Mannose-binding Lectin Gene Polymorphism Associatedwith the Patterns of Glomerular Immune Deposition inIgA NephropathyRujun Gong,1Zhihong Liu2 and Leishi Li2

From the 1Department of Medicine, Nanjing University School of Medicine and 2Research Institute of Nephrology, JinlingHospital, Nanjing University School of Medicine, Nanjing, People’s Republic of China

(Submitted September 17, 2000. Accepted for publication January 12, 2001)

Scand J Urol Nephrol 35: 228–232, 2001

Objective: To elucidate the genetic background underlying the diversity of mesangial immune deposition in IgA nephropathy(IgAN), we investigated the distribution of mannose-binding lectin (MBL) gene codon 54 polymorphism and serum MBLlevels in IgAN patients.Methods: Seventy-seven IgAN patients with glomerular IgA and C3 deposits (Group A) and 70 with glomerular IgA, IgG,IgM, C3 and Clq deposits (Group AGM) were included in the present study. Control group consisted of 140 normal adults.MBL genotypes were investigated by polymerase chain reaction and restriction fragment length polymorphism. Serum MBLlevels with different genotypes were also assayed in some subjects.Results: The variant allele (GAC) was markedly associated with Group AGM (OR = 1.95, 95% C.I.: 1.06–3.58). In bothGroup A and Group AGM, more patients carrying the variant allele had episodes of upper respiratory or gastrointestinalinfections prior to onset or exacerbation of IgAN than wild homozygotes (GGC/GGC). In addition, a signi� cant difference inserum MBL level was also observed between wild homozygotes and heterozygotes (GGC/GAC) (GGC/GGC > GGC/GAC)(p < 0.0001) in all groups, while there was no difference for subjects with the same genotypes among the three groups(p > 0.05). Serum MBL levels of the rare variant homozygotes approached zero.Conclusions: Our � ndings provide evidence that the host defense molecule, MBL, may be involved in the formation of thediversity of glomerular immune deposition in IgAN. Genetic de� ciency of MBL may partially account for abundant immunedeposits in some IgAN patients.

Key words: mannose-binding lectin, gene polymorphism, IgA nephropathy.

Rujun Gong, MD, Department of Medicine, Nanjing University School of Medicine, 22 Hankou Road, Nanjing 210093,People’s Republic of China. E-mail: grb@mes.nju.edu.cn

IgA nephropathy (IgAN) is the most common form ofglomerular diseases world widely. Predominant de-position of IgA in mesangial area of glomeruli has longbeen accepted as the main diagnostic feature of IgAN(1). However, immunopathological studies reveal thatimmune deposits in IgAN are not homogeneous.Isolated IgA deposition occurs in approximately 7–15% of biopsy specimens, while the rest often haveconcomitant deposition of IgG and/or IgM. Accordingto the classes of immunoglobulins appearing inmesangia, the immune deposition in IgAN could bedivided into four patterns: IgA only, IgA ‡ IgG,IgA ‡ IgM and IgA ‡ IgG ‡ IgM (2). The mechanismaccounting for the diversity of immune deposition inIgAN still remains obscure.

Mannose-binding lectin (MBL), or mannose-bindingprotein, is a liver-derived acute phase protein (3),which belongs to the collectin family of proteins with aC-terminal lectin domain and a collagenous backbone

(4). MBL has high af� nity to terminal mannose andN-acetylglucosamine (GlcNAc) moieties (5). Thesestructures have been found common on glycoproteinsof various pathogens, including bacteria, mycobacteria,yeast, fungi and virus. Once binding to these carbohy-drates, MBL has the ability to activate the complementcascades as the key component of the lectin pathway(6). In addition, MBL can also act directly as anopsonin for pathogens through the collagen-likedomain, which can bind to a collectin receptor onmany phagocytes (7). Low serum MBL levels havebeen associated with a common opsonic defect as wellas recurrent infections in both children (8) and adults(9). The molecular basis of this de� ciency was found tobe a single nucleotide polymorphism (SNP) in codon54 (GGC ® GAC) of MBL gene (10).

