Diabetic kidney disease (DKD), also known as diabetic nephropathy, is the leading cause of renal impairment and end-stage renal disease. Patients with diabetes are at risk for DKD because of poor control of their blood glucose, as well as nonmodifiable risk factors including age, ethnicity, and genetics. This genome-wide association study (GWAS) was conducted for the first time in the Emirati population to investigate possible genetic factors associated with the development and progression of DKD. We included data on 7,921,925 single nucleotide polymorphism (SNPs) in 258 cases of type 2 diabetes mellitus (T2DM) who developed DKD and 938 control subjects with T2DM who did not develop DKD. GWAS suggestive results (P < 1 × 10–5) were further replicated using summary statistics from three cohorts with T2DM-induced DKD (Bio Bank Japan data, UK Biobank, and FinnGen Project data) and T1DM-induced DKD (UK-ROI cohort data from Belfast, UK). When conducting a multiple linear regression model for gene-set analyses, the CNR2 gene demonstrated genome-wide significance at 1.46 × 10–6. SNPs in CNR2 gene, encodes cannabinoid receptor 2 or CB2, were replicated in Japanese samples with the leading SNP rs2501391 showing a Pcombined = 9.3 × 10–7, and odds ratio = 0.67 in association with DKD associated with T2DM, but not with T1DM, without any significant association with T2DM itself. The allele frequencies of our cohort and those of the replication cohorts were in most cases markedly different. In addition, we replicated the association between rs1564939 in the GLRA3 gene and DKD in T2DM (P = 0.016, odds ratio = 0.54 per allele C). Our findings suggest evidence that cannabinoid signalling may be involved in the development of DKD through CB2, which is expressed in different kidney regions and known to be involved in insulin resistance, inflammation, and the development of kidney fibrosis.
Type 2 diabetes mellitus (T2DM) is a chronic metabolic disorder associated with hyperglycemia due to a combination of insulin resistance and β-cell impairment, influenced by genetic and environmental factors1. Among the different complications that arise with T2DM, progressive loss of renal function represents one of the most serious complications, characterized as diabetic kidney disease (DKD), with an estimated prevalence of 30–50%2. DKD is a multifactorial, heterogeneous condition that is considered as the leading cause of progressive renal impairment and end-stage renal disease3. This condition results in pathologic changes associated with glomerular structure and function, albuminuria, excessive extracellular matrix deposition resulting in mesangial matrix expansion, inflammation, and fibrosis4,5. In addition to poor glycemic control, which can be controlled by pharmaceutical and clinical interventions, there also are several nonmodifiable risk factors associated with DKD, such as age, sex, ethnicity, and genetics6. Familial aggregation of DKD has been demonstrated in different ethnic groups, demonstrating the importance of understanding how environmental and genetic factors interact resulting in distinct disease predispositions and mechanisms in different populations7,8,9,10,11,12,13.
The Middle East region has some of the highest prevalence of T2DM worldwide, with a disproportionally high prevalence of DKD, ranging between 10.8 and 61.2%, rendering this particular cohort interesting for genetic studies of this trait14,15. Genetic and epigenetic studies of DKD, using candidate gene approaches and genome-wide studies, have been undertaken to identify genetic loci conferring susceptibility or resistance to DKD. Currently, more than 150 genes have been associated with DKD using candidate gene approaches, while about 42 genetic loci have been associated with DKD susceptibility or for DKD indicators, such as albuminuria, through genome-wide association studies (GWAS)12,16,17,18,19,20,21,22,23. There is, however, a few robustly replicated loci that have emerged, with a knowledge gap on the functional role of the newly identified genes in DKD, which would require further replication studies and detailed analysis in experimental models. Furthermore, there are a limited number of GWAS studies that have been conducted in the Middle Eastern populations related to T2DM-associated DKD.
Though DKD demonstrates familial aggregation and clustering, and SNP heritability, the genetic factors influencing DKD remain largely unknown24,25,26,27. To further explore the genetic association of DKD in the Middle Eastern population with T2DM, we performed a GWAS to identify possible genetic risk factors and uncover mechanisms that contribute to the development and progression of DKD among an Emirati population. Additionally, we aimed to replicate the strongest signals from the current analysis in different populations to confirm this association, as well as in the GWAS Catalog (European ancestry cohorts). We also aimed to replicate previously identified genome-wide significant associations in the Middle Eastern population.
Recruitment and ethical consideration
This cross-sectional study recruited 1223 unrelated Emirati participants from Imperial College of London Diabetic Centre (ICLDC), Abu Dhabi, United Arab Emirates, between the period of April 2018 to January 2020. The participants were divided into two clinical groups: T2DM with DKD (n = 261) and T2DM without DKD (n = 962). The diagnosis of T2DM was made by a qualified physician based on WHO guidelines and criteria, with an a HbA1c ≥ 6.5 and were receiving treatment for their condition, whereas DKD was diagnosed as outlined previously28. In summary, DKD was defined as either decreased estimated glomerular filtration rate (eGFR < 60 mL/min/1.73 m2) with or without renal damage over a period of at least 3 months, or based on an albumin-to-creatinine ratio ≥ 30 mg/g, or proteinuria > 500 mg over a 24-h period in the setting of T2DM and/or abnormalities, as assessed by imaging or histology29. The inclusion criteria were: (1) UAE nationals, (2) clinically diagnosed with T2DM with or without DKD, (3) above 18 years old and (4) able to provide informed consent and complete the survey. Participants were excluded if they were diagnosed with a critical condition (e.g., cancer) or pregnancy.
