Nephrology Dialysis Transplantation

Transcription

1 Nephrol Dial Transplant (1998) 13: Original Article Nephrology Dialysis Transplantation DNA polymorphisms in the ACE gene, serum ACE activity and the risk of nephropathy in insulin-dependent diabetes mellitus Maria Beatriz S. Freire1,3, David J. van Dijk2, Arie Erman2, Geoffrey Boner2, James H. Warram1 and Andrzej S. Krolewski1 1Research Division, Joslin Diabetes Centre and Department of Medicine, Harvard Medical School Boston, Massachusetts, USA, 2Institute of Hypertension and Kidney Diseases, Rabin Medical Centre, Beilinson Medical Centre, Petach Tikwa and Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel, and 3Currently at Faculdade de Medicina de Jundiai, Jundiai, Brazil Abstract examined extensively, with the angiotensin I converting Background. To determine the relationship between enzyme (ACE) gene being the favourite candidate gene DNA polymorphisms in the angiotensin I converting for nephropathy [5,6 ]. This has occurred for several enzyme (ACE) gene, serum ACE activity and the risk reasons. First, it was reported a long time ago that of diabetic nephropathy. IDDM patients with diabetic nephropathy have elev- Methods. A case-control study was carried out in a ated serum levels of ACE [7 9]. Second, carriers of population of Jewish insulin-dependent diabetes mel- certain allele(s) in the ACE gene were found to have litus (IDDM ) patients. Cases (77 IDDM patients with high levels of serum ACE [10]. The same individuals diabetic nephropathy) and controls (89 IDDM patients were found in some studies to be at increased risk of with normoalbuminuria) were genotyped with PCR diabetic nephropathy [11,12]. Third, clinical trials have protocols for detecting two DNA polymorphisms in demonstrated the effectiveness of ACE inhibition in the ACE gene: one in intron 7 detected with the retarding the progression of diabetic nephropathy restriction enzyme PstI and the other in intron 16 [13,14]. identified as an insertion/deletion ( I/D). ACE is an ectoenzyme anchored to the plasma Results. The risk of nephropathy was increased only membrane with the bulk of its mass exposed on the in patients homozygous for the allele with the PstI extracellular surface of endothelial, epithelial and neursite. These homozygotes had a nephropathy risk that onal cells [15,16]. It cleaves the carboxyl-terminal was 2.3 times (95% C.I.: ) that of the other dipeptide from angiotensin I to produce angiotensin genotypes. Furthermore, these individuals did not have II, and inactivates bradykinin (the most favoured elevated serum ACE activity. substrate) by the sequential removal of two carboxyl- Conclusions. The results of this study are evidence that terminal dipeptides [15,16 ]. Cloning of ACE cdna the risk of diabetic nephropathy in IDDM is influenced revealed that the enzyme is composed of two homologby genetic variability at the ACE locus, but the responsous domains: the amino and carboxyl domain [17]. ible variant is not the I/D polymorphism in intron 16. Each of these domains contains enzymatically active Our findings require further studies in other sites that catalyse physiological substrates at slightly populations. different rates [16]. ACE is encoded by a single gene located on the long arm of chromosome 17 [17,18]. The gene has 26 exons: exons 1 12 encode for the amino domain and exons Introduction encode for the carboxyl domain. Several different DNA polymorphisms have been identified in this Only a subset of insulin-dependent diabetes mellitus gene [10,19,20]. A frequent insertion/deletion ( I/D) ( IDDM) patients (about 30%) is susceptible to diabetic polymorphism, located in intron 16 has been the one nephropathy [1], and familial clustering of this com- most studied [10]. The D allele is associated with plication points to genetic factors as the determinants higher plasma levels of ACE than the I allele, and the of susceptibility [2 4]. So far, genes encoding for D allele also seems to be associated with increased proteins of the renin-angiotensin system have been myocardial infarction risk in diabetic and non-diabetic individuals [10,21 23]. It is not clear, however, that Correspondence and offprint requests to: Andrzej S. Krolewski, MD, PhD, Joslin Diabetes Centre, One Joslin Place, Boston, MA 02215, the increased risk of myocardial infarction is related USA. to the elevated plasma ACE levels [24]. In some 1998 European Renal Association European Dialysis and Transplant Association

