Angiotensin-converting enzyme (ACE) haplotypes and cyclosporine A (CsA) response: a model of the complex relationship between ACE quantitative trait locus and pathological phenotypes

Please download to get full document.

View again

of 11
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Angiotensin-converting enzyme (ACE) haplotypes and cyclosporine A (CsA) response: a model of the complex relationship between ACE quantitative trait locus and pathological phenotypes
Document Share
Document Tags
Document Transcript
  Angiotensin-converting enzyme ( ACE  ) haplotypesand cyclosporine A (CsA) response: a model of thecomplex relationship between ACE  quantitative traitlocus and pathological phenotypes Paolo Catarsi 2, * , Roberto Ravazzolo 3,7 , Francesco Emma 4 , Doriana Fruci 5 , Livio Finos 6 ,Andrea Frau 2 , Giacomo Morreale 2 , Alba Carrea 2 and Gian Marco Ghiggeri 1,2 1 DepartmentofNephrology,  2 Laboratory of Pathophysiology of Uremia and  3 Laboratory of Molecular Genetics, G. GasliniInstitute, Genova,  4 Division of Nephrology and Dialysis and  5 Laboratory of Cell Biology, Bambin Gesu` Hospital, Romaand  6 Faculty of Medicine, D.S.B.T.A. Section of Human Physiology, University of Ferrara, Ferrara,  7 Department ofPediatrics and CEBR, University of Genova, Italy Received May 6, 2005; Revised and Accepted June 29, 2005 It is highly controversial to define the role of angiotensin-converting enzyme (ACE) polymorphisms inessential hypertension. We studied a group of patients in whom hypertension was the major side effect oftreatmentbycyclosporineA(CsA).Thisstudygroupcomprised 227Italianpatientswithnephroticsyndrome,103 of which were treated with CsA and had different outcome. Forty-nine patients developed serious hyper-tension that was reversed after withdrawal of drug.  ACE   haplotypes were determined by a combination ofmolecular and statistical methods after verifying genotypes of six intragenic single nucleotide polymorp-hisms in 304 Italian blood donors and assembling them in clades (A, B, C) that include 95% of observedhaplotypes. The association between  ACE   clade combinations and serum enzymatic levels confirmed theprevious results about a role of an unidentified genetic variant at the 5 0 of the intragenic recombinationsite located near intron 7.  ACE   clades were then determined in patients, and regression methods wereused to analyze variables associated with CsA responsivity and progression to renal failure.  ACE   genotypeand responsiveness to CsA were strictly associated, because homozygosis for  ACE   B clade was able toinfluence CsA sensitivity. This highlights the role of 5 0 variants, which differentiate clades B and C. Othergenetic markers were tested to search for possible additive effects. We found that  PAI-1 4G   allele was asso-ciated withprogressiontorenalfailureinthegroupofCsA-treatedpatients.Ourresultsareinagreementwiththe hypothesis, raised after experimental results obtained in mouse models, that the effect of  ACE  polymorphisms on blood pressure is detectable once environmental factors, like CsA treatment in ourcase, overcome physiological homeostatic mechanisms. INTRODUCTION Angiotensin II (AngII) is a potent vasoactive and fibrogenic peptide with wide implications in human physiology and patho-logy. Almost 60% of the biologically active AngII is derived from the conversion of AngI into the active peptide by ACE(1), a large protein encoded by the  ACE   gene (MIM106180).AngII levels depend therefore, to some extent, on ACElevels that are, in part, determined by genetic factors.Definition of   ACE   genotypes, which influence levels, and their association with human diseases such as hypertension,myocardial infarction, renal fibrosis and so on were a logicalconsequence of the above assumption, but clear proof for adirect relationship is still lacking. The starting point wasthe recognition of a major quantitative trait locus (QTL)in  ACE   gene in 1990 where the presence of a 287 bp # The Author 2005. Published by Oxford University Press. All rights reserved. For Permissions, please email:  *To whom correspondence should be addressed at: Laboratory of Pathophysiology of Uremia, G. Gaslini Institute, Largo Gaslini 5, 16147 Genova,Italy. Tel:  þ 39 010 5636 419; Fax:  þ 39 010 395214; Email:  Human Molecular Genetics, 2005, Vol. 14, No. 16   2357–2367  doi:10.1093/hmg/ddi238 Advance Access published on July 7, 2005   b  y g u e  s  t   onF  e  b r  u a r  y2 4  ,2  0 1  6 h  t   t   p :  /   /  h m g . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   insertion/deletion polymorphism in intron 16 (I/D; rs1344744)was identified: the  D  allele was codominant in raising ACE plasma levels in Europeans (2–4). New polymorphisms atthe 5 0 in strong linkage disequilibrium with the I/D locuswere identified and it was proposed a bifunctional QTLmodel in which two different causative variants act together in determining ACE level variability (5). Finally it was intro-duced a haplotype-based model in which an ancestral recom- bination breakpoint at intron 5–exon 8 interval determines ahaplotype cluster carrying the  D  variant but still differing atthe 5 0 of the gene (6–8). Data on  ACE   haplotypes have beenextended to different populations in the last few years, consid-ering several single nucleotide polymorphisms (SNPs) withthe conclusion that a cladistic approach better fits statisticalanalysis in Europeans. An approach based on a simple cladis-tic structure was first proposed by Keavney  et al.  (7) in a studyon a British white population. Clades A, B and C were defined on the basis of evolutionary information and they account for about 90% of the observed haplotypes. This concept wasfurther extended to Africans and led to consider a few other SNPs as causative of ACE levels: at the promoter region(rs4291), at exon 17 (rs4343) (9) and at the 3 0 end of intron25 (rs4363) (9–11). Cox  et al.  (12) defined a multiple, inde- pendent effect of four variants in Nigerian families, a minor one (rs4459609) in the 5 0 flanking region at  2 5499 from theATG starting codon and three located downstream of therecombination site (rs4343, 31839delC, rs4363). Takentogether, the aforementioned studies demonstrated that theI/D polymorphism is indeed associated with serum level, buta cladistic approach based on haplotype analysis represents amore efficient determinant of   ACE  -related phenotypic traits.Although D/D genotype is invariably linked to high circu-lating levels, data associated with cardiovascular and renaldiseases are controversial (13). Moreover, association of   ACE   clades with human pathology is not a direct consequenceof levels, as hypothesized in the past. Results by Zhu  et al.  (9)contradicted, in fact, this possibility because they found anepistatic interaction between the aforementioned variantslocated in the promoter (rs4291 A/T) and in exon 17(rs4343 A/G), which affects blood pressure. Furthermore theauthors observed that the allelic effect of the promoter variantsis opposite in direction for ACE concentration and blood  pressure. This probably reflects a complex interaction between genetic loci not necessarily linked to ACE levels.Beside hypertension,  ACE   I/D has been implicated in theclinical outcome to renal failure in patients who developtubulo-interstitial fibrosis. Several genetic variants of mole-cules synergic with AngII in determining fibrosis may playadditive effects to  ACE   locus (14). We study the effect of three functional polymorphisms that belong to the renin-angiotensin system (RAS), i.e. M235T (C4027T) of angioten-sinogen (  AGT  ) (MIM106150), A1166C of angiotensin IItype 1 receptor (  AT1 ) (MIM106165) (15,16) and theK528R (A1583G) (17) in exon 11 of adipocyte-derived leucine aminopeptidase gene (  ALAP ) (MIM606832) plus theSNP 4G/5G (four or five guanines) in the promoter regionof plasminogen-activator inhibitor type 1 (  PAI-1 ) gene(MIM173360). The former three regulate AngII levels (  AGT  and   ALAP ) and function (  AT1 ), whereas  PAI-1  regulatesremoval of extracellular matrix (18).This work has two major aims: the first was to consider thecomplex interaction among the  ACE   genetic locus, enzymaticlevels and human hypertension in a clinical condition such ascyclosporine-induced hypertension in which AngII appearsmore directly involved in regulating blood pressure. A vastarea of clinical and experimental observations in animalsand humans support this possibility (19–25). The second was to consider the additive effects of synergic polymorp-hisms in determining evolution to end-stage renal failure(ESRF).For this reason, we have evaluated   ACE   haplotypes in agroup of patients with nephrotic syndrome, part of themtreated with cyclosporine. We found a strong association of different  ACE   clades with either protection or intoleranceto the drug, always manifested as severe hypertension and worsening of renal function. RESULTS Population genetics of   ACE   haplotypes The starting choice of genetic markers for defining  ACE   hap-lotypes was based on a critical review of data reported byRieder   et al.  (8) who defined haplotypes in Europeans and African-Americans on the basis of 52 SNPs. This originalreport, in agreement with the study by Keavney  et al.  (7) onEuropeans, supported the idea of an ancestral recombinationin  ACE   gene and validated the association of   ACE   haplotypesin three major clades (A, B and C) that include the five mostfrequent haplotypes as follows: H6 and H8 in clade A, H1 inclade B, H7 and H9 in clade C. Six of the srcinal 52 SNPsare sufficient to define haplotypes and therefore clades: threemarkers (rs4424958, rs4309 and rs4311) flank the recom- bination site, one upstream in intron 2 (rs4295) and two down-stream, in introns 16 (I/D) and 25 (rs4363) (see Table 1). Thecombination of single marker genotypes, which characterizeshaplotypes and clades, is given in Table 2. In agreementwith other studies in Europe (7,26,27), we could confirm anoverall frequency of the haplotypes described by Rieder inthe 94% of a normal Italian population. The finding of H1-6and the reciprocal C2 possible products of recombination between H1 and H6 haplotypes (see Table 2) allows to statethat recombination has occurred between rs4309 and rs4311,as already hypothesized (6). Recombination betweenrs4424958 and rs4309 could also have occurred although therecombination products reciprocal to H7 and H9 were notfound. As described in other studies (27,28), the presence of this ancestral recombination breakpoint does not influencethe high level of linkage disequilibrium across the gene(Table 3). Using the Arlequin 2000 software, we tested and excluded significant departure from the Hardy-Weinberg equi-librium for both genotypes and haplotypes frequencies. ACE haplotypes and levels ACE levels were determined in normal population with diffe-rent combinations of haplotypes belonging to A, B, C clades.In agreement with Danilov  et al.  (29), we could confirm astrong correlation between serum ACE activity and proteinlevels ( r  ¼ 0.9) in a sample of 16 individuals representative 2358  Human Molecular Genetics, 2005, Vol. 14, No. 16    b  y g u e  s  t   onF  e  b r  u a r  y2 4  ,2  0 1  6 h  t   t   p :  /   /  h m g . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   of different haplotype combinations. ACE levels in relation todifferent genotypes and haplotypes are reported in Figure 1and Table 4. Basically, serum ACE levels were differentamong carriers of I/I, I/D and D/D genotypes, however thestratification in haplotypes splits the two simple I/D and D/Dgenotypes of the previous classification in five possible combi-nations (2 for I/D and 3 for D/D) that present different ACElevels (Figure 1). The most remarkable difference is betweenB/B and B/C, both included in the previous D/D genotype.This finding supports the hypothesis that differences in ACEserum levels are associated to the I/D alleles, with anadditional contribution of a variant located at the 5 0 of therecombination site. Determinants of CsA responsivity and clinical outcome In a recent analysis of factors that influence survival in patients with steroid-resistant nephrotic syndrome, CsA wasfound as the major modifier, improving the outcome in a portion of patients with worse prognosis (30). We haveevaluated 99 patients treated with CsA considering severalvariables, which included genetic factors, by logisticregression analysis. Results relative to genetic variableswere given following a dominant model of allelic effect on phenotype except for   ACE   haplotype in the form B/B versusothers that gave most significant result in contingency tableanalysis (see below). Table 5 shows that homozygosis for   ACE   B clade was able to influence CsA sensitivity(  P ¼ 0.0139) in the overall cases and such influence was con-firmed when considering only steroid-resistant patients. CsAresponse was also influenced by response to steroid treatment(  P , 0.0001) but this analysis highlights the independence of the  ACE   haplotype for other clinical variables. Contingencytable analysis on all clade combinations shows a five timeshigher frequency of B/B in unresponsive versus responsive patients. This is counterbalanced