Kidney International, Vol. 57 (2000), pp. 1818–1835

PERSPECTIVES IN BASIC SCIENCE

Evidence for genetic factors in the development andprogression of IgA nephropathy

STEPHEN I-HONG HSU, SYLVIA B. RAMIREZ, MICHELLE P. WINN, JOSEPH V. BONVENTRE,and WILLIAM F. OWEN

Departments of Medicine and Pediatrics, Faculty of Medicine, National University of Singapore, and Research Division,National Kidney Foundation of Singapore, Singapore; Renal Unit and Department of Medicine, Massachusetts General Hospitaland Harvard Medical School, Charlestown, Boston, Massachusetts; Duke Institute for Renal Outcomes Research and HealthPolicy, Duke Clinical Research Institute, and Duke Center for Human Genetics, Duke University Medical Center, Durham,North Carolina, USA

disease/susceptibility genes. Alternatively, if the case-controlEvidence for genetic factors in the development and progres-study design is employed to identify associations between par-sion of IgA nephropathy.ticular candidate genes or markers and the development ofBackground. IgA nephropathy (IgAN) is the most commonIgAN, spurious associations caused by the effects of populationglomerulonephritis in the world among patients undergoingstratification should be ruled out by confirming the findingsrenal biopsy. Once considered a relatively benign condition,using powerful and sensitive family-based methodologies suchlongitudinal follow-up studies have revealed that in fact 9 toas the transmission/dysequilibrium test (TDT).50% of patients progress to end-stage renal disease within

20 years of disease onset. In the three decades since its firstdescription by Jean Berger and Nicole Hinglais, clinical, epide-miologic, and immunologic studies of the pathogenesis of pri- Over three decades have passed since Jean Berger andmary (idiopathic) mesangial glomerulonephritis with predomi-

Nicole Hinglais first reported the immunofluorescent de-nant IgA deposits have characterized the features of IgAN astection of IgA and IgG mesangial deposits in kidneya distinct glomerular disease entity. However, the basic molecu-

lar mechanism(s) underlying abnormal IgA deposition in the biopsy specimens from patients with persistent hematuriamesangium with ensuing extracellular matrix expansion and and proteinuria [1]. They subsequently proposed primarymesangial cell proliferation remains poorly understood. The (idiopathic) mesangial glomerulonephritis with predom-task of elucidating the molecular basis of IgAN is made espe-

inantly IgA deposits as a distinct entity [2]. The twocially challenging by the fact that both environmental and ge-decades following Berger’s seminal articles can be char-netic components likely contribute to the development and

progression of IgAN. acterized as a period of descriptive study and growingMethods and Results. We review here the evidence for ge- appreciation of the unique immunopathologic features

netic factors in the development and progression of IgAN, of both primary IgA nephropathy (pIgAN), the closelyincluding a reappraisal of earlier conflicting results from small

linked systemic illness Henoch-Schonlein purpura (HSP),immunogenetic case-control studies, the evidence for racialand a broad spectrum of glomerulopathies characterizeddifferences in the prevalence of IgAN, a detailed summary ofby IgA deposition such as that observed in lupus nephri-all reported occurrences of familial IgAN worldwide, and an

exhaustive review of new insights gained through the study tis and hepatic glomerulosclerosis [reviewed in 3, 4]. Thisof two murine models of hereditary IgAN: the ddY and the has prompted the widely held notion of IgAN as a syn-uteroglobin-deficient mouse. drome and has led to the clinicopathologic distinctionConclusions. With the development of powerful molecular

between primary and secondary forms of disease associ-genetic approaches to the study of both Mendelian and com-ated with IgA mesangial deposits [reviewed in 2, 3, 5–7].plex human genetic diseases, and the successful efforts of inves-

tigators to identify and clinically characterize large IgAN multi- pIgAN, defined as predominant IgA mesangial depositsplex families, we propose that genetic analysis of familial IgAN in the absence of clinical or laboratory evidence of otheris the most promising approach to the identification of IgAN associated systemic disease, has also been proposed to

include at least two distinct clinical entities (those withand without macroscopic hematuria) [5, 7].Key words: heredity, IgAN, familial nephropathy, glomerulonephritis,

chronic renal failure, hematuria. During this same time period, demographic studiesrevealed that IgAN was the most common glomerulone-Received for publication July 21, 1999phritis in the world among patients undergoing renaland in revised form October 15, 1999

Accepted for publication December 9, 1999 biopsy [8–11]; the prevalence has been observed to be asgreat as 50% of all glomerular diseases in Asia, especially 2000 by the International Society of Nephrology

1818

Hsu et al: Genetic factors in IgA nephropathy 1819

among the Japanese. The clinical significance of this dis- Bw35 [26, 27]. A strong association between benign(nonprogressive) IgAN and the HLA-DR4 antigen wasorder is further emphasized by its natural history. Longi-

tudinal follow-up studies of biopsy-proven cases of IgAN first reported in a Japanese cohort of 103 adult patients[28]. Although numerous independent groups have re-reveal that 9 to 50% of the patients progress to end-

stage renal disease (ESRD) within 20 years of disease ported these same results, neither Bw35 nor DR4 hasbeen consistently confirmed to be associated with pIgANonset [8–11]. Despite intensive efforts to elucidate the

pathogenesis of the IgA glomerulopathies, employing [reviewed in 22]. The explanation for the discrepancyamong these studies has been attributed to differencesboth immunologic and immunogenetic approaches, little

progress has been made in defining the precise molecular in policies governing diagnostic renal biopsy, the smallnumber of patients enrolled, racial differences that re-mechanism(s) of this common and clinically important

glomerular disease process [4, 12, 13]. While powerful flect the variation in genetic background of cohorts fromdifferent countries, genetic heterogeneity of IgAN dis-molecular genetic approaches have been applied to the

study of both Mendelian and complex human genetic dis- ease/susceptibility genes, and the effect of environmentalfactors on the expression of such genes (the genetic con-eases, leading to the identification of novel genes responsi-

ble for kidney diseases such as autosomal dominant poly- cept of epistasis) in different populations [22].The observed inconsistencies have also raised the pos-cystic kidney disease [14–16], juvenile nephronophthisis

type I [17], and congenital nephrotic syndrome of the sibility that the genes encoding the HLA antigens act asmarkers for a closely linked disease/susceptibility locus,Finnish type [18], progress has not been made in identi-

fying the genes involved in the IgA glomerulopathies. rather than being disease-causing genes themselves [21, 22].This explanation appeared to gain initial support withThis review summarizes the evidence for genetic fac-

tors in the development and progression of the IgA glom- the introduction of molecular biological techniques fordetermining MHC gene allelic variation through the anal-erulopathies. We begin with a reappraisal of earlier con-

flicting results from small case-control studies on the ysis of DNA restriction fragment length polymorphisms(RFLPs) and the use of allele-specific oligonucleotiderole of immune response genes [centered on the major

histocompatibility complex (MHC) genes] in light of (ASO) hybridization probes. These techniques overcamethe limitations of traditional HLA typing methods andmore recent studies using high-resolution DNA genotyp-

ing, as well as important new reports of gene polymor- allowed investigators to distinguish definitively betweentwo or more alleles for a single MHC gene. The earliestphisms associated with IgAN disease progression. This

is followed by a critical examination of the evidence case-control study employing RFLP typing was performedby Moore et al in 1990. These investigators initially dem-for racial differences in the prevalence of IgAN and a

detailed summary of all reported occurrences of familial onstrated an association between the MHC class II allelicpolymorphism HLA-DQB1 and pIgAN among BritishIgAN worldwide. Finally, we review new insights gained

about the pathogenesis of IgAN through the ongoing patients (Table 1) [29]. This finding led them to proposethat, since the IgA antibody response is T-cell depen-study of two murine models of hereditary IgAN: the

ddY and the uteroglobin (UG)-deficient mouse. The role dent, the MHC class II products encoded by the DP,DQ, and DR genes might play a crucial role in the presen-of environmental factors in the etiology of IgAN is also

discussed in the specific context of their impact on ge- tation of processed antigen to specific T-cells. The appar-ent DQB1 association could be of pathogenetic signifi-netic approaches to disease gene mapping [19].cance in pIgAN by determining the responder status toeither native or foreign antigen. This hypothesis possessed

CANDIDATE DISEASE GENEboth biologic plausibility and experimental support from

POLYMORPHISM STUDIESthe work of other investigators. Several autoimmune

Role of MHC class II gene polymorphisms in pIgAN diseases, including rheumatoid arthritis (RA) and insu-lin-dependent diabetes mellitus (IDDM), have also beenThe earliest studies of the role of genetic factors in

pIgAN were reported in the mid-1970s. These studies linked to allelic polymorphisms of the DQB1 product[reviewed in 29, 30]. However, the early studies of Moorefocused on the association between specific HLA anti-

gens encoded by the human MHC locus and the develop- et al relied on the presence of a combination of the Taq IRFLP fragments T2 (2.0 kb) and T6 (6.0 kg) to definement of pIgAN, and defined allelic variation (the HLA

genotype) by traditional serologic and cellular tech- a disease-associated DQB1 polymorphism, a techniquethat was subsequently revealed by polymerase chain re-niques that require the expression of membrane-bound

HLA proteins on lymphocytes [reviewed in 20–22]. A action (PCR)-RFLP (discussed later in this article) to beunable to distinguish between multiple allelic variationsstrong association between pIgAN and the HLA Bw35

antigen was initially reported in small Australian and at the HLA-DQB1 locus. For example, the T2 RFLP iden-tifies but cannot distinguish between the HLA-DQ*0302,French cohorts [23–25], and was supported by several

reports of twins with IgAN who shared the HLA-antigen *0303, and *0402 alleles [31]. Although a subsequent study

Hsu et al: Genetic factors in IgA nephropathy1820

Table 1. Genotypic studies of HLA-DQ alleles and their association with pIgAN

Patients ControlsFollow-up Subgroup

Ref. Year Race Ethnicity years Methods HLA comparisons P value OR% frequency

[32] 1990 Caucasian British 98 (50) 94 (16) RFLP DQB1 (T2/T6) ,0.0001 5.3[33] 1991 Caucasian British 35 (71) 1103 (28) RFLP/ASO DQB1*0301 ,0.001 6.4[23] 1993 Caucasian Italian 73 (29) 63 (32) RFLP DQB1 (T2/T6) NS

Finnish 47 (28) 34 (35) DQB1 (T2/T6) NS[25] 1993 Caucasian German 67 (43) 22 (23) RFLP DQA1*0101 ,0.05 2.6[38] 1995 Caucasian Australian 70 230 RFLP DQA/DQB typinga NS[42] 1991 Multiracial Italy

Turin 76 1108 PCR/ASO DQA1 (6 alleles) Sporadic NS(Ethically DQB1 (7 alleles)a NS

21 matched DQA1/DQB1 Familial NSPavia 94 controls) DQA1/DQB1 Sporadic NS

Japan 39 (18) ? (38) DQB1*0601 Sporadic ,0.05 5.18b

39 (13) ? (30) DQB1*0301 ,0.025 4.36b

[39] 1994 Caucasian American 64 75 PCR-RFLP DQB1 (17 alleles)a NS[40] 1995 Caucasian French 58 (26) 150 (16) PCR-RFLP DQB1*0301 All Pts vs. control NS

? 21 (62) 150 (29) Progressors vs.controls ,0.0007 5.85

? 21 (62) 37 (35) Progressors vs.non-progressors ? 2.89

[36] 1995 Caucasian French 62 170 PCR-RFLP DQB1 (7 alleles) NS[35] 1996 Caucasian British 105 111 PCR-RFLP DQB1 (14 alleles)a NS

Italian 71 63 NSFinnish 48 41 NS

[41] 1996 Asian Korean 30 46 PCR-RFLP DQB1 (10 alleles)a NS30 (80) 46 (50) DQA1*0301 0.04 4.0

[34] 1994 Asian Chinese 79 104 PCR-RFLP DQB1 (6 alleles)a All Pts NS79 (41) 104 (28) DQA2 U All Pts NS

$1 21 (67) 52 (27) DQA2 U Progressors vs.non-progressors 0.0016 5.4

$1 21 (67) 104 (28) DQA2 U Progressors vs.control ,0.0003 5.2

79 (33) 104 (20) DQA1*0201 All Pts NS$1 21 (14) 52 (40) DQA1*0201 Progressors vs.

non-progressors 0.03 0.2

Abbreviations are: Ref., reference; HLA, human lymphocyte antigen; RFLP, restriction fragment length polymorphism; ASO, allele-specific oligonucleotide; OR,odds ratio; Pts, patients.

aGenotyping panel included the DQB1*0301 allelebChi-square test

employing higher resolution RFLP/ASO genotyping of of follow-up and adequate statistical analysis for the com-parison of progressors versus nonprogressors was not pro-British patients confirmed that the specific allele HLA-

DQB1*0301 was strongly associated with pIgAN [32], vided, found an association between DQB1*0301 and dis-ease progression [36]. All six additional case-control studiesthe follow-up report of Moore failed to find any associa-

tion between HLA-DQB1 and IgAN in Italian and Finn- reported to date using standardized PCR-RFLP HLAgenotyping [35–41] failed to find an association betweenish patients [21]. Two additional RFLP studies in Ger-

man and Australian patients [20, 33] also failed to find a any allele at the DQB1 locus (including DQB1*0301)and the development or progression of pIgAN in multi-significant association at the HLA-DQB1 locus (Table 1).

With introduction of the PCR using sequence-specific ple racial and ethnic groups (lower half of Table 1). Inone Japanese cohort, the DQB1*0601 and *0301 allelesprimers followed by subsequent restriction enzyme diges-

tion (PCR-RFLP), HLA genotyping methods achieved the were reported to be less frequent (that is, protectivefor IgAN) among patients with pIgAN compared withability to discriminate up to 16 distinct alleles at the HLA-

DQB1 locus [34]. When Fennessy et al repeated the controls [37]. Finally, in an analysis of 21 Italian patientsbelonging to six families from Turin (familial IgAN), nogenotyping of British, Italian, and Finnish patients using

PCR-RFLP to examine the association between pIgAN association with any HLA allele was observed (includingHLA-DQB1*0301) [37]. Similarly, PCR-RFLP genotyp-and 14 distinct DQB1 alleles (including DQB1*0301), as

well as 8 distinct DQA1 alleles, no consistent association ing of 26 affected members belonging to 10 families withancestral roots in the Valsaviore valley in Northern Italyof any DQ alleles was apparent in any of the populations

studied (including the British patients) [35]. Only a small and comparison to either healthy family members, pa-tients with sporadic pIgAN, or healthy ethnically matchedstudy of French Caucasian patients, in which the length

Hsu et al: Genetic factors in IgA nephropathy 1821

controls found no compelling evidence for involvement Role of angiotensin-converting enzyme (ACE) geneof the HLA system in the pathogenesis of IgAN in af- polymorphisms in the progression of pIgANfected family members [42]. Increased production or activity of angiotensin II

Thus, despite intriguing early reports of a strong associa- (Ang II) and decreased levels of bradykinin have beention between the HLA system and pIgAN, case-control hypothesized to play a detrimental role in the glomerularstudies in the past decade, employing high-resolution response to injury. The proposed pathophysiology is me-HLA genotyping methodologies, demonstrate rather diated through adverse intraglomerular hemodynamicconsistently that at least among Caucasian patients of mechanisms [47, 48], promotion of inappropriate glo-multiple ethnic backgrounds with either sporadic or fa- merular hypertrophy and sclerosis [49], and/or the prolif-milial IgAN, allelic variation of HLA genes in the MHC erative effects of increased Ang II and decreased brady-class II region does not appear to be associated with the kinin on tubular, mesangial, and smooth muscle cellsdevelopment of pIgAN. These findings differ from those [50, 51]. Following the discovery of an insertion/deletionin Asian patients, in which several HLA-DQA alleles polymorphism in intron 16 of the human ACE gene, whichhave been reported to be either associated with the oc- accounts for half of the variation in serum ACE levelscurrence of pIgAN in Korean patients [41], or associated in a Caucasian study cohort [52], and the report of strongwith disease progression (DQA2 U associated with pro- associations between the D allele and common clinicalgression, DQA*0201 associated with benign course) in conditions such as cardiovascular disease (myocardial in-Chinese patients (Table 1) [40]. These small isolated farction, left ventricular hypertrophy, and hypertension)case-control studies are suggestive, but require confir- [reviewed in 53], end-organ complications of diabetesmation by prospective studies of larger cohorts or by the mellitus (diabetic retinopathy and nephropathy) [reviewedgenetic analysis of familial IgAN (discussed later in this in 54], and progression of renal disease (glomerular dis-article) before any conclusions can be drawn regarding ease, tubulointerstitial disease, polycystic kidney dis-the role of HLA gene polymorphisms in the pathogenesis ease) [55–57], eight major studies have explored the asso-and/or progression of pIgAN in different ethnic groups.

ciation between the ACE gene insertion/deletion (I/D)polymorphism and the development and/or progressionRole of T-cell receptor (TCR) a and b chain geneof pIgAN [58–65]. Details of the design and relevantpolymorphisms in pIgANfindings from these studies are summarized in Table 2Hyperactivity of T cells and high serum levels of T-cell–and are critically evaluated later in this article.derived cytokines such as interleukin-2 (IL-2) are charac-

All but one of these studies employed a combinationteristic of pIgAN [43, 44]. This has prompted severalof the case-control design (to identify an association be-investigators to examine the association between TCRtween ACE gene polymorphisms and the developmentgene polymorphisms and the development and/or progres-of disease) and the retrospective cohort design (to iden-sion of IgAN. A case-control study of the TCR Cb poly-tify an association between ACE gene polymorphismsmorphism determined by RFLP analysis of Bgl II 10.0/9.2and the progression of disease). The exception is a clini-kb fragments in 40 German patients with biopsy-provencopathologic cross-sectional study [64]. These studies ex-pIgAN found no difference in allele frequencies com-hibit important differences with regard to racial/ethnicpared with healthy controls [20]. However, a cohort studystudy populations and differences in the length of follow-among Japanese patients with biopsy-proven pIgANup (range, 1 to $10 years), both of which may potentiallyfound that the 10 kb TCR Cb allele was associated withlead to conflicting results. The relevance of race is bestdisease progression and higher maximal proteinuria afterexemplified by the finding that the association betweena follow-up of an average of 3.5 to 4.0 years [45]. Athe ACE DD genotype and higher ACE activity wassingle case-control study of TCR Ca gene polymorphismssignificant in French and American Caucasians (DD vs.determined by RFLP analysis of Taq I 7/2 kb fragmentsII, P 5 0.0001) and a Japanese cohort [52, 66, 67], butin 53 Chinese patients with biopsy-proven pIgAN foundnot in an African American cohort [66]. The relevancea significantly higher frequency of the 7 kb TCR Ca alleleof the length of clinical follow-up is raised by the factin patients versus healthy controls [60.3% (32 out of 53)that although none of the studies found a significantvs. 38.8% (26 out of 67), P , 0.05] [46]. There was nodifference in ACE I/D genotype frequencies betweenassociation of any allele with either histologic gradingpatients with pIgAN compared with healthy ethnicallyor disease progression in this cohort. These isolated stud-matched controls, the majority of these studies also com-ies are suggestive of a role for germline polymophismspared ACE I/D genotype frequencies between cohortsof the unrearranged TCR Ca and Cb genes in the devel-with stable renal function (defined variously as mainte-opment and progression of pIgAN, respectively. Thenance of creatinine #1.5 mg/dL or failure to exhibit aTCR genes should be considered candidate disease/sus-doubling of creatinine) versus patients with progressiveceptibility loci and further evaluated in larger prospec-

tive studies or by the genetic analysis of familial IgAN. deterioration of renal function over time (including de-

Hsu et al: Genetic factors in IgA nephropathy1822

Tab

le2.

Ang

iote

nsin

conv

erti

ngen

zym

e(A

CE

)ge

nepo

lym

orph

ism

sas

soci

ated

wit

hpI

gAN

dise

ase

prog

ress

ion

Pol

ymor

phis

m%

Fre

quen

cyof

case

s/co

ntro

lsF

ollo

w-u

pSu

bgro

upco

mpa

riso

nsR

ef.

Yea

rR

ace

Eth

nici

tyye

ars

Cas

esC

ontr

ols

DD

IDII

ID/I

Ior

mea

sure

men

tsP

valu

eO

R

[46]

1995

Cau

casi

anSc

otti

sh.

110

010

040

/39

41/4

219

/17

—P

tsvs

.con

trol

sN

S10

0—

23.

22

0.6

21.

3—

Dec

line

inre

nal

fxna

0.00

530

—47

3320

—F

req

ofE

SRD

NS

30—

3742

48—

Med

ian

age

atE

SRD

NS

[47]

1995

Asi

anJa

pane

se.

1028

2542

/12

29/4

429

/44

58/8

8P

rogr

esso

rsvs

.non

-pro

gres

sors

,0.

053.

94b

2846

42/1

129

/44

29/4

558

/89

Pro

gres

sors

vs.c

ontr

ol,

0.00

54.

62b

2546

12/1

144

/44

44/4

588

/89

Non

-pro

gres

sors

vs.c

ontr

olN

S21

—2

1.02

20.

101

0.25

—E

ffec

tof

AC

EI

ther

apy

,0.

05c

[48]

1995

Asi

anJa

pane

se$

1025

2412

/040

/58

48/4

2—

Pro

gres

sors

vs.n

on-p

rogr

esso

rsN

S24

100

0/6

58/5

542

/39

Non

-pro

gres

sors

vs.c

ontr

olN

S25

100

12/6

40/5

548

/39

Pro

gres

sors

vs.c

ontr

olN

S[4

9]19

95C

auca

sian

Ger

man

$5

8212

246

/34

36/4

218

/24

—E

SRD

vs.n

on-p

rogr

esso

rsN

SA

ustr

alia

n12

223

434

/33

42/5

024

/17

Non

-pro

gres

sors

vs.c

ontr

olN

SIt

alia

n82

234

46/3

336

/50

18/1

7E

SRD

vs.c

ontr

ol0.

06A

ustr

ian

[50]

1996

Cau

casi

anA

mer

ican

Med

ian

7.4

Nor

mal

Init

ial

Cr

(1to

15.3

)18

2333

/444

/61

22/3

5—

Pro

gres

sors

vs.n

on-p

rogr

esso

rs0.

057

Med

ian

6.9

Nor

mal

Cr/

BP

/Upr

otei

n

(1–1

3)5

1560

/040

/47

0/53

—P

rogr

esso

rsvs

.non

-pro

gres

sors

0.00

9[5

1]19

97C

auca

sian

Can

adia

nM

ean

168

100

33/3

048

/49

19/2

1A

llP

tsvs

.con

trol

sN

S6.

116

8—

26.

62

6.0

26.

6D

eclin

ein

rena

lfx

ndN

S(a

ll$

2)60

102

26/

27

23/

28

AG

TM

Mvs

.MM

/MT

geno

type

e0.

008

[52]

1997

Asi

anJa

pane

seB

xst

udy

9771

16/1

349

/49

35/3

8A

llP

tsvs

.con

trol

sN

S97

—0.

561.

02U

rine

prot

einf

0.12

55.2

61.8

Mes

angi

alpr

olif

erat

ion

NS

13.8

14.9

Cre

scen

tsN

S2.

58.

0C

apsu

lar

adhe

sion

s0.

017

1.4

5.1

Scle

rosi

s0.

28[5

3]19

99C

auca

sian

Ital

ian

$10

8150

43/4

237

/40

20/1

8—

Pts

vs.c

ontr

ols

NS

1829

5282

Ren

alsu

rviv

alra

teg

0.02

5.41

h

0.04

93.

07i

4130

49/3

039

/43

12/2

7—

ESR

Dvs

.non

-pro

gres

sors

NSj

39.6

46M

ean

age

aton

set

HD

(yea

rs)

NS

a The

rate

ofde

clin

eof

rena

lfu

ncti

ongi

ven

inL

/mm

ol/d

ayb O

dds

rati

osof

the

DD

geno

type

rela

tive

toth

eco

mbi

nati

onof

IDan

dII

unad

just

edfo

rag

ean

dse

xc Si

gnif

ican

tde

clin

ein

urin

ary

prot

ein

excr

etio

non

lyin

the

DD

geno

type

grou

paf

ter

48w

eeks

ofA

CE

inhi

bito

rth

erap

yin

21pa

tien

tsm

atch

edfo

rag

e,se

x,ba

selin

ebl

ood

pres

sure

,ur

inar

ypr

otei

nex

cret

ion

rate

,ser

umC

r,an

dur

inar

yC

rcl

eara

nce;

DD

(N5

9),I

D(N

56)

,II

(N5

6)d T

hera

teof

decl

ine

ofre

nal

func

tion

give

nin

mL

/min

/yea

re M

ultiv

aria

tean

alys

isof

the

rate

ofde

clin

eof

rena

lfu

ncti

ongi

ven

inm

L/m

in/y

r,in

pati

ents

wit

hco

mbi

natio

nsof

AC

ED

Dan

dID

/II

geno

type

san

dA

CE

AG

TM

Man

dM

T/T

Tge

noty

pes;

MM

and

DD

(N5

16),

MM

and

ID/I

I(N

544

),M

T/T

Tan

dD

D(N

539

),M

T/T

Tan

dID

/II

(N5

69)

f Uri

nary

prot

ein

excr

etio

nat

tim

eof

biop

sy(g

/day

/m2

body

surf

ace

area

)g K

apla

n-M

eier

anal

ysis

of$

10ye

arfo

llow

-up

for

pati

ents

wit

hse

rum

Cr

leve

lsof

#1.

5m

g/dL

atth

eti

me

ofbi

opsy

,com

pari

ngth

eti

me

from

rena

lbio

psy

toin

itia

tion

ofdi

alys

is(o

rdo

ublin

gof

seru

mC

rle

vels

)in

DD

vs.I

D/I

Ih C

hi-s

quar

ean

alys

isi H

azar

dra

tio

(HR

)by

Cox

prop

orti

onal

haza

rdm

ulti

vari

ate

anal

ysis

adju

stin

gfo

rth

eco

vari

ates

ofse

x,ag

eat

biop

sy,H

TN

,hea

vypr

otei

nuri

a,an

dA

CE

inhi

bito

rth

erap

yj M

ulti

vari

ate

logi

stic

regr

essi

onan

alys

isad

just

ing

for

the

cova

riat

esof

sex,

age

atbi

opsy

,HT

N,h

eavy

prot

einu

ria,

and

AC

Ein

hibi

tor

ther

apy

Hsu et al: Genetic factors in IgA nephropathy 1823

velopment of ESRD). An inadequate length of clinical other possible risk factors for disease progression suchas hypertension and/or heavy proteinuria at presentationobservation would tend to misclassify individuals with

the DD genotype, who may be destined to experience (Table 2). The ACE genotype was found to be an inde-pendent risk factor for disease progression upon adjustingdisease progression, to the stable renal function group

(“observation bias”). An additional bias favoring the for blood pressure and proteinuria (incidence relative risk,IRR 5 7.5, 95% CI, 1.3 to 44.9, P 5 0.027). In a relativelyunderrepresentation of patients with the DD genotype

among progressors is suggested by a recent meta-analy- large study by Pei et al, univariate analysis found noassociation between ACE genotype frequencies and dis-sis, which supports an association of the D allele with

myocardial infarction [65]. Since none of the case-control ease progression or proteinuria in patients with pIgAN[63]. Multivariate analysis, however, revealed an adversestudies in pIgAN patients reported the frequency of the

DD genotype as a function of the duration of deteriorat- influence on disease progression of the ACE DD geno-type in patients with the concomitant angiotensinogening renal function or chronic dialysis, an “observation

bias” might eliminate those patients at exceedingly high MM genotype (P 5 0.008; discussed later in this article),after adjusting for age and serum creatinine at presenta-risk for death caused by cardiovascular disease (“com-

peting risks”) or for death caused by the combination tion. Mean arterial blood pressure at presentation orat last follow-up was not found to be an independentof cardiovascular disease and chronic uremia (“additive

risks”) from consideration in the statistical analysis. predictor of disease progression in this study. In a rela-tively large cross-sectional study comparing morphologicAnother important limitation of the majority of these

studies is evident in the comparison of ACE I/D genotype findings and urinary protein excretion at the time ofbiopsy with ACE genotype in children (age #15), Ta-frequencies between progressors and nonprogressors.

Few studies attempted to account for the effects of inde- naka et al reported that the ID/DD genotypes were asso-ciated with indices of chronic injury such as capsularpendent poor prognostic indicators for disease progres-

sion such as hypertension and heavy proteinuria (nor adhesions, glomerulosclerosis, and higher urinary pro-tein excretion [64]. They concluded that the ACE genethe influence of the ACE D allele in subsets defined

by age, sex, race, tobacco use, or presence of familial ID/DD genotypes are associated with pathologic param-eters of known prognostic significance in pIgAN. Finally,hyperlipidemia) by multivariate analysis. Inadequate

sample size may also contribute to inconsistencies in the study by Stratta et al adjusted for a number of clinicalcovariates potentially associated with disease progressionthe results of these studies, particularly if several major

susceptibility loci are hypothesized to determine a multi- such as hypertension, heavy proteinuria (.3 g/day), andACE inhibitor therapy, and found a strong associationfactorial event such as disease progression [65]. A final

critical difference in study methodology relates to the between the DD genotype and 5- and 10-year renal sur-vival rates [65]. There were no significant differencesuse of an insertion-specific PCR primer pair [68] in all

studies beginning in 1996 (the four most recent studies), between the ACE genotype subgroups in age, sex, bloodpressure, proteinuria, ACE inhibitor therapy, and renalwhich eliminated the possibility of preferential amplifi-

cation of the D allele. Thus, in earlier studies, ID hetero- histopathologic lesions at the time of biopsy. The ACEDD genotype was found to be an independent risk factorzygotes may have been mistyped as D homozygotes,

introducing another source of observation bias. In gen- for decreased renal survival (HR 5 3.6; 95% CI, 1.1 to12); the coexistence of heavy proteinuria increased theeral, the use of insertion-specific PCR primers would be

expected to improve the accuracy of genotyping and the risk even further (HR 5 9.16; 95% CI, 1.8 to 15.7).Overall, the type of methodologic weaknesses inher-validity of the study results. Given the potential for bias

inherent in the earlier studies that predate the use of ent in these studies, such as failure to account for in-creased mortality associated with the ACE DD genotypeinsertion-specific PCR primers, it is not unexpected that

two studies reported association [58, 59], while two stud- caused by competing risks, would tend to bias in favorof the null hypothesis for an association between theies reported no association [60, 61], between ACE I/D

genotype and disease progression (Table 2). The remain- ACE D allele and disease progression in pIgAN. Thefinding of a positive association in the majority of recentder of the discussion focuses on a critical appraisal of

the four most recent studies [62–65]. studies that employed improved genotyping methodolo-gies and more strict stratification of patients into sub-The retrospective cohort study by Hunley et al first

examined the association between ACE I/D genotype groups with well-defined covariates of disease progres-sion strongly supports the association between ACE genefrequencies and disease progression in a small group of

patients with pIgAN who presented with normal serum D allele and the progression of established pIgAN. Largeprospective studies of well-defined ethnic/racial cohorts,creatinine at the time of biopsy [62]. They found a trend

toward increased frequency of the DD genotype among however, with sufficient follow-up, adjustment for drop-out because of death from competing risks such as car-progressors (P 5 0.057), which became significant (P 5

0.009) when they excluded patients who presented with diac disease, and randomization according to both treat-

Hsu et al: Genetic factors in IgA nephropathy1824

ment with ACE inhibitors (and other therapeutic agents) servation that deserves further investigation and con-firmation.and the ACE genotype, are required to implicate fully

a role for ACE I/D polymorphisms in the progressionof IgAN.

RACIAL DIFFERENCESRole of other renin-angiotensin system gene Incidence and prevalence figures for pIgAN suggestpolymorphisms in the progression of pIgAN that it is the most common glomerulonephritis worldwide

[8–11]. The true population incidence rates for pIgANPolymorphic variants of genes encoding two additionalare unknown since definitive diagnosis is dependent oncomponents of the renin-angiotensin system, angioten-policies governing renal biopsy, which vary considerablysinogen, and Ang II type 1 receptor (ATR1) have beenfrom country to country. For example, the high incidencereported to be associated with cardiovascular diseasefigures reported for Singaporean Asian males undergoing[69, 70]. A study from the Toronto Glomerulonephritisrenal biopsy likely reflect the practice of mandatory rou-Registry tested the hypothesis that the frequencies oftine screening for urinary abnormalities of all army re-polymorphic variants in the angiotensinogen gene (AGTcruits [71, 72]. In addition, a large proportion of clinicallyM235→T), the ATR1 gene (A1166→C), and the ACE genesilent and undetected IgAN likely exists, as suggested(I/D), may modify renal disease progression in pIgANby a control necropsy study of 200 consecutive patients[63]. Genotype frequencies for each gene were consistentwho died of traumatic injuries and had no clinical historywith Hardy–Weinberg equilibrium in 168 Caucasian pa-of renal disease or other organic disease discovered attients and were similar to those of 100 Caucasian controlthe time of necropsy [73]. Mesangial IgA was the pre-subjects. However, multivariate analysis revealed an as-dominant immunoglobulin found in 4% of these appar-sociation between the AGT MT/TT genotype (comparedently “normal” cases from the general population (5 outwith MM) and a progressive decline in renal functionof 153 males and 3 out of 47 females). Racial differences[odds ratio (OR) 5 2.86, P , 0.001], independent of thein disease incidence, which have been advanced in supporteffects of hypertension. This result would implicate theof a genetic basis for susceptibility to the development ofAGT T235 variant as a risk factor for disease progressionhypertensive nephrosclerosis and diabetic nephropathyin pIgAN. A large proportion of the subjects in theamong African Americans [74], remain largely unknownToronto study were treated with antihypertensive medi-for pIgAN. The one exception is a relatively low preva-cations at the time of diagnosis, thereby limiting thelence of pIgAN and HSP reported for both African andability to detect an association between the AGT T235American Blacks that cannot readily be accounted forvariant and hypertension. Thus, it remains possible thatby differences in the racial distribution of hospital admis-the adverse influence of this allelic variant is mediated

through a direct effect on blood pressure rather than sions or renal biopsies performed [75, 76]. These findingsthrough an independent effect on deterioration of renal provide indirect support for the existence of genetic fac-function. This possibility is supported by an earlier study tors that determine susceptibility to the development ofthat failed to find an association between the AGT T235 pIgAN.variant and disease progression in a subset of patientswithout clinical risk factors for progression (elevated cre-

FAMILIAL IgAN AND ASSOCIATEDatinine, hypertension, and heavy proteinuria) at the timeOVERLAP SYNDROMESof diagnosis [62]. Alternatively, the AGT T235 variant

The strongest evidence from human studies of a rolemay contribute to the progression of renal failure byfor genetic factors in the development and progressionits effects on intrarenal Ang II production, glomerularof IgAN is provided by descriptive reports of familialhypertrophy, and tubular and mesangial cellular prolifer-aggregations of pIgAN. Affected sib pairs and parent-ation and matrix deposition, as reviewed in the previouschild pairs have been reported from a multitude of ethnicsection. Finally, there remains the possibility that thebackgrounds (Table 3). In addition, one large extendedAGT T235 variant may simply be a marker in linkagemultiplex American family of Scottish, Irish, and Englishdisequilibrium with an as of yet unidentified neighboringdescent and a five-generation Australian Aboriginal fam-disease-progression gene [63].ily have been studied in detail (Table 4). There is alsoAlthough univariate analysis failed to demonstrate anreported familial aggregation in multiple, unrelated mul-association between ACE or ATR1 gene polymorphismstiplex American Zuni Indian families from New Mexicoand disease progression in the study of Pei et al, multivar-and Northern Italian families from the Valsaviore valleyiate analysis revealed that the ACE DD genotype ad-of the district of Brescia. Familial syndromes with otherversely affected disease progression only in patients withprimary features but having associated mesangial IgAthe AGT MM genotype (P 5 0.008; Table 2). This sug-deposition as a clinical finding (overlap syndromes) havegests an interaction between specific allelic variants in

two components of the renin-angiotensin system, an ob- also been reported (Table 5).

Hsu et al: Genetic factors in IgA nephropathy 1825

Table 3. Familial primary IgAN simplex kindreds

Ref. Year Relationship of affected probands Race Ethnicity Other affected family membersa

[29] 1978 HLA-identical brothers Caucasian Lebanese ND[30] 1979 HLA-identical brothers Caucasian French ND[79] 1980 HLA-unrelated siblings Caucasian Spain —[80] 1981 HLA-unrelated father and son Caucasian Spain —[81] 1984 HLA-DR4-related siblings Asian Japanese ND[82] 1986 HLA-identical sisters Asian Japanese Father, sister with microhematuria[83] 1988 HLA-identical brother Caucasian American —[86] 1989 34 families with 2–3 affecteds Caucasian French —[84] 1992 2 HLA-identical brothers, Caucasian Yugoslavian ND

1 HLA-unrelated brother[85] 1996 Identical twin sisters Asian Japanese ND[87] 1997 HLA-identical siblings Caucasian Turkish ND

HLA-unrelated brothersHLA-unrelated siblings

ND is not determined.aDisease status determined by screening urinalysis

Familial IgAN simplex kindreds ico in the United States [88]. A high inbreeding coeffi-cient of 0.00797 has been estimated for this entire popula-Table 3 summarizes the reported cases of IgAN amongtion, compared with 0.00084 for the affected membersaffected sib pairs and parent-child pairs [26, 27, 77–85].of the Kentucky kindred [86, 89]. Although 50% of theAlthough HLA identity among the sibs was advancedZuni ESRD population carried a diagnosis of diabetesas an argument in support of early case-control studiesmellitus up to 1989 [90], mesGN was found in 94% (50implicating specific HLA alleles such as DR4 in associa-out of 53) of all Zuni Indian renal biopsies performedtion with IgAN, not all sib pairs were HLA identical, andsince 1971 [88]. An unusual feature of the mesGN amongno single HLA allele was reported to be associated withthe Zuni Indians is the marked familial clustering amongfamilial IgAN in these families. Furthermore, an HLA-cases and the finding of disparate morphologic and im-unrelated brother of affected HLA-identical brothersmunofluorescence (IF) findings on renal biopsy amongfrom Yugoslavia also had IgAN [82].affected family members (Table 4). Mesangial IgA de-

Familial IgAN multiplex kindreds posits were observed in 64% of 49 cases, usually in associ-ation with electron dense mesangial deposits in well-To date, four large extended multiplex families of di-established disease, whereas IgM and IgG deposits wereverse racial/ethnic backgrounds have been described thatpresent in 94 and 45% of cases, respectively; IgA deposi-exhibit familial clustering of IgAN with differing clinicaltion alone was not observed in any of the cases [88].features and at least two different putative genetic inheri-

A predominance of IgA deposition was associatedtance patterns (Table 4). Julian et al first described 14with more severe clinical disease, but not with a greaterpatients with biopsy-proven IgAN from three potentiallyincidence of renal failure in a Zuni biopsy case seriesrelated pedigrees of Scottish, Irish, and English descent[91]. Of the three renal biopsies in family 2 (F-II), onlyfrom central and eastern Kentucky in the United Statesone showed mesGN with deposition of IgA, IgG, and[86]. An additional 17 members of the extended pedigreeIgM, whereas the remainder showed typical focal seg-had clinical glomerulonephritis, and 6 had “chronic ne-mental glomerulosclerosis (FSGS) without Ig deposition.phritis” noted on their death certificates. Assuming thatIn contrast, of the 10 renal biopsies in F-V, 5 showedall 35 patients with evidence of renal disease in this familymesGN with IgA deposition (with or without IgG and/orhad IgAN nephropathy, the presence of male-to-maleIgM), but cases of isolated IgG mesangial deposition andtransmission and an equal distribution of affected malesFSGS were also observed. The light microscopic findingsand females is consistent with the segregation of a simpleof either FSGS or crescents appeared to be associatedautosomal dominant (for example, monogenic) trait forwith the development of nephrotic syndrome and pro-IgAN in this kindred [87]. Notably, no single HLA haplo-gression to ESRD. Isolated IgM mesangial deposits weretype was found to be associated with IgAN in this multi-observed in affected members of F-I and F-VI. Nephroticplex family.syndrome associated with either isolated IgG or IgMSegregation of an apparently autosomal dominantmesangial deposits have been previously reported [92, 93].trait for an unusual variant of mesangial proliferativeMembranoproliferative GN (MPGN) was the biopsy find-glomerulonephritis (mesGN) has been observed among

the Zuni Indians, who comprise an isolated and highly ing in 3 Zuni patients in a reported biopsy series of 44patients with nondiabetic renal disease, and was associ-consanguinous Pueblo community in western New Mex-

Hsu et al: Genetic factors in IgA nephropathy1826

Table 4. Familial IgAN multiplex kindreds

No. of No. of No. of Other biopsy M:F ratio Putative pattern ofRef. Year affectedsa assessedb generations findingsc Race Ethnicity affectedsd inheritanced

[88] 1985 35:14:6 176 8 — Caucasian American 18:17 Autosomal dominant[96] 1987 17:6 49 3 1 non-IgA mesGN, Aboriginal Australian 14:3 X-linked recessive

1 ?HTN/DM[89] 1989 Six families Amerindians Zuni

F-I 8:2:1 23 5 1 IgM 4:5 Autosomal dominantF-II 7:3:0 11 3 2 FSGS 4:3 Autosomal dominantF-III 5:1:1 9 3 1 IgA/FSGS 3:3 Autosomal dominantF-IV 3:2:1 11 4 2 Crescents 3:1 ?F-V 31:10:9 66 4 1 normal, 20:20 Autosomal dominant

1 amyloid/DM,2 IgG, 1 FSGS,1 IgA/FSGS

F-VI 13:7:6 29 5 1 IgM, 2 FSGS, 7:12 ?1 IgG/IgM/MGN

[45] 1990 10 families Caucasian ItalianF-I 5:3 16 3 4:1 X-linked recessiveF-II 8:5 44 7 7:1 X-linked recessiveF-III 5:2 47 12 3:2 ?F-IV 6:4 55 11 5:1 X-linked recessiveF-V 2:2 8 2 2:0 ?F-VI 3:2 5 2 2:1 ?F-VII 3:2 5 2 3:0 ?F-VIII 2:2 6 2 2:0 ?F-IX 2:2 12 2 1:1 ?F-X 3:2 23 4 2:1 ?

aBiopsy-proven or clinical evidence of renal disease:Biopsy-proven disease:Living subjects with unconfirmed but suspected renal disease or deceased subject with“nephritis” noted on their death certificates

bFamily members for which renal status was known from biopsy, urinary abnormalities consistent with glomerulonephritis, or history of renal diseasecPredominant immunoglobulin associated with mesangioproliferative glomerulonephritis in the absence of IgA deposits is indicated. Abbreviations are: FSGS,

focal segmental glomerulosclerosis; DM, diabetic changes; MGN, membranous glomerulonephropathy; GN, glomerulosclerosisdAnalysis included all patients with biopsy-proven, clinical evidence of renal disease, and suspected but unconfirmed renal disease

ated with severe clinical disease and a generally poor this “syndrome” of renal disease exists is most evidentin F-V, in which 10 of 31 patients underwent renal biopsy,outcome [91]. All three cases presented with nephrotic

range proteinuria. Two progressed to ESRD, and one the biopsy findings were variable, but segregation of thedisease trait fits a simple autosomal dominant inheri-patient had three family members with IgA-predominant

mesGN. The variability of IF findings in the reported tance pattern (Table 4).In the case of a large five-generation Australian Ab-Zuni families may be explained if this variant of familial

renal disease manifests early as IF-negative or isolated original kindred exhibiting clustering of IgAN, the pat-tern of inheritance is uncertain. Table 4 summarizes theIgM deposition disease, with IgA, IgG, and complement

deposition variably superimposed later by Ig class switch- original pedigree consisting of a father with biopsy-provenIgAN and a history of significant alcohol use (.160 g/day)ing in response to provoking antigenic stimuli. The ap-

pearance or disappearance of IgA in sequential biopsies but no overt liver disease, a mother with renal disease butbiopsy findings more consistent with chronic ischemicof several subjects has been noted [88, 91]. While mesan-

gial IgA deposits have rarely been reported to decrease changes and diabetic nephropathy, and eight children(3 biopsy-proven IgAN, 1 biopsy-proven IgA-negativeor disappear in serial renal biopsies in patients with pIgAN

[138, 139], the appearance of mesangial IgA deposits mesGN) and seven grandchildren with hematuria [94].Both parents and two of their children had type 2 diabe-in a second renal biopsy specimen has been previously

reported for several patients presenting with proteinuria tes mellitus, but the children had biopsy-proven IgAN.The marked male predominance (only 2 affected daugh-and hematuria in whom the initial biopsy failed to detect

IgA deposits [140, 141]. The additional finding of isolated ters) suggests an X-linked recessive inheritance pattern[87]. However, for this to be the case, the mother (whoFSGS and MPGN among members of Zuni families af-

fected predominantly with mesGN raises the possibility had no biopsy evidence of IgAN) would have to be anasymptomatic carrier of one X-linked disease allele forthat other genetic and/or environmental etiologic factors,

distinct from those responsible for familial IgAN with IgAN. The father was subsequently found to be the patri-arch of a branch of the third generation of a five-genera-the typical clinical features of pIgAN, are involved in

the pathogenesis of familial renal glomerular disease tion extended pedigree. All seven affected members inthe fifth generation who were diagnosed with hematuriaamong the Zuni Indians. That a strong genetic basis for

Hsu et al: Genetic factors in IgA nephropathy 1827

Table 5. Familial syndromes with associated IgAN

Other affected Putative modeRef. Year Syndrome Relationship of affected probands Race Ethnicity family membersa of inheritance

[90] 1984 Overlapping IgA and Brothers Caucasian American ND ?membranous GN

[91] 1988 Spastic paraplegia, Mother with a son and a son-daughter Caucasian British Grandmother and ?bilat. sensorineural sib pair from 2 marriages grandaunt ofhearing loss, mental affected sibsretardation, and IgAN and half-sibs

[92] 1992 Bilat. sensorineural Brother-sister sib pair and a son and Caucasian Spanish Hearing loss in 1st Autosomalhearing loss and IgAN a daughter of the brother two generations; dominant

urine not examined[93] 1989 HSP and IgAN Mother and two daughters Asian Japanese Maternal grandmother ?

with proteinuria[86] 1989 HSP and IgAN 19 families (2–4 affecteds): 13 with two Caucasian French Various ?

1st degree relatives; 1 with two 2nd Algeriandegree relatives; 5 with multigenera- Moroccontional affecteds Madagascar

[79] 1990 HSP and IgAN Father and son Caucasian Spanish ND ?[94] 1998 Thin basement Three sisters Asian Japanese Father with ?

membrane and IgAN hematuria, nl bx.aDisease status of family members determined by screening urinalysis

were #10 years of age. The pattern of disease segregation as the etiology of disease in that family. However, neitherthe high rate of alcohol intake nor prevalence of diabetesin the additional affected members did not clearly follow

any readily identifiable pattern of inheritance. mellitus among the Zuni Indians and the Australian Ab-originals can explain the observed clustering of IgAN inAt least a subset of pIgAN families (F-I, F-II, and

F-IV; Table 4) with ancestral clustering from the isolated the remaining families, since nonalcoholic and nondia-betic family members (especially children) were oftenValsaviore valley of the district of Brescia in Northern

Italy also exhibited a male predominance among affected affected. In addition, the age of onset (age at biopsyvaried over the range of 1 to 66 years in the Zuni families;members and appear to exhibit an X-linked recessive

trait [42]. Ancestral clustering of birthplace in this subset age at diagnosis varied over the range of 5 to 76 yearsin the Italian families) and the time intervals separatingof Valsaviore families is also consistent with a genetic

“founder” effect [42]. F-V through F-IX are two-genera- the apparent onset of disease (8 months to 13 years inthe Italian families), which varied widely among familytional families with a predominance of affected males

but without a sufficient number of family members to members, are not consistent with the effects of a commonenvironmental exposure [42, 88, 94].suggest clearly a discrete pattern of inheritance. How-

ever, F-VII and F-VIII exhibit father-to-son transmis- Finally, a Japanese case-control study of risk factorsfor IgAN, with patient characteristics regarding educa-sions, which would exclude X-linked recessive inheritance

in these families (although matings of a carrier female tional level, food intake, frequency of respiratory infec-tions (colds), exposure to solvents, and family history ofand an affected male can give the appearance of male-

to-male transmission) [87]. Nonetheless, the apparent “chronic GN” obtained through a questionnaire, foundmodestly increased or decreased ORs for various envi-X-linked recessive mode of inheritance in F-I, F-II, and

F-IV suggests that the putative disease locus is distinct ronmental exposures (range 0.42 to 4.73) when univari-ate analysis was performed. However, a family historyfrom those segregating in the Kentucky and Zuni families.

A common environmental exposure was not identified of “chronic GN” in a parent, first-degree relative, orfirst- and second-degree relatives had an OR of 16 (95%in any of these reported familial clusterings of IgAN or

IgA-associated mesGN in the case of the Zuni Indians. CI, 2 to 128), 6.67 (95% CI, 1.84 to 24.2), and 4.8 (95%CI, 1.69 to 13.6), respectively, for the increased risk ofSince not all family members were affected, phenotypic

expression of disease caused by exposure to a presump- developing biopsy-proven IgAN [19]. Despite inherentweaknesses in this type of study, including recall bias,tive environmental stimulus could arise only as a result

of epistatic effects on a disease susceptibility gene. Alco- differences in exposure ascertainment among cases andcontrols, selection bias caused by healthier controls, in-holic liver disease has been reported to be associated

with secondary mesGN with IgA and IgM deposits on creased a error secondary to multiple comparisons, andthe absence of multivariate analysis to identify indepen-renal biopsy [95]. Indeed, both biopsy-proven cases of

IgAN in F-IV had concomitant alcoholic liver disease, dent risk factors, the robust OR for family history of“chronic GN” argues strongly for a genetic componentsuggesting secondary IgAN rather than familial IgAN

Hsu et al: Genetic factors in IgA nephropathy1828

in the pathogenesis of biopsy-proven pIgAN among Jap- strain [105] and the mink affected with Aleutian disease[106]. In addition, the pathologic features of IgAN haveanese.recently been reported in the UG knockout mouse and

Familial overlap syndromes with associated IgAN UG antisense transgenic mouse models [142]. We reviewthe pathologic, immunologic, and immunogenetic char-The study of overlap syndromes has the potential to

provide clues regarding disease-associated genetic loci acterization of hereditary IgAN in the ddY and UG-deficient mouse models.when a consistent chromosomal abnormality results in

the loss of multiple contiguous genes and gives rise to aSpontaneous IgAN in the ddY mousecomplex clinical phenotype. Often, these chromosomal

abnormalities are cytogenetically detectable as deletions, Imai et al reported that ddY mice derived from nonin-bred dd-stock mice known to carry a retrovirus causingtranslocations, monosomies, or polysomies. A paradigm

for this type of contiguous disease genes locus is the mammary gland tumors developed spontaneous IgA-predominant mesangial deposits [105]. A 59-week naturalWAGR region, encoding at least three genes (WT1,

AN2, and 239FB), which are mutated or deleted in the history study of the clinical phenotype and renal pathol-ogy of these animals revealed many of the hallmarks ofsyndrome of Wilms’ tumor, aniridia, genitourinary ab-

normalities, and mental retardation [96–99]. classic pIgAN in humans. Mild to moderate albuminuriawas detectable after 28 weeks and increased progres-IgAN has been reported as a familial overlap syn-

drome in a single family with associated bilateral sensori- sively with age. Mesangial cell proliferation became evi-dent at 16 weeks and progressed to overt mesangioproli-neural hearing loss that is distinguishable from Alport

syndrome (Table 5) [100]. This four-generation Spanish ferative GN by 59 weeks. IgG and IgM mesangialdeposition was predominant until 28 weeks, whereas IgAfamily has three members in the last two generations

with bilateral sensorineural hearing loss and microscopic deposition became predominant at 40 weeks and per-sisted at 59 weeks. Serum polyclonal IgA in ddY micehematuria; two had biopsy-proven IgAN. All four gener-increased by 850% at 59 weeks. Serum creatinine in 48-ations have affected members with a history of hearingweek-old ddY mice was approximately 2.5-fold higherloss, but renal disease status was not ascertained for thethan 6-week-old ddY mice [107]. No antinuclear antibod-majority of family members, as few were screened fories were detected, and amyloid staining was negativeurinary abnormalities. Both hearing loss and nephritis[105]. ddY mice have IgG antihistone autoantibodiesappear to segregate in an autosomal dominant fashionthat react with proteins extracted from renal tissue thatin this family. Genes for autosomal dominant (DFNA11)have been identified as histones H3 and H1 [108]. Theand recessive (DFNB2) nonsydromic sensorineural deaf-pattern of early mesangial proliferative changes, withness have been mapped to chromosome 11q12.3-q21subsequent development of mesangial deposits consisting[101], while a nonsyndromic X-linked form of sensori-predominantly of IgA with associated massive accumula-neural deafness (DFN4) has been mapped to Xp21.2tion of electron dense materials in the mesangium and[102]. These loci may represent candidate regions con-subendothelial space and the development of azotemia,taining IgAN disease/susceptibility genes. An additionalmimics the morphological and clinical features of humanfamily with the syndrome of spastic paraplegia, bilateralIgA nephritis. Hematuria, however, was noticeably ab-sensorineural deafness, mental retardation, and IgANsent. Although ddY mice have been reported to develophas also been reported [103]. However, linkage studiesmammary gland tumors at the age of 40 weeks, whichof genes implicated in various forms of syndromic andcoincides with the onset of high serum polyclonal IgAnonsyndromic spastic paraplegia, sensorineural deaf-levels, none of the animals sacrificed at 59 weeks exhib-ness, and mental retardation do not indicate a commonited mammary tumor formation or any other kinds ofchromosomal region of clustered known disease genestumors. Thus, the mammary gland is unlikely to be thethat could explain this constellation of clinical findingssource of IgA overproduction in ddY mice.[101, 102, 104].

Noting that the incidence of mesangioproliferativeIgAN in the ddY strain was not very high (approximately

ANIMAL MODELS OF IgAN 10%), likely because of its heterogeneous genetic back-The spontaneous development of glomerular or tubu- ground as derived from outbred dd-stock mice, Miya-

lar disease in certain strains of mice and rats has provided waki et al isolated a strain of ddY mice designated HIGAuseful models for the study of the pathogenesis, molecu- (“high IgA”) by selective breeding of individual ddYlar genetics, and response to experimental therapy of mice with the highest serum IgA levels [109]. The re-important renal diseases such as diabetic nephropathy, sulting strain, maintained as brother-sister matings afterhypertensive nephropathy, polycystic kidney disease, F2, is characterized by mean IgA levels at four monthsand lupus nephritis. Two animal models of spontaneous of age that are nine and five times that of the founders

in females and males, respectively. Approximately 50%IgA nephritis have been reported to date: the ddY mouse

Hsu et al: Genetic factors in IgA nephropathy 1829

of HIGA mice exhibit evidence of early onset (at 25 tion, but not the level of serum IgA production. Consis-tent with this hypothesis, the administration of a mono-weeks) moderate to severe mesangial IgA deposition,

and more than 70% show moderate to severe mesangio- clonal antibody to murine CD4 molecules resulted inselective reduction in the number of CD41 T cells andproliferative GN and massive matrix expansion at 40

weeks. The latter was associated with a high incidence a decrease in IgA mesangial deposition, but no differencein the levels of serum IgA, urinary protein, or averageof global glomerulosclerosis in aged mice, a rare finding

in the original ddY strain. Such lesions would be ex- number of intraglomerular cells between treated andcontrol ddY mice [115]. In contrast, treatment with anti-pected to lead to a more marked decline in renal func-

tion. However, assessment of renal function was not re- CD8 monoclonal antibody resulted in a reduction inmesangial hypertrophy in ddY mice, but had no effectported in this study. Notably, the HIGA mice manifested

only mild proteinuria (at most 300 mg/dL) and no hema- on proteinuria or serum IgA levels [116]. These resultssuggest that while CD41 T cells mediate mesangial depo-turia. Approximately 25% of the HIGA mice had only

a mild IgA mesangial deposition at 60 weeks of age, and sition of IgA, CD81 T cells mediate mesangial prolifera-tion and disease progression in ddY mice.serum IgA levels varied widely even after four genera-

tions of selection. The relative contribution of genetic Nagasawa et al have documented a unique TCR gchain in ddY mice by RFLP analysis [117]. Since T cellsfactors to the variation in serum IgA levels was estimated

to be about 40% [109]. These findings suggest that persis- bearing TCR g chain are highly represented in tonsillarand gut-associated lymphoreticular tissues indispensibletent genetic or environmental heterogeneity contributes

to the development of IgAN in HIGA mice. Indeed, for IgA production, a TCR c polymorphism could under-lie abnormal T-cell–regulated IgA production in theseboth ddY and HIGA mice, like many other strains of

experimental mice, are known to harbor retroviral infec- tissues. TCR c polymorphisms have not yet been exam-ined in human IgAN.tions. Takeuchi et al have documented increased mesan-

gial deposition of immune complexes containing retrovi- The disappearance of mesangial IgA deposits afterallogeneic bone marrow transplantation in a patient withral envelope antigen gp70 in ddY in an age-dependent

manner in mice over 24 weeks old, in parallel with the IgAN and chronic myelogenous leukemia prompted Im-asawa et al to investigate a role for hematopoietic stemextent of IgA and IgG mesangial deposition [110]. In

contrast, no gp70 deposits were observed in 12-week- cells in the pathogenesis of IgAN. T-cell–depleted bonemarrow cells from ddY mice were transplanted intoold mice without mesangioproliferative GN.C57BL/6j (B6) mice pretreated with cyclophosphamide.

Immunopathology and immunogenetics of IgAN in B6 recipients of T-cell–depleted allogeneic bone marrowthe ddY mouse cells from ddY mice exhibited significantly higher serum

IgA levels, higher relative proportions of polymeric IgA,Serum IgA levels and the polyclonal IgA response oflymphocytes isolated from Peyer’s patches (but not from and more intense mesangial IgA and C3 deposits com-

pared with B6 mice receiving syngeneic bone marrowspleen) are significantly higher in ddY mice with glomer-ular IgA deposits [111]. These immunologic studies sug- cells from B6 mice [118]. Allogeneic bone marrow trans-

plants from normal mice to ddY mice resulted in de-gest that gut-associated lymphoreticular tissue plays animportant role in the pathogenesis of high IgA serum creased serum levels of polymeric IgA and amelioration

of nephritis. These results suggest that aberrant serumresponse and glomerular IgA deposition in ddY mice.Selective glomerular deposition of more acidic IgA mole- IgA production in ddY mice is quantitatively and qualita-

tively determined by signals originating from bone mar-cules among the polyclonally expanded IgA has also beenobserved in old ddY mice [112]. In addition, the severity row stem cells.of GN is associated with a predominance of the poly-

Role of cytokines that stimulate mesangial cellmeric form of serum IgA. This is consistent with theproliferation and extracellular matrix deposition in thefinding that IgA eluted from the kidneys of patients withpathogenesis of IgAN in the ddY mouseIgAN is predominantly polymeric and anionic (pI 5 4.5

to 6.8) [113]. Platelet-derived growth factor (PDGF), transforminggrowth factor-b (TGF-b), IL-1, IL-6, IL-6 receptorEvidence for the role of aberrant cell-mediated immu-

nity in the pathogenesis of IgAN in ddY mice includes (IL-6R), and the endothelins (ETs) have been implicatedin the mitogenic and fibrogenic response leading to mes-a significant reduction in mesangial IgA deposition fol-

lowing removal of thymus-derived T cells by neonatal angial cell proliferation and accumulation of extracellu-lar matrix (ECM) in IgAN and other glomerulopathiesthymectomy [114]. Since serum IgA levels and the rela-

tive proportion of serum polymeric IgA in thymecto- [119–123]. In ddY mice, PDGF-A chain and PDGF-Bchain messenger RNA (mRNA) levels increase progres-mized animals were not different from control animals,

hypofunction of thymus-derived T cells appears to play sively in an age-dependent fashion and are 2.2-fold and2.4-fold higher at 48 weeks compared with New Zealanda role in determining the extent of mesangial IgA deposi-

Hsu et al: Genetic factors in IgA nephropathy1830

White (NZW) control mice (P , 0.05) [124]; this increase ddY mice compared with age-matched ICR control mice(P , 0.01 for all outcomes) [122]. Treatment of 24-week-is paralleled by a twofold increase in proliferating cell

nuclear antigen (PCNA) mRNA, a marker of cell prolif- old ddY mice (prior to onset of histologic evidence ofIgAN) for 36 weeks with the ET receptor A antagonisteration. Similarly, TGF-b1 and TGF-b2 mRNA levels

increase in an age-dependent manner and are signifi- FR 139317 resulted in a significant attenuation of protein-uria, a reduction in renal impairment, and a reduction ofcantly higher in ddY mice with nephropathy compared

with their littermates without nephropathy [125]. Fur- mesangial hypercellularity compared with age-matchedsaline-treated ddY mice (P , 0.01) and ICR controlthermore, TGF-b2 mRNA levels correlate positively

with the mesangial matrix expansion index (P , 0.05), mice (P , 0.05). Thus, ET family members also appearto play an important pathogenic role in the establishedraising the possibility that TGF-b2 may be an important

mediator of ECM production in the mesangioprolifera- phase of IgAN in ddY mice.tive IgAN affecting ddY mice.

Uteroglobin-deficient mouse modelsMessenger RNA transcripts encoding ECM compo-nents such as type I and type IV collagen a1 chains, A recent series of reports describing two independent

genetic models of uteroglobin (UG) deficiency has unex-laminin A, B1, and B2 chains, fibronectin (FN), andheparan sulfate proteoglycan (HSPG) have also been pectedly revealed a novel role for this cytokine-like se-

cretory protein in the prevention of IgAN in mice [142,reported to be up-regulated in ddY mice compared withage-matched control animals, in parallel with the extent 143]. Both UG2/2 and UG1/2 knockout mice were origi-

nally reported to manifest abnormal glomerular deposi-of IgA and C3 mesangial deposition [126, 127]. However,only TGF-b, type 4 collagen a1 chains, FN, and HSPG tion of FN and collagen, leading to renal failure [143].

Subsequently, both the UG gene-disrupted mice and twohave been demonstrated by IF to be components of theincreased mesangial matrix of ddY mice [128] and HIGA independent lines of UG-antisense transgenic (AS-UG)

mice were found to have high circulating levels of IgA-mice [127]. Given the likely selection of “high secretor”IgA-bearing B-cell clones and evidence of increased mes- FN complexes and abnormal glomerular deposits of IgA

(massive in the case of UG2/2 mice), FN, collagen, andangial ECM production in HIGA mice, Muso et al havespeculated that these phenomena may both be related C3. In addition to these features, which closely mimic

the histopathologic findings in human pIgAN, UG2/2to the observed local up-regulation of TGF-b in themesangium. TGF-b may play a prominent role in pro- animals developed significant microhematuria (.200 red

blood cells per high-power field) and massive proteinuriamoting ECM production while also acting as an immuno-regulatory cytokine in inducing the IgA isotype switching [142, 143]. Notably, UG was shown to prevent both the

formation of IgA-FN complexes in vitro and the glomer-of mature IgA-producing B cells and in suppressing IgGand IgM secretion [127]. ular deposition of exogenously administered IgA in vivo.

Thus, in mice, UG plays an essential role in preventingNephritogenic IgA-immune complexes (IC) havebeen shown to activate human mesangial cells to produce the abnormal glomerular accumulation of IgA [142]. The

role of UG in the prevention of human pIgAN remainscytokines such as IL-1 and IL-6 [129]. The role of IL-1in the progression of established IgAN in ddY mice was to be determined.investigated by administration of IL-1 receptor antago-

Limitations of murine models of IgANnist (IL-1ra), an 18 kD glycoprotein with homology toboth IL-1a and IL-1b, that competitively inhibits the Although study of the ddY and the UG-deficient

mouse models may provide important clues about theeffects of IL-1 receptor in vivo [107]. Forty-week-oldddY mice treated weekly for two months showed reduc- pathogenesis of human pIgAN, several potentially im-

portant phenotypic differences between murine IgANtions in proteinuria, serum creatinine, mesangial cell pro-liferation, and glomerular expression of collagen type in these models and human pIgAN may limit the ability

to infer mechanisms derived from studies of the former,IV, FN, laminin, IL-6, as well as decreased glomerularC3 deposition compared with age-matched saline- to pathogenesis of disease in the latter. The ddY and

HIGA mice exhibit an absence of hematuria and demon-treated ddY mice (P , 0.001 for all outcomes). Thus,IL-1 appears to play a key pathogenic role in the estab- strate only modest proteinuria [105]. In addition, humans

have two IgA subclasses (IgA1 and IgA2), while micelished phase of IgAN in ddY mice.Endothelins may contribute to the progression of glo- have only one IgA subclass. In human pIgAN, the IgA

deposited in the mesangium is almost exclusively IgA1,merular disease either by inducing local vasoconstrictionor by stimulatory effects on mesangial cell proliferation with only infrequent and low-level deposition of IgA2

[140]. In contrast to IgG, IgM, and IgA2, the humanand overproduction of ECM components [130, 131]. Sig-nificantly increased mRNA levels of renal ET-1 (6.6- IgA1 isotype contains a unique hinge region between

the CH1 and CH2 domains of the a1 heavy chain. Recentfold) and ET receptor A (4.8-fold), and B (3.6-fold), butnot ET-3, were observed from 8 to 60 weeks of age in studies have demonstrated that circulating immune com-

Hsu et al: Genetic factors in IgA nephropathy 1831

plexes isolated from the sera of IgAN patients are com- alcohol abuse among the Zuni Indians and AustralianAboriginals, that may contribute to or modify pheno-posed of aberrantly undergalactosylated O-linked gly-

cans in the hinge region and increased amounts of IgG typic expression of an inherited disease trait. Sib pairanalysis [136], affected relative pair analysis [137], and(and less frequently, naturally occurring IgA1) directed

at such galactose-deficient antigenic determinants, as the transmission/dysequilibrium test (TDT) [146] havealso been successfully applied when the underlying ge-compared with healthy controls [144, 145]. The forma-

tion of these immune complexes has been hypothesized netic model is not evident. Such nonparametric modelstake advantage of the simplicity of family structure andto interfere with the hepatic clearance of abnormal IgA1

molecules, leading to their accumulation in the serum well-developed statistical methodologies. Candidate geneapproaches looking at polymorphisms of putative dis-and eventual deposition in the glomerulus [144, 145].

Since the single IgA isotype in mice lacks a hinge region, ease genes or markers in linkage disequilibrium withdisease genes are also best conducted in the setting ofno existing murine model can explain this proposed

mechanism of human IgA1-mediated glomerulopathy. family based studies such as the TDT. Unlike population-based studies such as the case-control comparison, whichNonetheless, since human pIgAN is likely to be a geneti-

cally heterogeneous entity and not all patients with can give rise to spurious associations caused by popula-tion stratification (the existence of multiple populationpIgAN have abnormally O-glycosylated IgA1, ddY and

UG-deficient mice may prove to be useful models for subtypes), the TDT considers parents who are heterozy-gous for an allele associated with disease and evaluatesthe study of alternative mechanisms of IgA-associated

glomerular disease. the transmission of that allele to a single affected off-spring [146]. The TDT can be a more powerful test todetect linkage than the affected sib pair linkage tests,

APPROACHES TO THE IDENTIFICATION OFparticularly for disease alleles that contribute small ge-

IgAN DISEASE/SUSCEPTIBILITY GENESnetic effects.

Evidence derived from the study of the immunogenet- Finally, the ddY and HIGA mouse models of sponta-ics, racial differences in prevalence, and familial occur- neous IgAN and the UG-deficient mouse models haverence of IgAN, as well as the existence of a mouse model yielded a large body of information with possible rele-of spontaneous IgAN, strongly supports the influence of vance to both the immunology and immunogenetics ofgenetic factors in the development and progression of human IgAN and suggest experimental therapeutic mo-the IgA glomerulopathies. This evidence also suggests dalities that deserve further investigation in human clini-a number of feasible approaches for the identification cal trials. The role of UG deficiency in the developmentand positional cloning of putative IgAN disease/suscepti- of IgAN in mice has been convincingly demonstrated,bility genes. Case-control studies are inherently limited but requires confirmation as a mechanism of human dis-in their ability to prove an association between a candi- ease. Identification and positional cloning of disease/date gene and the development or progression of IgAN. susceptibility loci in the ddY and HIGA mouse modelsGiven the successful efforts of a number of investigators have not been reported to date. The outbred nature ofto obtain clinical samples from members of IgAN multi- the ddY mouse strain lends itself to such an analysis.plex families, along with powerful new molecular genetic We speculate that genetic analysis of the ddY mousetools and statistical methods for gene mapping of both will reveal a gene(s) responsible for the development ofMendelian and complex genetic disease traits, we pro- IgAN, while analysis of both the ddY and HIGA micepose that genetic analysis of familial IgAN is the most will provide insight into the genes that mediate progres-promising approach to the identification of IgAN dis- sion of established disease. Such genes likely functionease/susceptibility genes. upstream of the regulatory pathway(s) leading to overex-

Recent success with linkage mapping of independent pression of such immunoregulatory growth factors anddisease genes causing familial FSGS provides a paradigm cytokines as PDGF, TGF-b, IL-1, and ET, which havefor such an approach [132–134]. These studies demon- been shown to mediate the glomerular abnormalitiesstrate the feasibility of disease gene mapping of Mende- characteristic of IgAN in both humans and the ddYlian traits in single large kindreds. Genetic heterogeneity mouse.is unlikely to be a confounder in the analysis of individuallarge kindreds or closely related families assumed to

POSTSCRIPThave a Mendelian disease. In the absence of readily iden-tifiable patterns of inheritance characteristic of complex We are a group of clinical and molecular epidemiolo-

gists, geneticists, and molecular nephrologists who arehuman diseases, genetic epidemiologic methods [135]employing stringent, conservative clinical classification actively recruiting families with familial IgAN in Singa-

pore and the United States for the purpose of gene map-schemes can be used to analyze polygenic interactionsas well as the effect of environmental factors such as ping of IgAN disease/susceptibility genes. We invite phy-

Hsu et al: Genetic factors in IgA nephropathy1832

19. Wakai K, Kawamura T, Matsuo S, Hotta N, Ohno Y: Risksicians from the international medical community whofactors for IgA nephropathy: A case-control study in Japan. Am

are caring for patients from such families to join the In- J Kidney Dis 33:738–745, 199920. Rambausek MH, Waldherr R, Ritz E: Immunogenetic findingsternational Collaborative Group for the Study of IgAN

in glomerulonephritis. Kidney Int 43(Suppl 39):S3–S8, 1993(http://www.nkfs.org/homepage.html). Please contact21. Moore R: MHC gene polymorphism in primary IgA nephropathy.

Stephen I. Hsu, M.D., Ph.D. at the address below. Kidney Int 43(Suppl 39):S9–S12, 199322. Schena FP: Immunogenetic aspects of primary IgA nephropathy.

Kidney Int 48:1998–2013, 1995ACKNOWLEDGMENTS23. MacDonald IM, Dumble LJ, Kincaid-Smith P: HLA and glo-

merulonephritis, in HLA and Disease, Paris, INSERM, 1976, pThe authors wish to thank Alfredo Zannier, Isabelle Contreras, andMichael J. G. Somers for providing bibliographic materials, and Ali 203

24. Berthoux FC, Gagne A, Sabatier JC, Le Petit JC, MarcellinGharavi and Chorh Chuan Tan for insightful discussions.M, Mercier B, Brizard CP: HLA-Bw35 and mesangial IgA glo-merulonephritis. N Engl J Med 298:1034–1035, 1978Reprints requests to Stephen I. Hsu, M.D., Ph.D., Department of

Medicine, National University Hospital, 5 Lower Kent Ridge Road, 25. Noel LH, Deschamps J, Jungers P, Back JF, Busson M, SuetC, Hors J, Daussete J: HLA antigen in three types of glomerulo-Singapore 119074.

E-mail: mdchsus@nus.edu.sg nephritis. Clin Immunol Immunopathol 10:19–23, 197826. Tolkoff-Rubin NE, Cosimi AB, Fuller T, Rubin RH, Colvin

RB: IgA nephropathy in HLA-identical siblings. TransplantationREFERENCES 26:430–433, 1978

27. Sabatier JC, Genin C, Assenat H, Colon S, Ducret F,1. Berger J, Hinglais N: Les depots intercapillaires d’IgA-IgG.Berthoux FC: Mesangial IgA glomerulonephritis in HLA-identi-J Urol Nephrol 74:694–695, 1968cal brothers. Clin Nephrol 11:35–38, 19792. Berger J: IgA glomerular deposits in renal disease. Transplant

28. Hiki Y, Kobayashi Y, Tateno S, Sada M, Kashiwagi N: StrongProc 1:939–944, 1969association of HLA-DR4 with benign IgA nephropathy. Nephron3. Clarkson AR, Woodroffe AJ, Bannister KM, Lomax-Smith32:222–226, 1982JD, Aarons I: The syndrome of IgA nephropathy. Clin Nephrol

29. Moore RH, Hitman GA, Lucas KY, Richards NT, Venning21:7–14, 1984MC, Papiha S, Goodship THJ, Fidler A, Awad J, Festenstein H,4. Galla JH: IgA nephropathy. Kidney Int 47:377–387, 1995Cunningham J, Marsh FP: HLA DQ region gene polymorphism5. Beukhof JR, Ockhuizen T, Halie LM, Westra J, Beelen JM,associated with primary IgA nephropathy. Kidney Int 37:991–995,Donker AJM, Hoedemaeker PJ, van der Hem GK: Subentities1990within adult primary IgA-nephropathy. Clin Nephrol 22:195–199,

30. Levy M, Lesavre P: Genetic factors in IgA nephropathy (Berger’s1984disease). Adv Nephrol 21:23–51, 19926. Emancipator SN, Gallo GR, Lamm ME: IgA Nephropathy: Per-

31. Bontrop RE, Carpenter CB, Walford R, Sekiouchi S, Bignonspectives on pathogenesis and classification. Clin Nephrol 24:161–JD, Cohen D: RFLP standardisation report for DQ Beta/Taq I179, 1985in Immunobiology of HLA, New York, Springer Verlag, 1989,7. Beukhof JR, Kardaun O, Ockhuizen T, van der Hem GK:pp 620–622Kidney survival in IgA nephropathy: Multiple regression analysis

32. Li PKT, Burns AP, So AKL, Pusey CD, Feehally J, Reesof genetically differing subpopulations: Is IgA nephropathy a realAJ: The DQw7 allele at the HLA-DQB locus is associated withdisease entity? Semin Nephrol 7:367–369, 1987susceptibility to IgA nephropathy in Caucasians. Kidney Int8. D’Amico G: The commonest glomerulonephritis in the world:39:961–965, 1991IgA nephropathy. Q J Med 245:709–727, 1987

33. Knight JF, Falk MC, Ng G, Fanning GC, Artlett CM, Roy9. Levy M, Berger J: Worldwide perspective of IgA nephropathy.LP: IgA nephropathy: Immunogenetic studies of Australian pa-Am J Kidney Dis 12:340–347, 1988tients, in Contributions in Nephrology, edited by Clarkson AR,10. Julian BA, Waldo FB, Rifai A, Mestecky J: IgA nephropathy,Woodroffe AJ, Basel, Karger, pp 18–23the most common glomerulonephritis worldwide: A neglected

34. Noreen H, Hors J, Rønningen KS: HLA-DQA1 and -DQB1disease in the United States? Am J Med 84:129–132, 1988polymorphism using polymerase chain reaction (PCR) oligotyp-11. Schena FP: A retrospective analysis of the natural history ofing, in HLA 1991: Proceedings of the Eleventh International Histo-primary IgA nephropathy worldwide. Am J Med 89:209–215, 1990compatibility Workshop and Conference, edited by Tsuji K, Ai-12. Williams DG: Pathogenesis of idiopathic IgA nephropathy. Pedi-zawa M, Sasazuki T, Oxford, Oxford University Press, 1991, ppatr Nephrol 7:303–311, 1993477–48413. Endo Y, Kanbayashi H: Etiology of IgA nephropathy syndrome.

35. Fennessy M, Hitman GA, Moore RH, Metcalfe K, MedcraftPathol Int 44:1–13, 1994J, Sinico RA, Mustonen JT, D’Amico G: HLA-DQ gene poly-14. Consortium TEPKD: The polycystic kidney disease 1 gene en-morphism in primary IgA nephropathy in three European popula-codes a 14 kb transcript and lies within a duplicated region ontions. Kidney Int 49:477–480, 1996chromosome 16. Cell 77:881–894, 1994

36. Raguenes O, Mercier B, Cledes J, Whebe B, Ferec C: HLA class15. Consortium TIPKD: Polycystic kidney disease: The completeII typing and idiopathic IgA nephropathy (IgAN): DQB1*0301, astructure of the PKD1 gene and its protein. Cell 81:289–298, 1995possible marker of unfavorable outcome. Tissue Antigens 45:246–16. Mochizuki T, Wu G, Hayashi T, Xenophontos SL, Veldhuisen249, 1995BJJ, Saris JJ, Reynolds DM, Cai Y, Gabow PA, Pierides A,

37. Martinetti M, Cuccia M, Amoroso A, Borelli I, Cappello N,Kimberling WJ, Breuning MH, Deltas CC, Peters DJ, SomloDaielli C, D’Alfonso S, Dondi E, Ducret F, Dugoujon JM,S: PKD2, a gene for polycystic kidney disease that encodes anHiratska T, Lefranc MP, Luger AM, Mazzola G, Naito S,integral membrane protein. Science 272:1339–1342, 1996Richiardi P, Salvaneschi L, Walker REW, Zhang XM: Idio-17. Hildebrandt F, Otto E, Rensing C, Nothwang HG, Vollmerpathic IgA mesangial nephropathy: An immunogenetic study, inM, Adolphs J, Hanusch H, Brandis M: A novel gene encodingHLA, 1991: Proceedings of the Eleventh International Histocom-an SH3 domain protein is mutated in nephronophthisis type I.patibility Workshop and Conference, edited by Tsuji K, AizawaNat Genet 17:149–153, 1997M, Sasazuki T, Oxford, Oxford University Press, pp 749–75518. Kestila M, Lenkkeri U, Mannikko M, Lamerdin J, McCready

38. Luger AM, Komathireddy G, Walker RE, Pandey JP, HoffmanP, Putaala H, Ruotsalainen V, Morita T, Nissinen M, HervaRW: Molecular and serologic analysis of HLA genes and immuno-R, Kashtan CE, Peltonen L, Holmberg C, Olsen A, Tryggva-globulin allotypes in IgA nephropathy. Autoimmunity 19:1–5,son K: Positionally cloned gene for a novel glomerular protein—1994nephrin—is mutated in congenital nephrotic syndrome. Mol Cell

1:575–582, 1998 39. Berthoux P, El Deeb S, Boulharouz R, Cecillon S, Le Petit

Hsu et al: Genetic factors in IgA nephropathy 1833

J-C, Berthoux FC: HLA molecules and genes in primary IgA 59. Yoshida H, Mitarai T, Kawamura T, Kitajima T, Miyazaki Y,Nagasawa R, Kawaguchi Y, Kubo H, Ichikawa I, Sakai O:nephritis with end-stage renal failure, in IgA Nephropathy: Patho-

genesis and Treatment, edited by Clarkson AR, Woodroffe AJ, Role of the deletion polymorphism of the angiotensin convertingenzyme gene in the progression and therapeutic responsivenessBasel, Karger, 1995, pp 24–28

40. Li PKT, Poon ASY, Lai KN: Molecular genetics of MHC class of IgA nephropathy. J Clin Invest 96:2162–2169, 199560. Fukushima T, Nomura S, Kawai S, Osawa G: ACE genotypeII alleles in Chinese patients with IgA nephropathy. Kidney Int

46:185–190, 1994 and progression of IgA nephropathy. Lancet 346:571–572, 199561. Schmidt S, Stier E, Hartung R, Stein G, Bahnishce J, Wood-41. Jin DK, Kohsaka T, Koo JW, Ha IS, Cheong HI, Choi Y: Com-

plement 4 locus II gene deletion and DQA1*0301 gene: Genetic roffe AJ, Clarkson AR, Ponticelli C, Campise M, Mayer G,Ganten DE, Ritz E: No association of converting enzyme inser-risk factors for IgA nephropathy and Henoch-Schonlein nephritis.

Nephron 73:390–395, 1996 tion/deletion polymorphism with immunoglobulin A glomerulo-nephritis. Am J Kidney Dis 26:727–731, 199542. Scolari F, Amoroso A, Savoldi S, Mazzola G, Prati E, Valz-

orio B, Viola BF, Nicola B, Movilli E, Sandrini M, Campanini 62. Hunley TE, Julian BA, Phillips JA III, Summar ML, YoshidaH, Horn RG, Brown NJ, Fogo A, Ichikawa I, Kon V: Angioten-M, Maiorca R: Familial clustering of IgA nephropathy: Further

evidence in an Italian population. Am J Kidney Dis 33:857–865, sin converting enzyme gene polymorphism: Potential silencer mo-tif and impact on progression in IgA nephropathy. Kidney Int1999

43. Schena FP, Mastrolitti G, Jirillo E, Munno I, Pellegrino N, 49:571–577, 199663. Pei Y, Sholey J, Thai K, Suzuki M, Cattran D: Association ofFracasso AR, Aventaggiato L: Increased production of interleu-

kin-2 and IL-2 receptors in IgA nephropathy. Kidney Int 33:875– angiotensinogen gene T235 variant with progression of immuno-globulin A nephropathy in Caucasian patients. J Clin Invest879, 1989

44. Lai KN, Leung JCK, Li PKT, Lui SF: Cytokine production by 100:814–820, 199764. Tanaka R, Iijma K, Murakami R, Koide M, Nakamura H, Yoshi-peripheral blood mononuclear cells in IgA nephropathy. Clin

Exp Immunol 85:240–245, 1991 kawa N: ACE gene polymorphism in childhood IgA nephropathy:Association with clinicopathologic findings. Am J Kidney Dis45. Nagasawa R, Matsumura O, Maruyama N, Mitarai T, Isoda

K: T-cell receptor beta-chain gene polymorphism and the progno- 31:774–779, 199865. Stratta P, Canavese C, Ciccone C, Barolo S, Dall’Omo AM,sis of IgA nephropathy in Japanese patients. Nephron 70:502–503,

1995 Fasano ME, Mazzola G, Berutti S, Fop F, Curtoni ES, PiccoliG: Angiotensin I-converting enzyme genotype significantly affects46. Li PKT, Poon P, Poon ASY, Szeto CC, Yu AWY, Lai KN:

Association of IgA nephropathy with T-cell receptor constant progression of IgA glomerulonephritis in an Italian population.Am J Kidney Dis 33:1071–1079, 1999alpha chain gene polymorphism. Am J Kidney Dis 30:260–264,

1997 66. Bloem LJ, Manatunga AK, Pratt JH: Racial difference in therelationship of the angiotensin I-converting enzyme gene polymor-47. Jardine AG: Angiotensin II and glomerulonephritis. J Hypertens

13:487–493, 1995 phism to serum angiotensin I-converting enzyme activity. Hyper-tension 27:62–66, 199648. Andreucci VE, Gallioni M, Brancaccio D: ACE inhibitors

and progression of chronic renal insufficiency: A contribution of 67. Fujimura T, Yokota M, Kato S, Hirayam H, Tsunekawa A,Inagaki H, Takatsu F, Nakashima N, Yamada Y: Lack of associa-Italian clinical research. J Nephrol 11:105–109, 1998

49. Fogo A, Yoshida Y, Yared AI, Ichikawa I: Importance of angio- tion of angiotensin converting enzyme gene polymorphism orserum enzyme activity with coronary artery disease in Japanesegenic action of angiotensin II in the glomerular growth of matur-

ing kidneys. Kidney Int 38:1068–1074, 1990 subjects. Am J Hypertens 10:1384–1390, 199768. Lindpainter K, Pfeffer MA, Kreutz R, Stampfer MJ, Grod-50. Johnson RJ, Alpers CE, Yoshimura A, Lombardi D, Prizl

P, Floege U, Schwartz SM: Renal injury from angiotensin II- stein F, Lamotte F, Muring J, Hennekens CH: A prospectiveevaluation of an angiotensin-converting-enzyme gene polymor-mediated hypertension. Hypertension 19:464–474, 1992

51. Wolf G, Neilson EG: Angiotensin II as a renal growth factor. phism and the risk of ischemic heart disease. N Engl J Med 332:706–711, 1995J Am Soc Nephrol 3:1531–1540, 1993

52. Rigat B, Hubert C, Alhenc-Gelas F, Cambien F, Corvol P, 69. Bonnardeaux A, Davies E, Jeunemaitre X, Fery I, Charru A,Clauser E, Tiret L, Cambien F, Corvol P, Soubrier F: Angioten-Soubrier F: An insertion/deletion polymorphism in the angioten-

sin I-converting enzyme gene accounting for half the variance of sin II type 1 receptor gene polymorphisms in human essentialhypertension. Hypertension 24:63–69, 1994serum enzyme levels. J Clin Invest 86:1343–1346, 1990

53. Schunkert H: Polymorphism of the angiotensin-converting en- 70. Katsuya T, Koike G, Yee TW, Sharpe N, Jackson R, NortonR, Horiuchi M, Pratt R, Dzau V, MacMahon S: Association ofzyme gene and cardiovascular disease. J Mol Med 75:867–875,

1997 angiotensinogen gene T235 variant with increased risk of coronaryheart disease. Lancet 345:1600–1603, 199554. Vleming LJ, van der Pijl JW, Lemkes HHPJ, Westendorp RGJ,

Maassen JA, Daha MR, van Es LA, van Kooten C: The DD 71. Sinniah R, Pwee HS, Lim CH: Glomerular lesions in asymptom-atic microscopic hematuria discovered on routine medical exami-genotype of the ACE gene polymorphism is associated with pro-

gression of diabetic nephropathy to end stage renal failure in nation. Clin Nephrol 5:216–228, 197672. Sinniah R, Law CH, Pwee HS: Glomerular lesions in patientsIDDM. Clin Nephrol 51:133–140, 1999

55. van Essen GG, Rensma PL, de Zeeuw D, Sluiter WJ, Scheffer with asymptomatic persistent and orthostatic proteinuria discov-ered on routine medical examination. Clin Nephrol 7:1–14, 1977H, Apperloo AJ, de Jong PE: Association between angiotensin-

converting-enzyme gene polymorphism and failure of renoprotec- 73. Sinniah R: Occurrence of mesangial IgA and IgM deposits in acontrol necropsy population. J Clin Pathol 36:276–279, 1983tive therapy. Lancet 347:94–95, 1996

56. McLaughlin KJ, Harden PN, Ueda S, Boulton-Jones JM, Con- 74. Freedman BI, Soucie JM, McClellan WM: Family history ofend-stage renal disease among incident dialysis patients. J Amnell JMC, Jardine AG: The role of genetic polymorphisms of

angiotensin-converting enzyme in the progression of renal dis- Soc Nephrol 8:1942–1945, 199775. Galla JH, Kohaut EC, Alexander R, Mestecky J: Racial differ-eases. Hypertension 28:912–915, 1996

57. Baboolal K, Ravine D, Daniels J, Williams N, Holmans P, ence in the prevalence of IgA-associated nephropathies. Lancet2:522, 1984Coles GA, Williams JD: Association of the angiotensin I con-

verting enzyme gene deletion polymorphism with early onset 76. Jennette JC, Wall SD, Wilman AS: Low incidence of IgA ne-phropathy in blacks. Kidney Int 28:944–950, 1985of ESRF in PKD1 adult polycystic kidney disease. Kidney Int

52:607–613, 1997 77. Montoliu J, Darnell A, Torras A, Ercilla G, Valles M,Revert J: Familial IgA nephropathy: Report of two cases and58. Harden PN, Geddes C, Rowe PA, McIlroy JH, Boulton-Jones

M, Rodger RSC, Junor BJR, Briggs JD, Connell JMC, Jardine brief review of the literature. Arch Intern Med 140:1374–1375,1980AG: Polymorphisms in angiotensin-converting-enzyme gene and

progression of IgA nephropathy. Lancet 345:1540–1542, 1995 78. Millet VG, Palacios JJN, Ruano RO, Guzman AB, Sarraga

Hsu et al: Genetic factors in IgA nephropathy1834

GU, Carles CP, Montalban MA, Rodicio JL: Familial and he- 101. Tamagawa Y, Kitamura K, Ishida T, Ishikawa K, Tanaka H,reditary mesangial glomerulonephritis with IgA deposits. Med Tsuji S, Nishizawa M: A gene for a dominant form of non-Clin (Barc) 76:1–7, 1981 syndromic sensorineural deafness (DFNA11) maps within the

79. Nomoto Y, Endo M, Miura M, Suga T, Tomino Y, Sakai H, region containing the DFNB2 recessive deafness gene. Hum MolNose Y, Tsuji K: IgA nephropathy associated with HLA-DR4 Genet 5:849–852, 1996antigen. Am J Med 4:184–187, 1984 102. Lalwani AK, Brister JR, Fex J, Grundfast KM, Pikus AT,

80. Miura M, Tomino Y, Yagame M, Suga T, Endoh M, Nomoto Y, Ploplis B, San Agustin J, Skarka H, Wilcox ER: A new nonsyn-Sakai H: Siblings with IgA nephropathy and diffuse proliferative dromic X-linked sensorineural hearing impairment linked toglomerulonephritis (PGN) associated with identical HL-A anti- Xp21.2. Am J Hum Genet 55:685–694, 1994gens. Tokai J Exp Clin Med 11:323–327, 1986 103. Fitzsimmons JS, Watson AR, Mellor D, Guilbert PR: Familial

81. Brensilver JM, Mallat S, Scholes J, McCabe R: Recurrent spastic paraplegia, bilateral sensorineural deafness, and intellec-IgA nephropathy in living-related donor transplantation: Recur- tual retardation associated with progressive nephropathy. J Medrence or transmission of familial disease? Am J Kidney Dis 12:147– Genet 25:168–172, 1988151, 1988 104. Kobayashi H, Garcia CA, Alfonso G, Marks HG, Hoffman

82. Grcevska L, Polenakovi M, Kolevski P, Stavric D: IgA ne- EP: Molecular genetics of familial spastic paraplegia: A multitudephropathy in two HLA-identical brothers and acute, diffuse pro- of responsible genes. J Neurol Sci 137:131–138, 1996liferative glomerulonephritis in the third, HLA-different brother. 105. Imai H, Nakamoto Y, Asakura K, Miki K, Yasuda T, MiuraNephron 61:479–480, 1992 AB: Spontaneous glomerular IgA deposition in ddY mice: An

83. Masuda J, Shiiki H, Fujii Y, Dohi K, Harada A: Identical twin animal model of IgA nephritis. Kidney Int 27:756–761, 1985sisters with IgA nephropathy. Nippon Jinzo Gakkai Shi 38:52–56, 106. Portis JL, Coe JE: Deposition of IgA in renal glomeruli of mink1996 affected with Aleutian disease. Am J Pathol 96:227–236, 1979

84. Levy M: Familial cases of Berger’s disease or of Berger’s disease 107. Chen A, Sheu LF, Chou WY, Tsai SC, Chang DM, Liang SC,and Henoch-Schonlein purpura: French cooperative group of the Lin FG, Lee WH: Interleukin-1 receptor antagonist modulatesSociety of Nephrology. Nephrologie 10:175–182, 1989 the progression of a spontaneously occurring IgA nephropathy

85. Kabasakal C, Keskinoglu A, Mir S, Basdemir G: IgA nephropa- in mice. Am J Kidney Dis 30:693–702, 1997thy occurring in two siblings of three families. Turk J Pediatr 108. Wakui H, Imai H, Nakamoto Y, Kobayashi R, Itoh H, Miura39:395–401, 1997 AB: Antihistone autoantibodies in ddY mice, an animal model

86. Julian BA, Quggins PA, Thompson JS, Woodford SY, Gleason for spontaneous IgA nephritis. Clin Immunol Immunopathol 52:K, Wyatt RJ: Familial IgA nephropathy: Evidence of an inherited 248–256, 1989mechanism of disease. N Engl J Med 312:202–208, 1985 109. Miyawaki S, Muso E, Takeuchi E, Matsushima H, Shibata Y,

87. Speer MC: Basic concepts in genetics, in Approaches to Gene Sasayama S, Yoshida H: Selective breeding for high serum IgAMapping in Complex Human Diseases, edited by Haines JL, levels from noninbred ddY mice: Isolation of a strain with anPericak-Vance MA, New York, Wiley-Liss, 1998, pp 17–52 early onset of glomerular IgA deposition. Nephron 76:201–207,

88. Hoy WE, Smith SM, Hughson MD, Megill DM: Mesangial 1997proliferative glomerulonephritis in Southwestern American Indi- 110. Takeuchi E, Doi T, Shimada T, Muso E, Maruyama N, Yoshidaans. Transplant Proc 21:3909–3912, 1989H: Retroviral gp70 antigen in spontaneous mesangial glomerulo-89. Woolf CM, Dukepoo FC: Hopi Indians, inbreeding, and albinism.nephritis of ddY mice. Kidney Int 35:638–646, 1989Science 164:30–37, 1969

111. Kawaguchi S: Immunopathological study of glomerular IgA de-90. Hoy WE, Megill DM: End-stage renal disease in Southwesternposition in ddY mice. Acta Pathol Jpn 38:1–10, 1988native Americans, with special focus on the Zuni and Navajo

112. Muso E, Yoshida H, Takeuchi E, Shimada T, Yashiro M, Sugi-Indians. Transplant Proc 21:3906–3908, 1989yama T: Pathogenic role of polyclonal and polymeric IgA in a91. Hughson MD, Megill DM, Smith SM, Tung KSK, Miller G,murine model of mesangial proliferative glomerulonephritis withHoy WE: Mesangiopathic glomerulonephritis in Zuni (New Mex-IgA deposition. Clin Exp Immunol 84:459–465, 1991ico) Indians. Arch Pathol Lab Med 113:148–157, 1989

113. Monteiro RC, Halbwachs-Mecarelli L, Roque-Barreira MC,92. Sepandj F, McFarlane C, Trillo A: Nephrotic syndrome second-Noel L-H, Berger J, Lesavre P: Charge and size of mesangialary to primary immunoglobulin-G mesangioproliferative glomer-IgA in IgA nephropathy. Kidney Int 28:666–671, 1985ulonephritis. Nephrol Dial Transplant 13:1889–1890, 1998

114. Nagasawa R, Mitarai T, Utsunomiya Y, Yoshida H, Kitamura93. Cohen AH, Border WA, Glassock RJ: Nephrotic syndromeM, Yamura W, Maruyama N, Isoda K, Sakai O: Neonatalwith glomerular mesangial IgM deposits. Lab Invest 38:610–619,thymectomy diminished renal IgA deposition in IgA nephropa-1978thy-prone ddY mice. Nephron 66:326–332, 199494. O’Connell PJ, Ibels LS, Thomas MA, Harris M, Eckstein RP:

115. Tomino Y, Shimizu M, Koide H, Abe M, Shirai T: Effect ofFamilial IgA nephropathy: A study of renal disease in an Aborigi-monoclonal antibody CD4 on glomerulonephritis of ddY mice,nal family. Aust N Z J Med 17:27–33, 1987a spontaneous animal model of IgA nephropathy. Am J Kidney95. Smith SM, Hoy WE: Frequent association of mesangial glomeru-Dis 21:427–432, 1993lonephritis and alcohol abuse: A study of 3 ethnic groups. Mod

116. Shimamine R, Shibata R, Ozono Y, Harada T, Taguchi T,Pathol 2:138–143, 1989Hara K, Kono S: Anti-CD8 monoclonal antibody protects against96. Gessler M, Simola KO, Bruns GA: Cloning of breakpoints ofspontaneous IgA nephropathy in ddY mice. Nephron 78:310–318,a chromosome translocation identifies the AN2 locus. Science1998244:1575–1578, 1989

117. Nagasawa R, Takahashi T, Mitarai T: Polymophism of T cell97. Gessler M, Poustka A, Cavenee W, Neve RL, Orkin SH, Brunreceptor genes and hyper-production of cytokines: Their possibleGA: Homozygous deletion in Wilms tumours of a zinc-fingercontribution to the pathogenesis of IgA nephropathy in ddYgene identified by chromosome jumping. Nature 343:774–778, 1990mouse. Nippon Rinsho 50:3051–3055, 199298. Pelletier J, Bruening W, Li FP, Haber DA, Glaser T, Housman

118. Imasawa T, Utsunomiya Y, Kawamura T, Nagasawa R, Maru-DE: WT1 mutations contribute to abnormal genital system devel-yama N, Sakai O: Evidence suggesting the involvement of hema-opment and hereditary Wilms’ tumour. Nature 353:431–434, 1991topoietic stem cells in the pathogenesis of IgA nephropathy. Bio-99. Schwartz F, Eisenman R, Knoll J, Gessler M, Gruns G: cDNAchem Biophys Res Commun 249:605–611, 1998sequence, genomic organization, and evolutionary conservation

119. Gesuoldo L, Pinzani M, Floriano JJ, Hassan MO, Nagy NU,of a novel gene from the WAGR region. Genomics 29:526–532,Schena FP, Emancipator SN, Abboud HE: Platelet-derived1995growth factor expression in mesangial proliferative glomerulone-100. Chahin J, Ortiz A, Mendez L, Gallego E, Garcia-Perez J,phritis. Lab Invest 65:160–167, 1991Garcia-Castro G, Julian BA, Egido J: Familial IgA nephropa-

120. Niemir ZI, Stein H, Noronha IL, Kruger C, Andrassy K, Ritzthy associated with bilateral sensorineural deafness. Am J KidneyDis 19:592–596, 1992 E, Waldherr R: PDGF and TGF-b contribute to the natural

Hsu et al: Genetic factors in IgA nephropathy 1835

course of human IgA glomerulonephritis. Kidney Int 48:1530–1541, Slotterbeck BD, Smith AH, Graham FL, Bember MB, QuarlesLD, Pericak-Vance MA, Vance JM: Linkage of a gene causing1995

121. Taniguchi Y, Yorioka N, Oda H, Yamakido M: Platelet-derived familial focal segmental glomerulosclerosis to chromosome 11 andfurther evidence of genetic heterogeneity. Genomics 58:113–120,growth factor, interleukin (IL)-1b, IL-6, IL-6R and tumor necrosis

factor-a in IgA nephropathy. An immunohistochemical study. 1999134. Winn MP, Conlon PJ, Lynn KL, Howell DN, Gross DA, Ro-Nephron 74:652–660, 1996

122. Nakamura T, Ebihara I, Fukui M, Tomino Y, Koide H: Effects gala AR, Smith AH, Graham FL, Bembe M, Quarles LD, Peri-cak-Vance MA, Vance JM: Clinical and genetic heterogeneityof a specific endothelin receptor A antagonist on glomerulone-

phritis of ddY mice with IgA nephropathy. Nephron 72:454–460, in familial focal segmental glomerulosclerosis. Kidney Int 55:1241–1246, 19991996

123. Stein-Oakley AN, Maguire JA, Dowling J, Perry G, Thomson 135. Schildkraut JM: Examining complex genetic interactions, in Ap-proaches to Gene Mapping in Complex Human Diseases, editedNM: Altered expression of fibrogenic growth factors in IgA ne-

phropathy and focal and segmental glomerulosclerosis. Kidney by Haines JL, Pericak-Vance MA, New York, Wiley-Liss, 1998,pp 379–410Int 51:195–204, 1997

124. Tsukasa N, Ebihara I, Nagaoka I, Tomino Y, Koide H: Renal 136. Goldgar DE: Sib pair analysis, in Approaches to Gene Mappingin Complex Human Diseases, edited by Haines JL, Pericak-platelet-derived growth factor gene expression in NZB/W F1 mice

with lupus and ddY mice with IgA nephropathy. Clin Immunol Vance MA, New York, Wiley-Liss, 1998, pp 273–304137. Haines JL: Affected relative pair analysis, in Approaches to GeneImmunopathol 63:173–181, 1992

125. Ye XF, Yorioka N, Oda H, Taniguchi Y, Yamakido M: Role Mapping in Complex Human Diseases, edited by Haines JL,Pericak-Vance MA, New York, Wiley-Liss, 1998, pp 305–322of transforming growth factor-b1 and -b2 in ddY mouse. J Int

Med Res 25:141–154, 1997 138. Cheng IKP, Chan K-W, Chan M-K: Mesangial IgA nephropathywith steroid-responsive nephrotic syndrome: Disappearance of126. Tomino Y, Nakamura T, Ebihara I, Funabiki K, Yaguchi Y,

Shimizu M, Shirato I, Koide H: Altered steady-state levels of mesangial IgA deposits following steroid-induced remission. AmJ Kidney Dis 14:361–364, 1989mRNA coding for extracellular matrices in renal tissues of ddY

mice, an animal model for IgA nephropathy. J Clin Lab Anal 139. Yoshikawa N, Iijima K, Matsuyama S, Suzuki J, Kameda A,Okada S, Nakamura H: Repeat renal biopsy in children with5:106–113, 1991

127. Muso E, Yoshida H, Takeuchi E, Yashiro M, Matsushima H, IgA nephropathy. Clin Nephrol 33:160–167, 1990140. Vangelista A, Frasca GM, Mondini S, Bonomini V: IdiopathicOyama A, Suyama K, Kawamura T, Kamata T, Miyawaki S, Izui

S, Sasayama S: Enhanced production of glomerular extracellular IgA mesangial nephropathy: Immunohistological features. Con-trib Nephrol 40:167–173, 1984matrix in a new mouse strain of high serum IgA ddY mice. Kidney

Int 50:1946–1957, 1996 141. Julian BA, Cannon VR, Waldo FB, Egido J: Macroscopic he-maturia and proteinuria preceding renal IgA deposition in pa-128. Tsushima Y, Tomino Y, Ohmura H, Takeuchi K, Kuramoto

T, Shirato I, Koide H: IgA deposits might not influence the tients with IgA nephropathy. Am J Kidney Dis 17:472–479, 1991142. Zheng F, Kundu GC, Zhang Z, Ward J, Demayo F, Mukherjeeproduction of extracellular matrix in glomeruli of ddY mice, a

spontaneous animal model for IgA nephropathy. Nephron 66:81– AB: Uteroglobulin is essential in preventing immunoglobulin Anephropathy in mice. Nat Med 5:1018–1025, 199986, 1994

129. Chen A, Chen WP, Sheu LF, Lin CY: Pathogenesis of IgA 143. Zhang Z, Kundu GC, Yuan CJ, Ward JM, Lee EJ, Demayo F,Westphal H, Mukherjee AB: Severe fibronectin-deposit renalnephropathy: In vitro activation of human mesangial cells by IgA

immune complex leads to cytokine secretion. J Pathol 173:119– glomerular disease in mice lacking uteroglobin. Science 276:1408–1412, 1997126, 1994

130. Simonson MS, Wann S, Mene P, Dubyac GR, Kester M, Naka- 144. Tomana M, Novak J, Julilan BA, Matousovic K, Konecny K,Mestecky J: Circulating immune complexes in IgA nephropathyzato Y, Sedor JR, Dunn MR: Endothelin stimulates phospholi-

pase C, Na1/H1 exchange, c-fos expression, and mitogenesis in consist of IgA1 with galactose-deficient hinge region and antigly-can antibodies. J Clin Invest 104:73–81, 1999rat mesangial cells. J Clin Invest 83:708–712, 1989

131. Orisio S, Benigni A, Bruzzi I, Corna D, Perico N, Zoja C, 145. Allen AC, Bailey EM, Barratt J, Buck KS, Feehally J: analysisof IgA1 O-glycans in IgA nephropathy by fluorophore-assistedBenatti L, Remuzzi G: Renal endothelin gene expression is

increased in remnant kidney and correlates with disease progres- carbohydrate electrophoresis. J Am Soc Nephrol 10:1763–1771,1999sion. Kidney Int 43:354–358, 1993

132. Mathis BJ, Kim SH, Calabrese MH, Seidman JG, Seidman CE, 146. Spielman RS, McGinnis RE, Ewens WJ: Transmission test forlinkage dysequilibrium: The insulin gene region and insulin-Pollack MR: A locus for inherited focal segmental glomerulo-

sclerosis maps to chromosome 19q13. Kidney Int 53:282–286, 1998 dependent diabetes mellitus (IDDM). Am J Hum Genet 52:506–516, 1993133. Winn MPPJ, Conlon PJKL, Lynn KLDN, Howell DNBD,