Ayo Doumatey - Academia.edu (original) (raw)

Scientific Articles by Ayo Doumatey

Obese individuals without expected metabolic co-morbidities are referred to as metabolically heal... more Obese individuals without expected metabolic co-morbidities are referred to as metabolically healthy obese (MHO). The molecular mechanisms underlying this phenotype remain elusive. MicroRNAs may be involved in the MHO phenotype. To test this hypothesis, we screened 179 serum miRNAs in 20 African-American women (10 MHOs and 10 metabolically abnormal obese individuals-MAO). We identified 8 differentially expressed miRNAs (DEMs) with validation in an independent sample of 64 MHO and 34 MAO. Of the eight DEMs in the screening phase (p ≤ 0.05), miR-374a-5p remained significant (p = 0.04) with directional consistency in the validation sample. Ingenuity Pathway analysis revealed that miR-374a-5p putatively targeted 37 mRNAs (e.g. chemokines and transcription factors) which are members of canonical pathways involved in inflammation (IL-17A signaling) and lipid metabolism. Analysis restricted to adipocytes, the main source of circulating miRNAs in obesity, identified 3 mRNAs (CCL2, STEAP2, EN1) as the main target of miR-374a-5p. Evaluation of the 3 mRNAs in an independent sample showed that CCL2 was significantly downregulated (p = 0.0005). In summary, MiR-374a-5p is upregulated in MHO compared to MAO individuals and appears to show association with downregulation of pro-inflammatory markers that are linked to insulin resistance. Given the correlative nature of our findings, functional studies are needed. Obesity is characterised by excessive fat accumulation that is mostly the result of excess energy intake relative to levels of physical activity 1. However, the development of obesity is also influenced by lifestyle-independent factors like genetic background and hormone dysfunction 1. Obesity has been shown to alter metabolic function. At the onset of obesity, adipose tissue expands through either hypertrophic expansion, which occurs when the volume of pre-existing fat cells is increased, or hyperplasia, which occurs when new small fat cells are generated 2. Hypertrophic adipose tissue has been shown to be highly correlated with metabolic complications compared to hyperplastic adipose tissue 3. Also, excessive abdominal fat impacts metabolic health more adversely than subcu-taneous fat 4. Furthermore, obesity is a chronic low-grade inflammatory disease state that contributes to insulin resistance (IR), type 2 diabetes (T2D), and cardiovascular disease 5. While most obese persons display metabolic risk factors such as high blood pressure, hyperglycaemia and dyslipidaemia and are thus classified as metabolically abnormal individuals with obesity (MAO), a small subset of obese individuals appears to be metabolically healthy. These persons are referred to as metabolically healthy individuals with obesity (MHO) 6. While the concept of MHO is clear, there is a lack of consensus regarding the criteria used for the definition of the MHO phenotype. This has resulted in a wide variation in the reported prevalence , depending on the stringency of the definition used. A recent review by Rey-Lopez et al., reports MHO prevalence ranging from as low as 6% to as high as 75% 7. There is accumulating evidence suggesting that the MHO phenotype may represent a transitional physiologic state in which obese individuals are not protected from metabolic consequences later in life 8,9. However, it has also been shown that a subset of healthy obese individuals remain healthy into their elderly years, as MHO do not have a higher risk of all-cause mortality when compared to their normal weight counterparts 10. Understanding this subgroup of obesity at the molecular and cellular level

