 H 55:: 5E 55 m535> 50:205..” 5550 552, 5:0 _A0x\0E N .5 03 mmI :0 A0V<0E 05 M 5:50:00 .Qobcoov 0VO mmI 5 M 5:50:00 055025EU< 505m 00 555E55Q 0512—00552 03:0 _5E0505=2 52l— 5E0m NH mjm UEBEEEUd. 3mm .5 whoomEmhmo mcfizonmuoi GEO 35.8222 .531. oEom .3 MIKE; 168 35.53% Sod v & bucmoEEEm 5.5 53:; .550 :5 “Sod A 3 58.4 93255 9:63 _o.5coo 3 5:86 2525353 5305 GE . a_E\mm_oEc 5555 955.08: 0 n m $.qu Aw 5cmocoOv E H n: 3 H RR 3 H 3 3 H 98 85 H and mm H 8m 35 H EH -.q..m..~.m.s.m.N 233 AM emcoocoOv E H :2 3 H ER 3 H 6.2 62 H 6.: 86 H 34 on H NR 33 H in Jeanninw Amid Aw. emcmmcoOV on H 65: 34 H non no H 6.2 .1 H :N 35 H 84 2 H NE N3 H 8H -.m..s..N.s.m.N amqv Q. cmcmocoUv o: H NE on H 9% 6.3 H 2 3 H 16.2 85 H 86 «N H 63 :3 H Bu -.m..~.m.6.~ 83 3 H 23 3 H a: 3 H 3 no H N: 85 H one 3 H 62 8.6 H ad .6680 0mm— mmmuOJUOp D D XGCC nmmmhmvmcm: ammo. 138.65 mam—xcumEoc 332 E: 93.6. ACE . 5 35:55:. 4.30.5020 no.5»; ocmio 25:23 03393 550.508me 599.0 do: 62665 -3628 -6554 025 -Iao omhooflEEcd. 35m Ho 2558950 GEE—055.2 9:0 _mEomo5_§ 52:. mEom ca UTE/5r 169 Congener Z was the most effective inducer of this activity (16-fold) while congener 1 was almost non-effective (Table 14). Congeners 2 and 9 caused intermediate increases. Results of assays for microsomal p-nitrophenol-UDP-glucuronyltrans- ferase activity were also in agreement with those for the AHH induction (Table 14). To further evaluate the pattern of induction, microsomal proteins were analyzed by SDS-polyacrylamide gel electrophoresis. Microsomes from rats administered PB or MC were also included for reference. The proteins in the 50,000 daltons molecular weight region are the ones of interest (Figure 21). Like MC, but even more effectively, congener Z induced protein bands IV (55,000 daltons) and III (53,000 daltons), both of which (especially band IV) were apparently not induced by PB. The same treatment, however, did not seem to cause any increase in band 11 (51,000 daltons) which may be composed of at least two proteins of similar molecular weight (within 1000 daltons of each other). Band II seemed to specifically respond to PB but not to MC induction. Both congeners 1 and 2 caused a similar pattern of induction. Like PB, these congeners mainly induced band II. Congener _5_ was rather unique since it resembled both PB and MC in inducing bands II, III, and IV. Band I (49,000 daltons) which appears to specifically respond to PB but not to MC, was responsive to all PBB treatments to varying degrees. DISCUSSION Firemaster, a heterogeneous mixture composed primarily of twelve PBB congeners, is immuno- and hepatotoxic, and a mixed-type inducer of hepatic microsomal drug metabolizing enzymes. The pattern of induction closely resembles that resulting from the combined treatment by both PB and MC (Troisi, 1975; Dent g a_l., 1976a,b; Moore _ei a_l., 1978b, 1979; Dannan, 1978). The 170 FIGURE 21. SDS-POLYACRYLAMIDE GEL ELECTROPHORESIS PROFILE OF LIVER MICROSOMAL PROTEINS FROM RATS ADMINISTERED CONGENERS 1, 5,1, OR 9 Samples from left to right include: (1) mixture of 2 pg of each of bovine serum albumin (66,000), catalase (58,000), E; gcfl glutamate dehydrogenase (53,000), and egg albumin (45,000); or 20 ug of liver microsomal protein from (2) control; (3) PB-; (4) MC-; (5) congener 1-; (6) congener _5_-; (7) congener Z-; or (8) congener 9-treated rat. 171 e m a . T... _ _ IEJMumfixxmfi JVQOékfiw JmCYLWYmd JmfiWQvN 02 mm JOEPZOO wom zsxtssfi4h3'? zaAdiZRTAL5L 3,4,5,3',4',5'- 179 FIGURE 22 180 In female rat liver microsomes, four protein bands in the 50,000-60,000 dalton molecular weight region were evaluated for their apparent preferential induction by Firemaster, PB, MC, or PB + MC (Dent _e_t 21., 1978). The lowest molecular weight band (band 1), 49,000, was also assigned to epoxide hydrolase since it had an identical electrophoretic mobility to pure enzyme. Therefore, band Iof this study should be identical to band 1 of Dent g 31. (1978). Band 11 seems to match band 2 of Dent e_t_ a_l. study, and both results agree that this band is present in all microsomes even though it seems to specifically respond to PB-type inducers (PB, Firemaster, or PB + MC). This band also seems to specifically respond to the di-9_I_‘_t11_9_ substituted congeners 1, 3, 2, and 12, all of which are strict PB-type inducers (Figures 21 and 22). Band 3 of the study by Dent gt a1. has an apparent molecular weight of 54,000, in close agreement with band 111 (53,000). However, besides Firemaster (which Dent _e_t a1. found as the only effective inducer of this band) MC and especially congeners 2, 2, Z and HBB effectively intensified this band. Band IV (55,000) may correspond to band 4 (58,000) of Dent e1 g1. (there was no 58,000 daltons protein band that responded to any treatment in this study). However, while Dent e_t 31. found this band to specifically respond to MC (or PB + MC) only, Firemaster, or especially congeners 2, g, Z, and HBB were very effective inducers of this band as well. These findings strengthen the assumption that induction by the mixture of PBB is partially similar to MC and should be mainly caused by congeners such as 2, g, and Z. These data, therefore, do not support the conclusion of Dent e1 a_l. (1978) who suggested that PBB induce the d_e Q9112 synthesis of a control like hemoproteins rather than inducing the major MC- inducible hemoproteins. The present results, therefore, still agree with the general classification of Firemaster as a mixed-type inducer (Dent g; _a_1., 1976a,b). The results of the microsomal drug metabolizing enzyme studies and those of the electrophoretic separations of microsomal proteins (Tables 9 and 14, and 181 Figures 16, 21 and 22) seem to indicate that epoxide hydrolase activity is not so closely associated with the mixed-function oxidase enzymes which are induced by PB or PB-Iike inducers. Epoxide hydrolase is commonly believed to be one of several enzymes that are induced by PB or PB—like inducing compounds including some PCB or PBB congeners which are strict PB-type inducters (Bresnick g a_l_., 1977; Kohli e_t _a__l_., 1979; Moore e1g1., 1978b, 1979; Poland _e_tgl_., 1980). In general, the induction of epoxide hydrolase correlates well with the extent of induction in cytochrome P-450 form(s), that are normally induced by PB and measured by aminopyrine-N-demethylase activity, and the cytochrome P450 associated reductase enzyme commonly known as NADPH-cytochrome P-450 reductase (Moore £21., 1978b, 1979; Poland e_t_ Q1, 1980). However, the results of the present studies with congeners 1, and Z do not seem to favor such a correlation. Congener Z, mainly an MC-type inducer (Table 14 and Figures 21, and 22) is a somewhat better inducer of epoxide bydrolase than the strict PB-type inducer congener 1. Similar findings have also been obtained with congener 2, also predominantly an MC-type inducer (Table 9 and Figures 16 and 22). These findings may indicate that epoxide hydrolase induction may not necessarily be coordinately expressed with the PB-type cytochrome P-450 hemoprotein(s) and NADPH- cytochrome P-450 reductase. Somewhat similar findings have recently been reported when a number of PCB congeners were compared for their abilities to induce microsomal epoxide hydrolase and cytosolic glutathione-S-transferase in mice (Ahotupa, 1981). In that study the effects on epoxide hydrolase by 3,4,3',4'- tetrachlorobiphenyl, a known strict MC-type inducer (Goldstein _e_t 21., 1977), did not match those of MC since MC slightly depressed this activity while the PCB congener seemed to induce it by nearly 50%. Furthermore, 2,4,6,2',4',6'- hexachlorobiphenyl, which according to Goldstein e_t a_l. (1977) was only a weak PB- type inducer, caused comparable induction in epoxide hydrolase to those of PB or 182 other strong PB-type PCB inducers (Ahotupa, 1981). Further studies will be required to evaluate the relationship between the mixed-function oxidase enzymes and epoxide hydrolase, particularly with regard to the expression of their structural genes and the requirements for inducing them. The extent of MC-like induction arising from various halogenated aromatic hydrocarbons is commonly considered a rather sensitive measure of TCDD-like toxicity, even though no immediate cause and effect relationships are known to exist between MC-type induction and other symptoms of toxicity (Poland g 311., 1979). Since several parameters are known to be diagnostic of a TCDD-like toxicity it was of interest to evaluate one or more of these parameters to see if certain PBB congeners can cause other TCDD-like toxic responses. There appears to be good correlations between the toxicity of a PBB congener and its ability to cause an MC-type induction of hepatic microsomal drug metabolizing enzymes. Congener 1, for example, had little effect on the liver or the lymphoid organs and it had no apparent MC-like induction activity. Congener Z, on the other hand, was responsible for the most severe toxicity (as evidenced by thymus involution and depressed body weight gain, and by the severe hepatic Ultrastructural changes) and it appeared to be the most potent MC-type inducer of all the PBB congeners of Firemaster. Even though the thymus of rats treated with congener 2 appeared normal there was a slight decrease in the weight of this tissue accompanied by a slight reduction in body weight gain. The humoral immune system, however, was apparently a rather sensitive measure of toxicity since there was a 70% reduction in response to SRBC by rats treated with congener 2 (Table 11). This congener was even more immunotoxic than Firemaster FF-l which only caused a 50% reduction in the ability of the T-helper cells and B-cells to respond to the same antigen. Similar qualitative effects were also seen when mice were fed diets containing various concentrations of Firemaster FF-l (Fraker, 1980). 183 Since events occurring in the spleen are generally representative of the types of changes taking place in the lymph nodes and other lymphatic organs, the 50-70% reduction in splenic plaque forming capacity by exposure to the mixture or pure congener 2 may represent an appreciable loss in host defense capacity against disease and infection. The uniform reduction in the percentage of directs and indirects suggests that both the mixture and congener 2 substantially impaired both the T-helper cells and B-cells involved in antibody mediated responses. At the present time it is uncertain whether these immunotoxic effects are reversible. If such was the case the time required for a complete recovery also remains to be determined. It should be mentioned that rats fed diets containing 10 or 100 ppm HBB, a strict MC-type inducer, suffered thymic involution and declined body weight gain in a dose-dependent manner (Render, 1980). Results of the hepatic histopathological and ultrastructural studies are consistent with the results of other toxicity studies including those on the induction of the drug metabolizing enzymes. The most severe histopathological changes, including individualization of hepatocytes and cytoplasmic vacuolation, were seen in livers of rats administered congeners 2, g and Z all of which caused relatively strong MC-type induction. These congeners also seemed to increase the numbers of lipid droplets and to cause a marked increase in RER and disorganization in the normal stacking pattern of these membranes. They also seemed to give rise to concentric arrays of laminated ER membranes around a few of the lipid vacuoles. TCDD has also been shown to cause similar changes including the formation of similar ultrastructural configurations which have also been called whorls or myelin figures (Norback and Allen, 1973; Gasiewicz _e_t gl_., 1980). Similar ultrastructural changes were also reported after HBB, an extremely toxic congener not present in Firemaster, was fed to rats (Render, 1980). Such changes were not seen, however, with any of the relatively non-toxic PBB congeners that were strict PB-like 184 inducers including congeners 1, 2 and 12 of these studies and congener 3 of other studies (Render, 1980; Gupta e_t_ _a_1_., 1981). Congener 2, a di-gflhg brominated congener that was a mixed-type inducer, seemed to cause somewhat intermediate hepatic histopathological and ultrastructural changes. Firemaster, can also cause hepatic ultrastructural changes similar to those seen with congeners 2, 2, g, and Z (Sleight and Sanger, 1976; Gupta and Moore, 1979; Gupta e_t_ £11., 1981; and these studies). A relationship may, therefore, exist between the ability of a PBB congener to cause an MC-like induction response and the formation of cytoplasmic lipid droplets which are occasionally surrounded by laminated ER membranes (whorls). Even though the significance of these ultrastructural changes is not yet understood, the increased number of lipid droplets may indicate an abnormality in lipid metabolism. Gasiewicz e_t_ a_l. (1980) have recently suggested that a similar accumulation in lipid granules following TCDD treatment may be due to an alteration in the control of lipid biosynthesis from carbohydrates. However, further research may be required for understanding the etiology of this phenomenon. Hepatomegaly is commonly seen in rats given Firemaster or pure PBB congeners regardless of the type of induction they may cause. However, congener 1 failed to increase the weight of the liver although it caused a considerable PB- Iike induction of the microsomal drug metabolizing enzymes. This may be because this compound, unlike most other congeners, can be appreciably metabolized by liver microsomes, and therefore it must be less persistent than the other congeners of these studies (Dannan, 1978; Moore e_t a_l., 1980). 