Human Molecular Genetics Advance Access originally published online on February 5, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Human Molecular Genetics, 2004, Vol. 13, Review Issue 1 R91-R102
DOI: 10.1093/hmg/ddh074
Genetic dissection of myocilin glaucoma
1Osteoporosis Research Center, Creighton University Omaha, NE 68131, USA, 2Division of Ophthalmology, Creighton University, Omaha, NE 68131, USA and 3Office of the President, Texas Tech University Health Sciences Center, Lubbock, TX 79430, USA
Received December 24, 2003; Revised January 12, 2004; Accepted January 26, 2004
 |
ABSTRACT |
|---|
TOP
ABSTRACT
INTRODUCTIONPrimary open-angle glaucoma (POAG) is a complex disease with unknown causes. However, in the past decade, POAG has been linked to six chromosomal regions, of which the gene MYOC encoding myocilin and the gene OPTN encoding optineurin have been identified to harbor causal mutations (disease-causing variants, DCV). POAG caused by DCV at MYOC has been termed ‘myocilin glaucoma’. Clinically, DCV at MYOC may manifest as a typical POAG, normal tension glaucoma, or ocular hypertension without glaucoma. Individuals with the Arg46Stop mutation that almost knocks out the entire coding sequence may have severe glaucoma or no glaucoma. Genetically, myocilin glaucoma follows autosomal dominant, recessive or no pattern of inheritance. DCV at MYOC cause POAG in interaction with environmental factors and DCV at other loci. Most DCV at MYOC are relatively young, and the Gln368Stop mutation is exclusively European in origin. The overall frequency of DCV at MYOC is similar among African, Caucasian and Asian probands with POAG. Because of this fact and the higher prevalence of POAG in African descendants compared with Caucasians or Asians, the overall frequency of DCV at MYOC is several-fold higher in the general population of African descendants, which is in part responsible for their higher prevalence of POAG. Although the Arg46Stop mutation was often observed in normal controls, Arg46Stop carriers tend to have higher risk of developing POAG. Polymorphisms at several loci including MYOC are associated with POAG, and play an important role in the pathogenesis of POAG.
|
INTRODUCTION |
|---|
Glaucoma researchers have enjoyed success in identifying genomic variations that contribute to the pathogenesis of complex diseases. To date, six loci (GLC1A–E) (1–6) have been linked to primary open angle glaucoma (POAG) alone. Among them, two genes, MYOC encoding myocilin and OPTN encoding optineurin, have been identified as harboring causal mutations. Several other genes harboring mutations causing primary congenital glaucoma (7), autosomal dominant optic atrophy (8,9) and other eye diseases (10) have also been mapped or identified. More than 70 mutations in the MYOC gene that encodes a protein composed of only 504 amino acids have been found to contribute to the pathogenesis of POAG. Numerous polymorphisms at MYOC, notably single nucleotide polymorphisms (SNPs), which may or may not cause glaucoma, have been reported. MYOC and its mutations have been extensively investigated in different racial/ethnic populations, in animal models, and in cell cultures. Many findings appear to be of interest not only to geneticists and ophthalmologists but also to anthropologists, sociologists and the public at large.
The discoveries of glaucoma-causing mutations are of great interest to the public, because the disease affects 
67 million people worldwide with about 6.7 million suffering from bilateral blindness as a result (11). POAG accounts for half of all cases (12). Depending on the definition of POAG, the prevalence is 2.8–8.8% in African descendants, much higher than that in Caucasians (1.1–2.1%) and Asians (2.6%) (13–15). It is hoped that discovery of genes and causal mutations for POAG may help the development of more effective drugs for treatment of POAG.
One of the most interesting findings is that the prevalence of MYOC mutations is similar among Caucasian, Asian and African POAG probands (16). This has raised a question as to whether genetic factors are responsible for the higher prevalence of POAG among African descendants (17), and added fuel to the debate of whether ‘race’ is a valid concept (18–21). Some investigators have suggested that environmental factors may play a more important role (17), while others have argued that ‘physicians must assess people on an individual basis, not on preconceived ideas of racial prevalences’ (17).
Many findings are of interest to researchers working on the genetic aspect of other diseases for their general implications. For example, common diseases may not necessarily be caused by a few common variants (22) because many rare mutations are found to cause a relatively common disease. After the initial discovery of a POAG-causing mutation, the mutation was subsequently observed in controls. Some researchers consider such mutations non-disease-causing (23) polymorphisms. This raises the question of how to define a mutation and a polymorphism.
Glaucoma has the typical feature of complex traits: individuals with the putative mutation may not have the disease (24–26), while those without it may have the disease (25). Lander and Schork (27) define a complex trait as any phenotype that does not exhibit classic Mendelian recessive or dominant inheritance attributable to a single locus. Complex traits have several or all of the following features: locus heterogeneity (mutations at many loci can cause the same disease independently), polygenic inheritance (more than one mutation are required to cause the disease), phenocopies (those with the disease caused by environmental factors), and incomplete penetrance (not all mutation-carriers have the disease) (27). Glaucoma has all these features. The term ‘complex trait’ is used in the context of genetic diseases. In fact, almost all traits have a genetic component (non-zero heritability) (28), while few genetic diseases are caused by genetic factors alone. Even AIDS and diarrhea can be considered as ‘genetic’ diseases since a few normal (or rather, ‘abnormal’) individuals with cells resistant to HIV entry are resistant to AIDS (29), and genetic factors are involved in the pathophysiology of diarrhea (30,31). The border between ‘genetic’ and ‘non-genetic’ diseases is blurred. Nonetheless, glaucoma appears to be more of a genetic disease based on our current knowledge.