An informed consent was obtained from all participants. The study was conducted in accordance with the Declaration of Helsinki and has been approved by the Imperial College London Diabetes Centre (ICLDC) ethical committee (IREC 036).
All participants completed a validated screening questionnaire that includes details on the demographic characterization, risk factors, symptomatology, and family history of diseases. Medical history, medication and management methods were provided by ICLDC. Anthropometric parameters and blood pressure were collected and assessed during the hospital visit by trained nurses using calibrated stadiometer, weight scale and sphygmomanometer. Biochemical data such as chemistry and hematology parameters, to assess the presence of diabetic complications, were obtained from ICLDC.
DNA extraction, genotyping, imputation, and quality control
Total of 4 ml of blood sample was collected from each participant into EDTA anti-coagulant tubes. Genomic DNA was extracted using the automated MagPurix 24 system (LTF Labortechnik, GmbH&Co. KG, Hattnaeur, Wasserburg, Germany) according to the manufacturer’s instructions. DNA was quantified using DS-11 Series of Spectrophotometer/Fluorometer (DeNovix Inc. Wilmington, USA).
Genotyping was done using Infinium Global Screening Array (Illumina, Inc., San Diego, CA, USA), which contained 654,027 genetic markers and developed by Avera Institute for Human Genetics (South Dakota, USA). The Infinium Global Screening Array-24 v3.0 BeadChip would be best suited for this ethnically diverse cohort because it features a broad spectrum of diverse exonic content, including both cross population and population specific markers (EUR: 52,980 markers; EAS: 31,375 markers; AMR: 45,977 markers; AFR: 43,122 markers; SAS: 40,298 markers). Genotypes were exported, in Genome Reference Consortium human build 37 (GRCHb37) and Illumina ‘source’ strand orientation. The raw data was processed using Genome Studio (Version 2, Illumina, Inc.) genotyping module to generate the data for further analysis.
Stringent quality control measures were performed for samples and SNPs to ensure subsequent robust association tests. Samples that failed to reach a 98.0% call rate were reanalyzed. Samples were excluded if they had discordant sex information (n = 0) or outlying heterozygosity rate (n = 0). We also applied identity by state (IBS) and identity by descent (IBD) to exclude duplicated individuals or those who are first degree relatives (pairwise identity-by-state (PI_HAT) > 0.5, and accordingly, we excluded 27 individuals: three cases and 24 controls. Quality control measures for SNP callings were missing per individual (> 10%), missingness per marker (> 5%), minor allele frequency (MAF > 1%), Hardy–Weinberg Equilibrium (HWE > 0.00001). After quality control, a total of 443,465 variants passed filters.
Genotypes were prephased and imputed with untyped markers (∼39 M) using the Phase 3 multi-ethnic 1000 Genomes Projects panel, as the reference based on the human genome assembly hg19 (https://mathgen.stats.ox.ac.uk/impute/1000GP_Phase3.html), and was carried out by using BEAGLE, using standard protocols and recommended settings. Imputation was performed on both autosomal and sex chromosomes. Post imputation quality control measures (genotyping rate < 95%, call rate < 98%, MAF > 1%, HWE > 0.00001) were conducted, and a total of 7,921,925 variants passed filters.
Continuous variables were presented as means and standard deviations, and categorical data were calculated as proportions and percentages. When needed, Kolmogorov–Smirnov test was used to test the normality of the data. student’s t-test (unpaired t-test) was used to test the significant differences of the pre-laboratory measurements between cases and controls. The significance level for pre-laboratory measurements was set at p < 0.05. These analyses were done using SPSS 22 (SPSS Inc., IBM Company, Chicago, Illinois, USA).
In the GWAS and replication analyses, we used a multivariate logistic regression model, assuming additive allelic effects for genotyped and imputed SNPs, to assess the statistical significance of each SNP with adjustment for age, gender, log BMI, and the first ten principal components as covariates. The significance levels for the GWAS were 6.3 × 10−9 (0.05/7,921,925) in the analysis following the Bonferroni correction for multiple testing. Odds ratios and 95% confidence intervals were calculated using the minor allele genotype as a reference. Principal component analysis (PCA) was performed using the case–control samples in this GWAS and a reference panel of populations in the 1000 Genome project (Fig. 1). Meta-analysis studies were performed using PLINK 1.09. Haploview v4·1 was used to conduct subsequent association analysis. Functional Mapping and Annotation of genetic associations (FUMA) was used to perform visualization Manhattan plots30. To assess the association of a given SNP with T2DM and DKD, the allelic frequencies were compared by means of a χ2 statistic, which yielded an individual p-value for each combination of SNP and allelic model. A quantile–quantile (Q-Q) plot analysis was carried out to check whether the distribution of the inflation p-values deviated from the expected distribution under the null hypothesis of no genetic association and investigate if the overall significance of the genome-wide associations is due to potential impact of population stratification. Systematic bias and the impact of population stratification was evaluated by calculating the genomic control inflation factor [λ GC] and noted for each analysis. A Manhattan plot was generated with −log10 p-values.
