HapMap Provides Potential to Identify Genetic Markers of Respiratory Diseases

Key Point

The large HapMap database provides researchers with information that may help identify genetic markers for respiratory disease.

SAN FRANCISCO —Genetic association mapping could change the way medicine is practiced in the near future. The International HapMap Project—a collaboration of scientists from Japan, the United Kingdom, Canada, China, Nigeria, and the United States that is developing a catalog of common genetic variants that occur in humans—has published its phase I findings and completed the genotyping for phase II.

Over the course of many generations, segments of the ancestral chromosomes in an interbreeding population are shuffled through repeated recombination events. Some of the segments of the ancestral chromosomes are haplotypes that occur as regions of DNA sequences that have not been broken up by recombination and are shared by multiple individuals. The goal of the HapMap Project is to compare the genetic sequences of different individuals to identify chromosomal regions where genetic variants are shared.

While the HapMap Project, which began in 2002, does not establish connections between genetic variants and diseases, it provides a huge library of information for researchers to sift through to identify genetic variants that may increase the disease risk—thus leading to breakthroughs in the prevention, diagnosis, and treatment of some common illnesses. At the 103rd American Thoracic Society International Conference held in San Francisco, a program featuring six speakers presented a comprehensive overview of the HapMap Project and its implications for respiratory diseases.

A GLOBAL STUDY OF HUMAN GENETIC VARIANTS

Daniel O. Stram, PhD, Professor at the University of Southern California Keck School of Medicine in Los Angeles, provided information on the biostatistical underpinnings of the program and spoke of the scope of the program. According to Dr. Stram, samples of trios (two parents and one offspring) were collected from the Yoruba in Western Africa and persons of northern and western European heritage in Utah, as well as from unrelated individuals in Beijing and Tokyo. A total of 270 people were genotyped, yielding six million single nucleotide polymorphisms (SNPs). The HapMap Web site provides access to the data for those who wish to conduct genetic studies.

“Across the human genome, there are areas of stability that resist recombination and areas that are recombination hotspots,” Dr. Stram noted. Recombination helps to create the diversity among humans.

Older populations, such as those centered in Africa, have had more time for recombination and thus have fewer stable block structures in their genes and more areas of recombination than populations outside Africa, he explained. In contrast, newer populations appeared to have gone through a “bottleneck” as they left Africa, which narrowed their genetic pool. These groups are essentially a subset of the haplotypes found in Africa. The HapMap Web site provides detailed information on the haplotype structure of the human genome and Dr. Stram’s presentation gave examples of how the site can be used to help design genetic studies in various populations.

LOOKING FOR THE VARIANTS INVOLVED IN COMMON DISEASE

Christopher Carlson, PhD, of the Fred Hutchinson Cancer Research Center at the University of Washington in Seattle, spoke on the inherited basis of common disease and complex traits. “This is a rapidly evolving field,” he said. “We’ve had findings on prostate cancer, type 2 diabetes—the Genome Project is finally starting to pay off.”

Many common diseases have both common and rare genetic variants underlying them. “In cystic fibrosis, for example, a single allele accounts for 75% of the disease. The disease may be rare, but within that disease, the allele that causes it is common. For other diseases, everything is idiosyncratic. Common diseases will follow a similar pattern—there will be common risk variants and some rare risk variants,” he stated. Although a particular allele may increase or decrease the risk of getting a common disease, such as prostate cancer, it does not cause the disease by itself. Because the statistical power to detect the rare alleles is limited, current investigations have focused on the common alleles.

In trying to determine the genetic roots of disease, it is necessary to look at both SNPs and haplotype, which is a block of SNPs on a gene that has not been broken up by recombination. One way such blocks are defined is through the use of linkage disequilibrium (LD) methods that define blocks as areas of high pairwise LD as compared with surrounding areas of low pairwise LD. LD is the nonrandom association of haplotypes at two linked loci. Over time, random combinations of alleles gradually erode the ancestral LD.

