Association between Angiotensin-Converting

Association between Angiotensin-Converting Enzyme Gene Polymorphisms and Diabetic Nephropathy: Case-Control, Haplotype, and Family-Based Study in Three European Populations Samy Hadjadj,* Lise Tarnow,† Carol Forsblom,‡ Gbenga Kazeem,§ Michel Marre,储 Per-Henrik Groop,‡ Hans-Henrik Parving,† François Cambien,¶ David A. Tregouet,¶ Ivo G. Gut,** Alexandre The´va,¶ Dominique Gauguier,§ Martin Farrall,§ Roger Cox,†† Fumihiko Matsuda,** Mark Lathrop,** FinnDiane Study Group, and Nathalie Hager-Vionnet;¶ for the EURAGEDIC (European Rational Approach for the Genetics of Diabetic Complications) Study Group *CHU de Poitiers, Department of Diabetology, and INSERM ERM 324, Poitiers University Hospital, Poitiers, France; † Steno Diabetes Center, Copenhagen, Denmark; ‡Folkha¨lsan Institute of Genetics, Folkha¨lsan Research Center, Biomedicum Helsinki, and Helsinki University Central Hospital, Department of Medicine, Division of Nephrology, Helsinki, Finland; §Wellcome Trust Centre for Human Genetics, Oxford, United Kingdom; 储INSERM U695, Xavier Bichat University of Medicine, and Department of Diabetology, Bichat Hospital, Paris, France; ¶INSERM and Universite´ Pierre et Marie Curie-Paris 6, UMR S 525, Paris, France; **Centre National du Genotypage, Evry, France; and ††Mammalian Research Council, Oxford, United Kingdom Angiotensin 1-converting enzyme gene (ACE) is a risk factor for diabetic nephropathy (DN) in patients with type 1 diabetes. The selection of this candidate gene is supported by cross-sectional and follow-up studies, but no convincing family-based studies are available. Recruited were 1057 patients (with DN: persistent albuminuria with or without renal failure) and 1127 control subjects (long-standing [>15 yr] normoalbuminuric patients with type 1 diabetes) in Denmark, Finland, and France and 532 family trios that were composed of 244 trios with DN probands and 288 trios with non-DN probands. Five ACE polymorphisms were studied. In the case-control analysis, the rs1800764-C, rs4311-T, Insertion/deletion (I/D or rs1799752)-D, rs4366-G, and rs12449782-G alleles were associated with an increased risk for DN, homogeneously across populations, with allelic odds ratios of 1.11 (95% confidence interval 1.00 to 1.22), 1.18 (1.04 to 1.33), 1.13 (1.02 to 1.23), 1.10 (0.99 to 1.20), and 1.12 (1.01 to 1.23), respectively. Haplotype analysis further demonstrated that the haplotype defined by the D, rs4366_G and rs12449782_G alleles was associated with a greater risk for DN. Even though no significant allelic overtransmission to DN or non-DN probands was detected, the family-based study provided consistent results with the case-control analysis. In a large case-control study, it was shown that the ACE polymorphisms were associated with DN; these findings were not confirmed in a family-based association study. This study population is suitable to search for additional candidate genes for DN. J Am Soc Nephrol 18: 1284 –1291, 2007. doi: 10.1681/ASN.2006101102

D

iabetic nephropathy (DN) is associated with a risk for end-stage renal failure and accounts for the reduced life expectancy of patients with type 1 diabetes (1). Although almost all patients with type 1 diabetes are affected by diabetic retinopathy over time, only 25 to 35% of these patients develop DN (2–5). Chronic hyperglycemia cannot fully account for the pathogenesis of renal complications. Genetic factors play an important role, as shown by familial studies (6 – 8).

Received October 10, 2006. Accepted February 2, 2007. Published online ahead of print. Publication date available at www.jasn.org. Address correspondence to: Dr. Samy Hadjadj, Department of Diabetology, Poitiers University Hospital, BP 577, 86021 Poitiers Cedex, France. France: ⫹335-49-44-39-00; Fax: ⫹33-5-49-44-40-06; E-mail: [email protected] Copyright © 2007 by the American Society of Nephrology

Angiotensin-converting enzyme gene (ACE) is a risk factor for DN. Its plasma levels have been reported to be associated with DN but not with diabetic retinopathy in type 1 diabetes patients (9). ACE modulates the generation of angiotensin II, which increases intraglomerular hydraulic pressure (10), leading to glomerulopathy. ACE inhibition strongly modifies renal hemodynamics in animals (11), and the course of DN can be considerably improved by treatment with ACE inhibitors, in patients with type 1 diabetes (12,13). Plasma ACE concentrations are stable in individuals (14) and are partly under genetic control (15). The ACE gene is located on chromosome 17, at q23, and contains 26 exons that span a total of 21 kb, and ACE levels are controlled primarily by the ACE region that maps to an 18-kb interval that is flanked by ISSN: 1046-6673/1804-1284

J Am Soc Nephrol 18: 1284 –1291, 2007

ACE Genotype and Nephropathy in Patients with Type 1 Diabetes

two intragenic ancestral recombination breakpoints (16). An insertion/deletion (I/D; rs1799752) polymorphism in intron 16 of ACE accounts for a large proportion of interindividual variation for ACE serum concentrations (17). Several case-control studies have examined the association between the ACE I/D polymorphism and DN. As shown in a recent meta-analysis, a deleterious effect of the D allele was evidenced (18). Longitudinal follow-up studies recently provided further evidence for the deleterious effect of the D allele (19 –21). However, one family-based study discordantly showed that the I allele was overtransmitted in cases of DN (22). We aimed to investigate the role of ACE in DN, using two complementary approaches—a case-control study and a family-based study— on a large population that was recruited from three European countries. In addition to the I/D variant, which is widely investigated in studies of the genetics of DN, we chose to study four single-nucleotide polymorphisms (SNP; rs1800764, rs4311 T, rs4366, and rs1244978) because they were shown to characterize the haplotypic structure of ACE (Supplementary Figure 1) in European populations (16). Indeed, the genetic structure of ACE is made up of three ancestral regions. The chosen SNP were selected in these three regions because they tagged the haplotypic structure of European individuals, according to a previous quantitative trait locus study (16).

