BETHESDA, Md.—Taking aim at a disease that could afflict one in three adults by the year 2050, a group of National Human Genome Research Institute (NHGRI) researchers has captured what they call the most comprehensive snapshot to date of DNA regions that regulate genes in human pancreatic islet cells, a subset of which produces insulin.
The study highlights the importance of genome regulatory sequences in human health and disease, they say, particularly type 2 diabetes, which affects more than 20 million people in the United States and 200 million people worldwide. The findings appeared recently in Cell Metabolism.
"This study applies the power of epigenomics to a common disease with both inherited and environmental causes," says Dr. Daniel Kastner, scientific director of the NHGRI, which is a part of the National Institutes of Health. "Epigenomic studies are exciting new avenues for genomic analysis, providing the opportunity to peer deeper into genome function, and giving rise to new insights about our genome's adaptability and potential."
According to Kastner, epigenomic research focuses on the mechanisms that regulate the expression of genes in the human genome. Genetic information is written in the chemical language of DNA, a long molecule of nucleic acid wound around specialized proteins called histones. Together, they constitute chromatin, the DNA-protein complex that forms chromosomes during cell division.
The researchers used DNA sequencing technology to search the chromatin of islet cells for specific histone modifications and other signals marking regulatory DNA. Computational analysis of the large amounts of DNA sequence data generated in this study identified different classes of regulatory DNA.
"This study gives us an encyclopedia of regulatory elements in islet cells of the human pancreas that may be important for normal function and whose potential dysfunction can contribute to disease," says senior author and NIH Director Dr. Francis S. Collins. "These elements represent an important component of the uncharted genetic underpinnings of type 2 diabetes that is outside of protein-coding genes."
According to Dr. Michael L. Stitzel, a post-doctoral fellow in Collins' lab, the study used chromatin-based approaches to predict thousands of regulatory elements that may control gene expression in unstimulated human pancreatic islets.
"Several of these elements are unique to the islet and do not appear to be utilized in several other distinct cell types," he says.
This catalog of regulatory elements is important for a few reasons, including revealing molecular mechanisms which underlie the genetic basis of type 2 diabetes and related traits.
"Our study suggests that these elements may play a role in diabetes susceptibility for seven of 18 type 2 diabetes-associated regions," Stitzel says. "Further molecular investigation of these seven will uncover exactly how they influence diabetes. Also, as ongoing genome-wide association analyses for diabetes- or islet-related physiologic traits identify new associated genetic variants, we and others in the islet biology research community can use this catalog to identify regulatory elements that may be disrupted by these variants, leading to islet dysfunction and disease."
Additionally, it can help to improve understanding of transcriptional regulation in the human islet.
"We show that nearly 35 percent of regulatory elements are unique to the islet compared to four other cell types. These islet-unique elements are often clustered together, potentially working in a cooperative manner," Stitzel says. "We and other islet biologists are now well positioned to investigate these elements further to assess how they contribute to islet identity, function and dysfunction."
Moreover, the catalog can create a reference map for future comparative epigenomic studies.
"Our study has generated a reference map of certain epigenomic marks that are present in the resting human islets of healthy individuals," says co-lead author Dr. Praveen Sethupathy, an NHGRI postdoctoral fellow. "Similar profiles can be generated in the future from islets of diabetic individuals, and can be compared to this reference map to identify disease-associated epigenomic marks. This is of significant interest in the field of disease genomics because epigenomic marks can be modified with drugs."
Also, several groups are working very hard to derive insulin-producing islets from other cell types, including embryonic stem cells and induced pluripotent stem cells, as a therapeutic approach for both type 1 and type 2 diabetes.
"We expect that one of the criteria for measuring success will be to ensure that these derived cells have similar epigenomic profiles to the reference map we have created," Sethupathy says.
Among the results, the researchers detected about 18,000 promoters, which are regulatory sequences immediately adjacent to the start of genes. Promoters are like molecular on-off switches and more than one switch can control a gene. Several hundred of these were previously unknown and found to be highly active in the islet cells.
"Along the way, we also hit upon some unexpected but fascinating findings," Sethupathy points out. "For example, some of the most important regulatory DNA in the islet, involved in controlling hormones such as insulin, completely lacked typical histone modifications, suggesting an unconventional mode of gene regulation."
The researchers also identified at least 34,000 distal regulatory elements, so called because they are farther away from the genes. Many of these were bunched together, suggesting they may cooperate to form regulatory modules. These modules may be unique to islets and play an important role in the maintenance of blood glucose levels.