ETHNIC DIVERSITY AND DISEASE

Kathleen C. Barnes, PhD, an Associate Professor at the Johns Hopkins University School of Medicine in Baltimore, presented information on the impact of ethnic diversity on disease. A population structure study using 377 autosomal microsatellite loci in 1,056 individuals from 52 populations found that self-reported population ancestry serves as a suitable proxy for genetic ancestry in most cases and can help determine membership in genetic clusters. She noted that such information also may prove useful to health care practitioners in assessing risk for common diseases among their patients. However, while differences among major population groups accounts for 3% to 5% of human diversity, it is individual differences that account for up to 95% of variations.

“So what about ethnicity and race and their biological relevance to the study of disease?” Dr. Barnes asked. “In complex lung disease we see striking differences in prevalence and risk in different groups. Cystic fibrosis is more prevalent among white people, whereas sickle cell disease with its subset of acute chest syndrome occurs primarily in those of African descent. Diffuse form scleroderma has a very high prevalence among African-American women as compared to whites.”

She added that in those of African heritage, the risk of dying from acute lung injury is considerably higher than it is among those of European heritage. Even in COPD, which is traditionally considered to be more prevalent among white people, the death rate is higher among those who are African-American and have the disease. Asthma is more prevalent, more severe, and causes greater mortality among African-Americans, as well.

Despite more than a dozen genome linkage studies in asthma, little evidence has been uncovered with regard to identifying markers, although studies do suggest that genetic control differs among ethnic minority groups. Identifying the multiple genes likely to be responsible for asthma and asthma risk is a slow process. However, a new tool that has been developed recently—SNP chip technology—promises to speed up the research.

“We’re currently performing a genome-wide study for asthma in two populations of individuals of African descent. We’ve genotyped 650,000 SNPs for each of 2,000 individuals—that’s over 1.3 billion data points. We’ve used tagging SNPs taken from the HapMap data and the SNP panels that were developed commercially from that as well. The first panels were developed from SNPs identified in Utah whites, but fortunately for our group, the company that developed the original 550K panel enriched it with 100K SNPs from the Yoruban group,” Dr. Barnes said.

She noted that the 550K panel does not provide adequate coverage of nonwhite populations. Furthermore, there has been some question as to whether data for the Yoruban population can stand in for African-Americans, as there is tremendous genetic diversity in West Africa and African-Americans likely have genetic links to many different African countries. However, to date is appears that the allelic frequency of African-American to Yoruban samples shows a nice concordance.

Several factors complicate the search for relevant polymorphisms in this population. In general, populations with significant African heritage have reduced levels of LD (and conversely, higher levels of genetic variation), making identification of haplotype blocks more difficult and limiting the usefulness of the genome-wide approach, Dr. Barnes noted. Genetic/environmental interactions, as suggested by the finding that host defense gene polymorphisms appear to be strongly associated with asthma, also complicate the picture in all populations, regardless of ethnicity. Furthermore, environmental factors can vary significantly within and between groups, independent of genetic background.

However, African-Americans with asthma demonstrate different patterns of allergic sensitization, immunoglobulin E levels, and bronchial hyperresponsiveness than European Americans with the disease, suggesting that more than differences in environmental, social, cultural, and economic factors are involved. Data from the HapMap Project have significantly advanced our understanding of genetic variations in asthma. But it is still uncertain as to whether there are polymorphisms for asthma risk that are unique to populations of African descent, Dr. Barnes noted.

APPLICATION OF HAPMAP TO LUNG DISEASE

“The HapMap is one of a series of the most extraordinary tools that have just become available to us in recent years,” said William O. Cookson, MD, PhD, Chair in Respiratory Genetics at the National Heart and Lung Institute and Professor of Respiratory Genetics at Imperial College in London. In 1973, he pointed out, the entire knowledge of human genetic polymorphisms consisted of 27 Mendelian markers and about 57 markers measurable in proteins. Today, the HapMap database contains six million common variants.

“This represents enormous change. The whole of the genome has been sequenced; virtually all 30,000 human genes are known. The HapMap has made virtually all human genetic polymorphisms easily available. We have 19 mammalian genomes, 61 bacterial pathogen genomes, and eight fungal genomes. It’s an extraordinary time,” he said.