2 2554 M. B. S. Freire et al. populations, the association of the D allele with between 30 and 300 mg/24 h were considered microalbuminmyocardial infarction has not been confirmed [25]. uria, and values 300 mg/24 h were considered overt pro- The role of the I/D polymorphism at the ACE gene teinuria. Individuals were considered cases if the AER was has also been evaluated in the development of kidney 30 mg/24 h in at least two of the three 24-h collections; all others were considered as controls. diseases. The presence of the D allele predisposes to progression of IgA nephropathy [ 26,27], but findings with regard to diabetic nephropathy have been contradictory. Whereas Marre et al. [11,12] reported that the Determination of serum ACE activity I/D polymorphism was associated with the developand centrifuged at 2000 g for 10 min at 4 C. Serum samples Blood samples were collected in non-heparinized glass tubes ment of diabetic nephropathy, several other studies were frozen and assayed within 2 weeks. Serum ACE activity did not confirm this finding [28,29]. In our research in was assayed by a radiometric method using [3H]hippuryl- Boston, while we found an effect of sequence differ- glycyl-glycine as substrate, as described previously [9]. ences at the ACE locus on the risk of diabetic nephro- Briefly, serum samples (10 ml ) were incubated with 800 nmol pathy, the effect was associated with a polymorphism substrate ( c.p.m., mg) for 60 min at 37 C in in intron 7 of the gene rather than the I/D in intron HEPES buffer (0.05 mol/l, ph 8.0) in a final volume of 16 [30]. 0.1 ml. The reaction was terminated with 1.0 ml of HCl (0.1 The present study was undertaken to study the mol/l ); the medium was extracted with 1.5 ml ethyl acetate, relationship between polymorphisms in the ACE gene centrifuged and the radiolabelled hippuric acid present in the in a Jewish population of IDDM patients and to organic phase was counted by liquid scintillation spec- examine whether the association between the specific trometry. ACE activity was expressed as nmol/ml/min. Interassay CV of 6.1% and 4.2% were obtained for serum polymorphisms in the ACE gene and diabetic nephroactivity of 88 and 184 nmol/ml/min, respectively (n=15), pathy can be accounted for by elevated serum ACE and 5% for intra-assay CV. activities. Seven patients who had nephropathy and were taking an ACE inhibitor were excluded from the statistical analyses related to ACE activity and blood pressure. Research design and methods DNA analysis Study population DNA was extracted from leukocytes by the phenol chloro- Patients selected for this study included IDDM patients form technique according to standard protocols. Each patient attending the outpatient clinic at the Rabin Medical Centre, had genotypes determined at two loci of the ACE gene one Beilinson Campus, a tertiary referral centre in Petach Tikwa, in the intron 7 (amino domain) and the other in the intron Israel. This group of patients includes some of the patients 16 (carboxyl domain). with IDDM described previously [31]. For the present study A DdeI Polymorphism in intron 7 of the ACE gene was only patients with at least 10 years duration of IDDM were first identified by us as a three allele system (+,,=) with selected. The study protocol was approved by the Human the denaturing gradient gel electrophoresis (DGGE) blotting Subjects Committee at the Rabin Medical Centre. technique [33]. Molecular characterization demonstrated that an A to G substitution in intron 7 of the + and = alleles Patient examination gives them a PstI restriction site which is absent in the allele [26 ]. Therefore, the three allele DdeI polymorphism Each patient had a physical examination performed that can be reduced to a two allele system that can be genotyped included height, weight and blood pressure measurements. with a PCR-based protocol. Flanking primers con- A standardized questionnaire was used to obtain the medical taining DNA sequences from intron 7 and exon 8 of history and demographic information. In addition retinopathe gene (5 -CTCGGCTTGGGACTTCTA-3, upstream; thy status was assessed by reviewing medical records. 