Materials and Methods Patient Populations Three European coordinating centers in Denmark, Finland, and France contributed to the case-control and trio studies. The recruitment characteristics and the general research strategy are described elsewhere (EURAGEDIC 1; Tarnow et al., submitted for publication). The patients participated in a case-control study (1057 case patients and 1127 control subjects). A complementary family-based study included 244 case patients and 288 control subjects as probands (532 family-based patients) with both parents recruited or, if missing, at least one parent and a sibling to infer probabilities for the missing parental genotype. In the Finnish and Danish populations, both parents were recruited in the studied pedigrees. The majority of pedigrees from the French population are made of trios with both parents recruited. We had the opportunity to infer the genotype of the missing parent in 16 pedigrees (three with missing mothers and 13 with missing fathers). Case patients had established DN— defined as persistent albuminuria (⬎300 mg/24 h or ⬎200 ␮g/min or ⬎200 mg/L) in two of three consecutive measurements on sterile urine samples—with or without renal failure (serum creatinine ⬎150 ␮mol/L). Control subjects with type 1 diabetes had long-standing diabetes (disease duration ⱖ15 yr) and persistently normal albumin excretion rate (⬍30 mg/24 h or ⬍20 ␮g/min or ⬍20 mg/L) in two of three consecutive sterile urine samples) and were not treated with ACE inhibitors or angiotensin receptor blockers. Microalbuminuric patients were not retained because they could not be surely classified as case patients or control subjects.

Genotyping Genomic DNA was isolated from human leukocytes by standard methods. Genotyping was carried out with various automated highthroughput methods (see Supplementary Table 1). Each polymorphism was determined in the same way for all of the populations studied. The ID polymorphism in ACE was determined by PCR with the ACE45 (PF)

1285

and ACE46 (PR) primers (16), with subsequent separation of the fragments by electrophoresis on 2% agarose gels. The rs4311 and rs4366 SNP were determined by matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry of the primer extension products that were generated by PCR, as described previously (23). The rs12449782 and rs1800764 SNP were determined by the TaqMan method, using the Applied Biosystems protocol with end-point reading with an ABI 7009HT. The genomic control markers were characterized using the SNPlex genotyping technology (Applied Biosystems, Courtaboeuf, France). Quality control for genotyping was ensured by testing for Hardy Weinberg (HW) equilibrium on every 384-well plate and by genotyping 192 replicates.

Statistical Analyses Allele frequencies were calculated by the gene-counting method, and deviation from HW equilibrium was tested by means of a ␹2 test with one degree of freedom. In each population, association between studied SNP and DN was assessed by standard logistic regression analysis, with adjustment for gender, smoking status, diabetes duration, and glycosylated hemoglobin concentration. The Mantel-Haenszel statistic was then used to investigate possible heterogeneity across populations and, when appropriate, to combine the results that were obtained for each population into a weighted analysis that provided an overall estimate of the allelic odds ratio (OR). Pairwise linkage disequilibrium (LD) between polymorphisms was estimated and expressed in terms of the D⬘ and r2 statistics, as implemented in THESIAS software (24). THESIAS software was also used for haplotype analysis of the 5 ACE polymorphisms with respect to DN. In each population, we compared the haplotype frequency distributions of case patients and control subjects by means of a likelihood ratio test (␹2 with m ⫺ 1 degrees of freedom for m haplotypes), and haplotype effects (95% confidence interval [CI]) were expressed as haplotypic OR, assuming additive effects on a logistic scale. For trio data analysis, a family-based test of association (in the presence of linkage) of the pedigree data was carried out for each population to account for possible differences in allele frequencies and associations with DN among the three populations. The transmission of alleles to affected (case patients) or unaffected (control subjects) patients with type 1 diabetes was analyzed using the transmission disequilibrium test (TDT). TRANSMIT software, which incorporates cases of uncertain transmission (25), was used for the analysis. The proportion T of “overtransmitted” or “high-risk” alleles from informative (i.e., heterozygous) parents was estimated by counting informative transmissions. For each polymorphism, the results from the three populations were combined to obtain an overall estimate of the effect that was associated with each polymorphism.

Combination of Results from Case-Control and Trio Studies For the trio data, the OR for each marker was estimated from the transmission counts (T) that were obtained from TDT analysis, as follows: ORtdt ⫽ T/(1 ⫺ T) (26). The SE can be easily obtained because the logarithm of the OR is asymptotically normally distributed (27). These OR were then combined using a weighted analysis method (27). Statistical significance was set at 0.05.

Results Genotyping success rates of between 90 and 98.4% were obtained and no mismatches were detected among the 192 replicate samples. All polymorphisms were in HW equilibrium. Each pair of polymorphisms was in linkage disequilibrium (see Supplementary Table 2). The main clinical and biologic char-

1286

Journal of the American Society of Nephrology

J Am Soc Nephrol 18: 1284 –1291, 2007

acteristics of the patients with type 1 diabetes are described in Table 1.

