* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Polygenic inheritance and micro/minisatellites
Ridge (biology) wikipedia , lookup
Epigenetics in learning and memory wikipedia , lookup
Oncogenomics wikipedia , lookup
Saethre–Chotzen syndrome wikipedia , lookup
Human genome wikipedia , lookup
Copy-number variation wikipedia , lookup
Non-coding DNA wikipedia , lookup
Genetic engineering wikipedia , lookup
Transposable element wikipedia , lookup
Biology and consumer behaviour wikipedia , lookup
Gene therapy of the human retina wikipedia , lookup
Genomic imprinting wikipedia , lookup
Dominance (genetics) wikipedia , lookup
Gene therapy wikipedia , lookup
Epigenetics of diabetes Type 2 wikipedia , lookup
Vectors in gene therapy wikipedia , lookup
Epigenetics of neurodegenerative diseases wikipedia , lookup
History of genetic engineering wikipedia , lookup
Neuronal ceroid lipofuscinosis wikipedia , lookup
Public health genomics wikipedia , lookup
Gene desert wikipedia , lookup
Gene nomenclature wikipedia , lookup
Genome evolution wikipedia , lookup
Gene expression programming wikipedia , lookup
Genome editing wikipedia , lookup
Quantitative trait locus wikipedia , lookup
Epigenetics of human development wikipedia , lookup
Genome (book) wikipedia , lookup
Nutriepigenomics wikipedia , lookup
Point mutation wikipedia , lookup
Site-specific recombinase technology wikipedia , lookup
Gene expression profiling wikipedia , lookup
Helitron (biology) wikipedia , lookup
Therapeutic gene modulation wikipedia , lookup
Artificial gene synthesis wikipedia , lookup
Molecular Psychiatry (1998) 3, 21–31 1998 Stockton Press All rights reserved 1359–4184/98 $12.00 PERSPECTIVE Polygenic inheritance and micro/minisatellites DE Comings Department of Medical Genetics, City of Hope Medical Center, Duarte, CA 91010, USA While it has often been stated that the identification of the genes involved in complex polygenic traits may be extremely difficult, the principles learned in the past century about single gene–single disease inheritance may not be relevant to polygenic inheritance. A new paradigm specific to complex disorders may be needed. It is proposed that micro- and minisatellite polymorphisms play a role in the expression of many genes. As a result, these genes exist in the population with many functional alleleomorphic variants. While each variant has only a modest effect on a given phenotype, because the variants are common, and most quantitative traits are controlled by a number of genes, there is a reasonable probability that an individual will inherit a threshold number of functional variants beyond which there is an appreciable effect on the phenotype. Twelve different aspects of such a new model for complex inheritance, some corollary implications, and three examples of its immediate application, are presented with the hope that the model may allow an acceleration of the identification of the genes involved in complex polygenic disorders. Keywords: microsatellites; minisatellites; polymorphisms; psychiatric genetics; Z-DNA; behavior; evolution Since the re-discovery of Mendel’s Laws in the early part of this century, the emphasis in human genetics has been on single gene disorders inherited in a dominant or recessive fashion. In the past two decades a large proportion of the genes for these disorders have been identified, localized, cloned, and sequenced. As the number of such genes remaining to be identified has decreased there has been an increased interest in the more common polygenic disorders. It has often been suggested that the genes involved in these disorders will be far more difficult to identify. This difficulty is well illustrated by the psychiatric disorders. Despite large numbers of linkage studies of manicdepressive disorder, schizophrenia, Tourette syndrome, panic disorder, autism, and others, with the possible exception of bipolar disorder,1 there have been few replicated findings. Part of the problem may be the very success of the studies of single gene disorders that are caused predominantly by exon mutations. In recessive disorders such mutations result in a total cessation of gene function, and in dominant disorders the effects are also profound. This single gene–single disease model has often led those studying polygenic disorders to focus on identifying mutations in the exons of candidate genes, either by sequencing2–4 or the use of special electrophoretic techniques.5 I propose here that the identification of the mutant genes involved in polygenic disorders could be hinCorrespondence: DE Comings, MD, Department of Medical Genetics, City of Hope Medical Center, Duarte, CA 91010, USA. E-mail: dcomingsKmindspring.com Received 17 January 1997; revised and accepted 15 April 1997 dered by the attempts to transfer the model of single gene inheritance to polygenic inheritance, and that a new paradigm, specific to polygenic disorders, may be needed to aid in the identification of the genes involved in these common entities. The following is a practical example of some of the issues involved. In 1990 Blum et al6 reported an association between the Taq A1 D2A1 allele of the DRD2 gene and severe alcoholism. This polymorphism is 39 to the DRD2 reading frame. This was a controversial finding because some confirmed it while others did not. Assuming that alcoholism is a complex, multifactorial, polygenic disorder, and that the DRD2 gene is just one of many genes that could be involved, this discrepancy in findings was not surprising. Given the role of dopamine as a modulator of many different behaviors,7 we8 suspected that the D2A1 allele might also be associated with other disorders as well and in recent years numerous studies have reported a significant association between alleles of the DRD2 gene and cocaine addiction, polysubstance abuse, smoking, attention deficit hyperactivity disorder (ADHD), Tourette syndrome, visual-perceptual disorders, conduct disorder, posttraumatic stress disorder, pathological gambling, and compulsive eating.9,10 Despite these associations, largely as a result of the application of the single gene–single disease model, many have remained skeptical of the role of the DRD2 gene in human behavioral disorders because sequencing studies have failed to find any mutations in the exons that could explain these findings. These findings could be explained if the D2A1 allele was in linkage disequilibrium with an unknown non-exon mutation that played a role in the regulation of DRD2 function.8 Polygenic inheritance and micro/minisatellites DE Comings 22 We and others have examined other genes (DRD1, DRD3, DRD4, OB, TDO2, MAOA, TRH and HTR2A) that also show significant associations between nearby polymorphisms and different behaviors, despite an absence of meaningful exon mutations. Micro/minisatellite polymorphisms in psychiatric genetics Minisatellites have been defined as repeat sequences of up to 65 base pairs in length.11 Microsatellites consist of shorter repeats variously defined as 2–5 bp in length. For the purposes of the present paper, unless specifically stated, we will use the term micro/minisatellites to cover both. If psychiatric disorders are polygenic, and each gene contributes to less than 10% of the variance of a given quantitative trait, then association studies may provide greater power than most linkage techniques.12 Although this technique can produce false positives due to population stratification, this can be minimized using the haplotype relative risk procedure13 with large numbers of subjects.14 One of the major impediments to the wider use of association studies is the lack of availability of suitable polymorphisms