The etiology of POAG is unknown. However, it has long been recognized that patients with the disease are often found to cluster in families (32–34). Some 30–56% of patients with glaucoma or ocular hypertension (OHT) had a positive family history (35,36). First-degree relatives of POAG patients or suspects are seven to 10 times more likely to suffer from POAG compared with the general population (37,38). These facts suggest that genetic factors are involved in the pathogenesis of POAG, and perhaps play the central role. With the advent of modern biotechnology, mutations for many diseases have been identified and cloned. Even some good mutations for longevity (39), high bone strength (40) and reduced risk for glaucoma (26) have been mapped (39), cloned (40) or suggested (26). The MYOC gene is the first discovered to be linked to POAG.
Sheffield et al. (1) first mapped juvenile-onset open angle glaucoma (JOAG) to chromosome 1q21–q31. It was further fine-mapped to a small region (41). The locus harboring mutation causing POAG was termed GLC1A (‘GLC’ stands for glaucoma, ‘1’ for primary open-angle, and ‘A’, the first gene discovered). Others subsequently mapped JOAG in different families to the same locus (42–47). Stone et al. (48) eventually identified mutations in the MYOC gene causing JOAG. Later studies showed that MOYC mutations also caused adult-onset POAG (49–52). The MYOC gene with unknown function was independently discovered by Kubota et al. (53) and Nguyen et al. (54). It bears significant homology with genes encoding myosin and olfactomedin (49,53,54). The gene was also known as TIGR (trabecular meshwork inducible glucocorticoid response) (54). Fingert et al. (55) termed POAG caused by mutations at the MYOC gene ‘myocilin glaucoma’.
A wealth of data has been accumulated in the past decade, which has greatly enhanced our understanding of the nature of POAG. Meanwhile, discrepancies and confusions arise due to the complexity of glaucoma. We wish to integrate isolated reports to draw an overall picture of the genetic complexity of myocilin glaucoma, which may be beneficial to future investigations. Necessary meta-analyses are undertaken to explore further implications of previous findings.
|
FACTS |
|---|
Prevalence of MYOC mutations in POAG probands from random samples in different racial/ethnic populations
Fingert et al. (16) set criteria for probable disease-causing mutations (DCM) as those that: (i) alter myocilin amino acid sequence; (ii) are present in one or more glaucoma subjects; (iii) are present in less than 1% of the general population; and (iv) are absent in normal controls. In Tables 1–
4, we list genomic variations previously considered as DCM by any authors (23,24,26,56–83), even if they were found later in normal controls, because the number of DCMs decreases with time, and eventually none will meet the above criteria. In Tables 1–3, frequencies of DCMs at MYOC are categorized by race, which is another aspect of genetic complexity of glaucoma that must be considered.
|
|
|
|
A total of 73 DCMs have been found in the MYOC gene (Tables 1–4). These DCMs at MYOC are found in 3.86% (155/4107) of Caucasian probands with POAG (including normal tension glaucoma) or ocular hypertension, 3.30% (18/545) of probands of African descendants (including African Americans and black residents in Africa) and 4.44% (36/810) of Asian probands (Tables 1–3). The rates are similar to the initial estimates by Fingert et al. (16). Normal tension glaucoma and ocular hypertension appear to be independently associated with MYOC mutations, which are found in 3.88% (5/129) of Caucasian probands with normal tension glaucoma, 2.08% (6/288) of Caucasians with ocular hypertension and 2.52% (3/119) of Asian probands with normal tension glaucoma.
The most common DCMs are the Gln368Stop (1.65%) and the Arg46Stop (0.99%) mutations. No mutation was found in more than two POAG probands in African descendants because of smaller sample size. The Arg46Stop, Arg158Gln, thr353Ile, Asp208Glu and Pro481Ser mutations initially reported as DCMs in Asian populations were later found in normal controls (26,72). If we include only those mutations based on the criteria of Fingert et al. (16), then we would have to exclude 65 POAG probands with Gln368Stop, because Gln368Stop was also found in two normal controls in case-control studies (23,24) and more in family pedigree studies (49,50,82,83). We will discuss whether these mutations are disease-causing later.
Distribution of MYOC mutations
Three DCMs at MYOC (Gly252Arg, Gly367Arg and Pro370Leu) were found in Asians and Caucasians, and three (Thr293Lys, Thr377Met and Glu352Lys) in Africans and Caucasians (Table 1). No single mutation was shared by all three major human populations. Most DCMs at MYOC exist only in a specific racial group. For example, the most frequent mutation Gln368Stop was present only in European descendants (with one exception of an African American), and the second most frequent mutation Arg46Stop was shared only by Asian populations. Many mutations were found only in a specific region. For example, the Asn480Lys mutation was found in two geographically close European countries but not anywhere else. Remarkably, all the 71 patients with POAG from six unrelated families from northern France investigated by Brezin et al. (69) carried the same Asn480Lys mutation. One Asn480Lys carrier from a random sample consisted of 123 unrelated Caucasian POAG patients was also from northern France. The Gln48His mutation was found only in India (56), and the Cys433Arg mutation only in Brazil (63). Interestingly, this Cys433Arg mutation was found in five of 13 white and two of 11 reportedly black unrelated Brazilians with POAG.
The distribution of mutations provides information about population history, structure, movement, relation, ancestry and diversity, which is useful not only to anthropologists but also to geneticists. For example, the most common mutations, Gln368Stop and Arg46Stop, must have arisen after the divergence between Asians and Caucasians because they are not common to both populations. Distribution of genomic variations and their frequencies are the bases of the ‘Out of Africa’ (84) and, more recently, the ‘Out of Ethiopia’ theory (85). Mutation distribution is also the keystone of the coalescence approach to determine the age of most recent common ancestors (86). The ‘Out of Africa’ theory is, in turn, the basis of the common disease/common variants hypothesis for association studies with SNPs (22).
Fingert et al. (16) reported that the Gln368Stop mutation was found at least once in all populations (including one out of 312 African Americans with POAG) except the Japanese. They concluded that the Gln368Stop comes from a common founder. Sale et al. (68) and Baird et al. (25) suggested that this mutation in the single African American individual was of European origin due to racial admixture. This is an important issue because it would imply that this mutation is very old if African descendants in general share this mutation. Subsequently, Sale et al. investigated 46 glaucoma patients from Uganda, Eastern Africa and found no such mutation. Because of small sample size, they suggested further investigation. Challa et al. (66) investigated 90 patients with POAG from Ghana, West Africa, and found MYOC mutations (4.4%, 4/90) other than Gln368Stop. However, absence of evidence is not yet a proof. Mathematical analysis may provide some insights. Since African Americans tend to have an average of 17% Caucasian genetic background (87), and the Gln368Stop mutation was found in 26/1284 Caucasian POAG probands (16), the expected number of African Americans with POAG and the mutation due to racial admixture is (26/1284)(0.17)(312)=1.08. This estimate is very close to the actual number in the studies of Fingert et al. (16).
Evidence from haplotype-sharing studies also supports the view of Sale et al. and Baird et al. Fingert et al. (16) showed that markers closely franking the MYOC gene were shared 100% by all POAG probands with the Gln368Stop mutation, suggesting a close genetic relationship among the subjects including the African American individual. Faucher et al. (24) showed that MYOC mutation carriers share similar haplotypes over a long stretch of DNA in five (including Gln368Stop) of the six mutations examined. These findings suggest that Gln368Stop is exclusively European in origin and is a relatively young mutation. Other mutations such as the Asn480Lys also come from one single founder (49,69). However, the Pro370Leu was more likely to have occurred twice (a recurrent mutation) (49). The few mutations shared by two major human populations are likely to have arisen from independent mutation events or gene flow because they are extremely rare in either population.
The overall allele frequency of MYOC mutations in the general populations
Fingert et al. (16) reported two MYOC mutation carriers in 505 Caucasian subjects from the general population, and 0 in 50 subjects from the African American general population. The number of MYOC mutation carriers was too small to estimate the allele frequency in the general populations. Therefore, we estimate below the overall frequency (f) of MYOC mutations with an average penetrance k in the general population based on the overall allele frequency (fG) of MYOC mutations in the probands with POAG, and the known prevalence (p) of POAG.
 |
The constant c accounts for possible imprecision in estimating p, fG or k. Since fG is similar among the three major human populations and p is about 4-fold greater in African descendants than in Caucasians, f is about 4-fold greater in African descendants if k is similar. Let the lowest average of k (=0.5 based on Table 5) for Gln368stop represent k of all DCMs at MYOC in the Caucasian population, and k=1 in African descendants, then f is still 2-fold greater in African descendants than in Caucasians. We conclude that the higher frequency of MYOC mutations in the general population of African descendants is in part responsible for its higher prevalence of POAG. Further investigations on penetrance levels of MYOC mutations in Africans as well as in other populations are crucial for a more accurate estimate of f.
|
|
MUTATION OR POLYMORPHISM? |
|---|
In the literature, when a variant was found only in subjects with POAG, it was reported as a mutation. When it was also found in normal controls it was called (or suspected as) a polymorphism. Two groups (23,24) have each observed a normal control carrying the most frequent mutation Gln368Stop in case–control studies. Thus, Gln368Stop was not regarded as a DCM but a modifier (23). In familial studies, as high as 84.2% (16/19) Gln368Stop-carriers did not have POAG (50). Also, Pang et al. (26) reported that the most frequent mutation Arg46Stop in Asians was even (slightly) more often found in normal controls (2.24%, 9/402) than in POAG probands (1.99%, 4/201). Are Gln368Stop and Arg46Stop mutations, polymorphisms or modifiers?
Mutation is a process that changes chromosomal DNA sequence. The variant created by mutation is often called a mutation or a mutant as opposed to the wild-type. When the frequency of a mutant(s) approaches that of the wild-type, these alternative forms (mutants and the wild-type) are often called a polymorphism. These are concepts in terms of frequency rather than function. However, historically a polymorphism was often thought of as a ‘normal’ genomic variation, while a mutation was often associated with a disease in one's mind. In fact, a rare variant (i.e. a mutation) may not necessarily cause any diseases (88), while a common variant of a polymorphism associated with a disease in association studies is either disease causing or is in linkage disequilibrium with a disease-causing variant (27). Polymorphism is created by the act of mutation (89). Every nucleotide in the human genome was once a mutation since the very first DNA molecule had emerged on Earth. For example, the mutant 
4 allele in the apolipoprotein E gene (APOE) associated with Alzheimer's disease used to be the wild type, while 1, 2 and 3 were mutations some 200 000 years ago (90). Alward et al. (61) recently used the term ‘disease-causing variations’ (DCV) instead of mutations, which avoids confusion.
Because of its nature (incomplete penetrance, polygenic inheritance, etc.), a DCV for complex diseases cannot be defined as a variant that is present only in cases but not in controls. It is often determined statistically by comparing the frequencies of the mutant in question in cases (fd) versus controls (fn). A more sensitive approach is to compare the absolute risk (m) for mutant carriers versus the absolute risk (w) for wild-type carriers to develop POAG in familial studies, where m is the ratio of the number of mutant carriers with POAG to the total number of mutant carriers. Another advantage is that mutant carriers are more readily available in families with the mutant than in random samples. Since wild-type carriers with POAG are rarely observed in familial studies with a few exceptions (25), we may derive w from the general population:
 |
 |
Since p of POAG is small (0.02 in Caucasians, 0.026 in Asians and 0.08 in Africans) (13), w (=0.0204, 0.0267 and 0.0870, respectively) is not significantly different from p. Both fd and fn are often too small to affect w significantly in equation (2). That is, the risk of non-mutant carriers to developing POAG is approximately equal to p.
In a family with the lowest penetrance of Gln368Stop reported (50), only three in 19 members with Gln368Stop had POAG (m=3/19=0.158). However, the relative risk (RR) of Gln368Stop carriers to developing POAG is RR=m/w=0.158/0.0204=7.7 (P=0.05). Note that w=0.0204 in Caucasians. In other words, Gln368Stop carriers had a risk of developing POAG more than 7-fold greater than non-carriers under the worst-case scenario. With data pooled from the three families in the above study (50), m=8/30=0.267 and RR=13.07 with 95% confidence interval between 5.31 and 20.83 (P<0.0015). Therefore, there is no doubt that Gln368Stop is a DCV. Lam et al. (91) and Yoon et al. (70) reported that two in 14 Arg46Stop carriers had POAG in two Asian families. Thus, we have m=(2/14)=0.143, w=0.0267 for Asians, and RR=5.35 (P=0.107), showing a trend for Arg46Stop carriers to develop POAG. Further investigations in more families with the Arg46Stop mutation are needed for a definitive conclusion. Other questionable mutations can also be more readily resolved through familial studies.
A variant conveying significant risk for a disease is often called a risk allele, a risk factor or a susceptibility gene. In many cases a risk allele is irrelevant to the pathogenesis of a disease, because it may be in linkage disequilibrium with a DCV. However, since MYOC is well established to harbor DCV, it is unlikely that POAG in these families is caused by a DCV in linkage disequilibrium with Gln368Stop or Arg46Stop. That is, Gln368Stop and possibly Arg46Stop are DCV.
|
INCOMPLETE PENETRANCE |
|---|
Very often, a DCV can cause the disease in some but not all individuals. Some normal DCV carriers may never develop the disease later in life. This phenomenon is called incomplete penetrance, which is expressed as the ratio of the number of DCV carriers with the disease to the total number of DCV carriers. In linkage analysis, penetrance is routinely estimated for a better LOD score. We list published penetrance estimates for different MYOC mutations in Table 5. The penetrances of DCVs at MYOC are age-dependent. All DCVs at MYOC have incomplete penetrance at age 30. Thr377Met has the highest penetrance, 88% at 30 years of age (92) followed by Cys433Arg (93). Gln368STOP has the lowest. The underlying mechanisms for incomplete or complete penetrance are not well understood. Several lines of evidence suggest that increase in penetrance with age may be related to accumulation of environmental exposure with age, age-related gene expression, and gene by gene and/or gene by environment interaction.
|
GENE BY GENE AND GENE BY ENVIRONMENT INTERACTIONS |
|---|
In a single extended family (139 members genotyped) reported by Baird et al. (25), nine individuals with either POAG or OHT and 10 normal individuals carried Gln368STOP. Interestingly, 31 members with POAG or OHT did not carry the Gln368STOP mutation. Baird et al. suggested that other factors were involved. It is very likely that these family members carry another mutation. Therefore, there is a greater chance of some family members carrying both the unknown DCV and the MYOC mutation. As a result, those with both mutations had earlier onset, while those carrying only the Gln368STOP mutation had a later onset. Consistent with this hypothesis, Vincent et al. (60) reported a family with POAG in which both MYOC and CYP1B1 mutations segregated with the disease. The mean age at onset was 51 years in MYOC mutation (Gly399Val) carriers versus 27 years in carriers of both the MYOC and CYP1B1 mutations. This is an unequivocal piece of evidence showing the interaction between MYOC and another gene in causing POAG. This may also serve as one of the mechanisms for incomplete penetrance.
