Abstract
Despite the extensive efforts of scientists, the genetic background of diabetic nephropathy (DN) has not yet been clarified. To elucidate the genetic variants that predispose to the development of DN, we conducted a systematic review and meta-analysis of all available genetic association studies (GAS) of DN. We searched in the Human Genome Epidemiology Navigator (HuGE Navigator) and PubMed for available GAS of DN. The threshold for meta-analysis was three studies per genetic variant. The association between genotype distribution and DN was examined using the generalized linear odds ratio (ORG). For variants with available allele frequencies, the examined model was the allele contrast. The pooled OR was estimated using the DerSimonian and Laird random effects model. The publication bias was assessed with Egger’s test. We performed pathway analysis of significant genes with DAVID 6.7. Genetic data of 606 variants located in 228 genes were retrieved from 360 GASs and were synthesized with meta-analytic methods. ACACB, angiotensin I-converting enzyme (ACE), ADIPOQ, AGT, AGTR1, AKR1B1, APOC1, APOE, ATP1B2, ATP2A3, CARS, CCR5, CGNL1, Carnosine dipeptidase 1 (CNDP1), CYGB-PRCD, EDN1, Engulfment and cell motility 1 (ELMO1), ENPP1, EPO, FLT4, FTO, GLO1, HMGA2, IGF2/INS/TH cluster, interleukin 1B (IL1B), IL8, IL10, KCNQ1, KNG, LOC101927627, Methylenetetrahydrofolate reductase, nitric oxide synthase 3 (NOS3), SET domain containing seven, histone lysine methyltransferase (SETD7), Sirtuin 1 (SIRT1), SLC2A1, SLC2A2, SLC12A3, SLC19A3, TCF7L2, TGFB1, TIMP1, TTC39C, UNC13B, VEGFA, WTAPP1, WWC1 as well as XYLT1 and three intergenic polymorphisms showed significant association with DN. Pathway analysis revealed the overrepresentation of six signalling pathways. The significant findings provide further evidence for genetic factors implication in DN offering new perspectives in discovery of new therapies.
Keywords: association, diabetic nephropathy, genetic, meta-analysis, polymorphisms, systematic review
INTRODUCTION
Diabetic nephropathy (DN) has complex aetiology due to synergistic interplay between many factors, modifiable or not. Among them, prominent contribution is attributed to glycaemic and haemodynamic factors, as well as the genetic background [1, 2]. Despite the extensive research, the genetic architecture is poorly understood [3, 4]. Although many genetic variants have been implicated in DN susceptibility, the results are not replicable and create inconsistency [4].
Family-based studies through linkage analysis suggest familial clustering in many populations, but this type of method has the potential to detect rare variants with large effects [5–9]. Unlike non-DN and apolipoprotein L1 (APOL1) confirmed susceptibility, in the case of DN such established genetic contribution has not been achieved [10].
Population-based studies, the well-known genetic association studies (GAS), tried to shed light on genetic background on many common diseases [11]. In particular, when the first wave of genome-wide association studies (GWAS) appeared, the researchers were extremely optimistic for the potential of these [12], but even GWAS results have failed to be replicated. In DN, many GWAS have been conducted with promising results, but the burden of DN still reaches alarming proportions [13–24].
In an effort to increase power and enlighten the genetic architecture of DN, meta-analysis of genetic studies promises more convincing evidence. Meta-analysis of genome-wide linkage studies of DN [25] or renal function traits [26] has been performed, indicating suggestive evidence for many cytogenetic locations. We conducted a field synopsis including all the case–control GAS that examine the association between genetic variants and DN.
MATERIALS AND METHODS
Identification and eligibility of relevant studies
Τo elucidate the contribution of genetic background in development of DN, we conducted a systematic review and meta-analysis of GASs in DN. In the meta-analysis, studies in English recorded in the Human Genome Epidemiology (HuGE) Phenopedia (last update on 3 July 2019) regarding the disease term ‘diabetic nephropathies’ were included. We also retrieved articles from GWAS in HuGE Publit and the National Human Genome Research Institute (NHGRI) Catalog of Published GWAS (http://www.genome.gov/gwastudies/). We cross-checked manually these findings against those indexed in PubMed using the search terms [(‘diabetic nephropathy’ OR ‘diabetic kidney disease’ OR ‘diabetic end stage renal disease’) AND ( ‘genetic association’ OR ‘gene’ OR ‘variant’ OR ‘polymorphism’)] (accessed on 7 July 2019). Finally, any meta-analyses and the references of the eligible articles were also screened. We did not request unpublished data from any author.
The included studies should meet the following criteria: (i) they involved cases with persistent micro/macroalbuminuria with or without diabetic retinopathy; (ii) they involved diseased controls with diabetes and normoalbuminuria or normal renal function and/or healthy controls; (iii) they provided full genotypic data, either genotype counts or allele frequencies, excluding the articles with results after having merged genotypes; and (iv) they included human subjects. The diabetes could be either Type I diabetes mellitus (T1DM) or Type II diabetes mellitus (T2DM). No study involving exclusive cases with not persistent microalbuminuria was included.
Studies examining disease progression, phenotype modification, response to treatment or survival were excluded. Case reports, editorials, reviews, non-English articles, unpublished studies as well as studies with other study designs, such as family-based studies, were also excluded. The eligibility of the articles was assessed independently by two investigators (M.T. and E.Z.), the results were compared and any disagreements were resolved by reaching consensus.
Data extraction
From each article, the following information was extracted: first author, year of publication, ethnicity, PubMed ID (PMID) , type of diabetes and the phenotype. For cases and controls, we recorded their number and the selection criteria. With regard to the genotypic data, we extracted, if available, the full genotype counts or allele frequencies.
Data synthesis and analysis
The association between genotypes and DN was examined using the generalized linear odds ratio (ORG) [27, 28]. For the variants with available allele frequencies, the examined model was the allele-contrast. The threshold for meta-analysis was the presence of three studies. The pooled OR was estimated using DerSimonian and Laird random-effects model [29]. The associations are presented with ORs with corresponding 95% confidence intervals (CIs). We tested for between-study heterogeneity with Cochran’s Q statistic (considered statistically significant at P < 0.10) and assessed its extent with the I2 statistic [30, 31]. ORG was calculated using generalized odds ratio methodology for the analysis and meta-analysis of GAS (ORGGASMA) (http://biomath.med.uth.gr) [27, 28]. Furthermore, we conducted a subgroup analysis regarding the DM type and ethnicity in case of existence of >10 studies per genetic polymorphism.
For each study, we examined if controls confronted with Hardy–Weinberg equilibrium (HWE) predicted genotypes using Fisher’s exact test. For studies providing only allele counts, we relied on the authors' assessment of deviations from HWE. We also tested for ‘small-study effect’ with the Egger test [32].
Pathway analysis
We performed pathway analysis of statistically significant genes with DAVID version 6.7 to identify in the signalling pathways that are overrepresented by the significant genes [33, 34].
RESULTS
Study characteristics
The literature search retrieved 3697 records after removing of duplicates. When an article provided data for different populations, then each population was considered as a different study. Different ethnic descents were categorized as Caucasians, Asians, Africans and mixed. Figure 1 presents a flowchart of retrieved articles and excluded articles with specification of reasons for exclusion. The threshold for a variant in order to be meta-analysed was three studies per variant. Overall, 227 candidate genes and 606 polymorphisms were investigated in 356 articles. The characteristics of each study and their references are shown in Supplementary data, Table S1. The studies were published between 1994 and 2019.
FIGURE 1.
Flowchart of retrieved articles with specifications of reasons for exclusion. RCT: randomized controlled trial; TDT: transmission disequilibrium test.
Main meta-analysis results
Tables 1–3 show the statistically significant results of meta-analyses exploring the presence of association between the relevant genetic variants and DN based on genotype counts. An overview of pooled ORG of the statistically significant variants is shown in Figures 2–5. The statistically significant results of meta-analyses exploring the presence of association between the genetic variants and DN based on allele counts are presented in Table 4. An overview of the meta-analysis results based on allele counts is shown in Figure 5. The meaning of the acronyms of statistically significant genes is shown in Supplementary data, Table S2. The non-significant results from meta-analysis are presented in Supplementary data, Tables S3–S7. Supplementary data, Table S8 summarizes information on the functional implications of the genetic variants associated with DN, indicating whether the gene variants are associated with increased or decreased expression of the gene or increased or decreased activity or levels of the gene product, if this is known.
