Organisms ranging from bacteria to humans all rely on circadian rhythms as their molecular clocks. This complex network of molecular feedback loops not only helps to control wake-sleep patterns, but also helps to regulate body temperature, hormone levels, and metabolism, along with other systems. Whether it is from the disruption of sleep patterns through jetlag or seasonal affective disorder from the lack of sunlight, external stimuli can cause physical disturbances that are easily observable and can have serious outcomes. Along with these, perturbations can also occur that directly affect the molecular gears that keep the system running smoothly.
Although established Mendelian disorders such as familial advanced sleep-phase syndrome have been linked to circadian genes, complex disorders have remained more elusive. Recently, 3 papers published in Nature Genetics [1-3] reported such a link: a variant found within the melatonin receptor 1B (MTNR1B) is associated with both type 2 diabetes and fasting blood glucose levels.
Last summer, the first genome-wide association studies (GWAS) for fasting blood glucose levels were published using longitudinal cohorts.[4-6] Three markers were significantly associated. All were located in genes that were not surprising. The variants existed in the genes of glucokinase, glucokinase regulatory protein, and islet-specific glucose-6-phosphatase or glucose-6-phosphatase catalytic unit 2.[4-6] These 3 genes represented the most significant findings, but several other variants were near statistical significance after multiple hypothesis corrections for hundreds of thousands of tests. To gain statistical power, all 3 groups performed typing on an additional set of samples, thereby increasing their ability to discover alleles of lower penetrance. The 3 published papers represent the results of typing more than 60,000 individuals, leading to enough power to discern an association in MTNR1B as a true variant for fasting glucose levels.
Two papers, one by Lyssenko and colleagues[1] and the other by Prokopenko and colleagues on behalf of MAGIC (the Meta-Analysis of Glucose and Insulin-related traits Consortium),[2] along with several other groups who shared region-specific data, homed in on the same single nucleic polymorphism (SNP) within the MTNR1B locus. The SNP, rs10830963, maps to an 11.5-kilobase intron (regulatory region, not a protein-coding element) and does not appear to affect any known transcription factor-binding sites or cryptic splice sites. The effect of the variant reported from MAGIC and the additional cohorts of its meta-analysis, based on data from 35,812 subjects, had a per-allele effect of 0.072 mmol/L glucose and a P value of 3.2 x 10-50.[2] The 2 other studies, although they used slightly smaller datasets, showed a similar effect size and significance.
One difference in the study by Bouatia-Naji and colleagues[3] was that their highest association came from a different SNP, rs1387153. Although the strength of the association was very similar, this is a good example why functional analysis is the critical next step for all GWAS to determine the true causal variants. For MTNR1B, the true signal for association could be from either of the SNPs because they are in moderate linkage disequilibrium. Alternatively, the association could be from a separate variant for which the markers are both tagging. It is unlikely that the SNPs represent 2 separate signals because neither is significant after conditionally correcting for each other.
The ultimate goal of all GWAS is translation of the findings to a clinical setting, where they can have a positive impact on patients. However, this requires much more than just statistical significance. The current value of GWAS lies with the biological networks that the variant exposes as an underpinning of the phenotype. Both Lyssenko and colleagues[1] and Bouatia-Naji and colleagues[3] demonstrated direct expression of the MTNR1B gene in pancreatic beta cells. This finding, independently shown by both groups, reverses previous observations.[7] Preliminary data suggest that the G allele is linked with increased fasting blood glucose levels and may affect the MTNR1B transcription levels in the pancreatic islets of subjects older than 45 years of age.[1]
These 3 papers provide another variant for fasting blood glucose levels, bringing the total attributable variance accounted for by the 4 known markers to around 1.5%.[2] It is of interest that while the association to fasting blood glucose levels is strong, the variant's link to type 2 diabetes is still very weak. This could be because conversion to type 2 diabetes status occurred in few patients in the longitudinal cohort during the study collection and follow-up periods. It could also be a hint that there are possible subtypes of diabetes that are manifested through different pathways, all of which lead to the same end-phenotype. Type 2 diabetes may actually represent several subtypes at the molecular, root level. Regardless, these findings strengthen prior research that our circadian rhythms can have strong influences on our metabolic system. Understanding and targeting malfunctions within our internal clock could prove a viable therapeutic strategy for type 2 diabetes
Although established Mendelian disorders such as familial advanced sleep-phase syndrome have been linked to circadian genes, complex disorders have remained more elusive. Recently, 3 papers published in Nature Genetics [1-3] reported such a link: a variant found within the melatonin receptor 1B (MTNR1B) is associated with both type 2 diabetes and fasting blood glucose levels.