Although six loci have so far been linked to POAG, the likelihood for simultaneous presence of any two DCVs at these loci in one POAG patient is very low. However, some common variants with functional significance have more chance to interact with a major DCV or with many other common variants to cause the disease, although they themselves alone may be neither sufficient nor necessary to cause POAG. Morton termed such more frequent variants with small effect ‘polygenes’ (94). A common allele at the APOE gene is reported to interact with an SNP in the MYOC promoter, increasing intra ocular pressure (IOP) (95). Such an interaction might be one of the mechanisms for the occurrence of OHT.
Yoon et al. (70) reported that a homozygote with the Arg46Stop mutation had POAG but heterozygotes did not, suggesting autosomal recessive inheritance. However, Pang et al. (26) reported a case of a 77-year-old woman homozygous for the Arg46Stop mutation who had no POAG, while a heterozygote suffered from POAG. One of the explanations for these findings is that this particular variant may require unknown necessary mutations or genetic backgrounds (polygenic inheritance) or certain environmental exposure for its expression of the disease phenotype. POAG appears to be more complex than complex traits defined (27) because it does follow classic Mendelian autosomal recessive or dominant inheritance in some cases (1,70).
Certain environmental factors are involved in the pathogenesis of POAG. The effect of age may reflect age-related accumulation of environmental exposure (96), which may interact with MYOC mutations to cause POAG. Although glucocorticoid can induce glaucoma, it did not appear to interact with MYOC mutations (61). Certain behavioral factors (e.g. high salt intake, stress, lack of exercise and obesity) are known to play a role in the etiology of hypertension and diabetes (97), which are possible risk factors for POAG (13). Thus, it is possible that these environmental and behavioral factors may cause POAG indirectly by interaction with mutations at MYOC and/or other genes.
It was reported recently that the prevalence of POAG in black residents in South Africa was much lower (2.9%) (98) than that observed in Barbados or St Lucia (13,15). Thus, the frequency of MYOC mutations in the general South African black population should not be much higher than that of Caucasians. Both environmental and genetic factors may account for the finding. Since Sub-Saharan Africans are the most diversified genetically in the world (84), different genetic background may be partly responsible for the lower prevalence. Geographic locations, climate, culture and diet and their interaction with genetic factors may also play a role.
|
GENETIC HETEROGENEITY AND PHENOCOPY |
|---|
Locus heterogeneity of POAG is evident; at least six loci can cause the same disease and more than 70 mutations in the MYOC gene alone are POAG-causing variants. Two different MYOC mutations have been reported to be present in one individual, which caused severe POAG. Two polymorphisms, (C/T) and MYOC.mt1, at the promoter region (–153) and (–1000) of the MYOC gene have been reported to be associated with normal tension glaucoma and increased IOP levels, respectively (99–101). Polymorphisms in other genes, including OPA1 (102), APOE (103) tumor necrosis factor
308 (104) and atrial natriuretic peptide (105), are also associated with POAG. Currently, whether the association of MYOC.mt1 with POAG is real is inconclusive since others have found no association (61). Inconsistent results are very common in association studies in part due to interference of genetic heterogeneity and environmental factors (106). Thus, it is very important to reduce their interference in study design in order to increase the chance of detecting an association. For example, Colomb et al. (100) checked all patients for the absence of MYOC mutations known to cause POAG or to increase IOP. This may be one of the reasons for their success. Inclusion of subjects with known MYOC mutations might be partly responsible for a negative result (no association detected) (61). In contrast to other studies (61), Colomb et al. (100) investigated only one ethnic group (the French), which also reduces genetic heterogeneity. Inclusion of subjects from different racial/ethnic groups in association studies (61) has another disadvantage: it may cause a spurious association (a false positive result) (27) unless data are analyzed separately in different races.
One may also reduce the interference of certain factors statistically. In the studies of Polansky et al. (101), many risk factors including age, family history, initial drug and surgical treatments, diabetes, gender, myopia and initial disease severity were included in their analysis of association between MYOC.mt1 and POAG. They adjusted IOP and visual field for all these factors, reducing their interference. Although a better way is to investigate subjects of the same age, sex, race and treatment (or without it), such ‘ideal subjects’ are hard to find. Nonetheless, once an association is found it needs to be tested in multiple samples (107), as does the association between MYOC.mt1 and IOP and visual field.
Although six loci have been mapped, they may account for only a small fraction of glaucoma cases. Many may suffer from glaucoma as the result of interaction among polygenes or their interaction with environmental factors and/or with major genes. Thus, it is very important to identify polygenes by association studies (94). However, because of inconsistency in results due to different designs, a journal sets criteria for a rapid rejection of association studies on complex traits (108). While recognizing the validity of editorial scrutiny, researchers appeal for freedom of association so that results from association studies of high quality can be published (109).
|
PLEIOTROPIC EFFECTS |
|---|
The same mutation may cause different diseases. Sometimes, it may be deleterious to one disease and beneficial (protective) to another. This phenomenon is called pleiotropic effect. The association of
4 with both normal tension glaucoma (103) and Alzheimer's disease (110) is a typical example. Currently, we are not quite sure whether OHT, normal tension glaucoma and POAG are the same or different genetic entities. MYOC mutations appear to be associated with all these phenotypes. Also, MYOC mutations were found in patients with exfoliative glaucoma (61). However, Jansson et al. (23) recently reported no disease-causing mutations at MYOC in exfoliative glaucoma patients. Gln368Stop has the highest frequency among MYOC mutations. In addition to the relatively low penetrance level (which increases fitness), one may speculate that this mutation might have some advantages for survival, similar to mutations for sickle cell anemia (111) and cystic fibrosis (30). Identification of other effects of MYOC mutations will depend on further investigations at molecular as well as clinical levels.