Case-Control Study The allele frequencies of the five ACE polymorphisms are given in Table 2. In the French population, all of the polymorphisms were associated with DN, with an OR for the risk allele ranging from 1.26 (95% CI 1.06 to 1.42) to 1.35 (95% CI 1.08 to 1.72) for the rs4366C and rs4311T variants, respectively. Even though these associations with DN failed to reach statistical significance in the Danes and Finns, the same trends were observed as in the French population. Consistently, the OR across the three populations were not found to be significantly heterogeneous for any of the five markers studied. In the combined population, the rs4311-T allele, the ACE_ID D allele, and the rs12449782-G allele were significantly associated with the risk for DN (OR 1.18 [95% CI 1.04 to 1.33; P ⫽ 0.01], 1.13 [95% CI 1.02 to 1.23; P ⫽ 0.024], and 1.12 [95% CI 1.01 to 1.23; P ⫽ 0.039], respectively). Some patients (298 Danish and 417 French) were already analyzed in previous reports on ACE (28,29). To limit the impact of already analyzed people, we restricted the study to newly recruited patients. It had no major effect on the statistical significance of the alleles that were associated with DN risk (rs4311, P ⫽ 0.023; ACE_ID, P ⫽ 0.055; and rs4366, P ⫽ 0.086). The association of the 94 genomic control markers was compatible with expectations of null hypothesis of no association (data not shown), which indicates that stratification within one or more of the populations is unlikely to cause positive association results. Multivariate adjustment for clinical factors showed that the adjusted OR (Table 3) were similar to those for the nonadjusted data presented in Table 2.

Haplotype Analysis As a consequence of the strong LD among the five polymorphisms (see Supplementary Table 2), four common haplotypes with a frequency ⬎5% could be inferred, accounting for approximately 90% of all chromosomes (Table 4). The haplotype structure that was derived from these five poly-

morphisms was highly homogeneous across populations. The I allele and the rs 4366 C and rs12449782A alleles were in very strong LD with each other and defined a unique (⫺)ICA haplotype. Conversely, the DGG haplotype was subsequently split into three haplotypes according to the rs1800764 and rs4311 polymorphisms: (CC)DGG, (CT)DGG, and (TT)DGG. Although the association tests for a global haplotypic effect were not statistically significant, haplotype analysis confirmed the results that were obtained in univariate analysis: The TCICA haplotype was less frequent in case patients than in control subjects. Consequently, the (⫺)DGG haplotypes tended to be more frequent in case patients than in control subjects, being then associated with a greater risk for DN than the (⫺)ICA haplotype. However, because of this pattern of LD, we were unable to identify which of the I/D, rs4366 and rs12449782 polymorphisms was per se associated with DN.

Family-Based Study Nine families were excluded a priori because of familial inconsistencies. In total, 288 trios with non-nephropathic probands and 244 trios with nephropathic probands were studied. The results of the TDT analysis of the pedigree of nephropathic probands are presented in Table 5. No specific allele was significantly overtransmitted, and no heterogeneity was observed across the three populations. Similar results were obtained when probands that were already included in previous reports were excluded. If we considered only non-nephropathic trios, no significant allelic overtransmission was detected for the ACE polymorphisms (data not shown). Haplotype analyses using TRANSMIT identified similar haplotype structure and frequencies as given in the casecontrol samples. No evidence was found for significant overtransmission of common haplotypes (frequency higher than 2%) to DN.

Combined Analysis The combined results of case-control and trio studies are presented in Table 5. No heterogeneity was found between data

Table 1. Clinical and biological characteristics of patients with type 1 diabetes according to DNa

Case-control patients Family-based patients Country of origin (DK/FI/FR; n) Gender (male/female; n) Age (yr) Diabetes duration (yr) BMI (kg/m²) SBP (mmHg) DBP (mmHg) HbA1c (%) a

Cases (n ⫽ 1301)

Controls (n ⫽ 1415)

P

1057 244 524/442/335 757/544 42.0 ⫾ 10.2 28.4 ⫾ 8.7 24.4 ⫾ 3.6 145 ⫾ 21 83 ⫾ 11 9.0 ⫾ 1.6

1127 288 466/616/333 671/744 44.8 ⫾ 11.0 29.2 ⫾ 9.0 24.6 ⫾ 3.3 132 ⫾ 17 76 ⫾ 9 8.3 ⫾ 1.2

— — ⬍0.001 ⬍0.001 ⬍0.001 ⬍0.001 0.15 ⬍0.001 ⬍0.001 ⬍0.007

BMI, body mass index; DBP, diastolic BP; DK, Denmark; DN, diabetic nephropathy; FI, Finland; FR, France; HbA1c, glycosylated hemoglobin; SBP, systolic BP.


1287

0.09 0.29 0.11 0.13 0.12

from the case-control and family-based studies. All of the risk alleles were associated with an increase of approximately 15% in the risk for DN (Table 6).

0.47 0.45 0.54 0.51 0.55

1.30 (1.10 to 1.46) 1.35 (1.08 to 1.72) 1.30 (1.11 to 1.45) 1.26 (1.07 to 1.42) 1.29 (1.10 to 1.44)

1.11 (1.00 to 1.22) 1.18 (1.04 to 1.33) 1.13 (1.02 to 1.23) 1.10 (0.99 to 1.20) 1.12 (1.01 to 1.23)

0.060 0.010 0.024 0.076 0.039

Discussion

b

a

ACE, angiotensin-converting enzyme gene; CI, confidence interval; OR, odds ratio. Allele 2 was considered to be the risk allele. c Allelic frequency of allele 2. d P value for test of homogeneity of OR across the three populations.