Table 1.
Statistically significant results from meta-analysis between diseased controls (diabetics with normoalbuminuria) versus cases (diabetics with diabetic nephropathy) based on genotype counts
| Gene | Variant | RS | Studies (n) | Cases/controls (n) | RE ORG (95% CI) | I 2 (%) | PQ | PE |
|---|---|---|---|---|---|---|---|---|
| ACACB | (C > T) | rs2268388 | 10 | 3222/2881 | 1.41 (1.09–1.80) | 81.65 | 0 | 0.3 |
| All in HWE | 7 | 2424/2190 | 1.30 (1.02–1.65) | 72.54 | 0 | 0.48 | ||
| ACE | I > D | 65 | 10787/10404 | 1.21 (1.09–1.34) | 74.61 | 0 | 0.70 | |
| All in HWE | I > D | 55 | 8733/8267 | 1.26 (1.14–1.39) | 63.85 | 0 | 0.54 | |
| AGT | M235T | rs699 | 26 | 5015/5253 | 1.21 (1.01–1.45) | 82,45 | 0,00 | 0.84 |
| All in HWE | 19 | 3181/3655 | 1.09 (0.92–1.31) | 72.76 | 0.00 | 0.95 | ||
| AKR1B1 | C-106T | rs759853 | 7 | 1149/1068 | 1.53 (1.15–2.02) | 65.11 | 0.01 | 0.3 |
| All in HWE | 6 | 1064/922 | 1.51 (1.08–2.10) | 70.38 | 0 | 0.33 | ||
| APOC1 | C > T | rs4420638 | 3 | 1526/1620 | 1.47 (1.27–1.70) | 0 | 0.61 | 0.21 |
| CARS | (G > A) | rs451041 | 5 | 3061/3170 | 1.22 (1.07–1.38) | 47.41 | 0.11 | 0.06 |
| All in HWE | 4 | 2571/2230 | 1.25 (1.07–1.45) | 50.50 | 0.11 | 0.06 | ||
| CCR5 | −59029 A > G | rs1799987 | 8 | 2125/2127 | 0.69 (0.53–0.91) | 83.21 | 0.00 | 0.04 |
| All in HWE | 6 | 1789/1780 | 0.79 (0.60–1.03) | 80.27 | 0.00 | 0.07 | ||
| CNDP1 | [CTG]5 > [CTG]6/7 | D18S880 | 11 | 4064/7318 | 1.38 (1.17–1.64) | 64.78 | 0.002 | 0.01 |
| All in HWE | 10 | 3184/3316 | 1.47 (1.19–1.82) | 66.46 | 0.001 | 0.002 | ||
| ELMO1 | rs741301 | 5 | 1072/984 | 1.27 (1.04–1.56) | 32.92 | 0.20 | 0.83 | |
| All in HWE | 4 | 1.23 (0.91–1.66) | 44.24 | 0.15 | 0.40 | |||
| ENPP1 | K173Q (A > C) | rs1044498 | 4 | 744/982 | 1.40 (1.03–1.91) | 62.33 | 0.05 | 0.21 |
| All in HWE | 3 | 1.27 (0.90–1.81) | 59.93 | 0.08 | 0.25 | |||
| EPO | G > T | rs1617640 | 3 | 1618/954 | 1.64 (1.43–1.89) | 0.00 | 0.78 | 0.03 |
| HMGA2 | (G/C) | rs1531343 | 3 | 1233/2125 | 1.49 (1.23–1.81) | 0 | 0.41 | 0.45 |
| IL1B | −511C/T | rs16944 | 3 | 774/665 | 1.66 (1.38–2.01) | 0 | 0.86 | |
| IL8 | −251 T > A | rs4073 | 5 | 661/703 | 1.40 (1.07–1.73) | 36.43 | 0.18 | 0.49 |
| All in HWE | 4 | 611/565 | 1.28 (0.98–1.67) | 38.13 | 0.18 | 0.26 | ||
| IL10 | −1082 A > G | rs1800896 | 4 | 677/761 | 1.23 (1.01–1.49) | 0 | 0.56 | 0.63 |
| All in HWE | 2 | 610/690 | 1.25 (1.02–1.53) | 0 | 0.62 | NA | ||
| KCNQ1 | rs2237897 | 4 | 1539/2208 | 1.30 (1.11–1.53) | 30.60 | 0.23 | 0.12 | |
| LOC101927627 | G > A | rs1411766 | 6 | 3780/3561 | 1.30 (1.19–1.42) | 0 | 0.64 | 0.77 |
| All in HWE | 5 | 1.31 (1.19–1.45) | 0 | 0.53 | 0.82 | |||
| MTHFR | C677T | 26 | 4246/4380 | 1.43 (1.16–1.77) | 83.91 | 0 | 0.49 | |
| All in HWE | 21 | 3163/3085 | 1.28 (1.03–1.59) | 78.12 | 0 | 0.51 | ||
| NOS3 | T-786C | rs2070744 | 7 | 1934/1710 | 1.21 (1.07–1.36) | 0.56 | 0.42 | 0.49 |
| All in HWE | 5 | 1672/1418 | 1.22 (1.04–1.44) | 33.45 | 0.20 | 0.17 | ||
| rs1866813 | 5 | 1798/1831 | 1.40 (1.18–1.67) | 26.09 | 0.25 | 0.02 | ||
| SDC2 | BamHI T > G | 3 | 485/231 | 0.70 (0.52–0.93) | 0 | 0.45 | 0.14 | |
| SETD7 | rs2592970 | 3 | 1939/2222 | 0.90 (0.81–0.997) | 0 | 0.44 | 0.3 | |
| SIRT1 | rs4746720 | 6 | 2785/2484 | 0.88 (0.80–0.97) | 0 | 0.55 | 0.08 | |
| All in HWE | 6 | |||||||
| SLC12A3 | +1870 G/A | 3 | 1070/650 | 0.39 (0.28–0.53) | 0 | 0.73 | 0.01 | |
| All in HWE | 2 | 558/334 | 0.36 (0.23–0.57) | 0 | 0.52 | NA | ||
| SLC2A1 | XbaI(+) > XbaI(−) | rs841853 | 10 | 1438/1331 | 1.44 (1.06–1.96) | 78.03 | <0.001 | 0.13 |
| All in HWE | 7 | 938/840 | 1.71 (1.19–2.45) | 75.80 | <0.001 | 0.30 | ||
| UNC13B | rs13293564 | 4 | 1573/1910 | 1.27 (1.13–1.42) | 0 | 0.54 | 0.08 |
RS: SNP identifier, RE ORG: random effects odds ratio generalized, I2: I2 statistic, NA: non applicable, PQ: P-value from heterogeneity testing, PE: P-value from Egger’s test.
Table 3.