5- AGAGCTGGTCCATCGTGA-3, downstream) produce Proliferative retinopathy was diagnosed if a patient had laser 800 bp fragments which, after PstI digestion, permit recognitreatment or had proliferative retinopathy diagnosed by a tion of two alleles + (which includes the minor DdeI allele retinal specialist. A blood sample was drawn for biochemical = with the DdeI + allele) and. After genotyping half determinations and for isolation of DNA. of the PstI + patients by DGGE to recognize the DdeI = allele, the distinction was found not to be informative (data not shown), so this analysis is based only on the PstI Determination of the albumin excretion rate (AER) polymorphism. The PCR mixture (total volume of 50 ml ) contained 100 ng The albumin excretion rate was determined in three 24-h of genomic DNA together with 0.25 mm of each primer, urine collections performed at least 1 month apart. The urine 2.0 mm MgCl,5ml 10 PCR buffer (GeneAmp PCR Core 2 samples were stored in polystyrene tubes at 4 C and assayed Reagents, Perkin Elmer, Roche), 0.2 mm of each dntp, and within 1 week. Urinary albumin was determined by RIA as 1.25 units of AmpliTaq DNA polymerase (Perkin Elmer, previously described [32]. Human albumin antibodies were Roche). The PCR mixture was denatured at 94 C for 1 min purchased from Bioyeda ( Rehovot, Israel ). Human serum followed by annealing at 58 C for 1 min and extension at albumin was obtained from Sigma (Holon, Israel ). 125I- 72 C for 1.5 min. Amplification was carried out for 30 cycles Labelled human albumin (specific activity 10 mci/mg) was followed by a final extension period of 10 min. The PCR purchased from the Radiochemical Centre (Negev, Israel ). amplified fragments were digested with PstI (New England Measurements of the AER were classified into three cat- Biolabs, Beverly, MA) and electrophoresed on a 1.5% agarose gel stained with ethidium bromide. The allele egories: values <30mg/24 h were considered normal, values was

3 ACE gene polymorphisms and nephropathy 2555 seen as a single band (800 bp), and the + was seen as two Table 1. Characteristics of the study groups bands (600 bp and 200 bp, approximately). The I/D polymorphism in intron 16 was determined by a Characteristic Controls Cases PCR based protocol [34]. The PCR mixture (total volume P value 50 ml ) contained 100 ng of genomic DNA together with 0.2 mm of each primer, 3.0 mm MgCl,5ml 10 PCR buffer 2 Males (%) (GeneAmp PCR Core Reagents, Perkin Elmer, Roche), Ashkenazi (%) mm of each dntp, and 1.25 units of AmpliTaq DNA Age at onset of IDDM (years) 11±7 10± Polymerase (Perkin Elmer, Roche). Previously published Duration of IDDM (years) 17±6 18± sequences were used as primers [34]. The PCR mixture was Mean blood pressure (mmhg) 78±11 85± denatured at 95 C for 30 s followed by annealing at 58 C Anti-hypertensive therapy (%) 0 17 < for 1 min and extension at 72 C for 2 min. Amplification Mean AER (mg/24 h) 13±4 673±1640 was carried out for 30 cycles followed by a final extension ACE activity (nmol/ml/min) 114±33 131± period of 10 min. The products were electrophoresed on a Proliferative retinopathy (%) < % agarose gel and stained with ethidium bromide. Fragments of 190 bp and 490 bp indicate the presence of the Values are means±s.d. D/I, respectively. To exclude the possibility of mistyping I/D heterozygotes as D/D, all individuals with D/D were geno- Table 2. Distribution of PstI genotypes and alleles in the study typed once again by a Southern blot protocol described groups elsewhere [30]. Briefly, this protocol included digestion of genomic DNA with XbaI restriction enzyme, electrophoresis Controls Cases of the digested DNA on an 0.8% agarose gel, blotting onto n (%) n (%) a nylon membrane and hybridization with a cdna probe (exons 15 19) labelled with radioactive 32P. Blots were hybridized for 16 h, washed and exposed to Kodak XAR-5 PstI Genotypesa / 18 (20.2) 16 (20.8) film. The allele corresponding to a deletion was identified as /+ 50 (56.2) 29 (37.6) a 3.15 kb band, and the allele corresponding to an insertion +/+ 21 (23.6) 32 (41.6) was identified as 3.4 kb band. Total individuals PstI alleles Statistical analysis 86 (48) 61 (39.6) The distribution of genotypes and alleles were compared between cases and controls using x2 tests. The association between genotypes and diabetic nephropathy was evaluated by computing odds ratios (ORs) and 95% confidence intervals [35]. Analysis of variance was used to compare ACE