at or near the many candidate genes that have been cloned and sequenced.15 Because of the high frequency of micro/minisatellite repeats throughout the genome, we and others have often used these polymorphisms in association studies. Our initial assumption was that like the neutral or silent single base pair polymorphisms, the micro/minisatellite alleles would be in linkage disequilibrium with other ‘critical’ mutations that affect gene function. However, after working with these polymorphisms for several years we began to suspect that the micro/minisatellites themselves might be the ‘critical’ mutations. There are two aspects of these micro/minisatellite polymorphisms that are relevant – their mutation rate and their size. Mutation rate The mutation rate of micro/minisatellite alleles is higher than for non-repeat sequences.16 Lack of exchange of flanking markers suggests the new mutations are due to replication slippage or complex conversion-like events.17 A high mutation rate could make these polymorphisms less valuable for association studies since over many generations there would be too much noise introduced into linkage disequilibrium relationships which require that two different polymorphisms remain in phase for many generations. However, if the micro/minisatellite mutations were themselves the ‘critical’ alleles, they would actually be more powerful for association studies, despite the higher mutation rate. Allele size If linkage disequilibrium was involved in the association of micro/minisatellite alleles with specific phenotypes, each time a new mutation in the satellite occurred there would be a specified chance it was occurring on the same chromosome as the ‘critical’ alleles. However, on average there should be no trend for the longer vs the shorter mutant alleles to preferentially be in linkage disequilibrium with these ‘critical’ alleles. However, if the size of the repeats played a role in gene regulation (see below) there should be a trend for an association of quantitative traits with groups of different sized alleles rather than with specific individual alleles or allele sequences. If there is a major peak of repeat gene frequency, both the shorter and the longer alleles may be associated with phenotypic effects. In our published18–23 and unpublished results of studies of behavioral phenotypes associated with micro/minisatellite polymorphisms at different neuropsychiatric candidate genes we have found a significant association between the shorter or longer alleles with various quantitative behavioral traits and mini- or microsatellites at the following genes: MAOA, MAOB, HTR1A, DAT1, DRD4, HRAS, HTT, OB, CNR1, GABRA3, GABRB3, FRAXA, and NO. Some examples are shown in Figure 1. Others have reported significant phenotypic behavioral effects with specific size alleles of the same polymorphisms of the DAT124,25, DRD4,26– 29 HRAS,30–32 HTT,33,34 INS35–38 and DBH39 genes. While these studies do not rule out the presence of an important role of single base pair changes in a subset of these length variants, (see below and28,40,41), they are consistent with the possibility that both length and sequence can be involved, and that the different sized micro/minisatellite alleles might have an effect on gene function. This hypothesis would be much more plausible if different micro/minisatellite alleles could be shown to have an effect on the rate of transcription or translation of the genes. There are, in fact, many examples of this. Evidence that minisatellites affect gene function The trinucleotide repeats The most dramatic evidence for an association between repeat sequences and disease comes from the involvement of long triplet repeats in a variety of neurological disorders42 including fragile-X syndrome, Huntington’s disease,43 myotonia dystrophica, Kennedy’s disease, Friedreich’s ataxia,44 and others.42 At least five of these disorders involve intronic GAG repeats producing polyglutamine tracts in the amino acid sequence of the respective gene products. A clue to the mechanism by which polyglutamine tracts affect cell function was provided by the observation that the hungingtin protein binds other cellular proteins, such as hungingtin associated protein 1 (HAP1).45 Glyceraldehyde-3 phosphate dehydrogenase, a critical enzyme in glycolysis, was also found to bind to the huntingtin polyglutamine tracts.46 This enzyme is of particular importance for neurological disorders because the brain relies almost exclusively on glucose for its energy. Other proposed mechanisms by which increasing lengths of trinucleotide repeats affect gene function include decreases in mRNA and protein production,47 Polygenic inheritance and micro/minisatellites DE Comings increased DNA methylation,48 effects on nucleosome assembly,49 decreased RNA polymerase efficiency,50 competition between DNA binding proteins and transcription factors,51 regulatory squelching,52 gene silencing,53 and the formation of complex, non-B DNA three-dimensional structures.50,54,55 Some of the remarkable aspects of the above observations are that the CTG repeat silenced the growth inhibitory factor/metallothionein III gene promoter whether it was upstream or at a distance downstream from the promoter,53 the efficiency of nucleosome formation increased according to the size of CTG repeats suggesting such blocks may repress transcription through the creation of stable nucleosomes,49 and the reduction of androgen gene expression remained proportionate to the number of CAG repeats even over the range of normal alleles.47 While I have not included the extremely long triplet repeats in the proposed hypothesis of polygenic inheritance, these studies do introduce the concept that the varying length of repeat alleles, even when they are well into ‘normal’ range, may affect gene function through a wide variety of mechanisms. The binding of nuclear proteins by minisatellites. The HRAS gene The first evidence that minisatellite repeats could affect the function of neighboring genes by binding transcription factors came from the studies of Krontiris and colleagues of the minisatellite located near the HRAS gene.56–58 This 28-bp repeat is located 1 kb 39 to the HRAS gene.59 Rare alleles at this locus are significantly associated with a number of cancers.56 This minisatellite displays transcriptional enhancer activity60,61 which occurs because it binds four members of rel/NF-kB transcriptional factors.56,58 Since C → G exchanges occurred throughout the minisatellite, the increased enhancer activity of some alleles could have been due to sequence differences producing secondary structure variations, as well as length differences.57 Studies also showed that the enhancer activity of the HRAS minisatellite could be dramatically upregulated by interaction with a transcription factor unrelated to rel/NF-kB.57 While the studies of Krontiris were based on the unique sequence of rare variants, we have shown that when the common HRAS alleles were grouped simply on the basis of size, there was an effect of the resultant genotypes on behavioral variables.19 IGH gene In an extension of these studies, Trepicchio and Krontiris62 also examined the minisatellite located in the DHJH junction of the human immunoglobulin heavy chain (IGH) gene. This minisatellite, located 1500 bp upstream of the JH genes, was found to bind a nuclear protein factor present in a wide variety of cell types. The binding site contained the mycHLH motif, CACGTG, closely related to the USF/MLTF binding site adjacent to the adenovirus major late promoter. The binding of the nuclear protein factor resulted in its sequestration to an inactive