12 M evidence for the low persistence of this congener has also been provided by its preferential disappearance from milk or liver of rats pretreated with Firemaster FF-l (Dannan, 1978). Proliferation of the SER is another common change that appears to be 185 evoked by all inducing congeners regardless of the type of induction. Various other chemicals are also known to cause SER proliferation (Trump and Jones, 1978). Under the conditions of these experiments (two weeks after a single ip injection of 90 mg/kg) congener 2 or Firemaster FF-l were not toxic enough to cause any hematologic changes or hepatocellular necrosis (Table 10). Previous studies have also shown no alterations in erythrocyte count, hemoglobin or serum alkaline phosphatase and glutamate-oxaloacetate transaminase in rats fed diets containing Firemaster BP—6 (Sleight and Sanger, 1976). Evidence for a hepatocellular damage was seen, however, when certain serum enzymes were elevated in rats administered TCDD (Zinkl e1 a_l., 1973). Hematologic changes indicative of hemoconcentration and decreased platelet counts were also observed in these rats (Zinkl 9131., 1973). The degree of planarity that can be assumed by polyhalogenated biphenyls, shown to be mainly influenced by the number of _o_rt[_i_g halogens, may determine their toxicity and type of microsomal enzyme induction (Goldstein, 1979; Poland e1 gl_., 1979; Moore e_t_ 11., 1980). Polyhalogenated aromatic compounds that can assume a planar conformation and are approximate isostereomers with TCDD are commonly thought to interact with the cytoplasmic TCDD-receptor before their toxicity and MC-type induction effects are expressed (Poland and Glover, 1977, 1980; Poland 3 g1, 1979). Certain polyhalogenated biphenyls with no halogen substituents adjacent to the biphenyl bridge (2,2',6 or 6' positions) are considered capable of assuming a planar conformation since they are quite toxic and MC-like inducers, and can serve as ligands to the TCDD-receptor (Goldstein gt g1, 1977; Poland and Glover, 1977, 1980; Poland 9111., 1979; Moore e_t {a_l., 1980). Many other congeners with two or possibly more 9LT“). halogens evoke no toxicity or MC-type induction response and do not serve as ligands to the TCDD-receptor (Goldstein e_t gl_., 1977; Poland and Glover, 1977, 1980; Poland e_t_ a_l., 1979; Moore a a_l., 1980). 186 The question of whether an 2292 halogenated biphenyl can resemble TCDD in being able to evoke similar toxicity and microsomal enzyme induction (MC-type) effects has been a controversial one. The original structure-activity relationship rules did not allow any M substituted congener to have any of these effects (Goldstein e_t a_l., 1977; Poland and Glover, 1977; Goldstein, 1979). Adding an m bromine to the brominated analogue of 3,4,3',4'-tetrachlorobiphenyl, a highly toxic and an MC-type inducer (Goldstein e_t fl" 1977; Poland and Glover, 1977), yields the brominated analogue 2,4,5,3',4'-pentabrominated biphenyl (congener 2). Congeners g and Z, 2,4,5,3',4',5'-, and 2,3,4,5,3',4'-hexabromobiphenyl, respectively, are mono-gr_t_h_o_ substituted derivatives of the possibly very potent MC-type inducer 3,4,5,3',4'-p‘entabrominated biphenyl whose chlorinated analogue is considered the most toxic and MC-like inducer of all PCB congeners (Yoshimura gt a_l., 1979). Congeners 2, g, and Z share certain common substitution features where each has two w bromines, two or three _rfle_ta_ bromines, and a single M bromine. Congeners 2, g, and Z seem to retain a sufficient degree of planarity to presumably interact with the TCDD-receptor protein to ultimately cause toxicity and MC-type induction. However, as shown by comparing the effects of HBB and congener 2 (Table 12) the mono-92mg substituted congeners may be at least an order of magnitude less toxic than their related non-951119 halogenated compounds. In addition, these congeners seem to cause a variable degree of PB-like induction. These compounds, therefore, may be considered intermediate (at various degrees) between the non-M halogenated compounds that are highly toxic and potent MC-type inducers and the PB-like congeners that have two (or possibly more) o_rt_h_c_)_ halogens. Results of the chemical studies (Chapter 1) also agree that the mono- ortho brominated PBB congeners have somewhat intermediate chemical properties 187 between the non-212319 substituted congeners such as 3,4,5,3',4',5'-hexabromo- biphenyl (HBB) and the di-w substituted congeners exemplified by 2,4,5,2',4',5'- hexabromobiphenyl (congener _4). Results of several other recent studies have also shown that the mixed-type inducers among halogenated biphenyls are not that uncommon. A number of mono- _o_1;t_h_q substituted PCB congeners had, at variable degrees, the effects of PB and MC (Yoshihara g gl_., 1979; Parkinson e_t_ 21., 1979, 1980b,c). These workers have also recognized that a biphenyl with halogens at both w carbons and at two or more awe—ta carbons can cause an MC-type induction and a TCDD-like toxicity even if another halogen occupies one of the carbons m to the biphenyl bridge. Certain di'-91‘_t_f_12 substituted congeners have also been reported to evoke an induction response similar to both PB and MC. The earlier findings of Stonard and Greig (1976) regarding mixed-type induction by 2,3,4,2',3',4'- and 2,3,4,2',4',5'- hexachlorobiphenyl, the latter compound being the chlorinated analogue of congener 2, have more recently been confirmed by Parkinson g a_l. (1980a, 1981). Based on the results of induction by these and other PCB congeners it was suggested that di-M chlorinated PCB congeners that are substituted at both Lag positions and at least at two aria positions, and have a 2,3,4-trichloro substitution pattern on one ring are mixed-type inducers (Parkinson 21 _a_l_., 1981). In this study the mixed-type induction effects of congener 2 are in agreement with the similar results on the chlorinated analogue, 2,3,4,2',4',5'-hexachlorobiphenyl (Stonard and Greig, 1976; Parkinson g1 a_l_., 1980a, 1981). The slight degree of MC- type induction by congener 2 seems also to match the extent of MC-type induction due to the chlorinated analogue, 2,3,4,5,2',3',4'-heptachlorobiphenyl (Parkinson a 331., 1981). However, as mentioned earlier, the nearly 3% contamination due to congeners 2 and 2 may, at least partly, have been responsible for the slight MC- type induction by congener 2. It must be pointed out that several congeners that 188 have been classified as mixed-type inducers did not seem to exactly reproduce the effects of the combined PB and MC treatments, since they tended to predominantly resemble one or the other of the two prototype inducers (Parkinson _e_t__a_l_., 1980a,b,c, 1981). About the same time these studies were being communicated (Dannan and Aust, 1980; Dannan g gl_., 1980) synthetically prepared congeners 2 and Z were reported to cause a mixed-type and an MC-type induction, respectively (Robertson e_t_ a_l_., 1980, 1981a). In the first study (Robertson _e_t; a_l., 1980) microsomal cytochrome P-450 reductase and aminopyrine-N-demethylase activities appeared significantly less induced by congener 2 than by PB (administered alone or with MC). However, benzo(a)pyrene hydroxylase was induced to the same extent by congener 2 or MC. The results presented in this thesis and those of Robertson a 21. (1980), therefore, agree that congener 2 is more like an MC than a PB-type inducer. The results of Robertson _e_t; a_l. (19813) with synthetically prepared and 90% pure 2,3,4,5,3',4'-hexabromobiphenyl also agree that congener Z is mainly an MC-type inducer and could only be a subtle or very weak PB-type inducer. As discussed earlier, the pharmacotoxicological properties of TCDD and similar halogenated aromatic hydrocarbons are thought to be mediated by a high affinity specific binding to hepatic cytosolic protein(s) named the TCDD-receptor (Poland g 31., 1979). The avidity by which the receptor binds the various congeners can be expressed by their equilibrium dissociation constant (KD) which can be measured by an _i_n_ ELM competitive binding assay (Poland _e_t a_l., 1976a). The ability of a congener to displace the binding of 3H-TCDD correlates well with the extent of its TCDD-like pharmacotoxicological effects and AHH induction (Poland e_t a_l., 1979; Poland and Glover, 1980). Since several PBB congeners of Firemaster caused various degrees of MC (TCDD)-like induction and toxicity it was of interest to compare their binding affinities to the cytosolic receptor species. 189 The results of such studies, which were performed in Dr. Alan Poland's laboratory in the University of Wisconsin, are summarized in Table 15. At a maximal 6 M (2000 x the concentration of 3H-TCDD) the KO of most of the congeners as well as Firemaster could not be exactly determined since concentration of 1.1 x 10- they tended to be rather insoluble in aqueous solutions. The cytosolic binding affinity could only be determined for three congeners one of which was HBB whose KD (2.0 x 10.8 M) was nearly 100 times higher than that of TCDD (KD = 0.27 x 10'9 M). Congeners 2 and 2 were nearly 10-f01d less competitive than HBB in displacing 3H-TCDD (Table 15). These results seem to agree with those of the drug metabolizing enzymes (Table 12) where congener 2 seemed to be nearly 20 times less-potent as an MC-type inducer than HBB. Furthermore, the binding affinity of congener 2 seems to also agree with its moderate potency as an MC- type inducer (Table 14). The remaining congeners and Firemaster FF-l seemed to have lower binding affinities (K > 4.9x 10-7) that could not be determined D because of limited solubility. All of these congeners, with the exception of congener é, displayed no MC-type induction activity. Congener é, however, would have been expected to express a binding affinity similar to that of congeners 2 and 2 since all three congeners seemed to evoke a nearly comparable MC-type induction response. Failure to demonstrate a specific binding affinity for congener 2 may have been due to solubility problems or an experimental error especially since this congener was assayed only once. At the time these experiments were performed congener Z was not available, and therefore it was not tested. However, Robertson 91 a_l. (1981a) have recently attempted to measure the cytosolic binding affinity of a synthetically prepared congener Z but their efforts were also hampered by solubility problems that precluded such measurements. While a comprehensive assessment of the relationship between the structure of a PBB congener and its induction pattern is beyond the scope of these studies it 190 TABLE 15 Competition by Brominated Biphenyl Congeners for the Specific Binding of [3H ]-TCDD to Hepatic Cytosol The binding affinities for PBB congeners by hepatic cytosol protein were estimated by the capacity of these compounds to compete with 3H-TCDD (at 0.55 x 10'9 M) for specific binding sites in C57BL/6J mouse liver cytosol. A maximum concentration of 1.1 x 10"6 M was used of each PBB. These experiments were done by Dr. Alan Poland as previously described (Poland g gl_., 1976a) where the equilibrium dissociation constant (KB) of TCDD was estimated to be 0.27 nM. PBB Binding Affinity Number of Congener KD (nM) Experiments 3,4,5,3',4',5'- ' (HBB) 20 2 2,4,5,3',4'- (2) 130 2 2,3,4,2',4',5'- (2) 210 2 2,4,5,2',5'- (1) >490 1 2,4,5,2',4',5- (g) > 490 1 2,4,5,3',4',5'- (_6_) > 490 1 2,3,4,5,2',4',5'- (g) > 490 1 2,3,4,5,2',3',4'- (g) > 490 2 2,3,4,5,2',3',4',5'- (g) > 490 1 Firemaster FF-1 > 490 2 191 is rather important to consider the following matters. First of all, with only one exception the PCB and PBB congeners that share an identical halogenation pattern on the biphenyl nucleus seem to evoke similar (but not necessarily quantitatively identical) induction patterns of microsomal enzymes. 