0.56 0.52 0.63 0.59 0.63 1.04 (0.83 to 1.22) 1.18 (0.96 to 1.43) 1.09 (0.90 to 1.25) 1.09 (0.89 to 1.25) 1.09 (0.89 to 1.26) 0.48 0.48 0.56 0.55 0.57 0.49 0.52 0.58 0.57 0.59 1.03 (0.81 to 1.21) 1.05 (0.86 to 1.30) 1.04 (0.82 to 1.21) 0.99 (0.76 to 1.17) 1.01 (0.78 to 1.20) 0.45 0.48 0.51 0.52 0.51 T/C C/T I/D C/G A/G rs 1800764 rs 4311 ID rs 4366 rs 12449782

0.46 0.49 0.52 0.51 0.52

Heterogeneity Test (P)d P OR (95% CI) OR (95% CI)

Casesc Controlsc (n ⫽ 385) (n ⫽ 468)

OR (95% CI)

Casesc Controlsc (n ⫽ 277) (n ⫽ 273)

OR (95% CI)

All Populations Pooled France Finland Denmark

Controlsc Casesc (n ⫽ 380) (n ⫽ 382) Allele (1/2)b ACE Marker

Table 2. Association between genetic variants of the ACE gene and DN: Case-control analysis in each study population and in the pooled study populationa

J Am Soc Nephrol 18: 1284 –1291, 2007

In this international, large-scale, multicenter study from three European populations, we combined a family-based approach and a case-control analysis to analyze the role of several polymorphisms in ACE on DN. We found that DN in patients with type 1 diabetes was associated with the studied polymorphisms in this gene. This association was not limited to the ACE I/D polymorphism. Univariate and haplotype analysis suggested that this association was mainly related to the haplotype that carries the ACE_ID D allele. In the European Rational Approach for the Genetics of Diabetic Complications (EURAGEDIC) program, we used a research strategy that consists of a candidate gene approach with a case-control design combined with a familial transmission analysis. This strategy to analyze trios with DN probands but also trios with non-DN probands was recently presented as relevant for diabetic kidney disease (30). In the studied populations, the risk for any founder effect is small as a result of the international recruitment of patients. No heterogeneity was found across the three populations, one of which came from Finland, a country that often is considered a genetic isolate. In addition, a genomic control allowed us to rule out stratification bias. The overall results for the ACE genetic variants tested were positive, although some variation between populations was possible. No statistically significant heterogeneity was detected, but it should be noted that most of the positive results were obtained for the French population, with no contribution from the Finnish population. Despite the detection of a positive association in the casecontrol analysis, we found no association with ACE in the family-based study. Unlike Krolewski (22), we observed no overtransmission of the ACE I allele in patients with type 1 diabetes and nephropathy. The recruitment of the trios must be discussed. As reported elsewhere (EURAGEDIC 1; Tarnow et al., submitted for publication), the probands who had type 1 diabetes and for whom the parents were also recruited differed in clinical characteristics relating to age and diabetes duration from patients who had type 1 diabetes and for whom the parents were not recruited. Because DN in probands was associated with a high risk for premature mortality and cardiovascular disease in both the patients and their parents, the parents of case patients were less likely to be alive than the parents of control subjects. However, although this recruitment bias must be recognized, it is unlikely to have led to spurious results. We recruited more than 500 trios, corresponding to one of the largest family-based studies of DN in patients with type 1 diabetes. However, only informative trios can be considered in transmission analysis, and transmission was analyzed separately in the 244 trios with DN and the 288 trios without DN. In addition, the power calculation, assuming an additive effect, shows that we would need 526 trios to detect with 85% power significant associations (P ⱕ 0.05) with an OR of 1.3 for a gene

1288


J Am Soc Nephrol 18: 1284 –1291, 2007

Table 3. Logistic regression analysis of the association between ACE genetic polymorphisms and DNa ACE Marker

rs 1800764 rs 4311 ID rs 4366 rs 12449782

Denmark

Allele (1/2)b

T/C C/T I/D C/G A/G

Finland

France

OR

95% CI

OR

95% CI

OR

95% CI

1.05 1.07 1.06 1.02 1.09

0.83 to 1.33 0.84 to 1.36 0.84 to 1.34 0.81 to 1.28 0.86 to 1.38

1.04 1.14 1.09 1.12 1.14

0.83 to 1.31 0.91 to 1.42 0.89 to 1.35 0.90 to 1.38 0.91 to 1.44

1.35 1.28 1.36 1.30 1.34

1.03 to 1.75 1.00 to 1.65 1.05 to 1.75 1.01 to 1.68 1.04 to 1.73

a

The clinical and biological factors that were included in the logistic regression were gender, smoking, diabetes duration, and HbA1c. b Allele 2 was considered to be the risk allele.

with minor allelic frequency of 0.45. The number rises to 3980 for an OR at 1.1. It is interesting that the combination of cross-sectional and trio analyses gave similar results to the case-control study. However, it must be recognized that the relative sample size of trios compared with case-control samples leads to a relatively small contribution of trios to the combined statistics. The results of our study are not consistent with those for the family-based study reported by Krolewski (22), which supported a deleterious effect of the I allele. No combined analysis of case-control and family-based studies (27) has been reported before regarding ACE. This new statistical tool confirmed the results that were obtained elsewhere (18,19). Our data for the I/D polymorphism are consistent with the results of cross-sectional studies, analyzed in a recent meta-

analysis (18), and with three prospective follow-up studies (19 –21). We found that the ACE D allele is a risk factor for DN. This is at variance with one of the first published studies on the relationship between the ACE I/D polymorphism and DN (28). However, in this later report, glycosylated hemoglobin was significantly lower in control subjects than in case patients. In addition, diabetic retinopathy was less frequently encountered in control subjects than in case patients. Altogether, this suggests that the glycemic exposure in control subjects was not sufficient, leading to an unbalanced matching. The inclusion of a large-scale recruitment from different populations allowed us to correct this probable lack of statistical power and recruitment bias. The ID polymorphism can be considered a “reference” polymorphism because it has been determined in the majority of the studies in this area. The magnitude of the effect of this poly-

Table 4. Association between ACE genetic polymorphisms and DN: Haplotype analysis of ACE in each study population Polymorphism Population C/Ta

C/Tb

D/I

G/Cc

A/Gd

T

C

I

C

A

C

C

D

G

G

C

T

D

G

G

T

T

D

G

G

Likelihood ratio test with 3 df

a

rs 1800764. rs 4311. c rs 4366. d rs 12449782. b

Haplotype Frequency Controls

Denmark Finland France Denmark Finland France Denmark Finland France Denmark Finland France Denmark Finland France