form, causing a suppression of transcriptional activation. Insulin gene minisatellite Another example of the role of repeats in gene regulation comes from studies of a 14-bp repeat minisatellite located 596 bp 59 to the human insulin gene. In family studies, Bennett et al35 observed that of the Class I minisatellite alleles the most common 814 mobility unit (mu) allele was infrequently transmitted to diabetic offspring, while the shorter 641–756 mu and the longer 828 mu alleles were transmitted significantly more often. While there were also numerous single base pair RFLPs in the region, the studies indicated the most significant effects were seen with the minisatellite alleles. The primary involvement of the minisatellite was supported by gene expression studies of pancreatic RNA from several subjects with different class I alleles. Kennedy et al36 examined the transcriptional effects of alleles of this minisatellite in further detail. They examined the shorter class I alleles (average 40 repeats) and the longer class III alleles (average 157 repeats) using transvection of plasmid constructs linked to CAT or luciferase to assay transcriptional activity. This showed dramatic differences based both on the length and the sequence of the repeat. Alleles where the repeat sequences differed by a single base pair substitution varied by as much as 10–260% of the rate of transcription of the standard allele. The insulin minisatellite contains high-affinity binding sites for the transcription factor, Pur-1,36 and is capable of adopting a number of variant DNA conformations.63,64 Recently Vafiadis et al38 and Pugliese et al37 have shown that the transcription levels of the INS gene in the thymus correlate with alleleic variation of the VNTR-IDDM2 locus and suggest this plays a role in the selection of insulin-specific T-lymphocytes which in turn play a role in the pathogenesis of type-1 diabetes. Msbp-4 A minisatellite-binding protein called Msbp-4 was isolated from mouse testis and tumor cells. Msbp-4 binding favored multiple copies of the PC-2 repeat sequence, GGCAGGA, and showed sequence and strand specificity.65,66 Other minisatellite binding proteins have also been reported.67 Effect of 39 UTR repeats on translation Repeat sequences in the 39 untranslated regions of the 15-lipoxygenase (LOX) gene were shown to bind a 48kDa reticulcyte protein that specifically inhibited the translation of LOX.68 Other examples of repeat sequences playing a role in regulation of translation have been reported.69 Evidence that microsatellites affect gene function An important aspect of sequences containing alternating purines and pyrimidines, such as (CA)n microsatellites, is that they are capable of forming left-handed or Z-DNA under physiological conditions.70,71 Z-DNA is 23 Polygenic inheritance and micro/minisatellites DE Comings 24 Figure 1 Illustrative examples of the association between alleles of repeat polymorphisms at a variety of genes and quantitative behavioral traits. (a) Monoamine oxidase A gene (MAOA) and VNTR polymorphism alleles divided into four groups. There was a significant association with a number of quantitative variables relating to specific symptoms. The results are shown for a manic symptoms score in 351 Tourette syndrome probands, their relatives and controls. There was a significant increase in the score in subjects carrying the longest alleles.23 (b) Use of the same polymorphism in a group of substance abusers and controls. There was again a significant association with the longest alleles and a drug abuse score.23 (c) Association between a quantitative score for anxiety, based on the SCL-90 test, and the obesity gene (OB) microsatellite polymorphism alleles divided into subjects homozygous for the ,208-bp alleles (left) and individuals homozygous or heterozygous for the $208-bp alleles (right).20 (d) Association between the amplitude of the event-related potential (ERP) P300 wave (in uvolts) and the CNR1 cannabinoid receptor gene microsatellite polymorphism alleles divided into subjects homozygous for the $5 repeat alleles (right) and individuals homozygous or heterozygous for the ,5 repeat alleles (left).21 (e) Association between the Brown Adult ADD score and the GABAA B3 (GABRB3) receptor gene microsatellite polymorphism alleles divided into subjects homozygous for $185-bp alleles (left) and those homozygous or heterozygous for the ,185-bp alleles. (f) Association between performance IQ and the normal alleles of the FRAXA locus in random adults assessed for IQ.104 characterized by a zig-zag conformation of the phosphate backbone (Figure 2).72 The major groove is absent in Z-DNA and the minor groove is much deeper than in B-DNA, extending to the axis of the helix and exposing the bases.72 Hybridization studies have shown that the human genome contains up to 100 000 regions of poly (CA)n sequences, and that such sequences are common in all eukaryotes including Drosophila and yeast.73 Studies of 137 human DNA genes showed that Z-DNA forming Polygenic inheritance and micro/minisatellites DE Comings Figure 2 Conformation of Z-DNA and B-DNA showing the zig-zag nature of the phosphate backbone (dark spheres) and the exposure of the base pairs (shaded spheres) in the deep groove of Z-DNA (from Ref. 76 by permission). sequences were non-randomly located with 35% occurring at the 59 end of genes, often in promotor regions, compared to 3% at the 39 end. Of the remaining sequences, 47% were found in introns vs 15% in exons.74 The number of Z-DNA domains associated with each gene ranged from 0 to 14. When the 13 000bp a-globin gene cluster, and the 73 326-bp of the bglobin cluster were examined, there was a dramatic tendency for Z-DNA regions to cluster around sites of initiation of transcription, and a virtual absence of ZDNA regions in intergenic sequences and at pseudogenes. These studies also showed that Z-DNA motifs may be present up to 1.5 kb 59 upstream of the transcribed sequences. Z-DNA and the regulation of transcription A number of other studies have also suggested a role of Z-DNA in gene function. DNAse I hypersensitivity is associated with transcriptionally active chromatin. Fine structure mapping of the DNAse I hypersensitive sites of the enhancer region of the SV40 minichromosome showed the presence of DNAase I cleavage approximately 25 bp on either side of the segments forming Z-DNA.75 The position of these hypersensitive sites may have been determined, in part, by Z-DNA binding proteins.75 A survey of other transcriptional enhancers showed that pairs of Z-DNA segments were present in a number of transcriptional enhancers with 50–80 bp between the Z-DNA segments.70,75 Rich et al76 suggested that the spontaneous conversion of the ZDNA segments to B-DNA could produce a local increase in negative supercoiling that may alter local chromatin structure and facilitate entry of RNA polymerase into the molecule. Hamada et al77 also proposed that by reversibly interconverting between the B and the Z forms in vivo the (CA)n sequences could distort DNA and result in the activation or inactivation of genes.77 Banerjee et al78 examined the ability of various synthetic polymers to affect the transfection of herpes simplex virus thymidine kinase in LtK cells. Potential ZDNA forming polymers cotransfected with the tk gene decreased the level of Tk+ transformed colonies, while cotransfection with polymers that do not form Z-DNA had no effect. They found the effect was dependent upon the length of the microsatellites linked to the 39 or 59 end of the tk gene. Only the longer sequences, capable of forming Z-DNA, had an effect. In a subsequent report, Banerjee and