2,4,5,3',4',5'-Hexabrominated biphenyl (congener _6_) has been classified as a mixed-type inducer (Dannan 91 a1, 1978a; and present studies) while its chlorinated analogue has been considered strictly a PB-type inducer (Parkinson 21 a_l., 1980b). Interestingly, the latter finding represents an exception to the structure-activity rules for microsomal enzyme induction by PCB congeners that were reported later by the same group (Parkinson _e1 a_l., 1980c). Secondly, since an "absolute” definition of the structure- function relationship rules may require the testing of all 209 possible PCB or PBB congeners, any conclusions based on studying a fraction of them may not be conclusive. Therefore, such rules as originally set down by Goldstein and coworkers (Goldstein e_t_ gl_., 1977; Goldstein, 1979) and Poland and Glover (1977) could not account for the type of induction caused by a number of mono-and di- o_rt_h_o_ halogenated PCB and PBB congeners (Dannan _e1 gl_., 1978a; Yoshihara e_t_ 21., 1979; Parkinson g; gl_., 1979, 19803,b,c, 1981; Robertson e1 _a_1., 1980; and present studies). Thirdly, in order to arrange the polyhalogenated biphenyls in the order of their MC-type or PB-type induction potency there is an obvious need for studying the dose-response relationships so that a median effective dose (EDSO) can be determined for each congener. The completion of such a task will obviously require large amounts of pure congeners, and will also require one or more sensitive and reliable enzymatic assays for measuring the extent of MC-type induction. However, it generally appears that the efficiency of either type of induction may depend on a number of factors including the stereochemical and electronic properties of the congeners, as well as their i_r_i 11129 persistence. The stereochemical conformation of a polyhalogenated biphenyl molecule may mainly 192 be determined by the degree of halogenation at carbons o_rt_f2 to the biphenyl bridge (see Chapter 1), while its electronic properties may be influenced the most by m_et_a; and gala halogenation. The stereochemical and electronic properties may in turn determine the recognition and binding of these congeners by their specific receptors that are allegedly involved in mediating the biological effects of the two types of congeners. However, while the stereochemical and electronic properties may determine both the type and potency of induction, the extent of induction may also be determined by the persistence of the congeners. Certain PCB or PBB congeners can be metabolized and eliminated at considerable rates, and therefore are appreciably less persistent than others (Matthews and Kato, 1979; Dannan e_t gl_., l978b;~Moore _e_g £11., 1980). Differences in rates of metabolism and excretion should determine their effective _ig M concentrations, and, consequently, their availability for binding by the induction receptors. The experiments reported in this Chapter have determined some of the pharmacotoxicological properties of seven brominated biphenyls (1, Z, _5_, 2, Z, 2, and E) totalling nearly 20% and 30% of Firemaster FF-l and Firemaster BP-6, respectively (the relative concentrations of these congeners in both mixtures are summarized in Table 4). Including congeners _4 and _8_ (Moore _e_t. _a_l., 1978b, 1979) nearly 97% of the PBB congeners in Firemaster have thus been studied for their induction effects on rat liver microsomal drug metabolizing enzymes. Congeners 2, 2, g and Z are the only congeners of Firemaster that resemble TCDD in causing an MC-type induction and in certain cases toxicity (from their structures congeners _3_, 12 and _11 are not expected to cause any measurable TCDD-like effects). The four PBB congeners, however, are apparently not equipotent. Even though dose- response studies were not performed the relative TCDD-like potency of these congeners as well as HBB may be ranked in the following order: TCDD >> HBB > Z > 2 or g > 2. Congeners 2, 2, g and Z seemed to also cause a variable degree of 193 PB-type induction that in the case of congeners 2 and Z was very weak or subtle. These four congeners total nearly 15% and 25% of Firemaster FF-l and Firemaster BP-6, respectively and are, therefore, expected to account for most if not all of the TCDD—like effects of these mixtures. These predictions will be further investigated in Chapter 3. CHAPTER 3 RECONSTITUTION OF SOME BIOCHEMICAL AND TOXICOLOGICAL EFFECTS OF COMMERCIAL MIXTURES OF POLYBROMINATED BIPHENYLS 194 195 ABSTRACT Two experiments were performed to test the hypothesis that the TCDD-like toxic and induction effects of crude Firemaster mixtures are mainly due to 2,4,5,3',4'-penta-, 2,3,4,2',4',5'-hexa-, 2,4,5,3',4',5'-hexa-, and 2,3,4,5,3',4'—hexa- bromobiphenyl (congeners 2, 2, g and Z, respectively.) One week before sacrifice, male rats were administered a single ip injection (90 mg/kg) of Firemaster FF-l, Firemaster BP-6, or a reconstituted mixture formulated to resemble BP-6 by mixing nine purified polybrominated biphenyl (PBB) congeners, totalling nearly 97% of either Firemaster mixture. These treatments were compared for their effects on liver microsomal drug metabolizing enzymes and on body weight gain and on the weights of selected organs. In Experiment I, the effects of Firemaster FF-l and Firemaster BP-6, containing nearly 15% and 25%, respectively, of congeners 2, 2, _6_ and Z, were compared. Enzyme activities, including NADPH-cytochrome P-450 reductase, aminopyrine-N-demethylase, and epoxide hydrolase, that are typically induced by phenobarbital (PB) were similarly induced by both mixtures. The content of cytochrome P-450 hemoproteins was nearly quadrupled by both treatments; however, the absorption maxima (Amax) of the difference spectra were shifted more by Firemaster BP-6 than by Firemaster FF-l (A max = 449 and 449.5 nm, respectively). In agreement with this observation, Firemaster BP-6 was also a significantly better inducer of the ethyl isocyanide 455:430 nm difference spectral ratio, and the activities of benzo(a)pyrene hydroxylase, and p-nitrophenol- UDP-glucuronyltransferase, all of which are typically induced by 3-methylcholan- threne (MC). Both mixtures had no effect on body weight gain or on the weights of the thymus or spleen, but liver enlargement was evident. While these results agree that a mixed (PB + MC)-type induction is evoked by both mixtures, Firemaster BP-6 which contains nearly 65% more of congeners 2, 2, g and Z is a more potent 196 MC-type inducer than Firemaster FF-l. In Experiment II, the effects of Firemaster FF-l and the reconstituted PBB mixture, which resembles Firemaster BP-6 in composition, were compared. Like the results of Experiment I, both mixtures caused comparable inductions in the drug metabolizing enzyme activities that are normally induced by PB, but the extent of MC-type induction was significantly higher in rats administered the reconstituted mixture. Furthermore, the effects of the reconstituted mixture on almost all the parameters examined seemed to be nearly identical to those due to Firemaster BP-6. Therefore, it may be concluded that the TCDD-like toxic and induction effects of Firemaster (FF-1 or BP-6) must be due to the additive similar effects of congeners 2, 2, g, and Z, and that no other component in these mixtures, a brominated biphenyl or otherwise, may significantly contribute to these effects. INTRODUCTION The accidental contamination of Michigan's environment by a crude mixture of polybrominated biphenyls (Firemaster) has prompted an extensive amount of research on their biological and toxicological effects. Thus far, most of these studies have been performed with the crude mixture itself which has been shown to resemble TCDD in causing a number of toxic responses including immunosuppres- sion, thymus involution, hepatomegaly, porphyria, and loss of body weight (Sleight and Sanger, 1976; Gupta and Moore, 1979; Fraker, 1980; Gupta gt a1, 1981). In addition, Firemaster induces the microsomal enzymes normally induced by PB, and therefore it is commonly referred to as a mixed-type inducer (Dent £91., 1976a,b). Assessment of toxicity and/or enzyme induction by PBB mixtures may be complicated since various commercial preparations may significantly be different in their PBB composition. For example, the relative concentrations of the twelve 197 major PBB congeners that are common between Firemaster FF-l and Firemaster BP-6 (lot numbers FH7042 and 6224-A, respectively) are appreciably different between these two mixtures (Figure 1, and Table 4). Furthermore, even though the highly toxic brominated-dibenzofurans or brominated dibenzo-E-dioxins could not be detected (at < 0.5 ppm) in Firemaster BP-6 (Hass a 31., 1978) it remained possible that one or more of the nearly twenty minor non PBB impurities, including nearly 200 ppm of penta- and hexa-brominated naphthalenes, could be responsible for at least part of the toxicity of the mixture. Each of the PBB congeners in Firemaster has at least one £1112 bromine, but according to the original structure- activity relationship rules (Goldstein e_t_ 31., 1977; Goldstein, 1979) none would have been expected to cause any MC-type induction effects. Various indirect attempts have been made to identify the toxic MC-like inducers in Firemaster following its fractionation by column chromatography. A number of the resulting crude fractions caused a variable degree of toxicity and/or MC-type induction (Safe 51 a_l., 1978; Hass 3g a_l., 1978; Kimbrough e_t_ 33., 1977; Moore Si 31., 1980; Robertson 31 33., 1981b). However, none of these studies could conclusively identify or account for all of the TCDD-like toxins in the mixture. A more direct approach has been adopted where nine purified PBB congeners totalling nearly 97% of PBB in Firemaster FF-l and Firemaster BP-6 have been examined for some of their pharmacotoxicological effects (Moore e_t a_l., 1978b, 1979; and Chapter 2). Three of these congeners (_2_, 2, and Z) were mono-313m brominated and all three were toxic and induced hepatic microsomal drug metabolizing enzymes typically induced by MC (Chapter 2). To a lesser and variable degree some of the liver microsomal enzymes that PB induces were also induced by these congeners. A di-9_r_tm brominated congener (2) was also classified as a mixed-type inducer since it caused nearly equal MC- and PB-like induction responses (Chapter 2). The remaining five congeners, totalling nearly 82% and 72% 198 of Firemaster FF-l and Firemaster BP-6, respectively, were di-p_rlli_g brominated and all five were strict PB-type inducers (Moore 31 31., 1978b, 1979; and Chapter 2). Since the four PBB congeners that cause some toxicity and MC (TCDD)-like induction effects constitute nearly 15% and 25% of all PBB in Firemaster FF-l and Firemaster BP-6, respectively (Table 4), it is likely that they account for all of the similar effects of the crude mixtures. However, it is also likely that a trace contaminant(s) that is (are) hard to detect may also be responsible for at least part of these effects. This possibility would be evaluated by recombining the purified congeners to a reconstituted Firemaster BP-6-like mixture. Some of the biological effects of the reconstituted mixture were compared to those of crude Firemaster mixtures. The objective was to determine if some minor yet very toxic component(s) of Firemaster, a brominated biphenyl or otherwise, would significantly contribute to the TCDD-like toxicity of the crude mixtures. MATERIALS AND METHODS Materials All nine PBB congeners that were employed in this study were purified from Firemaster as described in Chapter 1. These congeners (1, 2, fl, 2, _6_, Z, 2, 2, and 12) were mixed in the same proportions as found in Firemaster BP-6. The resulting reconstituted mixture is shown in Figure 23 which also shows the composition of Firemaster FF-l. Samples of Firemaster BP-6 (lot no. 6224-A) and Firemaster FF-l (lot no. FH7042, the lot which was allegedly involved in Michigan's incident) were gifts from Michigan Chemical Corporation (St. Louis, Michigan) and Farm Bureau Services (Lansing, Michigan), respectively. In Table 4 are listed the percent 199 FIGURE 23. GAS CHROMATOGRAPHIC ELUTION PROFILES AND STRUCTURES OF POLYBROMINATED BIPHENYLS IN FIREMASTER FF-l AND A MIXTURE RECONSTITUTED WITH PURE CONGENERS Injector port, column (3% OV-l), and detector temperatures were 2800, 2450, and 350°C, respectively. The amount of sample injected of each mixture was not determined. RESPONSE 200 FIGURE 23 RECONSTITUTED MIX I L L 1 1 4 1 1 O 4 8 I2 IS 20 24 28 RETENTION TIME (min) 201 composition of all PBB congeners in the three mixtures. Other chemicals and reagents are the same as those described in Chapter 2. Animals Outbred male Sprague-Dawley rats weighing 100-125 g were purchased from spartan Research Animals, Haslett, Michigan. Before being given any treatment they were allowed an acclimation period of 48 hr. Twelve rats were employed for each of two experiments. In the first experiment (Experiment 1), two groups of rats (4 each) received an ip injection of 90 mg/kg body weight of Firemaster FF-l or Firemaster BP-6, each dissolved in 3 ml PEG. A third group (control) was administered 3 ml PEG/kg. In the second experiment (Experiment II) twelve rats were also equally divided into three groups. Each of the reconstituted PBB mixture or Firemaster FF-l was administered (90 mg/kg, ip) into four rats, while control rats received 3 ml PEG/kg body weight. After treatment, all rats were maintained for one week on regular feed and water until the night before sacrifice when feed was taken away. One week after treatments rats were weighed, then killed by decapitation. At necropsy, weights of the liver, spleen and thymus were recorded for each rat. Preparation of Microsomes Microsomes from individual livers were isolated, washed, and stored as described in Chapter 2. Enzyme Assays All enzyme assays were essentially done as described in Chapter 2. 