activity + 92 (52) 93 (60.4) Total chromosomes ain the control group, the genotype distribution is consistent with Hardy Weinberg equilibrium. The x2 for departure from equilibrium was 1.39 (1 d.f., P=0.24). In the group of cases, the x2 for departure and blood pressure according to genotypes and nephrobetween nephropathy and genotype =7.15 (2 d.f., P=0.028). from equilibrium was 3.49 (1 d.f., P=0.06). x2 for association pathy status. Since differences in the distribution of alleles and genotypes among study groups could result from unrecognized popula- the other genotypes (odds ratio 2.3, 95% CI: ). tion stratification the ethnic origin of study patients was This pattern of association was similar in Ashkenazi defined as previously described [31]. In the analysis, the distributions of the I/D orpsti polymorphisms in the study groups were also examined according to the ethnic origin of the patients. Results and non-ashkenazi patients (odds ratios 2.1 and 2.6, respectively). While the association resulted in a higher frequency of the + allele in the group of cases than controls, the difference in allele frequency did not reach statistical significance with this sample size ( x2=3.03, 1 d.f. P=0.08) since only homozygotes were at extra risk. The distributions of genotypes for the I/D poly- Selected characteristics of cases and controls are summarized in Table 1. The groups had similar distribu- morphism in the ACE gene differed only slightly tions according to sex, age at onset of IDDM, diabetes (Table 3). The group of cases had a deficiency of duration, and ethnic origin ( proportion Ashkenazi). heterozygotes and an excess of homozygotes in comparison In comparison with controls, the group of cases had with controls, but these differences were significantly higher means for the albumin excretion not statistically significant ( x2=1.532, 2 d.f. n.s.). rate, mean blood pressure, ACE activity and proportion Moreover, the two groups had almost identical fre- of patients with proliferative retinopathy. In the quencies of the D allele ( 63.5% and 63.6%). following analysis, the associations between ACE gene ACE activity in serum was examined according to polymorphisms in intron 7 (PstI) and in intron 16 (I/D) genotype at each of these loci. Activity was significantly and diabetic nephropathy are examined separately. higher in patients with nephropathy than in those For the PstI polymorphism, the genotype distribu- without renal complications ( Table 1). Even after tions in cases and controls were significantly different adjusting for this difference between groups with ( x2=7.153, 2 d.f. P=0.028) ( Table 2). The difference ANOVA, significant variation was also found according was due to an excess of PstI +/+ homozygotes, to genotype ( Table 4). ACE activity was highest implying a risk of nephropathy 2.3 times higher than in patients homozygous for the PstI / and lowest

4 2556 M. B. S. Freire et al. Table 3. Distribution of I/D genotypes and alleles in the study groups pathy and the PstI polymorphism in intron 7 of the ACE gene, which was described by us previously Controls Cases [30,33]. Interestingly, IDDM patients in the current n (%) n (%) study who are homozygous for the PstI allele + had a higher risk of diabetic nephropathy and a lower ACE Genotypesa activity level than patients with other genotypes. In I/I 10 (11.2) 12 (15.6) this study, we did not find association between the D I/D 45 (50.6) 32 (41.6) D/D 34 (38.2) 33 (42.8) allele in intron 16 of the ACE gene and diabetic Total individuals 89 (100) 77 (100) nephropathy. This latter result is in agreement with Alleles other studies that were not able to demonstrate associ- I 65 (36.5) 56 (36.4) ation between this allele and diabetic nephropathy D 113 (63.5) 98 (63.6) [28,29]. Total chromosomes 178 (100) 154 (100) Alleles of the PstI and I/D polymorphism are in linkage disequilibrium (data not shown). Almost all of ain the control group, the genotype distribution is consistent with the alleles of the PstI polymorphism in intron 7 are Hardy Weinberg equilibrium. The x2 for departure from equilibrium was 0.73 (1 d.f., P=0.39). In the group of cases the x2 for departure on chromosomes with the deletion in intron 16, while from equilibrium was 0.80 (1 d.f., P=0.37). x2 for association only a few percent of them are on chromosomes with between nephropathy and