|
ISOLATED POPULATIONS |
|---|
Genetic studies in isolated populations (e.g. Finland) have several advantages such as the relatively high degree of genetic homogeneity (112). Mutations in isolated populations tend to have a single or few founder(s), reducing genetic heterogeneity. For this reason, many disease-causing mutations were first discovered in the Finnish population (113). In addition to the Finnish, Old Order Amish, Hutterites, Sardinian and Jewish communities (114), multiple genetic isolates exist in India (115) and China. The distribution of MYOC mutations suggests different degrees of isolation among human populations that are not traditionally considered as isolated populations. Because of the relative isolation and inbreeding, homozygotes of the Arg46Stop mutation were identified in Asian populations (26,70,91). Such relatively isolated populations may also provide opportunities for genetic research into glaucoma.
|
THE COMMON DISEASE–COMMON VARIANTS HYPOTHESIS |
|---|
It was hypothesized several years ago that a common disease is caused by one or few common variants (116,117), which is called the Common Disease–Common Variants hypothesis (22). The basis is the ‘Out of Africa’ theory (22). All modern humans are descendants of about 10 000 (effect size) founders (118,119) 100 000–160 000 years ago (84,85). The term ‘common’ implies that the disease is both prevalent and widespread (22). Mutations for common diseases are likely to have arisen before the divergence of peoples because the diseases are now widespread globally (22). Pritchard (120) proposed that rare mutations at many loci might explain the high prevalence of common diseases. Results from studies on MYOC mutations show that a great number of rare mutations contribute to a relatively common disease, POAG. Many common variants (e.g. SNPs) are also associated with POAG (95,99–105) and with more prevalent diseases such as hypertension and Alzheimer's disease (22,110). Some of the SNPs may be causal. Thus, many common and/or many rare variants in interaction among themselves and with environmental factors provide a better explanation for the high prevalence of common diseases.
In this context, if one can demonstrate that a common variant (e.g. SNP) is associated with a disease in the three major human populations, then the common variant and the putative DCV must have been in linkage disequilibrium before the divergence of peoples 100 000–160 000 years ago. An association detected in modern times implies that recombination events in 5000–8000 meioses (assuming 20 years per generation) have not separated the common variant and the putative DCV on the same ancestral chromosome sufficiently for them to reach equilibrium. This further implies that the two loci must be very closely linked (no more than 330 kb) (121,122). Therefore, association studies in three major human populations are capable of mapping DCV loci responsible for common and complex diseases to a small region (<0.3 cM and, in some case, <0.03 cM) (122). If SNPs at MYOC and other loci (99–105) are associated with POAG in the three major human populations independently, then DCVs are within these candidate genes rather than in genes at a distance.
|
PHENOTYPE DEFINITION |
|---|
Geneticists are powerless without sufficient knowledge of clinical aspects of the disease of interest, especially in phenotype definition (123,124). Now, ophthalmologists have recognized different forms and subtypes of glaucoma, which is very helpful for genetic studies. However, confusion still exists as to whether normal tension glaucoma is a subtype of POAG or a different type of glaucoma. Within the subtype of POAG, different clinical manifestations such as myopia, disc hemorrhage, reduced corneal thickness, migraine and vasospasm might represent different genetic bases (13). For example, abnormal electroencephalogram rather than epilepsy is linked to a gene harboring a mutation for epilepsy (123). Had researchers not paid attention to a subtle clinical sign, another gene for epilepsy would not have been located (124). Stoilova et al. (125) ensured phenotypic homogeneity at clinical level before linkage analysis, which may be partly responsible for their success in localizing GLC1B (125). With sufficient clinical knowledge, some genes harboring DCV are ‘mapped’ by clinical diagnosis (126).
|
MOLECULAR AND CELLULAR MECHANISMS |
|---|
Morissette (127) reported several unaffected homozygotes for the Lys423Glu mutation in a family in which heterozygotes are affected with POAG. Wiggs and Vollrath (128) reported a clinically normal patient with only one functional copy of MYOC. These findings suggest that the cause of POAG is not due to haploinsufficiency, but gain of function (127,128). Consistent with clinical findings, MYOC-knockout mice grow normally with normal IOP and ocular morphology (129), suggesting that total absence of myocilin is harmless to IOP and visual field. At the cellular level (130), little or no myocilin was secreted from cultured cells with mutant MYOC, while wild-type myocilin was secreted, suggesting that MYOC glaucoma may be partly due to reduction in myocilin secretion. Myocilin can form homo- or hetero-multimers and interact with myosin regulatory light chain in vivo (131,132). mRNA of myocilin have been recently detected in myelinated nerves as the structural component of the myelin sheath in peripheral nerves (133). The exact mechanisms whereby MYOC mutations cause POAG will rely on further investigation of signaling pathways of the MYOC gene.
|
RACE, ETHNICITY AND POAG |
|---|
Diversity and polymorphisms are a great asset to the human population. A polymorphism with functional significance may be deleterious or beneficial depending on the historical clock, environment and individuals. Without diversity, the whole European population, rather than one third of it, might have been wiped out by ‘the Black Death’ (the Plague). Diversity and polymorphism are also important in genetic studies such as admixture mapping (87), avoidance of spurious association (27), determination of the age of most recent common ancestors (86) and fine mapping of susceptibility genes (121,122). Environmental factors, including social–economic status, and their interaction with genetic factors certainly play an important role in disease prevalence in general. However, exact mechanisms whereby particular environmental factors cause POAG and their interaction with genetic factors are still not well understood. Sometimes, genetic factors confound environmental factors. Identification of factors for health disparity among different populations is a great challenge to geneticists, clinicians and sociologists alike (21).