Grunberger79 found that certain polynucleotides, capable of forming Z-DNA, enhanced the transcription of the chloramphenicol acetyltransferase gene when cotransfected into murine L cells. ¨ Wolfl et al80 reported studies of the corticotropinreleasing hormone (CHR) gene with an intron that contained 100 bp of alternating purine/pyrimidine (CA)n residues. These studies showed that the changes in ZDNA formation were linked to the rate of transcription of the CHR gene and the action was gene specific. The authors concluded that the presence of a Z-DNA motif in the non-coding parts of all sequenced mammalian CHR genes make it reasonable to consider that Z-DNA may contribute to the regulation of the CHR gene.80 RNA polymerase cannot transcribe through ZDNA,81,82 and topoisomerase II and other proteins show enhanced affinity for Z-DNA.76,83,84 A number of nuclear proteins bind to Z-DNA but not to B-DNA.85 This may be related to the different conformation of the two forms. The Z-DNA conformation exposes the bases to interaction with proteins, while the B-form does not (Figure 2). It has been suggested80,86 that ZDNA motifs at the 59 end of genes play an important role in increasing the spacing of RNA polymerases and that correct spacing is required to prevent transplicing or erroneous splicing, thus maximizing the efficiency of the RNA splicing process. Five other studies have also indicated a role of (CA)n repeats in gene regulation. First, recent studies have shown that 59-UTR Z-DNA forming poly(TG)n tracts play a role in the regulation of RNA editing at the GLuR gene.87,88 These results suggest a significant role of intronic sequences in controlling gene expression.87 Second, studies by Epplen and colleagues89 showed that (CA)n microsatellites bind nuclear proteins in humans and that binding vs non-binding is dependent upon the number of repeats. They found that (CA)6 did not bind nuclear proteins, (CA)13 bound with moderate intensity, and (CA)25 bound nuclear proteins with the greatest intensity. They suggested that simple repeats could have subtle effects on gene expression via their differential protein binding with respect to the genetic background of multifactorial diseases. Third, a study of 125 unrelated healthy individuals showed a signifi- 25 Polygenic inheritance and micro/minisatellites DE Comings 26 cant (P , 0.0001) correlation between the serum levels of the dopamine b-hydroxylase and homozygosity for two different sized alleles of an associated (CA)n repeat polymorphism.39 Fourth, a (CA)n repeat upstream of the rat prolactin gene forms Z-DNA and inhibits transcription.90 Fifth, deletion of the (CA)n repeat at the human cytosolic phospholipase A2 gene showed it had an inhibitory effect on promoter activity.91 In summary, there are many examples where micro/minisatellite sequences play a significant role in gene regulation by altering the primary, secondary or tertiary structure of DNA, by binding to transcription or translation factors, or by affecting RNA editing. In many studies the length of the repeat sequences affected the affinity for proteins involved in transcription or translation, determined the probability that a sequence would form Z-DNA, or determined whether an enhancer sequence would adopt a conformation suitable for the entry of RNA polymerase. For example, in the study by Schroth et al,74 using a computer algorithm to evaluate a Z-score indicating the potential to form Z-DNA, the score was 0.6 for (CA)6, 1.3 for (CA)7 and 2.4 for (CA)8, ie the score doubled with each additional dinucleotide. Modest variations in the sequence also produced major changes in the Z-score. For example, the synthetic sequence (CGCG)4 had a Zscore of 17 300, while the synthetic sequence (CGCC)4 had a Z-score of 5, or 3460 times lower than that of (CGCG)4. In addition, the sequence (TGTG)4 had a Zscore of 2.4, while the sequence (TGCG)4 had a Z-score of 135. This tetranucleotide sequence was also sensitive to length in that the Z-score for (TGCG)3 was 16.7, or eight times lower than that of (TGCG)4. These results, illustrated in Figure 3, support the potential for alleles of varying length and sequence of polymorphic micro- and minisatellite sequences to be associated with variations in the level of expression of specific genes. A general hypothesis of polygenic inheritance The various aspects of a micro/minisatellite hypothesis of polygenic inheritance are listed as follows: (1) Many micro/minisatellites have an effect on the expression of the genes with which they are associated. (2) Many genes are associated with one or more micro/minisatellite polymorphisms. (3) As a result of No. 1 and No. 2, many genes will present with a range of functional alleleomorphic variants. Those genes associated with several micro/minisatellite polymorphisms, each with multiple alleles, will present with an especially large number of functional haplotypes. While most genes will be associated with ±10–15% of the average level of gene activity, the range may be as high as ±40%, and if multiple micro/minisatellite polymorphisms are involved their effects can be additive. (4) Significant functional variants are common in the Figure 3 Z-DNA scores for repetitious DNA sequences based on variations in length and sequence. Based on data of Schroth et al.74 general population, with prevalences ranging from 1 to 100%. A 100% prevalence can occur if there is a bimodal distribution of alleles consisting entirely of hyperfunctional alleles (.10% of the average) and hypofunctional variants (,10% of the average). (5) Individually these functional variants have only a modest effect on the phenotype (0.5–8% of the variance), and rarely cause disease, ie they are generally not responsible for the classic one gene– one disease autosomal dominant or recessive disorders. (6) Since most of the micro/minisatellite polymorphisms are not in exons, the mutations in polygenic disorders are usually outside the exons and often outside the transcribed sequences. (7) Log score linkage studies lose power when more than six genes are involved in a given disorder.12,92 Linkage studies are based on the assumption that the presence of a given allele is associated with the presence of the disorder and the absence of the allele is associated with the absence of the disorder. By contrast, in polygenic inheritance the disorder is associated with the presence of a threshold number of alleles of several different genes. As such, many members of a pedigree may carry a given mutant gene but not have the disorder, because of the absence of the necessary threshold number of other mutant Polygenic inheritance and micro/minisatellites DE Comings (8) (9) (10) (11) (12) genes. Thus, many members of a family may carry a specific gene and not have the disorder, and other members of the family with the disorder may not carry the mutant allele.93 Non-parametric approaches to linkage94 are better suited to complex inheritance (but see95). However, when a given gene accounts for less than 8% of the variance, a large number of parent–child sets or sibpairs must be examined.96 Association studies, comparing the frequency of the mutant candidate genes in severely affected probands to totally unrelated, ethnically matched controls that are free of the disease, can identify these small effects.92,93,97 Both hyper- and hypo-functional alleles can have an effect on the phenotype. For example, in psychiatric genetics, both an increase in the expression of a given receptor gene leading to receptor supersensitivity, or a decrease in the expression of a receptor gene leading to receptor hyposensitivity, could be associated with an altered