202 RESULTS Some of the PBB congeners (including 2, 2, 3, and Z) are present at significantly different concentrations in Firemaster FF-l and Firemaster BP-6 (Figure 1 and Table 4). I, therefore, decided to see if these two mixtures may also cause significantly different toxic and MC-type induction effects. In Experiment I, one week following the administration of a single ip injection (90 mg/kg), both mixtures nearly quadrupled the total content of cytochrome P-450 hemoproteins, and doubled NADPH-cytochrome P-450 reductase (Table 16, Experiment I). Firemaster FF-l and Firemaster BP-6 shifted the cytochrome P-450 absorption maxima (Amax) of the CO-difference spectra to nearly 449.5 nm and 449 nm, respectively. These results seemed to agree with the respective increases of 47% and 74% of the absorbance ratios of the EtNC difference spectra of dithionite reduced microsomes. Aminopyrine-N-demethylase activity, which is specifically induced by PB but not by MC, was increased by nearly 2.5-fold by both mixtures. Epoxide hydrolase, another PB-inducible enzyme, was equally induced by both treatments by nearly 3.5-fold. The extent of MC-type induction was assessed by quantitating benzo(a)pyrene hydroxylase (AHH) and g-nitrophenol-UDP-glucuronyl- transferase activities. Both were more significantly induced by Firemaster BP-6 (13.5- and 2.6-fold, respectively) than by Firemaster FF-l (8.5- and 2.1-fold, respectively). A few toxicity parameters were also evaluated such that the degree of toxicity may be correlated with the extent of MC-type induction. As shown in Table 17 (Experiment I) neither mixture significantly reduced the average daily body weight gain or the weight of the spleen or thymus. However, Firemaster FF-1 and Firemaster BP-6 increased liver weights by 45% and 72%, respectively. 203 guy-0.. u;- --I'"I . 5 .51.“. .59 0 cv Om H C555 5 5555.205 5.5 550 .555. v.5 m. 5555855.“. 411.1. 555E557”. ho 0V<0E .mmn. Ho 55535.3. 5:255 5555.55.54. 35W. 00 m 00 5 .5 20558 0V 5595505 :6. 5 235.2 550 a? H 3: 5.5 H RR 15 H 5.2 3 H NS 62.5 H :4 S H N3 Rd H RR 63358865 H-00 .3 H N56 3 H 5.5m 5.5 H no 2 H 52 3.5 H mad 3 H Rs R5 H a: 286652.... S H 23 3 H 5.2 no H 2 5.5 H 5.: 8.5 H 55.5 S H a: B. H 5.5 6:85 : 5:58.590 6.33 5.65 65.5 H 52 5.5 H 5.5» 63 H NS 3 H 0% 68.5 H SH 8 H HR 2.5 H mm... 28882.... $.23 70... 5.5 H 5.55 3 H RR .2 H E E H 4.: 3.5 H 35 R H RR 3.5 H SM 28852: 83 NR H 5.3 S H 1; 5.5 H 3. .3 H :5 8.5 H Rd NN H .2 2.5 H 84 .9585 . 5:5EZ50xm. 555. 55555355.. 555505055 5555. $5.5»: 5555;55E55 A255 E: 09510 0:: .55 K 0 55.555: 1.55.5530 15.5.3 55.5.8 25:23 0985.: mEoucoouxO @0310 do: 8925 -3385 -oc_E< 52.5 .1092 9:256on 005 .5 55,555.). m:5..5> 55552:.EU< 55m “.5 5555555 00.0.5553. 0:50 .5E555502 52.1. 5Eom 0H Him/E. 205 TABLE 17 Body and Organ Weights of Rats Administered Various Mixtures of PBB Four rats per treatment were administered a single ip injection of 3 ml PEG/kg or 90 mg/kg of either of the PBB mixtures one week before sacrifice. Data are expressed as the mean 1 SD. Average daily body weight °/o of Body Weight Treatment gain (g) Liver Spleen Thymus Experiment I Control 3.14 i 0.30 3.83 i 0.18 0.42 : 0.05 0.45 i 0.05 Firemaster 2.80 i 0.62 5.57 : 0.37a 0.39 i 0.07 0.43 i 0.08 FF-l Firemaster 2.34 _+_ 0.62 6.58 : 0.328’b 0.35 2: 0.07 0.42 i 0.04 BP-6 Experiment 11 Control 5.07 i 0.44 4.11 r. 0.14 0.40 2: 0.02 0.43 i 0.03 Firemaster 2.97 : 0.58a 5.45 : 0.40a 0.43 i 0.07 0.37 i 0.05 FF-l Reconstituted 2.50 2: 0.66a 6.21 : 1.10"?"b 0.45 _+_ 0.09 0.32 : 0.04a PBB Mixture Significantly different from mean of control using Student's t-test (p< 0.01) 0 Significantly different than mean of Firemaster FF-l using Student's t-test (p < 0.01) 206 Experiment 11 was concerned with evaluating the effects of the reconstituted mixture representing at least 97% of Firemaster FF-l or Firemaster BP-6 (Figure 23 and Table 4). As shown in Table 16 (Experiment 11) the effects of the reconstituted mixture and Firemaster FF-l on cytochrome P-450 hemoproteins content and on NADPH-cytochrome P-450 reductase, aminopyrine-N-demethylase, and epoxide hydrolase activities were almost indistinguishable. Both treatments induced these parameters by 3.5-fold, 2.5-fold, nearly 2-fold, and 2.6-fold, respectively. However, there were slight but significantly different effects on the A max of the hemoproteins (449 nm and 449.5 nm), the EtNC difference spectral ratio (2- and l.7-fold increase), the induction of benzo(a)pyrene hydroxylase (10.5- and 7.7-fold), and the induction of p-nitrophenol-UDP-glucuronyltransferase (2- and l.6-fold), where in each case the reconstituted mixture was compared to Firemaster FF-l, respectively. In Table 17 the effects of Firemaster FF-l and the reconstituted PBB mixture on body weight gain and on the weights of certain vital organs are compared. The average daily body weight gain was similarly reduced by both treatments. Liver weight to body weight ratio was increased by both treatments, while thymus weight to body weight ratio was significantly depressed (to 75% of control) only by the reconstituted mixture. Spleen weight to body weight ratio was unaffected by either treatment. Results of the SDS-polyacrylamide gel electrophoresis showed no major differences in the profile of induced microsomal proteins from rats treated with Firemaster FF-l, Firemaster ESP—6, or the reconstituted PBB mixture (a photograph of the gel was not available). Like microsomes from rats pretreated with the combination of PB + MC all three PBB mixtures induced bands I-IV whose molecular weights range between 49,000-55,000. As discussed earlier (see Figures 16, 21 and 22) band 1 (49,000) should be due to epoxide hydrolase and it and band 11 (51,000) are normally induced by PB, while MC induces band 111 and IV (53,000 and 207 55,000, respectively). Firemaster BP-6 and the reconstituted PBB mixture, however, seemed to be somewhat better inducers of the last two bands than Firemaster FF-l. The profile of microsomal proteins from Firemaster FF-l treated rats was presented earlier (Figure 16). DISCUSSION Nine PBB congeners comprising nearly 97% of the P88 in Firemaster FF-l or Firemaster BP-6 have been purified and tested for some of their biological effects (Moore et a_l., 1978b, 1979; and Chapter 2). Three mono-9312112 substituted congeners ~(2, g, and _7_) resembled TCDD in causing an appreciable MC-like induction of the microsomal drug metabolizing enzymes and evoking one or more of the toxicity responses that are characteristic of TCDD. Among these toxic responses were decreased body weight gain, atrophy of the thymus, immuno- suppression, and some unique changes in the hepatic histopathology and Ultrastructure (Chapter 2). None of these changes could be seen with any of the di- _o_rt_h_o_ substituted congeners I, 4, g, 2 and l_2_ that were strict PB-type inducers of the hepatic drug metabolizing enzymes (Moore at a_l., 1978b, 1979; Render, 1980; Gupta g a_l., 1981; and Chapter 2). A unique di-_o_rt_l'£ brominated congener (2) caused a mixed-type induction and moderately severe hepatic histopathological and ultrastructural changes including increased lipid accumulation and proliferated and disorganized RER. The purpose of this study was to see if the toxic and MC-like inducing agents of Firemaster may be accounted for by these congeners. Two mixtures of Firemaster, FF-l and BP-6, containing nearly 15% and 25%, respectively, of the toxic congeners (g, 2, Q and Z) (Table 4) exhibited equal PB-like induction activity, but the mixture (SP-6) which contained more of these congeners was the more “My... 208 effective MC-like inducer (Table 16). Thus, the relative MC-like induction potencies of these mixtures seem to be consistent with the hypothesis that congeners 2, 2, _6_ and _7_ are the principal components that cause the toxicity and MC-like induction effects of the crude Firemaster mixtures. The results with the reconstituted PBB mixture were even more conclusive. Congeners 2, _5_, Q and Z made up nearly 23.5% of this mixture, very close to their total concentration (25%) in Firemaster BP-6 (Table 4). However, their concentration in Firemaster FF-l (nearly 15%) was significantly less (Table 4). If these four congeners represented the main toxic components then a nearly identical response should be evoked by Firemaster BP-6 and the reconstituted mixture. Firemaster FF-l should however, be somewhat less effective. This assumption is fully supported by the results of the experiments as shown in Tables 16 and 17. The significant (25%) reduction in thymic weight due only to the reconstituted mixture should not be considered anomalous since the other two treatments seemed to slightly but not significantly reduce the weight of this tissue as well. It is quite likely that the single 90 mg/kg dose of any of these mixtures may be the minimal dose required to affect the lymphatic organs including the thymus. Among all PBB treatments this tissue and to a lesser extent the spleen were most sensitive to congener Z which at a single dose (90 mg/kg) caused a 60% and 15%, respectively, reduction in their weights (Table 13). Doses larger than 90 mg/kg of the other toxic PBB congeners or of any of the three mixtures would be expected to cause similar effects in these tissues. Several attempts have recently been made at identifying the TCDD-like agents after fractionating Firemaster by alumina or Florisil column chroma- tography into two or more fractions of different polarity (Kimbrough _et a_l., 1977, Hass e_t gl_., 1978; Safe e_t a_l., 1978; Moore e_t a_l., 1980). Aside from the PBB congeners, any TCDD-like compounds, such as brominated dibenzo-p-dioxins or l—rui—._1| -—~_—i ,.—...I - 209 brominated dibenzofurans, that may be present in the mixture would be expected to elute in the most polar fraction(s). While these studies were not capable of fully accounting for the TCDD-like agents in Firemaster, at least one study concluded that the bromobiphenyls are most probably responsible for the TCDD-like toxicity and biological effects of the crude mixture (Hass g a_l., 1978). In the same study, no brominated dibenzofurans or dibenzo-p-dioxins could be detected at > 0.5 ppm (Hass fl a_l., 1978). The 200 ppm of brominated naphthalenes, estimated to be present in Firemaster (Hass _e_t _a_l., 1978), was also ruled out as being responsible for the MC-like induction effects of the mixture (Goldstein gt _a_l., 1979). In a recent study by Robertson e_t_ a_l. (1981b) Firemaster BP-6 was fractionated into fractions by alumina (fractions AA and AB) and Florisil (fractions FA, FB’ and FC) column chromatography. These fractions were administered to rats at 100, 6.7, 100, 0.6, or 0.4 mg/kg, respectively, and an attempt was made to correlate the extent of MC-type induction with the P88 composition of these fractions. (No explanation, however, was provided for the rationale behind selecting such doses). Throughout that study, even though congeners 2, 2, _6_, and 1 were precariously implicated as the agents that were responsible for the MC-type induction effects, the conclusion was reached that other yet unidentified components or synergistic effects may also be important (Robertson e_t _a__l_., 1981b). Since only four congeners of the reconstituted mixture were found to produce a toxicity resembling that caused by TCDD, it is very unlikely that other yet unidentified components may also significantly contribute to this effect of Firemaster. Although the synergistic effects between the components of Firemaster were not intentionally studied it would appear that such effects did not significantly contribute to the TCDD-like action of the mixture. The two mixtures of Firemaster were appreciably different with regard to the concentrations of congeners 2, _5_, Q, and Z and therefore other lots of 210 Firemaster or PBB-related environmental contaminants may also contain different amounts of these congeners. Provided that no other TCDD-like agents are present it may be possible to predict the toxicity and MC-like induction activity of such mixtures once the relative concentrations of these four congeners are measured. Since human exposure to PBB in Michigan occurred indirectly through contaminated meat and dairy products it is possible that the composition of