Papers by Ayo Doumatey

Genome Medicine, Dec 3, 2023

Scientific Reports, Nov 8, 2022

Adiponectin has been associated with cardiometabolic traits in observational studies across popul... more Adiponectin has been associated with cardiometabolic traits in observational studies across populations, yet it is unclear if these associations are causal. We performed Mendelian randomization (MR) analysis to assess the relationship between adiponectin and cardiometabolic traits in sub-Saharan Africans. We constructed a polygenic risk score (PRS) for adiponectin levels across 3354 unrelated sub-Saharan Africans. The PRS was used as the instrumental variable in two-stage leastsquares MR analysis to assess its association with insulin resistance, HDL, LDL, total cholesterol, triglycerides, blood pressure, Type 2 Diabetes (T2D), and hypertension. The adiponectin PRS was causally related with LDL (β = 0.55, 95%CI 0.07-1.04, P-value = 0.024) but not the other traits. This association was observed in both overweight/obese and normal weight individuals, but only reached statistical significance among overweight/obese individuals (β = 0.55, 95%CI 0.01-1.08, P-value = 0.045). In normal weight individuals, the adiponectin PRS was associated with T2D (OR = 0.13, 95%CI 0.02-0.73, P-value = 0.021), and in men with HDL (β = 1.03, 95%CI 0.14-1.92, P-value = 0.023). The findings of this first MR study in sub-Saharan Africans support a causal relationship of adiponectin with LDL, with T2D in normal weight individuals only, and with HDL in men only. These observations add to the small but growing literature on adiponectin MR studies. Nearly 80% of cardiovascular disease (CVD) related deaths occur in low-and middle-income countries 1. In sub-Saharan Africa, the rates of CVD risk factors, such as type 2 diabetes (T2D) and hypertension, are increasing rapidly. Sub-Saharan Africa has the highest projected increase in T2D of all world regions with a projected increase of 156% by 2045 2. The prevalence of hypertension in Africa has increased from 19.7% in 1990 to 30.8% in 2010 3. A rise in obesity is thought to underly the increases in these cardiometabolic outcomes. Due to epidemiological transition, including reduced levels of physical activity and increased consumption of unhealthy diets, the obesity prevalence in sub-Saharan Africa has increased from 3 to 11% between 1975 and 2014 4. However, a high prevalence of T2D has also been observed among normal weight Africans (Body Mass Index [BMI] < 25 kg/m 2) living in Africa and in Europe 5,6 , for reasons that are not yet fully understood. The biological mechanisms underlying the association of adiposity with cardiometabolic outcomes have been under intense study, especially with regard to the role of inflammation and of biomolecules secreted by adipose tissue, adipokines. Adipose tissue is an active organ involved in the production of a wide range of circulating proteins including adiponectin 7. Adiponectin is an anti-inflammatory adipokine, which regulates glucose levels and fatty acid metabolism 8. Higher circulating adiponectin levels have been negatively associated with obesity 9 , insulin resistance 9 , triglycerides 9 , and T2D risk 10 , and positively with high-density lipoprotein (HDL) cholesterol, and total cholesterol 9 in epidemiological studies across populations, including sub-Saharan African populations 9. However, these epidemiological studies are not able to determine whether changes in adiponectin levels are a cause or consequence of these cardiometabolic disorders. If the observed associations are causal and higher levels of adiponectin are indeed protective of cardiometabolic outcomes, there may be important therapeutic implications. Mendelian randomization (MR) is a means of investigating causality between two traits. In MR analyses, the exposure is defined by genetic variants that contribute to the distribution of the putative causal agent. The

Circulation, Sep 2, 2020

Background - Resistin, a protein linked with inflammation and cardiometabolic diseases, is one of... more Background - Resistin, a protein linked with inflammation and cardiometabolic diseases, is one of few proteins for which GWAS consistently report variants within and near the coding gene ( RETN ). Here, we took advantage of the reduced linkage disequilibrium in African populations to infer genetic causality for circulating resistin levels by performing GWAS, whole-exome analysis, fine-mapping, Mendelian randomization and transcriptomic data analyses. Methods - GWAS and fine-mapping analyses for resistin were performed in 5621 African ancestry individuals, including 3754 continental Africans (AF) and 1867 African Americans (AA). Causal variants identified were subsequently used as an instrumental variable in Mendelian randomization analyses for homeostatic modelling (HOMA) derived insulin resistance index, BMI and type 2 diabetes. Results - The lead variant (rs3219175, in the promoter region of RETN ) for the single locus detected was the same for AF ( P -value 5.0×10 -111 ) and for AA (9.5×10 -38 ), respectively explaining 12.1% and 8.5% of variance in circulating resistin. Fine-mapping analyses and functional annotation revealed this variant as likely causal affecting circulating resistin levels as a cis -eQTL increasing RETN expression. Additional variants regulating resistin levels were upstream of RETN with genes PCP2 , STXBP2 and XAB2 showing the strongest association using integrative analysis of GWAS with transcriptomic data. Mendelian randomization analyses did not provide evidence for resistin increasing insulin resistance, BMI or type 2 diabetes risk in African-ancestry populations. Conclusions - Taking advantage of the fine-mapping resolution power of African genomes, we identified a single variant (rs3219175) as the likely causal variant responsible for most of the variability in circulating resistin levels. In contrast to findings in some other ancestry populations, we showed that resistin does not seem to increase insulin resistance and related cardiometabolic traits in African-ancestry populations.

