Figures
Abstract
Aims/Hypothesis
Several studies have provided compelling evidence implicating the Wnt signalling pathway in the pathogenesis of diabetic nephropathy. Gene expression profiles associated with renal fibrosis have been attenuated through Wnt pathway modulation in model systems implicating Wnt pathway members as potential therapeutic targets for the treatment of diabetic nephropathy. We assessed tag and potentially functional single nucleotide polymorphisms (SNPs; n = 31) in four key Wnt pathway genes (CTNNB1, AXIN2, LRP5 and LRP6) for association with diabetic nephropathy using a case-control design.
Methods
SNPs were genotyped using Sequenom or Taqman technologies in 1351 individuals with type 1 diabetes (651 cases with nephropathy and 700 controls without nephropathy). Cases and controls were white and recruited from the UK and Ireland. Association analyses were performed using PLINK, to compare allele and haplotype frequencies in cases and controls. Adjustment for multiple testing was performed by permutation testing.
Results
Following logistic regression analysis adjusted by collection centre, duration of T1D, and average HbA1c as covariates, a single SNP in LRP6 (rs1337791) was significantly associated with DN (OR = 0.74; CI: 0.57–0.97; P = 0.028), although this was not maintained following correction for multiple testing. Three additional SNPs (rs2075241 in LRP6; rs3736228 and rs491347 both in LRP5) were marginally associated with diabetic nephropathy, but none of the associations were replicated in an independent dataset. Haplotype and subgroup analysis (according to duration of diabetes, and end-stage renal disease) also failed to reveal an association with diabetic nephropathy.
Conclusions/Interpretation
Our results suggest that analysed common variants in CTNNB1, AXIN2, LRP5 and LRP6 are not strongly associated with diabetic nephropathy in type 1 diabetes among white individuals. Our findings, however, cannot entirely exclude these genes or other members of the Wnt pathway, from involvement in the pathogenesis of diabetic nephropathy as our study had limited power to detect variants with small effect size.
Citation: Kavanagh DH, Savage DA, Patterson CC, McKnight AJ, Crean JK, Maxwell AP, et al. (2011) Association Analysis of Canonical Wnt Signalling Genes in Diabetic Nephropathy. PLoS ONE 6(8): e23904. https://doi.org/10.1371/journal.pone.0023904
Editor: Massimo Federici, University of Tor Vergata, Italy
Received: March 4, 2011; Accepted: July 28, 2011; Published: August 18, 2011
Copyright: © 2011 Kavanagh et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: David H. Kavanagh is supported by a Northern Ireland Kidney Research Fund (NIKRF) PhD studentship. We thank the NIKRF and Northern Ireland Research and Development Office for supporting this work. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Renal interstitial fibrosis and glomerular sclerosis are hallmarks of diabetic nephropathy (DN) and several studies have implicated the canonical Wnt pathway in this fibrotic process [1]–[2]. The Wnt/β-catenin signalling pathway modulates numerous developmental processes and mutations in Wnt/β-catenin pathway members have been implicated in multiple diseases [3]. In humans, 19 different specific Wnt ligands have been reported to induce Wnt/β-catenin signalling [3]. These ligands bind to transmembrane Frizzled (FZD) receptors and their co-receptors, low-density lipoprotein receptor related proteins 5 or 6 (LRP5/LRP6), forming a binding complex that recruits intracellular scaffolding protein Dishevelled (Figure 1). This leads to phosphorylation and sequestration of the Axin complex, which is composed of the proto-oncogene adenomatous polyposis coli, casein kinase 1 and glycogen synthase kinase 3, as well as the Axin protein itself. In the absence of Wnt, the Axin complex phosphorylates the cytoplasmic transcriptional coactivator β-catenin. This leads to subsequent recognition by β-transducin repeat-containing protein, which when ubiquitinated, promotes proteosomal degradation of β-catenin. When the Axin complex and Wnt ligands combine, β-catenin phosphorylation and ubiquitination ceases. The resultant stabilisation of intracellular β-catenin facilitates its translocation from the cytoplasm to the nucleus, where it interacts with transcription factors of the T-cell factor/lymphoid enhancer-binding factor family, initiating transcription of Wnt-responsive genes [3].
