Molecular Diagnosis Vol. 4 No. 4 1999
Populations and Genetic Polymorphisms WENDELL
W. W E B E R , P h D , M D
A n n Arbo~ Michigan
Background: Population frequencies of many polymorphic genes of pharmacogenetic interest depend on race or ethnic specificity. Association of these genes with person-to-person differences in drug effectiveness (hypersensitivity or resistance) and drug toxicity may also depend on the racial or ethnic characteristics of a population. Information about ethnic specificity is an integral part of pharmacogenetics because it can suggest a starting point for further study of these traits, tailoring drug therapy to the individual patient, and rational development and clinical trials of new drugs. Ethnic specificities of several medically important metabolic traits serve to illustrate these ideas. Among the traits considered is primaquine sensitivity, a sex-linked trait attributed to glucose-6-phosphate dehydrogenase deficiency that mainly affects males among African, Mediterranean, and Oriental people. Additional examples include the remarkable sensitivity of the Japanese to alcohol (ethanol) compared with whites; the ethnic specificity of the cytochrome P-450 enzyme CYP2D6* (debrisoquine/sparteine) polymorphism that results in poor, extensive, and ultrarapid metabolizers of at least 30 drugs; the CYP2C19* (mephenytoin) polymorphism that accounts for variable metabolism of proguanil, omeprazole, and certain barbiturates; and the polymorphic (NAT2*) acetylation of hydrazine and aromatic amine drugs, such as isoniazid, hydralazine, and sulfasalazine. Key words: ethnic specificity, pharmacogenetics, human drug response, human evolution, genetic polymorphism, racial variation.
Reports of racial differences in response to drugs and other exogenous chemicals appeared very infrequently in the medical literature before the 1920s. In one early study, Marshall et al. [1] reported that blacks were beyond a doubt much more resistant than whites to blistering of the skin from exposure to mustard gas, a substance introduced into the arsenal of biological weapons during World War I. In 1921, McGuigan [2], in reporting that small doses of atropine always caused a slowing of the pulse, briefly mentioned that blacks seemed to be much less susceptible to this effect than whites. In a fol-
low-up study, Paskind [3] confrmed this as a bonafide racial difference. A few years hence, Middleton and Chen [4] published ophthalmological observations that ephedrine dilated the pupils of Chinese individuals only slightly compared with whites. Because their findings were exceptional, they withheld publication until further study of ephedrine, as well as of some other mydriatics such as cocaine and pseudoephedrine, confirmed their findings in Chinese and whites and extended them to blacks [5]. During the 1930s, studies of taste blindness, a hereditary deficit in sensory perception, in several subpopulations of Africa, Asia, the Middle East, and Europe showed that the frequency of the nontaster phenotype in northwestern Europe is approximately 35% to 40%, but is appreciably less in Africans, Chinese, Japanese, South Amerinds, and also in Lapps (Table 1) [6]. The association of race or ethnicity with human responsiveness to a few drugs and exogenous chemicals suggested this might be a more general phenomenon af-
From the University of Michigan, Ann Arbor, Michigan. Supported in part by Public Health Service grant no. CA39018. Reprint requests: Wendell W. Weber, PhD, MD, Department of Pharmacology, 1301b-MSRB III, University of Michigan, Ann Arbor, MI 48109-0632. Email:
[email protected] Copyright © 1999 by Churchill Livingstone ® 1084-8592/99/0404-0006510.00/0
299
300
Molecular Diagnosis Vol. 4 No. 4 December 1999
Table 1. Racial Variation of Taste Blindness: Frequency of Phenylthiourea Nontasters Population Hindus Danish English Spanish Portuguese Negritos (Malaya) Malays Japanese Lapps West Africans Chinese South Amerinds (Brazil)
Number Tested
Nontasters (%)
489 251 441 203 454 50 237 295 140 74 50 163
33.7 32.7 31.5 25.6 24.0 18.0 16.0 7.1 6.4 2.7 2.0 1.2
Reprinted with permission [6].
fecting human responses to many other exogenous substances among different populations. During the 1950s, the development of new technologies enabled unique patterns of drug metabolites to be identified in different individuals that brought to light new relationships between the metabolic fate of exogenous chemicals and human drug response. Intense scrutinization of person-to-person variations in response to such drugs as primaquine, succinylcholine, and isoniazid combined with a more genetic approach led to the discovery of primaquine sensitivity (glucose-6-phosphate dehydrogenase [G6PD] deficiency), succinylcholine sensitivity, and isoniazid acetylation polymorphism, and saw the emergence of pharmacogenetics as a new experimental science [7]. Because these studies were often conducted on diverse populations by investigators in different parts of the world, striking ethnic specificities were observed. Gradually, the idea that ethnic variation could provide unique insights into hereditary and environmental factors affecting individual response to the therapeutic actions of drugs, dietary components, and other exogenous chemicals was established more by chance than design. Since then, the study of ethnogeographic variation has been regarded as an important part of comprehensive pharmacogenetic study.