Episodes of upper respiratory or gastrointestinalinfections usually proceed the onset and exacerbationof IgAN (11). And it has been accepted that immune

Ó 2001 Taylor & Francis. ISSN 0036–5599 Scand J Urol Nephrol 35

ORIGINAL ARTICLE

Scan

d J

Uro

l Nep

hrol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

UB

Kie

l on

10/2

6/14

For

pers

onal

use

onl

y.

deposition in IgAN may result from pathogens thatpenetrate into the body after mucosal infection. In viewof the status of MBL in host defense, we thereforehypothesized that the genetic variance of MBL mightin� uence the patterns of immune deposition in IgAN.In this study, we investigated the distribution of MBLgene polymorphism in IgAN patients with differentpatterns of immune deposition as well as Chinesenormal controls, and selected some subjects for furtherserum MBL assay to elucidate the genetic mechanismunderlying the diversity of immune deposition inIgAN.

MATERIALS AND METHODS

Patients and controls

A total of 628 patients with primary IgAN, whounderwent percutaneous renal biopsy at the JinlingHospital of Nanjing University School of Medicinesince 1992, were investigated. Diagnosis was made bystandard examination of the renal biopsy specimens bylight microscopy and immuno� uorescence. None ofthem had clinical or serological evidence of systemiclupus erythematosus, Henoch-Schonlein purpura, orliver diseases including liver cirrhosis. Among thesepatients, 88 had isolated IgA and C3 deposits inmesangia, 131 had IgA, IgG, C3 deposits, 94 hadIgA, IgM, C3 deposits and the remaining 315 hadmesangial deposition of IgA, IgG, IgM, C3 and C1q.Seventy-seven of the 88 patients with isolated deposi-tion of IgA and C3, and 70 of the 315 patients with IgA,IgG, IgM, C3 and C1q deposits, who were seenconsecutively in this hospital, were included in thepresent study. The relevant clinical parameters ascer-tained at the time of presentation included gender, age,duration of the disease, episodes of upper respiratory orgastrointestinal infections prior to the onset or exacer-bation of the disease, recurrent or isolated grosshematuria, and hypertension characterized by therequirement of regular medication or a blood pressure>140/90 mmHg on at least three occasions. Thelaboratory assessment at the time of biopsy includedroutine urinalysis, 24-h urine protein excretion, andserum creatinine concentrations. Serum immunoglo-bulin (IgA, IgG, IgM) levels were also measured.

One hundred and forty unrelated Chinese healthyadults were randomly recruited as normal controls. Allpatients and normal controls gave informed consentbefore entry into the study.

MBL genotyping

Genomic DNA was extracted from peripheral bloodleukocytes by standard techniques. MBL codon 54gene polymorphism was analyzed by polymerase chain

reaction and restriction fragment length polymorphism(PCR–RFLP). The MBL exon 1 sequences wereampli� ed by PCR. As primers a pair of oligonucleo-tides 5’-GTAGGACAGAGGGCATGCTC-3’ (upperstrand) and 5’-CAGGCAGTTTCCTCTGGAAGG-3’(lower strand) were used as previously described(10). PCR was performed in the presence of 1.5 mMMgCl2 in the following 35 cycles: 1 min at 94°C, 1 minat 58°C and 1 min at 72°C. This was preceded by a3 min denaturation step at 94°C and followed by 7 minextension at 72°C. The PCR product was digested withthe Ban I restriction enzyme (Promega, USA). Ban Irecognizes the sequence 5’-GGPyPuCC-3’, which ispresent in the wild allele but absent from the variantallele. Subsequent RFLP was analyzed by electrophor-esis of the digested products on 2.5% agarose gels andstained with ethidium bromide for visualization withultra violet light. Results of RFLP were represented byGGC for wild allele (328 bp) and GAC for variantallele (246 bp and 82 bp) (Fig. 1).

Assay for serum MBL

Taking into account that MBL is a serum acute phaseprotein, which may increase upon stress state, and thatsome immunosuppressants, e.g. dexamethasone, etc.,have been shown to decrease the serum MBL level(12), we selected 58 IgAN patients (30 from Group Aand 28 from Group AGM) and 32 normal subjects whohad no experience of stress states including infection,fever or trauma, and had received no immunosuppres-sive treatment in the latest 4 weeks prior to MBL assay.Serum samples were stored at ¡80°C until analysis.

Fig. 1. Restriction fragment length polymorphism (RFLP) ofpolymerase chain reaction (PCR) products of exon 1 sequences ofmannose-binding lectin (MBL) gene digested with Ban I. Lanes 1,4 and 6 contain GGC/GGC homozygous samples; lanes 3 and 5contain GGC/GAC heterozygous samples; lane 2 contains a GAC/GAC homozygous sample; lane M contains a 100 bp DNA ladder(Promega, USA).