Blood, Dec 6, 2014

Hemolysis as a suggested cause of hyperuricemia is based upon the fact that red blood cells conta... more Hemolysis as a suggested cause of hyperuricemia is based upon the fact that red blood cells contain uric acid. The development of hyperuricemia also generates from an increased synthesis of nucleic acids occurring as part of the erythropoietic response to hemolysis in hemoglobin disorders such as sickle cell anemia, α-thalassemia, and β-thalassemia. In addition, multiple genome-wide association studies (GWAS) have reported significant association between uric acid levels and specific genomic loci. However, the mechanism of hyperuricemia still remains controversial and it is also unknown whether African Americans have higher prevalence of hyperuricemia due to genetic vs environmental risk factors. Here, we used joint admixture mapping and association testing to identify genetic variants associated with serum uric acid levels in African American. Interestingly, we detected 6 SNPs (rs2855125, rs2855126, rs11036415, rs11036496, rs4320977, and rs4348933) in an intergenic region of the β–globin cluster on chromosome 11 are associated with high levels of serum uric acid in populations of African ancestry. Next, we explored the potential regulatory role of intergenic SNPs associated with hyperuricemia using luciferase reporter gene activity assays and electrophoretic mobility shift assays (EMSA). Each SNP-containing DNA fragment was amplified by PCR using human genomic DNA and inserted into a firefly luciferase reporter vector, pGL3-basic vector. 293T or K562 cells were co-transfected with these constructs and a Renilla luciferase vector to control for transfection efficiency. Expression of firefly luciferase driven by each SNP-containing DNA fragment was measured by a dual luciferase reporter assay and normalized by Renilla luciferase expression. SNPs rs2855126, rs11036496 and rs4348933 on chromosome 11 had significantly greater expression levels of firefly luciferase than pGL3-basic-transfected cells in both 293T and K562 cells. Of these, the SNP rs2855126 ancestral C allele (associated with higher serum uric acid levels) showed significantly higher luciferase activity than the derived G allele. Functionally, the luciferase activities from these constructs were determined to be very similar in both cell lines used. Alleles altering expression were further assessed for binding of nuclear extracted proteins by EMSA. We found specific gel shift bands for SNPs rs2855126, rs11036496, and rs4348933, suggesting these SNPs are situated in the binding site of potential transcription factors. These data provide new insights into the potential contribution of imbalanced β-globin gene expression to hyperuricemia. Disclosures No relevant conflicts of interest to declare.

Circulation, Feb 28, 2023

High blood pressure is a heritable risk factor for cardiovascular disease, stroke and other chron... more High blood pressure is a heritable risk factor for cardiovascular disease, stroke and other chronic conditions. There is an increasing prevalence of high blood pressure in sub-Saharan Africa. We aimed to identify genetic variants underlying blood pressure traits in Sub-Saharan Africa. We conducted a GWAS, meta-analyses and gene-set analyses of blood pressure traits (systolic, diastolic, pulse pressure), in African populations. We used the African Collaborative Center for Microbiome and Genomics Research data (ACCME, n=~15,000) for GWAS discovery and the Africa America Diabetes Mellitus study data (AADM, n=~5,200), for GWAS replication. Multiple SNPs were identified reaching genome-wide significance. The lead SNP from the elevated blood pressure group was on chromosome 16 (rs148403740, minor allelic frequency (MAF)=0.0176, p= 9.28x10 -9 , odds ratio(OR)= 2.58) nearest to the MGRN1 gene. The lead SNP from the hypertension stage 1 group was on chromosome 7 (rs4236415, MAF= 0.4912, p= 1.33x10 -8 , OR= 1.2). The index SNP within the pulse pressure group was on chromosome 10 (rs61836106, MAF= 0.1828, p= 7.4x10 -9 , OR= 5.7). The findings provide promising candidate SNPs for developing risk stratification models in African populations.