Concepts of Race and Ethnicity According to most commonly accepted recent views, the history of the modern human species (Homo) begins around 2 million years ago. Unless new findings come to light, Africa is believed to be the cradle of the most recent ancient ancestors of humans, and the divergence of modern humans (Homo sapiens) is estimated to have begun 100,000 to 150,000 years ago [8]. Since then, the human species has diverged further into Negroid, Mongoloid, and Caucasoid races, for which a race is defined
genetically as a large group of individuals with a significant fraction of its genes in common that can be distinguished by its gene pool [9]. Sometimes, in addition to these three major races, Capoids (Bushmen and Hottentots) and Australoids, two smaller groups, are added. Speaking very broadly, four criteria have been used to specify races: geographic, which is founded on the idea that the exchange of genes is reduced as distance between populations increases (until 150 years ago, most movements covered distances of no more than 150 miles, and only rarely did people venture beyond this short range); anthropologic, which focuses on similarity of height, weight, body build, and facial features as traits that could be easily perceived; linguistic,which takes account of relationships between 4,736 languages of the world (interestingly, there is a strong parallel between the evolution of linguistic and genetic trees); and ethnic, which attempts to take account of social, behavioral, and cultural characteristics. The order of importance for classification diminishes from geographic to ethnic. Biologists had long been aware of a certain amount of variation among and within human populations, but the remarkable extent of this variation was not fully appreciated until the detection of protein polymorphism by electrophoresis approximately 40 years ago and of DNA polymorphism by techniques of molecular genetic analysis more recently. First efforts to construct the history of human differentiation from genetic data were made approximately 25 to 30 years ago from restriction analysis of mitochondrial DNA. Within a few years, investigations of nuclear genes used D N A collected from 15 populations on five blood group genes (ABO, MN, Rh, Diego, and Duffy), comprising a total of 20 alleles. This and later research indicated that the need for a greater number of genes, as well as for a balanced sample of populations as close as possible to what may have been the aboriginal set of populations, was crucial to evolutionary analysis. Analysis of 42 aboriginal populations incorporating 120 alleles grouped into clusters representing 9 populations yielded the human phylogenetic tree shown in Fig. 1 [8]. This is the best, most up-to-date tree that could be obtained using present methods.
What Should We Know About Drug-related Ethnic Specificity? The existence of ethnic specificities in the human response to environmental substances raises a number of questions. How frequent are ethnic specificities of pharmacogenetic interest? Are they of such a nature that might suggest a starting point for further investigation into the basis of the biological action of drugs, carcinogens, and other environmental chemicals? Would such differences be important in the development and clinical
Populations and Polymorphisms 0.102
use
I
I
•
Weber
301
New Guinean and Australian 0.044
Pacific Islander Southeast Asian
a0~
d~ r l D ~
~oss[ 0.0211
=t
Northeast Asian o.o46
ArcticNortheast Asian Amerind
I
o.ols European
E-2_
Io.o1=Non-European Caucasoid African 0.20 I
0.15 I
0.10 !
0.05 I
0.00 I Genetic Distance
150,000 years Fig. 1. Human evolutionary tree based on the average of 42 populations in nine clusters. (Modified and reprinted with permission [8].)
trials of new drugs? Do they occur frequently enough to alter the clinical care of a substantial proportion of patients? Does the application of this information in practice significantly improve patient care? Evidence is now available for many traits of pharmacogenetic interest that provide at least a partial answer to many of these questions.
How Frequent Are Ethnic Specificities of Pharmacogenetic Interest? E t h n i c S p e c i f i c i t y in G 6 P D D e f i c i e n c y Discovery of this trait is linked to sudden hemolytic reactions first observed among World War II servicemen to a number of antimalarial drugs (primaquine, pamaquine, pentoquine, isopentoquine). These reactions were high (5% to 10%) in dark-skinned races, whereas whites contracted hemolysis only rarely. Males of African, Mediterranean, and Oriental descent are particularly susceptible to this deficiency, which is estimated to affect more than 400 million people worldwide. In Africans, two types of variants are found by biochemical analysis: G6PD A and G6PD A - . The first produces normal levels of red cell activity and the second is unstable in vivo, yielding only approximately 10% of normal activity levels. Recombinant DNA analysis indicates that the enzyme deficiency that characterizes G6PD deficiency, G6PD A - , is caused by a Val--*Met substitution at codon 68 (G202-~A) [10]. A second mutation that
causes G6PD deficiency, A542~T, is found in the African variant, G6PD Santamaria. This mutation causes an Asp~Val substitution at codon 181. In Mediterranean people, a C563--*T change results in Ser-* Phe substitution at codon 188. Less is known about deficiency mutations in Orientals than in Mediterraneans, but one of the more common Oriental variants, G6PD Canton, has an Arg-~Leu substitution at codon 459. Thus, the ethnic specificities in G6PD deficiency are caused by different molecular changes (for a more comprehensive list of G6PD variants, see [11]).