Gene names represented: Dishevelled (DSH), adenomatous polyposis coli (APC), low-density lipoprotein receptor related proteins (LRP), casein kinase 1 (CK1a), glycogen synthase kinase 3 (GSK3- β), β-transducin repeat-containing protein (β-TrCP), T-cell factor (TCF)/lymphoid enhancer factor (LEF), Wilms tumor-suppressor (WTX), Frizzled (FZD), dickkopf homolog 1.
TGFβ1-induced epithelial-to-mesenchymal cell transition (EMT) promotes renal fibrosis [4], a characteristic pathological feature of DN. TGFβ1, integrin-linked kinase and Wnt pathways converge upon activation of β-catenin and EMT is initiated, implicating β-catenin as a master controller of multiple pathways. Furthermore, membrane bound E-cadherin/β-catenin complexes, which form the cell-to-cell junctions typical of an epithelial cell type, are lost during TGF-β1 and IGF-I induced EMT [5]. In addition, inhibition of GSK3β has been shown to prevent mesenchymal transition in human embryonic stem cells [6]. It is estimated that there are more than 1800 Wnt pathway gene targets [7] including transcription factor c-Myc, matrix metalloprotease MMP-7, endothelin-1, and fibronectin (a marker of fibrosis) [8]. Increased expression of Wnt-4 is associated with the deposition of fibronectin [9].
Variation in expression profiles of many Wnt ligands, FZD receptors and β-catenin have been reported in the unilateral ureteral obstruction (UUO) mouse model of renal injury and reduction in interstitial injury identified in response to dickkopf homolog 1 (DKK-1), a Wnt signalling antagonist [1]. Independently, DKK-1 was also shown to promote hyperglycaemia-induced mesangial matrix accumulation and renal dysfunction in rat mesangial cells [2]. Comparison of injury models is difficult; nonetheless existing evidence implicates Wnt signalling in the pathology of DN.
Our study was designed to assess association of common single nucleotide polymorphisms (SNPs) in four key genes of the canonical Wnt/β-catenin signalling pathway (AXIN2, CTNNB1, LRP5 and LRP6) with DN using a case-control design involving 1467 individuals with type 1 diabetes. These genes were selected on the basis of their functional significance as key components of the Wnt signalling pathway and from differential expression profiles derived from human kidney biopsy material from DN cases compared to no nephropathy control samples. CTNNB1 encodes the β-catenin protein, the major effector of the pathway responsible for transducing the Wnt activated signal from the cytoplasm to the nucleus and subsequent transcriptional activation of Wnt responsive genes. Both LRP5 and LRP6 are co-receptors specific to the canonical Wnt pathway enabling detection of the Wnt ligands. AXIN2 is a key component of the destruction complex, which in the absence of a Wnt ligand, marks β-catenin for ubiquitin-dependent proteolytic degradation, thus preventing its translocation from the cytoplasm to the nucleus and subsequent transcriptional activation.
Methods
Participants
Ethics approval was obtained from the appropriate Research Ethics Committees, and written informed consent obtained prior to participation. All recruited individuals were white, had type 1 diabetes mellitus (T1D) diagnosed before 32 years of age and were born in the UK or Ireland. Patients (n = 718) and controls (n = 749) were from the Warren 3/UK Genetics of Kidneys in Diabetes (GoKinD) and all-Ireland collections [10]. The definition of DN in cases was based on development of persistent proteinuria (>0.5 g protein/24 h) at least 10 years after diagnosis of T1D, hypertension (blood pressure>135/85 mmHg or treatment with antihypertensive agents) and associated diabetic retinopathy. Controls were individuals with T1D for at least 15 years with normal urinary albumin excretion rates and no evidence of microalbuminuria on repeated testing. In addition, control subjects had not been prescribed antihypertensive drug treatment avoiding possible misclassification of diabetic individuals as ‘control phenotypes’ when the use of antihypertensive treatment may have reduced urinary albumin excretion into the normal range. Individuals with microalbuminuria were excluded from either case or control groups since it is not possible to confidently assign a case or control definition upon such individuals with microalbuminuria as their urinary albumin excretion may either regress or progress over time [11].