The gene-set analysis was conducted using the FUMA software, an online platform for functional mapping of genetic variants from GWAS summary30. The gene-set analyses was produced in Multi-marker Analysis of GenoMic Annotation (MAGMA), in which it examines sets of biologically related genes that are strongly associated to the disease of interest31. The MAGMA tool uses multiple linear regression models to assess whether genes in each gene set are associated to a polygenic trait, after correcting for linkage disequilibrium between variants and gene size.
Based on the top identified SNP in this cohort, further replication data was obtained from three datasets: DKD summary results of Biobank Japan32, the FinnGen Project33 (cohort name: finngen_R5_DM_NEPHROPATHY_EXMORE), and the UK Biobank34 (cohort name: E1122). The GWAS study for the Japanese population contains 220 cases with DKD vs. 132,764 controls. The Finnish samples included 3283 cases with DKD vs. 181,704 controls, and the UK Biobank contains 58 cases vs. 8444 controls. Replication analysis in type 1 diabetes-induced DKD was done in the UK-ROI cohort (Belfast, UK), white individuals with T1D, diagnosed before 31 years of age, whose parents and grandparents were born in the UK and Ireland19. The UK-ROI cohort case group comprised individuals with persistent proteinuria (> 500 mg/24 h) developing more than 10 years after the diagnosis of diabetes, hypertension (> 135/85 mmHg and/or treatment with antihypertensive medication), and retinopathy; ESRD (27.2%) was defined as individuals requiring renal replacement therapy or having received a kidney transplant. Absence of DKD was defined as persistent normal urine albumin excretion rate (AER; 2 out of 3 urine albumin to creatinine ratio [ACR] measurements, 20 mg of albumin/mg of creatinine) despite duration of T1D for at least 15 years, while not taking an antihypertensive medication, and having no history of treatment with ACE inhibitors. In the analysis in this study, there were 823 DKD/ESRD cases and 903 controls. DNA from individuals in the UK-ROI collection were genotyped using the Omni1-Quad array (Illumina, San Diego, CA, USA).
A replication analysis was conducted in this cohort on previously identified variants (SNPs and candidate genes) associated with DKD, totaling 49 SNPs in 42 gene loci. The direction of effect and strength of association was measured in this cohort.
Study cohort characteristics
All study participants had T2DM, and both cases and controls (with or without DKD) demonstrated poor glycemic control indices (mean HbA1c > 7% for both groups). Moreover, those with DKD were mostly men (p < 0.001), older in age (p < 0.001), and had higher rates of associated comorbidities, such as hypertension (p < 0.001) and dyslipidemia (p < 0.001) (Table 1). The renal function (eGFR, p < 0.001; creatinine, p < 0.001), urea, p < 0.001; vitamin D, p = 0.002) and lipid profile (cholesterol total, p < 0.001; HDL, p = 0.002; LDL, p < 0.001) were significantly lower in participants with DKD than non-DKD patients. BMI demonstrated no significant association (p = 0.392) between cases and controls.
GWAS of DKD
To identify genetic variants that contribute to the susceptibility of DKD in the UAE population, we conducted a genome-wide association study (GWAS) using 258 cases with T2DM who developed DKD, and 938 control subjects who had T2DM but did not develop DKD. The samples were genotyped using the Illumina Infinium Global Screening Array (“Methods”). We selected 7,921,925 SNPs for further association analysis with DKD after applying stringent quality control (QC) filtering and imputation analyses (“Methods”). As shown in Supplementary Fig. 1, all cases and controls were representative with the structure of the UAE population as determined by PCA35,36. The full GWAS results of all suggestive loci are presented in the Supplementary Table 1. A Quantile–quantile (Q-Q) plot with the genomic control inflation factor [λ GC] of 1 indicated a low possibility of population stratification (Fig. 1A).
Based on the association analysis, twelve loci showed suggestive associations with DKD (p = 10–5, Fig. 1C and Supplementary Table 1). The leading SNPs of PPP1R9A gene, located in locus 7q21.3, was the most significant SNP at rs78595611 (p = 9.73 × 10–7) and SNP rs76361654 (p = 7.43 × 10–6). PPP1R9A gene is involved in neurite formation, is associated to coronary artery disease, and is a prognostic marker in renal cancer37,38. Two leading SNPs of the cytochrome P450 enzymes (CYP2B7P, rs4001941, p = 1.12 × 10–6; CYP4F24P, rs117074522, p = 2.27 × 10–6) that are involved in drug pharmacokinetics and response have also been associated with DKD in T2DM patients39. Gene OR10H2 located in locus 19p13.12 and gene STX18-AS1 in locus 4p16.2 are associated with acute myeloid leukemia and congenital heart disease, and the SNPs that have been associated to DKD are rs11664515 (p = 2.27 × 10–6) and rs16836018 (p = 2.27 × 10–6), respectively40,41,42. The RB1 gene located at 13q14.2 has been associated with diabetic cardiomyopathy, chronic kidney disease, and proteinuria, all phenotypes that are associated with kidney function and DKD. The leading SNP of RB1 gene is (p = 5.49 × 10–6)43,44,45. The expression of gene MOXD1 located in locus 6q23.2 for SNP rs9493286 (p = 7.25 × 10–6) has been identified to be associated with diabetic nephropathy and obesity-related traits46,47,48.Gene MACROD2 located in locus 20p12.1 have been associated with chronic kidney disease, hypertension, glomerular filtration rate, with the leading SNP rs11696648 demonstrating an association (p = 9.05 × 10–6)41,49,50,51. The CNR2 gene in locus 1p36.1 (SNP rs542405361, p = 9.43 × 10–6) has been associated with diabetic nephropathy, kidney fibrosis and diabetes-induced cardiac dysfunction, all phenotypes associated with DKD52,53,54. Additionally, CNR2 have been identified to alleviate inflammation, oxidative stress and fibrosis, suggesting an important role in kidney function and cardiovascular complications of diabetes55.