Ethnic S p e c i f i c i t y in C y t o c h r o m e P - 4 5 0 Enzymes: CYP2D6 and CYP2C19 Polymorphisms At least 30 different human P-450 enzymes, now designated CYP450, have been purified, cloned, sequenced, and characterized [12]. Approximately a half dozen of these, namely 1A2, 2A6, 2C19, 2D6, 2E6, and 3A4, account for the oxidation of most drugs and other substrates in the human environment. Evidence for heritable polymorphisms is well established for 1A1, 2A6, 2C9, 2C19, and 2D6 and is rapidly accumulating for 1A2, 2A6, and 3A4. The data available on ethnic specificities of C Y P 2 D 6 and CYP2C19 phenotypes are particularly apropos of this article. The C Y P 2 D 6 * (debrisoquine/sparteine oxidation) polymorphism results in three separable phenotypes: poor metabolizers, extensive metabolizers, and ultrarapid metabolizers. Poor metabolizers are homozygous for an inactive or deficient CYP2D6 enzyme caused by trunca-
302
Molecular Diagnosis Vol. 4 No. 4 December 1999
Table 2. Polymorphisms in Cytochrome CYP2D6 Nucleotide Changes and Their Locations CYP2D6 Allele *2 *4A *4B *10A *10B *10C "17
C~T
C~A
A~G
C~G
188 188 188 188 188
1062 1062
1072 1072
1085 1085
C~T
C~T
G~C
G~C 1749 1749
G~A
C~T
G-~C
People From
2938
4268 4268 4268 4268 4268 4268 4268
Europe Europe Europe Japan China China A~ica
1934 1934
1749 1749 1749
1127 1127 1111
1726
2938
The variants *4A and *4B represent enzymes without activity.All others have reduced activity compared with the wild type. Compiled from data in [17,18] by W. Kalow (personal communication, May 1999).
tion or missense mutations of the CYP2D6* gene. It has been widely documented that the oxidizing capacity of CYP2D6* poor metabolizers is impaired for more than 30 widely used therapeutic agents, including beta-adrenergic blockers, antidepressants, antiarrhythmics, neuroleptics, and various miscellaneous drugs, such as phenformin, dextromethorphan, and codeine. Among Northern European populations, more than 95 % of the deficient alleles responsible for CYP2D6* poor metabolizers have been identified [13]. In contrast, ultrarapid metabolizers possess an enhanced capacity to metabolize because of gene amplification (CYP2D6L*) [14]. Ultrarapid metabolizers may fail to respond to drugs inactivated by CYP2D6* (e.g., the antianxiety agent, nortriptyline) or may show exaggerated responses to agents activated by CYP2D6* (e.g., the analgesic, codeine). Many population studies performed since the discovery of CYP2D6* polymorphism in the 1970s show that the prevalence of CYP2D6 phenotypes varies considerably from one population to another. For instance, the poor metabolizer phenotype varies from approximately 1% in Japanese and Egyptians to 5% to 10% among whites of Europe and North America, and one study reported an absence of this phenotype in Cunas Amerindians of Panama [15]. In Europeans, poor metabolizers of debrisoquine are also poor metabolizers of sparteine, but among Ghanians, poor metabolizers of debrisoquine (and phenformin) are not poor metabolizers of sparteine [16]. Table 2 [W. Kalow, personal communication, May 1999] concisely describes the extent of ethnic variation of people from Africa, Asia, and Europe, with 11 nucleotide changes belonging to seven CYP2D6 allelic variants [17,18]. It can be seen that certain mutations (G4268-+C) may be shared by people from all three geographic regions, suggesting the mutation probably occurred before evolutionary separation. In contrast, other mutations occur in people within particular regions, such as the C l 1 2 7 ~ T in China or the G1726~C in Africa. The latter examples suggest that those mutations probably occurred after evolutionary separation.