0.411 0.377 0.386 0.047 0.081 0.059 0.364 0.347 0.317 0.092 0.094 0.116

Haplotypic OR (95% CI)

Cases

0.382 Reference 0.351 Reference 0.300 Reference 0.056 1.22 (0.72 to 2.08) 0.073 1.00 (0.64 to 1.55) 0.066 1.30 (0.75 to 2.26) 0.347 0.96 (0.74 to 1.24) 0.376 1.20 (0.93 to 1.56) 0.389 1.44 (1.08 to 1.93) 0.106 1.16 (0.78 to 1.74) 0.113 1.31 (0.89 to 1.91) 0.114 1.16 (0.75 to 1.81) ␹2 ⫽ 1.38, P ⫽ 0.71 ␹2 ⫽ 3.30, P ⫽ 0.35 ␹2 ⫽ 8.26, P ⫽ 0.04

J Am Soc Nephrol 18: 1284 –1291, 2007


1289

Table 5. Transmission disequilibrium test for five genetic variants of ACE in trios with DNa Denmark ACE Markers

Allele (1/2)

N

Tb

OR

Finland 95% CI

rs 1800764 T/C 117.18 0.560 1.27 0.88 to 1.82 rs 4311 C/T 102.02 0.555 1.25 0.85 to 1.85 ID I/D 93.02 0.554 1.25 0.83 to 1.89 rs 4366 C/G 65.32 0.558 1.26 0.77 to 2.04 rs 12449782 A/G 89.05 0.517 1.06 0.70 to 1.61

France

Pooled

N

Tb

OR

95% CI

N

Tb

OR

95% CI

OR

95% CI

Heterogeneity Test (P)c

64.26 69.00 63.00 58.02 62.00

0.457 0.493 0.508 0.495 0.474

0.84 0.97 1.03 0.98 0.90

0.51 to 1.37 0.61 to 1.56 0.62 to 1.69 0.59 to 1.64 0.55 to 1.49

39.67 45.26 44.37 42.01 43.24

0.509 0.534 0.500 0.541 0.546

1.04 1.15 1.00 1.18 1.20

0.56 to 1.92 0.64 to 2.06 0.56 to 1.79 0.64 to 2.17 0.66 to 2.17

1.09 1.13 1.12 1.14 1.04

0.83 to 1.41 0.87 to 1.47 0.85 to 1.47 0.84 to 1.54 0.78 to 1.38

0.414 0.711 0.768 0.769 0.755

a

N is the estimated number of informative transmissions (in terms of heterozygous parent transmitting to an affected child), and T is the proportion of transmitted allele to diseased offspring (probands with DN) in transmission disequilibrium test analysis. mc⬎OR ⫽ T/(1 ⫺ T) is an estimate of the allelic odds ratio under the assumption of additive effect (on a logistic scale). Combined OR across populations according to Mantel-Haenszel statistic. b Transmission and OR of allele 2 (considered to be the risk allele). c P value for test of homogeneity of OR across the three populations.

morphism must be interpreted with caution. The relative risk for DN for the ACE I/D polymorphism was approximately 1.30 in the recent meta-analysis (18) and a large-scale long-term prospective study (19). This value is higher than the combined OR of 1.13 reported here. However, we considered transmission of the D allele, whereas most studies compared the genotype of the patients (18,19), with II genotype patients as the reference group. Because the ACE gene is long, with its 26 exons spanning 21 kb, we restricted our study to five genetic markers near the 18-kb region that are strongly associated with ACE concentration (16). This limitation should be recognized because we concentrated on this region of interest rather than analyze the whole gene. Because genetic variants in the vicinity of the critical 18-kb region have been shown to influence ACE concentrations (16), which determine the progression of DN (20), we compared the effects of the whole haplotype and of the single ACE_ID variant on the risk for DN. Haplotype analysis showed that ⫺DGG was associated with increased risk for DN, compared with the ⫺ICA reference haplotype. However, even if the D allele was associated with DN in many publications, our data did not allow us to identify it clearly as the putative risk allele. Because of the strong LD, the association could also be due to the G alleles of the rs4366 and rs 12449782 polymorphism. Accordingly, we speculate that

the polymorphisms in the 5⬘ region are unlikely to be responsible for the predisposition to DN. In the Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) cohort, the rs 1800764 polymorphism was also analyzed (19). In the genotype analysis, CT patients had a higher risk for developing microalbuminuria, compared with TT patients. In the haplotype analysis, patients who carried the T allele were at lower risk for persistent microalbuminuria or severe nephropathy. We were not able to replicate this result, because we found no effect of the polymorphism in the 5⬘ region. However, in the univariate analysis, only heterozygous patients with a CT genotype had a higher risk for severe nephropathy, suggesting a genotype rather than an allele effect. This discordant result might be due to the different geographic origin of the two study populations and also to the design (case-control in our study and longitudinal prospective study for the DCCT/EDIC). Our finding is at variance with a recently published study in which a risk haplotype in ACE was identified, even though the risk haplotype also carried the ID_ D allele (31). However, even when studying polymorphisms located in the same key regions of ACE, we did not select the same SNP. In addition, the authors studied American patients with type 2 diabetes, with control subjects having a relatively short diabetes duration, whereas we studied European patients with long-term type 1

Table 6. Association between ACE genetic polymorphisms and DN: Results of the combined analysis (case-control and family-based studies) ACE Allele Polymorphism (1/2)a

rs 1800764 rs 4311 ID rs 4366 rs 12449782 a

T/C C/T I/D C/G A/G

Case-Control OR (95% CI)

Trio OR (95% CI)

Combined OR (95% CI)

1.11 (1.00 to 1.22) 1.18 (1.04 to 1.33) 1.13 (1.02 to 1.23) 1.10 (0.99 to 1.20) 1.12 (1.01 to 1.23)

1.09 (0.83 to 1.41) 1.13 (0.87 to 1.47) 1.12 (0.85 to 1.47) 1.14 (0.84 to 1.54) 1.04 (0.78 to 1.38)

1.11 (1.00 to 1.22) 1.17 (1.04 to 1.31) 1.13 (1.03 to 1.24) 1.10 (1.00 to 1.22) 1.11 (1.01 to 1.22)

Allele 2 was considered to be the risk allele. P value for test of homogeneity of OR across case-control and trio analyses.

b

P Combined Heterogeneity OR Test (P)b

0.040 0.007 0.013 0.051 0.035

0.99 0.96 0.99 0.98 0.89

1290


diabetes. These points can possibly account for the discrepant findings.