It should be noted that the higher frequency of DCV at MYOC in African descendants (including African Americans, Afro-Caribbeans and black residents in Africa) does not serve as evidence that the high prevalence of POAG in Blacks is attributable to genes, because DCV at MYOC account for only a small fraction of the disease population, and much less in the general population. In the future, it is important to examine subgroups of African descendants to better understand the complex nature of POAG.
|
ACKNOWLEDGEMENT |
|---|
This publication was supported by revenue from the Nebraska Tobacco Settlement awarded to Creighton University by the State of Nebraska, LB692. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the State of Nebraska.
|
FOOTNOTES |
|---|
* To whom correspondence should be addressed at: Osteoporosis Research Center, Creighton University School of Medicine, 601 North 30th Street, Suite 6730, Omaha, NE 68131, USA. Tel: +1 4022804283; Fax: +1 4022805173; Email: gdgong{at}creighton.edu
|
REFERENCES |
|---|
- Sheffield, V.C., Stone, E.M., Alward, W.L., Drack, A.V., Johnson, A.T., Streb, L.M. and Nichols, B.E. (1993) Genetic linkage of familial open angle glaucoma to chromosome 1q21–q31. Nat. Genet., 4, 47–50.[CrossRef][ISI][Medline]
- Stoilova, D., Child, A., Trifan, O.C., Crick, R.P., Coakes, R.L. and Sarfarazi, M. (1996) Localization of a locus (GLC1B) for adult-onset primary open angle glaucoma to the 2cen-q13 region. Genomics, 36, 142–150.[CrossRef][ISI][Medline]
- Wirtz, M.K., Samples, J.R., Kramer, P.L., Rust, K., Topinka, J.R., Yount, J., Koler, R.D. and Acott, T.S. (1997) Mapping a gene for adult.onset primary open-angle glaucoma to chromosome 3q. Am. J. Hum. Genet., 60, 296–304.[ISI][Medline]
- Trifan, O.C., Traboulsi, E.I., Stoilova, D., Alozie, I., Nguyen, R., Raja, S. and Sarfarazi, M. (1998) A third locus (GLC1D) for adult-onset primary open-angle glaucoma maps to the 8q23 region. Am. J. Ophthalmol., 126, 17–28.[CrossRef][ISI][Medline]
- Sarfarazi, M., Child, A., Stoilova, D., Brice, G., Desai, T., Trifan, O.C., Poinoosawmy, D. and Crick, R.P. (1998) Localization of the fourth locus (GLC1E) for adult.onset primary open-angle glaucoma to the 10p15–p14 region. Am. J. Hum. Genet., 62, 641–652.[CrossRef][ISI][Medline]
-
Wirtz, M.K., Samples, J.R., Rust, K., Lie, J., Nordling, L., Schilling, K., Acott, T.S. and Kramer, P.L. (1999) GLC1F, a new primary open-angle glaucoma locus, maps to 7q35–q36. Arch. Ophthal., 117, 237–241.
[Abstract/Free Full Text] - Stoilov, I., Akarsu, A.N., Alozie, I., Child, A., Barsoum-Homsy, M., Turacli, M.E., Or, M., Lewis, R.A., Ozdemir, N., Brice, G. et al. (1998) Sequence analysis and homology modeling suggest that primary congenital glaucoma on 2p21 results from mutations disrupting either the hinge region or the conserved core structures of cytochrome P4501B1. Am. J. Hum. Genet., 62, 573–584.[CrossRef][ISI][Medline]
-
Eiberg, H., Kjer, B., Kjer, P. and Rosenberg, T. (1994) Dominant optic atrophy (OPA1) mapped to chromosome 3q region. I. Linkage analysis. Hum. Mol. Genet., 3, 977–980.
[Abstract/Free Full Text] - Delettre, C., Lenaers, G., Pelloquin, L., Belenguer, P. and Hamel, C.P. (2002) OPA1 (Kjer type) dominant optic atrophy, a novel mitochondrial disease. Mol. Genet. Metab., 75, 97–107.[CrossRef][ISI][Medline]
- Pang, C.P. and Lam, D.S.C. (2002) Differential occurrence of mutations causative of eye diseases in the Chinese population. Hum. Mut., 19, 189–208.[CrossRef][ISI][Medline]
-
Quigley, H.A (1996) Number of people with glaucoma worldwide. Br. J. Ophthal., 80, 389–393.
[Abstract/Free Full Text] - Goldberg, I. (2000) How common is glaucoma worldwide? In Weinreb, R.N., Kitazawa, Y. and Krieglstein, G. (eds), Glaucoma in the 21st Century. Mosby International, London, pp. 3–8.