Statistically significant results from meta-analysis between healthy controls versus cases (diabetics with diabetic nephropathy) based on genotype counts
| Gene | Variant | RS | Studies (n) | Cases/controls (n) | RE ORG (95% CI) | I 2 (%) | PQ | PE |
|---|---|---|---|---|---|---|---|---|
| ACE | I > D | 30 | 3690/4927 | 1.24 (1.02–1.52) | 83.20 | 0.00 | 0.03 | |
| All in HWE | I > D | 29 | 3283/4695 | 1.26 (1.02–1.55) | 82.87 | 0.00 | 0.01 | |
| APOE | E3 > e2/e4 | 7 | 1109/1371 | 1.68 (1.11–2.52) | 73.91 | 0 | 0.14 | |
| ELMO1 | rs1345365 | 3 | 1204/1241 | 0.79 (0.69–0.91) | 0 | 0.92 | 0.55 | |
| All in HWE | 3 | |||||||
| MTHFR | C677T | 14 | 1652/2603 | 1.79 (1.23–2.60) | 88.10 | 0 | 0.91 | |
| All in HWE | 8 | 847/1295 | 1.41 (0.94–2.12) | 80.38 | 0 | 0.36 | ||
| NOS3 | G894T | rs1799983 | 9 | 1941/2222 | 1.69 (1.15–2.48) | 85.21 | 0 | 0.18 |
| All in HWE | 7 | 1826/1852 | 1.84 (1.26–2.7) | 83.62 | 0 | 0.04 | ||
| SLC2A1 | XbaI(+) > XbaI(−) | rs841853 | 7 | 1.87 (1.11–3.15) | 89.74 | 0 | 0.89 | |
| All in HWE | 7 | |||||||
| TCF7L2 | C > T | rs7903146 | 4 | 1262/2526 | 2.19 (1.69–2.83) | 70.75 | 0.02 | 0.44 |
| All in HWE | 3 | 685/1930 | 2.46 (2.00–3.01) | 27.03 | 0.25 | 0.43 | ||
| TGFB1 | T869C | rs1800470 | 6 | 814/1450 | 1.30 (0.86–1.96) | 83.64 | 0 | 0.18 |
| All in HWE | 4 | 706/1103 | 1.73 (1.46–2.04) | 0 | 0.41 | 0.21 |
RS: SNP identifier, RE ORG: random effects odds ratio generalized, I2: I2 statistic, NA:nonapplicable,PQ: P-value from heterogeneity testing, PE: P-value from Egger’s test.
FIGURE 2.
Forest plot of diseased controls (diabetics with normoalbuminuria) and cases (diabetics with diabetic nephropathy) displaying only significant results based on genotype counts.
Table 4.
Statistically significant results from meta-analysis between diseased controls (diabetics with normoalbuminuria) versus cases (diabetics with diabetic nephropathy) based on allele counts
| Gene | Variant | RS | Studies (n) | Cases/controls (n) | RE OR (95% CI) | I 2 (%) | P | PE |
|---|---|---|---|---|---|---|---|---|
| rs4972593 | 4 | 1582/3499 | 1.23 (1.06–1.43) | 27.22 | 0.25 | 0.07 | ||
| ACE | T8968C | rs4311 | 3 | 1042/1123 | 0.86 (0.76–0.98) | 14.88 | 0.31 | 0.27 |
| ADIPOQ | −11391G > A | rs17300539 | 6 | 2629/3039 | 1.18 (1.04–1.35) | 0 | 0.46 | 0.38 |
| ATP1B2 | rs1642763 | 3 | 1176/1323 | 1.17 (1.03–1.34) | 0 | 0.45 | 0.14 | |
| rs1642764 | 3 | 1176/1323 | 1.15 (1.02–1.28) | 0 | 0.81 | 0.07 | ||
| ATP2A3 | rs1062683 | 3 | 1176/1323 | 1.18 (1.02–1.37) | 0 | 0.54 | 0.3 | |
| rs8068346 | 3 | 1176/1323 | 1.17 (1.00–1.38) | 0 | 0.39 | 0.3 | ||
| CGNL1 | rs16977473 | 3 | 1176/1323 | 1.299 (1.161–1455) | 0 | 0.399 | 0.51 | |
| CYGB. PRCD | rs895157 | 3 | 1176/1323 | 1.214 (1.114–1.323) | 0 | 0.726 | 0.68 | |
| EDN1 | rs1794849 | 3 | 1176/1323 | 1.16 (1.02–1.31) | 0 | 0.62 | 0.08 | |
| ELMO1 | A > G | rs741301 | 3 | 1526/1563 | 0.793 (0.661–0.951) | 56.19 | 0.102 | 0.09 |
| FLT4 | rs2242221 | 3 | 1176/1323 | 1.14 (1.01–1.29) | 0 | 0.38 | 0.43 | |
| FTO | rs56094641 | 3 | 1176/1323 | 1.187 (1.110–1.269) | 0 | 0.783 | 0.51 | |
| GLO1 | rs7769206 | 3 | 1176/1323 | 1.22 (1.02–1.47) | 0 | 0.73 | 0.07 | |
| IGF2/INS/TH cluster | rs1004446 | 3 | 1176/1323 | 1.16 (1.03–1.31) | 0 | 0.49 | 0.22 | |
| rs4320932 | 3 | 1176/1323 | 0.84 (0.73–0.96) | 0 | 0.43 | 0.06 | ||
| KNG | rs1656922 | 3 | 1057/1127 | 1.13 (1.003–1.27) | 0 | 0.90 | 0.36 | |
| LOC107986400 | 3 | 1.570 (1.277–1.929) | 0 | 0.614 | 0.42 | |||
| – | – | rs11225445 | 3 | 1640/1770 | 1.26 (1.14–1.39) | 0.27 | 0.27 | |
| – | rs610950 | 3 | 1640/1770 | 0.55 (0.44–0.68) | 0.36 | 0.36 | ||
| – | rs613804 | 3 | 1640/1770 | 0.58 (0.46–0.72) | 0.37 | 0.37 | ||
| SLC2A2 | rs5400 | 3 | 1057/1127 | 0.82 (0.69–0.98) | 0 | 0.54 | 0 | |
| SLC19A3 | rs12694743 | 5 | 2086/603 | 0.750 (0.564–0.997) | 37.91 | 0.169 | 0.21 | |
| rs6713116 | 5 | 2086/602 | 0.759 (0.631–0.914) | 0 | 0.498 | 0.23 | ||
| TIMP1 | ss95210393 | 3 | 1176/1323 | 0.88 (0.78–0.99) | 0 | 0.73 | 0.46 | |
| rs6520277 | 3 | 1176/1323 | 1.18 (1.05–1.33) | 10.75 | 0.33 | 0.3 | ||
| TTC39C | rs16940484 | 3 | 1176/1323 | 1.132 (1.066–1.202) | 0 | 0.988 | 0.25 | |
| UNC13B | rs13285401 | 3 | 1176/1323 | 0.88 (0.79–0.99) | 0 | 0.49 | 0.1 | |
| rs10972365 | 3 | 1176/1323 | 0.82 (0.72–0.93) | 0 | 0.46 | 0.41 | ||
| rs17360668 | 3 | 1176/1323 | 0.80 (0.68–0.93) | 24.65 | 0.27 | 0.45 | ||
| VEGFA | C > A | rs2146323 | 3 | 1176/1323 | 0.85 (0.76–0.95) | 0.2 | 0.2 | |
| WTAPP1 | rs756544 | 3 | 1640/1770 | 0.82 (0.75–0.90) | 0.46 | 0.46 | ||
| rs11225434 | 3 | 1640/1770 | 0.83 (0.75–0.91) | 0.46 | 0.46 | |||
| rs1034375 | 3 | 3280/3540 | 0.76 (0.64–0.91) | 0.38 | 0.38 | |||
| rs7926920 | 3 | 820/885 | 0.82 (0.75–0.90) | 0.46 | 0.46 | |||
| WWC1 | rs78954674 | 3 | 1176/1323 | 1.188 (1.100–1.284) | 0 | 0.597 | 0.25 | |
| XYLT1 | c.343G > T | 3 | 501/485 | 1.77 (1.08–2.91) | 0 | 0.682 | 0.49 |
RS: SNP identifier, RE ORG: random effects odds ratio generalized, I2: I2 statistic, NA:nonapplicable,PQ: P-value from heterogeneity testing, PE: P-value from Egger’s test.
FIGURE 5.
Forest plot of diseased controls (diabetics with normoalbuminuria) and cases (diabetics with diabetic nephropathy) displaying only significant results based on allele counts.
FIGURE 3.
Forest plot of healthy controls versus diseased controls (diabetics with normoalbuminuria) versus cases (diabetics with diabetic nephropathy) displaying only significant results based on genotype counts.
FIGURE 4.
Forest plot of healthy controls and cases (diabetics with diabetic nephropathy) displaying only significant results based on genotype counts.
Table 2.