Conclusion Our data, obtained using two complementary approaches, confirm the role of ACE in DN risk. This large-scale study produced interesting results with very narrow CI. These results for the EURAGEDIC study may increase interest in future candidate gene analyses for DN.

Acknowledgments The EURAGEDIC study was made possible through funding from the European Commission through contract QLG2-CT-2001-01669. All type 1 diabetes patients and their family are sincerely acknowledged. The Danish Diabetes Association and the Sehested Hansen Foundation are thanked for their continued support to this study. The French Diabetes Association is thanked for its support to this study (bourse AFD recherche 2000). The list of participating centers in France and Belgium is available in references (29,32). A. Boland and A. Lemainque (CNG) are acknowledged for handling DNA bank and for microsatellite and Taqman genotyping; V. Laneuze and C. Demer are thanked for secretarial assistance. Oluf Pedersen (Steno Diabetes Center, Gentofte, Denmark) and Ludovic Drouet (Hoˆpital Lariboisie`re, Paris, France) are kindly acknowledged for providing nondiabetic individuals from population-based studies. The FinnDiane study is supported by grants from the Folkha¨lsan Research Foundation, Samfundet Folkha¨lsan, the Wilhelm and Else Stockmann Foundation, the Liv och Ha¨lsa Foundation, and the Finnish Medical Society (Finska La¨karesa¨llskapet). We acknowledge all of the physicians and nurses at each center participating in the collection of patients ( † indicates deceased person). Anjalankoski Health Center: S. Koivula, T. Uggeldahl; Central Finland Central Hospital: T. Forslund, A. Halonen, A. Koistinen, P. Koskiaho, M. Laukkanen, J. Saltevo, M. Tiihonen; Central Hospital of Aland Islands: A.-C. Blomqvist, M. Forsen, H. Granlund, B. Nyroos; Central Hospital of Kanta-Hame: P. Kinnunen, A. Orvola, T. Salonen, A. Va¨ha¨nen; Central Hospital of Kymenlaakso: R. Paldanius, M. Riihela¨, L. Ryysy; Central Hospital of Lansi-Pohja: P. Nyla¨nden, A. Sademies; Central Ostrobothnian Hospital District: S. Anderson, B. Asplund, U. Byskata, T. Virkkala; City of Espoo Health Center: (Espoonlahti): A. Nikkola, E. Ritola; (Leppa¨vaara): L. Penttinen, H.-L. Siekkinen; City of Tampere Health Center: P. Alarotu, S. Gummerus, A. Kallio, H. Kirkkopelto-Jokinen, E. Kujansuu, M. Kuortti, T. Maijala, T. Niskanen, T. Saaristo, A. Vadeń; City of Vantaa Health Center: (Korso): R. Toivonen, H. Virtanen; (Martinlaakso): M. Laine, T. Pellonpaä¨, R. Puranen; (Myyrma¨ki): A. Airas, J. Laakso, K. Rautavaara; (Rekola): M. Erola, E. Jatkola; (Tikkurila): R. Lo¨nnblad, A. Malm, J. Ma¨kela¨, E. Rautamo; Helsinki University Central Hospital (Department of Medicine, Division of Nephrology): C.-G. af Bjo¨rkesten, J. Fagerudd, S. Fro¨jdo¨, R. Kilpikari, S. Lindh, M. Parkkonen, K. Pettersson-Fermholm, M. Rosengård-Ba¨rlund, H. Rosvall, M. Ro¨nnback, R. Sallinen, A. Sandelin, M. Saraheimo, L .Sjo¨lind, L. Thorn, T. Vesisenaho, J. Wadeń; Iisalmi Hospital: E. Toivanen; Jokilaakso Hospital, Ja¨msa¨: A. Parta, I. Pirttiniemi; Jorvi Hospital, Espoo: S. Aranko, S. Ervasti, R. Kauppinen-Ma¨kelin, A. Kuusisto, T. Leppa¨la¨, K. Nikkila¨, L. Pekkonen; Kainuu Central Hospital: S. Jokelainen, P. Kemppainen, A.-M. Mankinen, M. Sankari; Kerava Health Center: H. Stuckey, P. Suominen; Kivela¨ Hospital, Helsinki: A. Aimolahti, E. Huovinen; Koskela Hospital, Helsinki: V. Ilkka, M. Lehtima¨ki; Kouvola Health Center: E. Koskinen, T. Siitonen; Kuopio University Hospital: E. Huttunen,