Recently, individuals with enhanced capacities to metabolize drugs, the ultrarapid metabolizers [19], have been shown to possess alleles with more than one copy of the CYP2D6 gene. The discovery and initial characterization of the first cases and families harboring this phenotype were attributed to CYP2D6L* [14]. Further studies have identified subjects with 2, 3, 4, 5 and 13 copies of the CYP2D6 gene in one allele. The ethnic distribution of ultrarapid metabolizer phenotypes in Northern European populations, summarized by Ingelman-Sundberg (Table 3) [20], shows that Swedish and German populations carry 1% to 2% of the CYP2D6L ultrarapid phenotype, whereas the frequency of ultrarapid carriers in Ethiopia [22] and in Saudi Arabia (20%; not shown in Table 3) [23] is very high. Spain also has a high proportion of individuals carrying amplified CYP2D6L* genes [24,25]. Ingelman-Sundberg [20] proposed the intriguing hypothesis that dietary components have exerted selective pressure to explain the preservation of alleles with multiple CYP2D6* copies. He suggested that the presence of CYP2D6* alleles, which have a very high affinity for various plant-derived alkaloids, could provide a high detoxification potential for individuals exposed to diets containing these substances. In support of this hypothe-
Table 3. Ethnic Distribution of CYP2D6 Ultrarapid Metabolizers Allele Frequency (%) CYP2D6 Allele "1 ×2 *2 ×2 *2 ×3 *2 ×4 *2 ×5 *2 ×13
Swedes [13] 0 1.5 0.2 0 0 0.1
Germans Spaniards [21] [24] 0.5 1.3 0 0 0 0
ND 3.5-5 ND ND ND ND
Ethiopians [22] ND 13 1.6 0.8 0.4 0
ND, not determined. Data compiled from references 13, 21, 24, and 25 by M. Ingelman (personal communication, May 1999).
Populations and Polymorphisms
sis, he noted that very few CYP2D6* with inactivating mutations are found in Ethiopia and Saudi Arabia. Interethnic allelic frequencies of several poor-metabolizer variants of CYP2C19*, another CYP450 isozyme of pharmacogenetic importance, also vary quite remarkably. CYP2C19* polymorphism, originally called mephenytoin polymorphism, is of major pharmacogenetic importance because the elimination or activation of such drugs as the antiulcer drug, omeprazole; the antimalarial, proguanil; and a number of barbiturates depend on this trait.The incidence of CYP2C19* poor metabolizers is much greater in Oriental (13% to 23%) than in white (2% to 5%) populations. The data on specific CYP2C19 variants indicate the frequency of poor metabolizers with the m l (CYP2C19"2) variant is high in the Japanese compared with whites. Another variant, m2 (CYP2C19"3), occurs in Japanese and Africans but has not been detected in whites. The wild-type (wt), ml, and m2 variants account for almost 100% of the variation in Chinese and Japanese poor metabolizers, but these and additional allelic variants account for only approximately 92% of the variation in whites [26].
Ethnic Specificity in Acetylation
q
*
Weber
303
HI
IU _J UI _J .J
IL
0 >
0"5
I,J Ui
t
0m W LL 0, N
$
LATITUDE Fig. 2. A cline of slow acetylator allelic frequency. (Redrawn and reprinted with permission [27].)
(NAT2*) Polymorphism Acetylation polymorphism is attributed to hereditary variation in levels of the conjugating enzyme, N-acetyltransferase (NAT), present in liver and certain other tissues. It confers rapid or slow acetylator status on individuals and is widely acknowledged as a valuable prognosticator of individual susceptibility to the effective and safe use of numerous drugs (isoniazid, sulfonamides, procainamide, hydralazine, sulfasalazine, etc), to toxicity from industrial/occupational chemicals with carcinogenic potential, and possibly to dietary mutagens. Studies of the global distribution of acetylator gene frequencies encompass measurements on more than 10,000 individuals in dozens of different populations since the trait was discovered in the 1950s [27]. The percentage of slow acetylators is found to range from 80% or more of Egyptians and certain other Middle Eastern populations to 20% or less of Japanese and Canadian Eskimos. Populations of Europeans and African origin have, with few exceptions, intermediate percentages of slow acetylators. The cline in slow acetylator frequencies among populations of the Northern hemisphere provides one of the most dramatic illustrations of ethnic variation for any human trait described so far (Fig. 2). Population surveys of slow acetylator allelic frequencies conducted by Sunahara et al. [28] initially disclosed a trend of increasing slow acetylator allelic frequencies from northern to southern latitudes among the Japanese. Estimates of ethnic variation in acetylator allelic frequencies based on additional global observations by Karim et al. [29]