In addition to the six million SNPs that can be searched, Dr. Cookson noted that the technology also allows researchers to look for copy number variations, deletions, and micro-rearrangements, which are of particular relevance to cancer, including lung cancer. However, SNP genotyping technologies are the most economical way of searching for mutations in alleles that may be responsible for disease.

In conjunction with Professor Erika von Mutius of Ludwig Maximilians University in Munich, Dr. Cookson is coordinating the GABRIEL project. GABRIEL was launched in May of 2006 and involves more than 150 scientists from 14 countries who are examining genetic and environmental factors in more than 40,000 children and adults with asthma. The aim of the project is to identify how genetics and the environment interact to cause the development of asthma, with the ultimate goal of preventing and treating the illness. One key component of GABRIEL is a genome-wide association study.

Dr. Cookson noted that numerous polymorphisms on chromosome 17q21 have been identified as being associated with asthma and that these findings have been replicated in at least three populations to date. Variations in this gene or gene cluster appear to be associated primarily with childhood asthma and may affect one-third of children with the disease. These early results certainly appear promising.

SEPSIS AND SNPS

Mark M. Wurfel, MD, PhD, Assistant Professor of Medicine in the division of Pulmonary and Critical Care Medicine at Harborview Medical Center, University of Washington in Seattle, pursues basic and translational research on host susceptibility to sepsis and acute lung injury with a focus on genetic factors. He noted that sepsis ranks 10th as a cause of death in the United States. “The mortality with sepsis is high, and there are no good molecular markers of risk,” he emphasized.

The complex pathophysiology of sepsis begins with recognition of bacterial products by the innate immune system, which initiates an inflammatory cascade and induction of multiple processes, resulting in multiple organ failure and death. Toll-like receptor (TLR) pathways appear to play a crucial role in activating the downstream cascade and initiating transcription and secretion of inflammatory mediators. Several TLR pathway genes have been shown in gene association studies to be predictors for the development of sepsis and subsequent mortality.

Dr. Wurfel’s group generated ex vivo inflammatory phenotypes from a large population of healthy volunteers of European-American and African-American heritages. They have genotyped them using a tag SNP approach in multiple TLR genes and identified associations between tag SNPs in TLR genes and response to TLR agonists.

“What we have found is there are highly divergent responses that appear to be influenced by genetic factors. The strongest association was observed between SNPs within TLR-1 and inflammatory responses to bacterial lipopeptides. These SNPs were associated with similar hyperresponsive patterns in both African-American and European American populations,” said Dr. Wurfel.

An SNP just upstream of the coding region of TLR-1 was most highly associated with hyperresponsiveness. Patients who carry this SNP predisposing to elevated TLR-1–mediated responses are more likely to develop poor outcomes in the setting of sepsis, according to Dr. Wurfel.

“If we looked at severe sepsis and septic shock patients in terms of survival, the TLR-1 SNP was associated with higher 28-day mortality. Organ dysfunction showed a similar pattern, with the TLR-1 SNP being associated with fewer days free of organ dysfunction consistently across systems,” he noted. Interestingly, certain TLR-1 alleles appeared to make it more likely that a patient would be infected with a gram-positive rather than a gram-negative organism. “So there appears to be a correlation between the level of response and the organism contracted,” he said.

INDIVIDUALIZED DRUG THERAPY

“Variability of drug response has historically been attributed to differences in kidney and liver function, drug interactions, age, disease state, and the disease phenotype. However, more recently the focus has shifted to genetic contributors to variability in response,” said Julie Johnson, PharmD, Professor and Chair in the Department of Pharmacy Practice at the University of Florida in Gainesville.

There is tremendous potential for the discoveries generated by HapMap and the ensuing genetic studies to have a great impact on the use of drug therapies in clinical practice. With any drug, response can be grouped as good or poor (ie, those who do not respond and need a change of medication), Dr. Johnson explained. It may be possible in the near future to eliminate the current trial and error method of choosing medication for patients and use genetic tests to better tailor therapy to the individual.