Scand J Urol Nephrol 35

MBL gene polymorphism and IgAN 229

Scan

d J

Uro

l Nep

hrol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

UB

Kie

l on

10/2

6/14

For

pers

onal

use

onl

y.

Serum MBL concentrations were assessed by asandwich enzyme linked immunosorbent assay (ELI-SA) described previously (13), which was based on acommercially available MBL ELISA kit (StatensSerum Institut, Denmark).

Statistics

Genotype frequencies and allele frequencies werecalculated. Results from the controls and the testgroups were compared using Chi-squared analysis(with Yates’s correction) for statistical signi� cance.In each group, the observed distribution of the threegenotypes perfectly � t the expectations based onHardy-Weinberg equilibrium analysis. Odds ratios(ORs) were calculated for signi� cant associations andexpressed with 95% con� dence intervals (95% C.I.).An odds ratio was considered to be signi� cant if thelower 95% con� dence limit did not fall below l.0.Clinical or laboratory measurement and enumerationdata between different genotypes were compared by t-test and Chi-squared test, respectively. Unpaired two-sample t-tests (two-tailed) were used to compare theserum MBL levels between different genotypes orbetween different groups. The difference was consid-ered to be signi� cant if the p value was less than 0.05.Measurement data were presented as mean values§ SD.

RESULTS

Distribution of MBL gene polymorphism in IgANpatients with different patterns of immune deposition.

Genotype frequencies and allele frequencies observedin all groups were summarized in Table I. Thegenotype frequency of GGC/GAC heterozygotes inGroup AGM was signi� cantly higher compared withGroup A (41.4% vs 19.5%, p < 0.01) and normalsubjects (41.4% vs 26.4%, p < 0.05), while the fre-quency of GGC/GGC homozygotes was signi� cantlylower than Group A (55.7% vs 76.6%, p < 0.01) andnormal controls (55.7% vs 72.1%, p < 0.05). The allelefrequency of GAC was also higher in Group AGM than

Group A (0.24 vs 0.14, p < 0.05) and normal subjects(0.24 vs 0.15, p < 0.05). The variant allele (GAC) wasclosely associated with Group AGM (OR = 1.95, 95%C.I.: 1.06–3.58), whereas, there was no difference inthe distribution of MBL genotypes between Group Aand normal controls.

Correlation of MBL genotypes to clinical andlaboratory � ndings

Compared with patients with GGC/GGC genotype,more patients carrying the variant allele had episodesof upper respiratory or gastrointestinal infections priorto onset or exacerbation of IgAN in both Group A andGroup AGM (p < 0.05). No other features werededuced in connection with MBL genotypes fromthese data (Table II).

Effect of MBL gene variation on serum MBL levels inIgAN patients and normal controls

The serum MBL concentrations for GGC/GGC andGGC/GAC genotypes in the three groups were shownin Table III. A signi� cant difference (p < 0.0001) inserum MBL levels was observed between GGC/GGCgenotype and GGC/GAC genotype in all groups, withthe MBL levels of the latter genotype being 1/6–1/8 ofthe former one. Since the variant homozygotes (GAC/GAC) were very rare, there was no statistical sig-ni� cance. But the MBL concentrations for all GAC/GAC homozygotes were extremely low (5 mg/ml and7 mg/ml for 2 subjects in normal control group; 3 mg/mland 22 mg/ml for 2 patients in Group A; 14 mg/ml forone patient in Group AGM) and much lower than themean concentrations for the other two genotypes. Nodifference in the serum MBL levels was found for thesame genotypes among the three groups (p > 0.1).

DISCUSSION

IgAN is not a single disease entity, but consists of avariety of diseases with the same immunopathologicalcharacteristics (11). IgAN patients vary not only inclinical manifestation and prognosis but also in the

Table I. Distribution of MBL gene polymorphism in IgAN patients with different patterns of glomerular immune deposition

Genotype frequency Allele frequency

n GGC/GGC GGC/GAC GAC/GAC GGC GAC

Normal controls 140 101 (72.1%) 37 (26.4%) 2 (1.4%) 0.854 0.146IgAN patients

Group A 77 59 (76.6%) 15 (19.5%) 3 (3.9%) 0.864 0.136Group AGM 70 39 (55.7%)* 29 (41.4%)* 2 (2.9%) 0.764 0.236†

* p < 0.05, Group AGM vs Normal controls; p < 0.01, Group AGM vs Group A; † p < 0.05, Group AGM vs Normal controls; p < 0.05,Group AGM vs Group A.