have fully borne out the idea of Sunahara et al. [28] of a relationship between latitude and acetylator status and extended them to a number of other populations of the Northern hemisphere. We do not know the natural or endogenous substrate for N-acetyltransferase (later shown to be NAT2*, described next), so we do not understand the origin of this relationship, but it suggests that the capacity for rapid acetylation (or its absence) may have conferred a selective advantage on populations evolving in more northerly latitudes. Nearly 30 years elapsed before recombinant D N A analysis of the acetylation polymorphism, but in 1989, molecular genetics showed humans encode two functional NATs, NATI* and NAT2*. The loci that encode NATI* and NAT2*, NATI* and NAT2*, respectively, are both polymorphic [30]. NATI* and NAT2* are both implicated in the disposition of numerous therapeutic agents and various environmental/occupational chemicals, but data on ethnicity in this article mainly concern NAT2* genotypes and phenotypes. A complete account of the molecular genetics of NAT2* appears elsewhere [30,31] (the nomenclature for NAT polymorphisms is explained in [32]), but some background and a few remarks about the allelic variation at the human NAT2* locus are in order. More than 20 alleles for NAT2* have been reported. The method used for this purpose has usually been polymerase chain reaction (PCR) amplification of the coding
304
Molecular Diagnosis Vol. 4 No. 4 December 1999
region and restriction endonuclease digestion, nested PCR followed by endonuclease digestion, or allele-specific PCR. It is apparent that the coding region of the gene has been the site of multiple mutations, but analysis of regions upstream and downstream are very limited. Nine nucleotide changes have been found in the coding region. The large number of alleles (>20) resulting from these mutations occur singly or in combinations of two and three. Seven of the nine nucleotide changes result in changes in the deduced amino acid, some of which may strongly affect acetylating activity. The NAT2*4 allele is taken as the wild-type because it is found in individuals phenotyped as rapid acetylators. The other alleles are associated with slow acetylation when any two are paired in an individual. Ethnicity data indicate the distribution patterns of NAT2* allelic frequencies across different populations are neither uniformly nor randomly distributed. For example, among American and European whites, the frequency of the rapid NAT2*4 allele is approximately 20% to 25%; among blacks, 36%; among Hispanics, 42%; and among Asiatics, i.e., Hong Kong Chinese, 66%; Koreans in America, 66%; and native Japanese, greater than 70%. The distribution of slow NAT2* alleles shows that for whites, three alleles (4, 5B, and 6A) account for approximately 95% of all NAT2* alleles. In Oriental populations, the 5B allele becomes rare (5% in Hong Kong Chinese to < 1 % in the Japanese), and 6A (20% to 30%) and 7A/7B (7% to 16%) are most frequent [31].
Do Ethnic Specificities Suggest a Starting Point for Further Study? It is obvious from the quantitative and qualitative variation described that extrapolation across different ethnogeographic groups for a given trait for a given drug may not be permissible. Conversely, one should remain alert to the possibility that observations on ethnic specificity may be useful in other ways, e.g., to improve diagnosis and clinical care of patients and to provide a fuller understanding of a given trait. For example, although experience indicates that African, Mediterranean, and Oriental males affected by G6PD deficiency may be more susceptible to hemolysis induced by exposure to more than 200 drugs, other ethnic studies have shown that G6PD deficiency with different characteristics can occur among whites. Thus, it has also been shown that several agents (e.g., trinitrotoluene, quinidine, nitrofurazone) may induce hemolytic reactions of greater severity and longer duration among whites than among G6PD-deficient blacks [7]. Another example concerns ethnic differences in response to alcohol. Screening of various populations indicates that alcohol dehydrogenase (ALDH2) deft-
ciency occurs with varying frequencies (8% to 45%) in populations of Mongoloid origin, but is not found in Caucasoid or Negroid populations. Facial flushing, an acute vasomotor dilation in response to ethanol, has attracted attention by its association with variant forms of ALDH2. Among the Japanese, homozygotes and most heterozygotes for the atypical (Asian) A L D H 2 are flushers, whereas those homozygous for the usual A L D H 2 are nonflushers. Nearly 86% of Japanese subjects who always experienced facial flushing have inactive ALDH2, whereas infrequent or absence of flushing is associated with active A L D H 2 [33]. As a consequence of the aversive vascular effects of ethanol, Japanese men and women with the Asian form of A L D H 2 drink significantly less alcohol than those with the Caucasian form and are more highly protected from alcoholism [34]. Thus, ethnic variation has shown that flushing may act as a deterrent to ethanol abuse but may also serve as a useful biomarker of the Asian A L D H 2 phenotype that is easily perceived by patients and physicians. A patient who shows an unexpected clinical response to a drug that has a disposition and metabolism dependent on a known pharmacogenetic trait