genotype =1.532 (2 d.f., P=0.47). the insertion. On the other hand, only two thirds of the + alleles of the PstI polymorphism are coupled Table 4. ACE activity according PstI and I/D genotypes in the with an insertion allele, while the remaining third are study groups coupled with the deletion allele in intron 16. The small sample size does not permit any conclusion about ACE activity (nmol/ml/min) whether there is any differences in the risk of diabetic Controls Cases nephropathy associated with the latter two haplotypes. Since the PstI + allele is located in an intron [33], it is presumably a marker of susceptibility rather than PstI genotypesa a direct participant in the development of nephropathy. / 121.8± ±23.1 / ± ±31.8 More DNA sequence analysis of the PstI + chromo- +/+ 98.0± ±26.9 somes is required to identify the nephropathy-causing I/D genotypesb polymorphism/mutations. The responsible polymorph- I/I 81.1± ±39.4 ism/mutations may be located in exons that encode I/D 116.6± ±29.4 for either the amino or carboxyl domain, and their D/D 121.1± ±22.6 impact may be on the catalytic activity of these domains with respect to the conversion of angiotensin Values are means±s.d. aanova F=6.54 P= banova I to angiotensin II, or the inactivation of bradykinin F=3.53 P=0.03. as well as other substrates. While the putative disease-causing polymorphism/ for those +/+, with heterozygotes having inter- mutation may be present in the promoter or regulatory mediate values (ANOVA F=3.53, P=0.03). The parts of introns and influence expression of the ACE higher ACE activity among cases persisted when the gene and levels of the enzyme in the cells as well as in comparison was stratified according to PstI genotype serum, this possibility is not supported by our findings. (F=11.14, P=0.001). A similar pattern was seen for Homozygotes for the PstI + allele were at high risk the association between ACE activity and the I/D of diabetic nephropathy and had lower serum ACE polymorphism. ACE activity was lowest in patients activity than other genotypes. On the other hand, homozygous for I/I, highest in D/D homozygotes and homozygotes for the insertion allele had similarly low intermediate in heterozygotes (ANOVA F=6.54, P= serum levels of ACE but were not at particularly high ). These differences among genotypes, however, risk of diabetic nephropathy. This implies that the could not account for the higher ACE activity in cases. putative nephropathy-causing DNA sequence differ- The difference in the ACE activity between cases and ence/mutation in the ACE gene must influence some controls was consistent across genotypes (ANOVA aspect of the biology of ACE other than its serum F=11.55, P=0.0009). When blood pressure was anaactivity. lyzed in the same way as ACE activity, significantly The findings of the present study agree with our higher diastolic and systolic blood pressures were found previous report in which we showed that the DdeI = in patients with nephropathy than in those without. allele (in the current study included as part of the PstI However, blood pressure was not influenced by PstI + allele) was associated with increased risk of diabetic or I/D genotypes (data not shown). nephropathy [30]. The extra risk of nephropathy associated Discussion with these alleles, however, was moderate in both studies. This indicates, perhaps, that the risk of nephropathy attributable to genetic variability at the In this study of Jewish IDDM patients, we found an ACE locus is relatively small. The small attributable association between prevalence of diabetic nephro- risk, as well as the undefined nature of the putative

5 ACE gene polymorphisms and nephropathy 2557 polymorphism/mutations, presumably contributed to factors determine the development of diabetic nephropathy in the discrepancies among studies reported so far patients with IDDM. Diabetologia 1996; 39: Schmidt S, Ritz E. Genetics of the renin-angiotensin system and [11,12,28 30]. If the degree of linkage disequilibrium renal disease: a progress report. Curr Opin Neph Hypert 1997; between the disease locus and the I/D polymorphism 6: varies among populations, positive association will be 6. Fogarty DG, Krolewski AS. Genetic susceptibility and the role found in populations with tighter linkage but not in of hypertension in diabetic nephropathy. Curr Opin Neph Hypert 1997; 6: populations with weaker linkage. However, in these 7. Liebermann J, Sastre A. Serum angiotensin-converting enzyme: latter populations, association with the PstI poly- Elevation in diabetes mellitus. Ann Int Med 1980; 93: morphism in intron 7 may have been easily demon- 8. Hallab M, Berrut G, Bouhanick B, et al. Increase of activity of strable if it had been examined. To reconcile the angiotensin-converting enzyme in insulin-dependent diabetic existing discrepancies, further studies are needed in patients with permanent microalbuminuria. Arch Mal Coeur Vaiss 1992; 85: which diabetic patients with and without nephropathy 9. van Dijk DJ, Erman A, Erman T, Chen-Gal B, Sulkes J, are genotyped for other markers at the ACE locus Boner G. Increased serum angiotensin converting enzyme activity in type I insulin-dependent diabetes mellitus: its relation to besides the I/D. Finally some limitations of the study will be men- metabolic control and diabetic complications. Eur J Clin Invest 1994; 24: tioned. First, since prevalence cases were used in this 10. Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, study, one should be concerned with survival bias. For Soubrier F. An insertion/deletion polymorphism in the example, individuals with certain alleles may be at angiotensin-i converting enzyme gene accounting for half the higher risk of developing ESRD and death and, there- variance of serum enzyme levels. J Clin Invest 1990; 86: fore, be unable to come to the centre where recruitment Marre M, Bernadet P, Gallois Y, et al. Relationship between for this study took place. The young mean age at angiotensin I converting enzyme gene polymorphism, plasma examination of cases and controls (27 and 28 years, levels and diabetic retinal and renal complications. Diabetes respectively) significantly reduces the plausibility of 1994; 43: this scenario. Second, the difference in the distribution 12. Marre M, Jeunemaitre X, Gallois Y et al. Contribution of of genotypes between cases and controls could have genetic polymorphism in the renin-angiotensin system to the development of renal complications in insulin-dependent diaresulted from unrecognized population stratification. betes: Genetique de la Nephropathie Diabetique (GENEDIAB) Although this possibility can be excluded only in a study group. J Clin Invest 1997; 99: family based study, in this case-control study we found 13. Lewis EJ, Hunsicker LG, Bain RP, Rohde RD, for the similar association between homozygosity for PstI + Collaborative Study Group. The effect of angiotensin-converting enzyme inhibition on diabetic nephropathy. N Engl J Med 1993; allele and diabetic nephropathy in IDDM patients of 329: Ashkenazi and non-ashkenazi origin. Third, although 14. Viberti GC, Mogensen CE, Groop LC, Pauls JF, for the we examined the ACE locus and found association European Microalbuminuria Captopril Study Group. Effect of between diabetic nephropathy and the PstI + allele captopril on progression to clinical proteinuria in patients with in intron 7, the disease-causing DNA sequence differ- insulin-dependent diabetes mellitus and microalbuminuria. JAMA 1994; 271: ence may be located outside the ACE gene. The 15. Ballerman BJ, Zeidel ML, Gunning ME, Brenner BM. plausibility of this alternative is bolstered by recent Vasoactive peptides and the kidney. In: Brenner BM, Rector sequence data. The exons and promoter region of the FC, ed. The Kidney. 4th ed, Saunders, Philadelphia: 1991: ACE gene of two D/D homozygotes and two I/I pp homozygotes were sequenced by Villard et al. [36]. No 16. Corvol P, Michaud A, Soubrier F, Williams TA. Recent advances in knowledge of the structure and function of the DNA sequence differences that could be considered angiotensin I converting enzyme. J Hypert 1995; 13 [Suppl. biologically important were found. The importance of 3]: S3 S10 these results is qualified by the lack of sequence data 17. Hubert C, Houot AM, Corvol P, Soubrier F. Structure of the on PstI + chromosomes. angiotensin I converting enzyme gene. Two alternate promoters correspond to evolutionary steps of a duplicated gene. J Biol Chem 1991; 266: Acknowledgements. This study was supported by grant NIH DK 18. Mattei MG, Hubert C, Alhenc-Gelas F, Roeckel N, Corvol P, M.B.S. Freire was supported by a post-doctoral fellowship Soubrier F. Angiotensin-I converting enzyme is on chromosome from