Editorial
In Search of Perverse Polymorphisms
N Engl J Med 1998; 338:122-124January 8, 1998
- Article
By the time the Beagle left the waters of the remote Galapagos Islands, Charles Darwin had still not realized that the natural variations within a species would provide the means for its evolution. Later, the Galapagos finches furnished ideal support for his theory of natural selection, because after a long separation from the finches of the rest of the world, they had become highly inbred. Geographic isolation had restricted their gene pool, so that the subtle differences Darwin saw among the island finches must have resulted from genetic changes, or polymorphisms, that arose spontaneously in individual birds and were then bequeathed to subsequent generations through germ-line DNA.
Genetic polymorphisms underlie the diversity of finches and humans alike. Each of us carries a vast library of different polymorphisms. Such inherited changes in DNA structure are generally neutral when they occur in the long stretches of DNA that lie outside the genes, and they are often benign when they affect the function of genes such as those that determine the color of a son's eyes or the shape of a daughter's chin. These alternative versions of a gene are called alleles. The frequency of polymorphic changes in a gene is between 10-4 and 10-7 each generation. This means that about 1 in every 10 persons is likely to have acquired a new allele in a particular gene from one parent — a rough estimate, because of “hot spots” in which the frequency of polymorphisms is relatively high.
Occasionally, a polymorphism produces an allele that causes a fatal defect or increases susceptibility to a particular disease, in which case it is considered a mutation. Harmful mutations are relatively rare within a species, whereas benign polymorphisms are common. Inherited diseases due to specific mutations represent extreme cases within a spectrum of conditions that may be caused or influenced by natural genetic variation. The clinical implication of a mutation must therefore be measured against the background noise of harmless polymorphisms. Attempts to link a particular change in a patient's DNA to a disease are often plagued by this difficulty: when is a polymorphism benign, and when is it perverse?
Two papers in this issue of the Journal illustrate the problem. The first, by Iacoviello et al.,1 concerns the relation of myocardial infarction to variants of the factor VII gene that have been linked to the plasma level of activated factor VII (factor VIIa). In a previous study of nearly 1500 men who were followed for an average of 16 years, an elevated plasma level of factor VIIa was a strong risk factor for fatal myocardial infarction.2 Moreover, other risk factors for myocardial infarction, such as a high-fat diet and diabetes, raise plasma levels of factor VIIa.3,4 Factor VII is central to blood clotting, but powerless without tissue factor.5 Normally, tissue factor lies hidden beneath the endothelium, but in the wake of trauma or inflammation, it enters the zone of injury and initiates the conversion of the endothelium from an anticoagulant to a procoagulant surface by activating factor VII. In a cut vessel this change is salubrious, but in a coronary artery it is a Jekyll-to-Hyde transformation sparked by the rupture of an atherosclerotic plaque.6
Iacoviello et al. studied two factor VII polymorphisms. In one, amino acid residue 353 of factor VII is either glutamine or arginine (Q or R, respectively, to use the chemist's code). About 20 percent of the population has the RQ or QQ genotype. As compared with pooled plasma, plasma of QQ carriers has about one third less factor VIIa, and plasma of heterozygotes (RQ ) has about 20 percent less.7 The other polymorphism involves three variants within intron 7 of the factor VII gene (the gene has eight of these noncoding regions): H5 (very rare), H6 (common), and H7 (intermediate). People with the H7H7 genotype have the lowest plasma levels of factor VIIa.8 Iacoviello et al. determined the proportions of these polymorphisms among patients with and those without a history of myocardial infarction in a case–control study.
The key finding was that the genotypes linked to the lowest levels of factor VIIa were less frequent among case patients than among controls: 1 of 164 case patients had the QQ genotype, as compared with 10 of 224 controls, and 12 of 165 case patients had the H7H7 genotype, as compared with 31 of 225 controls. These differences are statistically significant, but the wide confidence intervals found when the genotypes were assessed as risk factors for myocardial infarction raise a cautionary flag. A curious aspect of the group with myocardial infarction was that the QQ and H7H7 genotypes were not associated with particularly low levels of factor VII. The authors attribute this anomaly to an insufficient number of subjects. Perhaps so, but the possibility of unrecognized confounders or selection bias must be acknowledged. It is well to remember that only about one third of the variation in factor VII levels is due to polymorphisms of the factor VII gene.9
In previous work, Wang et al.10 could not relate the QQ genotype to the number and severity of diseased coronary arteries, assessed angiographically, in 545 Australian patients, and Lane et al.11 found that the RQ genotype influences plasma levels of factor VIIa but not the risk of myocardial infarction. Others have also questioned the link between polymorphisms of factor VIIa and nonfatal myocardial infarction.2 For all these reasons, the study by Iacoviello et al. requires conservative interpretation.