characterized by ethnic specificities poses another situation of therapeutic interest. It should raise the question of whether the ethnicity of the patient suggests a basis for the response. Consider the CYP2D6* polymorphism as the trait of interest. As we know [7], this trait has three separable phenotypes, the elimination or activation of many widely used therapeutic agents (>30) are subject to this polymorphism, and the differences in response to medicines among the extreme CYP2D6* phenotypes, the poor and ultrarapid phenotypes, can be quite dramatic. Because the frequency of CYP2D6* poor metabolizers is significantly greater among Africans or Caucasians versus Asians, it follows that an unexpected response to a given drug might be expected to occur more frequently than in Asians. Thus, the failure of poor metabolizers to experience analgesia from codeine [35] and to be protected against dependence on the oral opiates such as codeine would be more likely to occur among Africans and Caucasians than Asians [36]. Still further, because poor metabolizers are more likely to experience interactions with other drugs, as has been widely shown (see [37] and references there), and to experience the neurotoxic effects of amphetamine analogs such as 3,4-methylenedioxymethamphetamine (MDMA, also referred to as Ecstasy) [38], these unexpected responses would also be expected to occur more frequently among Africans and Caucasians than Asians. Similar considerations may apply to the analysis of the basis for an unexpected response to a given drug in connection with ethnic differences in the ultrarapid metabolizer phenotype. Thus, the unexpected failure to respond to nortriptyline [19] or of an unexpected exaggerated response of CYP2D6* ultrarapid
Populations and Polymorphisms
metabolizers to codeine [39] would both be expected to occur more frequently among African or Saudi Arabian patients compared with Asian patients. Thus, when a patient experiences an unexpected response to a given drug, one should ask whether the ethnic or geographic origin of the patient suggests a basis for the response. Further study of genetically variable phenotypes showing ethnic specificity may also provide new information about the structures of the genes and regulatory pathways that may be responsible for unexpected or unusual drug responses. Thus, studies in various ethnic groups have shown that phenotypes caused by inactive or low-activity enzymes or to other defective protein variants may be explained alternatively as a deletion of an entire gene [40,41], truncated genes [42,43], and missense mutations of the coding region [44,45]. Examples involving ethnicity in which phenotypes are caused by high-activity enzymes may be explained as regulatory variants [46], kinetic variants [47,48], and duplicated genes [14,49].
Are Ethnic Specificities Important in New Drug Development and Testing? The ultimate purpose of pharmacogenetic investigation is to gather information that provides clues to the basis for an unexpected drug response and enables the expression of a given trait in susceptible persons to be avoided or managed safely and effectively. In essence, the genetics (mode of inheritance, allelic frequencies, ethnic and geographic specificities), molecular basis (genes responsible, and their mutation spectrum), and significance constitute the three types of information that characterize a given pharmacogenetic trait [7]. Clearly, information on ethnographic variation in the response of individuals, illustrated by the several examples chosen for presentation here, contributes to each of these three categories of information. Lacking this information would severely compromise global efforts toward optimizing the development and for clinical testing of new drugs.
Summary Human evolutionists have used geographic, anthropologic, linguistic, and ethnic criteria to divide human populations into three major races, Negroid, Mongoloid, and Caucasoid. Population frequencies of many traits of pharmacogenetic interest depend on ethnic specificity and differ greatly between and within these racial groups. Currently, the pharmacogenetic literature focuses mainly on metabolic traits affecting the disposition and response to drugs and other chemical substances in the hu-
•
Weber
305
man environment, but it clearly indicates that if pharmacogenetic polymorphisms occur across races and ethnogeographic populations, allelic frequencies for these traits are likely to differ. It also shows that allelic variants may be shared before evolutionary separation, but may be specific for a population if they occurred after evolutionary separation. Whereas allelic variants could only be inferred before the advent of molecular genetics, now they can be explained. Even though population frequencies of many traits of pharmacogenetic interest depend on ethnic specificity, the hypothesis that these ethnic specificities are clinically relevant remains to be rigorously tested. Received May 27, 1999. Received in revised form June 1, 1999. Accepted June 22, 1999.