Statistically significant results from meta-analysis between healthy controls versus diseased controls (diabetics with normoalbuminuria) versus cases (diabetics with diabetic nephropathy) based on genotype counts
| Gene | Variant | RS | Studies (n) | Cases/diseased controls/healthy (n) | RE ORG (95% CI) | I 2 (%) | PQ | PE |
|---|---|---|---|---|---|---|---|---|
| AGT | M235T | rs699 | 9 | 1156/1350/1243 | 1.21 (1.02–1.45) | 65.91 | 0 | 0.11 |
| APOE | E3 > e2/e4 | 6 | 743/571/1171 | 1.57 (1.03–2.40) | 86.26 | 0 | 0.4 | |
| MTHFR | C677T | 10 | 1345/1306/1903 | 1.44 (1.13–1.83) | 84.45 | 0 | 0.3 | |
| All in HWE | 6 | 1.14 (0.98–1.32) | 26.51 | 0.24 | 0.49 | |||
| NOS3 | G894T | rs1799983 | 6 | 1152/811/117 | 1.26 (0.94–1.70) | 77.82 | <0.001 | 0.95 |
| G894T | rs1799983 | 5 | 1085/775/1027 | 1.40 (1.11–1.77) | 58.93 | 0.05 | 0.16 | |
| T-786C | rs2070744 | 3 | 1.23 (1.09–1.38) | 0 | 0.76 | 0.83 | ||
| All in HWE | 2 | 1.22 (1.08–1.38) | 0 | 0.59 | – | |||
| SLC2A1 | XbaI(+) > XbaI(−) | rs841853 | 7 | 1.50 (1.05–2.15) | 89.18 | 0 | 0.49 | |
| TGFB1 | T869C | rs1800470 | 5 | 770/787/1332 | 1.36 (1.08–1.70) | 73.19 | 0 | 0.24 |
| All in HWE | 4 | 706/727/1103 | 1.47 (1.24–1.75) | 45.66 | 0.14 | 0.15 |
RS: SNP identifier, RE ORG: random effects odds ratio generalized, I2: I2 statistic, NA:nonapplicable,PQ: P-value from heterogeneity testing, PE: P-value from Egger’s test.
The pathway analysis was performed for the genes with significant association with DN and revealed the overrepresentation of six signalling pathways: the cytokine–cytokine receptor interaction, the pyruvate metabolism, T2DM, the adipocytokine signalling pathway, the renal cell carcinoma and the renin–angiotensin system pathway. The results of the statistically significant genes are presented based on the relevant signalling pathway (Table 5). Supplementary data, Table S9 illustrates where in the pathway the genes sit.
Table 5.
Results from pathway analysis of statistically significant genes
| Category | Term | Genes | Count | % | P-value | Benjamini |
|---|---|---|---|---|---|---|
| KEGG_PATHWAY | Cytokine–cytokine receptor interaction | CCR5, EPO, FLT4, IL1B, IL10, TGFB1, VEGFA | 7 | 1.2 | 4.4E-3 | 3.0E-1 |
| KEGG_PATHWAY | Pyruvate metabolism | ACACB, AKR1B1, GLO1 | 3 | 0.5 | 2.4E-2 | 6.3E-1 |
| KEGG_PATHWAY | T2DM | ADIPOQ, IFG2, INS, SLC2A2 | 3 | 0.5 | 3.3E-2 | 5.9E-1 |
| KEGG_PATHWAY | Adipocytokine signalling pathway | ACACB, ADIPOQ, SLC2A1 | 3 | 0.5 | 6.2E-2 | 7.3E-1 |
| KEGG_PATHWAY | Renal cell carcinoma | SLC2A1, TGFB1, VEGFA | 3 | 0.5 | 6.7E-2 | 6.8E-1 |
| KEGG_PATHWAY | Renin–angiotensin system | ACE, AGT | 2 | 0.4 | 9.9E-2 | 7.5E-1 |
Genetic variants related to cytokine–cytokine receptor interaction
The polymorphism –59029A/G in CCR5 showed significant association with DN in comparison between diabetics with normoalbuminuria and cases with DN, with a pooled ORG of 0.69 (95% CI 0.53–0.91) . This association was also significant in analysis of only studies with controls in HWE.
EPO rs1617640 polymorphism showed significant association with DN in comparison between diabetics without nephropathy and cases with DN with a pooled ORG of 1.64 (95% CI 1.43–1.89).
Fms related receptor tyrosine kinase 4 (FLT4) rs2242221 polymorphism was also associated with DN in analysis based on allele counts, with a pooled ORG of 1.14 (95% CI 1.01–1.29).
The polymorphism –511C/T in interleukin 1B (IL1B) showed significant association with DN with an ORG of 1.66 (95% CI 1.38–2.01).
The polymorphism –251T/A in IL10 in analysis between diabetics with normoalbuminuria and cases with DN was also associated with DN, with a pooled ORG of 1.40 (95% CI 1.07–1.73).
The polymorphism T869C in TGFB1 also showed significant association with DN in both analysis between three groups and in analysis between healthy controls and cases with DN, with a pooled ORG of 1.36 (95% CI 1.08–1.70) and ORG of 1.73 (95% CI 1.46–2.04), respectively.
Rs2146323 polymorphism in vascular endothelial growth factor A (VEGFA) was significantly associated with DN, with a pooled OR of 0.85 (95% CI 0.76–0.95) when diseased controls were compared with cases with DN with the allele contrast model.
Genetic variants related with pyruvate metabolism
Two polymorphisms in the acetyl-CoA carboxylase beta (ACACB) gene, rs2268388 and rs5186, showed significant association with DN when diabetics without DN were compared with cases with DN, with a pooled ORG of 1.41 (95% CI 1.09–1.80) and 1.59 (95% CI 1.28–1.98), respectively. The aforementioned associations were also significant in meta-analysis with studies in HWE.
In a comparison of diseased controls versus cases, the AKR1B1 rs759853 variant was significantly associated with DN, with a pooled ORG of 1.53 (95% CI 1.15–2.02). The association was also significant in meta-analysis of studies in HWE (1.51, 95% CI 1.08–2.10).
Rs7769206 in GLO1 showed a significant association with DN in comparison between diabetics with normoalbuminuria and cases with DN, with a pooled ORG of 1.22 (95% CI 1.02–1.47).
Genetic variants related with T2DM
The polymorphism –11391G/A in ADIPOQ showed a significant association with DN in comparison between diabetics with normoalbuminuria and cases with DN, with a pooled ORG of 1.18 (95% CI 1.04–1.35).
The polymorphisms rs10004446 and rs4320932 in the IGF2/INS/TH cluster showed a significant association with DN in comparison between diabetics with normoalbuminuria and cases with DN, with a pooled ORG of 1.16 (95% CI 1.03–1.31) and ORG of 0.84 (95% CI 0.73–0.96), respectively.
The polymorphism rs5400 in SLC2A2 was also significant when diabetics with normoalbuminuria and cases with DN were compared, with a pooled ORG of 0.82 (95% CI 0.69–0.98).
Genetic variants related with adipocytokine signalling pathway
Except ACACB and adiponectin (ADIPOQ) variants significant association, SLC2A1 rs841853 variant was also significantly associated with DN, with a pooled ORG of 1.44 (95% CI 1.06–1.96). This association remained significant in meta-analysis with the seven studies in HWE (1.71, 95% CI 1.19–2.45). Rs841853 was also significantly associated with DN in meta-analysis with three groups and between healthy controls and cases with DN, with a pooled ORG of 1.50 (95% CI 1.05–2.15) and 1.87 (95% CI 1.11–3.15).
Genetic variants related with renal cell carcinoma
In this pathway, SLC2A1, TGFB1 and VEGFA are involved, which were mentioned above.
Genetic variants related with renin–angiotensin system
In diseased controls versus cases, the deletion of the polymorphism rs179975 in angiotensin I-converting enzyme (ACE) was associated with DN, with a pooled ORG of 1.21 (95% CI 1.09–1.34) after meta-analysis conducted in 65 studies. The association was also significant in sensitivity analysis of 55 studies in HWE with an ORG of 1.26 (95% CI 1.14–1.39). The polymorphism rs179975 was also associated with DN, with an ORG of 1.24 (95% CI 1.02–1.52) when healthy controls were compared with cases with DN.