by a higher A/C (19%versus 11%) and C/C (8% versus 2%), respectively, in respon-sive and unresponsive to CsA patients (Table 6). Overall, thedifference is statistically significant. In untreated patients, fre-quencies of different clades were the same. The concomitanthigher C and lower B frequencies in responsive patientsexplains lack of significance using the I/D marker because both clades C and B include the  D  allele.Almost all patients who were treated and had a good response to CsA had a genotype different from B/B (23 outof 24) whereas in the subgroup of intolerant to CsA 13 outof 45 were B/B. Overall, 13 out of 14 B/B in the category Table 2.  Haplotype composition and frequencies in 304 Italian blood donorsHaplotypes(clades)rs4295 rs4424958 rs4309 rs4311 rs13447447 rs4363 Frequency(%)H1 (B)  C A C T D G   40H6 (A)  G G T C I A  35.9H7 (C)  G G C T D G   9.7H9 (C)  G G C C D G   6.9C2 (C)  G G T T D G   1.5H1-6  C A C C I A  1.5H8 (A)  G G C C I A  1.5Others 3 Table 3.  Pairwise linkage disequilibrium (  D 0 ) and physical distances (belowdiagonal)rs4295 rs4424958 rs4309 rs4311 rs13447447 rs4363rs4295 — 0.9865 0.9798 0.9171 0.8932 0.8409rs4424958 3326 bp — 0.9900 0.9260 0.9036 0.8514rs4309 296 bp — 0.9194 0.9284 0.9266rs4311 839 bp — 0.9922 0.9178rs13447447 5125 bp — 0.9859rs4363 8601 bp —  Figure 1.  Serum ACE activities in 170 blood donors. Individuals were subdi-vided into six groups corresponding to different  ACE   clade combinations.Classification based on I/D polymorphism is indicated in the upper portion.The use of clade-based genotypes leads to identification of groups withdifferent serum activities, although sharing D/D genotype (see also Table 4). Table 1.  Marker positions and allele frequencies in  ACE   geneSNPs PositionRieder (1999)Location Allele 1 Frequency(%)Allele 2 Frequency(%)rs4295 4504 Intron 2  G   56.9  C   43.1rs4424958 7831 Intron 7  G   56.6  A  43.4rs4309 8128 Exon 8  C   62.2  T   37.8rs4311 8968 Intron 9  T   53.9  C   46.1rs13447447 14094 Intron 16  D  60.7  I   39.3rs4363 22982 Intron 25  G   59.2  A  40.8 Table 4.  Post hoc  testsPairwisecomparisonsMean difference(U/ml)  P -valueA/A, A/C 1.35 0.0003A/C, A/B 0.113 0.7496A/B, B/C 0.867 0.0175B/C, B/B 0.951 0.0093 More explicative comparisons are reported (C/C genotype was excluded  because of small sample size).  Human Molecular Genetics, 2005, Vol. 14, No. 16   2359   b  y g u e  s  t   onF  e  b r  u a r  y2 4  ,2  0 1  6 h  t   t   p :  /   /  h m g . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   of patients treated with CsA were intolerant. Therefore ACEgenotype and responsiveness to CsA were strictly associated.We then looked at several clinical and genetic parametersinfluencing progression to ESRF in the context of a multi-variate approach (Table 7): the variables CsA treatment,  ACE   haplotype and   PAI-1  and   AT-1  genotypes were selected on the basis of a stepwise method. Also in this case, geneticvariables were given following a dominant model of alleliceffect on phenotype with exception of   ACE   haplotype, alsoin this case, in the form B/B versus others. This groupingwas based on both results from logistic regression and onthe observation that differences in ACE levels become remark-able upon splitting the D/D subgroups in B/B, B/C and C/C.Confirming previous data (30), progression was stronglyinfluenced by CsA response. In the subgroup with intoleranceto CsA, the outcome was not influenced by  ACE   genotype butonly by  PAI-1 4G   allele considered as dominant (Table 7 and Figure 2). This is most likely due to the strong association between diplotype and CsA responsivity, which masks thesignificance of the B/B diplotype in the survival regressionfor the treatment group. In the untreated patients, the CsAresponsivity cannot be considered and the variable  ACE  haplotype, although not significant, shows some relationwith the failure time.Therefore,  ACE   is strictly associated with responsivenessand/or intolerance to CsA and, without any treatment, isonly weakly associated to progression.  