References 1. Marshall E Jr, Lynch V, Smith H: On dichloroethylsulfide (mustard gas). II. Variations in susceptibility of the skin to dichloroethylsulphide. J Pharmacol Exp Ther 1918;12:291-301 2. McGuigan H: Racial differences as illustrated by the mydriatic action of cocaine, euphthalmine, and ephedrine. JAMA 1921;76:1338-1340 3. Paskind H: Some differences in response to atropine in white and colored races. J Lab Clin Med 1921;7:104-108 4. Middleton W, Chen K: Ephedrine, a clinical study. Arch Intern Med 1927;34:385-403 5. Chen K, Poth E: Racial differences as illustrated by the mydriatic action of cocaine, euphthalmine and ephedrine. J Pharmacol Exp Ther 1929;36:429-445 6. Harrison G, Weiner J, Tanner J, Barnicot N: Human biology. An introduction to human evolution, variation and growth. Oxford University Press, New York, 1964, p. 279 7. Weber W: Pharmacogenetics. Oxford University Press, New York, 1997 8. Cavalli-Sforza L, Menozzi R Piazza A: The history and geography of human genes. Princeton University Press, Princeton, 1994, p. 80 9. Motulsky A, Vogel F: Human genetics. SpringerVerlag, Berlin, 1986, p. 462 10. Hirono A, Beutler E: Molecular cloning and nucleotide sequence of cDNA for human glucose-6phosphate dehydrogenase variant A(-). Proc Natl Acad Sci U S A 1988;85:3951-3954 11. Beutler E:The molecular biology of G6PD variants and other red cell enzyme defects. Annu Rev Med 1992;43:47-59 12. Guengerich F: Human cytochrome P450 enzymes.
306
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
Molecular Diagnosis Vol. 4 No. 4 December 1999
In Ortiz de Montellano P (ed): Cytochrome P450, 2nd ed. Plenum, New York, 1995, pp. 473-535 Dahl M-L, Johansson I, Bertilsson L, IngelmanSundberg M, Sjoqvist F: Ultrarapid hydroxylation of debrisoquine in a Swedish population. Analysis of the molecular genetic basis. J Pharmacol Exp Ther 1995;274:516-520 Johansson I, Lundqvist E, Bertilsson L, Dahl M-L, Sjoqvist F, Ingelman-Sundberg M: Inherited amplification of an active gene in the cytochrome P450 CYP2D locus as a cause of ultrarapid metabolism of debrisoquine. Proc Natl Acad Sci U S A 1993;90: 11825-11829 Arias T, Jorge L, Inaba T: No evidence for the presence of poor metabolizers of sparteine in an Amerindian group: The Cunas of Panama. Br J Clin Pharmacol 1986;21:547-548 Woolhouse N, Eichelbaum M, Oates N, Idle J, Smith R: Dissociation of coregulatory control of debrisoquine/phenformin and sparteine oxidation in Ghanians. Clin Pharmacol Ther 1985;37:512-521 Daly A, BrockmOller J, Broly F, et al.: Nomenclature for human CYP2D6 alleles. Pharmacogenetics 1996;6:193-201 Marez D, Legrand M, Sabbagh N, et al.: Polymorphism of the cytochrome P450 CYP2D6 gene in a European population: Characterization of 48 mutations and 53 alleles, their frequencies and evolution. Pharmacogenetics 1996;7:193-202 Bertilsson L, Aberg-Wistedt A, Gustavsson L, Nordin C: Extremely rapid hydroxylation of debrisoquin: A case report with implication for treatment with nortriptyline and other tricyclic depressants. Ther Drug Monit 1985;7:478-480 Ingehnan-Sundberg M: The Gerhard Zbinden Memorial Lecture. Genetic polymorphism of drug metabolizing enzymes. Implications for toxicity of drugs and other xenobiotics. Arch Toxicol Suppl 1997;19:3-13 Sachse C, BrockmOller J, Bauer S, Roots I: Cytochrome P450 2D6 variants in a Caucasian population: Allele frequencies and phenotypic consequences. Am J Hum Genet 1997;60:284-295 Akillu E, Persson I, Bertilsson L, Johansson I, Rodrigues F, Ingelman-Sundberg M: Frequent distribution of ultrarapid metabolizers of debrisoquine in an Ethiopian population carrying duplicated and multiduplicated functional CYP2D6 alleles. J Pharmacol Exp Ther 1996;278:441-446 McClellan R, Oscarson M, Seidegard J, Evans D, Ingelman-Sundberg M: Frequent occurrence of CYP2D6 gene duplication in Saudi Arabians. Pharmacogenetics 1997;7:187-191 Agundez J, Ledesma M, Ladero J, Benitez J: Prevalence of CYP2D6 gene duplication and its reper-
25.
26.