The second paper, by Kuivenhoven et al.,12 deals with a possible genetic influence on the response to pravastatin: do polymorphisms of the gene for cholesteryl ester transfer protein (CETP) modify the effect of pravastatin against coronary artery disease? CETP uses high-density lipoprotein (HDL) and other lipoproteins to transfer cholesterol from tissues to the liver, whence the cholesterol exits through the bile. This mechanism can account for the antiatherogenic effect of HDL. However, a high plasma level of CETP can be too much of a good thing, because it depresses plasma concentrations of HDL.13 A polymorphism of the CETP gene termed TaqIb correlates with plasma levels of CETP and HDL. There are two TaqIb variants, B1 and B2; carriers of two B1 alleles have lower plasma concentrations of HDL and higher amounts of CETP than people with the B2B2 genotype.14
Kuivenhoven et al. examined these polymorphisms in 807 men who were enrolled in a placebo-controlled trial assessing the ability of pravastatin to reverse diffuse and focal angiographic changes indicating atherosclerosis in coronary arteries. In the placebo group, progression of coronary artery disease was most pronounced in B1B1 carriers, intermediate in B1B2 carriers, and least pronounced in B2B2 carriers. In the pravastatin group, B1B1 carriers fared better than those with the other two genotypes, and B1B1 carriers who took pravastatin had less progression of coronary artery disease than B1B1 carriers in the control group. The authors suggest that the CETP polymorphism can predict who will benefit from pravastatin. But we must note that the decreases in minimal luminal diameter after two years among B1B1 carriers in the placebo and pravastatin groups differed by 0.09 mm — smaller than the dot on this “i.”
What are clinicians to make of these studies? The real concern is not whether the correlation between a particular allelic configuration and a disease or response to treatment is statistically significant, but the practical value of the correlation. Polymorphisms are technically easy to study, but showing their clinical usefulness is far more difficult. At one end of the spectrum of cases a physician is likely to face is an allele that has a uniform link to a well-defined disease and that arises in a gene that is clearly involved in the pathogenesis of the disease. This kind of polymorphism can provide the means for diagnosis, prognosis, and even treatment. At the other extreme is an allele that may be frequent in a disorder but whose role is not manifest unless another “modifier” gene changes. Determining the clinical importance of a polymorphism in this case requires deep insight into the genetic makeup of the population of patients under study.
Several criteria must be met in establishing medically useful links between polymorphisms and disease. First, it is essential to show that the change in the gene under study causes a relevant alteration in the function or level of the gene product (which is a protein). Otherwise, if the variant is sufficiently prevalent within the test group, it could merely track with the disease but have no essential relation to it. The Iacoviello and Kuivenhoven studies meet this criterion only indirectly, since in both cases the readout of the allelic variation is a complex trait that cannot be attributed exclusively to the function of a single gene. Second, the number of cases associating an allele with a particular phenotype must be large enough to be convincing. The Iacoviello study includes only 12 persons with the “worst” combinations of alleles (H7H5 and H6H5, with the highest levels of factor VIIa), making confident extrapolations difficult. Third, the beneficial and harmful phenotypes being studied must have clear-cut clinical differences — so distinctive and characteristic that their assignment to one or another allelic group is unequivocal. It is uncertain whether the Kuivenhoven study meets this criterion, because of the very small differences in the angiographic measurements. Longer follow-up, however, may reveal clinically relevant differences. Last, the plausibility of the hypothesis must be convincing. The role of factor VII in myocardial infarction seems important, but it remains to be seen whether a 20 or 30 percent difference in the plasma level of the protein affects susceptibility to such a complex event.