phenotype. A polygenic disorder occurs when an individual inherits a threshold number of hypo- or hyperfunctional genes affecting the same phenotype. Thus, if 20 different polygenes play a role in a given quantitative variable, anyone with 10 or more of these hypo- or hyperfunctional variants would present with a significantly altered phenotype. The threshold number would vary according to the degree of hypo- or hyper-functionality of the alleles, and the severity of the altered phenotype would depend on the extent to which the number of polygenes exceeded a critical threshold number. Because micro/minisatellites are so common, the probability of No. 9 occurring is relatively high. Thus, polygenic disorders are much more common than single gene disorders. With the general population frequency of many polygenes ranging from 25% to over 50% the chance of inheriting a threshold number is much higher than for the single gene disorders. For a given micro/minisatellite the alleles can have a negative or positive effect on the phenotype depending upon the genetic background of other genes and on the nature of the phenotype. A practical example is the observation by Gelernter et al25 of an association between cocaine-induced paranoia and the 9 allele of the dopamine transporter gene (DAT1) gene. By contrast, Cook et al24 and Comings et al18 observed an association between the 10 allele of the DAT1 gene and attention deficit hyperactivity disorder. One might conclude from such divergent reports that one or the other observation must be incorrect. However, there is a reasonable probability that both are correct, and that different alleles of a given micro/minisatellite may be associated with different phenotypes. A reasonable approach to the study of polygenic disorders is to choose candidate genes and examine the effect of alleles of the nearest micro/minisatellite polymorphisms on relevant quantitative traits. Implications of the micro/minisatellite paradigm for polygenic inheritance The above proposed characteristics of polygenic inheritance carry a number of additional implications. These are as follows: Many genes can be involved in the polygenic inheritance of some trait If many micro/minisatellite polymorphisms have the potential of altering the rates of transcription or translation of the gene they are closest to, and if many genes are associated with at least one micro/minisatellite, then many and possibly most genes have the potential to contribute in some degree to the polygenic inheritance of the phenotypes they control. Exon mutations may provide only a minor contribution to polygenic inheritance Many studies of the potential role of specific genes in psychiatric or other disorders start with a search for mutations in the exons of a candidate gene. When such studies are negative it is often concluded that the gene has nothing to do with the disorder being examined. If the majority of the mutations involved in complex, polygenic traits are in non-exons, such conclusions would be invalid. The present hypothesis does not exclude a role for exon mutations that have only a minor effect on the function of the gene, it only emphasizes that these studies can be negative despite the presence of functionally important variants of that gene. The study of the additive effects of several genes will be informative If on average, each candidate gene contributes to only 0.5–8% of the variance of a given quantitative trait, depending upon the phenotype and the size of the study, the results may be of modest, borderline or negative significance. However, the examination of the additive or epistatic effect of two or more candidate genes may show a much more significant effect. We have observed this for the additive effect of the DRD2, DbH and DAT genes,18 the DRD1 and DRD2 genes,98 the OB and DRD2 genes,20 and other gene combinations (unpublished). Simply comparing allele frequencies in controls vs subjects may be inadequate Many association studies simply examine the frequency of specific alleles in subjects compared to controls. However, when a diagnosis is based on a complex set of symptoms, a given allele may show a significant association with several of the quantitative traits involved in the diagnosis, but not the diagnosis itself. For example, if schizophrenia is a polygenic disorder, the alleles of a specific candidate gene may show a significant association with negative symptoms (eg affective flattening), which are present in 27 Polygenic inheritance and micro/minisatellites DE Comings 28 some but not all cases, but no association with positive symptoms (eg delusions or hallucinations), or with the diagnosis of schizophrenia itself. Limiting the analysis to simple dichotomous diagnoses, rather than the examination of sub-syndromal quantitative traits, may miss genes that make an important contribution to a polygenic disorder. Heterosis may also allow a gene to have a significant effect on the phenotype in the absence of difference in the frequency of individual alleles in controls vs patients.99 Evolution may not need to wait for mutations, they are always there One of the tenets of evolution is a dependence upon the occurrence of just the right rare mutations that instead of being deleterious to the species, improve it, and allow selection to produce a new and better adapted species. This leaves open the question of how such rare events ie mutations that affect the right gene, are advantageous instead of deleterious, happen at the right time, and have occurred often enough to provide for the rich diversity of species. The concept of random drift helps some but that seems best suited to explain the accumulation and fixation over time of neutral base pair and amino acid changes. The concept of micro/minisatellite polymorphisms providing many alleleomorphic functional variants of each gene eliminates the need for this improbable sequence of events. Thus, new mutations or alleles do not have to occur, they are already present, and in great abundance, and affecting most genes. It is not necessary for that rare advantageous mutation to just happen to occur at the right time and in the right gene since all genes come in a rich array of functional variants. What is necessary to drive evolution is an environmental or ecological change to fuel the selection of a given set of polygenes better suited for the new environment. As an example, the evolution from primates to Homo sapiens required many different changes, including standing up-right, thumb-finger apposition, the development of speech centers, the expansion of the frontal lobes, and many others. There is evidence that a number of different hominid species were simultaneously and independently undergoing these changes.100 To accept that multiple advantageous new mutations of many different genes just happened to occur in the right temporal order, and in several different lineages, is difficult. However, if all the necessary mutations were already present, and environmental changes selected for the polygenic sets that were most advantageous, the parallel progression in multiple lineages is more readily understandable. This is not meant to imply that exon mutations in some critical genes do not also play a role in evolution, only that the presence of many functional alleles associated with micro-minisatellites may be a better explanation. Microsatellites and minisatellites may both involve ZDNA motifs While Z-DNA forming (CA)n repeats represent the major class of microsatellite polymorphisms, computer programs designed to identify Z-DNA motifs show that a wide range of sequences, in addition to dinucleotide