J Am Soc Nephrol 18: 1284 –1291, 2007

R. Ika¨heimo, P. Karhapaä¨, P. Keka¨laïnen, M. Laakso, T. Lakka, E. Lampainen, L. Mykka¨nen, L. Niskanen, U. Tuovinen, I. Vauhkonen, E. Voutilainen; Kuusamo Health Center: E. Isopoussu, H. Suvanto, E. Vierimaa; Kuusankoski Hospital: E. Kilkki, L. Riihela¨; Laakso Hospital, Helsinki: T. Merilaïnen, P. Poukka, R. Savolainen, N. Uhlenius; Lahti City Hospital: A. Ma¨kela¨, M. Tanner; Lapland Central Hospital: L. Hyva¨rinen, S. Severinkangas, T. Tulokas; Lappeenranta Health Center: P. Linkola, I. Pulli; Lohja Hospital: T. Granlund, M. Saari, T. Salonen; La¨nsi-Uusimaa Hospital, Tammisaari: I.-M. Jousmaa, J. Rinne; Malmi Hospital, Helsinki: H. Lanki, S. Moilanen, M. Tilly-Kiesi; Mikkeli Central Hospital: A. Gynther, R. Manninen, P. Nironen, M. Salminen, T. Va¨nttinen; North Karelian Hospital: U.-M. Henttula, A. Rissanen, H. Turtola, M. Voutilainen; Oulaskangas Hospital, Oulainen: E. Jokelainen, P.-L. Jylkka¨, E. Kaarlela, J. Vuolaspuro; Oulu Health Center: L. Hiltunen, R. Ha¨kkinen, S. Keina¨nen-Kiukaanniemi; Païja¨t-Ha¨me Central Hospital: H. Haapama¨ki, A. Helantera¨, H. Miettinen; Palokka Health Center: (Palokka): P. Sopanen, L. Welling; (Vaajakoski): K. Ma¨kinen, P. Sopanen; Pietarsaari Hospital: M.-L. Holmba¨ck, B. Isomaa, L. Sarelin; Pori City Hospital: P. Ahonen, P. Merensalo, K. Sa¨vela¨; Porvoo Hospital: M. Kallio, B. Rask, S. Ra¨mo¨; Rauma Hospital: K. Laine, K. Saarinen, T. Salminen; Riihima¨ki Hospital: E. Immonen, L. Juurinen; Salo Hospital: A. Alanko, J. Lapinleimu, P. Rautio, M. Virtanen; Satakunta Central Hospital: M. Juhola, P. Kunelius, M.-L. Lahdenma¨ki, P. Paä¨kko¨nen, M. Rautavirta; Savonlinna Central Hospital: T. Pulli, P. Sallinen, H. Valtonen, A. Vartia; Seinajoki Central Hospital: E. Korpi-Hyo¨va¨lti, T. Latvala, E. Leijala; South Karelia Hospital District: T. Hotti, R. Ha¨rko¨nen, U. Nyholm, R. Vanamo;† Tampere University Hospital: I. Ala-Houhala, T. Kuningas, P. Lampinen, M. Maä¨tta¨, H. Oksala, T. Oksanen, K. Salonen, H. Tauriainen, S. Tulokas; Turku Health Center: I. Ha¨ma¨laïnen, H. Virtamo, M. Va¨ha¨talo; Turku University Central Hospital: M. Asola, K. Breitholz, R. Eskola, K. ´ yra¨paä¨; Vammala Metsa¨rinne, U. Pietila¨, P .Saarinen, R. Tuominen, S. A Hospital: I. Isoma¨ki, R. Kroneld, M. Tapiolinna-Ma¨kela¨; Vasa Central Hospital: S. Bergkulla, U. Hautama¨ki, V.-A. Myllyniemi, I. Rusk.

Disclosures None.

References 1. Mogensen CE: Definition of diabetic renal disease in insulin-dependent diabetes mellitus based on renal function test. In: The Kidney and Hypertension in Diabetes Mellitus, 4th Ed., edited by Mogensen CE, Dordrecht, The Netherlands, Kluwer Acadmic Publishers, 1998, pp. 17–29 2. Orchard TJ, Dorman JS, Maser RE, Becker DJ, Drash AL, Ellis D, LaPorte RE, Kuller LH: Prevalence in complications in IDDM by sex and duration. Pittsburgh epidemiology of diabetes complications study II. Diabetes 39: 1116 –1124, 1990 3. Microalbuminuria Collaborative Study Group: Microalbuminuria in type I diabetic patients. Prevalence and clinical characteristics. Diabetes Care 15: 495–501, 1992 4. Hovind P, Tarnow L, Rossing P, Jensen BR, Graae M, Torp I, Binder C, Parving HH: Predictors for the development of microalbuminuria and macroalbuminuria in patients with type 1 diabetes: Inception cohort study. BMJ 328: 1105, 2004 5. Chaturvedi N, Bandinelli S, Mangili R, Penno G, Rottiers RE, Fuller JH: Microalbuminuria in type 1 diabetes: Rates, risk factors and glycemic threshold. Kidney Int 60: 219 –227, 2001 6. Borch-Johnsen K, Norgaard K, Hommel E, Mathiesen ER, Jensen JS, Deckert T, Parving HH: Is diabetic nephropathy an inherited complication? Kidney Int 41: 719 –722, 1992

J Am Soc Nephrol 18: 1284 –1291, 2007


7. Quinn M, Angelico MC, Warram JH, Krolewski AS: Familial factors determine the development of diabetic nephropathy in patients with IDDM. Diabetologia 39: 940 –945, 1996 8. Seaquist ER, Goetz FC, Rich S, Barbosa J: Familial clustering of diabetic kidney disease. Evidence for genetic susceptibility to diabetic nephropathy. N Engl J Med 320: 1161–1165, 1989 9. Marre M, Bernadet P, Gallois Y, Savagner F, Guyene TT, Hallab M, Cambien F, Passa P, Alhenc-Gelas F: Relationships between angiotensin I converting enzyme gene polymorphism, plasma levels, and diabetic retinal and renal complications. Diabetes 43: 384 –388, 1994 10. Hall JE, Guyton AC, Jackson TE, Coleman TG, Lohmeier TE, Tripoddo NC: Control of glomerular filtration rate by reninangiotensin system. Am J Physiol 233: F366 –F372, 1977 11. Zatz R, Dunn BR, Meyer TW, Anderson S, Rennke HG, Brenner BM: Prevention of diabetic glomerulopathy by pharmacological amelioration of glomerular capillary hypertension. J Clin Invest 77: 1925–1930, 1986 12. Should all patients with type 1 diabetes mellitus and microalbuminuria receive angiotensin-converting enzyme inhibitors? A meta-analysis of individual patient data. Ann Intern Med 134: 370 –379, 2001 13. Lewis E, Hunsicker L, Bain R, Rhode R: The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. N Engl J Med 329: 1456 –1462, 1993 14. Alhenc-Gelas F, Richard J, Courbon D, Warnet JM, Corvol P: Distribution of plasma angiotensin I-converting enzyme levels in healthy men: Relationship to environmental and hormonal parameters. J Lab Clin Med 117: 33–39, 1991 15. Cambien F, Alhenc-Gelas F, Herbeth B, Andre JL, Rakotovao R, Gonzales MF, Allegrini J, Bloch C: Familial resemblance of plasma angiotensin-converting enzyme level; the Nancy Study. Am J Hum Genet 43: 774 –780, 1988 16. Soubrier F, Martin S, Alonso A, Visvikis S, Tiret L, Matsuda F, Lathrop GM, Farrall M: High-resolution genetic mapping of the ACE-linked QTL influencing circulating ACE activity. Eur J Hum Genet 10: 553–561, 2002 17. Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, Soubrier F: An insertion deletion polymorphism in angiotensin I conversion enzyme accounting for half the variance of serum enzyme levels. J Clin Invest 86: 1343–1346, 1990 18. Ng DP, Tai BC, Koh D, Tan KW, Chia KS: Angiotensin-I converting enzyme insertion/deletion polymorphism and its association with diabetic nephropathy: A meta-analysis of studies reported between 1994 and 2004 and comprising 14,727 subjects. Diabetologia 48: 1008 –1016, 2005 19. Boright AP, Paterson AD, Mirea L, Bull SB, Mowjoodi A, Scherer SW, Zinman B: Genetic variation at the ACE gene is associated with persistent microalbuminuria and severe nephropathy in type 1 diabetes: The DCCT/EDIC Genetics Study. Diabetes 54: 1238 –1244, 2005 20. Hadjadj S, Belloum R, Bouhanick B, Gallois Y, Guilloteau G, Chatellier G, Alhenc-Gelas F, Marre M: Prognostic value of angiotensin-I converting enzyme I/D polymorphism for

21.

22. 23.

24.

25.

26.

27. 28.

29.

30.

31.

32.

1291

nephropathy in type 1 diabetes mellitus: A prospective study. J Am Soc Nephrol 12: 541–549, 2001 Rudberg S, Rasmussen LM, Bangstad H-J, Osterby R: Influence of insertion/deletion polymorphism in the ACE-I gene on the progression of diabetic glomerulopathy in type 1 diabetic patients with microalbuminuria. Diabetes Care 23: 544 –548, 2000 Krolewski AS: Genetics of diabetic nephropathy: Evidence for major and minor gene effects. Kidney Int 55: 1582–1596, 1999 Sauer S, Lechner D, Berlin K, Lehrach H, Escary JL, Fox N, Gut IG: A novel procedure for efficient genotyping of single nucleotide polymorphisms. Nucleic Acids Res 28: E13, 2000 Tregouet DA, Escolano S, Tiret L, Mallet A, Golmard JL: A new algorithm for haplotype-based association analysis: The Stochastic-EM algorithm. Ann Hum Genet 68: 165–177, 2004 Clayton D: A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission. Am J Hum Genet 65: 1170 –1177, 1999 Mitchell LE: Relationship between case-control studies and the transmission/disequilibrium test. Genet Epidemiol 19: 193–201, 2000 Kazeem GR, Farrall M: Integrating case-control and TDT studies. Ann Hum Genet 69: 329 –335, 2005 Tarnow L, Cambien F, Rossing P, Nielsen FS, Hansen L, Lecerf O, Poirier S, Danilov S, Parving H-H: Lack of relationship between an insertion/deletion polymorphism in the angiotensin I-converting enzyme gene and diabetic nephropathy and proliferative retinopathy in IDDM patients. Diabetes 44: 489 – 494, 1995 Marre M, Jeunemaitre X, Gallois Y, Rodier M, Chatellier G, Sert C, Dusselier L, Kahal Z, Chaillous L, Halimi S, Muller A, Sackmann H, Bauduceau B, Bled F, Passa P, AlhencGelas F: Contribution of genetic polymorphism in the renin-angiotensin system to the development of renal complications in insulin-dependent diabetes: Genetique de la Nephropathie Diabetique (GENEDIAB) study group. J Clin Invest 99: 1585–1595, 1997 Mueller PW, Rogus JJ, Cleary PA, Zhao Y, Smiles AM, Steffes MW, Bucksa J, Gibson TB, Cordovado SK, Krolewski AS, Nierras CR, Warram JH: Genetics of Kidneys in Diabetes (GoKinD) study: A genetics collection available for identifying genetic susceptibility factors for diabetic nephropathy in type 1 diabetes. J Am Soc Nephrol 17: 1782–1790, 2006 Ng DP, Placha G, Choo S, Chia KS, Warram JH, Krolewski AS: A Disease haplotype for advanced nephropathy in type 2 diabetes at the ACE locus. Diabetes 55: 2660–2663, 2006 Hadjadj S, Pean F, Gallois Y, Passa P, Aubert R, Weekers L, Rigalleau V, Bauduceau B, Bekherraz A, Roussel R, Dussol B, Rodier M, Marechaud R, Lefebvre PJ, Marre M: Different patterns of insulin resistance in relatives of type 1 diabetic patients with retinopathy or nephropathy: The Genesis France-Belgium Study. Diabetes Care 27: 2661–2668, 2004

Supplemental information for this article is available online at http://www.jasn.org/.

Access to UpToDate on-line is available for additional clinical information at http://www.jasn.org/