Variants in genes for drug metabolizing enzymes may be good targets, and there is excitement around two genes that may have implications for warfarin therapy. “We have a handful of candidate genes for drugs now, and I think we may soon have more. Candidate genes for drugs are easier to pick than genes for diseases, because we know the protein involved,” Dr. Johnson noted. Some disease states allow for easier testing as well. For example, in cancer the tissue of interest is easily available because it is often collected clinically. However, testing for genes involved in drug response in psychiatry is very difficult.

“SNP chip technology has great potential for high translational value in pharmacology,” Dr. Johnson stated and then went on to characterize the efforts of the NIH to capitalize on the many new genetic technologies available. In an effort to further the goal of providing more individualized drug therapy, the NIH supported the formation of the Pharmacogenetics Research Network in 2000. There are 11 research groups and one database group involved in the network, and all the research groups are using HapMap data. The current areas of focus are asthma, cardiovascular disease, nicotine addiction, cancer, and pharmacokinetics. 

Recent results of a trial of bucindolol in patients with heart failure demonstrated that the response to the drug was influenced by the b1-adrenergic receptor genotype. Patients who were Arg-389 homozygotes and received bucindolol had a 38% reduction in mortality and a 34% reduction in mortality or hospitalization as compared with placebo. However, patients who carried the GLY-389 polymorphism had no clinical response to bucindolol as compared with placebo. In asthma, response to albuterol has been shown to vary considerably, depending on the b2-adrenergic receptor polymorphism that the patient carried.

The ability to determine in advance which patients would be most likely to respond to therapy would doubtless save not only time and money spent in the current trial and error method but also, and more importantly, lives. This would be most evident in areas such as cancer therapy, where the failure to use the most effective therapy initially can result in a patient’s death. Haplotype may also affect metabolism and hence the dosing of a drug. In cancer therapy, it may also affect the likelihood of a patient developing a secondary cancer or myelosuppression in response to therapy.

“The HapMap data are a very important tool in pharmacogenetics, but we need to use a strong biological candidate gene methodology. Understanding the functional basis of the polymorphisms involved remains very important as well, and we also need to take a whole gene approach and look for multiple genes that can be pieced together. Warfarin is a great example. The drug metabolism gene explained a lot of the dose variability, and the drug target gene explained much more, and we’re now working on a third gene that will add even more. But in reality, most clinicians practice medicine in an evidence-based way. So we will have to document these results and use that evidence to improve patient outcomes,” Dr. Johnson stated.           

—Laurel McKee Ranger

Suggested Reading
Barnes KC, Grant AVG, Hansel NN, et al. African Americans with asthma: genetic Insights. Proc Am Thorac Soc. 2007;4(1):58-68.
European project on asthma causes launched. 15 May 2006. www.asthma.org.uk/news_media/news/european_project.html. Accessed September 13, 2007.
GABRIEL Summer 2007 Newsletter. www.gabriel-fp6.org. Accessed September 13, 2007.
International HapMap Project. www.hapmap.org. Accessed September 12, 2007.
Kohler JR, Cutler DJ. Simultaneous discovery and testing of deletions for disease association in SNP genotyping studies. Am J Hum Genet. 2007;81(4):684-699.
Liggett SB, Mialet-Perez J, Thaneemit-Chen S, et al. A polymorphism within a conserved b1-adrenergic receptor motif alters cardiac function and b-blocker response in human heart failure. Proc Natl Acad Sci U S A. 2006;103(30):11288-11293.
Mailund T, Besenbacher S, Schierup MH. Whole genome associating mapping by incompatibilities and local perfect phylogenies. BMC Bioinformatics. 2006;7:454.
Rosenberg NA, Pritchard JK, Weber JL, et al. Genetic structure of human populations. Science. 2002;298(5602):2381-2385.
Schwartz R, Halldórsson BV, Bafna V, et al. Robustness of inference of haplotype block structure. J Comput Biol. 2003;10(1):13-19.

Return to table of contents