"Genome-wide association studies have told us there are genetic differences between type 2 diabetic and non-diabetic individuals in specific regions of the genome, but substantial efforts are required to understand how these differences contribute to disease," Stitzel says. "Defining regulatory elements in human islets is a critical first step to understanding the molecular and biological effects for some of the genetic variants statistically associated with type 2 diabetes."
The researchers also found that 50 single nucleotide polymorphisms, or genetic variants, associated with islet-related traits or diseases are located within or very close to non-promoter regulatory elements. Variants associated with type 2 diabetes are present in six such elements that function to boost gene activity. These results suggest that regulatory elements may be a key component to understanding the molecular defects that contribute to type 2 diabetes.
The catalog of islet regulatory elements generated in the study provides an openly accessible resource for anyone to reference and ask whether newly emerging, statistically associated variants are falling within these regulatory elements. The raw data can be found at the National Center for Biotechnology Information's Gene Expression Omnibus using accession number GSE23784.
"These findings represent important strides that were not possible just five years ago, but that are now realized with advances in genome sequencing technologies," says NHGRI Director Dr. Eric D. Green. "The power of DNA sequencing is allowing us to go from studies of a few genes at a time to profiling the entire genome. The scale is tremendously expanded."
Stitzel explains that the idea to generate a chromatin map in human islets came fast on the heels of some of the initial ChIP-seq studies in human CD4 T Cells by Keji Zhao's group and in mouse pluripotent and lineage-committed cells by Brad Bernstein's group in 2007, but the results of the NHGRI research didn't happen overnight.
"It took some time to collect the appropriate biological materials, namely human pancreatic islets, and to optimize preparation and experimental methods for DNase and chromatin immunoprecipitation studies in those islets," he says. "We started the study in earnest in January 2009, so it has taken just under two years to complete."
While the team's ongoing research still is in its early stages—both in identifying the regulatory elements in islets and studying the effects that diabetes-associated variants have on their function—Sethupathy says the researchers believe identifying these elements will help identify important genes as well as molecular defects caused by diabetes-predisposing variants. This is an important step in developing therapeutics.
"For example, if a diabetes-predisposing variant causes increased activity of a certain gene through a particular regulatory element, we would screen for drugs that would decrease its activity," he says.
The role of genome regulatory sequences in diabetes remains up in the air, and the research team is only scratching the surface of the non-coding regulatory landscape of the human genome.
"In our study, the islet chromatin profiling approach has identified potential regulatory elements containing type 2 diabetes-associated SNPs for 7/18 loci investigated, so we speculate that it may play a role in up to 1/3 of the type 2 diabetes-associated regions," explains Stitzel. "As more studies investigate the role of regulatory variation in diabetes, we expect the numbers will rise. One recent example is the work of Garin et al., 2010 (PNAS), in which they identified pathogenic mutations in the Insulin (INS) promoter that cause neonatal diabetes. When we have a more complete picture of both the location of genetic variants predisposing individuals to diabetes and the regulatory elements used by islets and other diabetes-relevant cell types, we will be in better shape to make more definitive conclusions."
There is hope, Sethupathy notes, because epigenomic research is proving to be the path for new, effective diabetes drug candidates.
"We believe it is a critical approach to understand the molecular mechanisms underlying progression to diabetes, both as a tool to identify important regulatory elements in islets and other diabetes-relevant tissues (adipose, muscle, liver) and as a method to identify epigenomic changes that may contribute to disease," he notes.
As part of their ongoing efforts, the researchers have cataloged regions that are likely regulatory elements.
"Now we need to understand how they exert their function—that is, what factors activate these elements, and what genes these elements control," Stitzel says. "This is vital to be able to understand how diabetes-associated variants alter regulatory activity."
Going forward, Sethupathy says researchers anticipate this line of investigation will allow the team to elucidate both the regulatory mechanisms and the specific genes and pathways disrupted by diabetes-predisposing genetic variants.
"This should provide opportunities for therapeutic intervention in the future," he says. "We also hope that this study will allow others in the diabetes research community to assess the role of non-protein coding variation in diabetes and provide annotation to variants residing in these regions."
In addition to the NHGRI and the NIH Intramural Sequencing Center, researchers from Duke University in Durham, N.C. and the University of Michigan in Ann Arbor, contributed to the study.
Previously known as adult-onset diabetes or non-insulin dependent diabetes mellitus (NIDDM), type 2 diabetes usually appears after age 40, often in overweight, sedentary people. However, a growing number of younger people—and even children—are developing the disease.
Diabetes is a major cause of heart disease and stroke in U.S. adults, as well as the most common cause of blindness, kidney failure and amputations not related to trauma. Type 2 diabetes is characterized by the resistance of target tissues to respond to insulin, which controls glucose levels in the blood. This leads to gradual failure of insulin-secreting cells in the pancreatic islets.