consumed PBB could be different from that of the original mixture. The composition of PBB extracted from liver or milk of rats previously exposed to Firemaster was appreciably altered and seemed to have relatively more of the toxic congeners than Firemaster (Dannan e_t a_l., 1978b). The profile of extracted PBB from rats born to and nursed by dams fed PBB containing diets also seemed different than the original mixture (McCormack g a_l., 1981). Interestingly, all four toxic congeners seemed to increase relative to congeners 4 and 2, the two most abundant congeners in Firemaster (McCormack e; al., 1981). In Firemaster, only two congeners (2 and _3_) were metabolized by rat liver microsomes (Dannan e_t gl_., l978b; Moore 2211., 1980). Some other congeners, however, may also undergo i_n_ M metabolism at various but probably much slower rates. Therefore, significant differences may exist in the half lives as well as the i_n_ Mg concentrations of the various PBB congeners. If any or all of the toxic congeners (2, _5_, _6_ and 2) were similarly concentrated in cattle or dairy products then human exposure through the food chain could be more serious than might be predicted from studying the effects of Firemaster. This possibility seems to be supported by the findings of Aulerich and Ringer (1979) who in a long-term feeding study in mink found that PBB derived from contaminated beef and poultry tissues was more toxic than Firemaster FF-l which was experimentally incorporated into beef. GENERAL DISCUSSION A number of relatively important questions have been answered by research described in this thesis. Some of the points may be vitally important to the public and environment of Michigan. Others may be of interest to biochemists who are studying the mechanism(s) by which polyhalogenated aromatic hydrocarbons cause toxicity. Among the most urgent issues that needed to be resolved was the identification of the components that would account for the 2,3,7,8-tetrachloro- dibenzo-p-dioxin (TCDD)-like pharmacotoxicological effects of Firemaster. Most of the biological studies have thus far been performed with crude mixtures of PBB that are similar to that which mistakenly contaminated Michigan's environment. Those studies have generally shown that many of the biochemical and toxicological effects of TCDD can be reproduced by sufficiently large doses of Firemaster. As was discussed in the Literature Review, TCDD-like morphological and functional changes have been detected in the liver and lymphatic organs of animals administered Firemaster. Some of the hepatic changes included porphyria, proliferation and dilation of the SER, and increased formation of lipid droplets which occassionally were surrounded by concentric arrays of laminated ER membranes, commonly known as whorls or myelin figures (Sleight and Sanger, 1976; Gupta and Moore, 1979; Gupta e2 a_l., 1981). Even more severe changes such as necrosis, fibrosis, and neoplastic nodules were reported ten months after a single large oral dose (1 g/kg) of the PBB mixture (Kimbrough gt a}, 1977, 1978). Firemaster also evokes a number of biochemical responses in the liver including the 211 212 induction of microsomal drug metabolizing enzymes that are normally induced by TCDD and the polycyclic aromatic carcinogens such as MC. In addition, Firemaster induces the enzymes that are typically induced by barbiturates such as PB, hence the classification of Firemaster as a mixed-type inducer (Troisi, 1975; Dent _e_ga_l., 1976a,b). Firemaster also causes a number of TCDD-like immunotoxic responses including a loss in both the cell-mediated immunity and in the humoral immune function (Fraker, 1980; Luster e_t_ gl_., 1978, 1980). The TCDD-wasting syndrome and thymus atrophy can also be produced by relatively large doses of Firemaster (Sleight and Sanger, 1976; Gupta and Moore, 1979; Ringer, 1978; Fraker, 1980). In almost all cases the amount of Firemaster required to cause the typical TCDD-toxic responses appears to be several orders of magnitude larger than that of TCDD. It is becoming increasingly accepted that the mechanism by which TCDD (and similar compounds) cause these responses involves a liver cytosolic receptor protein where a pleiotrophic response results from translocation of the receptor-ligand complex to the nucleus and initiation of specific m-RNA and protein synthesis. There appears to be good correlations between the binding to the TCDD-receptor, the induction of enzymes usually induced by MC (or TCDD), and the TCDD-like toxic response (Poland and Glover, 1977, 1980; Poland e_t_ gl_., 1979; Greenlee and Poland, 1979). Planarity and dimensions of the ligand (fitting a rectangle of nearly 3 x 10 2) seem to be among the most important structural features that determine the ligand's affinity to the receptor and, therefore, the extent of biological responses. Rigidly planar compounds such as TCDD are the most active agonists, while certain halogenated biphenyls such as 3,4,5,3',4',5'- hexa-chloro (bromo) biphenyl are at least 100 times less potent (Poland and Glover, 1977; Poland g a_l., 1979). The polyhalogenated biphenyls, particularly the polychlorinated biphenyls (PCB), have generally been classified into one of two groups depending on their 213 type of microsomal enzymes induction and their biological effects (Goldstein e_t_ a_l., 1977; Goldstein, 1979; Poland and Glover, 1977; Poland gt gt, 1979). The first group, includes few congeners, one of which is 3,4,S,3',4',S'-hexahalobiphenyl, and these congeners bind to the TCDD-receptor, evoke a strict MC-type induction, and cause a TCDD-like toxicity. Congeners of the second group, which may be represented by 2,4,5,2‘,4',5'-hexahalobiphenyl, are not bound by the TCDD- receptor, are not toxic, and cause only a strict PB-type induction. The presence of halogens at carbons 0_I‘th_0 to the biphenyl bridge was considered responsible for the wide differences in biological response to halogenated biphenyls. Congeners of the first group, though not rigidly planar like TCDD, were considered capable of assuming planarity since free rotation around the biphenyl axis is rather unrestricted. Halogenation at one or more 9mg carbons was expected to sterically hinder rotation, and consequently result in a twisted halogenated biphenyl which would not be bound by the TCDD-receptor, and, therefore, would not evoke any TCDD-like effects. Chemical characterization of most of the PBB congeners (comprising nearly 98% of Firemaster) indicated that each congener should have at least one gttflg bromine (Sundstrom g_t a_l., 1976a; Jacobs gt gt, 1976; Moore g _a_l., 1980; and Chapter 1). According to the structure-activity relationship rules (Goldstein gt gt, 1977; Goldstein, 1979) all these congeners would have to belong to the second group of compounds, and therefore should cause a strict PB-type induction but should not evoke any TCDD-like effects. Furthermore, rigorous chemical analyses of Firemaster in this laboratory and elsewhere (Hass gt a_l., 1978; DeKok g g_l., 1977) have turned out no evidence for the presence of any highly potent TCDD-like impurities such as bromodibenzo-g—dioxins, bromodibenzofurans, or 3,4,5,3',4',5'- hexabromobiphenyl. Trace impurities of similar chlorinated compounds are present in certain crude mixtures of PCB, and are considered responsible for some of the 214 TCDD-like effects of those mixture (Hutzinger fl g_l., 1974). Several studies have been conducted to characterize the TCDD-like agents by fractionating Firemaster and testing the biological effects of several resulting crude fractions, however, the TCDD-like toxins could not be identified (Kimbrough gta_l., 1977; Safe gta_l., 1978; Hass g_t gl_., 1978; Moore _t_ gt, 1980; Robertson, fl gt, 1981b). This thesis has provided direct answers to questions regarding the identity of the TCDD-like chemicals of Firemaster. Four PBB congeners comprising nearly 15% and 25% of all PBB in Firemaster FF-l and Firemaster BP-6, respectively, evoke an MC-type induction and one or more morphological and functional changes that are indicative of a TCDD-like toxic response. However, each of these congeners has at least one g_r_t_l'lg bromine. Three are mono-g_rttlg brominated, and bromines in these congeners are arranged on the following carbons; 2,4,5,3',4‘ (congener 2), 2,4,5,3',4',5' (congener g), and 2,3,4,5,3',4‘ (congener _7_). The most noticeable common feature among these congeners is that they are derived by adding one g_th_o bromine to 3,4,3',4'-tetrabromobiphenyl (to yield congener 2) and to 3,4,5,3',4'-pentabromobiphenyl (to give congeners _6_, and Z). 3,4,3',4'-Tetra- and 3,4,5,3',4'-penta-bromobiphenyl are, like 3,4,5,3',4',5'-hexabromobiphenyl (HBB), supposed to be the most potent TCDD-like congeners of all possible 209 structures since their chlorinated analogues are also the most potent of all PCB congeners (Goldstein g_t fl” 1977; Yoshimura gt gl_., 1979). However, they are expected to be nearly IOU-fold less potent than TCDD (Poland and Glover, 1977; Poland _e_t 5‘!» 1979; and Table 15). Their mono-g_tflg brominated derivatives (congeners 2, g, and 2) still appear to be nearly 10-20 times less potent (Tables 12 and 15). Therefore, the introduction of a single gttig bromine into the highly TCDD-like toxic congeners weakens but does not abolish their toxicity. According to their TCDD- like potency these congeners appear to rank as follows; congener Z > congener 2 > congener é. Interestingly these congeners appeared to acquire a variable degree of 215 PB-type activity with congener _6_ > congener 2 > congener 2, almost in the opposite order of their MC-type potency (congener 2 seems to be a very subtle PB-type inducer). Congener 2 (2,3,4,2’,4',5'-hexabromobiphenyl), which has two m bromines, was the fourth congener to exhibit an MC-type induction properties. It also seemed to evoke a nearly equivalent PB-type induction, and therefore was classified a mixed-type inducer. This congener seemed to match congener g in its effectiveness both as an MC-type and as a PB-type inducer. Besides the induction of microsomal enzymes and activities that are typically induced by MC (or TCDD), these congeners also seemed to intensify the two microsomal proteins (bands III and IV whose molecular weights are 53,000 and 55,000 daltons, respectively) which apparently are the ones known to be induced by MC or TCDD (Atlas gt gl_., 1977; Guenthner and Nebert, 1978). Band 111 may be that of cytochrome P-448 which is associated with acetanilide-4-hydroxylase activity, while band IV could be due to cytochrome Pl-450 which is linked to AHH activity, commonly measured by the rate of benzo(a)pyrene hydroxylation (Atlas gt g_t, 1977; Guenthner and Nebert, 1978). It is of interest to note that the intensity of protein band 111 seems to correlate very well with the extent of MC-type induction and toxicity (more so than band IV) (Figures 16, 21, and 22). Therefore, it may be necessary to measure the extent of induction of the acetanilide-4- hydroxylase activity, especially if it proves to be associated with protein band III and if it proves to be a more sensitive measure of the extent of MC (TCDD)-type induction than benzo(a)pyrene hydroxylation activity. The extent of AHH induction as measured by benzo(a)pyrene hydroxylation (Gielen g_t g_l., 1972) or by ethoxyresorufin-O-deethylation (Burke and Mayer, 1974) (data not shown) seemed to plateau or level off such that it may become almost impossible to distinguish which of two MC-type inducers, for example, is the more potent. Other parameters or activities that are also affected by MC or TCDD, such as the EtNC 216 difference spectral 455:430 nm ratio, may, however, be a more accurate measure of the extent of MC-type induction. (For example, in Table 12 compare the EtNC difference spectral ratio and benzo(a)pyrene hydroxylase activity of microsomes from rats treated with 2 or 30 mg HBB/kg). Similarly, while the intensity of band IV is induced by all MC-type inducers to a slightly variable extent, it does not seem to reflect the extent of MC-type induction. In conclusion, protein band III of SDS- polyacrylamide gel electrophoresed microsomes seems to be the most sensitive of all proteins to MC (TCDD)-type induction, and, therefore, it may be important to examine its activity which may be that of acetanilide-4-hydroxylase (Atlas _e_t _a_l_., 1977; Guenthner and Nebert, 1978). Since congeners 2, 2, g and _7_ (see Table 4 for their concentration in Firemaster FF-l and Firemaster BP-6) seemed to constitute nearly 15% and 25% of Firemaster FF-l and Firemaster BP-6, respectively, it would appear that their additive effects are responsible for most if not all of the MC (TCDD)-like induction and toxicity of Firemaster. This is supported by the fact that each by itself (at 90. mg/kg) seemed to evoke a stronger MC-type induction response than Firemaster FF-l (also at 90 mg/kg). In order to prove that these congeners are responsible for the effects of crude Firemaster mixtures, the ideal experiment would be to test the biological effects of a reconstituted mixture formulated by mixing pure PBB congeners to resemble their composition in Firemaster. This experiment has almost conclusively'proven that congeners 2, 2, g and _7_ are mainly responsible for the MC-type induction effects of Firemaster (Table 16). The reconstituted mixture which was formulated to resemble Firemaster BP-6 (Table 4) caused an almost identical MC-type induction response to that of Firemaster BP-6; while Firemaster FF-l which had nearly 65% less of congeners 2, 2, g, and 2 was a significantly less potent MC-type inducer than the first two mixtures. 