27. 28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
cussion on the oxidative phenotype in a white population. Clin Pharmacol Ther 1995;57:265-269 Bernal Ruiz M, Johansson I, Dfilen R Dahl M-L, Ingelman-Sundberg M, Bertilsson L: Gene duplication of CYP2D6 in a Spanish population. Pharmacogenetics (in press) Ibeanu G, Blaisdell J, Ghanayem B, et al.: An additional defective allele, CYP2C19"5, contributes to the S-mephenytoin poor metabolizer phenotype in Caucasians. Pharmacogenetics 1998;8:129-135 Weber W: The acetylator genes and drug response. Oxford University Press, New York, 1987 Sunahara S, Motoyuki U, Ogawa M: Genetical and geographic studies on isoniazid inactivation. Science 1961;134:1530-1531 Karim A, Elfellah M, Evans D: Human acetylator polymorphism: Estimate of allele frequency in Libya and details of global distribution. J Med Genet 1981;18:325-333 Grant D, Hughes N, Janezic S, et al.: Human acetyltransferase polymorphisms. Mutat Res 1997;376: 61-70 Levy G, Weber W: Interindividual variability of Nacetyltransferases in man. In Pacifici G, Pelkonen O (eds): Interindividual variability in drug metabolism in man. Taylor & Francis, London (in press) Vatsis K, Weber W, Bell D, et al.: Nomenclature for N-acetyltransferases. Pharmacogenetics 1995;5:117 Shibuya A, Yasunami M, Yoshida A: Genotypes of alcohol dehydrogenase and aldehyde dehydrogenase loci in Japanese alcohol flushers and nonflushers. Hum Genet 1986;82:14-16 Higuchi S, Muramatsu T, Shigmori K, et al.: The relationship between low Km aldehyde dehydrogenase phenotype and drinking behavior in Japanese. J Stud Alcohol 1992;53:170-175 Sindrup S, Brosen K, Bjerring R et al.: Codeine increases pain thresholds to copper vapor laser stimuli in extensive but not poor metabolizers of sparteine. Clin Pharmacol Ther 1991;49:686-693 Tyndale R, Droll K, Sellers E: Genetically deficient CYP2D6 metabolism provides protection against oral opiate dependence. Pharmacogenetics 1997;7: 375-379 Ozdemir V, Naranjo C, Herrmann N, Reed K, Sellers E, Kalow W: Paroxetine potentiates the central nervous system side effect of perphenazine: Contribution of cytochrome P4502D6 inhibition in vivo. Clin Pharmacol Ther 1997;62:334-347 Wu D, Otton V, Inaba T, Kalow W, Sellers E: Interactions of amphetamine analogs with human liver CYP2D6. Biochem Pharmacol 1997;53:1605-1612 Dalen R Frengell C, Dahl M-L, Sjoqvist F: Quick onset of severe abdominal pain after codeine in an
Populations and Polymorphisms
40.
41.
42.
43.
44.
45.
ultrarapid metabolizer of debrisoquine. Ther Drug Monit 1997;19:543-544 Gaedfgk A, Blum M, Gaedigk R, Eichelbaum M, Meyer U: Deletion of the entire cytochrome P450 CYP2D6 gene as a cause of impaired drug metabolism in poor metabolizers of debrisoquine/sparteine polymorphism. A m J H u m Genet 1991;48:943-950 Oscarson M, McClellan R, Gullst6n U, et al.: Characterisation and PCR-based detection of a CYP2A6 gene deletion found at a high frequency in a Chinese population. FEBS Lett 1999;448:105-110 Gonzalez F, Skoda R, Kimura S, et al.: Characterization of the common genetic defect in humans deficient in debrisoquine metabolism. Nature 1988; 331:442-446 Goldstein J, de Morais S: Biochemistry and molecular biology of the human CYP2C family. Pharmacogenetics 1994;4:285-299 Bertina R, Koeleman B, Koster T, et al.: Mutation in blood coagulation factor V associated with resistance to activated protein C. Nature 1994;369:64-67 Feder J, Gnirke A, Thomas W, et al.: A novel M H C class l-like gene is mutated in patients with
46.
47.
48.
49.
*
Weber
307
hereditary haemochromatosis. Nat Genet 196;13: 399-408 McCarver D, Byun R, Hines R, Hichme M, Wegenek W: A genetic polymorphism in the regulatory sequences of human CYP2EI: Association with increased chloroxazone hydroxylation in the presence of obesity and ethanol intake. Toxicol Appl Pharmacol 1998;152:276-281 Bt~hler R, Hempel J, Von Wartburg J-R J6rnvall H: Human liver alcohol dehydrogenase: The unique properties of the "atypical" isozyme [32, [32-Bern can be explained by a single base mutation. Alcohol 1985;2:47-51 McCarver D, Thomasson H, Martier S, Sokol R, Li T: Alcohol dehydrogenase-2*3 allele protects against alcohol-related birth defects among African Americans. J Pharmacol Exp Ther 1997;283:1095-1101 McClellan R, Oscarson M, Alexandrie A-K, et al.: Characterization of a human glutathione S-transferase ix cluster containing a duplicated GSTM1 gene that causes ultrarapid enzyme activity. Mol Pharmacol 1997;52:958-965