SNP selection and genotyping
Tag SNPs for each gene investigated were selected from Phase III, release 2 HapMap (http://www.hapmap.org) CEPH data (Utah residents with ancestry in northern and western Europe; CEU) using Haploview (http://www.broadinstitute.org/haploview) to visualise linkage disequilibrium. Tag SNPs were selected using multimarker tagging where r2>0.8 (LOD threshold 3.0) for all downloaded SNPs with a minor allele frequency (MAF) ≥5%, genotype call rate ≥95%, and no deviation from Hardy–Weinberg equilibrium (HWE). Additional potentially functional SNPs were identified using the Ensembl genome browser (www.ensembl.org).
Genotyping was performed by MassARRAY iPLEX (Sequenom, San Diego, CA, USA) or Taqman 5′ nuclease (Applied Biosystems, Foster City, CA, USA) assays according to the manufacturers' instructions. Quality filters for exclusion of SNPs included call rates below 95% and deviation from HWE (P<0.001). DNA samples were excluded if missing genotypes exceeded 10%. Other quality control measures included parent/offspring trio samples, duplicates on plates, random sample allocation to plates, independent scoring of problematic genotypes by two individuals and re-sequencing of selected DNAs to validate genotypes.
Statistical analysis
Clinical characteristics of cases and controls were compared using the z-test for large independent samples and the χ2 test. Association analyses were performed using PLINK (version 1.07; http://pngu.mgh.harvard.edu/~purcell/plink/). Initially a χ2 test for trend (1 df) was used with stratification by collection centre. Logistic regression analysis was performed on each SNP with terms for potential confounders (collection centre, sex, duration of T1D and HbA1c) included in the model. The level of statistical significance was set at 5% and adjustment for multiple testing performed by permutation test (n = 100,000). In silico replication of the most significant SNPs was sought through data extracted from the US Genetics of Kidneys in Diabetes (US GoKinD) study [12] available on dbGAP (http://www.ncbi.nlm.nih.gov/gap, dataset phs000018.v2) which was based on a stratified analysis of 935 DN cases and 944 no nephropathy controls. The US GoKinD genotyping was performed on the Affymetrix 5.0 SNP array (Affymetrix, Santa Clara, CA, USA). Although only one of the four most significant SNPs identified from this study was genotyped directly on this platform, surrogate markers in high LD based on 1000 Genomes pilot data (http://www.1000genomes.org/), were used as a proxies for the other three. Potential gene-gene interactions between the most significant SNPs were assessed using likelihood ratio χ2 tests in the logistic regression.
Results
The clinical characteristics of the DN cases (n = 651) and diabetic controls (n = 700) genotyped in this study which met quality control filters are listed in Table 1. The average genotyping rate was 98.6%. There were more males, higher mean HbA1c and blood pressure values (despite the use of antihypertensive treatment) in the case group compared with the control group. All comparisons were significant at P<0.001 with the exception of age at diagnosis which did not differ significantly between groups. Approximately one quarter of cases (25.4%) had end-stage renal disease (ESRD).
We excluded 116 samples (67 patients and 49 controls) from the analysis with ≥10% missing genotypes. A total of 31 SNPs were genotyped, 28 using MassARRAY iPLEX technology, and 3 SNPs by Taqman 5′ nuclease assay in 651 cases and 700 controls (Table 2). The average call rate for all SNPs analysed was 98.65%. The genotype distribution for each SNP did not deviate significantly from HWE in either cases or controls. No duplicate or Mendelian inconsistencies were observed.