Last summer, the first genome-wide association studies (GWAS) for fasting blood glucose levels were published using longitudinal cohorts.[4-6] Three markers were significantly associated. All were located in genes that were not surprising. The variants existed in the genes of glucokinase, glucokinase regulatory protein, and islet-specific glucose-6-phosphatase or glucose-6-phosphatase catalytic unit 2.[4-6] These 3 genes represented the most significant findings, but several other variants were near statistical significance after multiple hypothesis corrections for hundreds of thousands of tests. To gain statistical power, all 3 groups performed typing on an additional set of samples, thereby increasing their ability to discover alleles of lower penetrance. The 3 published papers represent the results of typing more than 60,000 individuals, leading to enough power to discern an association in MTNR1B as a true variant for fasting glucose levels.
Two papers, one by Lyssenko and colleagues[1] and the other by Prokopenko and colleagues on behalf of MAGIC (the Meta-Analysis of Glucose and Insulin-related traits Consortium),[2] along with several other groups who shared region-specific data, homed in on the same single nucleic polymorphism (SNP) within the MTNR1B locus. The SNP, rs10830963, maps to an 11.5-kilobase intron (regulatory region, not a protein-coding element) and does not appear to affect any known transcription factor-binding sites or cryptic splice sites. The effect of the variant reported from MAGIC and the additional cohorts of its meta-analysis, based on data from 35,812 subjects, had a per-allele effect of 0.072 mmol/L glucose and a P value of 3.2 x 10-50.[2] The 2 other studies, although they used slightly smaller datasets, showed a similar effect size and significance.
One difference in the study by Bouatia-Naji and colleagues[3] was that their highest association came from a different SNP, rs1387153. Although the strength of the association was very similar, this is a good example why functional analysis is the critical next step for all GWAS to determine the true causal variants. For MTNR1B, the true signal for association could be from either of the SNPs because they are in moderate linkage disequilibrium. Alternatively, the association could be from a separate variant for which the markers are both tagging. It is unlikely that the SNPs represent 2 separate signals because neither is significant after conditionally correcting for each other.
The ultimate goal of all GWAS is translation of the findings to a clinical setting, where they can have a positive impact on patients. However, this requires much more than just statistical significance. The current value of GWAS lies with the biological networks that the variant exposes as an underpinning of the phenotype. Both Lyssenko and colleagues[1] and Bouatia-Naji and colleagues[3] demonstrated direct expression of the MTNR1B gene in pancreatic beta cells. This finding, independently shown by both groups, reverses previous observations.[7] Preliminary data suggest that the G allele is linked with increased fasting blood glucose levels and may affect the MTNR1B transcription levels in the pancreatic islets of subjects older than 45 years of age.[1]
These 3 papers provide another variant for fasting blood glucose levels, bringing the total attributable variance accounted for by the 4 known markers to around 1.5%.[2] It is of interest that while the association to fasting blood glucose levels is strong, the variant's link to type 2 diabetes is still very weak. This could be because conversion to type 2 diabetes status occurred in few patients in the longitudinal cohort during the study collection and follow-up periods. It could also be a hint that there are possible subtypes of diabetes that are manifested through different pathways, all of which lead to the same end-phenotype. Type 2 diabetes may actually represent several subtypes at the molecular, root level. Regardless, these findings strengthen prior research that our circadian rhythms can have strong influences on our metabolic system. Understanding and targeting malfunctions within our internal clock could prove a viable therapeutic strategy for type 2 diabetes
ليست هناك تعليقات:
إرسال تعليق