217 Most of the PB-type inducing properties of Firemaster have previously (Moore gt_a_l., 1978b, 1979) been accounted for by congeners {t and 2 which together constitute at least 75% and 65% of Firemaster FF-l and Firemaster BP-6, respectively. Congeners 2, 2, and 22 comprising nearly 5-7% of both mixtures should also contribute to the PB-type induction effects. Even though the slight MC-type induction due to congener _9_ could have resulted from nearly 3% contamination by congeners 2 and 2 it is also possible that congener 2, by itself, may cause a very slight MC-type induction. Congener 2 which can be metabolized (Dannan _e_t gl_., 1978b) did not seem to cause hypertrophy of the liver even though it caused a PB-type induction. All other congeners don't seem to be significantly metabolized (Dannan gt _a_l., 1978b) and all, regardless of the type of induction, were capable of causing liver enlargement. Congener 2 which constitutes nearly 10% of either mixture, is a mixed-type inducer and therefore should also contribute to the PB-type induction effects of Firemaster. Congeners 2, 2g and 22 total nearly 2% of FF-l or BP-6, but since each should have three o_rtllg bromines they are believed to cause a strict PB-type induction. Congener 2, however, can also be metabolized (Dannan g_t _a_l., 1978b; Moore e_t gt, 1980), and therefore may not cause hypertrophy of the liver. Including congeners 63 and 6_b (both of which are believed to be peculiar to Firemaster FF-l) and some other minor GC peaks, nearly 2% or less of Firemaster (FF-l or BP-6) remains unaccounted for by structure or biological effects. Based on the results of the reconstituted PBB study, however, they should not significantly contribute to the toxicity of the mixture. The nearly 200 ppm of penta— and hexa- brominated naphthalenes, that were detected in Firemaster BP-6, may only slightly contribute to the toxicity and MC-type induction effects of Firemaster (Hass gt gt, 1978; Goldstein gt 3:12., 1979). Since dose-response studies were not performed it is hard to accurately arrange the congeners in the order of their MC-type induction potencies (i.e. 218 according to their EDSU’ the half maximal effective dose). However, as mentioned earlier, the four congeners of Firemaster may be ranked according to their MC- type induction potency as follows; congener Z > congener 2 > congener §_>_ congener 2. Of these compounds, congener 2 was the only congener to cause a severe atrophy of the thymus and a dramatic decline in body weight gain. To a lesser extent congener 2 and congener _6 (data not shown) seemed to slightly reduce the weight of the thymus. One function of the immune system (the humoral immunity) was also affected by congener 2 (both the T-helper and B-cells were severly impaired). Firemaster FF-l also impaired the humoral immune function, even though significantly less so than congener 2. Unfortunately, similar studies were not performed with the other congeners. The humoral immune system appears to be rather sensitive to toxicity, and therefore may be utilized to compare the toxicity of PBB congeners. Certain hepatic histopathological and ultrastructural changes may also be quite indicative of the toxic PBB congeners. Among these changes are the appearance of prominent cellular margins (individualized hepatocytes), increased lipid vacuolation, disorganization in the normal stacking pattern of RER, and formation of concentric arrays of laminated ER membranes around lipid droplets (myelin figures, or whorls). Among the least sensitive measures of acute PBB toxicity may be changes in the bone marrow function and in the blood clinical parameters. Hematologic changes may be produced, however, by toxic doses of TCDD particularly in chronic studies (Zinkl gt _a_l_., 1973; McConnell and Moore, 1979). Finally, it should be mentioned that these attempts to study the relationship between the structure of a PBB congener and one form of toxicity or another were rather exploratory in nature since background information on the toxicology of polyhalogenated biphenyl congeners has been rather scarce or non-existent. Attention has only very recently been given to studying the toxicological effects of pure polyhalogenated biphenyl congeners and 219 determining the relationships between structure and toxicity (Biocca g _a_l., 1981; Gupta fl gt, 1981). While such studies should be of significance to clinicians, toxicologists, and environmentalists, more research should be devoted to the biochemical mechanism of toxicity by TCDD and similarly planar halogenated aromatic hydrocarbons including PBB. For many of the morphological and functional changes may be linked by a single mechanism. The hypothesis that TCDD and similar agonists evoke their toxic effects after binding to the hepatic TCDD-cytosolic receptor and translocation to the nucleus is worthy of further investigation (Poland gt gl_., 1979; Greenlee and Poland, 1979). The pleiotrophic gene expression and possibly repression (notice in Table 12 that HBB caused nearly 75% reduction in one of the mixed-function oxidase activities) may be responsible for all of the typical signs and symptoms of toxicity. Ligand-receptor binding followed by translocation to the nucleus and eventually gene expression are not foreign concepts to biologists investigating the mechanism of action of steroids and similar hormones. However, while hormone stimulation of gene expression may normally be under tight regulation in terms of extent and duration, the gene expression effects of the TCDD-receptor complex may be more persistent or permanent in nature. One could envision that the toxic effects of TCDD and similar agonists are caused by their sustained occupancy of a biological receptor which normally is involved in binding a hormone (H) and regulating one or more key enzymes. In effect, the TCDD-receptor complex may interact with the same nuclear acceptor sites with which the H-receptor complex normally interacts. However, it is conceivable that the TCDD-receptor and H-receptor complexes have vastly different pharmacokinetic properties in terms of their interaction with the nuclear acceptor sites. (The affinity with which TCDD is bound by the cytosolic receptor is in the same range of affinities which steroid hormones have for their receptors (O'key gta_l., 1979)). Compared to the H-receptor complex, the 220 TCDD-receptor could have a more permanent binding affinity to the nuclear sites, which would result in a more sustained gene expression than normal. Such a lasting response (for instance enzyme induction) could eventually lead to toxicity. Alternatively, this toxin may be extremely stable against metabolism and elimination, or it may even be stabilized against metabolism by being bound to the receptor. Under both of these circumstances the receptor would be depleted in the cytosol but retained in the nucleus and toxicity would become manifest after a prolonged biochemical response. These hypothetical events are summarized in the scheme shown below. Cytoplasmic Receptor Level Nuclear Receptor Level """""" ---..-..--.. B3671?" Biochemical Response Time 221 If the hormone (or any endogenous chemical) can be identified this hypothesis can then be tested by performing proper kinetic studies, 23 2.1.19 and lg m, using radiolabelled TCDD and radiolabelled hormone. (It is rather important to note that the natural ligand for the TCDD-receptor, has not yet been identified (O'key gt a_l., 1979)). Therefore, while more than one regulatory mechanism (such as hormone metabolism and elimination) may be able to regulate and control the induction effects of hormones to eventually result in "deinduction", TCDD and similar agonists may cause irreversible induction effects. A modified version of the above cited model would involve the interaction of the TCDD-receptor complex with nuclear acceptor sites other than those with which the H-receptor complex normally interacts. Such interaction, whether prolonged or transient in nature, may result in expressing (or repressing) one or a number of genes, thereby resulting in toxicity. It is commonly believed that more than one type of nuclear acceptor site interacts with estrogen-receptor complexes where different uterine responses are regulated by different nuclear acceptors (Mukku gt a_l., 1981). It is interesting to note that wide variations in susceptibility to TCDD do exist between species (for example, the guinea pig is nearly 1000-fold more sensitive to TCDD than the hamster) and are also present within the same species (D2 mice are at least 10 times more resistant to TCDD than B6 mice). Such variations could be due to mutation in the biological receptor(s) leading to a decreased affinity for the ligand and/or a reduction in receptor(s) binding sites (Poland e_t_ a_l., 1974; Poland and Glover, 1975). However, it is also possible that these variations are due to differences in the rates of metabolism and elimination of those toxins by different animal species. All these considerations are worthy of further investigations since eventually they may contribute to fully explaining the mechanism of TCDD- toxicity. 222 O'key e_t g2. (1979) have recently suggested that the presence of the cytosolic receptors for TCDD and similar foreign polycyclic aromatic compounds may have originally evolved from mutation in the genes that code for steroid or glucocorticoid receptors. The later endogenous compounds, through their binding receptors, are thought to regulate cytochrome P-450 mediated monooxygenases that are involved in the metabolic biosynthesis and degradation of steroids. A mutation in those natural receptors may have resulted in cytosolic receptor(s) that can bind with high affinity to hydrophobic compounds including TCDD, MC, and benzo(a)pyrene. A pleiotrophic response then results in inducing certain forms of cytochrome P-450 that in turn detoxify these ingested or inhaled noxious compounds (O'key g_t a_l., 1979). From the studies involving the induction and toxic effects of PBB congeners it has become obvious that the. original structure-activity relationship rules (Goldstein g_t g_l., 1977; Goldstein, 1979) are not universally valid for all 209 possible polyhalogenated biphenyl congeners. Recent'studies have also shown that several mono-gthg chlorinated congeners, as well as certain di-9_r_tflg chlorinated congeners with a 2,3,4-trichlorination pattern on one or both rings can cause an MC-type induction (Parkinson e_t_ _a_l., 1980a,b,c, 1981; Yoshihara _et_a_l_., 1979). Most of those PCB congeners were classified as mixed-type inducers since they also evoked a PB-type induction response. All results including those presented in this thesis agree that the structure-activity relationships for toxicity and induction of the mixed-function oxidase enzymes are not as simple as originally predicted by Goldstein and coworkers (Goldstein _e_t _a_l_., 1977; Goldstein, 1979). Epoxide hydrolase induction seemed even harder to correlate with the structure of PBB congeners or their induction of the mixed-function oxidase enzymes. Some of the PBB congeners that were weak or subtle PB—type inducers (such as congeners 2 and 2) caused a marked induction of epoxide hydrolase as 223 evidenced by its i_n in}; ability to metabolize 3l—l-styrene oxide and by the increased intensity of the protein band that has an identical electrophoretic mobility to purified epoxide hydrolase. Either of these two congeners was even a better inducer of this activity than congener 2, a strict PB-type inducer. HBB, a strict MC-type inducer, had no effect on epoxide hydrolase. Several PCB congeners that can cause a PB-type induction were able to induce this activity while 3,4,5,3',4',5'-hexachlorobiphenyl (the chlorinated analogue of HBB), a strict MC-type inducer, was without effect (Kohli gt £12., 1979; Ahotupa, 1981). However, 2,4,6,2',4',6'-hexachlorobiphenyl, which is only a weak PB-type inducer (Goldstein gt g2, 1977), induced epoxide hydrolase to an extent similar to that of PB or other potent PBétype inducers such as 2,4,5,2',4',5'-hexachlorobiphenyl (Ahotupa, 1981). Therefore, epoxide hydrolase may not be coordinately expressed with the induction of the mixed-function oxidase