Single marker testing stratified by collection centre identified one SNP (rs1337791 in LRP6) significantly associated and three SNPs (rs2075241 in LRP6; rs3736228 and rs491347 both in LRP5) marginally associated with DN. Following logistic regression analysis adjusted by collection centre, duration of T1D, and average HbA1c as covariates, a single SNP (rs1337791) was significantly associated with DN (OR = 0.74; CI: 0.57–0.97; P = 0.028; Table 2), however this was not maintained following correction for multiple testing. Haplotype analysis did not improve association over single marker analysis. Subgroup analyses on the basis of ESRD status showed no stronger association with ESRD compared to those calculated for DN. Follow-up evaluation of the top 4 significant associations in this study were not supported within the independent US GoKinD dataset (Table 3).
Having no prior hypotheses about interactions between SNPs within the genes analysed in this study, we assumed a more stringent level of significance (P<0.01) in assessing interactions. No evidence supporting interaction was observed (Table 4 and Table S1).
Discussion
Expansion of the mesangium is a well recognised histological feature of DN that occurs early in renal dysfunction. Lin and colleagues [2] showed that reduced expression of the endogenous inhibitor, DKK-1, which binds Wnt co-receptors LRP5/6, decreased diabetes-induced glomerular injury preventing hyperglycaemia-induced mesangial cell dysfunction. This homeostatic interaction between β-catenin and DKK-1 in mesangial cell function may result in variable β-catenin activity under different physiological and pathological conditions in renal tissue. In addition, they reported modulation of DKK-1, TGF-β1, and fibronectin expression through RNA interference suggesting a novel potential therapeutic target for DN [2].
Our study focused on the canonical Wnt pathway although other Wnt pathways exist, including Wnt/Ca2+ pathway and the planar cell polarity pathway which could also conceivably contribute to the pathogenesis of DN. The genes assessed in this study were chosen on the basis of gene expression profiles derived from human kidney biopsy samples of DN (unpublished data). Although the four genes investigated did not demonstrate significant association following correction for multiple testing, independent replication or provide evidence of significant interaction, additional untested variants within the canonical and non-canonical Wnt pathways remain untested. To our knowledge, this is the first association study in DN to assess common variation within key genes regulating the Wnt pathway. Interestingly, novel splice variants were proposed to modulate Wnt pathway gene expression through mRNA stability and enhanced translational efficiency [13]. Since our study focused only on common variants, untyped, highly penetrant rare variants within these genes could also be associated with DN. Based on sample sizes used here and considering Bonferroni correction for 31 comparisons, our study provides 90% power to identify an allele with an odds ratio of 1.69, 1.50, 1.44 and 1.42 whose frequency in controls is 10%, 20%, 30% and 40% respectively. However, this study has insufficient power to detect effect sizes of smaller magnitude with odds ratios of 1.2/1.3 which are more often seen in common complex diseases (Table 5). Other factors such as copy number variation or epigenetic mechanisms (e.g. DNA methylation, microRNAs) may alter gene function affecting these pathways, contributing to disease risk.
In conclusion, we found no strong association between common variants in the CTNNB1, AXIN2, LRP5, and LRP6 genes and DN. Further work to investigate other members of the Wnt/β-catenin or non-canonical pathways may identify potential risk factors for genetic susceptibility to DN.
Supporting Information
Table S1.
Assessment of gene-gene pair-wise interactions. P values for gene-gene interactions were calculated from likelihood ratio χ2 tests in the logistic regression with adjustment for centre, gender, duration of type 1 diabetes and HbA1c. Three attained significance at the P<0.01 level and are underlined. Were the comparisons to have been independent then four would have been expected to attained significance at the P<0.01 level by chance purely as a consequence of multiple testing.
https://doi.org/10.1371/journal.pone.0023904.s001
(DOC)
Acknowledgments
We thank Dr Denise Sadlier, University College Dublin, for providing DNA samples from cases and controls from the Republic of Ireland.