Another gene implicated in the renin–angiotensin–aldosterone system with significant results in a meta-analysis of 26 studies is AGT, with an ORG of 1.21 (95% CI 1.01–1.45). The variant M235T (rs699) was also associated with DN in comparison between healthy controls versus diabetics without nephropathy versus cases with DN, with a pooled ORG of 1.21 (95% CI 1.02–1.45).
Some genes were not classified in signalling pathways and thus, these genes are discussed based on their biologic role and function.
Genetic variants involved in lipid metabolism
The polymorphism rs4420638 of the APOC1 gene was significantly associated with DN, with a pooled ORG of 1.47 (95% CI 1.27–1.70) in a comparison of diabetics without nephropathy and cases with DN. In a comparison of healthy controls versus diseased controls versus cases, and comparison of healthy controls versus cases with DN, variant E2 and E4 was significantly associated with DN, with a pooled ORG of 1.57 (95% CI 1.03–2.40) and 1.68 (95% CI 1.11–2.52), respectively.
Genetic variants involved in endothelial function
The polymorphism –786T/C in nitric oxide synthase 3 (NOS3) gene was significantly associated with DN with a pooled ORG of 1.21 (95% CI 1.07–1.36), and this association was also significant in meta-analysis of only studies in HWE. The polymorphism –786T/C in the NOS3 gene also showed a significant association with DN in a comparison of healthy controls versus diabetics without DN versus cases with DN, with an ORG of 1.23 (95% CI 1.09–1.38), which was also significant in meta-analysis with only studies in HWE.
Another polymorphism in NOS3 gene, G894T, was significantly associated with DN both in comparison with three groups and between healthy controls versus cases, with DN with a pooled ORG of 1.40 (95% CI 1.11–1.77) and 1.69 (95% CI1.15–2.48), respectively.
Genetic variants involved in epigenetic procedures
SET domain containing seven, histone lysine methyltransferase (SETD7) rs2592970 polymorphism showed a significant protective association with DN with a pooled ORG of 0.90 (95% CI 0.81–0.997).
Methylenetetrahydrofolate reductase (MTHFR) is critical for production of S-adenosyl-l-methionine, the principal methyl donor. The C677T polymorphism in MTHFR gene showed a significant association with DN, with a pooled ORG of 1.45 (95% CI 1.16–1.82). This significance was also significant in meta-analysis of only studies in HWE. C677T was also significantly associated with DN in comparisons of three groups and between comparison of healthy controls versus cases with DN, with a pooled ORG of 1.44 (95% CI 1.13–1.83) and 1.68 (1.15–2.47), respectively.
Sirtuin 1 (SIRT1), a NAD+-dependent deacetylase, and more specifically its variant, rs4746720, showed significant association with DN, with a pooled ORG of 0.89 (95% CI 0.81–0.99).
Genetic variants identified by GWAS
Rs451041 in the CARS gene was significantly associated with DN with a pooled ORG of 1.22 (95% CI 1.07–1.38). This association was also significant in meta-analysis of only studies in HWE.
Carnosine dipeptidase 1 (CNDP1) D18S880 variant was also associated with DN with a pooled ORG of 1.38 (95% CI 1.17–1.64), and this association was also significant in meta-analysis with studies in HWE. The Engulfment and cell motility 1 (ELMO1) rs741301 polymorphism was significantly associated with DN with a pooled ORG of 1.27 (95% CI 1.04–1.56). Solute carrier family 12 member 3 (SLC12A3) 1870G/A showed a significant protective effect against DN, with a pooled ORG of 0.39 (95% CI 0.28–0.53).
Other genes
K173Q in ENPP1, rs13293564 in UNC13B, rs1531343 in HMGA2, SDC2 BamHI T/G polymorphism, rs7903146 in TCF7L2, T869C in TGFB1, rs2237897 in KCNQ, rs1866813 and two loci, LOC101927627 and LOC105370358, showed significant association with DN. Regarding the allele contrast model, significant associations were observed for ACE, ADIPOQ, ATP1B2, ATP2A3, CGNL1, CYGB, PRCD, EDN1, ELMO1, FTO, KNG, LOC107986400, WTAPP1, SLC19A3, TIMP1, TTC39C, UNC13B, WWC1 and XYLT1. Genes ACE, ELMO1 and UNC13B were reproducibly associated in meta-analysis based on genotype counts.
Subgroup analysis
When there were >10 studies per variant, a subgroup analysis was conducted regarding the type of diabetes and the ethnicity of the participants. More specifically, subgroup analyses were conducted for variants in seven genes including ACE, AGT, AGTR1, APOE, MTHFR, NOS3 and TGFB1 (Table 6). The majority of the variants were reproducibly associated with DN in more than one subgroup analysis.
Table 6.
Subgroup analysis results based on type of diabetes and ethnicity
| Gene | Variant | RS | Studies (n) | Cases/controls (n) | RE ORG (95% CI) | I 2 (%) | PQ | ||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Disease controls–cases | |||||||||||
| ACE | All | I > D | 65 | 10787/10404 | 1.21 (1.09–1.34) | 74.61 | 0 | ||||
| All in HWE | 55 | 8733/8267 | 1.26 (1.14–1.39) | 63.85 | 0 | ||||||
| Caucasians | 25 | 4176/4008 | 1.1 (0.99–1.22) | 38.92 | 0.03 | ||||||
| Asians | 35 | 5228/5013 | 1.24 (1.05–1.45) | 78.59 | <0.001 | ||||||
| T1DM | 19 | 2852/2973 | 1.1 (0.96–1.25) | 43.68 | 0.02 | ||||||
| T2DM | 46 | 7935/7431 | 1.25 (1.10–1.43) | 79.10 | <0.001 | ||||||
| Three groups | |||||||||||
| ACE | All | I > D | 21 | 2973/2622/3668 | 1.12 (0.94–1.34) | 88.33 | 0 | ||||
| All in HWE | 20 | 2566/2437/3436 | 1.13 (0.94–1.37) | 88.32 | 0 | ||||||
| Caucasians | 7 | 865/691/1075 | 1.03 (0.82–1.3) | 73.52 | 0 | ||||||
| Asians | 12 | 1394/1429/1745 | 1.06 (0.91–1.23) | 68.48 | 0 | ||||||
| T1DM | 6 | 815/575/1024 | 1.01 (0.78–1.31) | 77.09 | 0 | ||||||
| T2DM | 15 | 2158/2047/2644 | 1.15 (0.92–1.44) | 89.81 | 0 | ||||||
| Healthy cases | |||||||||||