The study highlights the importance of genome regulatory sequences in human health and disease, they say, particularly type 2 diabetes, which affects more than 20 million people in the United States and 200 million people worldwide. The findings appeared recently in Cell Metabolism.
"This study applies the power of epigenomics to a common disease with both inherited and environmental causes," says Dr. Daniel Kastner, scientific director of the NHGRI, which is a part of the National Institutes of Health. "Epigenomic studies are exciting new avenues for genomic analysis, providing the opportunity to peer deeper into genome function, and giving rise to new insights about our genome's adaptability and potential."
According to Kastner, epigenomic research focuses on the mechanisms that regulate the expression of genes in the human genome. Genetic information is written in the chemical language of DNA, a long molecule of nucleic acid wound around specialized proteins called histones. Together, they constitute chromatin, the DNA-protein complex that forms chromosomes during cell division.
The researchers used DNA sequencing technology to search the chromatin of islet cells for specific histone modifications and other signals marking regulatory DNA. Computational analysis of the large amounts of DNA sequence data generated in this study identified different classes of regulatory DNA.
"This study gives us an encyclopedia of regulatory elements in islet cells of the human pancreas that may be important for normal function and whose potential dysfunction can contribute to disease," says senior author and NIH Director Dr. Francis S. Collins. "These elements represent an important component of the uncharted genetic underpinnings of type 2 diabetes that is outside of protein-coding genes."
According to Dr. Michael L. Stitzel, a post-doctoral fellow in Collins' lab, the study used chromatin-based approaches to predict thousands of regulatory elements that may control gene expression in unstimulated human pancreatic islets.
"Several of these elements are unique to the islet and do not appear to be utilized in several other distinct cell types," he says.
This catalog of regulatory elements is important for a few reasons, including revealing molecular mechanisms which underlie the genetic basis of type 2 diabetes and related traits.
"Our study suggests that these elements may play a role in diabetes susceptibility for seven of 18 type 2 diabetes-associated regions," Stitzel says. "Further molecular investigation of these seven will uncover exactly how they influence diabetes. Also, as ongoing genome-wide association analyses for diabetes- or islet-related physiologic traits identify new associated genetic variants, we and others in the islet biology research community can use this catalog to identify regulatory elements that may be disrupted by these variants, leading to islet dysfunction and disease."
Additionally, it can help to improve understanding of transcriptional regulation in the human islet.
"We show that nearly 35 percent of regulatory elements are unique to the islet compared to four other cell types. These islet-unique elements are often clustered together, potentially working in a cooperative manner," Stitzel says. "We and other islet biologists are now well positioned to investigate these elements further to assess how they contribute to islet identity, function and dysfunction."
Moreover, the catalog can create a reference map for future comparative epigenomic studies.
"Our study has generated a reference map of certain epigenomic marks that are present in the resting human islets of healthy individuals," says co-lead author Dr. Praveen Sethupathy, an NHGRI postdoctoral fellow. "Similar profiles can be generated in the future from islets of diabetic individuals, and can be compared to this reference map to identify disease-associated epigenomic marks. This is of significant interest in the field of disease genomics because epigenomic marks can be modified with drugs."
Also, several groups are working very hard to derive insulin-producing islets from other cell types, including embryonic stem cells and induced pluripotent stem cells, as a therapeutic approach for both type 1 and type 2 diabetes.
"We expect that one of the criteria for measuring success will be to ensure that these derived cells have similar epigenomic profiles to the reference map we have created," Sethupathy says.
Among the results, the researchers detected about 18,000 promoters, which are regulatory sequences immediately adjacent to the start of genes. Promoters are like molecular on-off switches and more than one switch can control a gene. Several hundred of these were previously unknown and found to be highly active in the islet cells.
"Along the way, we also hit upon some unexpected but fascinating findings," Sethupathy points out. "For example, some of the most important regulatory DNA in the islet, involved in controlling hormones such as insulin, completely lacked typical histone modifications, suggesting an unconventional mode of gene regulation."
The researchers also identified at least 34,000 distal regulatory elements, so called because they are farther away from the genes. Many of these were bunched together, suggesting they may cooperate to form regulatory modules. These modules may be unique to islets and play an important role in the maintenance of blood glucose levels.
"Genome-wide association studies have told us there are genetic differences between type 2 diabetic and non-diabetic individuals in specific regions of the genome, but substantial efforts are required to understand how these differences contribute to disease," Stitzel says. "Defining regulatory elements in human islets is a critical first step to understanding the molecular and biological effects for some of the genetic variants statistically associated with type 2 diabetes."