The Warren 3/UK GoKinD Study Group was jointly funded by Diabetes UK and the Juvenile Diabetes Research Foundation and includes the following individuals: Belfast: Professor A. P. Maxwell, Dr A. J. McKnight, Dr D. A. Savage; Edinburgh: Dr J. Walker; London: Dr S. Thomas, Professor G. C. Viberti; Manchester: Professor A. J. M. Boulton; Newcastle: Professor S. Marshall; Plymouth: Professor A. G. Demaine and Dr B. A. Millward; Swansea: Professor S. C. Bain.
Author Contributions
Conceived and designed the experiments: DAS APM JKC AJM. Performed the experiments: DHK AJM. Analyzed the data: DHK CCP GJM. Contributed reagents/materials/analysis tools: AJM DAS APM JKC. Wrote the paper: DHK DAS APM CCP GJM. Reviewed and approved the final manuscript: DHK DAS CCP AJM JKC APM GJM.
References
- 1. He W, Dai C, Li Y, Zeng G, Monga SP, et al. (2009) Wnt/beta-catenin signaling promotes renal interstitial fibrosis. J Am Soc Nephrol 20: 765–776.
- 2. Lin CL, Wang JY, Ko JY, Huang YT, Kuo YH, et al. (2010) Dickkopf-1 promotes hyperglycemia-induced accumulation of mesangial matrix and renal dysfunction. J Am Soc Nephrol 21: 124–135.
- 3. MacDonald BT, Tamai K, He X (2009) Wnt/beta-catenin signaling: components, mechanisms, and diseases. Dev Cell 17: 9–26.
- 4. Liu Y (2010) New insights into epithelial-mesenchymal transition in kidney fibrosis. J Am Soc Nephrol 21: 212–222.
- 5. Ivanova L, Butt MJ, Matsell DG (2008) Mesenchymal transition in the kidney collecting duct epithelial cells. Am J Physiol Renal Physiol 294: F1238–F1248.
- 6. Ullmann U, Gilles C, De Rycke M, Van de Velde H, Sermon K, et al. (2008) GSK-3-specific inhibitor-supplemented hESC medium prevents the epithelial-mesenchymal transition process and the up-regulation of matrix metalloproteinases in hESCs cultered in feeder-free conditions. Mol Hum Reprod 14: 169–179.
- 7. Vlad A, Rohrs S, Klein-Hitpass L, Muller O (2008) The first five years of the Wnt targetome. Cell Signal 20: 795–802.
- 8. Gradl D, Kuhl M, Wedlich D (1999) The Wnt/Wg signal transducer beta-catenin controls fibronectin expression. Mol Cell Biol 19: 5576–5587.
- 9. Surendran K, McCaul SP, Simon TC (2002) A role for Wnt-4 in renal fibrosis. Am J Physiol Renal Physiol 282: F431–441.
- 10. Kavanagh D, McKay GJ, Patterson CC, McKnight AJ, Maxwell AP, et al. (2011) Association analysis of Notch pathway signalling genes in diabetic nephropathy. Diabetologia 54: 334–338.
- 11. Perkins BA, Ficociello LH, Silva KH, Finkelstein DM, Warram JH, et al. (2003) Regression of microalbuminuria in type 1 diabetes. N Engl J Med 348: 2285–2293.
- 12. Pezzolesi MG, Poznik GD, Mychaleckyj JC, Paterson AD, Barati MT, et al. (2009) Genome-wide association scan for diabetic nephropathy susceptibility genes in type 1 diabetes. Diabetes 58: 1403–1410.
- 13. Hughes TA, Brady HJ (2005) Expression of axin2 is regulated by the alternative 5′-untranslated regions of its mRNA. J Biol Chem 280: 8581–8588.