repeats, also form Z-DNA.74 If these longer sequences were involved in minisatellite repeat polymorphisms, they could also play a significant role of gene regulation. The suggestion that minisatellites (and Alu repeats101) are the evolutionary progeny of microsatellites11 further testifies to their interrelatedness. Immediate practical applications A hypothesis is most useful if it has immediate heuristic value. The following are examples of such value. (1) Resolution of the statistical problem of highly polymorphic alleles One of the disadvantages of the use of highly polymorphic micro/minisatellite polymorphisms in association studies is that when each of the large number of alleles, and even larger number of potential genotypes is examined, the power of the study may be so seriously compromised as to render it useless. However, if the size of the repeats is the critical variable, even the most complex of polymorphisms can be reduced to three genotypes consisting of homozygosity for the shorter alleles, homozygosity for the longer alleles, and heterozygotes. We have found this approach of value for studies of the role of a number of genes (OB, MAOA, MAOB, CNR1, GABRA3, GABRB3, DBH, FRAXA, and NOS1) in behavioral traits. (2) The identification of polymorphisms for candidate genes using genomic clones While many of the potentially interesting candidate genes have been cloned and sequenced, for most, no polymorphisms have been reported in the Genome Data Base, making association studies impossible. Based on the present hypothesis, if the sequence is known, one method of identifying useful micro/minisatellites is to obtain large genomic clones carrying the gene of interest from commercial sources. These can then be screened for repeat sequences which may be highly informative in association studies. (3) The identification of polymorphisms for candidate genes using Z-DNA screening The present model suggests that the repeat sequences capable of forming ZDNA will prove to be the most valuable polymorphisms in studies of polygenic inheritance. This suggests that screening known sequences using the Zhunt-II program,74 could provide important clues about the location of the most informative polymorphic regions. To test this, we utilized this program to examine the sequence of NOS1.102 Based on the results of studies showing aggression in knock-out mice, without the NOS1 gene,103 we had examined various behavioral traits in humans using a repeat polymorphism we identified by visual examination of the sequence of the NOS1 gene.102 Since some associations were observed, we wondered if: (a) the polymorphism we were using would be identified by the Z-hunt II program; and (b) if there were other regions in the gene that might contain informative polymorphisms. Both predictions were Polygenic inheritance and micro/minisatellites DE Comings true. Based on the assumption that the various lengths of the NOS1 (CA)n alleles would play a role in the function of the NOS1 gene, we divided the alleles into two groups consisting of the shortest 50% of the alleles (,199 bp) and the longest 50% of the alleles ($199 bp). Preliminary studies suggest some associations with behavioral phenotypes. We then screened the published sequence of the NOS1 gene for Z-DNA regions using the Z-hunt-II algorithm.72 This showed that the polymorphism we were using was one of three regions of high Z-DNA content. A second region with an even higher Z-score identified a new previously undetected polymorphic region. These three examples illustrate in practical terms how the predictions of the model may lead to the acceleration of the identification of the genes involved in polygenic disorders. The use of this model in conjunction with the sequencing data soon to be available from the Human Genome Project may contribute greatly to our knowledge about polygenic disorders through an increased ability to identify the presence of micro/minisatellites in proximity to the known candidate genes. Acknowledgements I thank PS Ho, Oregon State University, Corvallis, OR for performing the Z-hunt-II screening. Supported in part by the National Institutes of Drug Abuse grant RO1-DA08417 and Tobacco Related Research Disease Program grant 4RT-0110. References 1 Risch N, Botstein D. A manic depressive history. Nature Genet 1996; 12: 351–353. 2 Sarkar G, Kapelner S, Grandy DK, Marchionni M, Civelli O, Sobell J, Heston L, Sommer SS. Direct sequencing of the dopamine D2 receptor (DRD2) in schizophrenics reveals three polymorphisms but no structural change in the receptor. Genomics 1991; 11: 8–14. 3 Liu Q, Sobell JL, Heston LL, Sommer SS. Screening the dopamine D1 receptor gene in 131 schizophrenics and eight alcoholics: identification of polymorphisms but lack of functionally significant sequence changes. Am J Med Gen (Neuropsych Genet) 1995; 60: 165–171. 4 Sobell JL, Heston LL, Sommer SS. Delineation of genetic predisposition to multifactorial disease: a general approach on the threshold of feasibility. Genomics 1991; 12: 1–6. 5 Grompe M. The rapid detection of unknown mutations in nucleic acids. Nature Genet 1993; 5: 111–117. 6 Blum K, Noble EP, Sheridan PJ, Montgomery A, Ritchie T, Jadadeeswaran P, Nogami H, Briggs AH, Cohn JB. Allelic association of human dopamine D2 receptor gene in alcoholism. J Am Med Assn 1990; 263: 2055–2059. 7 LeMoal M, Simon H. Mesocorticolimbic dopaminergic network: functional and regulatory roles. Physiol Rev 1991; 71: 155–234. 8 Comings DE, Comings BG, Muhleman D, Dietz G, Shahbahrami B, Tast D, Knell E, Kocsis P, Baumgarten R, Kovacs BW, Levy DL, Smith M, Kane JM, Lieberman JA, Klein DN, MacMurray J, Tosk J, Sverd J, Gysin R, Flanagan S. The dopamine D2 receptor locus as a modifying gene in neuropsychiatric disorders. J Am Med Assn 1991; 266: 1793–1800. 9 Blum K, Wood RC, Sheridan PJ, Chen T, Comings DE. Dopamine D2 receptor gene variants: association and linkage studies in impulsive, addictive and compulsive disorders. Pharmacogenetics 1995; 5: 121–141. 10 Blum K, Cull JG, Braverman ER, Comings DE. Reward Deficiency Syndrome. American Scientist 1996; 84: 132–145. 11 Wright JM. Mutation at VNTRs: are minisatellites the evolutionary progeny of microsatellites. Genome 1994; 37: 345–346. 12 Risch N, Merikangas K. The future of genetic studies of complex human diseases. Science 1996; 273: 1516–1517. 13 Falk CT, Rubinstein P. Haplotype relative risks: an easy reliable way to construct a proper control sample for risk calculations. Ann Hum Genet 1987; 51: 227–233. 14 Comings DE. The haplotype relative risk technique lacks power in polygenic inheritance. 1995 World Congress Psychiatric Genetics 1995; 5: 103. 15 Comings DE. Candidate genes and association studies in psychiatry. (Letter to the editor). Am J Med Gen (Neuropsych Genet) 1994; 54: 324. 16 Jeffreys AJ, Royle NJ, Wilson V, Wong Z. Spontaneous mutation rates to new length alleles at tandem-repetitive hypervariable loci in human DNA. Nature 1987; 332: 278–281. 17 Wolff RK, Plaetke R, Jeffreys AJ, White R. Unequal crossingover between homologous chromosomes is not the major mechanism involved in the generation of new alleles at VNTR loci. Genomics 1989; 5: 382–384. 18 Comings DE, Wu H, Chiu C, Ring RH, Dietz G, Muhleman D. Polygenic inheritance of Tourette syndrome, stuttering, ADHD, conduct and oppositional defiant disorder: the additive and subtractive effect of the three dopaminergic genes – DRD2, DbH and DAT1. Am J Med Gen (Neuropsych Genet) 1996; 67: 264–288. 19 Comings DE, Wi J, Chiu C, Muhleman D, Sverd J. Studies of cHarvey-Ras gene in psychiatric disorders. Psychiatry Res 1996; 63: 25–32. 