- Wilson, M.R. and Martone, J.F. (1996) The epidemiology of chronic open-angle glaucoma and ocular hypertension. In Ritch, R., Shields, M.B. and Kruptin, T. (eds), The Glaucomas. A Multi-volume Reference. Mosby-Year Book, St Louis, MO, pp. 753–768.
- Racette, L., Wilson, M.R., Zangwill, L.M., Weinreb, R.N. and Sample, P.A. (2003) Primary open-angle glaucoma in blacks, a review. Surv. Ophthal., 48, 295–313.
- Mason, R.P., Kosoko, O., Wilson, M.R., Martone, J.F., Cowan, C.L Jr, Gear, J.C. and Ross-Degnan, D. (1989) National survey of the prevalence and risk factors of glaucoma in St Lucia, West Indies. Part I. Prevalence findings. Ophthalmology, 96, 1363–1368.[ISI][Medline]
-
Fingert, J.H., Heon, E., Liebmann, J.M., Yamamoto, T., Craig, J.E, Rait, J., Kawase, K., Hoh, S.T., Buys, Y.M., Dickinson, J. et al. (1999) Analysis of myocilin mutations in 1703 glaucoma patients from five different populations. Hum. Mol. Genet., 8, 899–905.
[Abstract/Free Full Text] - Mader, N. (2003) Population at risk: glaucoma with focus on populations in the Americas. Ocular Surg. News, 22, 70–74.
-
Cooper, R.S., Kaufman, J.S. and Ward, R. (2003) Race and genomics. New Engl. J. Med., 348, 1166–1170.
[Free Full Text] -
Phimister, E.G. (2003) Medicine and the racial divide. New Engl. J. Med., 348, 1081–1082.
[Free Full Text] -
Burchard, E.G., Ziv, E., Coyle, N., Gomez, S.L., Tang, H., Karter, A.J., Mountain, J.L., Perez-Stable, E.J., Sheppard, D. and Risch, N. (2003) The importance of race and ethnic background in biomedical research and clinical practice. New Engl. J. Med., 348, 1170–1175.
[Free Full Text] - Wilson, M.R. (2003) The use of ‘race’ for classification in medicine: is it valid? J. Glaucoma, 12, 293–294.[CrossRef][ISI][Medline]
-
Doris, P.A. (2002) Hypertension genetics, single nucleotide polymorphisms, and the common disease:common variant hypothesis. Hypertension, 39, 323–331.
[Abstract/Free Full Text] - Jansson, M., Marknell, T., Tomic, L., Larsson, L.I. and Wadelius, C. (2003) Allelic variants in the MYOC/TIGR gene in patients with primary open-angle, exfoliative glaucoma and unaffected controls. Ophthal. Genet., 24, 103–110.[CrossRef][Medline]
-
Faucher, M., Anctil, J.L., Rodrigue, M.A., Duchesne, A., Bergeron, D., Blondeau, P., Cote, G., Dubois, S., Bergeron, J., Arseneault, R. et al. (2002) Founder TIGR/myocilin mutations for glaucoma in the Quebec population. Hum. Mol. Genet., 11, 2077–2790.
[Abstract/Free Full Text] - Baird, P.N., Craig, J.E., Richardson, A.J., Ring, M.A., Sim, P., Stanwix, S., Foote, S.J. and Mackey, D.A (2003) Analysis of 15 primary open-angle glaucoma families from Australia identifies a founder effect for the Q368STOP mutation of myocilin. Hum. Genet., 112, 110–116.[CrossRef][ISI][Medline]
-
Pang, C.P., Leung, Y.F., Fan, B., Baum, L., Tong, W.C., Lee, W.S., Chua, J.K., Fan, D.S., Liu, Y. and Lam, D.S. (2002) TIGR/MYOC gene sequence alterations in individuals with and without primary open-angle glaucoma. Invest. Ophthal. Visual Sci., 43, 3231–3235.
[Abstract/Free Full Text] -
Lander, E.S. and Schork, N.J. (1994) Genetic dissection of complex traits. Science, 265, 2037–2048.
[Abstract/Free Full Text] - Lynch, M. and Walsh, B. (1998) Genetics and Data Analysis of Quantitative Traits. Sinauer, Sunderland, MA.
- Fujii, N., Nakashima, H. and Tamamura, H. (2003) The therapeutic potential of CXCR4 antagonists in the treatment of HIV. Expert. Opin. Invest. Drugs, 12, 185–195.
-
Gabriel, S.E., Brigman, K.N., Koller, B.H., Boucher, R.C. and Stutts, M.J. (1994) Cystic fibrosis heterozygote resistance to cholera toxin in the cystic fibrosis mouse model. Science, 266, 107–109.
[Abstract/Free Full Text] - Wiuf, C. (2001) Do DF508 heterozygotes have a selective advantage? Genet. Res., 78, 41–47.[CrossRef][ISI][Medline]
- Benedict, T.W.G., (1842) Abhandlungen aus dem Gebiete der Augenheilkunde. Freunde, Breslau, Poland.
- Alward, W.L, Fingert, J.H., Coote, M.A., Johnson, A.T., Lerner, S.F., Junqua, D., Durcan, F.J., McCartney, P.J., Mackey, D.A., Sheffield, V.C. and Stone, E.M. (1998) Clinical features associated with mutations in the chromosome 1 open-angle glaucoma gene (GLC1