The researchers also found that 50 single nucleotide polymorphisms, or genetic variants, associated with islet-related traits or diseases are located within or very close to non-promoter regulatory elements. Variants associated with type 2 diabetes are present in six such elements that function to boost gene activity. These results suggest that regulatory elements may be a key component to understanding the molecular defects that contribute to type 2 diabetes.
The catalog of islet regulatory elements generated in the study provides an openly accessible resource for anyone to reference and ask whether newly emerging, statistically associated variants are falling within these regulatory elements. The raw data can be found at the National Center for Biotechnology Information's Gene Expression Omnibus using accession number GSE23784.
"These findings represent important strides that were not possible just five years ago, but that are now realized with advances in genome sequencing technologies," says NHGRI Director Dr. Eric D. Green. "The power of DNA sequencing is allowing us to go from studies of a few genes at a time to profiling the entire genome. The scale is tremendously expanded."
Stitzel explains that the idea to generate a chromatin map in human islets came fast on the heels of some of the initial ChIP-seq studies in human CD4 T Cells by Keji Zhao's group and in mouse pluripotent and lineage-committed cells by Brad Bernstein's group in 2007, but the results of the NHGRI research didn't happen overnight.
"It took some time to collect the appropriate biological materials, namely human pancreatic islets, and to optimize preparation and experimental methods for DNase and chromatin immunoprecipitation studies in those islets," he says. "We started the study in earnest in January 2009, so it has taken just under two years to complete."
While the team's ongoing research still is in its early stages—both in identifying the regulatory elements in islets and studying the effects that diabetes-associated variants have on their function—Sethupathy says the researchers believe identifying these elements will help identify important genes as well as molecular defects caused by diabetes-predisposing variants. This is an important step in developing therapeutics.
"For example, if a diabetes-predisposing variant causes increased activity of a certain gene through a particular regulatory element, we would screen for drugs that would decrease its activity," he says.
The role of genome regulatory sequences in diabetes remains up in the air, and the research team is only scratching the surface of the non-coding regulatory landscape of the human genome.
"In our study, the islet chromatin profiling approach has identified potential regulatory elements containing type 2 diabetes-associated SNPs for 7/18 loci investigated, so we speculate that it may play a role in up to 1/3 of the type 2 diabetes-associated regions," explains Stitzel. "As more studies investigate the role of regulatory variation in diabetes, we expect the numbers will rise. One recent example is the work of Garin et al., 2010 (PNAS), in which they identified pathogenic mutations in the Insulin (INS) promoter that cause neonatal diabetes. When we have a more complete picture of both the location of genetic variants predisposing individuals to diabetes and the regulatory elements used by islets and other diabetes-relevant cell types, we will be in better shape to make more definitive conclusions."
There is hope, Sethupathy notes, because epigenomic research is proving to be the path for new, effective diabetes drug candidates.
"We believe it is a critical approach to understand the molecular mechanisms underlying progression to diabetes, both as a tool to identify important regulatory elements in islets and other diabetes-relevant tissues (adipose, muscle, liver) and as a method to identify epigenomic changes that may contribute to disease," he notes.
As part of their ongoing efforts, the researchers have cataloged regions that are likely regulatory elements.
"Now we need to understand how they exert their function—that is, what factors activate these elements, and what genes these elements control," Stitzel says. "This is vital to be able to understand how diabetes-associated variants alter regulatory activity."
Going forward, Sethupathy says researchers anticipate this line of investigation will allow the team to elucidate both the regulatory mechanisms and the specific genes and pathways disrupted by diabetes-predisposing genetic variants.
"This should provide opportunities for therapeutic intervention in the future," he says. "We also hope that this study will allow others in the diabetes research community to assess the role of non-protein coding variation in diabetes and provide annotation to variants residing in these regions."
In addition to the NHGRI and the NIH Intramural Sequencing Center, researchers from Duke University in Durham, N.C. and the University of Michigan in Ann Arbor, contributed to the study.
Previously known as adult-onset diabetes or non-insulin dependent diabetes mellitus (NIDDM), type 2 diabetes usually appears after age 40, often in overweight, sedentary people. However, a growing number of younger people—and even children—are developing the disease.
Diabetes is a major cause of heart disease and stroke in U.S. adults, as well as the most common cause of blindness, kidney failure and amputations not related to trauma. Type 2 diabetes is characterized by the resistance of target tissues to respond to insulin, which controls glucose levels in the blood. This leads to gradual failure of insulin-secreting cells in the pancreatic islets.