the Brazilian government: CNPq Conselho Nacional de 17. (abstract) Cytogenet Cell Genet 1989; 51: 1041 Desenvolvimento Cientifico e Tecnologico. The Rabin Medical 19. Doria A, Warram JH, Rich SS, Krolewski AS. Determination Centre group would like to acknowledge the contribution of Mrs of DNA haplotypes in unrelated individuals using denaturing Batya Chen-Gal who performed the serum ACE activity studies. gradient gel blots: angiotensin I converting enzyme locus. Hum Genet 1994; 94: Villard E, Tiret L, Visvikis S, Rakotovao R, Cambien F, References Soubrier F. Identification of new polymorphisms of the angiotensin I converting enzyme (ACE) gene, and study of their 1. Krolewski AS, Warram JH, Rand LI, Kahn CR. Epidemiologic relationship to plasma ACE levels by two QTL segregationapproach to the etiology of type I diabetes mellitus and its linkage analysis. Am J Hum Genet 1996; 56: complications. N Eng J Med 1987; 317: Cambien F, Poirier O, Lecerf L, et al. Deletion polymorphism 2. Seaquist ER, Goetz FC, Rich S, Barbosa J. Familial clustering in the gene for angiotensin-i converting enzyme is a potent risk of diabetic kidney disease: evidence for genetic susceptibility to factor for myocardial infarction. Nature 1992; 359: diabetic nephropathy. N Engl J Med 1989; 320: Ruiz J, Blanche H, Cohen N, et al. Insertion/deletion poly- 3. Borch-Johnsen K, Norgaard K, Hommel E, et al. Is diabetic morphism of the angiotensin converting enzyme gene is strongly nephropathy an inherited complication? Kidney Int 1992; 41: associated with coronary heart disease in non-insulin-dependent diabetes mellitus. Proc Natl Acad Sci USA 1994; 91: Quinn M, Angelico MC, Warram JH, Krolewski AS. Familial 23. Oike Y, Hata A, Ogata Y, Numata Y, Shido K, Kondo K.

6 2558 M. B. S. Freire et al. Angiotensin converting enzyme as a genetic risk factor for converting enzyme gene and diabetic nephropathy and proliferative coronary artery spasm. Implication in the pathogenesis of retinopathy in IDDM patients. Diabetes 1995; 44: myocardial infarction. J Clin Invest 1995; 96: Doria A, Warram JH, Krolewski AS. Genetic predisposition to 24. Gardemann A, Weiss T, Schwarts O, et al. Gene polymorphism diabetic nephropathy: evidence for a role of angiotensin-i conbut not catalytic activity of angiotensin I converting enzyme is verting enzyme gene. Diabetes 1994; 43: associated with coronary artery disease and myocardial infarction 31. Kalter-Leibovici O, van Dijk DJ, Leibovici L et al. Risk factors in low-risk patients. Circulation 1995; 92: for development of diabetic nephropathy and retinopathy in 25. Lindpaintner K, Pfeffer MA, Kreutz R, et al. A prospective jewish IDDM patients. Diabetes 1991; 40: evaluation of an angiotensin-converting-enzyme gene poly- 32. Erman A, Rabinov M, Rosenfeld J. Albumin determination in morphism and the risk of ischemic heart disease. N Engl J Med frozen urines underestimated results. Clin Chim Acta 1998; 1995; 332: : Yoshida H, Mitarai T, Kawamura T, et al. Role of the deletion 33. Doria A, Warram JH, Krolewski AS. Molecular characterization of polymorphism of the angiotensin converting enzyme gene in of a DdeI melting polymorphism at the angiotensin I converting the progression and therapeutic responsiveness of IgA nephropathy. enzyme (ACE) locus. Hum Mut 1994; 4: J Clin Invest 1995; 96: Rigat B, Hubert C, Corvol P, Soubrier F. PCR detection of the 27. Hunley TE, Julian BA, Phillips III JA, et al. Angiotensin insertion/deletion polymorphism of the human angiotensin conconverting gene polymorphism: Potential silencer motif and verting enzyme gene. Nucleic Acids Res 1992; 20: 1433 impact on progression in IgA nephropathy. Kidney Int 1996; 35. Rothman KJ. Modern Epidemiology. Little, Brown & Company, 49: Boston: 1986: pp Schmidt S, Schone N, Ritz E and the diabetic nephropathy 36. Villard E, Tiret L, Visvikis S, Rakotovao R, Cambien F, study group. Association of ACE gene polymorphisms and Soubrier F. Identification of new polymorphisms of the angiotensin diabetic nephropathy? Kidney Int 1995; 17: I-converting enzyme (ACE) gene, a study of their relationdiabetic 29. Tarnow L, Cambien F, Rossing P, et al. Lack of relationship ship to plasma ACE levels by two-qtl segregation-linkage between Insertion/Deletion polymorphism in the angiotensin-i analysis. Am J Hum Genet 1996; 58: Received for publication: Accepted in revised form: