GENETICS OF NEPHROTIC SYNDROME IN SINGAPORE PAEDIATRICS PATIENTS

160 451 0
GENETICS OF NEPHROTIC SYNDROME IN SINGAPORE PAEDIATRICS PATIENTS

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

Thông tin tài liệu

GENETICS OF NEPHROTIC SYNDROME IN   SINGAPORE PAEDIATRICS PATIENTS        NG JUN LI  (BSC), NUS      A THESIS SUBMITTED FOR THE   DEGREE OF MASTERS OF SCIENCE      DEPARTMENT OF PAEDIATRICS  NATIONAL UNIVERSITY OF SINGAPORE        2012                Acknowledgements  I would like to express my thankfulness to:  My supervisor, Professor Yap Hui Kim, for her guidance throughout these years.    My co‐supervisor, Associate Professor Heng Chew Kiat, for his statistical advices.    Dr Ng Kar Hui, for her continuous help and support.    Dr Isaac Liu and Dr U Mya Than, for their help in collecting the clinical data of the  patients.    Dr Leow Koon Yeow, for his technical advices on the use of high resolution melting for  screening.    Chan Chang Yien, Joni, Chong, Liang Ai Wei, Lauretta Low, Sun Zijin, Seah Ching Ching  and Toh Xue Yun, for their company and help that they provided.    My family for their continuous support and encouragement.    This work was supported by grants NKFRC/2007/09 from National Kidney Foundation    and NMRC IRG07 May 104 from the National Medical Research Council, Singapore. i    Table of contents  1.  Introduction ................................................................................................................ 1  1.1.  Nephrotic syndrome ............................................................................................ 1  1.1.1.  Epidemiology and histological features of childhood NS ............................. 1  1.1.2.  Clinical course of childhood NS ..................................................................... 3  1.1.3.  Pathogenesis of nephrotic syndrome ........................................................... 7  1.2.  Podocytes: Their role in nephrotic syndrome .................................................... 11  1.2.1.  The slit diaphragm ...................................................................................... 12  1.2.2.  Cytoskeletal proteins .................................................................................. 18  1.2.3.  Cell matrix adhesion complexes ................................................................. 19  1.2.4.  Podocytes and Nephrotic syndrome .......................................................... 20  1.3.  Genetics of nephrotic syndrome ........................................................................ 22  1.3.1.  NPHS1: The gene encoding for nephrin ...................................................... 28  1.3.2.  NPHS2: The gene encoding for podocin ..................................................... 32  1.3.3.  The two closely related partners: NPHS1 and NPHS2 ................................ 39  1.4.  Gaps in current knowledge ................................................................................ 43  1.5.  Objectives of the study ...................................................................................... 44  2.  Methods and Materials ............................................................................................. 46  2.1.  Study Subjects .................................................................................................... 46  ii    2.2.  Data collection.................................................................................................... 47  2.3.  Genomic DNA (gDNA) extraction ....................................................................... 48  2.4.  Primer design and melting domain analysis ...................................................... 48  2.5.  PCR amplification ............................................................................................... 53  2.6.  High Resolution Melting ..................................................................................... 54  2.7.  Direct Sequencing .............................................................................................. 55  2.8.  Genotyping by tetra‐primers ARMS PCR ............................................................ 56  2.9.  Genotyping by RFLP ........................................................................................... 60  3.  2.10.  Statistical analysis ........................................................................................... 62  2.11.  Prediction of the effect of the genetic variants ............................................. 63  2.12.  Protein sequence alignment ........................................................................... 64  Results ....................................................................................................................... 66  3.1.  Clinical Characteristics ........................................................................................ 66  3.2.  Genetic variants in NPHS1 .................................................................................. 69  3.2.1.  Screening for variants in NPHS1 using HRM ............................................... 69  3.2.2.  Statistical analysis of the genetic variants in NPHS1 .................................. 74  3.2.3.  Prediction of the effect of the genetic variants .......................................... 81  3.3.  Genetic variants in NPHS2 .................................................................................. 85  3.3.1.  Screening of NPHS2 using direct sequencing ............................................. 85  iii    3.3.2.  Statistical analysis of the genetic variants in NPHS2 .................................. 87  3.3.3.  Prediction of the effect of the genetic variants .......................................... 91  3.4.  Phenotype‐genotype associations ..................................................................... 93  3.5.  Analysis of composite genotypes of genetic variants in NPHS1 and NPHS2 ..... 97  3.6.  Contribution of rare alleles to nephrotic syndrome ........................................ 101  4.  Discussion................................................................................................................ 112  4.1.  Polymorphisms and their association with nephrotic syndrome .................... 114  4.2.  Phenotype‐genotype correlations of the genetic variants .............................. 118  4.3.  Gene‐gene interaction of NPHS1 and NPHS2 .................................................. 120  4.4.  Contribution of rare variants to nephrotic syndrome ..................................... 122  5.  Conclusion ............................................................................................................... 127  6.  Reference ................................................................................................................ 128        iv    Summary   Background  Nephrotic syndrome (NS) is the most common glomerulonephritis in childhood, 20% of  which  are  steroid‐resistant.  Mutations  in  podocyte‐related  genes  have  been  shown  to  cause  idiopathic  NS  in  Caucasian  populations,  whereas  Asian  studies  are  lacking.  The  aims  of  this  study  are  firstly  to  identify  genetic  variants  in  nephrin(NPHS1)  and  podocin(NPHS2), two important podocyte‐associated genes, in paediatric patients with  idiopathic NS in Singapore, secondly to determine the presence of disease associations  with  these  genetic  variants,  and  lastly  to  investigate  if  there  are  any  gene‐gene  interactions.     Methods  Genetic  screening  of  NPHS1  and  NPHS2  was  performed  on  121  patients  (97Chinese,  24Malays) with disease onset at a mean age of 5.69±4.43 years. High Resolution Melting  was used to detect genetic variants in NPHS1, and these were validated by bi‐directional  sequencing. Direct sequencing was used to detect genetic variants in NPHS2. All genetic  variants  identified  were  genotyped  in  cord  blood  controls  (125Chinese;  96Malays)  by  Restriction  Fragment  Length  Polymorphism,  Tetra‐primer  Amplification  Refractory  Mutation  System  PCR  and  sequencing.    Data  from  the  Singapore  Genome  Variation  Project  involving  healthy  individuals  (96Chinese;  89Malays)  and  SNP  arrays  from  the  Genome  Institute  of  Singapore  involving  1682  Chinese  controls  were  used  for  our  analysis.  Statistical  analysis  was  performed  using  SNPstats  and  SPSS  v17.  Multiple  v    logistic  regression  analysis  (codominant,  dominant,  recessive  and  log‐additive  models)  was done to determine the significance of genetic variants and phenotype associations.  Computer  prediction  programs  were  used  to  predict  the  functional  effects  of  these  genetic variants.    Results  Sixteen  genetic  variants  in  NPHS1  were  found  in  our  cohort  of  nephrotic  patients,  of  which  5  was  novel.  c.294C>T,  and  c.2289C>T  were  significantly  associated  with  NS  in  both  Chinese  and  Malay  patients,  whereas  c.349G>A  was  significantly  associated  with  NS in only Chinese patients. c.349G>A resulted in an amino acid substitution of glutamic  acid  to  lysine  at  position  117.  c.294C>T,  c.349G>A  and  c.2289C>T  have  been  found  to  affect the ESE and/or ESS sites.    We found 8 genetic variants in NPHS2, of which 1 was novel. In our Chinese patients, c.‐ 51G>T was significantly associated with NS. c.288C>T was observed to have a protective  effect  against  NS,  whereas  c.‐51G>T  and  c.1038A.G  were  associated  with  steroid  resistance. c.1038A>G was also associated with cyclosporine resistance.      Binary  logistic  regression  showed  that  the  composite  genotypes  of  GG/CC/CT  and  GT/CC/CT  for  NPHS2  c.‐51G>T  and  c.288C>T  and  NPHS1  c.2289C>T  were  significantly  associated  with  NS.  The  composite  genotypes  of  GG/CC/TT  and  GT/CC/CC  were  also  associated with steroid resistance.  vi    Additionally,  we  have  shown  a  significant  accumulation  of  rare  variants  in  nephrotic  patients  compared  to  controls,  and  also  in  nephrotic  patients  with  poor  prognosis  compared to those who responded well to therapy.    Conclusion  Genetic  variants  in  podocyte‐related  genes  were  present  in  paediatric  patients  with  idiopathic  NS  in  Singapore,  and  they  occurred  at  frequencies  which  differed  from  Caucasian populations. Phenotype association studies implicated NPHS2 polymorphisms  with  steroid  and  cyclosporine  resistance,  whereas  rare  variants  accumulation  was  associated  with  poor  prognosis.  Additionally,  gene‐gene  interactions  between  NPHS1  and NPHS2 were observed in our nephrotic population. Functional studies are required  to better understand the mechanisms of the genetic variants in the pathogenesis of NS.       vii    List of Tables  Table 1: Summary of genetic involvement in pathogenesis of nephrotic syndrome. ...... 24  Table 2: NPHS1 and NPHS2 genetic variants identified in familial and sporadic nephrotic  syndrome. ................................................................................................................. 25  Table 3: Primers used for the screening of NPHS1 using HRM. ....................................... 50  Table 4: Primers used for the screening of NPHS2 using direct sequencing. ................... 52  Table 5: Primers used in tetra‐ARMS PCR for the genotyping of NPHS1 SNPs. ............... 58  Table 6: Restriction enzymes used for the genotyping of NPHS1 SNPs. .......................... 61  Table 7: General characteristics of Chinese and Malay patients with idiopathic sporadic  nephrotic syndrome and/or FSGS. ............................................................................ 68  Table 8: Genetic variants identified in NPHS1 using HRM. ............................................... 70  Table 9: Allele frequencies of NPHS1 SNPs in Chinese (n=97). ......................................... 77  Table 10: Genotype frequency for NPHS1 SNPs in Chinese (n=97). ................................. 77  Table 11: Association analysis of NPHS1 genotypes with nephrotic syndrome in Chinese.  ................................................................................................................................... 78  Table 12:Allele frequencies of NPHS1 SNPs in Malay (n=24). .......................................... 79  Table 13: Genotype frequencies of NPHS1 SNPs in Malay (n=24). .................................. 79  Table 14: Association analysis of NPHS1 genotypes with nephrotic syndrome in Malay  (n=24). ....................................................................................................................... 80  Table 15: Prediction of the effect of amino acid substitution of the various genetic  variants on the protein using SIFT and PolyPhen2. .................................................. 82  Table 16: Effect of the individual NPHS1 SNPs on ESE/ESS sites. ..................................... 84  Table 17: Genetic variants identified in NPHS2 using direct sequencing. ........................ 85  viii    Table 18: Allele frequencies of NPHS2 variants in Chinese (n=97). .................................. 89  Table 19: Genotype frequencies of NPHS2 variants in Chinese (n=97). ........................... 89  Table 20: Association analysis of c.‐51G>T and c.288C>T in Chinese patients (n=97). .... 90  Table 21: Effect of the individual NPHS2 SNPs on ESE/ESS sites. ..................................... 92  Table 22: Genotype‐phenotype association in Chinese patients for NPHS2. .................. 95  Table 23: Allele frequencies for NPHS2 SNPs with phenotype‐genotype associations in  Chinese patients. ....................................................................................................... 96  Table 24: Composite genotype analysis of patients, with race as co‐variant. ................. 98  Table 25: Composite genotype analysis for SRNS patients, with race as co‐variant...... 100  Table 26: Rare NPHS1 and NPHS2 variants identified in NS patients and their minor allele  frequencies in patients and controls. ..................................................................... 102  Table 27: Non‐synonymous NPHS1 and NPHS2 rare variants identified in nephrotic  syndrome patients and their predicted effects ...................................................... 103  Table 28: Rare variants accumulation in nephrotic syndrome patients and controls ... 104  Table 29: Accumulation of rare variants in patients with poor prognosis. .................... 106  Table 30: Clinical characteristics of patients with rare variants. .................................... 107    ix    List of Figures  Figure 1: Histology of the glomerulus in (A) MCNS and (B) FSGS (Periodic acid‐Shiff stain).  ..................................................................................................................................... 2  Figure 2: MCNS and FSGS are the most common histological patterns seen in steroid‐ resistant nephrotic syndrome. .................................................................................... 5  Figure 3: The glomerular filtration barrier. ......................................................................... 9  Figure 4: Distribution of the histological profile in the nephrotic patients. ..................... 69  Figure 5:  Normalized high resolution melting curves and the corresponding difference  plots. .......................................................................................................................... 71  Figure 6: Electrophoretograms of the various genetic variants. ...................................... 72  Figure 7: Genotyping results of c.2289C>T and c.3315A>G. ............................................ 73  Figure 8: Electrophoretograms of the various genetic variants. ...................................... 86  Figure 9: Pedigree of Patient 94. .................................................................................... 111               x    List of Abbreviations  ACTN4   Alpha actinin 4  AIC  Akaike Information    ARMS    Amplification refractory mutation system  CD2AP   CD2‐associated proteins  CNF    Congenital nephrotic syndrome of the Finnish type  CNI    Calcineurin inhibitors  CNS    Congenital nephrotic syndrome  CsA    Cyclosporine   DAG    Diacylglycerol  DMS    Diffuse mesangial sclerosis  EDTA    Ethylenediaminetetraacetic  ESE    Exon splicing enhancer  ESRD    End stage renal disease  ESS    Exon splicing silencer  FGS    Focal global sclerosis  FMPGN  Focal mesangial proliferative glomerulonephritis  FSGS    Focal segmental glomerulosclerosis  GBM    Glomerular basement membrane  gDNA    Genomic DNA  HRM     High resolution melting  IP3  Inositol 1,4,5‐triphosphate    xi    IQGAP‐1  IQ motif–containing GTPase‐activating protein 1  ISKDC    International Study of Kidney Diseases in Children  MCNS    Minimal change nephrotic syndrome  MPGN    Membranoproliferative glomerulonephritis  NS  Nephrotic syndrome    NUH    National University Hospital of Singapore  PCR    Polymerase chain reaction  PFS    Potentially Functional SNP  PI3K    Phosphoinositide 3‐OH kinase  RFLP    Restriction fragment length polymorphism  Rs number  RefSNP accession ID  SD  Slit diaphragm    SDNS    Steroid dependent nephrotic syndrome  SGVP    Singapore Genome Variation Project  SIFT    Sorting intolerant from tolerant  SNP    Single nucleotide polymorphism   SRNS    Steroid resistant nephrotic syndrome  SSNS    Steroid sensitive nephrotic syndrome  suPAR    Serum soluble urokinase receptor  TRP    Transient receptor potential  USF    Upstream stimulatory factor    xii    List of Appendices  Appendix I: Clinical characteristics and genotype profile of the patients……………………146  Appendix II: Hardy‐Weinberg Results for NPHS1 and NPHS2 variants………………………..158  Appendix III: Linkage disequilibrium results for NPHS1 and NPHS2 variants……………….163  Appendix IV: Protein sequence alignment for nephrin and podocin…………………………..167              xiii    List of conference abstracts and awards  6th Congress of Asian Society for Paediatrics Research, Taipei, 2010  Title: Comparison of NPHS2 polymorphisms and their associations with end‐stage renal  disease in Chinese and Malay children.  Award: Best Poster Award    11th Asian Congress of Paediatrics Nephrology, Fukuoka, 2011  Title: Nephrin gene variants may account for ethnic differences in susceptibility to  sporadic Nephrotic Syndrome in Singapore children.  Award: Travel grant    8th Congress of Asian Society for Paediatrics Research, Seoul, 2012  Title: Multiple Rare Alleles in Nephrin and Podocin Contribute to Poor Prognosis in  Childhood Nephrotic Syndrome in Singapore.  Nominated for Young Investigator Award  xiv    1. Introduction  1.1. Nephrotic syndrome  Nephrotic  syndrome  (NS)  is  a  common  cause  of  kidney  disease  in  children.    It  is  characterized  by  heavy  proteinuria,  hypoalbuminaemia,  oedema  and  hyperlipidaemia  (Yu et al., 2005). There have been various definitions used to describe nephrotic‐range  proteinuria,  including  urinary  protein  excretion  ≥3  g/day/1.73m2  or  a  spot  urinary  protein:creatinine ratio ≥0.2 g/mmol (Yap and Lau, 2008).    NS in children can be congenital, which is presented at birth or during first 3 months of  life,  or  presented  in  later  part  of  their  life.  Childhood  NS  is  due  to  either  primary  or  secondary causes. Some examples of primary causes are genetic disorders or defects in  the  glomerular  filtration  barrier.  Infections,  drug  usage  and  immunological  or  allergic  disorders are some of the secondary causes of childhood NS (Eddy and Symons, 2003).  In this literature review, the main focus would be primary NS in children.    1.1.1. Epidemiology and histological features of childhood NS  The  reported  incidence  of  idiopathic  NS  is  2  to  7  cases  per  100  000  children.  It  has  a  prevalence of approximately 16 cases per 100 000 (Eddy and Symons, 2003).  The two  distinct  histological  features  of  primary  idiopathic  NS  in  children  are  minimal  change  nephrotic syndrome (MCNS), and focal segmental glomerulosclerosis (FSGS) (Figure 1).       1      Fiigure 1: Histtology of the e glomerulus in (A) MCN NS and (B) FFSGS (Period dic acid‐Shifff  sttain).    MCNS,  M which h  is  characte erized  by  minor  m morphhological  alteration  in  p podocytes,  is  the  most  m commo on  pathologiical  type  in  childhood  nnephrotic  syyndrome.    (Zhu  et  al.,  2 2009,  Laahdenkari  et  al.,  2005).  On  light  microscopy,  m tthe  glomeru ulus  for  MCN NS  looks  no ormal,  whereas  w elecctron  microscopy  show ws  the  preseence  of  efffacement  off  the  glomeerular  epithelial foo ot processes during relap pses of the ddisease (Kum mar et al., 20 003)    FSSGS  is  characterized  byy  a  commo on  pattern  oof  glomerulaar  injury  that  is  defineed  by  se egmental sclerotic lesions invading  a populatioon of glomerruli (Tsukagu uchi et al., 2 2002).  This  is  often  associated  with  tubulaar  scarring  aand  interstittial  fibrosis.  Five  varian nts  of  FSSGS  have  been  defined d  according  to  their  paathology:  co ollapsing,  ceellular,  tip  leesion,  perihilar and  not‐otherw wise‐specified d (Thomas eet al., 2006)).  FSGS is also an impo ortant  mary nephrottic syndrome in adults. FFSGS carriess a poor proggnosis with more  caause of prim th han 25% of cchildren and d adults proggressing to eend stage reenal diseasee (ESRD) in w within  five years of  onset of NSS (McKenzie  et al., 20077). It has beeen observed d that there  is an  2    increased incidence of FSGS in both children and adults, especially in black individuals in  the United States (Eddy and Symons, 2003, Mollet et al., 2009).     1.1.2. Clinical course of childhood NS  Steroids are the mainstay of therapy for children with NS. The initial steroid treatment is  prednisone 60 mg/m2 per day for 4 weeks, with the maximum dosage to be 80mg. This  is followed by prednisone 40 mg/m2 on alternate days for 4 weeks, with a steroid taper  over  3  to  6  months  (Eddy  and  Symons,  2003).  Steroid  sensitive  nephrotic  syndrome  (SSNS) is defined as patients being able to enter remission in response to corticosteroid  treatment alone. Steroid resistant nephrotic syndrome (SRNS) is the inability to induce  remission  after  8  weeks  of  corticosteroid  treatment.    Steroid  dependent  nephrotic  syndrome  (SDNS)  is  the  initial  response  to  corticosteroid  treatment  by  entering  complete remission but the development of relapse either while still receiving steroids  or  within  2  weeks  of  discontinuation  of  treatment  following  a  steroid  taper.  Patients  with  SDNS  require  the  continued  low‐dose  treatment  with  steroids  to  prevent  development of relapse (Gbadegesin and Smoyer, 2008).     A  multicenter  International  Study  of  Kidney  Diseases  in  Children  (ISKDC)  study  was  performed on more than 521 children with idiopathic NS in the late 70s and early 80s.   The  study  revealed  that  approximately  80%  of  children  with  idiopathic  NS  respond  to  steroid therapy. Steroid responsiveness was observed in 93% of the children with MCNS,  compared to only 30% of children with FSGS (Gbadegesin and Smoyer, 2008).   3    Twenty percent of children with NS are resistant to steroid therapy. They fail to respond  to  corticosteroids  (Schwaderer  et  al.,  2008).  According  to  the  study  by  ISKDC,  for  the  103  children  who  were  steroid‐resistant,  MCNS,  FSGS  and  membranoproliferative  glomerulonephritis  (MPGN)  were  the  most  prevalent  lesions  present,  each  accounting  for approximately 25% of the histological findings (ISKDC, 1981) (Figure 2). In Singapore,  renal biopsies of 47 children with SRNS showed that MCNS (30%) and FSGS (49%) were  the  two  main  causes  of  SRNS  (Yap,  2005)  (Figure  2).  The  prognosis  for  children  with  SRNS is often poor. They usually rapidly progress to end stage renal disease within 10  years from diagnosis (Kitamura et al., 2006, Schultheiss et al., 2004).  These children are  also more prone to more complicated clinical and therapeutic course, which will result  in ESRD in 30 to 40% (Schwaderer et al., 2008) .        4      Fiigure 2: MC CNS and FSG GS are the m most commo on histologiccal patternss seen in steeroid‐ re esistant nep phrotic syndrome.   CGN:  Chronicc  glomerulonephritis.  DMH:  D Diffusee  mesangial  hypercellullarity.  FGS:  Focal  global  glome erulosclerosis.  FSGS:  Focal  segmenttal  glomeru ulonephritis.  MCNS:  Min nimal  ch hange  neph hrotic  syndrrome.  Mem mbGN:  Mem mbranous  glomerulonep phritis.  MessPGN:  Mesangial  M proliferativve  glomerulonephritiis.  MPGN N:  Membranoproliferrative  glomerulonep phritis.                       5    Despite  the  fact  that  most  patients  with  MCNS  are  steroid  sensitive,  the  treatment  of  MCNS  remains  a  great  challenge  to  nephrologists  as  up  to  70%  of  them  experienced  frequent  relapses.  Furthermore,  some  patients  could  not  achieve  complete  remission  with steroid treatment (steroid‐resistant) (Koskimies et al., 1982, Tarshish et al., 1997)  and  therefore  have  to  be  treated  with  immunosuppressive  drugs  such  as  tacrolimus,  cyclophoshamide and cyclosporine (Li et al., 2009, Hino et al., 1998, Gulati et al., 2008).  Nevertheless, the prolonged use of corticosteroids results in adverse side effects such as  growth  retardation,  obesity,  infections,  hypertension,  osteoporosis  and  cataracts.  In  addition,  use  of  alternative  drugs  such  as  alkylating  agents  and  calcinuerin  inhibitors  carry  with  it  the  potential  risk  of  infections,  infertility,  malignancy  and  nephrotoxicity.  This indicated a need for better therapeutic approaches (Kyrieleis et al., 2009, Hodson  et al., 2004, Gipson et al., 2009).    Interestingly, there is a small group of patients who responded well to steroids initially  but later became resistant to steroids in their clinical course. This late steroid resistance  is a rare phenomenon, and its pathophysiology is still not well studied. Patients with late  steroid resistance have clinical findings quite similar to that of SSNS rather than SRNS.   They  also  respond  to  cytostatic  drugs  with  good  prognosis  of  stable  maintenance  of  renal function (Schwaderer et al., 2008).      6    1.1.3. Pathogenesis of nephrotic syndrome  The cause of idiopathic NS is usually due to either structural defects or immunological  imbalance.     Immunology and nephrotic syndrome  Idiopathic NS was histologically thought to be an immunological disorder. Infections and  immunological disorders, such as Epstein‐Barr virus infection (Blowey, 1996, Araya et al.,  2006) and Hodgkins’ lymphoma (Peces et al., 1991) which causes lymphadenopathy and  lymphoproliferation,  were  found  to  be  associated  with  NS.  Efficacy  using  immunosuppressants to alleviate proteinuria and induce remission further suggested an  immunologic basis in the pathogenesis of NS.     MCNS  is  hypothesized  to  be  caused  by  immunological  imbalance  as  MCNS  patients  responded  robustly  to  immunosuppression  and  their  relapse  is  usually  due  to  immunological events such as infection and allergic reaction (Shono et al., 2009). The T‐ cell  mediated  immunity  is  hypothesized  to  be  involved  in  the  pathogenesis  of  MCNS.  The  imbalance  of  T‐helper  cells  in  MCNS  patients  leads  to  an  increase  in  cytokine  secretions.  These  cytokines  serve  as  circulating  factors  that  alter  the  glomerular  permeability of the glomerular capillary wall (Shono et al., 2009, Lahdenkari et al., 2004).  However, the molecular identification of these humoral factors and the mechanisms by  which they compromise the filtration barrier remain unknown (Shono et al., 2009).     7    Recurrence  of  proteinuria  occurs  in  approximately  30  to  40%  of  FSGS  patients  after  renal transplantation. Proteinuria can recur immediately after transplantation or can be  delayed  for  days  to  weeks.  Plasmapheresis  or  immunoadsorption  reduces  proteinuria  and  may  also  stabilize  renal  function  in  recurrent  FSGS.  Plasma  from  these  recurrent  FSGS  patients,  when  injected  into  experimental  animals  can  induce  proteinuria  or  albuminuria. All these suggested the presence of plasma circulating factor that may be  involved in glomerular damage and the initiation of glomerular events which contribute  to the development of proteinuria in FSGS (Sharma et al., 1999, Sharma et al., 2004) .  Cardiotrophin  like  cytokine‐1  has  been  identified  as  a  possible  candidate  for  the  FSGS  permeability  factor(McCarthy  et  al.,  2010).  More  recently,  serum  soluble  urokinase  receptor (suPAR) was identified as a circulating factor as a cause for both primary and  recurrent  FSGS.  Elevated  suPAR  was  observed  in  patients  with  FSGS  and  this  elevated  suPAR  concentration  was  associated  with  an  increased  risk  of  recurrent  FSGS  after  transplantation (Wei et al., 2011).    Glomerular filtration barrier and nephrotic syndrome  The glomerular filtration barrier is the principle structure within the glomerulus, which  mediates  the  filtration  of  proteins.  A  dysfunction  in  the  glomerular  filtration  barrier  results in the increased permeability to plasma proteins to the glomerular capillary wall.  This eventually leads to proteinuria, a hallmark of NS.      8    The glomerullar filtration barrier is m made up of thhree layers,  the glomeru ular endotheelium  ce ells, the glom merular basement mem mbrane (GBM M) and visceeral epitheliaal cells which are  also known as podocytess (Figure 3).      Fiigure 3: The glomerularr filtration baarrier.   The glomerullar filtration barrier is m made up of tthe capillaryy endotheliaal cells, GBM M and  w their  fo oot  processses.  The  sliit  diaphragm m  connectss  the  podo ocytes  podocytes  with  ogether.  The e  slit  diaphragm‐associiated  moleccules  includee  nephrin,  podocin,  CD D2AP,  to TRPC6 and NEPH.  The G GBM is comp posed of thee collagen tyype IV, lamiinin and hep paran  su ulphate  pro oteoglycan  agrin.  a Integrrins  are  hetterodimeric  transmemb brance  receptors  which specifi w cally connecct the TVP co omplex (talinn, vinculin and paxillin) tto laminin. TThe α  and β dystrogglycans conn nect utrophiin to agrin. ((Adapted fro om: Kriz, W.. 2005. TRPC C6 ‐ a  new podocyte gene invollved in focal segmental gglomeruloscclerosis. Tren nds Mol Med d, 11,  527‐30 (Kriz, 2005))        9    The  glomerular  endothelial  cells  are  flattened  and  highly  fenestrated.  They  form  the  interface  between  blood  and  tissue  components.  They  are  also  responsible  for  the  regulation  of  vasotone,  homeostasis  and  trafficking  of  leucocytes.  Their  role  in  the  selective  filtration  seem  to  be  insignificant  as  they  are  highly  fenestrated  and  highly  permeable  to  water  and  small  solutes  (Ballermann,  2005).  All  of  the  filtration  barrier  structures  (endothelium  cell  glycocalyx,  GBM  and  podocyte  glycocalyx)  demonstrate  anionic charges, indicating that all structures act together for the total selectivity of the  barrier.  This  means  that  the  glomerular  endothelium  could  have  a  function  for  glomerular  permeability.  Previous  studies  done  by  Jeansson  and  Haraldsson  indicated  that  digestion  with  chondroitinase  decreased  the  thickness  of  the  glomerular  endothelium glycocalyx. This resulted in increased fractional clearance for albumin due  to the change in the charge selectivity (Jeansson and Haraldsson, 2006).     The  layer  surrounding  the  endothelium  is  the  glomerular  basement  membrane.    The  GBM  is  thicker  compared  to  other  basement  membranes  because  of  the  dense  structure  of  the  extracellular  matrix  components.  This  is  to  provide  structural  support  for  the  capillary  wall  that  is  necessary  for  the  maintenance  of  the  local  high  blood  pressure.  The  GBM  is  made  up  of  mainly  collagen  type  IV,  laminins,  nidogen,  and  proteoglycans  that  contribute  to  the  selective  permeability  of  the  GBM  based  on  size  and  charge  (Levidiotis  and  Power,  2005).  Initially  the  GBM  was  thought  to  be  the  important  structure  of  the  glomerular  filtration  barrier.  Hence,  many  studies  were  focused on its structure. Structural abnormalities of the GBM may result in proteinuria  10    and  hematuria,  as  observed  in  the  X‐linked  form  of  Alport’s  syndrome  that  is  due  to  mainly collagen IVα5 chain mutations and also collagen IV α3 and α4 chains (Barker et  al., 1990). Severe or truncating mutations in the LAMB2 gene, which encodes for laminin  β2 in GBM, result in congenital NS (CNS) associated with microcoria (Zenker et al., 2004,  VanDeVoorde et al., 2006). In recent years, the identification of several novel proteins  involved  in  the  glomerular  permeability  has  identified  the  podocytes  as  an  important  constituent of the glomerular filtration barrier.     1.2. Podocytes: Their role in nephrotic syndrome  Podocytes  are  highly  specialized  and  terminally  differentiated  epithelial  cells  that  line  the  outer  surface  of  the  GBM.  Differentiated  podocytes  are  mesenchymal‐like  cells  which arise from epithelial precursors during renal development. They are made up of  three  morphological  and  functionally  different  segments,  namely  the  cell  bodies,  the  major processes and lastly the foot processes. The foot processes are derived from the  splitting of the major processes from the cell bodies. These foot processes consist of an  actin cytoskeleton which linked to the GBM in focal contacts. The foot processes form a  tight interdigitating network with their neighbouring podocytes foot processes that are  connected by a continuous membrane‐like structure known as the slit diaphragm (SD)  (Mundel and Kriz, 1995).     11    1.2.1. The slit diaphragm  The SD is an important structure that regulates the selectivity of size in the glomerular  filtration  barrier.    It  restricts  the  protein  passage  through  the  filter,  therefore  being  responsible for a largely protein free filtrate (Reiser et al., 2000). The three‐dimensional  molecular architectural of the SD observed under electron microscopy revealed that it is  made up of a convoluted network of irregularly shaped pores emanating from a central  dense region (Wartiovaara et al., 2004). The intercellular space connected by the SD is  30 to 40nm wide. The rectangular pores of the zipper have an approximate size of an  albumin  molecule  (Rodewald  and  Karnovsky,  1974).  The  SD  is  a  modified  adherens  junction,  sharing  some  typical  morphological  features,  such  as  wide  intercellar  gap,  presence of a central dense line in grazing sections, with an adherens junction (Reiser et  al., 2000).      Several  molecules  have  been  identified  to  be  associated  with  the  structure  of  the  SD.  The  first  molecule  that  was  reported  was  ZO‐1,  which  is  located  in  the  epithelial  foot  processes  at  the  points  of  insertion  of  the  SD  (Schnabel  et  al.,  1990).    Nephrin  was  discovered to be a component of the slit diaphragm (Patrakka et al., 2000). P‐cadherin  was detected at the slit diaphragm together with ZO‐1. Resier et al. suggested that P‐ cadherin  to  be  the  core  protein  of  the  slit  diaphragm  (Reiser  et  al.,  2000).  Other  molecules associated with the SD include Neph 1, FAT, TRPC6, coupled with podocin and  CD2‐associated proteins (CD2AP) linked to the actin cytoskeleton of podocytes (Figure 3)  (Reiser et al., 2000, Schwarz et al., 2001).   12    CD2AP  CD2AP  is  an  adapter  molecule  which  is  originally  identified  as  a  ligand  for  the  T‐cell  adhesion protein CD2. It is an 80kDa cytoplasmic protein that has three Src homology 3  (SH3) domains at the N terminus and a coiled domain at the C‐terminus. It is expressed  in all the tissues, except for the brain (Gigante et al., 2009). The expression of CD2AP is  limited to podocytes within the glomeruli. However, it is also expressed in the collecting  duct cells and some proximal tubular cells (Li et al., 2000).  Mice with CD2AP knockout  developed NS and later died of renal failure at 6 to 7 weeks of age (Shih et al., 1999).  Mice with CD2AP haploinsufficiency exhibited glomerular change at 9 month of age and  had  increased  susceptibility  to  glomerular  injury.  This  suggested  that  the  intracellular  degradation pathway could be impaired; implicating that CD2AP could be a determinant  of  human  susceptibility  to  glomerular  disease  (Kim  et  al.,  2003).    In  SD,  CD2AP  may  function as an adaptor protein to anchor the C‐terminal cytoplasmic domain of nephrin  and/or  podocin  to  the  actin  cytoskeleton.  CD2AP  is  also  involved  with  nephrin  and  podocin in cell signaling pathway (Mao et al., 2006).     Nephrin  Nephrin  is  a  transmembrane  protein  belonging  to  the  immunoglobulin  superfamily.  It  has  an  N‐terminal  signal  peptide,  an  extracellular  domain  that  containing  eight  immunoglobulin motifs, a fibronectin type III‐like domain, a transmembrane domain and  lastly a cytosolic C‐terminal domain (Kestila et al., 1998, Lenkkeri et al., 1999). Nephrin is  localized  within  normal  human  kidney  and  is  distinctly  localized  at  the  SD  of  the  13    glomerular podocytes (Holthofer et al., 1999, Patrakka et al., 2000). Mice with nephrin  knockout have no slit diaphragm and died at the age of birth with massive proteinuria  (Putaala et al., 2001).     Nephrin plays an important role in the signaling pathways of the podocytes which are  essential for podocytes function, survival and differentiation (Patrakka and Tryggvason,  2007).  Nephrin,  together  with  CD2AP,  associates  with  the  p85  regulatory  subunit  of  phosphoinositide  3‐OH  kinase  (PI3K).  Together  with  podocin,  the  PI3K‐dependent  AKT  signaling  was  stimulated,  controlling  the  growth,  migration  and  survival  of  podocytes  (Huber  et  al.,  2003a).  The  tyrosine  residues  at  the  cytoplasmic  end  of  nephrin  are  phosphorylated  by  Src  kinases  (Fyn‐dependent  phosphorylation).  This  signals  the  development  of  podocytes  foot  processes  which  involve  actin  microfilaments  polymerization and interconnection between proteins at slit diaphragm (Lahdenpera et  al.,  2003,  Li  et  al.,  2004).  Phosphorylation  of  nephrin  by  Fyn  initiates  a  cascade  of  effectors which inactivated pro‐apoptotic factor Bad (Huber et al., 2003a) and therefore  promotes  anti‐apoptotic  signals  (Asanuma  et  al.,  2007).  This  protects  the  podocytes  from apoptosis.     Podocin  Podocin is a 383 amino acid protein, which has a molecular size of 42kDa. It belongs to  the  stomatin  protein  family.  It  is  made  up  of  an  N‐terminal  domain,  a  short  transmembrane  domain  and  a  cytosolic  C‐terminal  domain,  forming  a  hairpin  like  14    structure (Boute et al., 2000). Podocin was found to be localized to the podocytes foot  processes membrane, at the insertion of the slit diaphragm (Schwarz et al., 2001, Roselli  et al., 2002).  Mice that were deficient in podocin exhibited massive proteinuria at birth.  The  podocyte  foot  processes  were  occasionally  observed.  These  mice  died  a  few  days  later from renal failure caused by massive mesangial sclerosis (Roselli et al., 2004). The  inactivation  of  podocin  in  the  mature  kidney  of  mouse  led  to  a  different  phenotype.  These  mice  had  progressive  renal  disease  and  showed  the  features  of  NS,  such  as  hyperlipidemia, hypertension and renal insufficiency (Mollet et al., 2009).     Podocin forms homo‐oligomers and associates with the lipid rafts (Schwarz et al., 2001).  Podocin acts as a scaffolding protein that mediates the recruitment of nephrin into the  specialized  microdomains  of  the  plasma  membrane  of  podocytes.  It  provides  nephrin  with  a  specialized  lipid  environment  which  is  necessary  for  nephrin  signal  transduction(Huber et al., 2003b).  Podocin has been found to increase the efficiency of  nephrin signaling without the recruitment of other signaling molecules. The cytoplasmic  tail  of  nephrin  binds  to  podocin  and  this  interaction  is  mediated  by  the  C‐terminal  domain of podocin (Huber et al., 2001). The phosphorylation of nephrin also increases  the binding of nephrin and podocin (Li et al., 2004).     TRPC6  TRPC6  is  a  member  of  the  transient  receptor  potential  (TRP)  superfamily  of  cation  selective ion channels. TRP (TRPC1 – TRPC7) subfamily is a group of calcium permeable  15    cation  channels  which  are  important  in  the  increase  of  intracellular  calcium  concentration  after  the  engagement  of  G‐protein  coupled  receptors  and  receptor  tyrosine kinase. They can also form homo‐ and hetero tetramers that can interact with a  variety  of  proteins.  TRPC6  is  expressed  in  podocytes  and  is  a  component  of  the  glomerular  SD  (Reiser  et  al.,  2005).    TRPC6  deficient  mice  have  been  generated.  However,  the  young  mice  that  are  deficient  in  TRPC6  did  not  show  an  overt  kidney  phenotype. This could be due to the late onset of TRPC6‐related NS; hence there could  be a need to study aged deficient mice. Another hypothesis is that the mice could only  develop  NS  in  a  stressed  condition  such  as  hypertension  in  the  absence  of  TRPC6  channel (Moller et al., 2009).     TRPC6 has been found to interact with nephrin and podocin but not CD2AP. TRPC6 was  found to be clustered and regulated by a podocin‐lipid complex. The absence of nephrin  in  a  nephrin  knockout  mouse  induced  the  expression  of  TRPC6  and  led  to  altered  cellular localization of TRPC6 (Reiser et al., 2005). TRPC6 is a target of phosphorylation  by  Fyn,  enhancing  its  channel  activity.  TRPC6  assembles  in  complex  with  nephrin  and  Fyn at the SD. This complex is partially regulated by tyrosine phosphorylation of nephrin  and/or  TRPC6  (Hisatsune  et  al.,  2004).  The  roles  of  TRPC6  in  the  podocin‐TRPC6  mechnosensor  complex  and  nephrin‐Fyn‐TRPC6  complex  suggest  that  TRPC6  is  important in the maintenance of the integrity of SD. These complexes relate the calcium  signaling cascade into the foot processes of podocytes that are essential in the dynamic,  contractile podocytes actin cytoskeleton.  16    PLCε1  PLCε1  belongs  to  the  phospholipase  C  family  of  proteins.  These  proteins  have  an  essential role in the coupling of G‐protein coupled receptors by catalyzing the hydrolysis  of polyphosphoinositides such as phosphatidylinositol‐4,5‐bisphosphate to generate the  second  messengers  inositol  1,4,5‐triphosphate  (IP3)  and  diacylglycerol  (DAG).  IP3  releases calcium from intracellular storage. The elevated calcium level leads to a series  of  signaling  cascades.  DAG  activates  protein  kinase  C,  which  in  turn  modulates  the  activity of many proteins by phosphorylation. All these are essential for cell growth and  differentiation.  PLCε1 has both RasGEF and Ras‐associating domains and may serve as  an activator and an effector of small GTPases (Zhou and Hildebrandt, 2009). PLCε1 was  found to be expressed in the glomeruli, dominantly in the cell bodies of podocytes. The  expression  of  PLCε1  appears  at  the  S‐shaped  stage  of  glomerular  development  and  is  highly expressed during the early capillary loop stage (Hinkes et al., 2006).      A  zebrafish  with  PLCε1  knockdown  model  was  generated  to  investigate  the  role  of  PLCε1  in  the  maintenance  of  the  podocyte  filtration  barrier  during  development.  Characteristic  pathological  features  of  NS,  together  with  foot  process  effacement  and  severe  disorganization  of  SD  were  observed  in  zebrafish  with  knockdown  PLCε1.  The  absence  of  PLCε1  has  been  found  to  be  associated  with  a  significant  reduction  in  nephrin expression and may arrest the development of glomerulus at the capillary loop  stage. PLCε1 has been identified to interact with IQ motif–containing GTPase‐activating  protein 1 (IQGAP‐1), which is known to interact and co‐localize with nephrin. IQGAP‐1 is  17    a  podocyte  cell  junction‐associated  protein.  Its  expression  is  observed  in  the  S‐staged  and capillary loop stages of glomerular development. This suggests that PLCε1 may have  a role as an assembly scaffold for the organization of a multimolecular complex involved  in  morphogenetic  processes  of  glomerular  development  at  the  capillary  loop  stage  (Hinkes et al., 2006).  1.2.2. Cytoskeletal proteins  Actin  cytoskeletal  proteins  play  an  important  role  in  the  regulation  of  the  plasticity  of  the  podocyte  cytoskeleton.  Therefore  they  are  important  in  the  maintenance  of  the  filtration barrier. Alpha actinin 4 (ACTN4) is an actin bundling protein that is responsible  for  the  integrity  of  the  podocyte  cytoskeleton  associated  with  cell  motility.  Actinin  proteins  are  widely  expressed  in  human  tissues,  but  only  ACTN4  is  significantly  expressed  in  the  human  kidney.  Kos  et  al.  generated  mice  that  were  ACTN4  deficient,  which  developed  severely  damaged  podocytes  and  progressive  kidney  disease.  The  deficiency in ACTN4 increased the fluidity of podocytes and also altered the cell motility  and  cell  adhesion.  This  study  showed  that  ACTN4  is  important  for  cell  movement  and  also  normal  podocyte  function  (Kos  et  al.,  2003).  Dandapani  et  al.  showed  the  interaction of ACTN4 with GBM molecules. ACTN4‐deficient mice had lesser amount of  podocytes  per  glomerulus  and  podocyte  markers  were  detected  in  their  urine.  The  podocyte  cells  line  derived  from  these  mice  also  showed  less  adherence  to  the  GBM  components,  collagen  IV  and  laminins‐10  and  laminin‐11.  ACTN4  maintains  the  glomerular architecture and prevents disease (Dandapani et al., 2007).    18    Cathepsin L is a lysosomal protease which has an important role in intracellular protein  degradation  and  activation  of  enzyme  precursors.  Podocytes  express  low  levels  of  cathepsin  L  in  their  lysosomes  under  normal  physiological  condition.  Resier  et  al.  reported that the increased expression and activity of capthepsin L was associated with  the onset of proteinuria. The increased activity of cathepsin L is functionally significant  as it is required for podocyte migration and detachment from the extracellular matrix.  Cathepsin L may alter the size‐ and charge‐selective properties of the filtration barrier. It  also may modify extracellular moieties to engage a subset of integrins or dystroglycans  on the foot processes of podocytes allowing the foot processes to spread on the GBM,  thus establishing podocyte effacement (Reiser et al., 2004).    1.2.3. Cell matrix adhesion complexes   The  two  major  adhesion  complexes  found  at  the  podocyte‐GBM  interface  are  α3β1‐ integrins (Adler, 1992), and α/β‐dystroglycans (Oh et al., 2004, Raats et al., 2000). Figure  3  illustrates  the  molecules  at  the  podocytes‐GBM  interfaces  and  linkage  to  the  foot  processes actin cytoskeleton.      Integrins  are  heterodimeric  transmembrance  receptors  which  anchor  the  cell  to  the  extracellular  matrix  by  binding  to  their  ligands  in  the  GBM.  They  play  a  role  in  signal  transduction, relaying information about adhesion to control cell growth and structure.  α3β1‐integrin  knockout  mice  were  observed  to  be  unable  to  form  mature  foot  processes, displaying GBM fragmentation and abnormal lung maturation (Kreidberg et  19    al.,  1996).    Since  no  podocyte  detachment  was  seen,  it  is  hypothesized  that  α3β1‐ integrin  does  not  have  a  primary  adhesion  function.  The  signaling  of  α3β1‐integrin  through kinases at the adhesion site also modulates actin polymerization (Blattner and  Kretzler, 2005).    Dystroglycans  are  expressed  specifically  at  the  basal  cell  membrane  of  the  foot  processes of podocytes. It is synthesized as a large precursor that is post‐translationally  cleaved  into  two  different  forms  that  are  noncovalently  linked  to  each  other:  α‐ dystroglycan and β‐dystroglycan (Oh et al., 2004). The cytoplasmic tail of β‐dystroglycan  directly interacts with actin filaments of the cytoskeleton. Regele et al. investigated the  expression  of  dystroglycan  in  MCNS  and  FSGS.  The  density  of  α‐dystroglycan  was  significantly  reduced  in  patients  with  MCNS,  but  normal  in  healthy  kidneys  and  FSGS.  The expression of β‐dystroglycan was reduced even more significantly in MCNS patients  (Regele  et  al.,  2000).  The  gene  expression  level  of  dystroglycan  was  also  found  to  be  reduced significantly in an IL‐13 overexpression rat model (Lai et al., 2007).    1.2.4. Podocytes and Nephrotic syndrome  The  phenotypes  of  podocytes  and  the  expression  of  SD  proteins  may  be  different  in  different  form  of  NS.  In  MCNS,  the  podocytes  have  a  switch  in  their  phenotype  from  cells  with  foot  processes  protruding  from  its  basal  surface  to  cells  without  basal  processes. They instead have acquired microvillus protrusions on their apical cell surface  (microvillus  transformation)  (Wiggins,  2007).  Horinouchi  et  al.  had  reported  that  the  20    expression of podocin did not decrease in patients with MCNS (Horinouchi et al., 2003).  Mao  et  al.  reported  that  although  a  normal  level  of  nephrin  mRNA  was  observed  in  patients with MCNS, the amount of nephrin in glomerulus was reduced. The expression  of nephrin was granular and the degree of granulation depends on the degree of foot  process effacement. In addition, a decreased level of CD2AP mRNA and protein was also  observed  in  MCNS  patients.  This  could  suggest  that  nephrin  and  CD2AP  are  two  important  molecules  in  the  complex  of  molecules  that  assemble  and  stabilize  the  SD  (Mao  et  al.,  2006).  There  was  no  loss  in  the  number  of  podocytes  present  in  MCNS  patients. This could explain why patients might have a good prognosis for renal function.  Since  the  podocytes  in  MCNS  only  undergo  a  change  in  the  phenotype,  patients  therefore  responded  well  to  glucocorticoids  as  podocytes  can  revert  back  to  their  normal phenotypes (Wiggins, 2007).    FSGS  is  a  heterogeneous  condition  that  is  commonly  associated  with  podocyte  injury.  Hara  et  al.  had  reported  in  their  study  urinary  loss  of  podocytes  in  FSGS  and  MCNS  patients. They observed that the urinary loss of podocytes in FSGS was higher compared  to those with MCNS. This suggested that the podocyte injury in FSGS patients was more  serious, explaining why the prognosis in FSGS was poorer. The effacement of podocytes  was identified as the initial event in FSGS. Podocytes have a limited ability to repair and  were  unable  to  replicate  postnatally.  This  is  why  there  is  no  increase  in  podocytes  numbers during post natal and compensatory growth (Hara et al., 2001). Horinouchi et  al.  reported  that  podocin  expression  in  74%  of  FSGS  cases  was  either  decreased  or  21    absent.  In  FSGS,  patients  with  sufficient  podocin  expression  tended  to  have  a  better  prognosis  that  those  without  podocin  expression  (Horinouchi  et  al.,  2003).  Over  the  years,  several  podocyte‐associated  genes  have  been  identified  to  contribute  to  the  development of FSGS.     1.3. Genetics of nephrotic syndrome  In recent years, podocyte‐associated proteins have come into focus with regards to the  pathogenesis  of  NS.  Different  genes  have  been  associated  with  different  forms  of  NS  (Table 1). Nephrin was first identified to be the causative gene for congenital nephrotic  syndrome of the Finnish type. Podocin was found to be associated with the autosomal  recessive  SRNS  and  also  sporadic  cases  of  SRNS.  CD2AP  mutations  were  found  to  be  associated  with  sporadic  NS  and  FSGS.  However,  the  reports  on  CD2AP  mutations  are  scarce  and  complete  segregation  data  are  unavailable  in  cases  with  heterozygous  mutations.  Hence,  the  exact  role  of  CD2AP  is  still  unknown.  Mutations  in  ACTN4  and  TRPC6  were  associated  with  the  autosomal  dominant  form  of  familial  FSGS.    Mutated  ACTN4  proteins  demonstrated  higher  affinity  to  F‐actin,  and  hence  may  change  the  mechanical  characteristics  of  podocytes  (Kaplan  et  al.,  2000).  Mutations  in  TRPC6  resulted in a gain of function.  Mutations in WT1 affect the DNA‐binding affinity of WT1  to the target gene. Heterozygous de novo mutations in WT1 resulting in the inability of  zinc fingers to bind to DNA, are seen in Denys‐Drash syndrome, a condition associated  with CNS, male pseudohermphroditism, ambiguous genitalia, 46XY karyotype and Wilms’  tumour (Natoli et al., 2002).  Mutations in the donor splice site in intron 9 of the WT‐1  22    gene  results  in  the  Frasier  syndrome,  and  leads  to  alternative  splicing  and  loss  of  the  +KTS isoform the protein. This syndrome is characterized by nephrotic syndrome due to  FSGS,  and  associated  with  male  pseudohermphroditism  with  normal  female  external  genitalia, streak gonads and 46XY karyotype, without the risk of Wilms’ tumor (Klamt et  al.,  1998).  PLCE1  is  identified  to  be  involved  in  familial  early‐onset  NS  and  ESRD.  Recently, Kopp et al. have identified MYH9 as a major effect risk gene for FSGS in African  Americans (Kopp et al., 2008). In this literature review, the role of NPHS1 and NPHS2 in  NS  will  be  discussed  in  detail.  Table  2  summarizes  some  common  genetic  variants  identified in NPHS1 and NPHS2 in studies carried out on patients with familial NS and/or  sporadic NS.      23    Slit diaphragm    Table 1: Summary of genetic involvement in pathogenesis of nephrotic syndrome.    Clinical Diseases  Mode of  Inheritance  Finnish type  Congenital NS  Autosomal  recessive  Autosomal  recessive  Autosomal  recessive  SRNS  FSGS  SR‐FSGS  Glomerular  Basement  membrane  Podocyte cell body  Familial FSGS  Autosomal  dominant  Autosomal  dominant  Autosomal  recessive  Early‐onset familial  NS  Denys‐Drash  de novo  syndrome and Frasier  dominant  syndrome  Autosomal  FSGS  dominant  Autosomal  FSGS  dominant  Pierson syndrome  Autosomal  recessive  Protein encoded  Gene  Gene Locus  Reference Authors  NPHS1  19q13.1  (Kestila et al., 1998)  NPHS2  1q25‐q31  (Boute et al., 2000)   CD2AP  6  (Gigante et al., 2009)  Transient receptor  Influx of Ca2+ activating signalling  potential cation channel  cascade  6  TRPC6  11q21‐q22  (Reiser et al., 2005)  Alpha actinin 4  Mediators of actin signalling cascade  ACTN4  19q13  (Kaplan et al., 2000)   Phospholipase C epsilon  Regulator of  Ca2+ influx from TRPC6  PLCE1  10q23  (Hinkes et al., 2006)  Wilms tumor 1  Transcription factor that regulates  podocyte proteins  WT1  11q13  (Kreidberg et al., 1993,  Natoli et al., 2002)   Synpo  Mediators of actin signalling cascade  and cell migration  Synpo  5q33.1  (Asanuma et al., 2005,  Dai et al., 2010)  Nonmuscle myosin11A  heavy chain  Actin‐based motility  MYH9  22q12.3  (Kopp et al., 2008)  Laminin beta2 chain  Glomerular development  LAMB2  3q21  (Miner et al., 2006,  Jarad et al., 2006)  Nephrin  Podocin  CD2 associated protein  Protein function  Key protein that mediate signalling at  podocyte slit diaphragm  Integral membrane protein that  facilitate nephrin signalling  Interacts with nephrin and binds actin  24    Table 2: NPHS1 and NPHS2 genetic variants identified in familial and sporadic  nephrotic syndrome.    Exon  Variant  Protein  Ethnic  Disease  Reference  NPHS1  2  c.65C>T  A22V  European  SRNS  (Schultheiss et al., 2004)  3  c.294C>T  I98I  Korean  SRNS  (Kitamura et al., 2006)  Japanese  3  c.349G>A  E117K  European  MCNS  (Lahdenkari et al., 2004)  Korean  SRNS  (Kitamura et al., 2006)  Japanese  7  c.791C>G  P264R  Caucasian  FSGS  (Lowik et al., 2008)  (Santin et al., 2009b)  10  c.1223G>A  R408Q  European  MCNS  (Lahdenkari et al., 2004)  FSGS  (Santin et al., 2009b)  17  c.2289C>T  V763V  European  MCNS  (Lahdenkari et al., 2004)  Korean  SRNS  (Kitamura et al., 2006)  Japanese  Sporadic NS  (Mao et al., 2007)  Chinese    18  c.2398C>T  R800C  European  MCNS  (Lahdenkari et al., 2004)  Chinese  Sporadic NS  (Mao et al., 2007)  18  c.2479C>A  R827X  European  SRNS  (Philippe et al., 2008)  FSGS  (Santin et al., 2009b)  22  c.2928G>T  R976S  European  SRNS  (Philippe et al., 2008)  FSGS  (Santin et al., 2009b)  24  c.3230A>G  N1077S European  MCNS  (Lahdenkari et al., 2004)  26  c.3315G>A  S1105S  European  MCNS  (Lahdenkari et al., 2004)  Korean  SRNS  (Kitamura et al., 2006)  Japanese  Sporadic NS  (Mao et al., 2007)  Chinese  NPHS2  Promoter  c.‐51G>T  ‐  Chinese  SRNS  (Yu et al., 2005)  European  MCNS  (Di Duca et al., 2006)  American    (McKenzie et al., 2007)  (Zhu et al., 2009)  1  c.59C>T  P20L  European  SRNS/SDNS  (Caridi et al., 2003)  American  (Ruf et al., 2004)  (McKenzie et al., 2007)  (Santin et al., 2011)  2  c.288C>T  S96S  European  Familial and  (Karle et al., 2002)  Chinese  sporadic NS  (Yu et al., 2005)  American  SRNS  (Gbadegesin et al.,  MCNS  2007)  25    Exon  Variant  Protein  Ethnic  Disease  3  c.413G>A  R138Q  Familial NS  FSGS  SRNS/SDNS  4  c.503G>A  R168H  5  c.538G>A  European  African  Chinese  V180M  European  African  American    5  c.686G>A  R229Q  European  American  African    c.725C>T  A242V  European  7  c.871C>T  R291W  European  American  African  European  American  African  SRNS  Reference  (McKenzie et al., 2007)  (Zhu et al., 2009)  (Boute et al., 2000)  (Karle et al., 2002)  (Tsukaguchi et al., 2002)  (Caridi et al., 2003)  (Ruf et al., 2004)  (Weber et al., 2004)  (McKenzie et al., 2007)  (Tonna et al., 2008)  (Lowik et al., 2008)  (Jungraithmayr et al.,  2011)  (Weber et al., 2004)  (Yu et al., 2004)  Familial NS  Sporadic NS  SRNS  FSGS  (Boute et al., 2000)  (Caridi et al., 2001)  (Karle et al., 2002)  (Ruf et al., 2004)  (Weber et al., 2004)  (Tonna et al., 2008)  Familial NS  (Karle et al., 2002)  Sporadic NS  (Tsukaguchi et al., 2002)  SRNS/SDNS  (Caridi et al., 2003)  (Weber et al., 2004)  (Schultheiss et al., 2004)  (Gbadegesin et al.,  2007)  (Tonna et al., 2008)  (Jungraithmayr et al.,  2011)  FSGS  (Tsukaguchi et al., 2002)  (Jungraithmayr et al.,  2011)  (Santin et al., 2011)  Familial NS  (Boute et al., 2000)  FSGS  (Karle et al., 2002)  SRNS  (Tsukaguchi et al., 2002)  (Ruf et al., 2004)  (Weber et al., 2004)  (McKenzie et al., 2007)  26    Exon  Variant  Protein  Ethnic  Disease  8  c.954T>C  A318A  European  Japanese  Chinese  American  Familial and  sporadic NS  SRNS  MCNS  8  c.1038A>G  L346L  European  Japanese  Chinese  American  Familial and  sporadic NS  SRNS  MCNS    Reference  (Santin et al., 2011)   (Karle et al., 2002)  (Maruyama et al., 2003)  (Yu et al., 2004)  (Yu et al., 2005)  (Di Duca et al., 2006)  (Mao et al., 2007)  (Gbadegesin et al.,  2007)  (McKenzie et al., 2007)  (Zhu et al., 2009)  (Karle et al., 2002)  (Maruyama et al., 2003)  (Yu et al., 2005)  (Di Duca et al., 2006)  (Mao et al., 2007)  (Gbadegesin et al.,  2007)  (McKenzie et al., 2007)  (Zhu et al., 2009)    27    1.3.1. NPHS1: The gene encoding for nephrin  NPHS1  is  located  on  the  chromosome  19q13.1  and  has  29  exons.  It  codes  for  nephrin  which has 1241 amino acids. The first exon codes for the signal peptide, and exon 2 to  20  code  for  the  region  containing  the  immunoglobulin  motifs.  Each  immunoglobulin  motif  is  encoded  by  two  exons  each,  with  exception  to  motif  Ig‐2,  that  is  encoded  by  three exons. The fibronectin type III‐like domain is encoded by exon 22 and 23. Exon 24  codes for the transmembrane domain. Exons 25 to 29 code for the cytosolic domain and  3’UTR  (Lenkkeri  et  al.,  1999).    Kestilä  et  al.  first  identified  mutations  in  NPHS1  in  congenital  NS  of  the  Finnish  type  (CNF)  by  positional  cloning  (Kestila  et  al.,  1998).  To  date, more than 90 mutations (missense, nonsense, deletion, insertion, splice site and  promoter) in NPHS1 have been identified (Santin et al., 2009b).     Nephrin and congenital nephrotic syndrome  Mutations  in  NPHS1  have  been  reported  to  be  the  common  cause  of  congenital  NS  (CNS).  CNS  is  defined  as  NS  which  occurs  in  utero  or  during  the  first  3  months  of  life.  CNF  is  an  autosomal  recessive  disease  that  occurs  with  an  incidence  rate  of  1:10000  births in Finland but less frequently in other countries (Kestila et al., 1998). Kestilä et al.  reported two major mutations in their study of 49 Finnish patients with CNF.  These 2  mutations  were  found  in  more  than  90%  of  their  subjects,  either  in  homozygous  or  heterozygous  state.  The  first  mutation  is  Finmajor  (nt121  (del2))  which  is  deletion  of  nucleotide 121‐122 CT in exon 2, resulting in a truncated protein of only 90 amino acids.  The second mutation is Finminor (R1109X) which is a nonsense mutation in exon 26 that  28    results in a truncated protein of 1109 amino acids where part of the intracellular domain  is lost (Kestila et al., 1998). Glomerular extracts from kidneys with these two mutations  were  stained  negative  for  nephrin  and  these  kidneys  also  lack  slit  diaphragm.  These  mutations  have  resulted  in  the  total  lack  of  functional  nephrin  in  affected  patients  (Patrakka et al., 2000).     Finmajor  and  Finmajor  mutations,  however,  are  not  as  frequent  in  other  Caucasian  populations (Gigante et al., 2002, Heeringa et al., 2008, Lenkkeri et al., 1999) and Asian  populations (Aya et al., 2000, Lee et al., 2009, Sako et al., 2005). This suggests that each  ethnic  group  may  have  their  own  unique  NPHS1mutatons.  An  example  is  seen  in  the  Mennonite  population,  which  have  2  unique  mutations  (c.1481delC  and  c.3250delG)  that are not seen in non‐Mennonite patients (Beltcheva et al., 2001). Another example  is seen in the Japanese population which has 2 unique mutations (c.2515delC and c.736  G>T) (Aya et al., 2009).    Many  NPHS1mutation  have  been  reported  among  the  Caucasian  populations  of  CNS.  Lenkkeri  et  al.  reported  32  novel  mutations  in  their  group  of  Finnish  and  non‐Finnish  patients (Lenkkeri et al., 1999). An Italian study of 15 CNS patients revealed all of them  have homozygous NPHS1 mutations (Gigante et al., 2002). A recent study by Heeringa et  al.  done  on  a  worldwide  cohort,  mainly  Europeans,  revealed  13  novel  mutations  (Heeringa et al., 2008). In contrast, the incidence of NPHS1 mutation is low in Japanese  29    patients with CNS, suggesting that NPHS1 is not an exclusive or a major cause of CNS in  Japanese children (Aya et al., 2009, Sako et al., 2005).      Nephrin and other variants of nephrotic syndrome  Although  nephrin  is  apparently  pivotal  in  development  of  CNF,  the  role  of  nephrin  in  other variants of NS is also of interest. Lahdenkari et al. investigated NPHS1 in patients  with childhood MCNS. Twelve genetic variants were identified, of which 4 of the amino  acid  substitutions  occur  in  Ig‐7  and  Ig‐8  domains.  They  observed  that  patients  with  complicated disease have more genetic variants compared to those with a mild disease.  They  also  hypothesized  that  a  missense  mutation  (p.R800C)  is  a  pathogenic  alteration  and affected the phenotype of an MCNS patient who had Finmajor mutation(Lahdenkari  et al., 2004).    Mao et al studied NPHS1 and NPHS2 in Chinese children with SSNS and SRNS. One novel  missense  mutation,  one  known  missense  mutation  and  three  known  polymorphisms  were identified in this group. No Finmajor or Finminor mutation was identified (Mao et al.,  2007).  Philippe  et  al.  reported  14  mutations  in  10  unrelated  patients  among  160  patients presenting with SRNS between 3 months and 18 years of age. NPHS2 and WT1  mutations were excluded from these patients and hence SRNS can be attributed to the  pathogenic NPHS1 mutations identified (Philippe et al., 2008). On the contrary, a study  looking  at  15  Korean  and  Japanese  families  with  SRNS  identified  only  several  known  single nucleotide polymorphisms (SNPs) and hence NPHS1 was not responsible for SRNS  30    in  these  families  (Kitamura  et  al.,  2006).  Santin  et  al.  examined  NPHS1  in  97  patients  with familial and sporadic SRNS, of whom 52 presented with SRNS after 18 years of age.  Compound heterozygous or homozygous NPHS1 mutations were identified in 5 familial  and 7 sporadic cases. The frequency of NPHS1 was 2% in adults as compared to 14% in  children, which is 7 times higher (Santin et al., 2009b).    Types of NPHS1 mutations  The types of mutations in a patient will determine the severity of the disease. Hinkes et  al. described that children who have mutations in NPHS1, NPHS2, WT1 and/or LAMB2  do  not  respond  well  to  steroid  therapy  (Hinkes  et  al.,  2007).  Finmajor  and  Finmajor  mutations  lead  to  the  absence  of  nephrin  in  the  podocyte  slit  diaphragm  and  cause  SRNS leading to early ESRD, whereas other mutations that result in the impairment of  nephrin  or  entrapment  of  neprhin  in  the  endoplasmic  reticulum  do  not  completely  abolish  the  function  of  nephrin  (Heeringa  et  al.,  2008).  Shono  et  al.  described  how  nephrin mutations could lead to dysfunction of the SD via three mechanisms. The first  mechanism  is  that  a  frameshift  mutation,  such  as  the  Finmajor  mutation,  can  disrupt  a  transcription  or  translation  process,  resulting  in  the  absence  of  functional  nephrin.  Hence, this leads to a severe, congenital phenotype. The second mechanism is seen in  patients  with  early‐onset  SRNS,  when  they  have  mutations  which  resulted  in  the  mis‐ trafficking  of  nephrin  that  are  trapped  in  the  endoplasmic  reticulum.  The  third  mechanism is the formation of mutant proteins that have the ability to reach the plasma  membrane but are defective in the process of assembling dynamic nephrin complexes.  31    Patients  with  these  mutations  usually  have  the  mildest  phenotypes  compared  to  the  previous  two  mechanism    (Shono  et  al.,  2009).  Philippe  et  al.  grouped  mutations  according to “mild” or “severe” according to predictive algorithms.  They observed that  patients  with  compound  heterozygous  mutations,  of  which  at  least  one  is  a  “mild”  mutation,  had  a  later  onset  and  a  milder  course  of  disease  (Philippe  et  al.,  2008).  Heeringa  et  al.  hypothesized  that  mutations  that  affect  certain  Ig‐like  domains  of  nephrin could lead to a more serve disease phenotype compared to mutations affecting  other domains.     1.3.2. NPHS2: The gene encoding for podocin  NPHS2 is located on the chromosome 1q25‐q31. The complete NPHS2 ORF is 1,149 base  pairs;  followed  by  a  635  base  pair  of  3’UTR  that  contains  an  atypical  polyadenylation  signal (AATTAAA) situated 13 nucleotides up stream of the poly (A) tail. The full length  cDNA  encodes  for  podocin,  which  is  a  383‐  amino  acid  protein  that  is  approximately  42kD (Boute et al., 2000). The N‐terminal of podocin is encoded by amino acid 15 to 89  and the C‐terminal is encoded by amino acid 135 to 383 (Roselli et al., 2002). Boute et al.  first  identified  mutations  in  NPHS2  to  be  associated  with  autosomal  recessive  SRNS  (Boute et al., 2000).  Autosomal recessive SRNS belongs to the heterogeneous group of  familial  NS,  which  is  characterized  by  childhood  onset  of  proteinuria  and  histological  findings of either FSGS or MCNS (Karle et al., 2002).       32    Podocin and steroid resistance  NPHS2 was first identified to be the causative gene in patients with familial FSGS. Ten  different  NPHS2  mutations  (nonsense,  frameshift  and  missense)  were  identified  to  be  associated with familial FSGS, indicating an important role for podocin in the function of  glomerular filtration barrier (Boute et al., 2000). Karle et al. also found NPHS2 mutations  and  polymorphisms  not  only  in  familial  SRNS  but  also  in  sporadic  SRNS  (Karle  et  al.,  2002). In an Italian cohort of patients with non‐familial SRNS, the frequency of NPHS2  mutation  was  approximately  20%.  Although  NPHS2  mutations  were  found  in  both  familial and sporadic SRNS, there was genetic heterogeneity between the two groups.  Wuber et al. did NPHS2 screening for three groups of patients: familial SRNS, sporadic  SRNS and diffuse mesangial sclerosis (DMS). No NPHS2 mutation was identified in DMS.  The  mutation  frequency  for  familial  SRNS  was  40%  and  10.5%  in  sporadic  SRNS.  The  lower detection rate among sporadic SRNS compared to familial SRNS could be due to  the  complex  inheritance  patterns  and  the  potential  gene‐environment  interactions  in  patients with sporadic SRNS (Weber et al., 2004).    Studies  were  carried  out  to  study  the  effect  of  NPHS2  mutations  or  polymorphisms  in  steroid resistance by comparing patients with SSNS and SRNS. Caridi et al. reported that  12% of their patients with SRNS were found to have homozygous NPHS2 mutations. All  these patients had severe proteinuria that occurred in early childhood and progressed  to  ESRD.  In  comparison,  6%  of  the  patients  with  SSNS  had  single  heterozygous  NPHS2  mutations,  with  no  homozygous  mutations  identified  (Caridi  et  al.,  2003).    Ruf  et  al.  33    hypothesized  that  NPHS2  mutations  might  be  the  cause  of  steroid  resistance.  Hence,  they did NPHS2 screening on SRNS patients with SSNS patients as controls. Homozygous  and heterozygous mutations were found only in SRNS patients (Ruf et al., 2004). These  data revealed that NPHS2 mutations could be a cause for the resistance to steroids in  certain  patients.  As  steroids  have  adverse  side  effects,  it  would  be  useful  to  screen  children with SRNS for NPHS2 mutations so as to avoid giving them the potentially toxic  treatment.  Caridi  et  al.  suggested  that  all  cases  with  FSGS  and  heavy  proteinuria   resistant  to  steroids,  should  undergo  NPHS2  genotyping  prior  to  planning  therapeutic  regimens with steroids and other immunosuppressive drugs (Caridi et al., 2001).     NPHS2  mutations  are  a  frequent  cause  of  sporadic  SRNS,  occurring  in  10.5  to  28%  of  children with SRNS in Europe, the Middle East and North America. However, this is not  observed  in  Asian  countries.  Yu  et  al.  ,the  first  group  to  study  NPHS2  mutations  in  Chinese  SRNS  patients,  only  found  a  heterozygous  mutation  in  one  out  of  23  children  with  sporadic  SRNS.  Seven  polymorphisms  were  also  identified  in  both  patients  and  controls  (Yu et al., 2005). The low frequency of NPHS2 mutations was also reflected in  another study performed by Mao et al. on Chinese patients with SRNS and SSNS. Only 3  out  of  22  patients  were  identified  with  heterozygous  mutations  (Mao  et  al.,  2007).  A  study done in Japan also reported a low frequency rate of 2% in Japanese children with  sporadic  SRNS  (Maruyama  et  al.,  2003).  Among  Korean  children  with  SRNS,  no  NPHS2  mutation  was  found    (Cho  et  al.,  2008).  All  these  data  demonstrated  that  ethnic  differences  might  play  a  role  in  the  occurrence  of  NPHS2  mutations  and  NPHS2  34    mutations  were  more  common  in  Caucasian  populations  as  compared  to  Asian  populations.    Other  than  just  looking  at  the  effect  of  NPHS2  mutations  in  SRNS,  Frishberg  et  al.  investigated  the  association  of  cardiac  disorders  with  SRNS.  They  had  shown  that  cardiac anormalies and SRNS as a result of NPHS2 mutations was not an association by  chance but rather co‐inherited. Podocin transcript is found in fetal heart but not adult  heart.  Its  presence  in  the  fetal  heart  may  interact  with  other  peptides  and  affect  the  overall signal transduction. Mutated podocin may lead to abnormal signaling, resulting  in  congenital  heart  defects.  Therefore,  there  would  be  a  need  to  perform  cardiac  evaluation for SRNS patients with NPHS2 mutations (Frishberg et al., 2006).     Podocin and post‐transplant recurrence  Post‐transplant  recurrence  in  FSGS  is  considered  one  of  the  most  traumatic  clinical  events  in  the  clinical  setting.  The  recurrence  of  proteinuria  following  renal  transplant  has been described in 30% of paediatric patients with SRNS due to FSGS (Billing et al.,  2004).  Studies  have  been  carried  out  to  investigate  the  association  of  post‐transplant  recurrence  in  FSGS  in  patients  with  NPHS2  mutations.  Bertelli  et  al.  studied  the  incidence of post‐transplant recurrence of FSGS in patients with NPHS2 mutations. They  reported  that  5  of  13  patients  with  NPHS2  mutations  presented  with  post‐transplant  recurrence of proteinuria, of whom 2 were proven histologically to have recurrence of  FSGS.  There  were  only  2  cases  that  had  homozygous  or  compound  heterozygous  35    mutations of NPHS2 (22% of those who received a renal graft) and these two cases had  mild clinical impact with good prognosis. However, they also observed that there were 3  patients  who  had  a  single  NPHS2  mutation,  of  whom  one  presented  with  a  poor  prognosis (Bertelli et al., 2003).  Billing et al. also reported that their patients developed  heavy  proteinuria  early  after  renal  transplant.  They  did  not  identify  any  anti‐podocin  antibodies  or  immunoglobin  deposits  in  renal  histology  in  patients  with  SRNS  due  to  NPHS2  mutations  who  presented  with  early  post‐transplant  recurrence  of  proteinuria.  This suggested that post‐transplant recurrence in these patients was unlikely due to an  antibody mediated mechanism. They concluded that patients with SRNS due to NPHS2  mutations  are  not  protected  from  the  recurrence  of  proteinuria  after  renal  transplant  (Billing et al., 2004).     In contrast, there are other studies indicating that patients with NPHS2 mutations have  a lower risk of post‐transplant recurrence. Ruf et al. observed that 7 out of 20 patients  (35%)  with  SRNS  but  no  NPHS2  mutations  had  post‐transplant  recurrence  of  FSGS,  whereas  only  2  of  24  patients  (8%)  with  SRNS  and  compound  heterozygous  or  homozygous NPHS2 mutations experienced recurrence. Proteinuria was observed in 1 of  these 2 patients, who responded to steroid therapy. There was no histological finding of  FSGS. Hence, they demonstrated that there is a significant lower risk of FSGS recurrence  after  a  renal  transplant  in  patients  with  compound  heterozygous  or  homozygous  mutations (Ruf et al., 2004). Weber et al. reported that out of 32 transplanted patients  with  2  pathogenic  NPHS2  mutations,  only  1  patient  exhibited  recurrent  FSGS  in  a  36    delayed  onset.  They  also  observed  that  3  of  5  patients  with  sporadic  SRNS  and  heterozygous  NPHS2  mutations  experienced  post‐transplant  recurrence.  These  observations suggested that heterozygous NPHS2 mutations may play a role in recurring  disease following renal transplantation (Weber et al., 2004). Caridi et al. recommended  that grafts from carriers of NPHS2 mutations, such as from parents or siblings, should be  avoided as there appeared to be a higher risk of post‐transplant recurrence in patients  with  heterozygous  NPHS2  mutations.  Similarly,  patients  with  heterozygous  NPHS2  mutations should be considered to be at risk for recurrence and be strictly monitored in  the post‐graft phase (Caridi et al., 2005).     Types of mutations  Many  NPHS2  genetic  variants  have  been  identified  in  Caucasian,  African,  Asian  and  other populations (Yu et al., 2005). The types of mutations identified include missense  mutations, splice‐site mutations, frameshift mutations and polymorphisms (Karle et al.,  2002).     The type and location of the NPHS2 mutations affect how podocin would be expressed.  Zhang et al. reported that deletion mutation of 1 to 2 bp would cause a frameshift and  lead to the production of truncated protein. A mutation in Exon 7 that resulted in the  loss of 98 amino acids of podocin had a normal RNA expression. However, the presence  of  the  C‐terminal  domain  cannot  be  detected.  In  contrast,  a  mutation  in  Exon  3  that  leads  to  the  truncation  of  238  amino  acids  resulted  in  a  reduced  expression  of  RNA  37    transcripts and N‐terminal of podocin. The authors have concluded based on their data  that  NPHS2  mutations  could  affect  the  localization  of  nephrin,  CD2AP  and  α‐actinin.  They noticed that these proteins instead of localizing along the GBM, they were found  to  be  in  the  podocyte  body  (Zhang  et  al.,  2004).  Mutations  that  are  associated  with  severe childhood onset are distributed throughout podocin. However, in general, most  of them are located in the N‐terminal. This might show that N‐terminal mutations may  represent  particular  alleles  causing  the  misfolding  of  protein  or  altered  protein  processing, leading to a more severe clinical phenotype (Tsukaguchi et al., 2002).    Weber et al. observed that the nature of mutations correlated to some extent with the  age  of  onset.  Patients,  which  presented  with  frameshift  and  protein  truncating  mutations, had an early onset of NS. The presence of two pathogenic NPHS2 alleles in  patients indicated a more severe prognosis that was characterized by an earlier onset.  Patients  with  only  a  single  mutation  tended  to  have  NS  in  the  later  stage  of  life. They  also described that the severity of the disease seemed to be determined by the impact  of  the  amino  acid  substitution  on  the  specific  functional  domains  and  on  the  intracellular  trafficking  of  podocin.  Some  missense  mutations  managed  to  reach  the  plasma membranes, whereas there are some which were retained in the endoplasmic  reticulum (Weber et al., 2004).     p.R229Q  is  one  of  the  most  commonly  reported  NPHS2  mutations  in  Caucasian  populations  (Table  2).  A  G  to  A  nucleotide  change  is  observed  at  the  position  686  in  38    exon 5 of NPHS2, which results in an arginine to glutamine substitution at codon 229.  This substitution had altered the binding of podocin to nephrin, suggesting that podocin  resulting  from  R229Q  is  biochemically  altered  (Caridi  et  al.,  2003).  p.R229Q  polymorphism in the heterozygous status has been reported that on its own is not a risk  factor  for  the  late‐onset  of  FSGS.  However,  in  combination  with  another  pathologic  mutation,  it  can  enhance  the  susceptibility  to  FSGS  or  SRNS  (Tsukaguchi  et  al.,  2002,  McKenzie et al., 2007). This means that heterozygous mutations could contribute to the  development  of  proteinuria  by  associating  with  other  factors,  such  as  presence  of  a  second mutation, and not directly determining it. Tsukaguchi et al. also described that  there  is  a  possibility  that  p.R229Q  homozygotes  may  have  a  mild  phenotype  that  generally escape medical attention (Tsukaguchi et al., 2002).    1.3.3. The two closely related partners: NPHS1 and NPHS2  Both  nephrin  (NPHS1)  and  podocin  (NPHS2)  are  two  important  molecules  in  the  SD.  They interact closely with one another to maintain the integrity of podocytes, protecting  against  the  leakage  of  macromolecules  into  the  urine.  Mutations  of  NPHS1  and/or  NPHS2 often lead to proteinuria and the development of NS. Huber et al had carried out  studies to understand the interaction of podocin with nephrin (Huber et al., 2001, Huber  et al., 2003b). They first described that the cytoplasmic tail of nephrin binds to podocin  and  this  interaction  is  mediated  by  the  C‐terminal  of  podocin.  The  authors  concluded  that nephrin and podocin form a signaling complex which supports the functional and  structural  integrity  of  podocytes  (Huber  et  al.,  2001).  Podocin  is  responsible  for  the  39    recruitment  and  the  stabilizing  of  nephrin  at  the  podocyte  foot  processes.  Podocin  is  localized to the plasma membrane, and form homo‐oligomers that involve the C‐ and N‐ terminal cytoplasmic domains. This association of podocin with the specialized lipid raft  microdomains of the plasma membrane is essential in the recruitment of nephrin into  the  rafts.  Huber  et  al.  also  investigated  two  NPHS2  mutations,  of  which  one  caused  podocin to be retained in the endoplasmic reticulum and the other caused the failure of  podocin  to  be  associated  with  the  lipid  rafts.  Both  mutations  failed  to  recruit  nephrin  into the lipid rafts and hence the signaling of nephrin is affected. This study showed the  importance of lipid rafts targeting in nephrin signaling (Huber et al., 2003b). This close  relationship  between  nephrin  and  podocin  arise  the  interest  of  investigators  to  study  the mutations in both genes and see how these two genes could be responsible for NS.     A triallelic hit in NPHS1 and NPHS2 was described in CNS by Kozeill et al.. They observed  that 50% of the patients with homozygous R1160X nonsense mutation have the milder  CNF  phenotype  and  interestingly  majority  of  the  mildly  affected  cases  were  females.  They  also  observed  that  NPHS1  and  NPHS2  mutations  co‐existed  in  all  the  congenital  FSGS patients, providing evidence of a functional relationship between these two genes.  They  described  a  triallelic  hit  in  patients  with  congenital  FSGS  who  have  mutations  in  NPHS1 and NPHS2, where a biallelic hit in either NPHS1 or NPHS2 might result in CNF.  The presence of the third allelic hit may serve as a modifier of disease expression of CNF  to FSGS. In summary, they had identified a specific di‐genic inheritance, which resulted  in  three  variant  alleles  being  associated  with  the  modification  of  CNF  to  FSGS.  These  40    data have demonstrated that inheritance of different alleles at independent genetic loci  may contribute to disease phenotype (Koziell et al., 2002).    On the contrary, Schultheiss et al. did not find any evidence for di‐genic inheritance of  NPHS1 and NPHS2 mutations as a tri‐allelic hit that would result in the modification of  phenotypes in their group of 75 patients with NS. The frequency of patients that were  presented  with  both  NPHS1  and  NPHS2  was  low  (5%).  Their  data  also  did  not  suggest  any  genotype/phenotype  correlations  in  their  patients  with  NPHS1  and  NPHS2  mutations.  They  also  observed  that  there  can  be  rare  cases  of  unusually  mild  phenotypes in patients with NPHS1 mutations  (Schultheiss et al., 2004).    Mao et al. screened for NPHS1 and NPHS2 screening on Chinese patients with SRNS and  SSNS.  They  identified  5  out  of  60  patients  (8%)  with  mutations  in  both  NPHS1  and  NPHS2.  However,  only  known  polymorphisms  with  no  amino  acid  substitution  were  identified. There was no genotype/phenotype correlation in NPHS1 and NPHS2 among  these  patients  (Mao  et  al.,  2007).    Another  study  group  in  Japan  carefully  selected  15  families with SRNS (10 Japanese and 5 Koreans) with phenotypes that were comparable  to  the  Caucasian  population.    They  performed  NPHS1,  NPHS2  and  NEP1  screening  on  these subjects. They only managed to identify known NPHS1 and NPHS2 polymorphisms,  suggesting  that  NPHS1  and  NPHS2  might  not  be  the  cause  of  SRNS  in  this  population.  They suggested that the genetic factors of FSGS may be different between the Asian and  41    Caucasian patients and that there might be unidentified genes that could be involved in  the pathogenesis of SRNS in the Asian patients (Kitamura et al., 2006).    Hinkes et al. screened for NPHS1, NPHS2, WT‐1 and LAMB2 for patients with NS in their  first  year  of  life.  NPHS1  mutations  were  identified  exclusively  in  CNS,  whereas  NPHS2  mutations  were  found  in  patients  with  CNS  and  infantile  onset  of  NS.  They  also  observed  that  the  progression  to  ESRD  was  more  rapid  in  children  with  NPHS1  mutations  than  those  with  NPHS2  mutations.  They  concluded  that  patients  with  causative mutations in any of the 4 genes do not respond to steroid therapy (Hinkes et  al., 2007).      Lowik et al. analyzed seven podocyte associated genes (NPHS1, NPHS2, ACTN4, CD2AP,  WT‐1,  TRPC6  and  PLCE1)  in  patients  with  non‐familial  childhood  onset  of  SRNS.  Their  results  suggested  that  combined  haploinsufficiency  in  two  podocyte  associated  genes  might be responsible for the development of FSGS. They also observed a tri‐allelic hit of  NPHS1 and NPHS2 in one of their patients. A single mutation in a recessive disorder may  be  unable  to  induce  pathologic  effect.  The  presence  of  a  second  mutation,  which  has  not been identified or present in another gene that was not screened, may produce an  additive  effect.  In  conclusion,  their  data  demonstrated  that  the  combined  genetic  defects in podocyte associated genes may play a role in the development of FSGS (Lowik  et al., 2008).     42    1.4. Gaps in current knowledge  Current  studies  on  NPHS1  and  NPHS2  have  been  well  established  in  the  Caucasian  populations. Mutations in NPHS1, such as Finmajor and Finminor mutations, have been well  recognized to be associated with the congenital nephrotic syndrome of the Finnish type.  The  occurrence  of  certain  NPHS2  polymorphisms,  such  as  R229Q,  is  more  frequent  in  the  Caucasian  populations  compared  to  the  Asian  populations.  Asian  studies  have  showed a lower incidence in NPHS2 mutations, demonstrating that ethnicity may play a  role in the difference of polymorphism/mutation frequency. It is also noted that there  are very few studies being performed in South East Asians.     Based  on  a  16‐year  retrospective  review  of  renal  biopsies  from  patients  with  SRNS  or  SSNS  done  at  the  Shaw‐NKF‐NUH  Children’s  Kidney  Centre,  University  Children’s  Medical  Institute,  National  University  Health  System,  there  seems  to  be  an  increased  prevalence of SRNS among Malay  patients, and a greater tendency towards end‐stage  renal  failure  (ESRF).  However,  genetic  studies  in  Malays  have  not  been  described.  Singapore,  being  a  regional  referral  center  in  South  East  Asia  with  a  multiracial  population will serve as an ideal and unique platform for the study of podocyte genetic  variants, in the different ethnic populations.    Functional polymorphisms or mutations in podocyte‐associated genes may account for  the  different  clinical  course.  It  is  hypothesized  that  there  are  important  gene‐gene  interactions  among  the  different  podocyte‐associated  genes  that  may  result  in  the  43    nephrotic  phenotype.  Hence,  the  examination  of  any  possible  NPHS1  and  NPHS2  variants in our population will help us to better understand the pathogenesis of NS.     1.5. Objectives of the study  This  study  therefore  aims  to  address  an  important  question  in  the  Southeast  Asian  region  ‐  why  is  there  such  a  great  difference  in  the  prevalence  and  outcome  of  NS,  especially  FSGS  among  Malay  patients  compared  to  the  Chinese?  Genetic  or  environmental  factors  might  be  the  explanation  for  inter‐ethnic  differences.    By  characterizing  the  polymorphisms  and  mutations  in  the  different  Asian  ethnic  groups,  we will have a better understanding of the pathogenesis of NS in Asians.     Singapore  offers  the  advantage  for  a  genetic  study  on  the  various  ethnicities  in  the  South‐East  Asian  region;  as  Chinese,  Malays  and  Indians  make  up  a  significant  proportion  of  the  patient  pool  in  local  hospitals.  Hence,  by  comparing  the  ethnicities  (Chinese and Malays) within our patient population, this study may provide some insight  into this inter‐ethnic difference.    Hence, the objectives of this study are:  • Identify novel and/or known polymorphisms/mutations of NPHS1 and NPHS2 genes  using high resolution melting and direct gene sequencing.  44    • Describe the spectrum and frequency of NPHS1 and NPHS2 polymorphisms and/or  mutations in our young Chinese and Malay patients who visited the Shaw‐NKF‐NUH  Children’s Kidney Centre, University Children’s Medical Institute.  • Determine  disease  associations  of  these  genetic  variants,  in  terms  of  the  clinical  features,  response  to  therapy  and  progression  to  end‐stage  renal  failure  in  the  Chinese and Malay patients.  • Investigate  if  there  are  any  gene‐gene  interactions  between  NPHS1  and  NPHS2  in  our patients.                       45    2. Methods and Materials  2.1. Study Subjects  All patients with idiopathic sporadic NS attending the Shaw‐NKF‐NUH Children’s Kidney  Centre, University Children’s Medical Institute, National University Health System were  included in this study.  The children were initially treated with the prednisolone therapy  (standard  regimen  recommended  by  the  International  Study  of  Disease  in  Children),  (ISKDC,  1981,  ISKDC,  1982),  and  were  classified  into  steroid  responsiveness  or  steroid  resistance according to whether there was complete remission within 8 weeks following  steroid therapy. Remission was defined as normal urinary protein excretion (Albustix ®  trace or negative for at least 3 consecutive days or C,  c.65C>T,  c.494C>T,  c.803G>A,  c.1233C>T,  c.2871G>A, c.3047G>A, c.3230A>G and c.3315G>A) were genotyped using tetra‐primers  amplification  refractory  mutation  system  (ARMS)  PCR  (Table  5).  This  method  adopts  principles  from  tetra‐primers  PCR  and  ARMS.  The  normal  ARMS  require  two  separate  PCR reactions, but tetra‐primers ARMS PCR is a single PCR reaction. It uses two pairs of  primers in one reaction; one pair of outer primers that serves as a control and the other  pair of inner primers which is allele specific. Allele specificity is conferred by a mismatch  between  the  3’  terminal  base  of  one  inner  primer  and  the  DNA  template.  A  second  mismatch is introduced at the position ‐2 from the 3’ terminus in the inner primers so as  to  enhance  allele  specificity.  Primers  were  designed  with  an  online  program  that  was  created by Ye et.al. (http://cedar.genetics.soton.ac.uk/public_html/primer1.html) (Ye et  al., 2001).   56    A reaction mixture (15μl) for tetra‐primers ARMS PCR contained 30 to 60ng of gDNA, 1x  PCR  buffer,  2mM  of  MgCl2,  0.2μM  of  dNTP,  0.2µM  of  each  outer  primer,  0.2μM  to  0.4μM  of  each  inner  primer,    1  unit  of  Taq  polymerase  (Promega)  and  nuclease  free  water.  Betaine  (Sigma  Aldrich)  was  added  for  the  amplification  of  certain  regions.  The  reactions  were  run  at  95°C  for  5  minutes,  followed  by  32  to  35  cycles  of  95°C  for  30  seconds,  annealing  temperature,  as  shown  in  Table  5,  for  30  seconds  and  72°C  for  1  minute.  A  final  9  minutes  of  extension  at  72°C  was  performed  after  the  last  cycle.  The  PCR products were run on a 2.5% gel to determine the genotypes.      57    Table 5: Primers used in tetra‐ARMS PCR for the genotyping of NPHS1 SNPs.    SNP  Primers (5’ to 3’)  Product size (bp)  c.‐170 T>C  Outer forward: AAGGAAAGAGTCTGAGATCAACCTGGC Outer reverse: CTTTCTCTGGGTCCCTCTCTGTGTGT  Inner forward: ATTGAGACTGAGAGCAAGACAGAGAGATAT  Inner reverse: TTTCCTCTTCCCCTCTTCCCTGTGATTG    Outer forward: GTGGGATCCCAGCCTTGTACCCAGCGCC Outer reverse: CCCAGGAGCAGCCCATCTTTGGCCCATT  Inner forward: GCACCGCGCTGTGTCCTCAGGCCTTGT  Inner reverse: GGGAACGGAGGCAGGAATCGCCAACTTCG    Outer forward: GTACCGAACTCCAATCTTCAAGTCCTTC Outer reverse: TAGAAGGGTACTGGTCAGGAACACACAC  Inner forward: CGTGGTCAACTGTGTGTCTGGGGAAGC  Inner reverse: AATGGTGATGTCAGGTGCTGGCTGCA    Outer forward: CAGGCCCACTGAGTGACTGATATCC Outer reverse:CTGGCCTCACGGTCATCACCAGCAC  Inner forward:AGCTTGGAGCTGCCGTGCGTGGCACA  Inner reverse: ACTGCAGTGTGGCTAAGGGATTACCCCATC    Outer forward: GGATGGACTCAGGCCTCTAGCACGA Outer reverse: AGTTTCTGGGCGGGATCTGGCG  Inner forward: ATTCCTGGCGCGGCGGGAGGACCAT  Inner reverse: TGAAGGCCTCACATGTGAGGGTCAGAACG    Outer forward: GCTTTCCCATCTTTCTCCCCATAGGC  Outer reverse: AAATTCAGGGAAGTGCCCTAGCCCAT  Inner forward: GTTGTGAGTCTGACCCCACACTCCTTG  Inner reverse: AAGCCAGGCTTCCACTCCAGCACT    Outer forward: ATCCCCCTGTCCTGCAGGTATGAGG  Outer reverse: CCTCTGAGGAATACTCCAACCTGCCCA  Inner forward: GATACAGGGTCTGGCTGCTGGCAAA  Inner reverse: GTCCACTGTCCCCCAAGGCATGAC    Outer forward: TCAGGGCTACACTTTCTCGGGGAGACCCA Outer reverse: CCACAGGGTTCCCTATCACCCTCGGGTC  Inner forward: GCTCTTGGGGGGCTTCTGCTCCTCTCAAA  Inner reverse: AGAGGACCCCCCCGACACAGGAGGAAC  Product size forT allele: 140 Product size for C allele: 182  Product size of two outer primers: 264    62 /32  Product size for T allele: 178 Product size for C allele: 138  Product size of two outer primers: 260    66 /35  Product size for C allele: 144 Product size for T allele: 189  Product size of two outer primers: 281  62 /32  Product size for A allele: 229 Product size for G allele: 150  Product size of two outer primers: 331  60 /35  Product size for T allele: 210 Product size for C allele: 164  Product size of two outer primers: 320    62 /32  Product size for G allele: 141 Product size for A allele: 103  Product size of two outer primers: 193  64 /32  Product size for A allele: 130 Product size for G allele: 160  Product size of two outer primers: 241  66 /32 Product size for A allele: 145 Product size for G allele: 175  Product size of two outer primers: 264  66/32  c.65C>T  c.494C>T  c.803G>A  c.1233C>T  c.2871G>A  c.3047G>A  c.3230A>G  Annealing temperature (°C)/Cycles 58    SNP  Primers (5’ to 3’)  Product size (bp)  Annealing temperature (°C)/Cycles     c.3315G>A  Outer forward: GTGGGGGGCTTGCATAGGGTCACTGAG   Outer reverse: TCAACCTGATGCTAACGGCAGGGCTTCA  Inner forward: CCCCACACCTTCATCCTGGAAGGGCA  Inner reverse: TCATATTCGTTCCTGACTCGGTCCTCTGCC  Product size for A allele: 177  Product size for G allele: 212  Product size of two outer primers: 333  64 /32 59    2.9. Genotyping by RFLP  Seven  genetic  variants  (c.151C>T,  c.294C>T,  c.349G>A,  c.1339G>A,  c.2223C>T,  c.2289C>T  and c.2398C>T) in NPHS1 were genotyped using restriction fragment length polymorphism  (RFLP).  A  web‐based  programme  NEBcutter  v2.0  (http://tools.neb.com/NEBcutter2/)  was  used to search for the presence of a restriction site near the polymorphic region.  PCR was  performed, using 30ng of gDNA to get the target region for enzyme digestion, as described  in section 2.5. Different restriction enzymes have different incubation conditions as shown  in Table 6. After 16 hours of incubation at the respective restriction enzymes’ temperature,  the  digested  products  were  run  on  a  2%  gel  to  differentiate  the  undigested  and  digested  samples.     Various  restriction  enzymes  from  New  England  Biolabs  were  used  in  the  genotyping  of  NPHS1  SNPs.  The  wildtype  genotype  of  c.151C>T  was  digested  by  AluI.  The  wildtype  genotype  of  c.294C>T  and  c.2289C>T  were  digested  by  TaqI.  The  wildtype  genotype  of  c.349G>A was digested by Hpy188I. The mutant genotype of SNP c.1339G>A was digested  by  MseI.  The  wildtype  genotype  for  SNP  c.2223C>T  was  digested  by  BtsCI.  The  wildtype  genotype for SNP c.2398C>T was digested by SfoI.         60  Table 6: Restriction enzymes used for the genotyping of NPHS1 SNPs.    SNP  Restriction  Incubation condition Product size (bp)  enzyme  151C>T  AluI  Buffer 4: 1µl AluI (2U): 0.2µl  Water: 1.8µl  Template: 7ul  Incubate 16 hours at 37◦C  294C>T    Buffer 3: 1µl BSA: 0.1µl  TaqI (1U): 0.1µl  Water: 1.8µl  Template: 7µl  Incubate 16 hours at 65◦C  TaqI  2289C>T  349G>A  Hpy188I  Buffer 4: 1µl Hpy188I(1U): 0.1µl  Water: 1.9µl  Template: 7µl  Incubate 16 hours at 37◦C  1339G>A  MseI  Buffer 4: 1µl MseI(1U): 0.1µl  Water: 1.9µl  Template: 7µl  Incubate 16 hours at 37◦C  2223C>T  BtsCI  Buffer 4: 1µl BtsCI(1U): 0.1µl  Water: 1.9µl  Template: 7µl  Incubate 16 hours at 50◦C  2398C>T  SfoI  Buffer 4: 1µl SfoI(1U): 0.1µl  Water: 3.9µl  Template: 5µl  Incubate 16 hours at 37◦C  Genotype CC:  116 + 147   Genotype CT:   116 + 147 + 263  Genotype TT:   263     Genotype CC:  466 + 66   Genotype CT:   466 + 66 + 532  Genotype TT:   532   Genotype CC:  131 + 126   Genotype CT:   131 + 126 + 257   Genotype TT:   257     Genotype GG:  412 + 119  Genotype GA:   532 + 412 + 119  Genotype AA:   532    Genotype CC:  201  Genotype CT:   201 + 156 + 45   Genotype TT:   156 + 45     Genotype CC:  76 + 65   Genotype CT:   141 + 76 + 65   Genotype TT:   141    Genotype CC:  145 + 67   Genotype CT:   212 + 145 + 67  Genotype TT:   212     61  2.10. Statistical analysis   A  web‐based  program  SNPstats  (http://bioinfo.iconcologia.net/SNPstats)  was  used  to  analyze the association of the various polymorphisms with childhood idiopathic NS between  patients and healthy controls. Data on Hardy‐Weinberg disequilibrium, allele and genotype  frequencies,  linkage  disequilibrium,  haplotype  frequency  estimation  and  analysis  of  association  of  haplotypes  with  a  particular  response  were  generated.  A  p‐value  less  than  0.05 was viewed to be statistically significant.    For the analysis of composite genotypes of NPHS1 and NPHS2 genetic variants with NS, SPSS  for Windows (version 17) was used. Binary logistic regression was performed. A p‐value less  than 0.05 was viewed to be statistically significant.    The allele frequencies of the genetic variants were used to determine if they were rare. Rare  variants were defined as having minor allele frequencies of less than 1% in controls. Fisher’s  exact  test  was  performed  to  compare  the  accumulation  of  rare  variants  between  NS  patients  and  controls  with  GraphPad  Prism  (version  5.02).  Rare  variant  accumulation  was  also analyzed between patients of different clinical phenotypes. A p‐value of less than 0.05  was viewed to be statistically significant.         62  2.11. Prediction of the effect of the genetic variants  For  the  non‐synonymous  genetic  variants  identified  in  NPHS1  and  NPHS2,  two  web‐based  programs  PolyPhen2  (http://genetics.bwh.harvard.edu/pph2/)  and  Sorting  intolerant  from  tolerant  (SIFT)  sequence  (http://sift.bii.a‐star.edu.sg/www/SIFT_seq_submit2.html)  were  used to predict the effect of the amino acid substitution.  PolyPhen2  predicts the effect  of  the amino acid substitution on the three‐dimensional and function of the protein based on  physical  and  comparative  considerations  (Sunyaev  et  al.,  2001).  SIFT  predicts  the  effect  of  the amino acid substitution on the function of the protein based on sequence homology and  the physical properties of the amino acid (Ng and Henikoff, 2003).      For  genetic  variants  that  were  identified  in  the  promoter  region,  a  web‐based  program,  ConSite  (http://asp.ii.uib.no:8090/cgi‐bin/CONSITE/consite)  was  used  to  predict  if  there  is  any  transcription  factor  binding  site  at  the  position  of  the  variants  using  phylogenetic  footprinting.    For all the SNPs identified, we would use the Potentially Functional SNP (PFS) search engine  (http://pfs.nus.edu.sg/)  to  look  up  the  RefSNP  accession  ID  (rs  number)  for  the  individual  SNP. For those SNPs with rs number, we would search the rs number against the PFS search  engine to identify the effect of the SNPs. The PFS search engine looks at the effect of coding  region SNPs on exon splicing enhancer (ESE) or exon splicing silencer (ESS) sites, the effect of  non‐synonymous SNPs on the protein and also the effect of synonymous SNPs on the codon  usage difference (top and last 5%) (Wang et al., 2011).      63  For genetic variants without rs number or not available in the PFS search engine, the effect  of  the  variants  on  ESE/ESS  was  predicted  using  three  web‐  based  programs,  RESCUE‐ESE  (http://genes.mit.edu/burgelab/rescue‐ese/),  ESE  finder  3.0  (http://rulai.cshl.edu/cgi‐ bin/tools/ESE3/esefinder.cgi?process=home)  and  PESX:  Putative  Exonic  Splicing  Enhancers/Silencers  (http://cubweb.biology.columbia.edu/pesx/).  RESCUE‐ESE  predicts  the  splicing phenotypes by identifying nucleotide changes that disrupt or alter the predicted ESE  (Fairbrother et al., 2004). ESE finder 3.0 analyzes the exon sequences and predicts if there  are any ESEs responsive to the human SR proteins (SF2/ASF, SC35, SRp40 and SRp55). It can  hence  predicts  if  the  exonic  mutations  will  affect  these  elements  (Cartegni  et  al.,  2003).  PESX is used to identify the presence of ESE and/or ESS in the exon sequences and whether  the mutations would affect the sites (Zhang and Chasin, 2004).     2.12. Protein sequence alignment  For genetic variants that were predicted to affect the protein structure or affect the ESE or  ESS, a comparative genomic survey was performed between different organisms to see how  strongly the sites are conserved. The nephrin and podocin human protein sequences were  compared  with  the  protein  sequences  of  Rattus  norvegicus  (Norway  rat)  (Nephrin:  NP_072150;  Podocin:  NP_570841),  Mus  musculus  (house  mouse)  (Nephrin:  NP_062332;  Podocin:  NP_569723),  Bos  Taurus  (cattle)  (Nephrin:  NP_001179441;  Podocin:  NP_001193036),  Pan  troglodytes  (chimpanzee)  (Nephrin:  XP_524228;  Podocin:  XP_003308663) and Canis familiaris (dog) (Nephrin: XP_541685; Podocin: XP_547443). The  protein  sequences  were  obtained  from  NCBI  protein  database  64  (http://www.ncbi.nlm.nih.gov/protein).  Alignment  of  the  multiple  protein  sequences  was  performed using Clustal Omega (http://www.clustal.org/omega/) (Sievers et al., 2011).          65  3. Results  3.1. Clinical Characteristics  Chinese patients  Ninety‐seven  Chinese  patients  with  sporadic  idiopathic  NS  and/or  FSGS  were  recruited  (Table  7).  The  mean  age  of  diagnosis  was  5.65±4.24  years  (range  0.9  to  19  years).  There  were  58  males  and  39  females.  Seventy  patients  (72.2%)  were  steroid‐responsive  and  28  (28.7%)  were  steroid‐resistant.    Forty‐nine  patients  (50.5%)  needed  steroid‐sparing  immunosuppression therapy. Sixty patients (61.9%) had a renal biopsy performed. Patients  were  classified  according  to  whether  they  had  minor  glomerular  abnormalities  (28.3%),  diffuse  mesangial  hypercellularity  (1.7%),  focal  mesangial  proliferative  glomerulonephritis  (FMPGN)  (11.7%),  focal  global  sclerosis  (FGS)  (11.7%),  IgM  nephropathy  with  mesangial  injury  (11.7%),  diffuse  mesangial  proliferative  glomerulonephritis  (1.7%),  and  focal  segmental glomerulosclerosis (FSGS) (33.3%) (Appendix I).  Figure 4A shows the distribution  of the histological profile in the Chinese nephrotic patients who underwent a biopsy.      Eight (8.2%) patients, all of whom were steroid and therapy‐resistant, progressed to ESRD.  Their mean age at diagnosis  was  8.48±5.76 years  (range  1.9 to  19.0 years), progressing to  ESRF between 0.8 to 9.0 years. Seven of these patients had FSGS on renal biopsy, while 1  had diffuse mesangial proliferative glomerulonephritis.     Forty‐three  patients  (44.3%)  had  been  treated  with  a  calcineurin‐inhibitor  (CNI),  either  cyclosporine and/or tacrolimus, and of these, 15 patients (34.8%) did not respond. Among  the  patients  with  minor  abnormalities,  57.5%  (23  out  of  40)  had  been  treated  with  66  cyclophosphamide  and  6  of  them  (26.1%)  were  resistant.  Those  patients  who  did  not  respond  to  cyclophosphamide  were  then  treated  with  cyclosporine,  of  whom  2  remained  resistant to cyclosporine. Patients with poor prognosis were defined as patients who were  resistant to both prednisolone and CNI therapy, or who had progressed to end stage renal  disease.  Eighteen patients (18.6%) in our cohort were defined as having poor prognosis. Malay patients  Twenty‐four Malay patients were included in this study (Table 7). The mean age at diagnosis  was  5.84±5.25  years  (range  1.61  to  22.0  years).  There  were  16  males  and  18  females.  Twelve (50%) were steroid‐resistant and 12 (50%) were steroid‐responsive. Fifteen patients  (62.5%)  needed  steroid‐sparing  immunosuppression.  Renal  biopsy  was  performed  on  18  patients  (75%).  Patients  were  classified  according  to  whether  they  had  minor  glomerular  abnormalities (33.3%),  FMPGN (5.6%), FGS (5.6%), IgM nephropathy with mesangial injury  (5.6%)  and  FSGS  (50.0%)  (Appendix  I).  Figure  4B  shows  the  distribution  of  the  histological  profile in the Malay nephrotic patients who received a biopsy.    Four (16.7%) patients, all of whom were steroid and therapy‐resistant, progressed to ESRD.  Their  age  at  diagnosis  was  12.25±9.64  years  (range  3  to  22  years),  of  whom  all  were  had  FSGS on biopsy.   Progression to ESRF occurred over 0.5 to 4.0 years.    Fourteen  patients  (25.9%)  had  been  treated  with  cyclosporine  and  seven  of  them  (50%)  were  not  responsive.  None  had  been  treated  with  tacrolimus.    Four  patients  with  minor  abnormalities (44.4%) had been treated with cyclophosphamide, of whom all responded to  therapy. Five patients (20.8%) were defined as having poor prognosis.      67  It  was  observed  that  the  percentage  of  Malay  patients  (50%)  who  were  steroid  resistant  were significantly higher compared to that of the Chinese patients (27.8%) (p=0.038) (Table  7).     Table  7:  General  characteristics  of  Chinese  and  Malay  patients  with  idiopathic  sporadic  nephrotic syndrome and/or FSGS. p‐value    Chinese patients  Malay patients  (n=97)  (n=24)  Male :Female ratio  3: 2  2: 1  0.536*  Age, mean ± SD, years  17.15±7.08  18.83±7.28  0.259** Age at onset,   5.65±4.24  5.84±5.25  0.628**  SRNS  28 (28.9%)  12 (50%)  0.0375*  FSGS (%)  20 (33.3%)  9(50%)  0.199*  5 (20.8%)  0.766  4 (16.7%) 0.252*** mean ± SD, years  Steroid and CNI resistance  18 (18.6%)  (%)  Renal failure (%)  8 (8.2%) * Chi‐square test; **Mann‐Whitney test; ***Fisher exact test  SRNS: Steroid resistant nephrotic syndrome; FSGS: Focal segmental glomerulonephritis; CNI:  Calcineurin inhibitor          68    Figure 4 4: Distribution of the h histological profile in tthe nephrottic patientss.  (A)  Chinese  ;(B)  Malay.  FGSS:  Focal  glo obal  glomeerulosclerossis.  FMPGN N:  Focal  mesangial  proliferrative glomeerulonephriitis. FSGS: FFocal segmeental glomeerulonephrittis. MCNS:  Minimal  change  nephroticc  syndrome.  Others:  Diffuse  mesangial  m hypercellularity  and  diffuse  mesanggial proliferaative glomeerulonephrittis    3.2. Geenetic varian nts in NPHSS1  3.2.1. Screening ffor variantss in NPHS1 u using HRM We  scrreened  for  variants  in n  the  prom moter  region  (~300bassepairs  upsstream),  all  the  29  exonic  regions and d 3’UTR of tthe gene ussing HRM. A As certain amplicons in n the promo oter and  3’UTR d did not have e the optim mal melting domains fo or analysis, d direct sequeencing was used to  screen  those particular region ns.  We ideentified a to otal of 16 geenetic variaants in our  patients  (Table 8 8), of which 5 (c.‐170T>>C, c.494C>>T, c.1233C>>T, c.2871G G>A and 304 47G>A) werre novel.    The  tem mperature  shifted  curvves  and  the  difference  plots  for  genetic  vaariants  identified  in  Exon  2  and  17,  ass  well  as  th he  correspo onding  DNA A  electroph horetogram ms  are  as  sh hown  in  Figure  5  and  Figure  6  respecctively.    Th he  genetic  variants  v weere  determined  by  thee  LC480  Gene  Scanning  sofftware  v1.5 5  (Roche  Diiagnostics) which  anallyzed  changges  in  fluorrescence  levels frrom the meelting curve. The variou us settings ffor the analysis of the melting currve were  69  considered  to  be  appropriate  if  the  temperature  shifted  melting  curve  and  the  difference  plot  corresponded  to  each  other.  Genetic  variants  determined  by  HRM  were  validated  by  direct sequencing. For certain polymorphic regions (promoter, exon 3, 17 and 26), RFLP or  tetra‐ARMS PCR were used to determine the genotypes for the patients.  Controls were also  genotyped for the various genetic variants using either RFLP or tetra‐ARMS PCR (Figure 7).     Table 8: Genetic variants identified in NPHS1 using HRM.  Position   SNP ref  Variant  Amino acid  Promoter  ‐  ‐170T>C  ‐  2  rs116617171  c.65C>T  p.Ala22Val  2  rs114385015 c.151C>T p.Leu51Leu  3  rs2285450 c.294C>T p.Ile98Ile  3  rs3814995  c.349G>A  p.Glu117Lys  4  ‐  c.494C>T  p.Ala165Val  7  rs115308424  c.803G>A  p.Arg268Glu  10  ‐  c.1233C>T  p.Asn411Asn  11  rs28939695  c.1339G>A  p.Glu447Lys  17  rs2073901  c.2223C>T  p.Thr741Thr  17  rs437168  c.2289C>T  p.Val763Val  18  rs114896482 c.2398C>T p.Arg800Cys 21  ‐ c.2871G>A p.Val957Val  22  ‐  c.3047G>A  p.Ser1016Asn  24  rs4806213  c.3230A>G  p.Asn1077Ser  26  rs2071327  c.3315G>A  p.Ser1105Ser  70  Figure  5:    Normalized  high  resolution  melting  cu urves  and  the  t corresp ponding  diffference  plots.   in  their  pe  by  theirr  relative  differences  d The  variants  are  differentiatted  from  the  wild‐typ ure shifted  and differe ence plots o of Exon 2 which has  normalized floresccence. (A‐B)) Temperatu pe genotypees for both variants  two rarre genetic variants (c.65C>T and c.151C>T). TThe wild‐typ are  indicated  in  blue,  b whereas  the  genetic  variantts  are  indiccated  in  bro own  (c.65C C>T)  and  mperature  shifted  an nd  differen nce  plots  of  o Exon  17 7  which  green  (c.151C>T).  (C‐D)  Tem contain ned  two  com mmon  SNPs  (c.2223C>>T  and  c.22 289C>T)  weere  identifieed  in  patien nts.  The  heterozzygous  genotypes  (red d  and  greeen)  are  weell  differenttiated  from m  the  homozygous  genotyp pes (blue). 71  Figure 6 6: Electroph horetogram ms of the vaarious genettic variants.   All  the  genetic  vaariants  are  named  acccording  to  their  t nucleotide  posittions  in  thee  coding  strand.  (A)  Wild‐tyype  C  allelee  and  variaant  T  allele  (heterozyggous  form)  of  c.65C>TT.  Codon  GCG codes for alan nine, while codon GTG codes for vvaline. (B) W Wild‐type C allele and vvariant T  heterozygou us form) of  c.151C>T.  Both CTG aand TTG cod de for leucine. (C) Wild d‐type C  allele (h allele  and  variant  T  allele  (heterozygous  form)  of  c.2223C>T..  Both  ACC  and  ACT  code  c for  us form)  threonine. (D) Wild‐type C allele and variant T allele (heterozygous and  homozygou 89C>T. Both h GTG and G GTT code fo or valine.   of c.228                 72    Figure 7 7: Genotypiing results o of c.2289C> >T and c.3315A>G.   (A)  Gen notyping  off  c.2289C>TT  performed d  using  RFLLP.  The  hom mozygous  wild‐type  w geenotype  was diggested by TTaqI generating a band d that contaains both the 131 bp  and 126 bp p bands.  The  homozygous  mutant  m gen notype  was  not  cut  byy  the  enzym me,  which  iss  representted  by  a  ous  genotype  is  represented  by  2  bands  (one  band  off  131  bp  257  bp  band.  The  heterozygo 6  bp  and  another  a ban nd  of  257  bp).  b (B)  Genotyping  off  c.3315A>G G  performeed  using  and  126 tetra‐ARMS PCR. A A control baand of 333  bp is seen  for all the  samples. TThe A‐allele  specific  ozygous wild‐type genotype. The  G‐allele sp pecific of  band off 177 bp is  observed in the homo 212  bp   is  observe ed  in  homo ozygous  mutant  genotype.  As  forr  the  heterrozygous  geenotype,  nd 212 bp b bands are ob bserved.   both the 177 bp an   73  3.2.2. Statistical analysis of the genetic variants in NPHS1   All  genetic  variants  were  consistent  with  the  Hardy‐Weinberg  equilibrium  in  both  the  Chinese and Malay controls, with the exception of c.294C>T in the Chinese controls, which  was most likely due to chance (Appendix II).     In  our  Chinese  population,  c.‐170T>C  was  in  moderate  linkage  disequilibrium  (LD)  with  c.65C>T (D’: 0.65, p=0.002), c.294C>T, c.349G>A and c.2223C>T (D’: 0.78, pT  and  c.2871G>A  (D’:  0.96,  p=0.023)  and  c.2289C>T  (D’:  0.86,  pT  was  in  weak  LD  with  c.151C>T  (D’:  0.23,  pT (D’: 0.58, p=0.023) but strong LD with c.2871G>A (D’: 0.97, pT  was in strong LD with c.1233C>T, c.2223C>T and c.2289C>T (D’: 0.96, pA was  in  strong  LD  with  c.65C>T,  c.151C>T,  c.294C>T,  c.2223C>T,  c.2289C>T  and  c.2398C>T  (D’:  0.99, pA was in strong LD with c.151C>T, c.349G>A and c.3315G>A (D’: 0.99,  pT (D’: 0.22, p=0.0038). c.1399G>A was in moderate LD  with  c.349G>A  (D’:  0.69,  p=0.024)  and  c.3315G>A  (D’:  0.55,  p=0.0055).  c.2223C>T  was  in  strong LD with c.2289C>T (D’: 0.99, pG and c.3315G>A was in strong LD  (D’: 0.99, p=0.006) (Appendix III).    In  our  Malay  patients,  c.‐170T>C  was  in  weak  LD  with  c.3315G>A  (D’:  0.32,  p=0.0032),  moderate  LD  with  c.3230A>G  (D’: 0.69,  pT,  c.349G>A  (D’: 0.84‐0.86, pT and 2289C>T (D’=0.95‐0.99, pT was in  strong LD with c.349G>A, c.3047A>G, c.3315G>A (D’=0.99, pA (D’: 0.89,    pT was in strong LD with c.349G>A, c.2223C>T and c.2289C>T (D’: 0.95‐ 0.99,  pA  was  in  strong  LD  with  c.803G>A,  2223C>T  and  c.2289C>T  (D’:  74  0.90‐0.99,  pA  was  in  strong  LD  with  c.3047G>A  and  c.3315G>A  (D’:  0.99,  pT was in weak LD with c.3315G>A (D’: 0.32, p=0.012), moderate LD with  c.3230A>G  (D’:  0.69,  p=0.0002)  and  strong  LD  with  c.2223C>T  and  c.3047G>A  (D’:  0.99,  pT and c.349G>A on exon 3, and c.2289C>T on exon 17.  Table  9  and  Table  12  show  the  allele  frequencies  of  the  significant  SNPs  in  Chinese  and  Malay patients and controls. Significant differences in allele frequencies were observed for  c.294C>T and c.2289C>T among the patients and various controls for both Chinese (p=0.016  and  p=0.0003  respectively)  and  Malays  (p=0.035  and  p=0.0024  respectively).  Allelic  differences for c.349G>A (p=0.018) was observed only in the Chinese patients (Table 9 and  Table 12).     Table  10  and  Table  13  showed  the  genotype  frequencies  of  the  significant  SNPs  in  the  Chinese  and  Malay  populations.  In  the  Chinese  and  Malay  patients,  c.294C>T,  and  c.2289C>T  were  both  significantly  associated  with  NS,  whereas  c.349G>A  was  significantly  associated  with  NS  in  only  the  Chinese  patients  (Table  11  and  Table  14).c.294C>T  was  significantly associated with NS under the dominant (OR: 1.74, 95% CI 1.04‐2.90, p=0.035),  and log‐additive (OR: 1.86, 95%CI 1.15‐3.02, p=0.012) models in Chinese patients and under  the dominant (OR: 3.06, 95% CI 1.19‐7.86, p=0.027), and over‐dominant (OR: 3.35, 95% CI  1.30‐8.68, p=0.018) models in Malay patients. c.349G>A was significantly associated with NS  under  the  codominant  (OR:  2.52,  95%  CI  1.22‐5.19,  p=0.043),  recessive  (OR:  2.02,  95%  CI  1.09‐3.75,  p=0.028)  and  log‐additive  (OR:    1.57,  95%  CI  1.09‐2.25,  p=0.014)  models  in  75  Chinese patients. c.2289C>T was significantly associated with NS under the codominant (OR:  2.24,  95%CI  1.47‐3.41,  p=0.0008),  dominant  (OR:  2.22,  95%CI  1.47‐3.34,  p=0.0002),  overdominant  (OR:  2.16,  95%CI  1.43‐3.26,  p=0.0003)  and  log‐additive  (OR:  1.83,  95%CI:  1.31‐2.55, p=0.0006) models in Chinese patients and under the codominant (C/T) (OR: 2.60,  95%CI  1.01‐6.70,  p=0.026),  codominant  (T/T)  (OR:  10.71,  95%CI  1.40‐82.00,  p=0.026),  dominant (OR: 3.06, 95%CI 1.26‐7.46, p=0.017), and log‐additive (OR: 2.88, 95% CI 1.37‐6.07,  p=0.0073) models in Malay patients.     Based on the HapMap data on c.349G>A, in the European population, G allele is the major  allele  (68%)  while  A  allele  is  the  minor  allele  (32%).  The  frequencies  of  G/G,  G/A  and  A/A  genotypes are 43%, 49% and 8% respectively. In contrast, in the Han Chinese population in  Beijing, A allele was identified to be the major allele (67%) while G allele is the minor allele  (33%).  Likewise,  the  frequencies  of  G/G,  G/A  and  A/A  genotypes  are  32%,  51%  and  17%  respectively. Our allele and genotype frequencies are in agreement to that observed in the  Han Chinese population in Beijing, with reference to the data described in Table 9 and Table  10.    76  Table 9: Allele frequencies of NPHS1 SNPs in Chinese (n=97).    c.294C>T  Controls  c.349G>A Patients    Patients   (n=221)  (n=97)  386  155  (87%)  (80%)  56  39  T  Controls  Patients (n=1903)  (n=97)  3169   142 (83%)  (73%)  637  52 (17%)  (27%)    (n=221)  (n=97)  172 95 C  G  C    Controls c.2289C>T (39%)  (49%)  270 99 A  T  (13%)  (20%)  (61%)  p=0.016    p=0.018 (51%)  p=0.0003    Table 10: Genotype frequency for NPHS1 SNPs in Chinese (n=97).  Controls  Patients  c.294C>T  Controls Patients c.349G>A  (n=221)  165  (n=97)  (n=221)  61  CC  28 (n=97)  22 GG  (75%)  (63%)  56  33  CT  (34%)  0  3  TT  (13%)  (23%)  116 51 (3%)  (n=1903)  (n=1903)  1320  49 (69%)  (51%)  529  44 (28%)  (45%)  54  4 (3%)  (4%)  CT  (52%)  (53%)  77 24 AA  (0%)  Patients CC  GA  (25%)  Controls  c.2289C>T  TT  (35%)  (25%)                      77  Table 11: Association analysis of NPHS1 genotypes with nephrotic syndrome in Chinese.  c.294C>T  c.349 G>A   Patients (n=97); Controls (n=221)  Patients (n=97); Controls (n=221)  c.2289C>T Patients (n=97); Controls (n=1903)  Codominant (C/T): Codominant (G/G):  OR: 2.24 (1.47‐3.41), p=0.0008  OR: 2.52 (1.22‐5.19), p=0.043  Dominant:  Patients   Dominant:   Recessive:  OR: 1.74 (1.04‐2.90), p=0.035  vs Controls  OR: 2.22 (1.47‐3.34), p=0.0002  OR: 2.02 (1.09‐3.75), p=0.028  Log‐additive:  Overdominant:  Log‐additive:    OR: 1.86 (1.15‐3.02), p=0.012  OR: 2.16 (1.43‐3.26), p=0.0003  OR:  1.57 (1.09‐2.25), p=0.014  Log additive:    OR: 1.83 (1.31‐2.55), p=0.0006                        78  Table 12:Allele frequencies of NPHS1 SNPs in Malay (n=24).  c.294C>T  Controls  c.2289C>T Patients    (n=185)  342  (n=24)  40  Patients (n=185)  (n=24)  333 36 (90%)  (75%)  37 12 (25%)  C  C  (92%)  (83%)  28  8  T    Controls   T  (8%)  (17%)  (10%)  p=0.035    p=0.0024   Table 13: Genotype frequencies of NPHS1 SNPs in Malay (n=24).  Control  Patient  c.294C>T  Control Patient (n=185)  (n=24)  c.2289C>T  (n=185)  159  (n=24)  16  CC  150 14 (0.81)  (0.58)  33 8 (0.18)  (0.33)  2 2 (0.01)  (0.08)  CC  (0.86)  (0.67)  24  8  CT  CT  (0.13)  (0.33)  2  0  TT  TT  (0.01)  (0)                                    79  Table 14: Association analysis of NPHS1 genotypes with nephrotic syndrome in Malay  (n=24).    c.294C>T  c.2289C>T  Co‐dominant (C/T):  OR: 2.60 (1.01‐6.70), p=0.026  Co‐dominant (T/T)  Dominant:  OR: 10.71 (1.40‐82.00), p=0.026  OR: 3.06 (1.19‐7.86),p=0.027    Over dominant:    OR: 3.35 (1.30‐8.68), p=0.018  Dominant:  Patients (n=24)   vs   Controls (n=185)  OR: 3.06 (1.26‐7.46), p=0.017  Log‐additive:  OR: 2.88 (1.37‐6.07), p=0.0073  80  3.2.3. Prediction of the effect of the genetic variants   Effect on function of protein  Out of the 16 genetic variants identified, 8 of them were non‐synonymous. The effect of the  amino acid substitution on protein function was predicted using two web‐based programs,  SIFT and Polyphen2. SIFT looks at whether the amino acid substitution is tolerable. It gives a  SIFT score which is a scaled probability that amino acid would appear at that position in the  alignment (Ng and Henikoff, 2003). Polyphen2 predicts whether a genetic variant is probably  damaging, possibly damaging or benign. When a genetic variant is predicted to be probably  damaging, the prediction is made with high confidence that the amino acid substitution will  affect the function or structure of the protein. A genetic variant that is possibly damaging is  supposed  to  affect  the  function  or  structure  of  the  protein.  A  genetic  variant  that  is  predicted to be benign is most likely not to have any phenotypic effect (Sunyaev et al., 2001).     Only  1  genetic  variant  (c.3230  A>G)  was  predicted  by  SIFT  to  affect  the  protein  function.  Polyphen2 predicted 3 variants (c.65C>T, c.803G>A and c.3047G>A) to be benign, 4 variants  (c.349G>A,  c.494C>T,  c.1339G>A  and  c.3230A>G)  to  be  possibly  damaging  and  1  variant  (c.2398  C>T)  to  be  probably  damaging  (Table  15).  c.349G>A,  c.494C>T,  c.1339G>A,  c.2398C>T and c.3230A>G were well conserved among nephrin of other mammalian species  (Appendix IV).            81  Table 15: Prediction of the effect of amino acid substitution of the various genetic variants  on the protein using SIFT and PolyPhen2.  Variant  Amino Acid  Prediction by SIFT Prediction by PolyPhen2 c.65C>T  p.Ala22Val  Tolerated  Benign      Score: 0.76  Score difference: NA  c.349G>A  p.Glu117Lys  Tolerated  Possibly damaged      Score: 0.48  Score difference: 1.609  c.494C>T  p.Ala165Val  Tolerated Possibly damaged      Score: 0.12  Score difference: 1.577  c.803G>A  p.Arg268Glu  Tolerated  Benign      Score: 0.16  Score difference: 1.362  c.1339G>A  p.Glu447Lys  Tolerated  Possibly damaged      Score: 0.16  Score difference: 1.544  c.2398C>T  p.Arg800Cys  Tolerated  Probably damaging      Score: 0.09  Score difference: 2.223  Tolerated Benign  Score: 0.17  Score difference: 1.270  Affect protein structure  Possibly damaged  Score: 0.00  Score difference: 1.838  c.3047G>A  p.Ser1016Asn      c.3230A>G  p.Asn1077Ser                  82  Presence of ESE and/or ESS sites  For all the genetic variants that are in the coding region, they were analyzed to determine if  the  change  in  nucleotide  would  change  the  ESE  and/or  ESS  sites  using  various  ESE/ESS  search engines. Out of the 15 genetic variants in the exon regions, 3 of them (c.1339G>A,  c.2398C>T and c.3047G>A) did not have any effect on the ESE/ESS sites for their nucleotide  change.  Five  genetic  variants  (c.65C>T,  c.151C>T,  c.803G>A,  c.2223C>T  and  c.2289C>T)  resulted in the loss of ESE/ESS sites in the presence of the mutant nucleotide. The change in  nucleotide for 6 genetic variants (c.294C>T, c.349G>A, c.494C>T, c.2871G>A, c.3230G>A and  c.3315G>A)  resulted  in  a  gain  of  ESE/ESS  sites.  The  change  in  nucleotide  for  c.1233C>T  resulted  in  both  gain  and  loss  of  ESE/ESS  sites  (Table  16).  c.65C>T,  c.151C>T,  c.294C>T,  c.803G>A,  c.2223C>T,  c.2289C>T  and  c.2871G>A  were  observed  to  be  well  conserved  among nephrin of other mammalian species (Appendix IV).         83  Table 16: Effect of the individual NPHS1 SNPs on ESE/ESS sites.  Position   Variant  Effect on ESE/ESS sites  2  c.65C>T (‐) SF2/ASF  2  c.151C>T  (‐)GTGGAGCT  3  c.294C>T  (+) ACATTGAGGC  3  c.349G>A  (+) CTAAGATG  4  c.494C>T  (+) GTGAAG  (+)TGAAGC  7  c.803G>A  (‐)SF2/ASF  10  c.1233C>T  (‐)ACAACG  (+)SRp40  (+)ATGGTC  11  c.1339G>A  ‐  17  c.2223C>T  (‐)ACCATCCG  17  c.2289C>T  (‐)SF2/ASF  18  c.2398C>T  ‐  21  c.2871G>A  (+)SRp55  22  c.3047G>A  ‐  24  c.3230A>G (+)TCCAGT  26  c.3315G>A (+)TCAGAA  (+)SRp40        84  3.3. Genetic variants in NPHS2   3.3.1. Screening of NPHS2 using direct sequencing  For NPHS2, direct sequencing was used to screen the promoter region (~1000basepairs  upstream) and all the eight exons. We have identified a total of 8 genetic variants in our  patients (Table 17), of which 1 (c.685C>A) was novel. The DNA electrophoretograms of the  various genetic variants are illustrated in Figure 8.    Table 17: Genetic variants identified in NPHS2 using direct sequencing.  Position  SNP ref  Variant  Amino acid  Promoter  ‐  c.‐670C>T ‐  Promoter  ‐  c.‐116C>T ‐  Promoter  rs12406197  c.‐51G>T   ‐  2  rs3738423  c.288C>T  p.Ser96Ser  5  ‐  c.685C>A  p.Arg229Arg  7  rs74315348  c.871C>T  p.Arg291Trp  8  rs1410592  c.954T>C  p.Ala318Ala  8  rs3818587  c.1038A>G  p.Leu346Leu        85    Figure 8 8: Electroph horetogram ms of the vaarious genettic variants.   (A‐B)  W Wild‐type  C  allele  and  variant  T  allele  (heteerozygous  and  a homozzygous  form m)  of  c.‐ 670C>TT  and  c.‐11 16C>T  (C)  Wild‐type  G  allele  and  variant  T  allele  (heterozygo ( ous  and  homozyygous form) of c.‐51G>>T (D) Wild‐‐type C allele and variaant T allele  (heterozyggous and  homozyygous form)) of c.288C>>T. Both TCC and TCT ccode for serrine. (E) Wild‐type C alllele and  variant  A allele (he eterozygous form) of cc.685C>A. B Both CGA aand AGA code for argin nine. (F)  Wild‐type  C  allele e  and  variant  T  allele  (heterozyggous  form)  of  c.871C> >T.  CGG  co odes  for  argininee,  while  TG GG  codes  for  tryptop phan.  (G)  Wild‐type  W T allele  an T  nd  variant  C  allele  (hetero ozygous and d homozygo ous form) of c.954T>C.. Both GCT  and GCC co ode for alan nine. (H)  Wild‐type A allele  and variant G allele (h heterozygou us and hom mozygous fo orm) of c.10 038A>G.  Both  CTA leucine. A  a and  CTG  code  fo or  86  3.3.2. Statistical analysis of the genetic variants in NPHS2  All  genetic  variants  were  consistent  with  the  Hardy‐Weinberg  equilibrium  in  both  the  Chinese and Malay controls (Appendix II).    In  our  Chinese  population,  c.‐670C>T,  c.‐116C>T  and  c.288C>T  was  in  strong  LD  with  c.‐ 51G>T (D’: 0.83, pT (D’: 0.85, pT  (D’: 0.99, p=0.0086), c.288C>T (D’: 0.82, pC (D’: 0.90, pT (D’: 0.67, p=0.0003) (Appendix III).    In our Malay population, c.‐670C>T was in moderate LD with c.1038A>G (D’: 0.74, p=0.0025)  and  strong  LD  with  c.‐116C>T,  c.‐51G>T  and  288C>T  (D’:  0.92‐0.99,  pT  was  in  moderate  LD  with  c.954T>C  (D’:  0.79,  pT  (D’:  0.99, p=0.0001). c.288C>T was in strong LD with c.1038A>G (D’: 0.88, pC  was in strong LD with c.1038A>G (D’: 0.73, p=0.0036) (Appendix III).    Significant  allelic  differences  were  observed  between  the  Chinese  patients  and  controls  in  c.‐51G>T  (p=0.0091)  and  c.288C>T  (p=0.0021)  (Table  18).  There  was  no  allelic  difference  observed  in  the  Malay  patients.  The  genotype  frequencies  for  the  significant  SNPs  were  illustrated in Table 19. In our Chinese patients, out of the 8 genetic variants,  c.-51G>T was significantly associated with NS under the dominant (OR:  2.11,  95%CI  1.11‐4.00,  p=0.021), and log-additive (OR:  2.19,  95%CI  1.20‐3.99,  p=0.0088) models in Chinese patients.  87  c.288C>T was observed to have a protective effect in NS  under the codominant (OR: 0.25, 95%CI 0.11-0.58, p=0.0011), dominant (OR:  0.28,  95%CI  0.13‐0.61,  p=0.0004), overdominant (OR: 0.25, 95%CI 0.11-0.58, p=0.0002) and log-additive (OR: 0.33, 95%CI 0.16-0.69, p=0.0009) models in Chinese patients  (Table  20).  There  was  no  association observed for the Malay patients.       88  Table 18: Allele frequencies of NPHS2 variants in Chinese (n=97).    c.‐51G>T  Controls  c.288C>T Patients (n=125)  (n=97)  229   162   Patients (n=221)  (n=97)  386 185  (87%)  (95%)  56  9  (5%)  C  G  (92%)  (84%)  21   32   T  T    Controls     (8%)  (16%)  (13%)  p=0.0091    p=0.0021     Table 19: Genotype frequencies of NPHS2 variants in Chinese (n=97).  Control  Patient Control Patient (n=221)  (n=97)  c.288C>T  c.‐51G>T  (n=125)  104  (n=97)  68  167 89 (0.76)  (0.92)  52 7 (0.24)  (0.7)  2 1 (0.01)  (0.01)  CC  CC  (0.83)  (0.70)  21  26  CT  CT  (0.17)  (0.27)  0  3  TT  TT  (0)  (0.03)  89  Table 20: Association analysis of c.‐51G>T and c.288C>T in Chinese patients (n=97).   Promoter: c.‐51G>T  Exon 2: c.288C>T  Patients (n=97); Controls(n=125)  Patients (n=97); Controls (n=125       Co‐dominant (C/T):  OR: 0.25 (0.11‐0.58), p=0.0011  Dominant:  Dominant:   OR: 0.28 (0.13‐0.61), p=0.0004  Patients vs  OR: 2.11 (1.11‐4.00), p=0.021  Overdominant:  Controls  Log‐additive:  OR: 0.25 (0.11‐0.58), p=0.0002  OR: 2.19 (1.20‐3.99), p=0.0088  Log‐additive:  OR: 0.33 (0.16‐0.69), p=0.0009      90  3.3.3. Prediction of the effect of the genetic variants   Presence of transcription factor binding sites   Three SNPs (c.‐51G>T, c.‐116C>T and c.‐670C>T) were identified in the promoter regions.  The web‐based program, ConSite, did not find any transcription factor binding sites for  c.‐116C>T  and  c.‐670C>T.  As  for  c.‐51G>T,  it  was  predicted  to  be  the  binding  site  (TCCCGTG)  for  upstream  stimulatory  factor  (USF)  in  homo  sapiens.  The  change  of  nucleotide  from  G  to  T  resulted  in  the  loss  of  the  binding  site  to  USF  and  hence  the  transcription of podocin would most likely be affected.     Effect on the function of protein  Out  of  the  8  genetic  variants  identified,  only  c.871C>T  denotes  a  non‐synonymous  amino acid substitution in Exon 7. This nucleotide change results in the change of amino  acid  from  arginine  which  is  polar  and  positively‐charged  into  tryptophan  that  is  non‐ polar and neutral‐charged.  c.871C>T was present in 2 of the Chinese patients and was  not present in the 125 Chinese cord blood controls.     SIFT predicted that R291W would affect the function of the protein with a SIFT score of  0.01.  This  prediction  is further  supported  by  PolyPhen2,  which  also  predicted  that  the  amino acid substitution is probably damaging to the protein, with a score difference of  2.439.  c.871C>T  was  well  conserved  among  podocin  of  other  mammalian  species  (Appendix IV).    91  Presence of ESE and/or ESS sites  Out of the 5 genetic variants in the coding regions, 2 of them (c.685C>A and c.871C>T)  did not have any effect on the ESE/ESS sites. The remaining 3 genetic variants (c.288C>T,  c.954T>C and c.1038A>G) resulted in the loss of ESE/ESS sites when there was a change  in  the  nucleotide  (Table  21).  c.288C>T,  c.954T>C  and  c.1038A>G  were  well  conserved  among podocin of other mammalian species (Appendix IV).    Table 21: Effect of the individual NPHS2 SNPs on ESE/ESS sites.  Position  Variant  Effect on ESE/ESS sites  2  c.288C>T  (‐)SRp55  5  c.685C>A  ‐  7  c.871C>T  ‐  8  c.954T>C  (‐) TGCTGCT  8  c.1038A>G  (‐)ACTGAA                    92  3.4. Phenotype‐genotype associations  NPHS1  and  NPHS2  SNPs  were  tested  for  genotype‐phenotype  associations  in  Chinese  and  Malay  patients.  Clinical  features  analyzed  included  nephrotic‐range  proteinuria,  progression  to  ESRD,  the  use  and  response  to  various  immunosuppressive  drugs  (namely  prednisolone,  calcineurin  inhibitors,  mycophenolate  mofetil  and  cyclophosphamide),  need  for  steroid‐sparing  immunosuppression,  ability  to  maintain  remission  for  at  least  five  years,  and  ability  to  be  weaned  off  all  immunosuppressants  for  more  than  five  years.  Several  SNPs  in  NPHS2  were  significantly  associated  with  certain  phenotypes  in  our  Chinese  patients  (Table  22  and  Table  23).  There  was  no  association  between  the  genotypes  and  phenotypes  observed  in  our  Malay  patients.  This could be due to the small sample size.    c.‐51G>T  was  found  to  be  associated  with  steroid  resistance  under  the  dominant (OR:  2.87  95%CI  1.13‐7.27,  p=0.026)  and  log  additive  model  (OR:  2.62  95%CI  1.17‐5.87),  p=0.018)  in  Chinese  patients  (Table  22).  Twenty‐eight  Chinese  patients  were  steroid  resistant, of whom 13 patients (46.4%) had the mutation for c.‐51G>T, as compared to  16 of 69 patients with steroid sensitive nephrotic syndrome (SSNS) (23.2%). Significant  allele difference was observed between SRNS and SSNS patients (p=0.014) (Table 23).      c.1038A>G,  was  associated  with  steroid  resistance  under  the  log‐additive  model  (OR:  3.09 95%CI 1.10‐8.70, p=0.030) (Table 22).  Eight out of 28 SRNS patients (28.5%) had the  93  mutation A/G or G/G, compared to 8 out of 69 SSNS patients (11.6%). Significant allele  difference was observed between SRNS and SDNS patients (p=0.016) (Table 23).      c.‐51G>T was also found to be associated with the use of cyclosporine (CsA) under the  dominant model (OR: 2.87, 95%CI 1.13‐7.27, p=0.026) and log‐additive model (OR: 2.62,  95%CI 1.17‐5.87, p=0.018) in Chinese patients (Table 22). The mutation was found in 18  of 44 patients who used CsA (40.9%), and in only 11 of 53 patients (20.8%) who did not  use the drug. There was significant allele difference observed (p=0.044) (Table 23).    Forty‐four patients were on CsA, of whom 15 were resistant to therapy. c.1038A>G was  associated  with  CsA  resistance  under  the  dominant  (OR:  5.78,  95%CI  1.19‐28.04,  p=0.024) and log‐additive (OR: 5.39, 95% CI 1.20‐24.30, p=0.016) models (Table 22). Six  of 15 CsA‐resistant patients (40%) had the mutation, as compared to only 3 of 29 CsA‐ responders  (10.7%).    There  was  significant  allele  difference  observed  (p=0.029)  (Table  23).                      94  Table 22: Genotype‐phenotype association in Chinese patients for NPHS2.  Steroid resistance    Resistant to  Cyclosporine  Use of cyclosporine       Dominant:  OR: 2.87 (1.13‐7.27),  p=0.026  c. ‐51G>T  Dominant:   OR: 2.64(1.08‐6.47),  p=0.031    Log additive:  OR: 2.62 (1.17‐5.87),  p=0.018  Log additive:  OR: 2.30 (1.04‐5.09),  p=0.034    No association  observed    Dominant:  OR: 5.78(1.19‐28.04),  p=0.024    Log‐additive:   OR: 3.29 (1.16‐9.31),  c.1038A>G  p=0.023  No association observed    Log‐additive:   OR: 5.39 (1.20‐24.30),  p=0.016                        95  Table 23: Allele frequencies for NPHS2 SNPs with phenotype‐genotype associations in  Chinese patients.      Steroid    Use of CsA    Resistant to CsA  resistance  c.‐51G>T      No  (n=69)    T  121  17   (88%)  Yes  (n=28)    G    G  T        No  94  12        (89%)  (11%) 68   20        (78%)  (22%)       (12%)  (n=53) 41   (73%)    15   Yes  (27%)  (n=44) p=0.014*    p=0.031*  c.1038A>G    A  G          A  G    130  8         No  55  3     No  (n=69)  (94%)  (6%)  Yes  9  47  (n=28)  (84%)            (16%)  p=0.016*  (n=29)  (95%)  (5%)  Yes  7  23  (n=15)  (77%)          (23%)  p=0.028**  * Chi‐square test **Fisher‐exact            96  3.5. Analysis of composite genotypes of genetic variants in NPHS1 and NPHS2  In  our  cohort  of  patients,  NPHS1  c.2289C>T  and  NPHS2  c.‐51G>T  were  individually  associated  with  NS,  while  c.288C>T  seemed  to  have  a  protective  effect.  Hence,  we  selected  these  3  SNPs  to  perform  an  association  analysis  of  the  three  combined  genotypes, using binary logistic regression. To increase the sample size analyzed, both  the  Chinese  and  Malay  populations  were  combined  and  analysis  was  performed  with  race as the covariate.      Composite genotypes referred to the combination of genetic variants of different genes  found  on  different  chromosomes.  A  total  of  13  different  combinations  of  composite  genotypes  (‐51/288/2289)  were  observed  in  controls  and  patients  from  the  3  polymorphisms.  The  wild‐type  (GG/CC/CC)  was  the  most  common  in  our  controls  and  patients  and  was  taken  to  be  the  reference  group.  When  the  patients  have  the  composite  genotype  (GG/CC/CT),  there  was  significant  association  with  NS  (OR:  3.53,  95%CI  1.82‐6.86,  pT, c.1233C>T,  c.2398C>T, c.2871G>A and c.3047G>A) were rare variants, with minor allele frequencies  of  less  than  1%  in  controls.  Of  the  8  genetic  variants  identified  in  NPHS2,  2  of  them  (c.685C>A and c.871C>T) were rare variants (Table 26).  Of the 8 rare variants, 5 of them  (c.65C>T,  c.494C>T,  c.2398C>T  and  c.3047G>A  (NPHS1)  and  c.871C>T  (NPHS2))  were  non‐synonymous variants. Polyphen2 predicted c.65C>T and c.3047G>A (NPHS1) to be  benign,  c.494C>T  (NPHS1)  to  be  possibly  damaging  and  c.2398C>T  (NPHS1)  and  c.871C>T (NPHS2) to be damaging (Table 27).                                                   101  Table 26: Rare NPHS1 and NPHS2 variants identified in NS patients and their minor  allele frequencies in patients and controls.   Amino acid  No. of  Race  NS  Controls  Rare variant   SNP  change  patients patients  MAF  reference  (n=221)  MAF  (n=121)          NPHS1      c.65C>T  rs116617171  p.Ala22Val  3  Chinese 0.01  0.002  c.494C>T  ‐  p.Ala165Val  1  Chinese 0.004  0  c.1233C>T  ‐  p.Asn411Asn  1  Chinese 0.004  0  c.2398C>T  rs114896482  p.Arg800Cys  2  Chinese 0.008  0.007  c.2871G>A  ‐  p.Val957Val  1   Chinese 0.004  0  c.3047G>A  ‐  p.Ser1016Asn  1  Malay  0.004  0                NPHS2              c.685C>A  ‐  p.Arg229Arg  1  Chinese 0.004  0  c.871C>T  rs74315348  p.Arg291Trp  2  Chinese 0.008  0  Boldface indicates minor alleles. SNP: single‐nucleotide polymorphism; NS: nephrotic  syndrome; MAF: minor allele frequency                        102  Table  27:  Non‐synonymous  NPHS1  and  NPHS2  rare  variants  identified  in  nephrotic  syndrome patients and their predicted effects  Gene  Nucleotide  Amino acid  Predicted  Reported functional  change  change  effect*  studies  NPHS1  c.65C>T  p.Ala22Val  Benign  ‐    c.494C>T  p.Ala165Val  p.Ser1016Asn Possibly  damaged  Probably  damaging  Benign    c.2398C>T  p.Arg800Cys    c.3047G>A    NPHS2  ‐  ‐          c.871C>T  p. Arg291Trp  Probably  damaging  (Zhang et al., 2004,  ‐  Nishibori et al., 2004)  *The effect of the amino acid substitutions were predicted using Polyphen2.     As there were no allele differences for the rare variants between the Chinese and Malay  controls,  comparisons  between  patients  and  controls  and  between  different  phenotypes  were  performed  on  the  combined  data  of  both  the  Chinese  and  Malay  patients.     There was a significant accumulation of rare variants in patients with NS compared to  healthy  controls  (12  total  variants  in  121  NS  diploid  genomes  compared  to  4  total  variants  in  221  control  diploid  genomes,  p=0.0021)  (Table  28).  This  corresponded  to  a  significantly  higher  carrier  frequency  of  7.4%  in  NS  patients  compared  to  1.8%  in  controls  (p=0.0151).  Analysis  of  only  non‐synonymous  rare  variants  revealed  a  significant  burden  of  9  variants  in  NS  patients  compared  to  4  variants  in  controls,  103  corresponding  to  a  significantly  higher  carrier  frequency  of  6.6%  in  NS  patients  compared to 1.8% in controls (p=0.0299).    Table 28: Rare variants accumulation in nephrotic syndrome patients and controls      NS Patients  Controls    Total no. of  242  442  alleles  All variants  NPHS1  9   4   (synonymous and  non‐synonymous)    NPHS2  3   0         Non‐synonymous  variants only    Total      12      NPHS1  7   4   NPHS2  2   0     Total  9  4      4  p=0.0021*  p=0.0161*    NS: nephrotic syndrome; * Fisher’s Exact test                  104  Comparison of the rare variant accumulation was also performed between patients with  different  clinical  phenotypes  such  as  drug  therapy,  drug  response,  whether  they  have  poor prognosis as defined in Section 3.1 and whether they progressed to ESRF.    Twenty‐three  of  the  121  patients  (19%)  were  defined  as  having  poor  prognosis.  A  significant  accumulation  of  rare  variants  was  identified  in  patients  who  had  poor  prognosis  compared  to  those  with  good  prognosis.  A  total  of  8  rare  variants  were  identified in 23 diploid genomes in patients with poor prognosis, as compared to 4 total  variant  in  98  diploid  genomes  in  patients  with  good  prognosis  (p=0.0003).  This  corresponded to a significantly higher carrier frequency of 21.7% in patients with poor  prognosis  compared  to  4.1%  in  patients  with  good  prognosis  (p=0.012).  When  the  analysis  was  only  performed  on  the  non‐synonymous  SNPs,  a  significant  burden  of  5  rare variants was observed in patients with poor prognosis compared to 4 rare variants  in  patients  with  good  prognosis  (p=0.014).  This  corresponded  to  a  higher  carrier  frequency  of  17.4%  in  patients  with  poor  prognosis  than  4.1%  in  patients  with  good  prognosis (p=0.042) (Table 29).                           105  Table 29: Accumulation of rare variants in patients with poor prognosis.      Patients with good  Patients with poor  prognosis  prognosis        Total alleles  196  46  All rare variants  NPHS1  0   3   (synonymous and  non‐synonymous)    NPHS2  4   5     Total  4  8  p=0.0003*      Non‐synonymous  rare variants only          NPHS1  0   2   NPHS2  4   3     Total  4  5      p=0.0140*  *Fisher’s exact test    Of  the  121  Chinese  and  Malay  patients,  heterozygous  rare  variants  were  found  in  9  patients. One patient (Patient 94) had 2 rare variants in NPHS1 and 2 in NPHS2 while the  rest  had  only  one  rare  variant  each.  Table  30  illustrates  the  clinical  characteristics  of  these 9 patients with rare variants. Six of the 9 patients (66.7%) were steroid‐resistant,  while 5 patients (55.5%) were resistant to both steroid and calcineurin inhibitor therapy.  Of  these  9  patients,  7  (77.8%)  had  a  renal  biopsy,  3  of  whom  (42.9%)  had  histology‐ proven FSGS (Table 30).   106  Table 30: Clinical characteristics of patients with rare variants.  Patient  Gender  Age at  Renal Biopsy  Steroid  no.  diagnosis  resistant  Progression  On  Poor  Rare  to ESRF  Cyclosporine  Prognosis  variants  treatment  identified  No  Yes  No  NPHS1  c.2398C>T  No  No  No  NPHS1  c.2398C>T  20  F  16  Not performed  No  80  F  3  No  83  M  5.47  Mesangial injury, GN  with IgM deposits and  global sclerosis  Focal Global Sclerosis  Yes  No  Yes*  Yes  64  F  5.77  Not performed  No  No  No  No  94 M  11.01  FSGS, scalloping of  glomerular basement  membrane on electron  microscopy  Yes  Yes  Yes*  Yes  M  3  FSGS  Yes  No  Yes  Yes    232  NPHS1  c.1233C>T  NPHS1  c.494C>T  NPHS1  c.65C>T  c.2871G>A NPHS2  c.685C>A  c.871C>T  NPHS1  c.65C>T    252  M  17  MCNS  Yes  No  Yes  No  NPHS1  c.65C>T    58  F  11  Diffuse Mesangial  Proliferative  Glomerulonephritis  with Global and  Yes  Yes  Yes*  Yes  NPHS2  c.871C>T  107  Patient  Gender  Age at  Renal Biopsy  no.  diagnosis  Steroid  resistant  Progression  On  Poor  Rare  to ESRF  Cyclosporine  Prognosis  variants  treatment  identified  Segmental Sclerosis  3  F  21  FSGS and tubular   Yes  No  Yes*  Yes  NPHS1  interstitial infiltrates  c.3047G>A GN: Glomerulonephritis; MCNS: Minimal change nephrotic syndrome; FSGS: Focal segmental glomerulosclerosis  *Resistant to Cyclosporine treatment  108  Two  of  the  9  patients  (22.2%)  progressed  to  ESRD  (Patients  94  and  58).  Interestingly,  these are the only 2 patients with c.871C>T rare variant in NPHS2. Patient 94 who had 4  rare  variants  (NPHS1:  c.65C>T  and  c.2871G>A;  NPHS2:  c.685C>A  and  c.871C>T)  presented with NS at 12 years of age, and renal biopsy then showed FSGS. He was the  first  child  of  non‐consanguinous  parents  and  had  no  family  history  of  renal  disease.  Mutational  analysis  showed  that  c.65C>T  (NPHS1)  and  c.871C>T  (NPHS2)  were  of  paternal  origin,  while  c.2871G>A  (NPHS1)  and  c.685C>A  (NPHS2)  were  of  maternal  origin (Figure 9). Of note, his siblings and parents had normal urine dipstick analysis. His  renal  function  was  normal  at  presentation,  and  he  was  treated  with  prednisolone,  cyclosporine and mycophenolate mofetil, to which he did not respond. He progressed to  chronic kidney disease stage 2 at 13 years of age, and reached end‐stage renal failure at  17 years, when he was started on peritoneal dialysis. He received a living‐related renal  transplant from his mother at 19 years old. He was managed with methylprednisolone,  azathioprine  and  basilixumab  perioperatively,  and  subsequently  maintained  on  prednisolone,  tacrolimus  and  mycophenolate  mofetil.  He  achieved  good  urine  output  and  renal  function  post‐transplant.  There  was  mild  proteinuria  of  1.35g/day/1.73m2  which  resolved  with  initiation  of  enalapril.  To  date  at  1.5  years  post‐transplant,  his  serum creatinine was 110‐120 μmol/L (eGFR 50‐60 ml/min/1.73m2) and his urine total  protein was 0.32g/day/1.73m2. No post‐transplant renal biopsy had been performed.    The other ESRD patient (Patient 58) had only a single rare variant NPHS2 c.871C>T. She  presented  at  3  years  of  age  with  mild  proteinuria  which  was  not  followed  up.  She  109  presented  again  at  11  years  old  with  NS  and  hypertension,  and  renal  biopsy  then  showed mesangial proliferative glomerulonephritis with global and segmental sclerosis.  She  was  resistant  to  steroids,  cyclophosphamide  and  cyclosporine,  and  renal  function  progressively  deteriorated.  She  reached  ESRD  at  21  years  of  age  and  was  put  on  peritoneal dialysis. She received a deceased donor renal transplant at 24 years of age,  which  was  complicated  by  severe  acute  tubular  necrosis.  She  received  methylprednisolone, azathioprine, cyclosporine , and anti‐thymocyte globulin. She had  no  recurrence  of  proteinuria.  To  date  at  8  years  post‐transplant,  she  has  been  maintained on prednisolone, mycophenolate and tacrolimus and continued to have no  proteinuria, but had chronic kidney disease stage 3 due to chronic allograft dysfunction.     110    Fiigure 9: Ped digree of Pattient 94.   All the other  A family members were te ested negatiive for proteeinuria. Geno otyping of th he  raare variants were perforrmed for the e family mem mbers of Pattient 94.  111    4. Discussion  Idiopathic NS is the most common glomerular disease in children, occurring in 20 to 30  per million children annually. Although the majority is due to MCNS with good prognosis,  up to 40% can be steroid‐dependent or steroid‐resistant. Progression to end stage renal  failure may occur, especially in those patients who are diagnosed with FSGS on biopsy.  Recent  molecular  studies  have  identified  several  podocyte‐associated  genes  to  be  implicated  in  the  pathogenesis  of  childhood  NS.  Some  examples  are  nephrin  (NPHS1),  podocin  (NPHS2),  wilm’s  tumor‐1  (WT‐1)  and  α‐actinin‐4  (ACTN4).  Geographical  and  ethnic differences in genotypes and clinical outcomes have been described, especially in  Caucasian populations (Caridi et al., 2003, Weber et al., 2004).     Mutations  in  NPHS1  and  NPHS2  have  been  associated  with  early  onset  of  heavy  proteinuria,  and  rapid  progression  to  end‐stage  renal  disease.  This  suggests  that  both  nephrin  (NPHS1)  and  podocin  (NPHS2)  have  an  important  role  in  the  maintenance  of  glomerular permeability (Huber et al., 2003b). NPHS1 and NPHS2 were the initial genes  to  be  identified  as  the  main  genetic  causes  of  congenital  NS  and  familial  autosomal  recessive  SRNS  respectively  (Lenkkeri  et  al.,  1999,  Boute  et  al.,  2000).  Subsequent  studies  have  also  identified  mutations  in  NPHS1  in  other  forms  of  NS  (Philippe  et  al.,  2008, Santin et al., 2009b). NPHS2 mutations were also found in patients with sporadic  NS (Karle et al., 2002, Weber et al., 2004) . Geographical and inter‐racial differences in  mutation  profile  have  been  widely  reported  in  NS.  This  is  particularly  evident  in  the  mutation spectrum of NPHS2, which is one of the most extensively studied genes in NS.  112  Some  common  genetic  variants,  such  as  R138Q  and  R229Q,  have  been  described  in  Caucasian nephrotic cohorts (Jungraithmayr et al., 2011, Weber et al., 2004, Caridi et al.,  2003) but have not been described in any of the Asian studies, as well as in our study  (Mao  et  al.,  2007,  Kitamura  et  al.,  2006,  Yu  et  al.,  2005,  Yu  et  al.,  2004).  For  studies  performed  on  sporadic  SRNS  in  the  European  cohort,  the  prevalence  of  NPHS2  mutations (homozygous or compound heterozygous) was 12 to 19% (Weber et al., 2004,  Ruf et al., 2004, Caridi et al., 2003). In contrast, a study performed by Maruyama et al.  did not find any NPHS2 mutation in 36 Japanese children with SRNS (Maruyama et al.,  2003).  Similar  studies  conducted  by  Yu  et  al.  and  Mao  et  al.  observed  a  mutation  prevalence of 4.3% and 6.7% respectively.     In  our  present  study,  which  aimed  to  explore  the  spectrum  and  frequency  of  genetic  variants  of  NPHS1  and  NPHS2  in  young  Singapore  Chinese  and  Malay  patients  with  idiopathic  NS,  we  also  observed  a  lower  prevalence  of  NPHS2  mutations  (2.1%)  in  Singapore Chinese children, but no mutations in our Malay patients. This low incidence  of  NPHS2  mutations  in  the  Asian  population  indicates  that  NPHS2  mutations  have  a  lower  genetic  contribution  to  disease  pathogenesis  in  the  nephrotic  cohort  of  Asian  origin.    113  4.1. Polymorphisms and their association with nephrotic syndrome  Studies have demonstrated that polymorphisms, even without amino acid substitution,  could result in phenotypic variations either by affecting the structure of mRNA or by the  inactivation of genes through the change in gene splicing machinery to skip the mutant  exons (Mao et al., 2007). Hence, polymorphisms could have a role in the development  of  idiopathic  NS  and  might  also  contribute  to  the  different  phenotypes  seen  in  these  patients.     Out of the 16 genetic variants identified in NPHS1, 10 were polymorphic, with a minor  allele frequency of more than 1% in controls. Two SNPs (c.294C>T and c.2289C>T) were  found  to  be  significantly  associated  with  NS  in  our  Chinese  and  Malay  patients.  c.349G>A was found to be significantly associated with NS in only the Chinese patients.  c.294C>T  and  c.2289C>T  does  not  result  in  any  amino  acid  substitution,  whereas  c.349G>A results in an amino acid substitution of glutamic acid to lysine at position 117.  This  amino  acid  change  has  been  predicted  to  be  tolerated  by  SIFT,  and  possibly  damaging by PolyPhen2.    c.294C>T  was  found  in  Japanese  SRNS  patients  with  a  minor  allele  frequency  of  15%.  c.349G>A was reported in European MCNS patients and Japanese SRNS patients with a  minor allele frequency of 48% and  37% respectively. c.2289C>T had been identified in  MCNS patients in the European cohort, Chinese patients with sporadic NS and Japanese  SRNS  patients,  with  a  minor  allele  frequency  of  4%,  18.3%  and  13%  respectively  114  (Lahdenkari et al., 2004, Kitamura et al., 2006, Mao et al., 2007). In our present study,  the minor allele frequency for c.294C>T was 20% in Chinese patients and 17% in Malay  patients.  This  was  similar  to  the  findings  in  the  Japanese  population.    However,  the  minor allele frequency for c.2289C>T was 27% in our Chinese patients, and 25% in our  Malay patients. This was higher than those in the previous studies, suggesting that this  SNP could be more important in our population.     c.294C>T, c.349G>A and c.2289C>T have been found to affect the ESE and/or ESS sites.  Variants  which  cause  the  disruption  in  ESE  sites  could  result  in  loss  of  splice  sites,  creation  of  splice  sites,  activation  of  the  cryptic  splice  site  and  interfering  with  the  splicing elements (Cartegni et al., 2003).  Although both c.294C>T and c.2289C>T do not  have  any  amino  acid  substitutions,  they  could  affect  the  expression  and  function  of  nephrin  through  the  ESE  and/or  ESS  sites.  Functional  studies  would  be  needed  to test  the effect of these variants on nephrin function.    Of  the  8  NPHS2  variants  that  were  identified  in  our  populations,  6  of  them  were  polymorphic, with a minor allele frequency of more than 1% in the controls. Moreover,  c.‐670C>T,  c.‐51G>T,  c.288C>T,  c.954T>C  and  c.1038A>G  have  been  reported  in  other  Chinese  populations  (Zhu  et  al.,  2009,  Mao  et  al.,  2007,  Yu  et  al.,  2005).  However,  no  genotypic  or  allelic  differences  had  been  reported  for  these  polymorphisms  in  the  reported studies.    115  On the other hand, c.‐51G>T was found to be associated with NS in our Chinese patients.  The frequency of the minor allele T was two‐fold higher in patients (16%) compared to  controls  (8%).  c.‐51G>T  has  been  shown  to  be  a  functional  polymorphism.  The  web‐ based  program,  Consite,  predicted  c.‐51G>T  to  be  the  binding  site  (TCCCGTG)  for  upstream  stimulatory  factor  (USF)  in  homo  sapiens.  Di  Duca  et  al.  reported  a  downregulation of the reporter gene expression when the podocytes were transfected  with  the  variant.  c.‐51G>T  binds  to  the  upstream  stimulatory  factor‐1  (USF‐1)  trans  element,  which  is  a  main  regulatory  factor  of  podocin  expression  in  human  glomeruli.  When USF‐1 RNA expression was silenced in podocytes transfected with the wild‐type  constructs  containing  ‐51G  sequence,  there  was  a  significant  reduction  of  luciferase  expression. This supported the concept that USF‐1 has a functional implication (Di Duca  et al., 2006).    c.288C>T  seemed  to  have  a  protective  effect  on  our  nephrotic  Chinese  patients.  The  minor allele frequency was more than two‐fold lower in the patients (5%) compared to  controls  (13%).  This  protective  effect  was  not  observed  in  other  studies  on  Chinese  patients from the Northern and Eastern China.  Yu et al. observed a minor frequency of  7%  in  their  SRNS  patients  compared  to  4%  in  controls,  whereas  Zhu  et  al  observed  a  minor allele frequency of 10% in their minimal change disease patients as compared to  7.4% in controls. Both studies did not observe any genotypic or allelic differences (Yu et  al.,  2005,  Zhu  et  al.,  2009).  This  disparity  could  be  due  to  either  difference  in  allelic  frequencies or genetic variations between the Northern and Southern Chinese. Chinese  116  from  Singapore  population  mainly  originated  from  the  Southern  provinces,  such  as  Guangdong  and Fujian.  Suo et al.  observed a north‐south cline in genetic variations in  China when they performed a systemic survey of the population genetics data from two  Han Chinese populations, Han Chinese from Beijing and Singapore Chinese with South  China  ancestries.  They  have  identified  genomic  regions  that  explained  the  observed  north‐south cline in genetic differences in China (Suo et al., 2012).    The  NPHS1  and  NPHS2  polymorphisms  were  significantly  associated  with  NS  under  different  inheritance  models.  To  select  for  the  best  model  statistically,  the  Akaike  information (AIC) that was provided by SNPstats can be used as a reference. The  best  model  would  be  the  model  with  the  lowest  AIC.  However,  the  best  model  cannot  be  selected simply based on the statistical results. The best model selected needs to fit the  underlying  biology.  The  association  analysis  was  based  on  cases  and  controls  (i.e.  2  phenotypes).  Both  the  co‐dominant  and  log‐additive  models  require  at  least  3  phenotypes  to  determine  if  it  is  the  best  model.  An  example  of  analysis  would  be  to  compare  the  controls,  mild  cases  (i.e.  proteinuria  only)  and  severe  cases  (i.e.  develop  nephrotic syndrome).  Hence, in this scenario of case‐control study, the model selection  should  be  limited  to  only  dominant,  recessive  or  over‐dominant.  Functional  studies,  such as gene expression study, can also be performed to see which model is the most  appropriate for the variant. A variant which leads to a loss‐of‐function can typically be  explained  by  the  recessive  model,  while  a  variant  that  causes  a  gain‐of‐function  can  typically be represented by the dominant model.   117  In our Malay patients, we did not observe any unique NPHS1 SNPs or any NPHS2 SNPs  that were associated with NS. These findings suggest that NPHS1 and NPHS2 may not be  the  genes,  accounting  for  NS  in  Malays,  nor  are  they  responsible  for  the  poorer  prognosis of NS seen in the Malays in our population. Other genes, such as MYH9 and  PLCE1, need to be investigated in order to identify the gene(s) that may be responsible  for the poorer outcomes in Malays.  Exome sequencing, which is an efficient method to  identify  candidate  genes  by  targeted  sequencing  of  all  the  protein  coding  regions  (exomes),  could  be  performed  on  our  Malay  patients  (Ng  et  al.,  2009).  Through  this  strategy, the unique genetic variants that could account for their phenotypes could be  identified.    4.2. Phenotype‐genotype correlations of the genetic variants  Studies have been carried out in various ethnic populations to investigate the effect of  NPHS1 variants on the disease phenotypes. Philippe et al identified NPHS1 mutations in  their  patients  who  presented  with  renal  disease  later  in  childhood,  were  steroid‐ resistant  and  had  a  histological  spectrum  ranging  from  MCD  to  FSGS.  Santin  et  al.  performed NPHS1 mutational analysis on their patients with familial and sporadic SRNS.  Both  groups  observed  that  patients  with  compound  heterozygous  NPHS1  mutations  with at least 1 mild mutation tended to have a later onset and milder course of disease.  Patients with 2 severe mutations are more likely to present with congenital onset of the  disease.  Such  genotype‐phenotype  correlations  highlighted  the  importance  of  NPHS1  118  mutation  screening  in  SRNS  patients  for  disease  prognostication  (Philippe  et  al.,  2008,  Santin et al., 2009b).   p.R229Q  is  the  most  commonly  described  NPHS2  polymorphism  in  NS,  and  has  been  found  in  both  familial  and  sporadic  SRNS,  as  well  as  in  patients  with  FSGS.  In  patients  with  NS,  this  polymorphism  has  been  found  to  be  in  the  homozygous,  compound  heterozygous or heterozygous state (Karle et al., 2002, Tsukaguchi et al., 2002, Caridi et  al.,  2003,  Ruf  et  al.,  2004,  Weber  et  al.,  2004).  The  arginine  residue  at  position  229  is  highly  conserved  and  in‐vitro  studies  have  shown  that  this  polymorphism  could  cause  decreased binding to nephrin. It has been hypothesized that p.R229Q may pre‐dispose  to  FSGS  or  SRNS  in  the  presence  of  a  second  pathologic  NPHS2  mutation.  p.R229Q  homozygosity has never been demonstrated in familial or sporadic FSGS. This suggests  that  p.R229Q  is  not  a  highly  pathogenic  mutation  and  is  less  likely  to  have  a  serious  consequence  in  its  heterozygous  state  in  the  absence  of  a  second  NPHS2  mutation  (McKenzie et al., 2007).      Resistance to therapy is a major characteristic in patients with poor prognosis (Caridi et  al.,  2009).    Steroids  are  the  mainstay  of  therapy  for  children  with  NS,  about  20%  of  whom  are  resistant  to  treatment.  SRNS  patients  are  at  a  significantly  higher  risk  of  developing  complications  of  the  disease,  and  progression  to  chronic  kidney  disease  or  end  stage  renal  failure.  The  patient’s  initial  response  to  corticosteroids  is  the  most  important  prognostic  factor  for  long‐term  remission  in  NS  (Gbadegesin  and  Smoyer,  2008).   119    Kitamura  et  al.  carefully  selected  15  families  with  SRNS  with  phenotypes  that  were  comparable with the Caucasian populations and screened them for NPHS1, NPHS2 and  NEP1. They only managed to identify known polymorphisms of NPHS1 and NPHS2, and  hence concluded that NPHS1 and NPHS2 were not the cause of SRNS in their population  (Kitamura et al., 2006).  Mao et al screened 60 Chinese patients for mutations in NPHS1  and NPHS2 but did not find any correlation of phenotype‐genotype in NPHS1 and NPHS2.  In  our  Chinese  patients,  genotype‐phenotype  analysis  showed  that  c.‐51G>T  and  c.1038A>G were associated with SRNS, using the log‐additive model. The minor allele in  both  SNPs  modified  the  risk  of  developing  steroid  resistance  in  an  additive  model.  Patients with 2 copies of the minor allele were at a higher risk than those with 1 copy of  the minor allele. Our study also demonstrated that c.1038A>G was associated with CsA  resistance under the dominant and log‐additive models. The G‐allele was about 4 times  higher in CsA‐resistant patients (23%) compared to CsA‐responsive patients (5%).    4.3. Gene‐gene interaction of NPHS1 and NPHS2  The identification of gene‐gene interactions will allow us to understand the mechanisms  through which genes act to control the expression of certain traits. The failure to model  a gene‐gene interaction in an analysis could result in incorrect conclusions with regard  to the mode of inheritance and the estimation of the degree of the genetic effects (Luo  et  al.,  2006).  To  investigate  if  there  were  any  gene‐gene  interactions  between  NPHS1  and NPHS2, we performed composite genotype analysis for 3 significant SNPs, namely  120  c.‐51G>T, c.288C>T (NPHS2) and c.2289C>T (NPHS1). We observed that this composite  genotype  (GG/CC/CT)  was  significantly  associated  with  NS  in  our  patients.  This  composite genotype was observed at a frequency of 27% in patients, which was twice  that  in  controls  (10%).  Another  composite  genotype  (GT/CC/CT)  was  significantly  associated with NS with an increased odds ratio compared to GG/CC/CT. The frequency  of GT/CC/CT was 3 times more in patients (12%) compared to controls (4%). Therefore  our data showed that c.2289C>T (NPHS1) in its heterozygous form can be a risk factor  for NS, and together with heterozygous c.‐51G>T (NPHS2), the risk for the development  of NS increased.     Composite  genotype  analysis  also  revealed  that  the  composite  genotype  (GT/CC/CC)  was associated with steroid resistance. The frequency of this genotype was more than  doubled in SRNS patients (18%) compared to SSNS patients (7%). Caridi et al. considered  that  clinical  phenotypes  associated  with  heterozygous  mutations  could  reveal  the  specific features of the disease and be of clinical importance (Caridi et al., 2009). In our  study,  we  have  shown  that  c.‐51G>T  in  its  heterozygous  form  was  associated  with  steroid  resistance.  Another  composite  genotype  (GG/CC/TT)  was  also  associated  with  steroid  resistance  in  our  patients,  although  individual  analysis  of  c.2289C>T  did  not  show any significance. The frequency in SRNS (10%) was 5 times that of SSNS patients  (2%).  These  findings  suggested  that  there  could  be  some  gene‐gene  interactions  between  NPHS1  and  NPHS2  that  could  lead  to  the  development  of  idiopathic  NS,  or  121  determine  the  response  to  therapy.  Functional  studies  are  needed  to  validate  this  hypothesis.     Mutational studies have been performed to investigate the inter‐relationship of NPHS1  and NPHS2 on NS in children. Kozeill et al first described the co‐existence of NPHS1 and  NPHS2  mutations  in  patients  with  congenital  FSGS.  They  observed  that  homozygous  mutations present in one gene, together with a third mutation in another gene, resulted  in  modification  of  the  disease  phenotype  from  congenital  nephrotic  syndrome  of  the  Finnish  type  to  congenital  FSGS.  This  observation  was  suggestive  of  a  tri‐allelic  hit,  indicating  that  there  could  be  a  functional  inter‐relationship  between  nephrin  and  podocin  (Koziell  et  al.,  2002).  On  the  other  hand,  Schultheiss  et  al.  did  not  find  any  evidence of di‐genic inheritance of NPHS1 and NPHS2 mutations as a tri‐allelic hit that  would result in the modification of phenotypes in their patients. Their data also did not  suggest  any  phenotype‐genotype  correlations  in  their  patients  with  mutations  in  both  NPHS1 and NPHS2 (Schultheiss et al., 2004).    4.4. Contribution of rare variants to nephrotic syndrome  We have shown a significant accumulation of rare variants in NS patients compared to  controls,  and  also  in  NS  patients  with  poor  prognosis  compared  to  patients  with good  prognosis. To our knowledge, this is the first study on the contribution of rare variants in  podocyte genes in NS.     122  Recently,  there  are  studies  reporting  the  accumulation  of  rare  variants  in  quantitative  traits  such  as  hypertriglyceridemia  and  hypercholesterolemia  ((Cohen  et  al.,  2004,  Johansen et al., 2010). Rare variant accumulation studies can implicate genes in disease  susceptibility  when  a  significant  burden  is  observed  in  patients  versus  controls.  Such  analyses  may  be  particularly  useful  for  candidate  genes  that  are  selected  based  on  experiments  other  than  genome‐wide  association  studies  (GWAS)  (Johansen  et  al.,  2012).     Genetic  variants  that  have  strong  phenotypic  effects  may  contribute  to  the  variations  observed  in  complex  traits.  However,  these  variants  are  more  likely  to  be  rare  individually. On the contrary, they may accumulate to cause variations in common traits  in the population (Cohen et al., 2004). By studying individuals with extreme phenotypes,  it  is  useful  to  identify  functional  rare  variants.  Using  accumulation  analysis  in  genes  defined as a priori as likely to contain rare variants, statistical analysis can be performed  to  quantify  the  burden  of  mutations  in  subjects  with  severe  phenotypes,  prior  to  the  functional assessment of each variant (Johansen et al., 2010). There has been increasing  evidence  that  NS  is  a  complex  trait  with  the  multi‐hit  hypothesis,  in  which  several  proteins involved in maintenance and function of the glomerular filtration barrier need  to be abnormal in order to give rise to a severe disease phenotype (Santin et al., 2009a).  Therefore, we hypothesized that rare DNA sequence variations in nephrin and podocin  collectively contribute to idiopathic NS in Singapore children.    123  In  our  cohort  of  patients  with  idiopathic  NS,  9  had  rare  variants.  A  total  of  8  variants  were  identified  amongst  these  patients,  of  which  5  variants  (NPHS1:c.494C>T,  c.1233C>T, c.2871G>A and c.3047G>A; NPHS2: c.685C>A) were found exclusively in our  population.     Five  of  the  variants  (NPHS1:  c.65C>T,  c.494C>T,  c.2398C>T  and  c.3047G>A;  NPHS2:  c.871C>T)  were  non‐synonymous,  resulting  in  amino  acid  substitution.  Of  these,  c.2398C>T  and  c.871C>T  were  predicted  to  be  probably  damaging  by  PolyPhen2.  c.2398C>T (p.R800C) (NPHS1) results in a change of amino acid from arginine to cysteine  at  the  position  800  of  nephrin.  Lahdenkari  et  al.  predicted  that  p.R800C  (NPHS1)  is  pathogenic  as  it  causes  the  formation  of  incorrect  disulfide  bonds  (Lahdenkari  et  al.,  2004). c.871C>T (NPHS2) results in the substitution of arginine to tryptophan at position  291  of  podocin.  p.R291W  (NPHS2)  had  reported  functional  effect.  The  substitution  of  arginine  with  bulky  aromatic  side  chain  tryptophan  caused  the  interruption  of  an  important  intra‐  or  intermolecular  interaction  within  or  between  mutant  podocin.   Podocin was found to be retained in the endoplasmic reticulum in the presence of this  mutation.  The  trafficking  of  nephrin  was  also  affected  as  nephrin  was  found  to  co‐ localize with the mutant proteins in the same intracellular compartment (Nishibori et al.,  2004).     There was a significantly higher carrier frequency of rare variants (both synonymous and  non‐synonymous)  in  NS  patients  compared  to  controls.  This  significant  higher  carrier  124  frequency  was  also  observed  even  when  analysis  was  restricted  to  non‐synonymous  variants.  This  showed  that  the  accumulation  of  rare  variants,  in  particular  non‐ synonymous variants, could contribute to the genetic aspects of NS in childhood.     Analysis  was  performed  to  see  if  there  was  any  association  of  accumulation  of  rare  variants  within  the  various  clinical  phenotypes.  In  childhood  NS,  there  are  various  factors which help in predicting the prognosis of the patients. Two major factors are the  response  to  drug  therapy,  both  steroids  and  cyclosporine,  and  also  the  progression  to  ESRD. Prednisone is the first line of drugs used in patients with NS. Calcineurin‐inhibitors  (CNI), in particular cyclosporine, are commonly used second‐line drugs for patients who  are  steroid  dependent  but  are  relapsing  frequently  and  also  for  steroid‐resistant  patients.  It  was  observed  that  20  to  30%  of  paediatrics  FSGS  are  responsive  to  cyclosporine  (Eddy  and  Symons,  2003).  Hence,  patients  who  were  resistant  to  both  prednisolone and CNI or progressed to ESRD are defined as having poor prognosis.     It  was  observed  that  the  carrier  frequency  of  rare  variants  (synonymous  and  non‐ synonymous)  was  significantly  higher  in  the  group  of  patients  defined  as  having  poor  prognosis.  This  significance  persisted  when  analysis  was  limited  to  only  non‐ synonymous  variants.  Therefore  the  accumulation  of  rare  variants,  especially  the non‐ synonymous ones, could contribute to the poor prognosis seen in our patients.    125  The pedigree of Patient 94 demonstrated the possibility of an interacting effect between  the  rare  variants  of  NPHS1  and  NPHS2.  Patient  94  was  the  only  member  in  his  family  that  had  4  rare  variants  and  he  developed  NS, while  the  other  members  in  the  family  pedigree had between 2 to 3 rare variants. A multi‐hit hypothesis involving interacting  genes  could  therefore  explain  why  only  the  index  patient  had  the  NS  phenotype.  Clinically significant disease could only occur with the unique allelic combination found  in the index patient, but not when the rare variants were expressed individually or in the  combination seen in the rest of the family members.        126  5. Conclusion  This  study  highlights  the  importance  of  genetic  screening  in  the  different  ethnic  populations  so  as  to  better  define  the  role  of  genetic  variants  in  the  pathogenesis  of  idiopathic nephrotic syndrome in these populations.  We have demonstrated that both  NPHS1  and  NPHS2  polymorphisms  contributed  to  the  pathogenesis  of  NS  in  both  our  Chinese  and  Malay  populations,  and  did  not  appear  to  be  responsible  for  the  poorer  prognosis observed in our Malay population. We would need to screen for other genes  that could be responsible for the poorer outcomes in Malays.     In  our  study,  NPHS2  polymorphisms  were  also  implicated  in  the  development  of  drug  resistance,  suggesting  that  screening  of  NPHS2  could  have  therapeutic  implications  in  children  with  INS.  We  have  also  shown  that  there  could  be  gene‐gene  interactions  between NPHS1 and NPHS2 genetic variants by performing composite genotype analysis.  Composite  genotypes  of  c.‐51G>T,  c.288C>T  (NPHS2)  and  c.2289C>T  (NPHS1)  could  be  associated with NS and also steroid resistance.     We  have  also  shown  that  accumulation  of  rare  variants  could  contribute  to  the  development  of  NS  in  Singapore  children  with  NS.  The  burden  of  rare  variants  was  particularly  observed  in  patients  with  poor  prognosis.  Functional  analysis  of  these  variants  may  accurately  define  the  effect  of  these  rare  variants  identified  in  patients  with NS and their role in causing the disease.     127  6. Reference  ADLER,  S.  1992.  Characterization  of  glomerular  epithelial  cell  matrix  receptors.  Am  J  Pathol, 141, 571‐8.  ARAYA, C. E., GONZALEZ‐PERALTA, R. P., SKODA‐SMITH, S. & DHARNIDHARKA, V. R. 2006.  Systemic  Epstein‐Barr  virus  infection  associated  with  membranous  nephropathy  in  children. Clin Nephrol. 2006/03/23 ed.  ASANUMA,  K.,  CAMPBELL,  K.  N.,  KIM,  K.,  FAUL,  C.  &  MUNDEL,  P.  2007.  Nuclear  relocation  of  the  nephrin  and  CD2AP‐binding  protein  dendrin  promotes  apoptosis  of  podocytes. Proc Natl Acad Sci U S A, 104, 10134‐9.  ASANUMA,  K.,  KIM,  K.,  OH,  J.,  GIARDINO,  L.,  CHABANIS,  S.,  FAUL,  C.,  REISER,  J.  &  MUNDEL, P. 2005. Synaptopodin regulates the actin‐bundling activity of alpha‐actinin in  an isoform‐specific manner. J Clin Invest, 115, 1188‐98.  AYA,  K.,  SHIMIZU,  J.,  OHTOMO,  Y.,  SATOMURA,  K.,  SUZUKI,  H.,  YAN,  K.,  SADO,  Y.,  MORISHIMA,  T.  &  TANAKA,  H.  2009.  NPHS1  gene  mutation  in  Japanese  patients  with  congenital nephrotic syndrome. Nephrol Dial Transplant, 24, 2411‐4.  AYA, K., TANAKA, H. & SEINO, Y. 2000. Novel mutation in the nephrin gene of a Japanese  patient with congenital nephrotic syndrome of the Finnish type. Kidney Int, 57, 401‐4.  BALLERMANN,  B.  J.  2005.  Glomerular  endothelial  cell  differentiation.  Kidney  Int,  67,  1668‐71.  BARKER, D. F., HOSTIKKA, S. L., ZHOU, J., CHOW, L. T., OLIPHANT, A. R., GERKEN, S. C.,  GREGORY, M. C., SKOLNICK, M. H., ATKIN, C. L. & TRYGGVASON, K. 1990. Identification  of mutations in the COL4A5 collagen gene in Alport syndrome. Science, 248, 1224‐7.  BELTCHEVA, O., MARTIN, P., LENKKERI, U. & TRYGGVASON, K. 2001. Mutation spectrum  in the nephrin gene (NPHS1) in congenital nephrotic syndrome. Hum Mutat, 17, 368‐73.  BERTELLI, R., GINEVRI, F., CARIDI, G., DAGNINO, M., SANDRINI, S., DI DUCA, M., EMMA,  F.,  SANNA‐CHERCHI,  S., SCOLARI,  F.,  NERI,  T.  M.,  MURER,  L.,  MASSELLA,  L.,  BASILE,  G.,  RIZZONI,  G.,  PERFUMO,  F.  &  GHIGGERI,  G.  M.  2003.  Recurrence  of  focal  segmental  glomerulosclerosis after renal transplantation in patients with mutations of podocin. Am  J Kidney Dis, 41, 1314‐21.  128  BILLING,  H.,  MULLER,  D.,  RUF,  R.,  LICHTENBERGER,  A.,  HILDEBRANDT,  F.,  AUGUST,  C.,  QUERFELD,  U.  &  HAFFNER,  D.  2004.  NPHS2  mutation  associated  with  recurrence  of  proteinuria after transplantation. Pediatr Nephrol, 19, 561‐4.  BLATTNER,  S.  M.  &  KRETZLER,  M.  2005.  Integrin‐linked  kinase  in  renal  disease:  connecting cell‐matrix interaction to the cytoskeleton. Curr Opin Nephrol Hypertens, 14,  404‐10.  BLOWEY, D. L. 1996. Nephrotic syndrome associated with an Epstein‐Barr virus infection.  Pediatr Nephrol. 1996/08/01 ed.  BOUTE, N., GRIBOUVAL, O., ROSELLI, S., BENESSY, F., LEE, H., FUCHSHUBER, A., DAHAN,  K., GUBLER, M. C., NIAUDET, P. & ANTIGNAC, C. 2000. NPHS2, encoding the glomerular  protein  podocin,  is  mutated  in  autosomal  recessive  steroid‐resistant  nephrotic  syndrome. Nat Genet, 24, 349‐54.  CARIDI, G., BERTELLI, R., CARREA, A., DI DUCA, M., CATARSI, P., ARTERO, M., CARRARO,  M.,  ZENNARO,  C.,  CANDIANO,  G.,  MUSANTE,  L.,  SERI,  M.,  GINEVRI,  F.,  PERFUMO,  F.  &  GHIGGERI,  G.  M.  2001.  Prevalence,  genetics,  and  clinical  features  of  patients  carrying  podocin mutations in steroid‐resistant nonfamilial focal segmental glomerulosclerosis. J  Am Soc Nephrol, 12, 2742‐6.  CARIDI,  G.,  BERTELLI,  R.,  DI  DUCA,  M.,  DAGNINO,  M.,  EMMA,  F.,  ONETTI  MUDA,  A.,  SCOLARI,  F.,  MIGLIETTI,  N.,  MAZZUCCO,  G.,  MURER,  L.,  CARREA,  A.,  MASSELLA,  L.,  RIZZONI,  G.,  PERFUMO,  F.  &  GHIGGERI,  G.  M.  2003.  Broadening  the  spectrum  of  diseases related to podocin mutations. J Am Soc Nephrol, 14, 1278‐86.  CARIDI, G., GIGANTE, M., RAVANI, P., TRIVELLI, A., BARBANO, G., SCOLARI, F., DAGNINO,  M., MURER, L., MURTAS, C., EDEFONTI, A., ALLEGRI, L., AMORE, A., COPPO, R., EMMA, F.,  DE  PALO,  T.,  PENZA,  R.,  GESUALDO,  L.  &  GHIGGERI,  G.  M.  2009.  Clinical  features  and  long‐term  outcome  of  nephrotic  syndrome  associated  with  heterozygous  NPHS1  and  NPHS2 mutations. Clin J Am Soc Nephrol, 4, 1065‐72.  CARIDI,  G.,  PERFUMO,  F.  &  GHIGGERI,  G.  M.  2005.  NPHS2  (Podocin)  mutations  in  nephrotic syndrome. Clinical spectrum and fine mechanisms. Pediatr Res, 57, 54R‐61R.  129  CARTEGNI, L., WANG, J., ZHU, Z., ZHANG, M. Q. & KRAINER, A. R. 2003. ESEfinder: A web  resource to identify exonic splicing enhancers. Nucleic Acids Res, 31, 3568‐71.  CHO, H. Y., LEE, J. H., CHOI, H. J., LEE, B. H., HA, I. S., CHOI, Y. & CHEONG, H. I. 2008. WT1  and  NPHS2  mutations  in  Korean  children  with  steroid‐resistant  nephrotic  syndrome.  Pediatr Nephrol, 23, 63‐70.  COHEN, J. C., KISS, R. S., PERTSEMLIDIS, A., MARCEL, Y. L., MCPHERSON, R. & HOBBS, H.  H. 2004. Multiple rare alleles contribute to low plasma levels of HDL cholesterol. Science,  305, 869‐72.  DAI, S., WANG, Z., PAN, X., WANG, W., CHEN, X., REN, H., HAO, C., HAN, B. & CHEN, N.  2010. Functional analysis of promoter mutations in the ACTN4 and SYNPO genes in focal  segmental glomerulosclerosis. Nephrol Dial Transplant, 25, 824‐35.  DANDAPANI, S. V., SUGIMOTO, H., MATTHEWS, B. D., KOLB, R. J., SINHA, S., GERSZTEN,  R.  E.,  ZHOU,  J.,  INGBER,  D.  E.,  KALLURI,  R.  &  POLLAK,  M.  R.  2007.  Alpha‐actinin‐4  is  required for normal podocyte adhesion. J Biol Chem, 282, 467‐77.  DI DUCA, M., OLEGGINI, R., SANNA‐CHERCHI, S., PASQUALI, L., DI DONATO, A., PARODI,  S.,  BERTELLI,  R.,  CARIDI,  G.,  FRASCA,  G.,  CERULLO,  G.,  AMOROSO,  A.,  SCHENA,  F.  P.,  SCOLARI,  F.  &  GHIGGERI,  G.  M.  2006.  Cis  and  trans  regulatory  elements  in  NPHS2  promoter: implications in proteinuria and progression of renal diseases. Kidney Int, 70,  1332‐41.  EDDY, A. A. & SYMONS, J. M. 2003. Nephrotic syndrome in childhood. Lancet, 362, 629‐ 39.  FAIRBROTHER, W. G., YEO, G. W., YEH, R., GOLDSTEIN, P., MAWSON, M., SHARP, P. A. &  BURGE,  C.  B.  2004.  RESCUE‐ESE  identifies  candidate  exonic  splicing  enhancers  in  vertebrate exons. Nucleic Acids Res, 32, W187‐90.  FRISHBERG,  Y.,  FEINSTEIN,  S.,  RINAT,  C.,  BECKER‐COHEN,  R.,  LERER,  I.,  RAAS‐ ROTHSCHILD, A., FERBER, B. & NIR, A. 2006. The heart of children with steroid‐resistant  nephrotic syndrome: is it all podocin? J Am Soc Nephrol, 17, 227‐31.  130  GBADEGESIN,  R.,  HINKES,  B.,  VLANGOS,  C.,  MUCHA,  B.,  LIU,  J.,  HOPCIAN,  J.  &  HILDEBRANDT, F. 2007. Mutational analysis of NPHS2 and WT1 in frequently relapsing  and steroid‐dependent nephrotic syndrome. Pediatr Nephrol, 22, 509‐13.  GBADEGESIN,  R.  &  SMOYER,  W.  E.  2008.  Nephrotic  Syndrome.  In:  GERAY,  D.  F.  &  SCHAEFER, F. (eds.) Comprehensive Pediatrics Nephrology. 1st ed.: Mosby Elsevier.  GIGANTE,  M.,  MONNO,  F.,  ROBERTO,  R.,  LAFORGIA,  N.,  ASSAEL,  M.  B.,  LIVOLTI,  S.,  CARINGELLA,  A.,  LA  MANNA,  A.,  MASELLA,  L.  &  IOLASCON,  A.  2002.  Congenital  nephrotic  syndrome  of  the  Finnish  type  in  Italy:  a  molecular  approach.  J  Nephrol,  15,  696‐702.  GIGANTE,  M.,  PONTRELLI,  P.,  MONTEMURNO,  E.,  ROCA,  L.,  AUCELLA,  F.,  PENZA,  R.,  CARIDI,  G.,  RANIERI,  E.,  GHIGGERI,  G.  M.  &  GESUALDO,  L.  2009.  CD2AP  mutations  are  associated  with  sporadic  nephrotic  syndrome  and  focal  segmental  glomerulosclerosis  (FSGS). Nephrol Dial Transplant, 24, 1858‐64.  GIPSON, D. S., MASSENGILL, S. F., YAO, L., NAGARAJ, S., SMOYER, W. E., MAHAN, J. D.,  WIGFALL,  D.,  MILES,  P.,  POWELL,  L.,  LIN,  J.  J.,  TRACHTMAN,  H.  &  GREENBAUM,  L.  A.  2009. Management of childhood onset nephrotic syndrome. Pediatrics, 124, 747‐57.  GULATI, S., PRASAD, N., SHARMA, R. K., KUMAR, A., GUPTA, A. & BABURAJ, V. P. 2008.  Tacrolimus: a new therapy for steroid‐resistant nephrotic syndrome in children. Nephrol  Dial Transplant, 23, 910‐3.  HARA,  M.,  YANAGIHARA,  T.  &  KIHARA,  I.  2001.  Urinary  podocytes  in  primary  focal  segmental glomerulosclerosis. Nephron, 89, 342‐7.  HEERINGA,  S.  F.,  VLANGOS,  C.  N.,  CHERNIN,  G.,  HINKES,  B.,  GBADEGESIN,  R.,  LIU,  J.,  HOSKINS, B. E., OZALTIN, F. & HILDEBRANDT, F. 2008. Thirteen novel NPHS1 mutations  in  a  large  cohort  of  children  with  congenital  nephrotic  syndrome.  Nephrol  Dial  Transplant, 23, 3527‐33.  HINKES, B., WIGGINS, R. C., GBADEGESIN, R., VLANGOS, C. N., SEELOW, D., NURNBERG,  G., GARG, P., VERMA, R., CHAIB, H., HOSKINS, B. E., ASHRAF, S., BECKER, C., HENNIES, H.  C., GOYAL, M., WHARRAM, B. L., SCHACHTER, A. D., MUDUMANA, S., DRUMMOND, I.,  KERJASCHKI, D., WALDHERR, R., DIETRICH, A., OZALTIN, F., BAKKALOGLU, A., CLEPER, R.,  131  BASEL‐VANAGAITE,  L.,  POHL,  M.,  GRIEBEL,  M.,  TSYGIN,  A.  N.,  SOYLU,  A.,  MULLER,  D.,  SORLI,  C.  S.,  BUNNEY,  T.  D.,  KATAN,  M.,  LIU,  J.,  ATTANASIO,  M.,  O'TOOLE  J,  F.,  HASSELBACHER,  K.,  MUCHA,  B.,  OTTO,  E.  A.,  AIRIK,  R.,  KISPERT,  A.,  KELLEY,  G.  G.,  SMRCKA,  A.  V.,  GUDERMANN,  T.,  HOLZMAN,  L.  B.,  NURNBERG,  P.  &  HILDEBRANDT,  F.  2006.  Positional  cloning  uncovers  mutations  in  PLCE1  responsible  for  a  nephrotic  syndrome variant that may be reversible. Nat Genet, 38, 1397‐405.  HINKES, B. G., MUCHA, B., VLANGOS, C. N., GBADEGESIN, R., LIU, J., HASSELBACHER, K.,  HANGAN, D., OZALTIN, F., ZENKER, M. & HILDEBRANDT, F. 2007. Nephrotic syndrome in  the  first  year  of  life:  two  thirds  of  cases  are  caused  by  mutations  in  4  genes  (NPHS1,  NPHS2, WT1, and LAMB2). Pediatrics, 119, e907‐19.  HINO,  S.,  TAKEMURA,  T.,  OKADA,  M.,  MURAKAMI,  K.,  YAGI,  K.,  FUKUSHIMA,  K.  &  YOSHIOKA, K. 1998. Follow‐up study of children with nephrotic syndrome treated with a  long‐term moderate dose of cyclosporine. Am J Kidney Dis, 31, 932‐9.  HISATSUNE, C., KURODA, Y., NAKAMURA, K., INOUE, T., NAKAMURA, T., MICHIKAWA, T.,  MIZUTANI, A. & MIKOSHIBA, K. 2004. Regulation of TRPC6 channel activity by tyrosine  phosphorylation. J Biol Chem, 279, 18887‐94.  HODSON, E. M., KNIGHT, J. F., WILLIS, N. S. & CRAIG, J. C. 2004. Corticosteroid therapy  for nephrotic syndrome in children. Cochrane Database Syst Rev, CD001533.  HOLTHOFER,  H.,  AHOLA,  H.,  SOLIN,  M.  L.,  WANG,  S.,  PALMEN,  T.,  LUIMULA,  P.,  MIETTINEN,  A.  &  KERJASCHKI,  D.  1999.  Nephrin  localizes  at  the  podocyte  filtration  slit  area and is characteristically spliced in the human kidney. Am J Pathol, 155, 1681‐7.  HORINOUCHI,  I.,  NAKAZATO,  H.,  KAWANO,  T.,  IYAMA,  K.,  FURUSE,  A.,  ARIZONO,  K.,  MACHIDA, J., SAKAMOTO, T., ENDO, F. & HATTORI, S. 2003. In situ evaluation of podocin  in normal and glomerular diseases. Kidney Int, 64, 2092‐9.  HUBER, T. B., HARTLEBEN, B., KIM, J., SCHMIDTS, M., SCHERMER, B., KEIL, A., EGGER, L.,  LECHA, R. L., BORNER, C., PAVENSTADT, H., SHAW, A. S., WALZ, G. & BENZING, T. 2003a.  Nephrin  and  CD2AP  associate  with  phosphoinositide  3‐OH  kinase  and  stimulate  AKT‐ dependent signaling. Mol Cell Biol, 23, 4917‐28.  132  HUBER,  T.  B.,  KOTTGEN,  M.,  SCHILLING,  B.,  WALZ,  G.  &  BENZING,  T.  2001.  Interaction  with podocin facilitates nephrin signaling. J Biol Chem, 276, 41543‐6.  HUBER, T. B., SIMONS, M., HARTLEBEN, B., SERNETZ, L., SCHMIDTS, M., GUNDLACH, E.,  SALEEM,  M.  A.,  WALZ,  G.  &  BENZING,  T.  2003b.  Molecular  basis  of  the  functional  podocin‐nephrin  complex:  mutations  in  the  NPHS2  gene  disrupt  nephrin  targeting  to  lipid raft microdomains. Hum Mol Genet, 12, 3397‐405.  ISKDC 1981. The primary nephrotic syndrome in children. Identification of patients with  minimal change nephrotic syndrome from initial response to prednisone. A report of the  International Study of Kidney Disease in Children. J Pediatr, 98, 561‐4.  ISKDC  1982.  Early  identification  of  frequent  relapsers  among  children  with  minimal  change  nephrotic  syndrome.  A  report  of  the  International  Study  of  Kidney  Disease  in  Children. J Pediatr, 101, 514‐8.  JARAD,  G.,  CUNNINGHAM,  J.,  SHAW,  A.  S.  &  MINER,  J.  H.  2006.  Proteinuria  precedes  podocyte  abnormalities  inLamb2‐/‐  mice,  implicating  the  glomerular  basement  membrane as an albumin barrier. J Clin Invest, 116, 2272‐9.  JEANSSON, M. & HARALDSSON, B. 2006. Morphological and functional evidence for an  important role of the endothelial cell glycocalyx in the glomerular barrier. Am J Physiol  Renal Physiol, 290, F111‐6.  JOHANSEN,  C.  T.,  WANG,  J.,  LANKTREE,  M.  B.,  CAO,  H.,  MCINTYRE,  A.  D.,  BAN,  M.  R.,  MARTINS,  R.  A.,  KENNEDY,  B.  A.,  HASSELL,  R.  G.,  VISSER,  M.  E.,  SCHWARTZ,  S.  M.,  VOIGHT,  B.  F.,  ELOSUA,  R.,  SALOMAA,  V.,  O'DONNELL,  C.  J.,  DALLINGA‐THIE,  G.  M.,  ANAND, S. S., YUSUF, S., HUFF, M. W., KATHIRESAN, S. & HEGELE, R. A. 2010. Excess of  rare  variants  in  genes  identified  by  genome‐wide  association  study  of  hypertriglyceridemia. Nat Genet, 42, 684‐7.  JOHANSEN, C. T., WANG, J., MCINTYRE, A. D., MARTINS, R. A., BAN, M. R., LANKTREE, M.  B., HUFF, M. W., PETERFY, M., MEHRABIAN, M., LUSIS, A. J., KATHIRESAN, S., ANAND, S.  S., YUSUF, S., LEE, A. H., GLIMCHER, L. H., CAO, H. & HEGELE, R. A. 2012. Excess of rare  variants  in  non‐genome‐wide  association  study  candidate  genes  in  patients  with  hypertriglyceridemia. Circ Cardiovasc Genet, 5, 66‐72.  133  JUNGRAITHMAYR, T. C., HOFER, K., COCHAT, P., CHERNIN, G., CORTINA, G., FARGUE, S.,  GRIMM,  P.,  KNUEPPEL,  T.,  KOWARSCH,  A.,  NEUHAUS,  T.,  PAGEL,  P.,  PFEIFFER,  K.  P.,  SCHAFER, F., SCHONERMARCK, U., SEEMAN, T., TOENSHOFF, B., WEBER, S., WINN, M. P.,  ZSCHOCKE,  J.  &  ZIMMERHACKL,  L.  B.  2011.  Screening  for  NPHS2  mutations  may  help  predict FSGS recurrence after transplantation. J Am Soc Nephrol, 22, 579‐85.  KAPLAN,  J.  M.,  KIM,  S.  H.,  NORTH,  K.  N.,  RENNKE,  H.,  CORREIA,  L.  A.,  TONG,  H.  Q.,  MATHIS, B. J., RODRIGUEZ‐PEREZ, J. C., ALLEN, P. G., BEGGS, A. H. & POLLAK, M. R. 2000.  Mutations  in  ACTN4,  encoding  alpha‐actinin‐4,  cause  familial  focal  segmental  glomerulosclerosis. Nat Genet, 24, 251‐6.  KARLE, S. M., UETZ, B., RONNER, V., GLAESER, L., HILDEBRANDT, F. & FUCHSHUBER, A.  2002. Novel mutations in NPHS2 detected in both familial and sporadic steroid‐resistant  nephrotic syndrome. J Am Soc Nephrol, 13, 388‐93.  KESTILA, M., LENKKERI, U., MANNIKKO, M., LAMERDIN, J., MCCREADY, P., PUTAALA, H.,  RUOTSALAINEN, V., MORITA, T., NISSINEN, M., HERVA, R., KASHTAN, C. E., PELTONEN, L.,  HOLMBERG, C., OLSEN, A. & TRYGGVASON, K. 1998. Positionally cloned gene for a novel  glomerular protein‐‐nephrin‐‐is mutated in congenital nephrotic syndrome. Mol Cell, 1,  575‐82.  KIM, J. M., WU, H., GREEN, G., WINKLER, C. A., KOPP, J. B., MINER, J. H., UNANUE, E. R. &  SHAW,  A.  S.  2003.  CD2‐associated  protein  haploinsufficiency  is  linked  to  glomerular  disease susceptibility. Science, 300, 1298‐300.  KITAMURA, A., TSUKAGUCHI, H., IIJIMA, K., ARAKI, J., HATTORI, M., IKEDA, M., HONDA,  M., NOZU, K., NAKAZATO, H., YOSHIKAWA, N., KAGAMI, S., MURAMATSU, M., CHOI, Y.,  CHEONG,  H.  I.  &  DOI,  T.  2006.  Genetics  and  clinical  features  of  15  Asian  families  with  steroid‐resistant nephrotic syndrome. Nephrol Dial Transplant, 21, 3133‐8.  KLAMT, B., KOZIELL, A., POULAT, F., WIEACKER, P., SCAMBLER, P., BERTA, P. & GESSLER,  M. 1998. Frasier syndrome is caused by defective alternative splicing of WT1 leading to  an altered ratio of WT1 +/‐KTS splice isoforms. Hum Mol Genet, 7, 709‐14.  KOPP, J. B., SMITH, M. W., NELSON, G. W., JOHNSON, R. C., FREEDMAN, B. I., BOWDEN,  D. W., OLEKSYK, T., MCKENZIE, L. M., KAJIYAMA, H., AHUJA, T. S., BERNS, J. S., BRIGGS,  134  W., CHO, M. E., DART, R. A., KIMMEL, P. L., KORBET, S. M., MICHEL, D. M., MOKRZYCKI,  M. H., SCHELLING, J. R., SIMON, E., TRACHTMAN, H., VLAHOV, D. & WINKLER, C. A. 2008.  MYH9 is a major‐effect risk gene for focal segmental glomerulosclerosis. Nat Genet, 40,  1175‐84.  KOS, C. H., LE, T. C., SINHA, S., HENDERSON, J. M., KIM, S. H., SUGIMOTO, H., KALLURI, R.,  GERSZTEN,  R.  E.  &  POLLAK,  M.  R.  2003.  Mice  deficient  in  alpha‐actinin‐4  have  severe  glomerular disease. J Clin Invest, 111, 1683‐90.  KOSKIMIES,  O.,  VILSKA,  J.,  RAPOLA,  J.  &  HALLMAN,  N.  1982.  Long‐term  outcome  of  primary nephrotic syndrome. Arch Dis Child, 57, 544‐8.  KOZIELL,  A.,  GRECH,  V.,  HUSSAIN,  S.,  LEE,  G.,  LENKKERI,  U.,  TRYGGVASON,  K.  &  SCAMBLER, P. 2002. Genotype/phenotype correlations of NPHS1 and NPHS2 mutations  in nephrotic syndrome advocate a functional inter‐relationship in glomerular filtration.  Hum Mol Genet, 11, 379‐88.  KREIDBERG, J. A., DONOVAN, M. J., GOLDSTEIN, S. L., RENNKE, H., SHEPHERD, K., JONES,  R. C. & JAENISCH, R. 1996. Alpha 3 beta 1 integrin has a crucial role in kidney and lung  organogenesis. Development, 122, 3537‐47.  KREIDBERG, J. A., SARIOLA, H., LORING, J. M., MAEDA, M., PELLETIER, J., HOUSMAN, D.  & JAENISCH, R. 1993. WT‐1 is required for early kidney development. Cell, 74, 679‐91.  KRIZ,  W.  2005.  TRPC6  ‐  a  new  podocyte  gene  involved  in  focal  segmental  glomerulosclerosis. Trends Mol Med, 11, 527‐30.  KUMAR,  J.,  GULATI,  S.,  SHARMA,  A.  P.,  SHARMA,  R.  K.  &  GUPTA,  R.  K.  2003.  Histopathological spectrum of childhood nephrotic syndrome in Indian children. Pediatr  Nephrol, 18, 657‐60.  KYRIELEIS, H. A., LOWIK, M. M., PRONK, I., CRUYSBERG, H. R., KREMER, J. A., OYEN, W. J.,  VAN DEN HEUVEL, B. L., WETZELS, J. F. & LEVTCHENKO, E. N. 2009. Long‐term outcome  of biopsy‐proven, frequently relapsing minimal‐change nephrotic syndrome in children.  Clin J Am Soc Nephrol, 4, 1593‐600.  135  LAHDENKARI,  A.  T.,  KESTILA,  M.,  HOLMBERG,  C.,  KOSKIMIES,  O.  &  JALANKO,  H.  2004.  Nephrin  gene  (NPHS1)  in  patients  with  minimal  change  nephrotic  syndrome  (MCNS).  Kidney Int, 65, 1856‐63.  LAHDENKARI, A. T., SUVANTO, M., KAJANTIE, E., KOSKIMIES, O., KESTILA, M. & JALANKO,  H.  2005.  Clinical  features  and  outcome  of  childhood  minimal  change  nephrotic  syndrome: is genetics involved? Pediatr Nephrol, 20, 1073‐80.  LAHDENPERA,  J.,  KILPELAINEN,  P.,  LIU,  X.  L.,  PIKKARAINEN,  T.,  REPONEN,  P.,  RUOTSALAINEN,  V.  &  TRYGGVASON,  K.  2003.  Clustering‐induced  tyrosine  phosphorylation of nephrin by Src family kinases. Kidney Int, 64, 404‐13.  LAI, K. W., WEI, C. L., TAN, L. K., TAN, P. H., CHIANG, G. S., LEE, C. G., JORDAN, S. C. &  YAP,  H.  K.  2007.  Overexpression  of  interleukin‐13  induces  minimal‐change‐like  nephropathy in rats. J Am Soc Nephrol, 18, 1476‐85.  LEE, B. H., AHN, Y. H., CHOI, H. J., KANG, H. K., KIM, S. D., CHO, B. S., MOON, K. C., HA, I.  S.,  CHEONG,  H.  I.  &  CHOI,  Y.  2009.  Two  Korean  infants  with  genetically  confirmed  congenital nephrotic syndrome of Finnish type. J Korean Med Sci, 24 Suppl, S210‐4.  LENKKERI, U., MANNIKKO, M., MCCREADY, P., LAMERDIN, J., GRIBOUVAL, O., NIAUDET,  P.  M.,  ANTIGNAC,  C.  K.,  KASHTAN,  C.  E.,  HOMBERG,  C.,  OLSEN,  A.,  KESTILA,  M.  &  TRYGGVASON, K. 1999. Structure of the gene for congenital nephrotic syndrome of the  finnish type (NPHS1) and characterization of mutations. Am J Hum Genet, 64, 51‐61.  LEVIDIOTIS,  V.  &  POWER,  D.  A.  2005.  New  insights  into  the  molecular  biology  of  the  glomerular filtration barrier and associated disease. Nephrology (Carlton), 10, 157‐66.  LI, C., RUOTSALAINEN, V., TRYGGVASON, K., SHAW, A. S. & MINER, J. H. 2000. CD2AP is  expressed with nephrin in developing podocytes and is found widely in mature kidney  and elsewhere. Am J Physiol Renal Physiol, 279, F785‐92.  LI, H., LEMAY, S., AOUDJIT, L., KAWACHI, H. & TAKANO, T. 2004. SRC‐family kinase Fyn  phosphorylates  the  cytoplasmic  domain  of  nephrin  and  modulates  its  interaction  with  podocin. J Am Soc Nephrol, 15, 3006‐15.  136  LI, X., LI, H., YE, H., LI, Q., HE, X., ZHANG, X., CHEN, Y., HAN, F., HE, Q., WANG, H. & CHEN,  J.  2009.  Tacrolimus  therapy  in  adults  with  steroid‐  and  cyclophosphamide‐resistant  nephrotic syndrome and normal or mildly reduced GFR. Am J Kidney Dis, 54, 51‐8.  LOWIK, M., LEVTCHENKO, E., WESTRA, D., GROENEN, P., STEENBERGEN, E., WEENING, J.,  LILIEN, M., MONNENS, L. & VAN DEN HEUVEL, L. 2008. Bigenic heterozygosity and the  development  of  steroid‐resistant  focal  segmental  glomerulosclerosis.  Nephrol  Dial  Transplant, 23, 3146‐51.  LUO,  X.,  KRANZLER,  H.  R.,  ZUO,  L.,  WANG,  S.,  SCHORK,  N.  J.  &  GELERNTER,  J.  2006.  Diplotype  trend  regression  analysis  of  the  ADH  gene  cluster  and  the  ALDH2  gene:  multiple significant associations with alcohol dependence. Am J Hum Genet, 78, 973‐87.  MAO, J., ZHANG, Y., DU, L., DAI, Y., GU, W., LIU, A., SHANG, S. & LIANG, L. 2007. NPHS1  and  NPHS2  gene  mutations  in  Chinese  children  with  sporadic  nephrotic  syndrome.  Pediatr Res, 61, 117‐22.  MAO,  J.,  ZHANG,  Y.,  DU,  L.,  DAI,  Y.,  YANG,  C.  &  LIANG,  L.  2006.  Expression  profile  of  nephrin,  podocin,  and  CD2AP  in  Chinese  children  with  MCNS  and  IgA  nephropathy.  Pediatr Nephrol, 21, 1666‐75.  MARUYAMA,  K.,  IIJIMA,  K.,  IKEDA,  M.,  KITAMURA,  A.,  TSUKAGUCHI,  H.,  YOSHIYA,  K.,  HOSHII, S., WADA, N., UEMURA, O., SATOMURA, K., HONDA, M. & YOSHIKAWA, N. 2003.  NPHS2 mutations in sporadic steroid‐resistant nephrotic syndrome in Japanese children.  Pediatr Nephrol, 18, 412‐6.  MCCARTHY, E. T., SHARMA, M. & SAVIN, V. J.  2010. Circulating permeability factors in  idiopathic  nephrotic  syndrome  and  focal  segmental  glomerulosclerosis.  Clin  J  Am  Soc  Nephrol, 5, 2115‐21.  MCKENZIE,  L.  M.,  HENDRICKSON,  S.  L.,  BRIGGS,  W.  A.,  DART,  R.  A.,  KORBET,  S.  M.,  MOKRZYCKI, M. H., KIMMEL, P. L., AHUJA, T. S., BERNS, J. S., SIMON, E. E., SMITH, M. C.,  TRACHTMAN, H., MICHEL, D. M., SCHELLING, J. R., CHO, M., ZHOU, Y. C., BINNS‐ROEMER,  E.,  KIRK,  G.  D.,  KOPP,  J.  B.  &  WINKLER,  C.  A.  2007.  NPHS2  variation  in  sporadic  focal  segmental glomerulosclerosis. J Am Soc Nephrol, 18, 2987‐95.  137  MINER,  J.  H.,  GO,  G.,  CUNNINGHAM,  J.,  PATTON,  B.  L.  &  JARAD,  G.  2006.  Transgenic  isolation  of  skeletal  muscle  and  kidney  defects  in  laminin  beta2  mutant  mice:  implications for Pierson syndrome. Development, 133, 967‐75.  MOLLER, C. C., FLESCHE, J. & REISER, J. 2009. Sensitizing the Slit Diaphragm with TRPC6  ion channels. J Am Soc Nephrol, 20, 950‐3.  MOLLET, G., RATELADE, J., BOYER, O., MUDA, A. O., MORISSET, L., LAVIN, T. A., KITZIS, D.,  DALLMAN, M. J., BUGEON, L., HUBNER, N., GUBLER, M. C., ANTIGNAC, C. & ESQUIVEL, E.  L.  2009.  Podocin  inactivation  in  mature  kidneys  causes  focal  segmental  glomerulosclerosis and nephrotic syndrome. J Am Soc Nephrol, 20, 2181‐9.  MUNDEL,  P.  &  KRIZ,  W.  1995.  Structure  and  function  of  podocytes:  an  update.  Anat  Embryol (Berl), 192, 385‐97.  NATOLI,  T.  A.,  LIU,  J.,  EREMINA,  V.,  HODGENS,  K.,  LI,  C.,  HAMANO,  Y.,  MUNDEL,  P.,  KALLURI, R., MINER, J. H., QUAGGIN, S. E. & KREIDBERG, J. A. 2002. A mutant form of the  Wilms' tumor suppressor gene WT1 observed in Denys‐Drash syndrome interferes with  glomerular capillary development. J Am Soc Nephrol, 13, 2058‐67.  NG, P. C. & HENIKOFF, S. 2003. SIFT: Predicting amino acid changes that affect protein  function. Nucleic Acids Res, 31, 3812‐4.  NG,  S.  B.,  TURNER,  E.  H.,  ROBERTSON,  P.  D.,  FLYGARE,  S.  D.,  BIGHAM,  A.  W.,  LEE,  C.,  SHAFFER, T., WONG, M., BHATTACHARJEE, A., EICHLER, E. E., BAMSHAD, M., NICKERSON,  D.  A.  &  SHENDURE,  J.  2009.  Targeted  capture and  massively  parallel  sequencing  of  12  human exomes. Nature, 461, 272‐6.  NISHIBORI,  Y.,  LIU,  L.,  HOSOYAMADA,  M.,  ENDOU,  H.,  KUDO,  A.,  TAKENAKA,  H.,  HIGASHIHARA,  E.,  BESSHO,  F.,  TAKAHASHI,  S.,  KERSHAW,  D.,  RUOTSALAINEN,  V.,  TRYGGVASON, K., KHOSHNOODI, J. & YAN, K. 2004. Disease‐causing missense mutations  in NPHS2 gene alter normal nephrin trafficking to the plasma membrane. Kidney Int, 66,  1755‐65.  OH,  J.,  REISER,  J.  &  MUNDEL,  P.  2004.  Dynamic  (re)organization  of  the  podocyte  actin  cytoskeleton in the nephrotic syndrome. Pediatr Nephrol, 19, 130‐7.  138  PATRAKKA, J., KESTILA, M., WARTIOVAARA, J., RUOTSALAINEN, V., TISSARI, P., LENKKERI,  U.,  MANNIKKO,  M.,  VISAPAA,  I.,  HOLMBERG,  C.,  RAPOLA,  J.,  TRYGGVASON,  K.  &  JALANKO,  H.  2000.  Congenital  nephrotic  syndrome  (NPHS1):  features  resulting  from  different mutations in Finnish patients. Kidney Int, 58, 972‐80.  PATRAKKA,  J.  &  TRYGGVASON,  K.  2007.  Nephrin‐‐a  unique  structural  and  signaling  protein of the kidney filter. Trends Mol Med, 13, 396‐403.  PECES, R., SANCHEZ, L., GOROSTIDI, M. & ALVAREZ, J. 1991. Minimal change nephrotic  syndrome  associated  with  Hodgkin's  lymphoma.  Nephrol  Dial  Transplant.  1991/01/01  ed.  PHILIPPE,  A.,  NEVO,  F.,  ESQUIVEL,  E.  L.,  REKLAITYTE,  D.,  GRIBOUVAL,  O.,  TETE,  M.  J.,  LOIRAT, C., DANTAL, J., FISCHBACH, M., POUTEIL‐NOBLE, C., DECRAMER, S., HOEHNE, M.,  BENZING,  T.,  CHARBIT,  M.,  NIAUDET,  P.  &  ANTIGNAC,  C.  2008.  Nephrin  mutations  can  cause  childhood‐onset  steroid‐resistant  nephrotic  syndrome.  J  Am  Soc  Nephrol,  19,  1871‐8.  PUTAALA,  H.,  SOININEN,  R.,  KILPELAINEN,  P.,  WARTIOVAARA,  J.  &  TRYGGVASON,  K.  2001. The murine nephrin gene is specifically expressed in kidney, brain and pancreas:  inactivation  of  the  gene  leads  to  massive  proteinuria  and  neonatal  death.  Hum  Mol  Genet, 10, 1‐8.  RAATS,  C.  J.,  VAN  DEN  BORN,  J.,  BAKKER,  M.  A.,  OPPERS‐WALGREEN,  B.,  PISA,  B.  J.,  DIJKMAN, H. B., ASSMANN, K. J. & BERDEN, J. H. 2000. Expression of agrin, dystroglycan,  and utrophin in normal renal tissue and in experimental glomerulopathies. Am J Pathol,  156, 1749‐65.  REGELE, H. M., FILLIPOVIC, E., LANGER, B., POCZEWKI, H., KRAXBERGER, I., BITTNER, R. E.  &  KERJASCHKI,  D.  2000.  Glomerular  expression  of  dystroglycans  is  reduced  in  minimal  change nephrosis but not in focal segmental glomerulosclerosis. J Am Soc Nephrol, 11,  403‐12.  REISER, J., KRIZ, W., KRETZLER, M. & MUNDEL, P. 2000. The glomerular slit diaphragm is  a modified adherens junction. J Am Soc Nephrol, 11, 1‐8.  139  REISER,  J.,  OH,  J.,  SHIRATO,  I.,  ASANUMA,  K.,  HUG,  A.,  MUNDEL,  T.  M.,  HONEY,  K.,  ISHIDOH,  K.,  KOMINAMI,  E.,  KREIDBERG,  J.  A.,  TOMINO,  Y.  &  MUNDEL,  P.  2004.  Podocyte  migration  during  nephrotic  syndrome  requires  a  coordinated  interplay  between cathepsin L and alpha3 integrin. J Biol Chem, 279, 34827‐32.  REISER, J., POLU, K. R., MOLLER, C. C., KENLAN, P., ALTINTAS, M. M., WEI, C., FAUL, C.,  HERBERT, S., VILLEGAS, I., AVILA‐CASADO, C., MCGEE, M., SUGIMOTO, H., BROWN, D.,  KALLURI, R., MUNDEL, P., SMITH, P. L., CLAPHAM, D. E. & POLLAK, M. R. 2005. TRPC6 is a  glomerular  slit  diaphragm‐associated  channel  required  for  normal  renal  function.  Nat  Genet, 37, 739‐44.  RODEWALD,  R.  &  KARNOVSKY,  M.  J.  1974.  Porous  substructure  of  the  glomerular  slit  diaphragm in the rat and mouse. J Cell Biol, 60, 423‐33.  ROSELLI, S., GRIBOUVAL, O., BOUTE, N., SICH, M., BENESSY, F., ATTIE, T., GUBLER, M. C.  & ANTIGNAC, C. 2002. Podocin localizes in the kidney to the slit diaphragm area. Am J  Pathol, 160, 131‐9.  ROSELLI, S., HEIDET, L., SICH, M., HENGER, A., KRETZLER, M., GUBLER, M. C. & ANTIGNAC,  C. 2004. Early glomerular filtration defect and severe renal disease in podocin‐deficient  mice. Mol Cell Biol, 24, 550‐60.  RUF, R. G., LICHTENBERGER, A., KARLE, S. M., HAAS, J. P., ANACLETO, F. E., SCHULTHEISS,  M.,  ZALEWSKI,  I.,  IMM,  A.,  RUF,  E.  M.,  MUCHA,  B.,  BAGGA,  A.,  NEUHAUS,  T.,  FUCHSHUBER, A., BAKKALOGLU, A. & HILDEBRANDT, F. 2004. Patients with mutations in  NPHS2 (podocin) do not respond to standard steroid treatment of nephrotic syndrome.  J Am Soc Nephrol, 15, 722‐32.  SAKO, M., NAKANISHI, K., OBANA, M., YATA, N., HOSHII, S., TAKAHASHI, S., WADA, N.,  TAKAHASHI,  Y.,  KAKU,  Y.,  SATOMURA,  K.,  IKEDA,  M.,  HONDA,  M.,  IIJIMA,  K.  &  YOSHIKAWA, N. 2005. Analysis of NPHS1, NPHS2, ACTN4, and WT1 in Japanese patients  with congenital nephrotic syndrome. Kidney Int, 67, 1248‐55.  SANTIN, S., ARS, E., ROSSETTI, S., SALIDO, E., SILVA, I., GARCIA‐MASET, R., GIMENEZ, I.,  RUIZ,  P.,  MENDIZABAL,  S.,  LUCIANO  NIETO,  J.,  PENA,  A.,  CAMACHO,  J.  A.,  FRAGA,  G.,  COBO, M. A., BERNIS, C., ORTIZ, A., DE PABLOS, A. L., SANCHEZ‐MORENO, A., PINTOS, G.,  140  MIRAPEIX,  E.,  FERNANDEZ‐LLAMA,  P.,  BALLARIN,  J.,  TORRA,  R.,  ZAMORA,  I.,  LOPEZ‐ HELLIN, J., MADRID, A., VENTURA, C., VILALTA, R., ESPINOSA, L., GARCIA, C., MELGOSA,  M., NAVARRO, M., GIMENEZ, A., COTS, J. V., ALEXANDRA, S., CARAMELO, C., EGIDO, J.,  SAN JOSE, M. D., DE LA CERDA, F., SALA, P., RASPALL, F., VILA, A., DAZA, A. M., VAZQUEZ,  M., ECIJA, J. L., ESPINOSA, M., JUSTA, M. L., POVEDA, R., APARICIO, C., ROSELL, J., MULEY,  R., MONTENEGRO, J., GONZALEZ, D., HIDALGO, E., DE FRUTOS, D. B., TRILLO, E., GRACIA,  S.  &  DE  LOS  RIOS,  F.  J.  2009a.  TRPC6  mutational  analysis  in  a  large  cohort  of  patients  with focal segmental glomerulosclerosis. Nephrol Dial Transplant, 24, 3089‐96.  SANTIN, S., GARCIA‐MASET, R., RUIZ, P., GIMENEZ, I., ZAMORA, I., PENA, A., MADRID, A.,  CAMACHO, J. A., FRAGA, G., SANCHEZ‐MORENO, A., COBO, M. A., BERNIS, C., ORTIZ, A.,  DE  PABLOS,  A.  L.,  PINTOS,  G.,  JUSTA,  M.  L.,  HIDALGO‐BARQUERO,  E.,  FERNANDEZ‐ LLAMA,  P.,  BALLARIN,  J.,  ARS,  E.  &  TORRA,  R.  2009b.  Nephrin  mutations  cause  childhood‐ and adult‐onset focal segmental glomerulosclerosis. Kidney Int.  SANTIN,  S.,  TAZON‐VEGA,  B.,  SILVA,  I.,  COBO,  M.  A.,  GIMENEZ,  I.,  RUIZ,  P.,  GARCIA‐ MASET, R., BALLARIN, J., TORRA, R. & ARS, E. 2011. Clinical value of NPHS2 analysis in  early‐  and  adult‐onset  steroid‐resistant  nephrotic  syndrome.  Clin  J  Am  Soc  Nephrol,  6,  344‐54.  SCHNABEL,  E.,  ANDERSON,  J.  M.  &  FARQUHAR,  M.  G.  1990.  The  tight  junction  protein  ZO‐1 is concentrated along slit diaphragms of the glomerular epithelium. J Cell Biol, 111,  1255‐63.  SCHULTHEISS,  M.,  RUF,  R.  G.,  MUCHA,  B.  E.,  WIGGINS,  R.,  FUCHSHUBER,  A.,  LICHTENBERGER,  A.  &  HILDEBRANDT,  F.  2004.  No  evidence  for  genotype/phenotype  correlation in NPHS1 and NPHS2 mutations. Pediatr Nephrol, 19, 1340‐8.  SCHWADERER, P., KNUPPEL, T., KONRAD, M., MEHLS, O., SCHARER, K., SCHAEFER, F. &  WEBER,  S.  2008.  Clinical  course  and  NPHS2  analysis  in  patients  with  late  steroid‐ resistant nephrotic syndrome. Pediatr Nephrol, 23, 251‐6.  SCHWARZ, K., SIMONS, M., REISER, J., SALEEM, M. A., FAUL, C., KRIZ, W., SHAW, A. S.,  HOLZMAN,  L.  B.  &  MUNDEL,  P.  2001.  Podocin,  a  raft‐associated  component  of  the  glomerular slit diaphragm, interacts with CD2AP and nephrin. J Clin Invest, 108, 1621‐9.  141  SHARMA,  M.,  SHARMA,  R.,  MCCARTHY,  E.  T.  &  SAVIN,  V.  J.  1999.  "The  FSGS  factor:"  enrichment and in vivo effect of activity from focal segmental glomerulosclerosis plasma.  J Am Soc Nephrol, 10, 552‐61.  SHARMA,  M.,  SHARMA,  R.,  MCCARTHY,  E.  T.  &  SAVIN,  V.  J.  2004.  The  focal  segmental  glomerulosclerosis permeability factor: biochemical characteristics and biological effects.  Exp Biol Med (Maywood), 229, 85‐98.  SHIH, N. Y., LI, J., KARPITSKII, V., NGUYEN, A., DUSTIN, M. L., KANAGAWA, O., MINER, J.  H. & SHAW, A. S. 1999. Congenital nephrotic syndrome in mice lacking CD2‐associated  protein. Science, 286, 312‐5.  SHONO, A., TSUKAGUCHI, H., KITAMURA, A., HIRAMOTO, R., QIN, X. S., DOI, T. & IIJIMA,  K.  2009.  Predisposition  to  relapsing  nephrotic  syndrome  by  a  nephrin  mutation  that  interferes with assembly of functioning microdomains. Hum Mol Genet, 18, 2943‐56.  SIEVERS,  F.,  WILM,  A.,  DINEEN,  D.,  GIBSON,  T.  J.,  KARPLUS,  K.,  LI,  W.,  LOPEZ,  R.,  MCWILLIAM,  H.,  REMMERT,  M.,  SODING,  J.,  THOMPSON,  J.  D.  &  HIGGINS,  D.  G.  2011.  Fast,  scalable  generation  of  high‐quality  protein  multiple  sequence  alignments  using  Clustal Omega. Mol Syst Biol, 7, 539.  STEGER, G. 1994. Thermal denaturation of double‐stranded nucleic acids: prediction of  temperatures  critical  for  gradient  gel  electrophoresis  and  polymerase  chain  reaction.  Nucleic Acids Res, 22, 2760‐8.  SUNYAEV, S., RAMENSKY, V., KOCH, I., LATHE, W., 3RD, KONDRASHOV, A. S. & BORK, P.  2001. Prediction of deleterious human alleles. Hum Mol Genet, 10, 591‐7.  SUO, C., XU, H., KHOR, C. C., ONG, R. T., SIM, X., CHEN, J., TAY, W. T., SIM, K. S., ZENG, Y.  X.,  ZHANG,  X.,  LIU,  J.,  TAI,  E.  S.,  WONG,  T.  Y.,  CHIA,  K.  S.  &  TEO,  Y.  Y.  2012.  Natural  positive  selection  and  north‐south  genetic  diversity  in  East  Asia.  Eur  J  Hum  Genet,  20,  102‐10.  TARSHISH,  P.,  TOBIN,  J.  N.,  BERNSTEIN,  J.  &  EDELMANN,  C.  M.,  JR.  1997.  Prognostic  significance  of  the  early  course  of  minimal  change  nephrotic  syndrome:  report  of  the  International Study of Kidney Disease in Children. J Am Soc Nephrol, 8, 769‐76.  142  TEO, Y. Y., SIM, X., ONG, R. T., TAN, A. K., CHEN, J., TANTOSO, E., SMALL, K. S., KU, C. S.,  LEE,  E.  J.,  SEIELSTAD,  M.  &  CHIA,  K.  S.  2009.  Singapore  Genome  Variation  Project:  a  haplotype map of three Southeast Asian populations. Genome Res, 19, 2154‐62.  THOMAS, D. B., FRANCESCHINI, N., HOGAN, S. L., TEN HOLDER, S., JENNETTE, C. E., FALK,  R.  J.  &  JENNETTE,  J.  C.  2006.  Clinical  and  pathologic  characteristics  of  focal  segmental  glomerulosclerosis pathologic variants. Kidney Int, 69, 920‐6.  TONNA, S. J., NEEDHAM, A., POLU, K., USCINSKI, A., APPEL, G. B., FALK, R. J., KATZ, A.,  AL‐WAHEEB, S., KAPLAN, B. S., JERUMS, G., SAVIGE, J., HARMON, J., ZHANG, K., CURHAN,  G. C. & POLLAK, M. R. 2008. NPHS2 variation in focal and segmental glomerulosclerosis.  BMC Nephrol, 9, 13.  TSUKAGUCHI,  H.,  SUDHAKAR,  A.,  LE,  T.  C.,  NGUYEN,  T.,  YAO,  J.,  SCHWIMMER,  J.  A.,  SCHACHTER, A. D.,  POCH, E., ABREU, P. F., APPEL, G. B.,  PEREIRA, A.  B., KALLURI, R.  &  POLLAK, M. R. 2002. NPHS2 mutations in late‐onset focal segmental glomerulosclerosis:  R229Q is a common disease‐associated allele. J Clin Invest, 110, 1659‐66.  VANDEVOORDE, R., WITTE, D., KOGAN, J. & GOEBEL, J. 2006. Pierson syndrome: a novel  cause of congenital nephrotic syndrome. Pediatrics, 118, e501‐5.  WANG,  J.,  RONAGHI,  M.,  CHONG,  S.  S.  &  LEE,  C.  G.  2011.  pfSNP:  An  integrated  potentially  functional  SNP  resource  that  facilitates  hypotheses  generation  through  knowledge syntheses. Hum Mutat, 32, 19‐24.  WARTIOVAARA, J., OFVERSTEDT, L. G., KHOSHNOODI, J., ZHANG, J., MAKELA, E., SANDIN,  S.,  RUOTSALAINEN,  V.,  CHENG,  R.  H.,  JALANKO,  H.,  SKOGLUND,  U.  &  TRYGGVASON,  K.  2004.  Nephrin  strands  contribute  to  a  porous  slit  diaphragm  scaffold  as  revealed  by  electron tomography. J Clin Invest, 114, 1475‐83.  WEBER, S., GRIBOUVAL, O., ESQUIVEL, E. L., MORINIERE, V., TETE, M. J., LEGENDRE, C.,  NIAUDET,  P.  &  ANTIGNAC,  C.  2004.  NPHS2  mutation  analysis  shows  genetic  heterogeneity  of  steroid‐resistant  nephrotic  syndrome  and  low  post‐transplant  recurrence. Kidney Int, 66, 571‐9.  WEI,  C.,  EL  HINDI,  S.,  LI,  J.,  FORNONI,  A.,  GOES,  N.,  SAGESHIMA,  J.,  MAIGUEL,  D.,  KARUMANCHI, S. A., YAP, H. K., SALEEM, M., ZHANG, Q., NIKOLIC, B., CHAUDHURI, A.,  143  DAFTARIAN,  P.,  SALIDO,  E.,  TORRES,  A.,  SALIFU,  M.,  SARWAL,  M.  M.,  SCHAEFER,  F.,  MORATH, C., SCHWENGER, V., ZEIER, M., GUPTA, V., ROTH, D., RASTALDI, M. P., BURKE,  G.,  RUIZ,  P.  &  REISER,  J.  2011.  Circulating  urokinase  receptor  as  a  cause  of  focal  segmental glomerulosclerosis. Nat Med, 17, 952‐60.  WIGGINS, R. C.  2007. The spectrum of podocytopathies: a unifying view of glomerular  diseases. Kidney Int, 71, 1205‐14.  YAP,  H.‐K.  2005.  Nephrotic  syndrome  presenting  in  the  teens  [Online].  Singapore:  National  University  of  Singapore.  Available:  http://www.med.nus.edu.sg/paed/academic/NEPH_NS_in_teens.htm [Accessed].  YAP,  H.‐K.  &  LAU,  P.  Y.‐W.  2008.  Hematuria  and  Proteinuria.  In:  GEARY,  D.  F.  &  SCHAEFER, F. (eds.) Comprehensive Pediatric Nephrology. 1 ed.: Mosby Elsevier.  YE, S., DHILLON, S., KE, X., COLLINS, A. R. & DAY, I. N. 2001. An efficient procedure for  genotyping single nucleotide polymorphisms. Nucleic Acids Res, 29, E88‐8.  YU,  Z.,  DING,  J.,  GUAN,  N.,  SHI,  Y.,  ZHANG,  J.,  HUANG,  J.,  YAO,  Y.  &  YANG,  J.  2004.  A  novel mutation of NPHS2 identified in a Chinese family. Pediatr Nephrol, 19, 1285‐9.  YU,  Z.,  DING,  J.,  HUANG,  J.,  YAO,  Y.,  XIAO,  H.,  ZHANG,  J.,  LIU,  J.  &  YANG,  J.  2005.  Mutations  in  NPHS2  in  sporadic  steroid‐resistant  nephrotic  syndrome  in  Chinese  children. Nephrol Dial Transplant, 20, 902‐8.  ZENKER, M., AIGNER, T., WENDLER, O., TRALAU, T., MUNTEFERING, H., FENSKI, R., PITZ,  S., SCHUMACHER, V., ROYER‐POKORA, B., WUHL, E., COCHAT, P., BOUVIER, R., KRAUS, C.,  MARK, K., MADLON, H., DOTSCH, J., RASCHER, W., MARUNIAK‐CHUDEK, I., LENNERT, T.,  NEUMANN,  L.  M.  &  REIS,  A.  2004.  Human  laminin  beta2  deficiency  causes  congenital  nephrosis with mesangial sclerosis and distinct eye abnormalities. Hum Mol Genet, 13,  2625‐32.  ZHANG,  S.  Y.,  MARLIER,  A.,  GRIBOUVAL,  O.,  GILBERT,  T.,  HEIDET,  L.,  ANTIGNAC,  C.  &  GUBLER, M. C. 2004. In vivo expression of podocyte slit diaphragm‐associated proteins  in nephrotic patients with NPHS2 mutation. Kidney Int, 66, 945‐54.  ZHANG,  X.  H.  &  CHASIN,  L.  A.  2004.  Computational  definition  of  sequence  motifs  governing constitutive exon splicing. Genes Dev, 18, 1241‐50.  144  ZHOU, W. & HILDEBRANDT, F. 2009. Molecular cloning and expression of phospholipase  C epsilon 1 in zebrafish. Gene Expr Patterns, 9, 282‐8.  ZHU, L., YU, L., WANG, C. D., LV, J. C., LI, G. S., ZHANG, H. & WANG, H. Y. 2009. Genetic  effect  of  the  NPHS2  gene  variants  on  proteinuria  in  minimal  change  disease  and  immunoglobulin A nephropathy. Nephrology (Carlton), 14, 728‐34.      145  [...]... nephrin  signal  transduction(Huber et al., 2003b).  Podocin has been found to increase the efficiency of nephrin signaling without the recruitment of other signaling molecules. The cytoplasmic  tail  of nephrin  binds  to  podocin  and  this  interaction  is  mediated  by  the  C‐terminal  domain of podocin (Huber et al., 2001). The phosphorylation of nephrin also increases  the binding of nephrin and podocin (Li et al., 2004).  ... 8th Congress of Asian Society for Paediatrics Research, Seoul, 2012  Title: Multiple Rare Alleles in Nephrin and Podocin Contribute to Poor Prognosis in Childhood Nephrotic Syndrome in Singapore.   Nominated for Young Investigator Award  xiv    1 Introduction  1.1 Nephrotic syndrome Nephrotic syndrome (NS)  is  a  common  cause  of kidney  disease  in children.    It  is  characterized  by  heavy  proteinuria,  hypoalbuminaemia,  oedema  and  hyperlipidaemia ... al.,  2003).    In SD,  CD2AP  may  function as an adaptor protein to anchor the C‐terminal cytoplasmic domain of nephrin  and/or  podocin  to  the  actin  cytoskeleton.  CD2AP  is  also  involved  with  nephrin  and  podocin in cell signaling pathway (Mao et al., 2006).     Nephrin  Nephrin  is  a  transmembrane  protein  belonging  to  the  immunoglobulin  superfamily.  It  has  an  N‐terminal  signal ... recurrent  FSGS  patients,   when  injected  into  experimental  animals  can  induce  proteinuria  or  albuminuria. All these suggested the presence of plasma circulating factor that may be  involved in glomerular damage and the initiation of glomerular events which contribute  to the development of proteinuria in FSGS (Sharma et al., 1999, Sharma et al., 2004) .  Cardiotrophin  like  cytokine‐1  has  been ... podocyte  injury.  Hara  et  al.  had  reported  in their  study  urinary  loss  of podocytes  in FSGS  and  MCNS  patients.  They observed that the urinary loss of podocytes in FSGS was higher compared  to those with MCNS. This suggested that the podocyte injury in FSGS patients was more  serious, explaining why the prognosis in FSGS was poorer. The effacement of podocytes  was identified as the initial event in FSGS. Podocytes have a limited ability to repair and ... focused on its structure. Structural abnormalities of the GBM may result in proteinuria  10    and  hematuria,  as  observed  in the  X‐linked  form  of Alport’s  syndrome that  is  due  to  mainly collagen IVα5 chain mutations and also collagen IV α3 and α4 chains (Barker et  al., 1990). Severe or truncating mutations in the LAMB2 gene, which encodes for laminin  β2 in GBM, result in congenital NS (CNS) associated with microcoria (Zenker et al., 2004, ... processes  of glomerular  development  at  the  capillary  loop  stage  (Hinkes et al., 2006).  1.2.2 Cytoskeletal proteins  Actin  cytoskeletal  proteins  play  an  important  role  in the  regulation  of the  plasticity  of the  podocyte  cytoskeleton.  Therefore  they  are  important  in the  maintenance  of the  filtration barrier. Alpha actinin 4 (ACTN4) is an actin bundling protein that is responsible ... (Yu et al., 2005). There have been various definitions used to describe nephrotic range  proteinuria,  including  urinary  protein  excretion  ≥3  g/day/1.73m2  or  a  spot  urinary  protein:creatinine ratio ≥0.2 g/mmol (Yap and Lau, 2008).    NS in children can be congenital, which is presented at birth or during first 3 months of life,  or  presented  in later  part  of their  life.  Childhood  NS  is ... domain  that  containing  eight  immunoglobulin motifs, a fibronectin type III‐like domain, a transmembrane domain and  lastly a cytosolic C‐terminal domain (Kestila et al., 1998, Lenkkeri et al., 1999). Nephrin is  localized  within  normal  human  kidney  and  is  distinctly  localized  at  the  SD  of the  13    glomerular podocytes (Holthofer et al., 1999, Patrakka et al., 2000). Mice with nephrin ... sensitive  nephrotic syndrome (SSNS) is defined as patients being able to enter remission in response to corticosteroid  treatment alone. Steroid resistant nephrotic syndrome (SRNS) is the inability to induce  remission  after  8  weeks  of corticosteroid  treatment.    Steroid  dependent  nephrotic syndrome (SDNS)  is  the  initial  response  to  corticosteroid  treatment  by  entering  complete remission but the development of relapse either while still receiving steroids  ... Mutations  in TRPC6  resulted in a gain of function.  Mutations in WT1 affect the DNA‐binding affinity of WT1  to the target gene. Heterozygous de novo mutations in WT1 resulting in the inability of ... C‐terminal  domain of podocin (Huber et al., 2001). The phosphorylation of nephrin also increases  the binding of nephrin and podocin (Li et al., 2004).     TRPC6  TRPC6  is  a  member  of the ... Nephrin and other variants of nephrotic syndrome Although  nephrin  is  apparently  pivotal  in development  of CNF,  the  role  of nephrin  in other variants of NS is also of interest Lahdenkari et al. investigated NPHS1 in patients

Ngày đăng: 07/10/2015, 10:02

Từ khóa liên quan

Tài liệu cùng người dùng

  • Đang cập nhật ...

Tài liệu liên quan