| ACE | All | I > D | 30 | 3690/4927 | 1.24 (1.02–1.52) | 83.2 | 0 | ||||
| All in HWE | 29 | 3283/4695 | 1.26 (1.02–1.55) | 82.87 | 0 | ||||||
| Caucasians | 10 | 1051/1565 | 1.14 (0.92–1.43) | 52.63 | 0.03 | ||||||
| Asians | 18 | 1925/2514 | 1.17 (0.96–1.41) | 64.06 | 0 | ||||||
| T1DM | 8 | 908/1314 | 1.1 (0.84–1.43) | 60.08 | 0.01 | ||||||
| T2DM | 22 | 2782/3613 | 1.29 (1–1.66) | 85.58 | 0 | ||||||
| Diseased controls–cases | |||||||||||
| NOS3 | All | 4 b > a | – | 17 | 3887/3196 | 1.09 (0.96–1.23) | 24.04 | 0.18 | |||
| All in HWE | 16 | 3824/3130 | 1.04 (0.94–1.15) | 0 | 0.73 | ||||||
| Caucasians | 7 | 2147/1659 | 1.19 (0.96–1.46) | 50.17 | 0.06 | ||||||
| Asians | 8 | 1035/939 | 1.04 (0.85–1.28) | 0 | 0.51 | ||||||
| T1DM | 5 | 1600/1209 | 1.24 (0.91–1.68) | 65.81 | 0.02 | ||||||
| T2DM | 12 | 2287/1987 | 1.04 (0.91–1.18) | 0 | 0.62 | ||||||
| Diseased controls–cases | |||||||||||
| NOS3 | All | G894T | rs1799983 | 19 | 4184/3330 | 1.19 (0.96–1.48) | 79.95 | <0.001 | |||
| All in HWE | 17 | 3952/3120 | 1.25 (1.00–1.56) | 79.95 | <0.001 | ||||||
| Caucasians | 4 | 1778/1105 | 0.93 (0.81–1.06) | 0 | 0.69 | ||||||
| Asians | 11 | 1586/1548 | 1.32 (0.90–1.94) | 83.08 | 0 | ||||||
| T1DM | 3 | 1404/870 | 0.89 (0.76–1.03) | 0 | 1 | ||||||
| T2DM | 16 | 2780/2460 | 1.28 (0.99–1.70) | 80.64 | 0 | ||||||
| Diseased controls–cases | |||||||||||
| MTHFR | All | C677T | 26 | 4246/4380 | 1.43 (1.16–1.77) | 83.91 | 0 | ||||
| All in HWE | 21 | 3163/3085 | 1.28 (1.03–1.59) | 78.12 | 0 | ||||||
| Caucasians | 9 | 1198/1202 | 1.2 (0.91–1.59) | 66.8 | 0 | ||||||
| Asians | 15 | 2853/2811 | 1.31 (1.05–1.64) | 77.38 | <0.001 | ||||||
| T1DM | 4 | 372/667 | 1.31 (0.88–1.95) | 64.75 | 0.03 | ||||||
| T2DM | 22 | 3874/3713 | 1.45 (1.14–1.84) | 85.58 | 0 | ||||||
| Three groups | |||||||||||
| MTHFR | All | C677T | 10 | 1345/1306/1903 | 1.44 (1.13–1.83) | 84.45 | 0 | ||||
| All in HWE | 6 | 1.14 (0.98–1.32) | 26.51 | 0.24 | |||||||
| Caucasians | 5 | 518/595/895 | 1.09 (0.95–1.25) | 64.84 | 0.37 | ||||||
| Asians | 4 | 734/444/608 | 1.76 (1.38–2.25) | 55.63 | 0.08 | ||||||
| T1DM | 2 | 153/238/447 | 1.3 (1.04–1.61) | 0 | 0.7 | ||||||
| T2DM | 8 | 1192/1068/1456 | 1.47 (1.09–1.98) | 87.47 | 0 | ||||||
| Healthy controls–cases | |||||||||||
| MTHFR | All | C677T | 13 | 1507/2503 | 1.68 (1.15–2.47) | 88.17 | 0 | ||||
| All in HWE | 7 | 1.21 (0.86–1.7) | 69.13 | 0 | |||||||
| Caucasians | 6 | 585/1167 | 1.31 (1.02–1.69) | 45.04 | 0.11 | ||||||
| Asians | 6 | 829/936 | 1.63 (0.97–2.73) | 82.88 | 0 | ||||||
| T1DM | 3 | 174/647 | 1.71 (1.24–2.37) | 83.06 | 0.34 | ||||||
| T2DM | 10 | 1333/1856 | 1.67 (1.03–2.71) | 90.93 | 0 | ||||||
| Diseased controls–cases | |||||||||||
| AGT | All | M235T | rs699 | 26 | 5015/5253 | 1.21 (1.01–1.45) | 82.45 | 0.00 | |||
| All in HWE | 19 | 3181/3655 | 1.09 (0.92–1.31) | 72.76 | 0.00 | ||||||
| Caucasians | 15 | 2875/2726 | 1.04 (0.94–1.16) | 14.79 | 0.29 | ||||||
| Asians | 1628/2122 | 1.37 (0.86–2.20) | 90.17 | <0.001 | |||||||
| T1DM | 10 | 1478/1482 | 1.12 (0.95–1.32) | 34.37 | 0.13 | ||||||
| T2DM | 15 | 2802/3220 | 1.31 (0.99–1.72) | 86.84 | <0.001 | ||||||
| Three groups | |||||||||||
| AGT | All | M235T | rs699 | 9 | 1156/1350/1243 | 1.21 (1.02–1.45) | 65.91 | 0.02 | |||
| Caucasians | 4 | 827/835/725 | 1.15 (1.03–1.3) | 0 | 0.94 | ||||||
| Asians | 5 | 329/515/518 | 1.23 (0.82–1.85) | 80.41 | 0 | ||||||
| T1DM | 3 | 517/482/495 | 1.19 (1.02–1.38) | 0 | 0.99 | ||||||
| T2DM | 6 | 639/868/748 | 1.21 (0.9–1.64) | 78.5 | 0 | ||||||
| Healthy cases | |||||||||||
| AGT | All | M235T | rs699 | 12 | 1399/1759 | 1.17 (0.87–1.58) | 76.47 | 0 | |||
| Caucasians | 6 | 941/1125 | 1.09 (0.88–1.36) | 43.62 | 0.11 | ||||||
| Asians | 6 | 458/634 | 1.38 (0.66–2.9) | 86.09 | 0 | ||||||
| T1DM | 4 | 538/695 | 1.1 (0.78–1.54) | 56.02 | 0.08 | ||||||
| T2DM | 8 | 861/1064 | 1.26 (0.79–2.01) | 82.46 | 0 | ||||||
| Diseased controls–cases | |||||||||||
| AGTR1 | All | A1166C | rs5186 | 24 | 6000/5020 | 1.05 (0.93–1.18) | 49.13 | 0.003 | |||
| All in HWE | 21 | 5260/4296 | 0.97 (0.87–1.08) | 49.13 | 0.13 | ||||||
| Caucasians | 13 | 3915/2993 | 0.94 (0.85–1.03) | 10.64 | 0.34 | ||||||
| Asians | 11 | 2085/2027 | 1.35 (1.14–1.61) | 21.14 | 0.24 | ||||||
| T1DM | 13 | 3915/2993 | 0.94 (0.85–1.03) | 10.64 | 0.34 | ||||||
| T2DM | 11 | 2202/2145 | 1.26 (1.04–1.53) | 40.46 | 0.07 | ||||||
| Diseased controls–cases | |||||||||||
| APOE | All | E3 > e2/e4 | 14 | 1770/1586 | 1.08 (0.85–1.37) | 57.41 | 0.00 | ||||
| All in HWE | 13 | 1341/1170 | 1.07 (0.81–1.41) | 60.62 | 0.00 | ||||||
| Caucasians | 5 | 589/481 | 1.15 (0.79–1.67) | 48.44 | 0.1 | ||||||
| Asians | 6 | 1018/897 | 1.15 (0.89–1.50) | 35.62 | 0.17 | ||||||
| T1DM | 9 | 760/633 | 1.06 (0.69–1.62) | 68.5 | 0 | ||||||
| T2DM | |||||||||||
| Diseased controls–cases | |||||||||||
| TGFB1 | All | T869C | rs1800470 | 11 | 2408/2452 | 1.16 (0.94–1.44) | 75.49 | 0 | |||
| All in HWE | 8 | 1930/1952 | 1.08 (0.84–1.38) | 77.06 | 0 | ||||||
| Caucasians | 5 | 1.05 (0.74–1.49) | 82.75 | 0 | |||||||
| Asians | 4 | 1.37 (1.05–1.79) | 56 | 0.08 | |||||||
| T1DM | 4 | 0.86 (0.74–1) | 0 | 0.43 | |||||||
| T2DM | 7 | 1.38 (1.09–1.75) | 64.49 | 0.01 | |||||||
RS: SNP identifier, RE ORG: random effects odds ratio generalized, I2: I2 statistic, PQ: P-value from heterogeneity testing.
The A1166C polymorphism of the AGTR1 gene, while not statistically significant in the main analysis, was revealed as significant in the subgroup analysis of the Asians and in the subgroup analysis of T2DM, as well as polymorphism T869C of the TGFB1 gene in the case of genotypic analysis in diabetics versus cases with DN where in the main analysis there was no significance, in the subgroup analyses of Asians and T2DM was statistically significant.
DISCUSSION
To the best of our knowledge, the present systematic review and meta-analysis are the most comprehensive in the field of DN as it examines for the first time such a large number of genes and genetic polymorphisms, and in particular synthesizes data of 228 genes (606 polymorphisms). Sixty-six genetic polymorphisms have been shown to be associated with DN. The aforementioned polymorphisms are harboured in 51 genes, while one is not close to any known gene, however this does not mean that it is of secondary importance. The specific polymorphisms belong to the following 51 genetic loci: ACACB, ACE, ADIPOQ, AGT, AGTR1, AKR1B1, APOC1, APOE, ATP1B2, ATP2A3, CARS, CCR5, CGNL1, CNDP1, CYGB-PRCD, EDN1, ELMO1, ENPP1, EPO, FLT4, FTO, GLO1, HMGA2, IGF2/INS/TH cluster, IL1B, IL8, IL10, KCNQ1, KNG, LOC101927627, MTHFR, NOS3, SDC2, SETD7, SIRT1, SLC2A1, SLC2A2, SLC12A3, SLC19A3, TCF7L2, TGFB1, TIMP1, TTC39C, UNC13B, VEGFA, WTAPP1, WWC1 as well as XYLT1 and three intergenic polymorphisms. Of the above genes, KNG and SETD7 genes proved marginally statistically significant.
The present systematic review and meta-analysis confirmed the statistical significance of ACE, AKR1B1, APOC1, APOE, CARS, CCR5, CNDP1, ELMO1, EPO, NOS3, UNC13B and VEGFA, which showed significance in meta-analyses [35, 36]. However, the present meta-analysis provides novel statistical significance in or near ADRB3, ATP1B2, ATP2A3, CGNL1, CYGB-PRCD, FLT4, LOC101927627, TIMP1, TTC39C, WTAPP1 and WWC1 genetic loci. Some of the novel findings are harboured in cytogenetic locations, which were revealed significant in a meta-analysis of genome-wide linkage studies, such as locations on chromosome 5 [25].
The pathway analysis of significant genes revealed overrepresentation of six signalling pathways: cytokine–cytokine receptor interaction, pyruvate metabolism, T2DM, adipocytokine signalling pathway, renal cell carcinoma and the renin–angiotensin system.
Taking into consideration the aforementioned pathways and the classification of genes based on biological role indicated the important role played by the renin–angiotensin system [37], angiogenesis and erythropoiesis [38], lipid metabolism [39], polyol pathway [40], inflammatory mechanisms [36, 41], oxidative stress [42], endothelium function [43] and extracellular matrix degradation [44], as well as epigenetic mechanisms [45] and glucose transport [46, 47]. Functional studies remain necessary to confirm the involvement of these mechanisms, in order to clarify the exact role of these polymorphisms and pathways in DN. From the genetic polymorphisms of the FRMD3, CARS, ELMO1, CPVL and CHN2 genes, which were first detected in large genetic correlation studies, only CARS and ELMO1 genes remained statistically significant after the meta-analysis. The exact role of these genes requires further elucidation, but certainly many of the polymorphisms that GWAS have demonstrated will prove to be not really responsible.
Subgroup analyses were performed purely in order to detect statistical significance for a particular polymorphism in a particular environment, either T1DM or T2DM, or Caucasians or Asians. One finding that is worth mentioning is that genes that have been statistically significant in subgroup analyses in the Caucasians have also emerged as significant in subgroup analyses of T1DM (MTHFR and AGT), while significant genes in sub-analysis of Asians were significant and in subgroup analyses by T2DM (ACE, AGTR1, MTHFR and TGFB1). This pattern of significance between Caucasians and T1DM, as well as between Asians and T2DM, should be investigated in further studies.
A difference of the present meta-analysis compared with other meta-analyses is the fact that three types of comparisons were made: healthy controls versus cases with DN, diabetic controls versus cases with DN, as well as healthy controls versus diabetic controls versus cases with DN, in order to separate genes whose aggravating role is independent of the presence of diabetic milieu. This information may highlight unknown aspects of the pathophysiology of the disease in the hope of leading the scientific community to discover new therapies that target pathways common to both T1DM and T2DM nephropathy. The field synopsis has some limitations, one of which is the publication bias. In the present field synopsis, we included only English articles published in scientific journals. In addition, in most meta-analyses, a small number of studies were included, so the results should be interpreted with caution. The majority of included studies also had insufficient statistical power in order to identify modest genetic effect which is believed to result from common variants.
The studies included in the meta-analyses differed with regard to the ethnicity of the individuals, the type of diabetes and the clinical phenotype regarding the presence or not of persistent albuminuria. Other studies included subjects with persistent proteinuria, and some included cases with persistent microalbuminuria. However, no study involving exclusively cases with not persistent microalbuminuria was included, so that there is no underestimation of the genetic effect since not persistent albuminuria is a potentially reversible condition. For this reason, the model of random effects was used in the meta-analysis in which the variability of the determinant result is due to both the variability of each study due to the fact that samples are used rather than source populations and variability between different studies. In order to take into account heterogeneity due to different origins and types of diabetes, the relevant subgroup analyses were performed.
The identification of the genetic polymorphisms that contribute to the genetic predisposition of the disease will lead to the discovery of new therapies, but also to more valid and effective means of prevention and prognosis. These discoveries will provide new perspectives in the field of personalized medicine based on the genetic background of each patient and will allow the provision of both preventative and therapeutic interventions at the individual genome level, increasing both the efficacy and the safety of therapies. It is very logical to investigate polymorphisms that have been shown to be statistically significant in association with DN, but agnostic studies that examine the genome as a whole are the most promising.
CONCLUSIONS
Sixty-six genetic polymorphisms have been shown to be associated with DN. These polymorphisms are harboured in 53 gene loc,i while one is not close to any known gene. These results should be interpreted with caution because the true susceptibility loci could be loci that are in linkage disequilibrium with the significant loci. It would be useful in the future to examine haplotypes, study microarray data and perform functional studies for clarification of the role of significant genes to produce newer data, and their combination may lead to the creation of the genetic map of DN.
CONFLICT OF INTEREST STATEMENT
The authors declare that there is no duality of interest associated with this manuscript.
Supplementary Material
REFERENCES
- 1. Thomas MC, Brownlee M, Susztak K. et al. Diabetic kidney disease. Nat Rev Dis Prim 2015; 1: 15070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Dronavalli S, Duka I, Bakris GL.. The pathogenesis of diabetic nephropathy. Nat Rev Endocrinol 2008; 4: 444–452 [DOI] [PubMed] [Google Scholar]
- 3. Thomas MC, Groop PH, Tryggvason K.. Towards understanding the inherited susceptibility for nephropathy in diabetes. Curr Opin Nephrol Hypertens 2012; 21: 195–202 [DOI] [PubMed] [Google Scholar]
- 4. Rich SS. Genetics of diabetes and its complications. J Am Soc Nephrol 2006; 17: 353–360 [DOI] [PubMed] [Google Scholar]
- 5. Seaquist ER, Goetz FC, Rich S. et al. Familial clustering of diabetic kidney disease. N Engl J Med 1989; 320: 1161–1165 [DOI] [PubMed] [Google Scholar]
- 6. Cowie CC, Port FK, Wolfe RA. et al. Disparities in incidence of diabetic end-stage renal disease according to race and typeof diabetes. N Engl J Med 1989; 321: 1074–1079 [DOI] [PubMed] [Google Scholar]
- 7. Pettitt DJ, Saad MF, Bennett PH. et al. Familial predisposition to renal disease in two generations of Pima Indians with type 2 (non-insulin-dependent) diabetes mellitus. Diabetologia 1990; 33: 438–443 [DOI] [PubMed] [Google Scholar]
- 8. Freedman BI, Tuttle AB, Spray BJ.. Familial predisposition to nephropathy in African-Americans with non-insulin-dependent diabetes mellitus. Am J Kidney Dis 1995; 25: 710–713 [DOI] [PubMed] [Google Scholar]
- 9. Quinn M, Angelico MC, Warram JH. et al. Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 1996; 39: 940–945 [DOI] [PubMed] [Google Scholar]
- 10. Genovese G, Friedman DJ, Ross MD. et al. Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science 2010; 329: 841–845 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Hirschhorn JN, Daly MJ.. Genome-wide association studies for common diseases and complex traits. Nat Rev Genet 2005; 6: 95–108 [DOI] [PubMed] [Google Scholar]
- 12. Klein RJ, Zeiss C, Chew EY. et al. Complement factor H polymorphism in age-related macular degeneration. Science 2005; 308: 385–389 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Shimazaki A, Kawamura Y, Kanazawa A. et al. Genetic variations in the gene encoding ELMO1 are associated with susceptibility to diabetic nephropathy. Diabetes 2005; 54: 1171–1178 [DOI] [PubMed] [Google Scholar]
- 14. Maeda S, Osawa N, Hayashi T. et al. Genetic variations associated with diabetic nephropathy and type II diabetes in a Japanese population. Kidney Int 2007; 72: S43–S48 [DOI] [PubMed] [Google Scholar]
- 15. Pezzolesi MG, Poznik GD, Mychaleckyj JC. et al. ; the DCCT/EDIC Research Group. Genome-wide association scan for diabetic nephropathy susceptibility genes in type 1 diabetes. Diabetes 2009; 58: 1403–1410 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Craig DW, Millis MP, DiStefano JK.. Genome‐wide SNP genotyping study using pooled DNA to identify candidate markers mediating susceptibility to end‐stage renal disease attributed to Type 1 diabetes. Diabet Med 2009; 26: 1090–1098 [DOI] [PubMed] [Google Scholar]
- 17. McDonough CW, Palmer ND, Hicks PJ. et al. A genome-wide association study for diabetic nephropathy genes in African Americans. Kidney Int 2011; 79: 563–572 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Germain M, Pezzolesi MG, Sandholm N. et al. SORBS1 gene, a new candidate for diabetic nephropathy: results from a multi-stage genome-wide association study in patients with type 1 diabetes. Diabetologia 2015; 58: 543–548 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Iyengar SK, Sedor JR, Freedman BI. et al. ; Family Investigation of Nephropathy and Diabetes (FIND). Genome-wide association and trans-ethnic meta-analysis for advanced diabetic kidney disease: family investigation of nephropathy and diabetes (FIND). PLoS Genet 2015; 11: e1005352. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Sandholm N, Salem RM, McKnight AJ. et al. ; DCCT/EDIC Research Group. New susceptibility loci associated with kidney disease in type 1 diabetes. PLoS Genet 2012; 8: e1002921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Van Zuydam NR, Ahlqvist E, Sandholm N. et al. ; Finnish Diabetic Nephropathy Study (FinnDiane). A genome-wide association study of diabetic kidney disease in subjects with type 2 diabetes. Diabetes 2018; 67: 1414–1427 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Taira M, Imamura M, Takahashi A. et al. ; The SUMMIT Consortium. A variant within the FTO confers susceptibility to diabetic nephropathy in Japanese patients with type 2 diabetes. PLoS One 2018; 13: e0208654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Guan M, Keaton JM, Dimitrov L. et al. ; FIND Consortium. Genome-wide association study identifies novel loci for type 2 diabetes-attributed end-stage kidney disease in African Americans. Hum Genomics 2019; 13: 21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Jeong KH, Kim JS, Woo JT. et al. ; Representing KNOW-CKD Study Group and KNDP Study Group. Genome-wide association study identifies new susceptibility loci for diabetic nephropathy in Korean patients with type 2 diabetes mellitus. Clin Genet 2019; 96: 35–42 [DOI] [PubMed] [Google Scholar]
- 25. Tziastoudi M, Stefanidis I, Stravodimos K. et al. ; Family Investigation of Nephropathy and Diabetes (FIND) Consortium. Identification of chromosomal regions linked to diabetic nephropathy: a meta-analysis of genome-wide linkage scans. Genet Test Mol Biomark 2019; 23: 105–117 [DOI] [PubMed] [Google Scholar]
- 26. Rao M, Mottl AK, Cole SA. et al. Meta-analysis of genome-wide linkage scans for renal function traits. Nephrol Dial Transplant 2012; 27: 647–656 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Zintzaras E. The power of generalized odds ratio in assessing association in genetic studies with known mode of inheritance. J Appl Statist 2012; 39: 2569–2581 [Google Scholar]
- 28. Zintzaras E. The generalized odds ratio as a measure of genetic risk effect in the analysis and meta-analysis of association studies Statistical Applications in Genetics and Molecular Biology. Vol. 9. Berlin, Germany: de Gruyter, 2010 [DOI] [PubMed] [Google Scholar]
- 29. DerSimonian R, Laird N.. Meta-analysis in clinical trials. Control Clin Trial 1986; 7: 177–188 [DOI] [PubMed] [Google Scholar]
- 30. Cochran WG. The combination of estimates from different experiments. Biometrics 1954; 10: 101–129 [Google Scholar]
- 31. Higgins JPT, Thompson SG.. Quantifying heterogeneity in ametaanalysis. Statist Med 2002; 21: 1539–1558 [DOI] [PubMed] [Google Scholar]
- 32. Egger M, Smith GD, Schneider M. et al. Bias in meta-analysis detected by a simple, graphical test. Bmj 1997; 315: 629–634 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33. Huang DW, Sherman BT, Lempicki RA.. Systematic and integrative analysis of large gene lists using DAVID Bioinformatics Resources. Nat Protoc 2009; 4: 44–57 [DOI] [PubMed] [Google Scholar]
- 34. Huang DW, Sherman BT, Lempicki RA.. Bioinformatics enrichment tools: paths toward the comprehensive functional analysis of large gene lists. Nucleic Acids Res 2009; 37: 1–13 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35. Mooyaart AL, Valk EJJ, Van Es LA. et al. Genetic associations in diabetic nephropathy: a meta-analysis. Diabetologia 2011; 54: 544–553 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36. Tziastoudi M, Stefanidis I, Hadjigeorgiou GM. et al. A systematic review and meta-analysis of genetic association studies for the role of inflammation and the immune system in diabetic nephropathy. Clin Kidney J 2017; 10: 293–300 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Roscioni SS, Heerspink HJL, De Zeeuw D.. The effect of RAAS blockade on the progression of diabetic nephropathy. Nat Rev Nephrol 2014; 10: 77–87 [DOI] [PubMed] [Google Scholar]
- 38. Tanaka T, Nangaku M.. Angiogenesis and hypoxia in the kidney. Nat Rev Nephrol 2013; 9: 211–222 [DOI] [PubMed] [Google Scholar]
- 39. Rutledge JC, Ng KF, Aung HH. et al. Role of triglyceride-rich lipoproteins in diabetic nephropathy. Nat Rev Nephrol 2010; 6: 361–370 [DOI] [PubMed] [Google Scholar]
- 40. He J, Gao HX, Yang N. et al. The aldose reductase inhibitor epalrestat exerts nephritic protection on diabetic nephropathy in db/db mice through metabolic modulation. Acta Pharmacol Sin 2019; 40: 86–97 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41. Navarro-González JF, Mora-Fernández C, De Fuentes MM. et al. Inflammatory molecules and pathways in the pathogenesis of diabetic nephropathy. Nat Rev Nephrol 2011; 7: 327–340 [DOI] [PubMed] [Google Scholar]
- 42. Singh DK, Winocour P, Farrington K.. Oxidative stress in early diabetic nephropathy: fueling the fire. Nat Rev Endocrinol 2011; 7: 176–184 [DOI] [PubMed] [Google Scholar]
- 43. Nakagawa T, Tanabe K, Croker BP. et al. Endothelial dysfunction as a potential contributor in diabetic nephropathy. Nat Rev Nephrol 2011; 7: 36–44 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44. Kure M, Pezzolesi MG, Poznik GD. et al. Genetic variation in the matrix metalloproteinase genes and diabetic nephropathy in type 1 diabetes. Mol Genet Metab 2011; 103: 60–65 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Kato M, Natarajan R.. Diabetic nephropathy—emerging epigenetic mechanisms. Nat Rev Nephrol 2014; 10: 517–530 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46. Zintzaras E, Stefanidis I.. Association between the GLUT1 gene polymorphism and the risk of diabetic nephropathy: a meta-analysis. J Hum Genet 2005; 50: 84–91 [DOI] [PubMed] [Google Scholar]
- 47. Stefanidis I, Tziastoudi M, Tsironi EE. et al. The contribution of genetic variants of SLC2A1 gene in T2DM and T2DM-nephropathy: association study and meta-analysis. Ren Fail 2018; 40: 561–576 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.