A gene-set analysis was conducted to identify genes that are associated with DKD, after implementing multiple regression model and multi-marker association. A Quantile–quantile (Q-Q) plot indicated low possibility of population stratification (Fig. 1B). When conducting a multiple linear regression model on the 18,124 protein coding genes, after correcting for linkage disequilibrium between the SNPs and gene size, the CNR2 gene demonstrated genome-wide significance at 1.46 × 10–6, as demonstrated in Fig. 1D.
Replication analyses—T2DM-induced DKD cohorts
SNPs with suggestive associations were selected for replication in three datasets: the DKD summary results for Biobank Japan32, the UK Biobank34, and FinnGen Project33. Consequently, SNPs in one locus, CNR2, were replicated in the Japanese samples, but not in the Finnish samples, with the leading SNP in rs2501391 in CNR2 gene which we could replicate the Japanese samples with Pcombined = 9.3 × 10–7, and OR = 0.67 (Table 2). The samples from FinnGen project showed very significant differences in allele frequencies in comparison to our samples as well as the Japanese samples (Table 2). The full replication analyses are presented in Supplementary Tables 1 to 4. It is interesting to note that rs2501391 was not the most prominent SNP associated with DKD in the CNR2 gene. However, the more significant SNPs, such as rs542405361, were not found in the Japanese study to replicate them (Supplementary Tables 1 to 3). Due to the different ancestral background of the populations, this variant may be monomorphic in different ethnic groups. However, independent variants associated with DKD may be present in the same region for different ethnicities.
Replication analyses—T1DM-induced DKD cohorts
To test for the existence of a shared pathway for DKD in both type 1 and type 2 diabetes, SNPs with suggestive associations were also selected for replication in replication analysis in type 1 diabetes-induced DKD in the UK-ROI cohort from Belfast, UK. This is a cohort of 823 DKD/ESRD cases and 903 controls of white individuals from Ireland and the UK (see “Methods”). As shown in Table 3, none of the SNPs showed significant associations with GWAS SNPs. The lead SNP in the CNR2 gene, rs2501391, was not available in the UK-ROI cohort; however, it is highly linked to SNP rs2502959 (r2 = 0.98, D′ = 1), suggesting no significant association with T1DM-induced DKD.
Association of CNR2-rs2501391 with T2DM
We further investigated whether the CNR2-rs2501391 polymorphism is associated with T2DM itself. A genotype analysis was performed on this SNP in a cohort of 364 patients with T2DM and 103 healthy controls in an independent cohort. In the association analysis, we found that this SNP had no association with T2DM (Table 4), which suggests that CNR2 may be contributing to DKD development, and not T2DM per se.
Replication analysis of previous studies on DKD
We also replicated previous associations with DKD or kidney function indicators of DKD from previous GWAS that collectively reported 49 SNPs in 42 gene loci12,17,19,20,21,22,23,56. Based on our dataset, we evaluated the associations of 16 SNPs, and we were only able to replicate the association of rs1564939 in the GLRA3 gene with DKD (p = 0.016, OR = 0.54 per allele C, Table 5). In most cases, there were marked differences between the allele frequencies of our cohort and those of the original cohorts at these SNPs.
We identified a gene-based genome-wide significant variant in CNR2 gene, located in 1p36.11, which encodes the cannabinoid receptor 2 (CB2), associated with DKD in an Emirati cohort. After a replication and meta-analysis of GWAS variants across ancestry groups, we identified a similar signal in the Japanese population, but no association in the Finnish or UK-ROI population. We have discovered an additional eleven variants (rs78595611 in PPP1R9A gene; rs4001941 inCYP2B7P; rs117074522 inCYP4F24P; rs275475 in RP11-332J15.1; rs11664515 in OR10H2; rs16836018 inSTX18-AS1; rs17349061 inRB1; rs9493286 inMOXD1; rs76361654 in PPP1R3A; rs141291445 in LINC00693; and rs11696648 inMACROD2) with suggestive association of DKD that will need to be replicated in a larger cohort, across multiple ancestral ethnic groups from Middle East populations.
When investigating non-genetic effects, several different factors contribute to the development of DKD in patients with T2DM, including hypertension and dyslipidemia with elevated cholesterol level, HDL, and LDL. These factors are reflected in this study and are independently associated with progressive DKD through lipid accumulation in tubular epithelial cells of diabetic kidneys57. A decline in renal function, urea, and vitamin D were also evident in those with DKD. Those observations are in line with previous study findings28,58,59,60,61. The presence of hyperglycemia initiates and sustains pathogenic pathways in the kidney that lead to the development and progression of DKD, but several other factors enhance this process even further. This is enhanced by several different factors that will facilitate the progression of the condition. Through optimized diabetes care, implementing pragmatic guidelines for therapeutic treatment, and appropriate targeting of education and support, the prognosis of patients with DKD may be dramatically improved62,63.
The CNR2 gene, encoded by CB2, demonstrated genome-wide significance (p = 1.46 × 10–6) at a gene-set level, and has been implicated in mediating anti-inflammatory effects and immunomodulatory potential64,65. GWAS have associated variations in the CNR2 gene with blood inflammatory cells, especially eosinophils and lymphocytes54,66,67. CB2 receptors may also regulate insulin secretion, insulin resistance, obesity-related inflammation, and metabolic changes such as nonalcoholic fatty liver disease68,69. In mouse and rat models, endocannabinoids generated in renal cells have demonstrated oxidative stress, inflammation, and renal fibrosis70. While there are conflicting reports regarding CB2's role in DKD development71,72, recent studies suggest that CB2 is upregulated by ischemia–reperfusion injury in mice and patients, which is associated to progressive kidney fibrosis73. Additionally, Zhou et al. demonstrated that genetic ablation and selective antagonists inhibiting CB2 were protective against kidney fibrosis, suggesting that CB2 plays an important role in kidney fibrosis pathogenesis74. This suggests that endocannabinoids through CB2 may actually play a dual conflicting role in fibrosis in T2DM. CB2 receptor-mediated responses contribute to both age-related and diet-induced insulin resistance, suggesting that these receptors may be useful therapeutic targets not only for kidney fibrosis, but also for obesity and diabetes as well75. Several factors, such as downstream cellular signaling, activation of b-catenin pathway, type and timing of kidney injury, and genetic variations in CNR2 gene, particularly among different ethnic groups, may influence the receptor function and lead to the various pathological changes observed in the kidney during DKD.
Replication results of previously reported loci that were associated with DKD in patients with T2DM was conducted. SNP rs1564939 in GLRA3 was significant (p = 0.016). This SNP has been associated with albuminuria, an early sign of DKD, in several publications conducted in European populations20. The GLRA3 plays a major role in the central nervous system and has also been linked to ischemic damage. Glycine may also increase effective renal plasma flow and GFR, as well as decrease proximal and distal tubular sodium reabsorption by increasing renal interstitial hydrostatic pressure76. The remaining genetic locus were not replicated in the Middle Eastern cohort. This may be due to ascertainment and selection bias across the studies, or the differences in genetic structure, as indicated by the marked differences in allele frequencies, in which the study was conducted. Emerging evidence suggests that individuals of Middle Eastern descent have a different genome structure due to consanguineous marriage, endogamous unions, tribal structure, and larger family sizes77; however, genome studies from ethnic groups from the region are scarce, representing only 0.08% of genome data in the public domain78. Therefore, a larger cohort study of Middle Eastern descent must replicate these previously reported loci to attribute the lack of association.
There are several limitations in our study. The power analysis indicated that the sample size is barely sufficient to identify genome-wide significant variants. However, regardless of this, after conducting a gene-set analyses on the GWAS data, we were able to identify a genome-wide significant marker associated to DKD. The association of CNR2 gene should be replicated in a larger, multi-ethnic cohort that would be more suited to investigate the genetic association, functional enrichment analysis and functional validation. Due to the lack of information available in this study to adjust for duration of diabetes, future studies should adjust for this confounding variable as it may serve as a genetic risk factor. While presentation, phenotypic profile and clinical course of DKD may be heterogeneous, the recruitment was conducted from one center to avoid ascertainment bias and limit misclassification in the case group. However, the heterogeneous nature of DKD, as well as differences in research design, method, ethnicity and gender compositions of the cohorts in the included studies, may have impacted the replication analysis. Future studies must be conducted in a larger cohort, with subgroup analysis and sensitivity analysis of the different phenotypic stages of DKD. Additionally, due to the lack of comparable studies and datasets available for DKD, we were unable to replicate our GWAS results in independent cohorts from the region; furthermore, we were unable to provide functional studies for the polymorphisms described in this study due to the absence of tissues and biopsies from patients. It is important to note that the replication analysis was performed on samples obtained from biobanks that employ different types of controls than were used in this study. According to the FinnGen or BBJ flow of studies, individuals with a specific phenotype or endpoint are treated as cases, while all other individuals without the phenotype are treated as controls, explaining the large number of controls compared to cases. Our definition of controls was therefore much narrower than in these cases, which may have resulted in a failure of replication of some signals in addition to the differences in genetic structure. Another limitation of this study is that DKD markers such as rs2501391 did not show an association with T2DM risk, or with a previously identified DKD loci. Due to the limited sample size and power, even though this might suggest that genes such as CB2 could contribute to DKD development, it does not exclude the possibility that they may also confer risk for T2DM itself. Lastly, the use of a Euro-centric GWAS array limits the possibility of detected targeted SNPs in the genome that is ethnic-specific79,80,81,82,83. However, to limit this error, the imputation of the genotypes increases the detection of genetic markers that were not directly genotypes and provides further information on the association of these markers, and their respective haplotypes, to the condition.
In summary, we have discovered eleven variants that may be associated with DKD pathogenesis in T2DM patients. However, a highly plausible and significant association was identified in CNR2 gene, suggesting that endocannabinoids signalling through CB2 receptor may be associated with DKD pathogenesis. Further functional studies are warranted to establish the role of CNR2 gene, to improve fundamental knowledge and underlying biological pathway of DKD heterogeneous phenotypes. Performing GWAS studies across different racial and ethnic populations allow the identification of genes and haplotypes associated with different clinical outcomes of DKD. Further studies must be conducted on a large-scale, multi-ethnic cohort to substantiate our current knowledge on DKD pathogenesis and facilitate the development of population-specific therapeutic advances.