PAI-1  is associated with progression to renal failure in the group of CsA-intolerant patients. DISCUSSION Regulation of blood pressure in humans is a classical exampleof complex trait resulting from the interplay of several homeo-static mechanisms. Clinical and experimental evidencessupport the general idea that hypertension arises when signifi-cant environmental events overcome the homeostatic potentialin a specific genetic background. This makes it difficult toapproach the genetic basis of hypertension in large populationstudies where environmental factors are not uniform.The definition of a homeostatic key role of the kidney inregulation of blood pressure and in hypertension represented a milestone in the understanding of pathogenetic mechanisms.This is essentially based on the regulation of sodium balance Table 6.  Contingency tables analysis for the relationship between  ACE   genotypes and CsA responsivityGenotype c SD þ SR patients (  N  ¼ 99) SR patients (  N  ¼ 67)Observed frequencies a Expected frequencies  b Observed frequencies a Expected frequencies  b CsA respons.(N ¼ 52)CsA intol. or resist.(N ¼ 47)CsA respons.(N ¼ 24)CsA intol. or resist.(N ¼ 43)A/A 5 (0.10) 6 (0.13) 0.11 2 (0.08) 6 (0.14) 0.12A/B 15 (0.29) 11 (0.23) 0.26 6 (0.25) 10 (0.23) 0.24A/C 10 (0.19) 5 (0.11) 0.15 6 (0.25) 4 (0.09) 0.15B/B 3 (0.06) 14 (0.30) 0.17 1 (0.04) 13 (0.30) 0.21B/C 15 (0.29) 10 (0.21) 0.25 6 (0.25) 9 (0.21) 0.22C/C 4 (0.08) 1 (0.02) 0.05 3 (0.13) 1 (0.02) 0.06Allele d  (N ¼ 104) (N ¼ 94) (N ¼ 48) (N ¼ 86)B 36 (0.35) 49 (0.42) 0.43 14 (0.29) 45 (0.52) 0.44C 33 (0.32) 17 (0.19) 0.25 18 (0.38) 15 (0.17) 0.25A 35 (0.34) 28 (0.39) 0.32 16 (0.33) 26 (0.30) 0.31 a Values are genotype and allele numbers observed (frequencies).  b Expected frequencies are calculated by 6  2 for genotypes or 3  2 for alleles, contingency tables. c SD þ SR (  P ¼ 0.0339;  x 2 ¼ 12.069; DF ¼ 5); SR (  P ¼ 0.0563;  x 2 ¼ 10.763; DF ¼  5). d  SD þ SR (  P ¼ 0.0247;  x 2 ¼ 7.400; DF ¼ 2); SR (  P ¼ 0.0118;  x 2 ¼  8.880; DF ¼ 2). Table 5.  Logistic regression coefficients and Wald test  P -value for CsA responsivity in steroid-treated patientsRegression coefficient (SE)/  P -value for   ACE   haplotype(B/B a )  PAI-1 (4G  b )AGT( T   b )  AT-1 ( C   b )  ALAP ( G   b )Steroid responsivity(resistance a )Cycloph.(Treatment a )SEX(M a )Age atonsetSR  c (67)  2 2.3 (1.11)0.04160.53 (0.66)0.42310.59 (0.66)0.36660.33 (0.58)0.57040.806 (0.78)0.3002 — 0.09 (0.59)0.87720.58 (0.60)0.3335 2 0.01 (0.02)0.7223SR  þ SD d  (99)  2 2.02 (0.81)0.01200.68 (0.58)0.24170.54 (0.57)0.34310.49 (0.53)0.35510.71 (0.73)0.3270 2 2.59 (0.66) , 0.00010.14 (0.53)0.78990.34 (0.53)0.5178 2 0.01 (0.02)0.8201 a Level of a dichotomous covariate at which are associated reported regression coefficients and   P -value.  b Allele considered dominant in the analysis. c Steroid-resistant patients. d  Steroid-dependant patients. 2360  Human Molecular Genetics, 2005, Vol. 14, No. 16    b  y g u e  s  t   onF  e  b r  u a r  y2 4  ,2  0 1  6 h  t   t   p :  /   /  h m g . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om   and is particularly evident in case of renal diseases and renalfailure. Elegant experiments with cross-renal transplants inanimals (31,32) and in humans (33) first demonstrated thekey role of kidney and recent genetic advances on tubular salt transporters (34) contributed to provide further rational proofs. Current view on the interplay between hypertensionand the kidney considers the possibility that acquired subtlerenal defects confer sensitivity to salt overload through the basic mechanism of local vasoconstriction. Johnson  et al. (35) recently reviewed the putative causes of initial subclinicaldamage and proposed two general physiology imbalances suchas hyperactive sympathetic nervous system and stimulated RAS and, in parallel, two environmental causes such as low- potassium diet and CsA use. CsA is an immunosuppressivedrug used in therapy of autoimmune diseases, but thistreatment is commonly associated with the development of hypertension and nephrotoxicity: evidence of the literatureindicates that hyperactivity of RAS and related systems isimplicated in such events. When we analyzed patients receiv-ing CsA for response and adverse effects, we realized thatinvestigating the genetic background of these effects wasworth to be done, based on the above assumption that anyimplication of genetic variants is amplified in such a particular environment. In other words, the systems directed at maintain-ing blood pressure homeostasis (36,37) are altered withcyclosporine, and the phenotypic effects of genetic variantsmay become more evident in this setting.Influence of genetic factors on blood pressure control has been investigated by several