Sorting through these issues is not trivial, because the effect (penetrance) of a mutation can differ widely. The good news is that a limited number of clinically relevant alleles occur in any given gene in the general human population — somewhere between 4 and 26. Even the extreme polymorphism of the HLA locus has revealed only a few alleles with a substantial link to pathology. In the future, with the entire human genome represented on microchips, it will be possible to screen collections of allelic variants in search of meaningful associations between specific allelic combinations and an increased (or decreased) susceptibility to a particular disease. Although human geneticists do not have the advantage of turning to an inbred population such as an isolated group of finches to pinpoint the mutation they suspect underlies a disorder, they may soon have molecular tools powerful enough to distinguish benign polymorphisms from the perverse polymorphic combinations that contribute to a disease in a single patient. Darwin would have a field day.
Nadia Rosenthal, Ph.D.
Robert S. Schwartz, M.D.- References
References
1
Iacoviello L ,Di Castelnuovo A ,De Knijff P , et al. Polymorphisms in the coagulation factor VII gene and risk of myocardial infarction. N Engl J Med 1998;338:79-85
Full Text | Web of Science | Medline2
Meade TW ,Mellows S ,Brozovic M , et al. Haemostatic function and ischaemic heart disease: principal results of the Northwick Park Heart Study. Lancet 1986;2:533-537
CrossRef | Web of Science | Medline3
Miller GJ ,Cruickshank JK ,Ellis LJ , et al. Fat consumption and factor VII coagulant activity in middle-aged men: an association between dietary and thrombogenic coronary risk factor. Atherosclerosis 1989;78:19-24
CrossRef | Web of Science | Medline4
Fuller JH ,Keen H ,Jarrett RJ , et al. Haemostatic variables associated with diabetes and its complications. BMJ 1979;2:964-966
CrossRef | Web of Science | Medline5
Banner DW . The factor VIIa/tissue factor complex. Thromb Haemost 1997;78:512-515
Web of Science | Medline6
Fuster V ,Badimon J ,Chesebro JH ,Fallon JT . Plaque rupture, thrombosis, and therapeutic implications. Haemostasis 1996;26:Suppl 4:269-284
Medline7
Green F ,Kelleher C ,Wilkes H ,Temple A ,Meade T ,Humphries S . A common genetic polymorphism associated with lower coagulation factor VII levels in healthy individuals. Arterioscler Thromb 1991;11:540-546
CrossRef | Medline8
Marchetti G ,Patracchini P ,Gemmati D , et al. Detection of two missense mutations and characterization of a repeat polymorphism in the factor VII gene (F7). Hum Genet 1992;89:497-502
CrossRef | Web of Science | Medline9
Bernardi F ,Marchetti G ,Pinotti M , et al. Factor VII gene polymorphisms contribute about one third of the factor VII level variation in plasma. Arterioscler Thromb Vasc Biol 1996;16:72-76
Web of Science | Medline10
Wang XL ,Wang J ,McCredie RM ,Wilcken DE . Polymorphisms of factor V, factor VII, and fibrinogen genes: relevance to severity of coronary artery disease. Arterioscler Thromb Vasc Biol 1997;17:246-251
CrossRef | Web of Science | Medline11
Lane A ,Green F ,Scarabin PY , et al. Factor VII Arg/Gln353 polymorphism determines factor VII coagulant activity in patients with myocardial infarction (MI) and control subjects in Belfast and France but is not a strong indicator of MI risk in the ECTIM study. Atherosclerosis 1996;119:119-127
CrossRef | Web of Science | Medline12
Kuivenhoven JA ,Jukema JW ,Zwinderman AH , et al. The role of a common variant of the cholesteryl ester transfer protein gene in the progression of coronary atherosclerosis. N Engl J Med 1998;338:86-93
Full Text | Web of Science | Medline13