20 Comings DE, MacMurray JP, Gade R, Muhleman D, Peters WR. Genetic variants of the human obesity gene: association with psychiatric symptoms and body mass index in young women, and interaction with the dopamine D2 receptor gene. Mol Psychiatry 1996; 1: 325–335. 21 Johnson JP, Muhleman D, MacMurray J, Gade R, Verde R, Ask M, Kelley J, Comings DE. Association between the cannabinoid receptor gene (CNR1), and the P300 event-related potential. Mol Psychiatry 1997; 2 169–171. 22 Comings DE, Muhleman D, Gade R, Johnson P, Verde R, Saucier G, MacMurray J. Cannabinoid receptor gene (CNR1): association with IV drug use. Mol Psychiatry 1997; 2 161–168. 23 Gade R, Muhleman D, Blake H, MacMurray J, Johnson P, Verde R, Saucier G, McGue M, Lykken D, Comings DE. Correlation of length of VNTR alleles are the X-linked MAOA gene and phenotypic effect in Tourette syndrome and drug abuse. Mol Psychiatry 1997; (submitted). 24 Cook EH, Stein MA, Krasowski MD, Cox NJ, Olkon DM, Kieffer JE, Leventhal BL. Association of attention-deficit disorder and the dopamine transporter gene. Am J Hum Genet 1995; 56: 993–998. 25 Gelernter J, Krazler HR, Satel SL, Rao PA. Genetic association between dopamine transporter protein alleles and cocaineinduced paranoia. Neuropsychopharmacology 1994; 11: 195–200. 26 Benjamin J, Li L, Patterson C, Greenberg BD, Murphy DL, Hamer DH. Population and familial association between the D4 dopamine receptor gene and measures of novelty seeking. Nature Genet 1996; 12: 81–84. 27 Ebstein RP, Novick O, Umansky R, Priel B, Osher Y, Blaine D, Bennett ER, Nemanov L, Katz M, Belmaker RH. Dopamine D4 receptor (D4DR) exon III polymorphism associated with the human personality trait of novelty seeking. Nature Genet 1996; 12: 78–80. 28 Grice DE, Leckman JF, Pauls DL, Kurlan R, Kidd KK, Pakstis AJ, Chang FM, Buxbaum JD, Cohen DJ, Gelernter J. Linkage disequilibrium of an allele at the dopamine D4 receptor locus with Tourette’s syndrome by TDT. Am J Hum Genet 1996; 59: 644–652. 29 Lahoste GJ, Swanson JM, Wigal SB, Glabe C, Wigal T, King N, Kennedy JL. Dopamine D4 receptor gene polymorphism is associated with attention deficit hyperactivity disorder. Mol Psychiatry 1996; 1: 121–124. ´ ´ ´ ¨ 30 Herault J, Perrot A, Barthelemy C, Buchler M, Cherpi C, Leboyer ¨ M, Sauvage D, Lelord G, Mallet J, Muh J-P. Possible association of C-Harvey-Ras-1 (HRAS-1) marker with autism. Psychiatry Res 1993; 46: 261–267. 29 Polygenic inheritance and micro/minisatellites DE Comings 30 31 Eggers B, Kurth MC, Kurth JH. Allele frequencies of dopamine receptors DRD1 and DRD2 in Parkinson’s disease populations. Am J Hum Genet 1995; 57: A162. 32 Thelu M-A, Zarski J-P, Froissart B, Rachail M, Seigneurin J-M. cHa-ras polymorphism in patients with hepatocellular carcinoma. Gastroenterol Clin Biol 1993; 17: 903–907. 33 Ogilvie AD, Battersby S, Bubb VJ, Fink G, Hamaar AJ, Goodwin GM, Smith CAD. Polymorphism in serotonin transporter gene associated with susceptibility to major depression. Lancet 1996; 347: 731–733. 34 Lesch K-P, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S, Benjamin J, Muller CR, Hamer DH, Murphy DL. Association of anxiety-related traits with a polymorphism in the serotonin transporter gene regulatory region. Science 1996; 274: 1527–1531. 35 Bennett ST, Lucassen AM, Grough SCL, Powell EE, Undlien DE, Pritchard LE, Merriman ME, Kawaguchi Y, Dronsfeld MJ, Pociot F, Nerup J, Bouzekri N, Cambon-Thomsen A, Ronning KS, Barnett AH, Bain SC, Todd JA. Susceptibility to human type 1 diabetes at IDDM2 is determined by tandem repeat variation at the insulin gene minisatellite locus. Nature Genet 1955; 9: 284–292. 36 Kennedy GC, German MS, Rutter WJ. The minisatellite in the diabetes susceptibility locus IDDM2 regulates insulin transcription. Nature Genet 1995; 9: 293–298. 37 Pugliese A, Zeller M, Fewrnandez A Jr, Zalcberg L, Bartlett RJ, Ricordi C, Pietropaolo M, Eisenbarth GS, Bennett ST, Patel DD. The insulin gene is transcribed in the human thymus and transcription levels correlate with allelic variation at the INV VNTRIDDM2 susceptibility locus for type 1 diabetes. Nature Genet 1997; 15: 293–297. 38 Vafiadis P, Bennett ST, Todd JA, Nadeau J, Grabs R, Goodyer CG, Wickramasinghe S, Colle E, Polychronakos C. Insulin expression in human thymus is modulated by INS VNTR alleles at the IDDM2 locus. Nature Genet 1997; 15: 289–292. 39 Wei J, Ramchand CN, Hemmings GP. Possible control of dopamine b-hydroxylase via a codominant mechanism associated with polymorphic (GT)n repeat at this gene locus in healthy individuals. Hum Genet 1997; 99: 52–55. 40 Lichter JB, Barr CL, Kennedy JL, Van Tol HHM, Kidd KK, Livak KJ. A hypervariable segment in the human dopamine receptor (DRD4) gene. Hum Molec Genet 1993; 2: 767–773. 41 Krontiris TG, DiMartino NA, Colb M, Parkinson DR. Unique allelic restriction fragments of the human Ha-ras locus in leukocyte and tumor DNAs of cancer patients. Nature 1985; 313: 369–374. 42 Caskey CT, Pizzuti A, Fu YH, Fenwick RG Jr, Nelson DL. Triplet repeat mutations in human disease. Science 1992; 256: 784–789. 43 Huntington’s Disease Collaborative Research Group. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington’s disease chromosomes. Cell 1993; 72: 971–983. ` ´ 44 Campuzano V, Montermini L, Molto MD, Pianese L, Cossee M, Cavalcanti F, Monros E, Rodius F, Ducios F, Monticelli A, Zara ˜ F, Canizares J, Koutnikoa H, Bidichandani SI, Gellera C, Brice A, Trouillas P, Michele GD, Filla A, Frutos RD, Palau F, Patel PI, DiDonato S, Mandel J-L, Cocozza S, Koenig M, Pandolfo M. Friedreich’s ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 1996; 271: 1423–1427. 45 Li X-J, Li S-H, Sharp AH, Nucifora FC Jr, Schilling G, Lanahan A, Worley P, Snyder SH, Ross CA. A huntingtin-associated protein enriched in brain with implications for pathology of Huntington’s disease. Nature 1995; 378: 398. 46 Burke JR, Enghild JJ, Martin ME, Jou YS, Myers RM, Roses AD, Vance JM, Strittmatter WJ. Huntington and DRPLA proteins selectively interact with the enzyme GAPDH. Nature Med 1996; 2: 347–350. 47 Choong CS, Kemppainen JA, Zhou A-X, Wilson EM. Reduced androngen receptor gene expression with first exon CAG repeat expansion. Molec Endocr 1996; 10: 1527–1535. 48 Pieretti M, Zhang F, Fu Y-H, Warren ST, Oostra BA, Caskey CT, Nelson DL. Absence of expression of the FMR-1 gene in fragile X syndrome. Cell 1991; 66: 817–822. 49 Wang YH, Amirhaeri S, Kang S, Wells RD, Griffith JD. Preferential nucleosome assembly at DNA triplet repeats from the myotonic dystrophy gene. Science 1994; 265: 669–671. 50 Wells RD. Molecular basis of genetic instability of triplet repeats. J Biol Chem 1996; 271: 2875–2878. 51 Levine M, Manley JL. Transcriptional repression of eukaryotic promoters. Cell 1989; 59: 405–408. 52 Cahill MA, Ernst WH, Janknecht R, Nordheim A. Regulatory squelching. FEBS Lett 1994; 344: 105–108. 53 Imagawa M, Ishikawa Y, Shimano H, Osada S, Nishihara T. CTG triplet repeat in mouse growth inhibitory factor/metallothionein III gene promoter represses the transcriptional activity of the heterologous promoters. J Biol Chem 1995; 270: 20898–20900. 54 Nadel Y, Weisman-Shomer P, Fry M. The fragile X syndrome single strand d(CGG)n nucleotide repeats readily fold back to form unimolecular hairpin structures. J Biol Chem 1995; 270: 28970– 28977. 55 Ohshima K, Kang S, Larson JE, Wells RD. Cloning, characterization and properties of seven triplet repeat DNA sequences. J Biol Chem 1996; 271: 16773–16783. 56 Krontiris TG, Devlin B, Karp DD, Robert NJ, Risch N. An association between the risk of cancer and mutations in the Hras 1 minisatellite locus. New Eng J Med 1993; 329: 517–523. 57 Green M, Krontiris TG. Alleleic variations of reporter gene activation by the HRAS1 minisatellite. Genomics 1993; 17: 429–434. 