enzymes by PB-like compounds. More genetic and biochemical research will be needed to determine if any of the genes of the mixed- function oxidase enzymes are coordinately regulated with the gene for epoxide hydrolase. Also more studies will be needed to determine the structural requirements for the induction of these and other microsomal enzymes. While certain chemical features of PBB congeners appear similar their chemical properties are by no means identical. Differences in their chemical properties should be responsible for the differences in their biological effects. This seems to be supported by the fact that correlations seem to exist between their chemical and biological properties. For instance, the mono-gm brominated congeners (2, g and Z) seemed to have intermediate UV-absorption spectral characteristics and chromatographic properties between the di-o_rt_h_o brominated compounds such as congener 4_ and the 922m unsubstituted ones exemplified by HBB. The crystal studies indicate that HBB is not absolutely planar; however, its dihedral angle (SO-520) is significantly lower than that (85-860) of the di- and tri- 224 giflg brominated congener 3 and 2,5,2',6'-tetrabromobiphenyl, respectively. It is feasible that the dihedral angles of the mono-95th}; brominated congeners, like other chemical and biological characteristics, are also intermediate. The UV- absorption spectra unequivocally indicated that a fair degree of conjugation is allowed between the two benzene rings of the mono-gttgg brominated congeners. Clearly, therefore, the two rings are not fixed in a rigidly twisted conformation. The findings of this research do raise serious questions regarding the validity of using the major congener (g) as a measure of the total PBB concentration in environmental or biological studies (Fries gt g_l., 1978; Rickert gt _a_l., 1978; Wolff gt g2., l978a,b; Corbett gt gt, 1978; Howard _e_tgt., 1980; McCormack gtg_l., 1980). Most of these studies have attempted to correlate certain biological responses or TCDD-like toxicity parameters with the concentration of congener ft only. These correlations may not be meaningful unless the concentrations of the other congeners, particularly 2, 2, _6_ and _7_ that are the most similar to TCDD, are also measured. REFERENCES Ahotupa, M. (1981). Biochem. Pharmacol. fl, 1866-1869. Aitio, A. (1978). Conjugation Reactions in Drug Biotransformation, Proceedings of a Symposium held in Turku, Finland in 1978, Elsevier, Amsterdam. Allen, J.R., Lambrecht, L.K., and Barsotti, D.A. (1978). J. Am. Vet. Med. Assoc. 222,1485-1489. Alvares, A.P., Bickers, D.R., and Kappas, A. (1973). Proc. Natl. Acad. Sci. U.S.A. 19,1321-1325. Alvares, A.P., and Kappas, A. (1977). J. Biol. Chem. 2_5_2_, 6373-6378. Atlas, S.A., and Nebert, D.W. (1976). Arch. Biochem. Biophys. 23, 495-506. Atlas, S.A., Boobis, A.R., Felton, J.S., Thorgeirsson, 5.8., and Nebert, D.W. (1977). J. Biol. Chem. 222, 4712-4721. Aulerich, R.J., and Ringer,- R.K. (1979). Arch. Environ. Contam. Toxicol. 2, 487- 498. Babish, J.G., Gutenmann, W.H., and Stoewsand, G.S. (1975). J. Agric. Food Chem. 22,879-882. Babish, J.G., Stoewsand, G.S., and Lisk, D.J. (1976). Fed. Proc. 22, 376. Bairstow, F., Hsia, M.T., Norback, D.H., and Allen, JR. (1978). Environ. Health Perspect. _2_3, 321-327. Bastiansen, O., Seip, H.M., and Boggs, J.E. (1971). In Perspectives in Structural Chemistry (J.D. Dunitz, and J.A. Ibers, eds.), Vol. IV, pp. 110-115. Beaven, G.H., James, A.T., and Johnson, E.A. (1957). Nature 27_9, 490-491. Beaven, G.H. (1958). Steric Effects in Conjugated Systems (G.W. Gray, ed.), Academic Press, New York. Billings, R.E., and McMahon, R.E. (1978). Molec. Pharmacol. 1_4, 145-154. Biocca, M., Gupta, N., Chae, K., McKinney, J.D., and Moore, J.A. (1981). Toxicol. Appl. Pharmacol. 3, 461-474. Bock, K.W., Frohling, W., Remmer, H., and Rexner, B. (1973). Biochem. Biophys. Acta 22_7_, 46-56. 225 226 Breckenridge, A. (1975). In Enzyme Induction (D.V. Parke, ed.), pp. 273-301, Plenum Press, New York. Bresnick, E., Mukhtar, H., Stoming, T.A., Dansette, P.M., and Jerina, D.M. (1977). Biochem. Pharmacol. 26, 891-892. Brinkman, U.A.Th., and DeKok, A. (1980). In Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products (R.D. Kimbrough, ed.), pp. 1-40, Elsevier, Amsterdam. Buening, M.K., Wislocki, P.G., Levin, W., Yagi, H., Thakker, D.R., Akagi, H., . Koreeda, M., Jerina, D.M., and Conney, A.H. (1978). Proc. Natl. Acad. Sci. U.S.A. _7_5_, 5358-5361. Burke, M.D., and Mayer, R.T. (1974). Drug Metab. Disp. 2, 583-588. Burke, M.D., Prough, R.A., and Mayer, R.T. (1977). Drug Metab. Disp. 2, 1-8. Carter, L.J. (1976). Science fl, 240-243. Cecil, H.C'., Harris, S.J., and Bitman, J. (1975). Arch. Environ. Contam. Toxicol. _3_, 183-192. Charbonneau, G.P., and Delugeard, Y. (1976). Acta Crystallogr. B32, 1420-1423. Claude, A. (1969). In Microsomes and Drug Oxidations (J.R. Gillette, ed.), pp. 3-39, Academic Press, New York. Conney, A.H. (1967). Pharmacol. Rev. 22, 317-366. Conney, A.H., and Burns, J.J. (1972). Science 222, 576-586. Coon, M.J., Ballou, D.P., Haugen, D.A., Krezoski, S.O., Nordblom, G.D., and White, R.E. (1977). In Microsomes and Drug Oxidations (J.R. Gillette, ed.), pp. 82- 94, Academic Press, New York. Corbett, T.H., Beaudoin, A.R., Cornell, R.G., Anver, M.R., Schumacher, R., Endres, J., and Szwabowska, M. (1975). Environ. Res. B, 390-396. Corbett, T.H., Simmons, J.L., Kawanishi, H., and Endres, J.L. (1978). Environ. Health Perspect. 2_3, 275-281. Creaven, P.J., and Parke, D.V. (1966). Biochem. Pharmacol. _1_2, 7-16. Dannan, G.A. (1978). M.S. Thesis, Michigan State University, East Lansing, Michigan. Dannan, G.A., Moore, R.W., Besaw, L.C., and Aust, S.D. (l978a). Biochem. Biophys. Res. Commun. 22, 450-458. Dannan, G.A., Moore, R.W., and Aust, S.D. (1978b). Environ. Health Perspect. Q, 51-61. 227 Dannan, G.A., and Aust, SD. (1980). Abstracts of Papers for the Nineteenth Annual Meeting of the Society of Toxicology, Washington, D.C., 261. Dannan, G.A., Sleight, S.D., and Aust, SD. (1980). The Pharmacologist _22, 158. Dehlinger, P.J., and Schimke, R.T. (1972). J. Biol. Chem. _2_4_7_, 1257-1264. Deitrich, R.A., Bludeau, P., Roper, M., and Schmuck, J. (1978). Biochem. Pharmacol. 22, 2343-2347. DeKok, J.J., DeKok, A., Brinkman, U.A.Th., and Kok, R.M. (1977). J. Chromatog. _14_2, 367-383. Dent, J.G., Netter, K.J., and Gibson, J.E. (1976a). Res. Comm. Chem. Pathol. Pharmacol. 22, 75-82. Dent, J.G., Netter, K.J., and Gibson, J.E. (1976b). Toxicol. Appl. Pharmacol. 28, 237-249. Dent, J.G., Cagen, S.Z., McCormack, K.M., Rickert, D.E., and Gibson, J.E. (1977). Life Sciences 29, 2075-2080. Dent, J.G., Elcombe, C.R., Netter, K.J., and Gibson, J.E. (1978). Drug Metab. Disp. _6_, 96-101. - DeVos, R.H., and Peet, E.W. (1971). Bull. Environ. Contamin. Toxicol. g, 164-170. DiCarlo, F.J., Seifter, J., and DeCarlo, V.J. (1978). Environ. Health Perspect. _22, 351-365. Dutton, 5.0. (1975). Biochem. Pharmacol. 22, 1835-1841. ' Ecobichon, D.J., and Comeau, A.M. (1975). Toxicol. Appl. Pharmacol. 2, 94-105. Eliel, E.L. (1962). Stereochemistry of Carbon Compounds, pp. 156-179, McGraw Hill Book Co., Inc., New York. El-Tantawy, M.A., and Hammock, B.D. (1980). Mutation Res. _7_9, 59-71. Farber, T.M., Balazs, T., Marks, E., and Cerra, F. (1976). Fed. Proc. _32, 376. Farber, T., Karza, L., Giovetti, A., Carter, C., Earl, F., and Balazs, T. (1978). Toxicol. Appl. Pharmacol. 22, 343. Finar, LL. (1975). Organic Chemistyr, Vol. 2, Stereochemistry and the Chemistry of Natural Products 5th Edition, pp. 215-233, Longman Group Limited, London. Fraker, P.J., Hass, S.M., and Luecke, R.W. (1977). J. Nutr. £2, 1889-1895. Fraker, P.J. (1980). Toxicol. Appl. Pharmacol. 22, 1-7. Fries, G.F., Marrow, G.S., and Cook, R.M. (1978). Environ. Health Perspect. 23_, 43-50. 228 Friel, J.P. (1974). Dorland's Illustrated Medical Dictionary, 25th Edition, pp. 1623, W.B. Saunders, Philadelphia. Gasiewicz, T.A., Holscher, M.A., and Neal, R.A. (1980). Toxicol. Appl. Pharmacol. _52, 469-488. Gielen, J.E., Goujog, F.M., and Nebert, D.W. (1972). J. Biol. Chem. E, 1125- 1137. Gillette, J.R. (1966). Adv. in Pharmacol. 3, 219-261. Gillette, J.R., Davis, D.C., and Sasame, H.A. (1972). Ann. Rev. Pharmacol. 22, 57- 84. Gillette, J.R., Mitchell, J.R., and Brodie, BB. (1974). Ann. Rev. Pharmacol. 23, 271-288. Goldstein, J.A., McKinney, J.D., Lucier, G.W., Hickman, P., Bergman, H., and Moore, J.A. (1976). Toxicol. Appl. Pharmacol. _3_6_, 81-92. Goldstein, J.A., Hickman, P., Bergman, H., McKinney, J.D., and Walker, M.P. (1977). Chem. Biol. Interact. _1_7_, 69-87. Goldstein, J.A., Hass, J.R., Linko, L., and Harvan, D.J. (1978). Drug Metab. Disp. 13., 258-264. Goldstein, J.A. (1979). In Health Effects of Halogenated Aromatic Hydrocarbons (W.J. Nicholson, and J.A. Moore, Eds.), Ann. N.Y. Acad. Sci. 22_0_, pp. 164-178, The New York Academy of Sciences, New York. Goldstein, J.A., Linko, P.C., Levy, L.A., McKinney, J.D., Gupta, B.N., and Moore, J.A. (1979). Biochem. Pharmacol. 2_8_, 2947-2956. Gozukara, E.M., Belvedere, G., Robinson, R.C., Deutsch, J., Coon, M.J., Guengerich, F.P., and Gelboin, H.V. (1981). Molec. Pharmacol. _1_2, 153-161. Greenlee, W.F., and Poland, A. (1979). J. Biol. Chem. 2_5_4_, 9814-9821. Grunberger, D., and Weinstein, B. (1979). Prog. Nucleic Acid Res. Molec. Biol. 2_3_, 105-149. Guengerich, F.P. (1977). J. Biol. Chem. 2_52, 3970-3979. Guengerich, F.P. (1979). Pharmacol. Ther. _Até, 99-121. Guengerich, F.P., and Martin, M.V. (1980). Arch. Biochem. Biophys. 205, 365-379. Guenthner, T.M., and Nebert, D.W. (1978). Eur. J. Biochem. 22, 449-456. Gupta, B.N., and Moore, J.A. (1979). Am. J. Vet. Res. 422, 1458-1468. Gupta, B.N., McConnell, E.E., Harris, M.W., and Moore, J.A. (1981). Toxicol. Appl. Pharmacol. 27_, 99-118. 229 Hass, J.R., McConnell, E.E., and Harvan, D.J. (1978). J. Agric. Food Chem. _2_6, 94- 99. Haugen, D.A., Coon, M.J., and Nebert, D.W. (1976). J. Biol. Chem. 2_52, 1817-1827. Heidelberger, C. (1975). Ann. Rev. Biochem. fl, 79-121. Hill, D.L., Shih, T.W., Johnston, T.P., and Struck, R.F. (1978). Cancer Res. _32, 2438-2442. Howard, S.K., Werner, P.R., and Sleight, SD. (1980). Toxicol. Appl. Pharmacol. 22, 146-153. Huckins, J.N., Stalling, D.L., and Petty, JD. (1980). J. Assoc. Off. Anal. Chem. _6_3_, 750-755. Hutzinger, 0., Safe, 8., and Zitko, V. (1974). The Chemistry of PCB's, CRC Press, Cleveland, Ohio. Ito, N., Nagasaki, H., Arai, M., Makiura, S., Sugihara, S., and Hirao, K. (1973). J. Natl.’ Cancer Inst. 22, 1637-1646. Ivanetich, K.M., and Bradshaw, J.J. (1977). Biochem. Biophys. Res. Commun. E, 317-322. ' Iyanagi, T., and Mason, H.S. (1973). Biochemistry 22, 2297-2308. Jackson, H.L., and McKusick, BC. (1963). In Organic Syntheses (N. Rabjohn, ed.), Vol. 4, pp. 438-440, John Wiley and Sons, Inc., New York. ' Jacobs, L.W., Chou, S.F., and Tiedje, J.M. (1976). J. Agric. Food Chem. 23, 1198- 1201. Jerina, D.M., and Daly, J.W. (1974). Science 222, 573-582. Kaplowitz, N., Kuhlenkamp, J., and Clifton, G. (1975). Biochem. J. 1_46, 351-356. Kappas, A., and Alvares, AP. (1975). Sci. Am. 23_2, No. 6, 22-31. Kateley, J.R., and Bazzell, S.J. (1978). Environ. Health Perspect. 22, 75-82. Kennedy, M.W., Carpentier, N.K., Dymerski, P.P., Adams, S.M., and Kaminsky, L.S. (1980). Biochem. Pharmacol. 22, 727-736. Kennedy, M.W., Carpentier, N.K., Dymerski, P.P., and Kaminsky, L.S. (1981). Biochem. Pharmacol. 2, 577-588. Kimbrough, R. (1974). CRC Crit. Rev. Toxicol. 2, 445-498. Kimbrough, R.D., Squire, R.A., Linder, R.E., Standberg, J.D., Montali, R.J., and Burse, V.W. (1975). J. Natl. Cancer Inst. 22, 1453-1459. Kimbrough, R.D., Burse, V.M., Liddle, J.A., and Fries, G.F. (1977). The Lancet 11, 602-603. 230 Kimbrough, R.D., Burse, V.W., and Liddle, J.A. (1978). Environ. Health Perspect. Q, 265-273. Kimbrough, R.D., Groce, D.F., Korver, M.P., and Burse, V.W. (1981). J. Natl. Cancer Inst. fl, 535-542. Klaassen, CD. (1970). J. Pharmacol. Expt. Therap. 2_7_5, 289-300. Kohli, K.K., Mukhtar, H., Bend, J.R., Albro, P.W., and McKinney, JD. (1979). Biochem. Pharmacol. 22, 1444-1446. Kuntzman, R. (1969). Ann. Rev. Pharmacol. 2, 21-36. Lambrecht, L.K., Barsotti, D.A., and Allen, JR. (1978). Environ. Health Perspect. _2_3, 139-145. Lind, C., and Ernster, L. (1974). Biochem. Biophys. Res. Commun. _52, 392-399. Lowry, D.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J. (1951). J. Biol. Chem. 222, 265-275. Lu, A.Y.H., Junk, K.W., and Coon, M.J. (1969). J. Biol. Chem. 2%, 3714-3721. Lu, A.Y.H., Levin, W., and Kuntzman, R. (1974). Biochem. Biophys. Res. Commun. _62, 266-272. Lu, A.Y.H. (1979). Drug Metabol. Rev. 222, 187-208. Lu, A.Y.H., and Miwa, G.T. (1980). Ann. Rev. Pharmacol. Toxicol. _2_0, 513-531. Lucier, G.W., Sonawane, B.R., and McDaniel, 0.5. (1977). Drug Metab. Disp. 2, 279-287. Luecke, R.W., Simonel, C.E., and Fraker, P.J. (1978). J. Nutr. 202, 881-887. Luster, M.I., Faith, R.E., and Moore, J.A. (1978). Environ. Health Perspect. 