Tall A ,Sharp D ,Zhong S , et al. Cholesteryl ester transfer protein and atherogenesis. Ann N Y Acad Sci 1997;811:178-182
CrossRef | Web of Science | Medline14
Kuivenhoven JA ,de Knijff P ,Boer JM , et al. Heterogeneity at the CETP gene locus: influence on plasma CETP concentrations and HDL cholesterol levels. Arterioscler Thromb Vasc Biol 1997;17:560-568
CrossRef | Web of Science | Medline
- Citing Articles (16)
Citing Articles
1
Stella Trompet, Anton J.M. de Craen, J. Wouter Jukema, D. Pons, P. Eline Slagboom, D. Kremer, Eduard L.E.M. Bollen, Rudi G.J. Westendorp. (2011) Variation in the CBP gene involved in epigenetic control associates with cognitive function. Neurobiology of Aging 32:3, 549.e1-549.e8
CrossRef2
Eric A. J. Hoste, Jan J. De Waele. (2008) Acute kidney injury in the intensive care unit: It’s the gene, stupid!*. Critical Care Medicine 36:12, 3266-3267
CrossRef3
Orfeas Liangos, Bertrand L. Jaber. (2008) Multiple Organ Dysfunction Syndrome in Children With Sepsis: Role of Genetic Factors. Seminars in Nephrology 28:5, 499-509
CrossRef4
P. Puddu, E. Cravero, G. M. Puddu, A. Muscari. (2005) Genes and atherosclerosis: at the origin of the predisposition. International Journal of Clinical Practice 59:4, 462-472
CrossRef5
Bertrand L. Jaber, Brian J.G. Pereira, Joseph V. Bonventre, Vaidyanathapuram S. Balakrishnan. (2005) Polymorphism of host response genes: Implications in the pathogenesis and treatment of acute renal failure. Kidney International 67:1, 14-33
CrossRef6
Kersten M. Small, Dennis W. McGraw, Stephen B. Liggett. (2003) P
HARMACOLOGY AND PHYSIOLOGY OF HUMAN ADRENERGIC RECEPTOR POLYMORPHISMS . Annual Review of Pharmacology and Toxicology 43:1, 381-411
CrossRef7
T. Yamamoto, T. Sato, M. Hosoi, K. Yoshioka, S. Tanaka, H. Tahara, Y. Nishizawa, S. Fujii. (2003) Aldose reductase gene polymorphism is associated with progression of diabetic nephropathy in Japanese patients with type 1 diabetes mellitus. Diabetes, Obesity and Metabolism 5:1, 51-57
CrossRef8
Jacquelyn Long, Chandice Covington, Virginia Delaney-Black, Beth Nordstrom. (2002) Allelic Variation and Environmental Lead Exposure in Urban Children. AACN Clinical Issues: Advanced Practice in Acute and Critical Care 13:4, 550-556
CrossRef9
Michael M Hoffmann, Bernhard R Winkelmann, Heinrich Wieland, Winfried März. (2001) The significance of genetic polymorphisms in modulating the response to lipid-lowering drugs. Pharmacogenomics 2:2, 107-121
CrossRef10
J. WOUTER JUKEMA, JOHN J.P. KASTELEIN. (2000) Tailored Therapy to Fit Individual Profiles: Genetics and Coronary Artery Disease. Annals of the New York Academy of Sciences 902:1, 17-26
CrossRef11
Karen Horsburgh, Mark O McCarron, Fiona White, James A.R Nicoll. (2000) The role of apolipoprotein E in Alzheimer’s disease, acute brain injury and cerebrovascular disease: evidence of common mechanisms and utility of animal models. Neurobiology of Aging 21:2, 245-255
CrossRef12
Adnan Kastrati, Josef Dirschinger, Albert Schömig. (2000) Genetic risk factors and restenosis after percutaneous coronary interventions. Herz 25:1, 34-46
CrossRef13
B. Kennon, J. R. Petrie, M. Small, J. M. C. Connell. (1999) Angiotensin-converting enzyme gene and diabetes mellitus. Diabetic Medicine 16:6, 448-458
CrossRef14
Ellis J. Neufeld. (1998) UPDATE ON GENETIC RISK FACTORS FOR THROMBOSIS AND ATHEROSCLEROTIC VASCULAR DISEASE. Hematology/Oncology Clinics of North America 12:6, 1193-1209
CrossRef15
Richard Mayeux, Ruth Ottman. (1998) Alzheimer's disease genetics: Home runs and strikeouts. Annals of Neurology 44:5, 716-719
CrossRef16
Altshuler, David, Kruglyak, Leonid, Lander, Eric, . (1998) Genetic Polymorphisms and Disease. New England Journal of Medicine 338:22, 1626-1626
Full Text