58 Trepicchio WL, Krontiris TG. Members of the rel/NF-kB family of transcriptional regulatory factors bind the HRAS1 minisatellite DNA sequence. Nucleic Acids Res 1992; 21: 977–985. 59 Capon DJ, Chen EY, Levinson AD, Seeburg PH, Goeddel DV. Complete nucleotide sequences of the T24 human bladder carcinoma oncogene and its normal homologue. Nature 1983; 302: 33–37. 60 Cohen J, Walter M, Levinson A. A repetitive sequence element 39 of the human c-Ha-ras1 gene has enhancer activity. J Cellular Physiol 1987; 5: 75–81. 61 Spandidos DA, Holmes L. Transcriptional enhancer activity in the variable tandem repeat DNA sequence downstream of the human Ha-ras-1 gene. FEBS Lett 1987; 218: 41–46. 62 Trepicchio WL, Krontiris TG. IGH minisatellite suppression of USF-binding-site-and Em-mediated transcriptional activation of the adenovirus major late promoter. Nucleic Acids Res 1993; 21: 977–985. 63 Hammond-Kosack MCU, Dobrinski B, Lurz R, Docherty K, Kilpatrick MW. The human insulin gene linked polymorphic region exhibits an altered DNA structure. Nucleic Acids Res 1992; 20: 231–236. 64 Hammond-Kosack MCU, Docherty K. A consensus repeat sequence from the human insulin gene linked polymorphic region adopts multiple quadriplex DNA structures. FEBS Lett 1992; 301: 79–82. 65 Yamazaki H, Nomoto S, Mishima Y, Kominami R. A 35-kDA protein binding to a cytosine-rich strand of hypervariable minisatellite DNA. J Biol Chem 1992; 267: 12311–12316. 66 Collick A, Dunn MG, Jeffreys AC. Minisatellite binding protein Msbp-1 is a sequence-specific single-stranded DNA-binding protein. Nucleic Acids Res 1991; 19: 6399–6404. 67 Wahls WP, Swenson G, Moore P. Two hypervariable minisatellite DNA binding proteins. Nucleic Acids Res 1991; 19: 3269–3274. 68 Ostareck-Lederer A, Ostareck DH, Standart N, Thiele BJ. Translation of 15-lipoxygenase mRNA is inhibited by a protein that binds to a repeated sequence in the 39 untranslated region. EMBO J 1994; 13: 1476–1481. 69 Curtis D, Lehman R, Zamore PD. Translational regulation in development. Cell 1965; 81: 171–178. 70 Nordheim A, Rich A. The sequence (dC-dA)n.(dG-dT)n forms lefthanded Z-DNA in negatively supercoiled plasmids. Proc Natl Acad Sci USA 1983; 80: 1821–1825. 71 Haniford DB, Pulleybank DE. Facile transition of poly[d(TG).d(CA)] into a left-handed helix in physiological conditions. Nature 1983; 302: 632–634. 72 Wang A, Quigley H-J, Kolpak GJ, Crawford FJ, van Boom JH, van der Marcl G, Rich A. Molecular structure of a left-handed double helical DNA fragment at atomic resolution. Nature 1979; 282: 680–682. 73 Hamada H, Kakunaga T. Potential Z-DNA sequences are highly dispersed in the human genome. Molec Cell Biol 1984; 4: 2610– 2621. Polygenic inheritance and micro/minisatellites DE Comings 74 Schroth GP, Chou P-J, Ho PS. Mapping Z-DNA in the human genome. J Biol Chem 1992; 267: 11846–11855. 75 Nordheim A, Rich A. Negatively supercoiled simian virus 40 DNA containing Z-DNA segments within transcriptional enhancer sequences. Nature 1983; 303: 674–679. 76 Rich A, Nordheim A, Wang AJ. The chemistry and biology of lefthanded Z-DNA. Annu Rev Biochem 1984; 53: 791–856. 77 Hamada H, Petrino MG, Kakunaga T. A novel repeated element with Z-DNA-forming potential is widely found in evolutionarily diverse eukaryotic genomes. Proc Natl Acad Sci USA 1982; 79: 6465–6469. 78 Banerjee R, Carothers AM, Grunberger D. Inhibition of the herpes simplex virus thymidine kinase gene transfection in Ltk-cells by potential Z-DNA forming polymers. Nucleic Acids Res 1985; 13: 5111–5126. 79 Banerjee R, Grunberger D. Enhanced expression of the bacterial chloramphenicol acetyltransferase gene in mouse cells cotransfected with synthetic polynucleotides able to form Z-DNA. Proc Natl Acad Sci USA 1986; 83: 4988–4992. 80 Wolfl S, Martinez C, Rich A, Majzoub JA. Transcription of the human corticotropin-releasing hormone gene in NPLC cells is correlated with Z-DNA formation. Proc Natl Acad Sci USA 1996; 93: 3664–3668. 81 Peck LJ, Wang JC. Transcriptional block caused by negative supercoiling induced structural change in an alternating CG sequence. Cell 1985; 40: 129–137. 82 Butzow JJ, Shin YA, Eichhorn GL. Effect of template conversion from the B to the Z conformation on RNA polymerase activity. Biochemistry 1984; 23: 4837–4843. 83 Arndt-Jovin DJ, Udvardy A, Garner MM, Ritter S, Jovin TM. ZDNA binding and inhibition by GTP of Drosophilia topoisomerase II. Biochemistry 1993; 32: 4862–4872. 84 Herbert A, Lowenhaupt K, Spitzner J, Rich A. Chicken doublestranded RNA adenosine deaminase has apparent specificity for Z-DNA. Proc Natl Acad Sci USA 1995; 92: 7550–7554. ¨ 85 Nordheim A, Tesser P, Azorin F, Kwon YH, Moler A, Rich A. Isolation of Drosophilia proteins that bind selectively to lefthanded Z-DNA. Proc Natl Acad Sci USA 1982; 79: 7729–7733. ¨ 86 Wittig B, Wolffl S, Dorbic T, Vahrson W, Rich A. Transcription of human c-myc in permeabilized nuclei is associated with formation of Z-DNA in three discrete regions of the gene. EMBO J 1992; 12: 4653–4663. 87 Herbert A. RNA editing, introns and evolution. Trends Genet 1996; 12: 6–9. 88 Herbert A, Rich A. The biology of left-handed Z-DNA. J Biol Chem 1996; 271: 11595–11598. ¨ 89 Epplen J, Kyas A, Maueler W. Genomic simple repetitive DNAs are targets for differential binding of nuclear proteins. FEBS Lett 1996; 389: 92–95. 90 Naylor LH, Clark EM. d(TG)n.d(CA)n sequences upstream of the rat prolactin gene form Z-DNA and inhibit gene transcription. Nucl Acids Res 1990; 18: 1595–1601. 91 Wu T, Ikezono T, Angus CW, Shelhamer JH. Characterization of the promoter for the human 85 kDA cytosolic phospholipase A2 gene. Nucl Acids Res 1994; 22: 5093–5098. 92 Weeks DE, Lathrop M. Polygenic disease: methods for mapping complex disease traits. Trends Genet 1995; 11: 513–519. 93 Comings DE. Polygenetic inheritance of psychatric disorders. In: Blum K, Noble EP, Sparks RS, Sheridan PJ (eds). Handbook of Psychiatric Genetics. CRC Press: Boca Raton, FL, 1996, pp 235– 260. 94 Weeks DE, Lange K. The affected-pedigree-member method: power to detect linkage. Am J Hum Genet 1988; 42: 315–326. 95 Greenberg DA, Hodge SE, Vieland VJ, Spence MA. Affecteds-only linkage methods are not a panacea. Am J Hum Genet 1996; 58: 892–895. 96 Carey G, Williamson J. Linkage analysis of quantitative traits: increased power by using selected samples. Am J Hum Genetics 1991; 49: 786–796. 97 Owen MJ, McGuffin P. Association and linkage: complementary strategies for complex disorders. J Med Genet 1993; 30: 638–639. 98 Comings DE, Gade R, Wu S, Chiu C, Dietz G, Muhleman D, Saucier G, Ferry L, Burchete R, Johnson P, Verde R, MacMurray JP. Studies of the potential role of the dopamine D1 receptor gene in addictive behaviors. Mol Psychiatry 1997; 2: 44–56. 99 Comings DE, MacMurray JM. Molecular heterosis. 1977 (submitted). 100 Wills C. The Runaway Brain. Basic Books: New York, NY, 1993, pp 1–358. 101 Arcot SS, Wang Z, Weber JL, Deininger PL, Batzer MA. Alu repeats: a source for the genesis of primate microsatellites. Genomics 1995; 29: 136–144. 102 Hall AV, Antoniou H, Wang Y, Cheung AH, Arbus AM, Olson SL, Lu WC, Kau CL, Marsden PA. Structural organization of the human neuronal nitric oxide synthase gene (NOS). J Biol Chem 1994; 269: 33082–33090. 103 Nelson RJ, Demas GE, Huang PL, Fishman MC, Dawson VL, Dawson TM, Snyder SH. Behavioral abnormalities in male mice lacking neuronal nitric oxide synthase. Nature 1995; 378: 383–386. 104 Comings DE, Wu S, Gonzalez N, Muhleman D, Gade R, Blake H, MacMurray JP, McGue M, Lykken D. Association of the normal FRAXA and HTR2A genes with performance IQ in the general population. 1997 (submitted). 31