means among which linkage toregions in chromosome 17 has been established (for areview, see Ref. (38)). A region on rat’s chromosome 10,syntenic to human chromosome 17, was first described. So,linkage between hypertension and markers in the interval60–67 cM from the proximal telomere of the human chromo-some 17 was identified (39,40). Wide genome scan in participants from the Framingham study confirmed this criticalarea (17q12-21) and described an additional minor associationwith a locus overlapping the  ACE   gene (41). A marginalassociation, restricted to males, was found between I/D and diastolic blood pressure in 3095 subjects (42) that was con-firmed in males of a large group of 1488 siblings byFornage  et al.  (43). Wu  et al.  (44) found an association between 17q23 and young onset hypertension in 59 Chinesefamily, using TDT analysis.Moreover, studies on  ACE   polymorphisms in human hyper-tension were initially based on the reasonable assumption of adirect implication of ACE levels suggested both by knowledgeof pathophysiological mechanisms and by clinical studies onthe beneficial anti-hypertensive effect of ACE inhibitors and AT1 antagonists but most investigation produced negativeresults (13).A recent publication suggested that different molecular vari-ants located in the 5 0 flanking region could be independentlyassociated to different physiological phenotypes: blood  pressure and ACE levels (9). With the purpose to verifysuch hypothesis, we studied a group of patients with nephroticsyndrome who received cyclosporine as a second line defenseto proteinuria. In a preceding paper (30), we reported thatthe long-term outcome in these patients is dependent onsensitivity or intolerance to this drug, hypertension and renal toxicity being the hallmarks of CsA toxicity. Clinicaland experimental observations in humans and rats support adirect relationship between AngII and hypertension due toCsA (20,22,24,25,45). Our major finding is that B/B genotypeis almost invariably associated with CsA toxicity, therebyimplying hypertension, whereas other categories associated with high ACE levels such as B/C and C/C are neutral or  positively linked with drug responsiveness.As clades B and C are differentiated by the presence of avariant localized upstream of the recombination site, one pos-sible interpretation of our findings is that this variant is a deter-minant of CsA hypertensive response in spite of moderatedifference in ACE levels. The mutual aspect is the association between clade C and CsA responsiveness because clade C is,respectively, more and less frequent in patients with sensitivityand intolerance to CsA. Therefore, we consider unlikely thatthe effects of genetic factors characterizing the differentclades can be explained by differential ACE enzymaticactivity and rather advance other hypothesis: (a) the 5 0 variant confers functional properties at the level of tissue(s)which influence blood pressure control; (b) more than onevariant in the 5 0 region could independently influence blood  pressure and ACE activity; (c) the 5 0 functional variant, inlinkage disequilibrium with those analyzed, is external to  ACE.  Regarding hypothesis (c), based on reports aboutlinkage disequilibrium in the genomic region of interest, thelocation of such variant can be approximately limited to aregion extending up to about 729 kb from the  ACE   gene Table 7.  Survival regression coefficients for the selected variables in patients with steroid-resistant nephrotic syndrome who were treated or not with cyclosporineLikelihood ratio  P -value (DF)Exponential coefficient (95%CI)/  P -value a for CsA responsivity(intol. or resist. c )  PAI-1 ( 4G   b )  ACE   haplotype(B/B c )AT-1( C   b )Treatment group (67)  , 0.0001(4) 15.95 (3.34–76.24) 3.35 (1.22–9.22) 1.68 (0.73–3.88) 1.91 (0.81–4.51)0.0005 0.0189 0.2232 0.1407Untreated patients (84) 0.2808 (3) — 1.52 (0.71–3.24) 1.96 (0.97–3.96) 1.14 (0.64–2.01)0.2789 0.0612 0.6529 a  P -values with one degree of freedom.  b Allele increases risk of renal failure. c Level of a covariate at which reported exponential coefficients and   P -value are associated.  Human Molecular Genetics, 2005, Vol. 14, No. 16   2361   b  y g u e  s  t   onF  e  b r  u a r  y2 4  ,2  0 1  6 h  t   t   p :  /   /  h m g . oxf   or  d  j   o ur n a l   s  . or  g /  D o wnl   o a  d  e  d f  r  om 
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks