Xenogenetics in multifactorial disease susceptibility

C O M P L E X

DISEASES

M a j o r advances have been made in the past 20 years in understanding the genetic basis of single gene disorders. However, common diseases, such as rheumatoid arthritis 1, neurodegenerative diseases z, many cancers3 and diabetes 4, are conditions in which a variety of genes interact. Understanding these diseases is a real challenge because environmental factors are also involved. An example of this is in insulin-dependent diabetes mellitus, where concordance between monozygotic twins has been demonstrated to be only 36°4 (Ref. 5). It is the genetic control of the response of an individual to foreign chemicals (xenobiotics) which is meant by 'xenogenetics'. Polymorphisms that affect the response of an individual to the environment are candidate risk factors in multifactorial diseases, and family studies 4 have the power to identify potential xenogenes, the products of which metabolize environmentvl chemicals. Hc,A'ever, to understand the genetic control of a response to an environmental insult, it is necessary to study disorders in which there is an identified environmental trigger. The association between exogenous chemicals and disease has a long history. A very early link was made in 1775, when Pott described scrotal cancer as 'Chimney sweep's disease '6 and his grandson even made an early reference to xenogenetics7'.., a constitutional predisposition is required which renders the individual susceptible to the action of soot' (H. Earle, 1823; Ref. 7). An environmental contribution to bladder cancer was identified more than 100years ago when the disease was associated with German dye workers 8. The cause of bladder cancer in these worker.-,, has been identified as a group of carcinogenic arylamines 9. The use of these carcinogens has been severely restricted il, the British dye and rubber industry for 40 years, although accidental exposure to these carcinogens still occurs through cigarette smoke, exhaust fumes and even diet l°. It is less easy to identify accidental exposure, although sensitive methods are available to detect chemical DNA adducts in human tumours n, and protein adducts are also considered to be an indicator of chemical exposure 12. In contrast, exposure to a particular group of xenobiotics, drugs, is usually deliberate. Both the individual and the environmental factor can be identified unambiguously. In some cases, the side effects of drug treatment induce symptoms that mimic idiopathic diseases. Throe examples of very different polygenic disorders with a known environmental aetiology are discussed here. Studying such multifactorial disorders should provide clues to the connections between genetic and environmental factors in common diseases. Parkinson disease Parkinson disease is characterized by a reduction of dopamine in the brain due to the death of dopaminergic neurones in the substantia nigra. A role for environmental compounds in the pathogenesis was shown around 20 years ago: a contaminant of illicit narcotics, 1-methyl-4-phenyl-l,2,3,6-tetrahydropyridine (MPTP) induced a disorder clinically indistinguishable from Parkinson disease in some drug addicts 13. There is additional evidence for a genetic component to MvrP

Xen0genetics in multifact0rial disease

susceptibility

EDITH SIM, LESLEYA. STANLEY,ANGELARISCHAND PETERTHYGESEN

Susceptibility to multifactorial disease includes both genetic and environmental components. These two aspects of susceptibility are interlinked through genetic control of an individual's response to the environment As a first step in L4entifying disease susceptibility genes that influence the response of an individual to foreign compounds (xenobtotics), it is necessary to study disorders in which there is en identified environmental trigger. Establishing a DNA resource from individuals with known environmental exposure ('a xenogenetic register') for diseases with an established environmental aetiology is an essential step in beginning to understand how environmental factors contribute to the susceptibility to polygenic diseases. A complementary approach to identification of environmentalfactors is suggested using a comparison of genetically homogeneous subdivisions of individuals with polygenic diseases where there is no clue to the environmenta! wlgger. induction of Parkinson disease from mouse model studies, where different mouse strains vary in susceptibilityTM. The neurotoxic effect of MPTP has been confirmed in mice and primates, and is due to a meLabolite, the 1-methyl-4-phenyl-pyridinium ions (MPP+), produced by the oxidation of MPTP by monoamine oxidase B in glial cell mitochondria. The MPP+ ion is selectively taken up into the neurones of the substantia nigra, where it inhibits mitochondrial NADH-ubiquinone oxidase (complex 1). The result is a decrease in mitochondrial respiratory activity that leads to neuronal death 15. In idiopathic Parkinson patients, a similar complex 1 deficiency in the dopaminergic neurones of the substantia nigra has been described 16, showing the importance of complex 1 activity in the disease pathogenesis. Complex 1 of the mitochondrial respiratory chain can also be inhibited by other neurotoy±~, including Nmeth~dated analogues of the MPP+ ion, which accumulate in the brain 15. A 5 kb deletion in the mitochondrial DNA that encodes subunits of complex 1 has been found in elderly Parkinson disease patients 13, but its role is controversial because it is not found in young idiopathic Parkinson disease patients 17. Monoamine oxidase B, which metabolizes MPTP, is known to be polymorphic in one of the introns. However, there is no apparent difference in the activity associated with different alleles, and Parkinson patients and controls are suggested not to differ in allelic distribution TM. In vitro studies have shown that MPTP, and the MPP + ion also, might be substrates of debrisoquine hydroxylase, a polymorphic cytochrome P450 monooxygenase (CYP2D6). The enzyme is produced in the striatum 19, a region of the brain with connections to the

TIG DECEMBER1995 VOL. 11 NO. 12 © 1995ElsevierScienceLid0168- 9525/95/$09.50

5 0 9

COMPLEX

DISEASES

including the anti-hypertensive drug hydralazine, the anti-rheumatic agent penicillamine, the anti-arrythmic agent procainamide 24,25and, the now discontinued, betablocker practolol z6. These side effects have been shown 90with varying degrees of severity and tissue involvement. With penicillamine, 26% of patients experience 80proteinuria, whereas practolol caused the extremely severe, but very rare, 'oculomucocutaneous syndrome', which results in blindness and acute intestinal obstruc70tion in some patients necessitating withdrawal of the drug. Although it appeared that metabolism of practolol 60was required at the time when these patients were identified 2"/, no DNA samples have been available. There have been investigations into associations of % 50MHC antigens with drug-induced autoimmunity, and the presence of null alleles at the C4 loci in the class III 40region have been identified as contributory, but not obligatory, factors in hydralazine-induced SLE (Ref. 28) and penicillamine-induced proteinuria 29. Although there 30are hints that variation in the ability to metabolize penicillamine affects susceptibility, it remains to be estab20 lished that there is a genetic basis to the variation in the ability to metabolize this sulphydryl drugD0. 10 Molecular mechanisms have been identified that contribute to penicillamine-induced proteinuria and drug-induced SLE. For penicillamine and hydralazine, 0 the drugs themselves are likely to promote immune Control Nonsmokers Smokers complex deposition z4. In the case of procainamide, smokers with cancer with cancer which is an arylamine, oxidation of the drug31 is likely to be important. Polymorphonuclear leucocytes oxidize FiGtme1. NAT2type in smoking-inducedbladder cancer. procainamide to the hydroxylamir, e derivative32, which The NAT2genotypehas been determined in bladder cancer could affect locally the ability of' these cells to ingest patients and control patients with no malignancywhose smoking immune complexes. These studies lay the foundation history is known. Fast acetylators are represented by the open for identifying candidate genes, which can be comcolumns, slow acetylatorsby the filled columns. pared between individuals who take procainamide and Adapted from Ref. 38. develop a condition resembling SLE, and those who develop anti-nuclear antibodies but who do not develop substantia nigra. The genetic basis of CYP2D6 poly- immune complex disease. morphism is known. The 'poor metabolizer (PM) phenoIn competition with the oxidation reaction, protype', which occurs in around 10% of the Caucasian cainamide is metabolized by arylamine N-acetyltranspopulation, is due to a series of mutations in the ferase 33. Arylamine N-acetyltransferase is encoded at CYP2D6 gene, leading to the absence of production of two loci, NAT1 and NAT2, and NAT2 is a polymorphic a functional protein 20. It has been suggested that a poss- xenogene with at least six different alleles, only one of ible function of CYP2D6 in the substantia nigra is the which confers 'fast acetylation', which is dominant34. detoxification of unidentified environmental chemicals, The other alleles confer the 'slow acetylation phenoperhaps related to MPTP, which could cause Parkinson type'. Slow acetylators develop procainamide-induced disease 21. There is a 2-3-fold increase in the frequency SLE more rapidly than fast acetylators, but the toxic of mutant CYP2D6 alleles in patients with Parkinson condition can occur both in fast and in slow acetylators. disease2L z2. As would be expected for a multifactorial It might be that fast acetylators metabolize less of the disease, the locus of the CYP2D6gene is not the only drug to the hydroxylamine form. Hydralazine is a determinant of genetic susceptibility and appears to hydrazine derivative that is also metabolized by NAT2. have a minor role in susceptibility to familial Parkinson In hydralazine-induced SLE, where the drag itself is disease 2,z3. However, it might be that the environmental likely to contribute to the side effects24, only slow aspect, of which CYP2D6 is an indicator, is less impor- acetylators develop immune complex disease25. The tant in the autosomal dominant form of the disorderl ability to acetylate the drug, particularly hydralazine, is the most important polymorphism identified to date. Immune.complex disease In many connective tissue diseases, such as rheuma- Bladder cancer toid arthritis and systemic lupus erythematosus (SLE), a Epidemiological studies, as well as the accumulated range of autoantibodies that are not tissue specific are clinical experience of urologists with an interest in occufound. Immune complexes become deposited at inap- pational health, have identified occupational exposure propriate sites and induce inflammation. Autoimmune to arylamines as a major environmental risk factor in symptoms, resembling immune complex diseases, have bladder carcinogenesis35. There is also epidemiological been described after treatment with a range of drugs, evidence that exposure to smoking contributes to 100"

TIG DECEMBER1995 VOL. 11 NO. 12

510

COMPLEX

DISEASES

induction of bladder cancer36. In some cases, exposure in a Japanese population 46 compared with a European Caucasian population 47. It highlights the importance of to arylamines in smokers has been confirmed through identification of protein adducts 12. Carcinogenic aryl- taking ethnic origin into consideration when compiling amines, such as benzidine and beta naphthylamine, a xenogenetic resource, which is particularly important in an ethnically mixed society. A genomic DNA bank, which are bladder carcinogens, are substrates for xenobiotic metabolizing enzymes, including glucuronosyl- which is available for distribution to investigators has been established from Caucasians with known levels of transferase, cytochromes P450, glutathione S-transferase, sulphotransferase, and NAT1 and NAT2. Investigations industrial exposure3S. The DN~ bank can be used to of polymorphisms affecting the cytochrome P450 en- investigate polymorphisms in other enzymes involved in metabolism of bladder carcinogens. coding gene, CYPIA1, showed no difference between Within bladder tumours, al!elic loss has been identibladder cancer patients and controls, although deletion fied on specific regions of several chromosomes includof the glutathione S-transferase M1 encoding gene was ing 9 (Ref. 48), 11 (Ref. 49) and 8 (Ref. 50). It is intrimore prevalent in the bladder cancer patients37. There is a large body of information on NAT2 polymorphism guing that one region identified on human chromosome 8 as showing loss of heterozygosity in highly in bladder cancer. Polymorphism in NAT2 has been investigated using enzymic function and allelic variants, invasive bladder tumours, is close to and might include and has demonstrated that the slow N-acetylation pheno- the loci for arylamine N-acetyltransferase51. The relationtype is a susceptibility factor contributing to industrial ship between NAT gene loss, the loss of a candidate bladder cancer in Caucasians (Table 1)38-42. In addi- tumour suppressor gene in the same region of chromosome 8 (Ref. 52) and tumour progression is certainly tion, a comparison of smokers and non-smokers who had bladder cancer, or who had other non-malignant worthy of ful"ther investigation. urological problems, showed clearly the importance of slow acetylation as a xenogenetic factor contributing to Conclusions and future possibilities The examples that have been chosen show how a smoking-related bladder cancer (Fig. 1). These case control studies provide a growing body study of exposure to environmental chemicals can proof evidence that slow N-acetylation is a contributory vide clues to potential susceptibility factors in multifactor in bladder cancer with no identifiable cause, that factorial diseases. Xenogenetics can, in turn, provide is where exposure to environmental chemicals is acci- insight into potential contributions to common diseases from accidental exposure. Chemical analyses of DNA dental but not easily identifiable by questionnaire, such as from passive smoking or traffic fumes. Identification and protein can be used in this context to confirm the identity of postulated environmental factors in the aetiof an urban population (Table 1) has been one approach that has been taken to try to account for ology or progression of multifactorial diseases. It is important, in studies of disorders in which there exposure to traffic fumes. The maximum relative risk that can be conferred by is an established environmental aetiology, that DNA from individuals with known environmental exposure slow acetylation in Caucasians is about two, because around 50% of the C~.ucasian population are slow- ('a xenogenetic register') is generated as has been done acetylators. However, the contribution of a susceptibil- in studies of industrial bladder cancer38. This is an essenity factor present in 50% of the population affecting tial step in generating the material to investigate the how individuals respond to environmental chemicals interplay between environmental and genetic factors. has to be taken seriously. The incidence of slow acetylTAm 1. Slow N-acetylation and bladder cancer in Caucasians ation in Orientals is very low and is due to a virtual absence of the Slow acetylators per total in study_ commonest Caucasian 'slow' allele Exposure of (no. ofindividu~ls) Slowacetylators (O,1o) ltefs cancer patients (NAT2*5B)34. In Orientals, slow acetylation does not appear to 22/23 95.7 39 Occupational contribute to occupational bladder 44/62 71.0 38 cancer43,44, which might be the 66/85 77.6 Total result of subtle differences found in the enzymic activities of the ex74/111 66.6 39 Urban pressed products 45 encoded by the 83/127 65.4 38 commonest NAT2 alleles found in 66/100 66.0 40 46/71 64.8 41 the two populations (NAT2*5B in 67/105 63.8 42 Caucasians cf. NAT2*7B in Japan336/514 65.4 Total ese). Bladder cancer is a multifactorial disease and these popu114/207 55.1 39 Controlsa lation differences might also be 26/59 44.1b 38 due to other inherent genomic dif38/74 51.4 41 ferences in the two populations, 54/100 54.0b 42 which include differences in other 232/440 52.7 Total drug metabolizing enzymes. Cultural factors might also be imporaFree of bladder cancer (i.e. healthy volunteers). tant in expl~iining the fourfold t.Non-tnaiignant urological patietgts. lower incidence of bladder cancer TIG DECEMBER1995 VOL. 11 NO. 12

511

COMPLEX

DISEASES

This DNA can be screened for other susceptibility genes once these are identified, in order to assess the precise relationship between the clinical conditions induced by a known environmental agent and idiopathic disease. Parkinson disease and the MPTP-induced condition would be ideal for such a comparison. The major difficulties in identification of xenogenetic contributions to susceptibility in multifactorial diseases are that multiple combinations of genes and environmental factors are involved. This is particularly the case when studying disease populations of different ethnic origin. The very nature of susceptibility to multifactorial diseases means that no one risk factor is sufficient to induce disease. However, the family studies in idiopathic disease (with no identifiable environmental factor), which are being carried out 4, have succeeded in identifying a range of susceptibility genes. Family studies theoretically have the ability to identify xenogenetic risk factors, but unless all patients under investigation have the same environmental trigger it is extremely unlikely that a xenogene would emerge as an important risk factor. Subdivision of currently identified clinical populations using, as a first approach, the creation of more genetically homogeneous disease groups would be more likely to result in the identification of xenogenes on further comparisons amongst groups. The determination of a xenogenetic susceptibility factor could, in turn, prompt identification of potential environmental triggers of common diseases.

Acknowledgements We thank Gary Brown for helpful comments, Aileen Watson for assistance and the Wellcome Trust, the Cancer Research Campaign, Ganstedfonden, Denmark and the Royal Danish School of Pharmacy for financial support. A.R. is a scholar of the Christopher Welch Trust,

References I Wordsworth, B.P. and Bell, J.I. (1992) Springersemin. lmmunopathol. 14, 59-78 2 Gasser, T. et at. (1994) Ann. Neurol. 36, 387-396 3 Smith, G. et al. Cancer Surv. (in press) 4 Davies, J.L. et al. (1994) Nature 371, 130-136 50imos, P. et al. (1988) Diabetologia 31,747-750 6 Pott, P. (1775) in Cbirurgical Observations Relative to Cataract, the Polypus of the Nose, the Cancer of the Scrotum, the Different Kinds of Ruptures and the Mcrt~cations of the Toes and Feet, pp. 63-68, Hawes,

E. SIM, L.A. STANLEYAND .4. RI$CH ARE IN THE DEPARTMENT OF PHARMACOLOGY, UNIVERSlTIr OF OXFORig JWANSFIELD

Cla:'ke & Collins 7 Earle, H. (1823) Medico-Chirurgical Trans. 12, part II, 296--307 8 Rehn, L. (1895) Arch. Kiln. Chir. 50, 588--600 9 Case, R.A., Hosker, M.E., McDonald, D.B. and Pearson, J.T. (1954) Br.J. Indust. med. 11, 75-104 I 0 IAn, D. et al. (1994) Cancer Res. 54, 4920-4926 11 Phillips, D.H. (1993) Br. Med.J. 306, 1444-1448 12 Vineis, P. et al. (1994) Nature 369, 154-156 13 Mizuno, Y. et al. (1994) Arch. Gerontol. Geriatrics 19, 105-121 14 Festing, M. (1991) Experentia 47, 990-998 15 Singer, T.P. and Ramsay, R.R. (1990) FEBSLett. 274, 1-8 16 Schapira, A.H.V. et al. (1990)J. Neurochem. 54, 823-827 17 Shah, D,E., Yeh, S.I., Wan, Y.C. and Wei, Y.H (1995) Acta Neurol. Scand. 91, 149-152 18 Ho, S.L. et al. (1994) Biogenic Amines 10, 579-585 19 Nisnik, H.B., Tyndale, R.F. and Sallee, R.F. (1990) Arch. Biochem. Biophys. 276, 424-432 TIG DECE?

20 Helm, M.H. and Meyer, U.A. (1992) Genomics 14, 49-58 2/ Smith, C.A.D. et al. (1992) Lancet339, 1375-1377 22 Armstrong, M. et al. (1992) Lancet 339, 1017-1018 23 Plante-Bordeneuve, V. et al. (1994)J. Neurol. Neurosurg. Psychiatry 57, 911-913 24 Sim, E. (1991) Biochem. Soc. Trans. 19, 164-170 25 Adams, L.E. and Mongey, A-B. (1994) Lupus 3, 443-447 26 Nicholls, J.T. (1976) Anal Clin. Res. 8, 229-231 27 Amos, H.E., Lake, B.G. and Artis,J. (1978) Br. Med. J. 402-404 28 Speirs, C. et al. (1989) Lancet i, 922-924 2.9 Edmonds, S.D. et al. Br.J. Rheumatol. (in press) 30 Gregory, W.L. (1993) Pharmacogenetics 5, 270-274 31 Sim, E., Stanley, L., Gill, E.W. and Jones, A. (1988) Biochem. J. 251,323-326 32 Jiang, x., Khursigam, G. and Rubin, R.L. (1994) Science 266, 810-813 33 Ohsako, S. and Deguchi, T. (1990)J. Biol. Chem. 265, 4630-4634 34 Vatsis, K.P. et al. (1995) Pharmacogenetics 5, 1-27 35 Wallace, D.M.A. (1988) Br.J. Urol. 61, 175-182 36 Doll, R, etal. (1994) Br. med.J. 309, 901-911 37 Katoh, T., Inatomi, H., Nagaoka, A. and Sugita, A. (1995) Carcinogenesis 16, 655-657 38 Risch, A., Wallace, D.M.A., Bathers, S. and Sim, E. (1995) Hum. Mol. Gen. 4, 231-236 39 Cartwright, R.A. et al. (1982) Lancet ii, 842-845 40 Evans, D.A.P., Eze, L.C. and Whibley, E.J. (1983) J. Med. Genet. 20, 330-333 41 Wolf, H., Lower, G.M. and Bryan, G.T. (1980) Scand.J. Urol. Nephrol. 14, 161-165 42 Mommsen, S., Barfod, N.M. and Aaraard, J. (1985) Carcinogenesis6, 199-201 43 Hayes, R.B. et al. (1993) Carcinogenesis 14, 675-678 44 Horai, Y., Fujita, K. and Ishizaki, T. (1989) Eur.J. Clin. Pharm. 37 581-587 45 Hickman, D., Palamanda, J., Unadkat, J. and Sim, E. (1995) Biochem. Pharmacol. 50, 697-703 46 Nakata, S. et al. (1994) Jap.J. Urol. 85, 1734-1742 47 Kiemeney, L.A.L.M. et al. (1994) Eur.J. Cancer30, 1134-1137 48 Ruppert, J.M., Tokino, K. and Sidransky, D. (1993) Cancer Res. 53, 5093--5095 49 Shipman, R. et al. (1993) Hum. Genet. 91,455-458 50 Knowles, M.A., Shaw, M.E. and Proctor, A.J. (1993) Oncogene 8, 1357-1364 51 Spun', N.K. et al. (1995) Cytogen. Cell. Genet. 69, 147-164 52 Fujiwara, Y. et al. (1995) Oncogene 10, 891-895

RoAIg O x f o l ~ UK OXI 3Q~, P. THgG£SEN IS IN THE DEPARTMENTOF PHARMACOLOGY,ROYAL DANISHSCHOOLOF PHaS~CY, 2 U N ~ s n m s P a m ~ N , DE-2100 COPENHAGEN, DENMARK.

50% off all new student subscriptions! Did you know that as a student you are entitled to a s p e c i a l d i s c o u n t on a personal subscription to Trends in Genetics? See the subscription order form for details.

~ 1995 VOL. 11 NO. 12

512