Genome Medicine, Oct 7, 2021

EBioMedicine, May 1, 2023

Diabetologia, May 2, 2019

Clinical Epigenetics, Jul 14, 2022

Background: African Americans have a high risk for type 2 diabetes (T2D) and insulin resistance. ... more Background: African Americans have a high risk for type 2 diabetes (T2D) and insulin resistance. Studies among other population groups have identified DNA methylation loci associated with insulin resistance, but data in African Americans are lacking. Using DNA methylation profiles of blood samples obtained from the Illumina Infinium ® HumanMethylation450 BeadChip, we performed an epigenome-wide association study to identify DNA methylation loci associated with insulin resistance among 136 non-diabetic, unrelated African American men (mean age 41.6 years) from the Howard University Family Study. Results: We identified three differentially methylated positions (DMPs) for homeostatic model assessment of insulin resistance (HOMA-IR) at 5% FDR. One DMP (cg14013695, HOXA5) is a known locus among Mexican Americans, while the other two DMPs are novel-cg00456326 (OSR1; beta = 0.027) and cg20259981 (ST18; beta = 0.010). Although the cg00456326 DMP is novel, the OSR1 gene has previously been found associated with both insulin resistance and T2D in Europeans. The genes HOXA5 and ST18 have been implicated in biological processes relevant to insulin resistance. Differential methylation at the significant HOXA5 and OSR1 DMPs is associated with differences in gene expression in the iMETHYL database. Analysis of differentially methylated regions (DMRs) did not identify any epigenome-wide DMRs for HOMA-IR. We tested transferability of HOMA-IR associated DMPs from five previous EWAS in Mexican Americans, Indian Asians, Europeans, and European ancestry Americans. Out of the 730 previously reported HOMA-IR DMPs, 47 (6.4%) were associated with HOMA-IR in this cohort of African Americans. Conclusions: The findings from our study suggest substantial differences in DNA methylation patterns associated with insulin resistance across populations. Two of the DMPs we identified in African Americans have not been reported in other populations, and we found low transferability of HOMA-IR DMPs reported in other populations in African Americans. More work in African-ancestry populations is needed to confirm our findings as well as functional analyses to understand how such DNA methylation alterations contribute to T2D pathology.