Access to raw data supporting the conclusions of this article will be made available by the corresponding author Dr Habiba Alsafar.
DeFronzo, R. A. et al. Type 2 diabetes mellitus. Nat. Rev. Dis. Primers. 1, 1–22 (2015).
Retnakaran, R. et al. Risk factors for renal dysfunction in type 2 diabetes: UK Prospective Diabetes Study 74. Diabetes 55, 1832–1839 (2006).
de Boer, I. H. et al. Temporal trends in the prevalence of diabetic kidney disease in the United States. JAMA 305, 2532–2539 (2011).
Molitch, M. E., DeFronzo, R. A., Franz, M. J. & Keane, W. F. Nephropathy in diabetes. Diabetes Care 27, S79 (2004).
Alicic, R. Z., Rooney, M. T. & Tuttle, K. R. Diabetic kidney disease: Challenges, progress, and possibilities. Clin. J. Am. Soc. Nephrol. 12, 2032–2045 (2017).
Narres, M. et al. The incidence of end-stage renal disease in the diabetic (compared to the non-diabetic) population: A systematic review. PLoS ONE 11, e0147329 (2016).
Freedman, B. I., Bostrom, M., Daeihagh, P. & Bowden, D. W. Genetic factors in diabetic nephropathy. Clin. J. Am. Soc. Nephrol. 2, 1306–1316 (2007).
Murea, M., Ma, L. & Freedman, B. I. Genetic and environmental factors associated with type 2 diabetes and diabetic vascular complications. Rev. Diabet. Stud. 9, 6 (2012).
Thomas, M. C., Groop, P.-H. & Tryggvason, K. Towards understanding the inherited susceptibility for nephropathy in diabetes. Curr. Opin. Nephrol. Hypertens. 21, 195–202 (2012).
McDonough, C. W. et al. A genome-wide association study for diabetic nephropathy genes in African Americans. Kidney Int. 79, 563–572 (2011).
Iyengar, S. K. et al. Genome-wide association and trans-ethnic meta-analysis for advanced diabetic kidney disease: Family Investigation of Nephropathy and Diabetes (FIND). PLoS Genet. 11, e1005352 (2015).
Van Zuydam, N. R. et al. A genome-wide association study of diabetic kidney disease in subjects with type 2 diabetes. Diabetes 67, 1414–1427 (2018).
Vijay, V., Snehalatha, C., Shina, K., Lalitha, S. & Ramachandran, A. Familial aggregation of diabetic kidney disease in type 2 diabetes in south India. Diabetes Res. Clin. Pract. 43, 167–171 (1999).
Federation, I. D. Prevalence of Diabetes, https://www.vision2021.ae/en/national-agenda-2021/list/card/prevalence-of-diabetes (2019).
Aldukhayel, A. Prevalence of diabetic nephropathy among Type 2 diabetic patients in some of the Arab countries. Int. J. Health Sci. 11, 1 (2017).
Gu, H. F. Genetic and epigenetic studies in diabetic kidney disease. Front. Genet. 10, 507 (2019).
Salem, R. M. et al. Genome-wide association study of diabetic kidney disease highlights biology involved in glomerular basement membrane collagen. J. Am. Soc. Nephrol. 30, 2000–2016 (2019).
Sandholm, N. et al. Genome-wide meta-analysis and omics integration identifies novel genes associated with diabetic kidney disease. medRxiv (2021).
Sandholm, N. et al. New susceptibility loci associated with kidney disease in type 1 diabetes. (2012).
Sandholm, N. et al. Confirmation of GLRA3 as a susceptibility locus for albuminuria in Finnish patients with type 1 diabetes. Sci. Rep. 8, 1–8 (2018).
Sandholm, N. et al. Chromosome 2q31.1 associates with ESRD in women with type 1 diabetes. J. Am. Soc. Nephrol. 24, 1537–1543 (2013).
Sandholm, N. et al. Genome-wide association study of urinary albumin excretion rate in patients with type 1 diabetes. Diabetologia 57, 1143–1153 (2014).
Taira, M. et al. A variant within the FTO confers susceptibility to diabetic nephropathy in Japanese patients with type 2 diabetes. PLoS ONE 13, e0208654 (2018).
Harjutsalo, V., Katoh, S., Sarti, C., Tajima, N. & Tuomilehto, J. Population-based assessment of familial clustering of diabetic nephropathy in type 1 diabetes. Diabetes 53, 2449–2454 (2004).
Quinn, M., Angelico, M., Warram, J. & Krolewski, A. Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 39, 940–945 (1996).
Seaquist, E. R., Goetz, F. C., Rich, S. & Barbosa, J. Familial clustering of diabetic kidney disease. N. Engl. J. Med. 320, 1161–1165 (1989).
Sandholm, N. et al. The genetic landscape of renal complications in type 1 diabetes. J. Am. Soc. Nephrol. 28, 557–574 (2017).
Osman, W. M. et al. Clinical and genetic associations of renal function and diabetic kidney disease in the United Arab Emirates: A cross-sectional study. BMJ Open 8, e020759 (2018).
Levey, A. S. et al. Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int. 67, 2089–2100 (2005).
Watanabe, K., Taskesen, E., Van Bochoven, A. & Posthuma, D. Functional mapping and annotation of genetic associations with FUMA. Nat. Commun. 8, 1–11 (2017).
de Leeuw, C. A., Mooij, J. M., Heskes, T. & Posthuma, D. MAGMA: Generalized gene-set analysis of GWAS data. PLoS Comput. Biol. 11, e1004219 (2015).
Sakaue, S. et al. A cross-population atlas of genetic associations for 220 human phenotypes. Nat. Genet. 53, 1415–1424 (2021).
Kurki, M. I. et al. FinnGen: Unique genetic insights from combining isolated population and national health register data. medRxiv (2022).
Sudlow, C. et al. UK biobank: An open access resource for identifying the causes of a wide range of complex diseases of middle and old age. PLoS Med. 12, e1001779 (2015).
Daw Elbait, G., Henschel, A., Tay, G. K. & Al Safar, H. S. A population-specific major allele reference genome from the United Arab Emirates population. Front. Genet. 12, 527 (2021).
Tay, G. K., Henschel, A., Daw Elbait, G. & Al Safar, H. S. Genetic diversity and low stratification of the population of the United Arab Emirates. Front. Genet. 11, 608 (2020).
Li, Q., Yang, W., Lu, M. & Zhang, R. Identification of a 6-gene signature associated with resistance to tyrosine kinase inhibitors: Prognosis for clear cell renal cell carcinoma. Med. Sci. Monit. Int. Med. J. Exp. Clin. Res. 26, e927078-927071 (2020).
Tcheandjieu, C. et al. Large-scale genome-wide association study of coronary artery disease in genetically diverse populations. Nat. Med. 28, 1679–1692 (2022).
Mittal, B., Tulsyan, S., Kumar, S., Mittal, R. D. & Agarwal, G. Cytochrome P450 in cancer susceptibility and treatment. Adv. Clin. Chem. 71, 77–139 (2015).
Lv, H. et al. Genome-wide haplotype association study identify the FGFR2 gene as a risk gene for acute myeloid leukemia. Oncotarget 8, 7891 (2017).
Lahm, H. et al. Congenital heart disease risk loci identified by genome-wide association study in European patients. J. Clin. Investig. 131, e141837 (2021).
Cordell, H. J. et al. Genome-wide association study of multiple congenital heart disease phenotypes identifies a susceptibility locus for atrial septal defect at chromosome 4p16. Nat. Genet. 45, 822–824 (2013).
Wuttke, M. et al. Genetic loci associated with renal function measures and chronic kidney disease in children: The Pediatric Investigation for Genetic Factors Linked with Renal Progression Consortium. Nephrol. Dial. Transplant. 31, 262–269 (2016).
Liu, H. et al. Epigenomic and transcriptomic analyses define core cell types, genes and targetable mechanisms for kidney disease. Nat. Genet. 54, 950–962 (2022).
Söderholm, M. et al. Genome-wide association meta-analysis of functional outcome after ischemic stroke. Neurology 92, e1271–e1283 (2019).
Chen, J. et al. Diabetic kidney disease-predisposing proinflammatory and profibrotic genes identified by weighted gene co-expression network analysis (WGCNA). J. Cell. Biochem. 123, 481–492 (2022).
Chen, C. & Qiu, X. Identification of Hub Genes in Diabetic Nephropathy by an Integrated Bioinformatic Analysis (2021).
Comuzzie, A. G. et al. Novel genetic loci identified for the pathophysiology of childhood obesity in the Hispanic population. PLoS ONE 7, e51954 (2012).
Slavin, T. P., Feng, T., Schnell, A., Zhu, X. & Elston, R. C. Two-marker association tests yield new disease associations for coronary artery disease and hypertension. Hum. Genet. 130, 725–733 (2011).
Keaton, J. M. et al. Genome-wide interaction with the insulin secretion locus MTNR1B reveals CMIP as a novel type 2 diabetes susceptibility gene in African Americans. Genet. Epidemiol. 42, 559–570 (2018).
Surapaneni, A. et al. Identification of 969 protein quantitative trait loci in an African American population with kidney disease attributed to hypertension. Kidney Int. 102, 1167–1177 (2022).
Jourdan, T. et al. Cannabinoid-1 receptor deletion in podocytes mitigates both glomerular and tubular dysfunction in a mouse model of diabetic nephropathy. Diabetes Obes. Metab. 20, 698–708 (2018).
Pacher, P., Steffens, S., Haskó, G., Schindler, T. H. & Kunos, G. Cardiovascular effects of marijuana and synthetic cannabinoids: The good, the bad, and the ugly. Nat. Rev. Cardiol. 15, 151–166 (2018).
Astle, W. J. et al. The allelic landscape of human blood cell trait variation and links to common complex disease. Cell 167, 1415-1429.e1419 (2016).
Rajesh, M. et al. Cannabinoid receptor 2 activation alleviates diabetes-induced cardiac dysfunction, inflammation, oxidative stress, and fibrosis. GeroScience, 44(3), 1727–1741 (2022).
Sandholm, N. et al. Genome-wide meta-analysis and omics integration identifies novel genes associated with diabetic kidney disease. Diabetologia, 65(9), 1495–1509 (2022).
Murea, M. et al. Lipotoxicity in diabetic nephropathy: The potential role of fatty acid oxidation. Clin. J. Am. Soc. Nephrol. 5, 2373–2379 (2010).
Jelinek, H. F. et al. Clinical profiles, comorbidities and complications of type 2 diabetes mellitus in patients from United Arab Emirates. BMJ Open Diabetes Res. Care 5, e000427 (2017).
Zoppini, G. et al. Lower levels of 25-hydroxyvitamin D3 are associated with a higher prevalence of microvascular complications in patients with type 2 diabetes. BMJ Open Diabetes Res. Care 3, e000058 (2015).
Hovind, P., Rossing, P., Johnson, R. J. & Parving, H.-H. Serum uric acid as a new player in the development of diabetic nephropathy. J. Ren. Nutr. 21, 124–127 (2011).
Forbes, J. M. & Cooper, M. E. Mechanisms of diabetic complications. Physiol. Rev. 93, 137–188 (2013).
Bojestig, M., Arnqvist, H. J., Hermansson, G., Karlberg, B. E. & Ludvigsson, J. Declining incidence of nephropathy in insulin-dependent diabetes mellitus. N. Engl. J. Med. 330, 15–18 (1994).
Hovind, P. et al. Decreasing incidence of severe diabetic microangiopathy in type 1 diabetes. Diabetes Care 26, 1258–1264 (2003).
Yuan, M. et al. Δ9-Tetrahydrocannabinol regulates Th1/Th2 cytokine balance in activated human T cells. J. Neuroimmunol. 133, 124–131 (2002).
Srivastava, M. D., Srivastava, B. & Brouhard, B. Δ9 tetrahydrocannabinol and cannabidiol alter cytokine production by human immune cells. Immunopharmacology 40, 179–185 (1998).
Chen, M.-H. et al. Trans-ethnic and ancestry-specific blood-cell genetics in 746,667 individuals from 5 global populations. Cell 182, 1198-1213.e1114 (2020).
Vuckovic, D. et al. The polygenic and monogenic basis of blood traits and diseases. Cell 182, 1214-1231.e1211 (2020).
Zhang, X. et al. Cannabinoid 2 receptor agonist improves systemic sensitivity to insulin in high-fat diet/streptozotocin-induced diabetic mice. Cell. Physiol. Biochem. 40, 1175–1185 (2016).
Deveaux, V. et al. Cannabinoid CB2 receptor potentiates obesity-associated inflammation, insulin resistance and hepatic steatosis. PLoS ONE 4, e5844 (2009).
Tam, J. The emerging role of the endocannabinoid system in the pathogenesis and treatment of kidney diseases. J. Basic Clin. Physiol. Pharmacol. 27, 267–276 (2016).
Lecru, L. et al. Cannabinoid receptor 1 is a major mediator of renal fibrosis. Kidney Int. 88, 72–84 (2015).
Barutta, F., Bruno, G., Mastrocola, R., Bellini, S. & Gruden, G. The role of cannabinoid signaling in acute and chronic kidney diseases. Kidney Int. 94, 252–258 (2018).
Zhou, S. et al. Cannabinoid receptor type 2 promotes kidney fibrosis through orchestrating β-catenin signaling. Kidney Int. 99, 364–381 (2021).
Zhou, L. et al. Targeted inhibition of the type 2 cannabinoid receptor is a novel approach to reduce renal fibrosis. Kidney Int. 94, 756–772 (2018).
Agudo, J. et al. Deficiency of CB2 cannabinoid receptor in mice improves insulin sensitivity but increases food intake and obesity with age. Diabetologia 53, 2629–2640 (2010).
Thomsen, K., Nielsen, C. B. & Flyvbjerg, A. Effects of glycine on glomerular filtration rate and segmental tubular handling of sodium in conscious rats. Clin. Exp. Pharmacol. Physiol. 29, 449–454 (2002).
Zayed, H. The Arab genome: Health and wealth. Gene 592, 239–243 (2016).
Teebi, A. S. Introduction: Genetic diversity among Arabs. Genet. Disord. Among Arab Popul. 3–34 (2010).
Li, Y. R. & Keating, B. J. Trans-ethnic genome-wide association studies: Advantages and challenges of mapping in diverse populations. Genome Med. 6, 1–14 (2014).
Rosenberg, N. A. et al. Genome-wide association studies in diverse populations. Nat. Rev. Genet. 11, 356–366 (2010).
Peterson, R. E. et al. Genome-wide association studies in ancestrally diverse populations: Opportunities, methods, pitfalls, and recommendations. Cell 179, 589–603 (2019).
Scott, E. M. et al. Characterization of Greater Middle Eastern genetic variation for enhanced disease gene discovery. Nat. Genet. 48, 1071–1076 (2016).
Mousa, M. et al. Genome-wide association study of hospitalized COVID-19 patients in the United Arab Emirates. EBioMedicine 74, 103695 (2021).
The authors acknowledge and thank all the patients who volunteered to participate in this study. We would like to thank ICLDC Nursing and Research team for their assistance with patient recruitment, sample collection, processing, and data extraction. We are also grateful to Ms. Khayce Juma, Ms. Suna Nazar, Ms. Hessa Alkhoori, Ms. Fathimathuz Waasia and Ms. Hema Vurivi who assisted in processing the samples in the center for Biotechnology at Khalifa University. This project was supported by grants from Aljalila foundation AJF201756.
The authors declare no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Osman, W., Mousa, M., Albreiki, M. et al. A genome-wide association study identifies a possible role for cannabinoid signalling in the pathogenesis of diabetic kidney disease. Sci Rep 13, 4661 (2023). https://doi.org/10.1038/s41598-023-31701-w