2_3_, 227-232. Luster, M.I., Boorman, G.A., Harris, M.W., and Moore, J.A. (1980). Int. J. Immunopharmacol. 2, 69-80. Lutz, R.J., Dedrick, R.L., Matthews, H.B., Elling, T.E., and Anderson, M.W. (1977). Drug Metab. Disp. 2, 386-396. Mason, H.S. (1957). Science 222, 1185-1188. Matthews, H.B., and Kato, S. (1979). In Health Effects of Halogenated Aromatic Hydrocarbons (W.J. Nicholson, and J.A. Moore, eds.), Ann. N.Y. Acad. Sci. 222, pp. 131-138, The New York Academy of Sciences, New York. McConnell, E.E., and Moore, J.A. (1979). Ibid, pp. 138-150. McConnell, E.E. (1980). In Halogenated BiphenylsL Terphenyls, Naphthalenes, Dibenzodioxins and Related Products (R.D. Kimbrough, ed.), Topics in Environmental Health, Vol. 4, pp. 109-150, Elsevier, Amsterdam. 231 McCormack, K.M., Kluwe, W.M., Rickert, D.E., Sanger, V.L., and Hook, J.B. (1977). Toxicol. Appl. Pharmacol. 42, 539-553. McCormack, K.M., Melrose, P., Rickert, D.E., Dent, J.G., Gibson, J.E., and Hook, J.B. (1979). Toxicol. Appl. Pharmacol. _4_7_, 95-104. McCormack, K.M., Braselton, W.E., Sanger, V.L., and Hook, J.B. (1980). Toxicol. Appl. Pharmacol. 22, 108-115. McCormack, K.M., Lepper, L.F., Wilson, D.M., and Hook, J.B. (1981). Toxicol. Appl. Pharmacol. 22, 300-313. McKinney, J.D., Chae, K., Gupta, B.N., Moore, J.A., and Goldstein, J.A. (1976). Toxicol. Appl. Pharmacol. E, 65-80. McMahon, B.M., and Sawyer, L.D. (eds.). (1971). Pesticide Analytical Manual, Vol. 1, Methods which detect multiple residues, U.S. Department of Public Health, Education, and Welfare, Food and Drug Administration. Mitchell, J.R., Potter, W.Z., Hinson, J.A., Snodgrass, W.R., Timbrell, J.A., and Gillette, JR. (1975). In Handbook of E2perimental Pharmacology, Vol. 28, part 3, pp. 383-419, Springer-Verlag, Heidelberg. Moore, R.W. (1978). Ph.D. Thesis, Michigan State University, East Lansing, Michigan. Moore, R.W., and Aust, SD. (1978). Biochem. Biophys. Res. Commun. 8i, 936-942. Moore, R.W., O'Connor, J.V., and Aust, S.D. (1978a). Bull. Environ. Contam. Toxicol. _zg, 478-483. 6 Moore, R.W., Sleight, S.D., and Aust, S.D. (1978b). Toxicol. Appl. Pharmacol. 242, 309-321. Moore, R.W., Dannan, G.A., and Aust, S.D. (1978c). Environ. Health Perspect. Q, 159-165. Moore, R.W., Sleight, S.D., and Aust, SD. (1979). Toxicol. Appl. Pharmacol. 32, 73-86. Moore, R.W., Dannan, G.A., and Aust, SD. (1980). In Molecular Basis of Environmental Toxicity (R.S. Bhatnager, ed.), pp. 173-212, Ann Arbor Science Publishers, Inc., Ann Arbor, Michigan. Moorhead, P.D., Willett, L.B., and Schanbacher, F.L. (1978). Environ. Health Perspect. Q, 111-118. Mountain, K.R., Hirsh, J., Gallus, A.S. (1970). Lancet2, 265-268. Mukku, V.R., Kirkland, J.L., and Stancel, G.M. (1981). Trends in Pharmacological Sciences 2, 98-101. Nash, T. (1953). Biochem. J. 2, 416-421. 232 Neal, R.A., Beatty, P.W., and Gasiewicz, T.A. (1979). In Health Effects of Halogenated Aromatic Hydrocarbons (W.J. Nicholson, and J.A. Moore, eds.), Ann. N.Y. Acad. Sci. _322, pp. 204-213, The New York Academy of Sciences, New York. Nebert, D.W., and Gielen, J.E. (1972). Fed. Proc. 22, 1315-1325. Newbold, R.F., and Brookes, P. (1976). Nature 262, 52-53. Norback, D.H., and Allen, J.R. (1973). Environ. Health Perspect. 2, 233-240. Nystrom, E., and Sjovall, J. (1975). In Methods in Enzymology, Lipids B (J.M. Lowenstein, ed.), 22, pp. 378-395, Academic Press, New York. Oesch, F., Jerina, D.M., and Daly, J. (1971). Biochem. Biophys. Acta _22_7, 685-691. Oesch, F. (1973). Xenobiotica 2, 305-340. O'Farrell, P.H. (1975). J. Biol. Chem. @, 4007-4021. O'key, A.B., Bondy, G.P., Mason, M.E., Kahl, G.F., Eisen, H.J., Guenthner, T.M., and Nebert, D.W. (1979). J. Biol. Chem. @, 11636-11648. Omura, T., and Sato, R. (1964). J. Biol. Chem. 23_2, 2370-2378. Osborne, MR. (1979). Trends Biochem. Sci., September issue, 213-215. Owens, 1.5., and Nebert, D.W. (1975). Pharmacologist Q, 217. Parke, D.V. (1975). In Enzyme Induction (D.V. Parke, ed.), pp. 207-271, Plenum Press, New York. Parkinson, A., Cockerline, R., and Safe, 8. (1979). Biochem. Pharmacol. Q, 259- 262. Parkinson, A., Robertson, L.W., and Safe, S. (1980a). Life Sci. Q, 2333-2337. Parkinson, A., Cockerline, R., and Safe, S. (1980b). Chem. Biol. Interact. Q, 277- 289. Parkinson, A., Robertson, L., Safe, L., and Safe, S. (1980c). Chem. Biol. Interact. 12, 271-285. Parkinson, A., Robertson, L.W., Safe, L., and Safe, 8. (1981). Chem. Biol. Interact. 15., 1'12. Pederson, T.C., and Aust, SD. (1970). Biochem. Pharmacol. 22, 2221-2230. Poland, A., and Glover, E. (1974). Molec. Pharmacol. _1_q, 349-359. Poland, A.P., Glover, E., Robinson, J.R., and Nebert, D.W. (1974). J. Biol. Chem. 2Q, 5599-5606. Poland, A., and Glover, E. (1975). Molec. Pharmacol. 22, 389-398. 233 Poland, A., Glover, E., and Kende, A.S. (1976a). J. Biol. Chem. 222, 4936-4946. Poland, A., Glover, E., Kende, A.S., DeCamp, M., and Giandomenico, C.M. (1976b). Science 2%, 627-630. Poland, A., and Glover, E. (1977). Molec. Pharmacol. 2_3_, 924-938. Poland, A., Greenlee, W.F., and Kende, A.S. (1979). In Health Effects of Halogenated Aromatic Hydrocarbons (W.J. Nicholson, and J.A. Moore, eds.) Ann. N.Y. Acad. Sci. 222, pp. 214-230, The New York Academy of Sciences, New York. Poland, A., and Glover, E. (1980). Molec. Pharmacol. Q, 86-94. Poland, A., Mak, I., Glover, E., Boatman, R.J., Ebetino, F.H., and Kende, A.S. (1980). Molec. Pharmacol. E, 571-580. Polin, D., Ringer, R.K., and Aust, SD. (1979). Proc. Soc. Expt. Biol. Med. 262, 44- 47. Posner, H.S., Mitoma, C., Rothberg, S., and Udenfriend, S. (1961). Arch. Biochem. Biophys. 23, 280-290. Rannug, U., Sundvall, A., and Ramel, C. (1978). Chem. Biol. Interact. 22, 1-16. Render, J. (1980). M.S. Thesis, Michigan State University, East Lansing, Michigan. Rickert, D.E., Dent, J.G., Cagen, S.Z., McCormack, K.M., Melrose, P., and Gibson, J.E. (1978). Environ. Health Perspect. Q, 43-50. Rickert, D.E., and Fouts, J.R. (1970). Biochem. Pharmacol. 22, 381-390. Ringer, R.K., and Polin, D. (1977). Fed. Proc. fl, 1894-1898. Ringer, R.K. (1978). Environ. Health Perspect. Q, 247-255. Robertson, GB. (1961). Nature 222, 593-594. Robertson, L.W., Parkinson, A., and Safe, S. (1980). Biochem. Biophys. Res. Commun. 9_2, 175-182. Robertson, L.W., Parkinson, A., Bandiera, S., and Safe, S. (1981a). Chem. Biol. Interact. 2, 13-24. Robertson, L.W., Parkinson, A., and Safe, S. (1981b). Toxicol. Appl. Pharmacol. _51, 254-262. Rutter, W.J. (1967). In Methods in Development Biology (F.H. Wilt, and N.K. Wessells, eds.), pp. 671-683, Thomas Y. Crowell Co., New York. Ryan, D.E., Thomas, P.E., Korzeniowski, D., and Levin, W. (1979). J. Biol. Chem. 223, 1365-1380. Safe, S., Kohli, J., and Crawford, A. (1978). Environ. Health Perspect. Q, 147-152. 234 Safe, S. (1980). In Halogenated Biphenyls, TerphenylsL Naphthalenes, Dibenzo- dioxins and Related Products (R.D. Kimbrough, ed.), Topics in Environmental Health, Vol. 4, pp. 81-107, Elsevier, Amsterdam. Sher, S.P. (1971). Toxicol. Appl. Pharmacol. 2_8_, 780-834. Shimada, T., and Sato, R. (1978). Biochem. Pharmacol. 22, 585-593. Sims, P., and Grover, P.L. (1974). Adv. Cancer Res. 2_0_, 165-274. Sissons, D., and Welti, D. (1971). J. Chromatog. ég, 15-32. Sladek, N.E., and Mannering, G.J. (1966). Biochem. Biophys. Res. Commun. 24, 668-674. Sleight, S.D., and Sanger, V.L. (1976). J. Am. Vet. Med. Assoc. _1_62, 1231-1235. Stalling, D.L., and Huckins, J.N. (1973). J. Assoc. Off. Anal. Chem. E, 367-372. Stonard, M.D., and Greig, J.B. (1976). Chem. Biol. Interact. 22, 365-379. Strik, J.J.T.W.A. (1978). Environ. Health Perspect. Q, 167-175. Strobel, H.W., Lu, A.Y.H., Heidema, J., and Coon, M.J. (1970). J. Biol. Chem. E, 4851-4856. Stross, J.K., Smokler, I.A., Isbister, J., and Wilcox, KR. (1981). Toxicol. Appl. Pharmacol. 22, 145-150. Studier, F.W. (1973). J. Mol. Biol. 22, 237-248. Sundstrom, G., Hutzinger, O., and Safe, S. (1976a). Chemosphere 2, 11-14. Sundstrom, G., Hutzinger, O., and Safe, S. (1976b). Chemosphere 2, 267-298. Tephly, T.R., and Mannering, G.T. (1968). Molec. Pharmacol. 2, 10-14. Thomas, P.E., Kouri, R.E., and Hutton, J.J. (1972). Biochem. Genet. _6_, 157-168. Thomas, P.E., Lu, A.Y.H., Ryan, D., West, 8.8., Kawalek, J., and Levin, W. (1976). Molec. Pharmacol. 22, 746-758. Thorgeirsson, 5.5., and Nebert, D.W. (1977). Adv. Cancer Res. 22, 149-193. Tierney, B., Weaver, D., Heintz, N.H., Schaeffer, W.I., and Bresnick, E. (1980). Arch. Biochem. Biophys. 2Q, 513-523. Troisi, C.L. (1975). M.S. Thesis, Michigan State University, East Lansing, Michigan. Trump, B.F., and Jones, R.T. (1978). Diagnostic Electron Microscopy, Vol. 1, John Wiley and Sons, New York. Tuey, D.B., and Matthews, H.B. (1980). Toxicol. Appl. Pharmacol. _5_3, 420-431. 235 Vos, J.G., Faith, R.E., and Luster, M.I. (1980). In Halogenated Biphenyls, Terphenyls, Naphthalenes, Dibenzodioxins and Related Products (R.D. Kimbrough, ed.), Topics in Environmental Health, Vol. 4, pp. 241-266, Elsevier, Amsterdam. Wada, F., Slibata, H., Goto, M., and Sakamoto, Y. (1968). Biochem. Biophys. Acta _16_2, 518-524. Warner, M., Vella La Marca, M., and Neims, A.H. (1977). Fed. Proc. E, 3804. Welti, D., and Sissons, D. (1972). Organic Magnetic Resonance 12, 309-319. Welton, A.F., and Aust, SD. (1974). Biochem. Biophys. Res. Commun. 2g, 898-906. Wolff, M.S., Aubrey, B., Camper, F., and Haymes, N. (1978a). Environ. Health Perspect. Q, 177-181. Wolff, M.S., Haymes, N., Anderson, H.A., and Selikoff, I.J. (1978b). Environ. Health Perspect. Q, 315-319. Wood, A.W., Levin, W., Lu, A.Y.H., Yagi, H., Hernandez, O., Jerina, D.M., and Conney, A.H. (1976a). J. Biol. Chem. 222, 4882-4890. Wood, A.W., Wislocki, P.G., Chang, R.L., Levin, W., Lu, A.Y.H., Yagi, H., Hernandez, O., Jerina, D.M., and Conney, A.H. (1976b). Cancer Res. 23, 3358-3366. Wyndham, C., Devenish, J., and Safe, S. (1976). Res. Comm. Chem. Pathol. Pharmacol. Q, 563-570. Yang, S.K., McCourt, D.M., Roller, P.P., and Gelboin, H.V. (1976). Proc. Natl. Acad. Sci. U.S.A. 7_3_, 2594-2598. Yang, S.K., McCourt, D.M., Leutz, J.C., and Gelboin, H.V. (1977). Science 226, 1199-1201. Yang, S.K., Deutsch, J., and Gelboin, H.V. (1978). In Polycyclic Hydrocarbons and Cancer (H.V. Gelboin, and P.D.P. Ts'O, eds.), Vol. I, pp. 205-231, Academic Press, New York. Yoshihara, S., Kawano, K., Yoshimura, H., Kuroki, H., and Masuda, Y. (1979). Chemosphere _8_, 531-538. Yoshimura, H., Yoshihara, S., Ozawa, N., and Miki, M. (1979). In Health Effects of Halogenated Aromatic Hydrocarbons (W.J. Nicholson, and J.A. Moore, eds.), Ann. N.Y. Acad. Sci. E2, pp. 179-192, The New York Academy of Sciences, New York. Zinkl, J.G., Vos, J.G., Moore, J.A., and Gupta, B.N. (1973). Environ. Health Perspect. 2, 111-118. APPENDIX APPENDIX LIST OF PUBLICATIONS As mentioned in the acknowledgements, some studies in this dissertation were performed in collaboration with other workers. Some of these collaborative efforts yielded a number of copyrighted journal articles which have been or will be published. References to these and other materials are given below. 1. Published papers Ghazi A. Dannan, Robert W. Moore, and Steven D. Aust. Studies on the microsomal metabolism and binding of polybrominated biphenyls (PBBs). ' Environ. Health Perspect. Q: 51-61 (1978) Robert W. Moore, Ghazi A. Dannan, and Steven D. Aust. Induction of drug metabolizing enzymes in polybrominated biphenyl-fed lactating rats and their pups. Environ. Health Perspect. Q: 159-165 (1978) Ghazi A. Dannan, Robert W. Moore, Lawrance C. Besaw, and Steven D. Aust. 2,4,5,3',4',5'-Hexabromobiphenyl is both a 3-methylcholanthrene and a phenobarbital-type inducer of microsomal drug metabolizing enzymes. Biochem. Biophys. Res. Commun. _8_5: 450-458 (1978) Robert W. Moore, Ghazi A. Dannan, and Steven D. Aust. Structure-function relationships for the pharmacological and toxicological effects and metabolism of polybrominated biphenyl congeners. In Molecular Basis of Environmental Toxicity (R.S. Bhatnagar, ed.). Ch. _8_: 173-212 (1980) Manuscripts submitted or being prepared for submission Steven D. Aust, Ghazi A. Dannan, Stuart D. Sleight, Pamela J. Fraker, Robert K. Ringer, and Donald Polin. Toxicology of polybrominated biphenyls. Presented at the Symposium on Health and Ecological Effects of Chlorinated Hydrocarbons: Hepatotoxicity of Halogenated Hydrocarbons. Second Chemical Congress of the North American Continent, Las Vegas, Nevada, August 24-29, 1980. To be published by Pergamon Press, M.A.L. Khan, Ed. Ghazi A. Dannan, Stuart D. Sleight, Pamela J. Fraker, Janver D. Krehbiel, and Steven D. Aust. Toxicity of 2,4,5,3',4'-pentabromobiphenyl. Accepted in Toxicol. Appl. Pharmacol. (1982). Ghazi A. Dannan, Gerald J. Mileski, and Steven D. Aust. Purification of polybrominated biphenyl congeners. Accepted in J. Toxicol. Environ. Health. (1982). 236 237 Ghazi A. Dannan, Stuart D. Sleight, and Steven D. Aust. Reconstitution of some toxicopharmacological effects of commercial mixtures of polybrominated biphenyls. Manuscript in preparation. H ”lllllllljljlll[llljilljllllllllllllES