Background/Purpose: Anti-nuclear autoantibodies are a hallmark of scleroderma with anti-centromer... more Background/Purpose: Anti-nuclear autoantibodies are a hallmark of scleroderma with anti-centromere antibody (ACA) recognizing centromeric antigens. ACA-positive patients have longstanding Raynaud\u27s, limited cutaneous disease and increased risk for pulmonary arterial hypertension. We investigated the role of HLA classical genes and alleles on risk for ACA-positive scleroderma in a large collection of patients with scleroderma and genetically matched controls. Methods: SNP genotypes of 723 scleroderma cases and 5,561 controls, all of European ancestry, were obtained from dbGaP. Classical HLA types were imputed with SNP2HLA using the Type I Diabetes Genetic Consortium reference of 5,225 individuals. Association of HLA classical alleles was tested by a dominant model regression analysis coding the HLA types as numeric values (1 for present, 0 for absent). Regression tests were corrected for genetic dissimilarity by including the top 5 principal components as covariates. Results: Of the 723 scleroderma cases, 238 (32.9%) were positive for ACA. The most significantly ACA-positive scleroderma-associated HLA allele was HLA-DRB1∗07:01, which was disease protective (P-value=1.8x10-18, odds ratio (OR)=0.11 (95%CI=0.05 to 0.22)). This allele was found in only 3.4% of the ACA-positive cases versus 23.6% of controls and 20.8% of the ACA-negative cases. Regression analysis conditioning on the disease-associated alleles identified HLADQB1∗ 05:01 as the most significantly associated disease risk allele (P-value= 3.3x10-08, OR=2.18 (1.66-2.86)) with additional independent risk effects of HLA-DQA1∗04:01 and HLA-DQA1∗03:01). A two-locus analysis of the DQB1∗05:01 disease-risk and the DRB1∗07:01 disease-protective alleles suggested the risk effect of DQB1∗05:01 is overridden by the protective effect of DRB1∗07:01(Figure 1). The odds ratio for ACA-positive disease in individuals carrying both alleles was 0.14, similar to that in individuals carrying DRB1∗07:01 without DQB1∗05:01 (0.15). No effect of DRB1∗07:01 or DQB1∗05:01 was found in the anti-topoisomerase I autoantibody subset (P=0.98 and P=0.054, respectively). For validation, 62 African American ACA-positive cases and 946 matched controls from the Genomic Research in African American Scleroderma Patients (GRASP) Collection, were similarly analyzed. DQB1∗05:01 was associated with disease risk (P=3.6x10-4, OR=2.68 (1.57-4.58)) and DRB1∗07:01 was protective (P=8.6x10-3, OR=0.33 (0.13-0.85)) in the African American sample. Conclusion: HLA-DQB1∗05:01 is associated with risk and HLA-DRB1∗07:01 is associated with protection for ACA-positive scleroderma. The mechanisms responsible for these effects could be exploited to prevent or treat scleroderma. (Figure Presented)

Diabetes Research and Clinical Practice, 2021

AIMS The glycated hemoglobin (HbA1c) test can be unreliable in the presence of hemoglobinopathies... more AIMS The glycated hemoglobin (HbA1c) test can be unreliable in the presence of hemoglobinopathies. The co-existence of type 2 diabetes (T2D) with sickle cell anemia calls for alternative tests. Therefore, we established a reference interval for serum fructosamine and evaluated its utility as a potential glycemic biomarker that is not affected by abnormal hemoglobin. METHODS The accuracies of serum fructosamine in monitoring and diagnosing T2D were evaluated using the Area under the Receiver Operating Characteristics and other measures in 618 Nigerians with or without sickle cell trait. The estimated diagnostic cut-off for serum fructosamine was then validated in an independent multi-ethnic cohort of 634 West Africans. RESULTS Serum fructosamine was similar between individuals with or without sickle cell trait (median: 287 vs 275 umol/L, p=0·11, respectively) despite statistically different HbA1c. Fructosamine was highly correlated with both HbA1c and fasting glucose independently of sickle cell trait. The areas under the curve (AUC) of serum fructosamine in identifying individuals with uncontrolled glycemia and individuals with T2D were similar and independent of sickle cell trait: 0·92 (95% confidence interval [95% CI ], 0·88-0·95 and 0.92 (95% CI, (0.89-0.95) respectively. CONCLUSIONS Serum fructosamine is a good alternative to HbA1c for monitoring and diagnosing T2D in the presence of sickle cell trait.

Primers used for cloning of each SNP fragment. (DOCX 20 kb)

Genome Medicine, 2021

Background A complex set of perturbations occur in cytokines and hormones in the etiopathogenesis... more Background A complex set of perturbations occur in cytokines and hormones in the etiopathogenesis of obesity and related cardiometabolic conditions such as type 2 diabetes (T2D). Evidence for the genetic regulation of these cytokines and hormones is limited, particularly in African-ancestry populations. In order to improve our understanding of the biology of cardiometabolic traits, we investigated the genetic architecture of a large panel of obesity- related cytokines and hormones among Africans with replication analyses in African Americans. Methods We performed genome-wide association studies (GWAS) in 4432 continental Africans, enrolled from Ghana, Kenya, and Nigeria as part of the Africa America Diabetes Mellitus (AADM) study, for 13 obesity-related cytokines and hormones, including adipsin, glucose-dependent insulinotropic peptide (GIP), glucagon-like peptide-1 (GLP-1), interleukin-1 receptor antagonist (IL1-RA), interleukin-6 (IL-6), interleukin-10 (IL-10), leptin, plasminogen...

SSRN Electronic Journal, 2020

Background: The glycosylated hemoglobin (HbA1c) test can be unreliable in the presence of hemoglo... more Background: The glycosylated hemoglobin (HbA1c) test can be unreliable in the presence of hemoglobinopathies and remains cost-prohibited and impractical in many developing countries. The increasing burden of type 2 diabetes (T2D) in sub-Saharan Africa (SSA) where sickle cell anemia, the most common inherited hemoglobinopathy, is prevalent calls for alternative tests for the management of T2D. In the present study, we evaluated the utility of serum fructosamine (FRA) as a biomarker of glycemic control that is not affected by abnormal hemoglobin or red blood cell (RBC) turnover. Methods: The accuracy of FRA in monitoring and diagnosing T2D was evaluated using the Area under the Receiver Operating Characteristics (AUROC) and other measures in a training dataset of 618 Nigerians [155 (25·1%) carrying sickle cell trait (HbAS) and 463 (74·9%) without sickle cell trait (HbAA)]. The FRA’s diagnostic cut-off was validated in an independent cohort of 634 West Africans. Findings: FRA was similar between HbAS and HbAA participants (Median = 287 vs 275 umol/L, p=0·11) despite statistically different HbA1c median among the two groups. FRA was highly correlated with both HbA1c (r HbAA = 0·80, r HbAS =0·82, p<0·001) and fasting glucose (FG) (r HbAA =0·72, r HbAS = 0·83, p<0·001) independent of sickle cell trait (SCT). With optimal thresholds of 352·5 umol/L and 309·5 umol/L, FRA had 81·5% [CI 95%:0·76-0·87] and 82·4%, [CI 95% 0·77-0·88] sensitivities and 97·4% [CI 95%:0·95-1·00] and 92·4% [95% CI 0·90-0·95] specificities for identifying individuals with controlled glycemic index and individuals with T2D respectively and independently of SCT. The diagnosis cut-off was validated in a multi-ethnic population with a good positive predictive value (94·6%). Interpretation: We showed that FRA is an effective alternative to HbA1c for monitoring and diagnosing T2D in populations where SCT is prevalent. Funding: National Institutes of Health. Conflict of Interest: The authors have no conflict of interest to disclose. Ethical Approval: The study protocol was approved by the institutional ethicsreview board (IRB) of the National Institutes of Health/National Human Genome Research Institute (protocol HG-09-N070) and the IRBs of each participating institution, including University of Lagos, Lagos, Nigeria; College of Medicine, University of Ibadan, Ibadan, Nigeria; University of Nigeria Teaching Hospital, Enugu, Nigeria; University of Science and Technology, Kumasi, Ghana; University of Ghana Medical School, Accra, Ghana. Written informed consent was obtained from each participant prior to enrollment.

Frontiers in Cellular and Infection Microbiology, 2020

Current Diabetes Reports, 2019

Journal of Allergy and Clinical Immunology, 2019

NPJ genomic medicine, 2018

Several clinical guidelines have been proposed to distinguish metabolically healthy obesity (MHO)... more Several clinical guidelines have been proposed to distinguish metabolically healthy obesity (MHO) from other subgroups of obesity but the molecular mechanisms by which MHO individuals remain metabolically healthy despite having a high fat mass are yet to be elucidated. We conducted the first whole blood transcriptomic study designed to identify specific sets of genes that might shed novel insights into the molecular mechanisms that protect or delay the occurrence of obesity-related co-morbidities in MHO. The study included 29 African-American obese individuals, 8 MHO and 21 metabolically abnormal obese (MAO). Unbiased transcriptome-wide network analysis was carried out to identify molecular modules of co-expressed genes that are collectively associated with MHO. Network analysis identified a group of 23 co-expressed genes, including ribosomal protein genes (RPs), which were significantly downregulated in MHO subjects. The three pathways enriched in the group of co-expressed genes ar...