Scand J Urol Nephrol 35

230 R. Gong et al.

Scan

d J

Uro

l Nep

hrol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

UB

Kie

l on

10/2

6/14

For

pers

onal

use

onl

y.

patterns of mesangial immune deposition (2). But whatcauses the diversity of the patterns of immunedeposition in IgAN still remains unanswered. Indivi-dual genetic background and environmental context aretwo important determinants of the clinical presentationof human diseases. Genetic dissection of complex traitsmay help � nd associations between genotypes andclinical phenotypes which could imply a role forpredicting the disease process (14).

MBL, a novel host defense molecule, is an importantconstituent of human innate immunity (15). It has beenproposed that MBL might be of particular importanceeither in the promotion of complement mediatedopsonization after the decay of passively acquiredmaternal antibody opsonins and before the maturationof the infant’s own antibody repertoire, or at the time ofprimary contact with a pathogen as a � rst line of

defense (16). In the present study, we found that thepatterns of mesangial immune deposition were closelyassociated with the codon 54 polymorphism. Hetero-zygous genotype (GGC/GAC) was more frequent inGroup AGM compared with Group A, while there wasno difference in the distribution of MBL gene poly-morphism between Group A and normal controls.IgAN with abundant immune deposition signi� cantlycorrelated to the variant allele (GAC). Besides, in bothGroup A and Group AGM, patients carrying the variantallele were more likely to have episodes of upperrespiratory or gut infections prior to onset or exacer-bation of IgAN than those without this allele. In all thethree groups, the mean serum MBL levels of the variantallele carriers were markedly lower than GGC homo-zygotes. The mean MBL level of the heterozygoteswere 1/6–1/8 of that of the GGC homozygotes and the

Table II. Correlation of MBL genotypes to clinical and laboratory � ndings in IgAN patients with different patterns of immune deposition

Group A (n = 77) Group AGM (n = 70)

Items GGC/GGC (n = 59)GGC/GAC orGAC/GAC (n = 18) p value GGC/GGC (n = 39)

GGC/GAC orGAC/GAC (n = 31) p value

Gender F: 22; M: 37 F: 8; M: 10 >0.05 F: 13; M: 26 F: 11; M: 20 >0.05Age (years) 29.48 § 5.66 28.64 § 6.22 >0.05 32.42 § 6.81 31.85 § 5.54 >0.05Duration of disease

(months)20.30 § 10.22 25.66 § 12.50 >0.05 23.55 § 8.36 25.42 § 10.80 >0.05

Clinical onset rGH: 10; iGH: 9;NS: 14; Uab: 21;HT: 5; AGN: 8;MP: 0

rGH: 4; iGH: 2;NS: 3;Uab: 6;HT: 2; AGN: 3;MP: 1

>0.05 rGH: 9; iGH: 7;NS: 8; Uab: 12;HT: 2; AGN: 6;MP: 1

rGH: 8; iGH: 2;NS: 7; Uab: 10;HT: 3; AGN: 5;MP: 1

>0.05

Episodes of infections 10 8 <0.05* 7 15 <0.01**Scr (mmol/l) 125.48 § 35.20 137.25 § 50.84 >0.05 125.35 § 22.60 142.50 § 38.55 >0.05IgA (g/l) 2.22 § 0.45 2.28 § 1.02 >0.05 2.15 § 0.62 2.30 § 0.75 >0.05IgG (g/l) 9.25 § 2.82 8.90 § 3.15 >0.05 9.18 § 3.05 9.22 § 3.25 >0.05IgM (g/l) 1.10 § 0.30 1.18 § 0.48 >0.05 1.15 § 0.55 1.05 § 0.42 >0.05C3 (g/l) 1.40 § 0.45 1.30 § 0.50 >0.05 1.45 § 0.38 1.36 § 0.52 >0.05C4 (g/l) 0.50 § 0.18 0.45 § 0.16 >0.05 0.62 § 0.20 0.53 § 0.15 >0.05Urinary protein (g/

24 h)1.20 § 0.64 1.45 § 0.75 >0.05 1.25 § 0.33 1.35 § 0.42 >0.05

Hematuria (£106,RBC/ml)

2.82 § 1.55 3.31 § 2.18 >0.05 3.31 § 2.18 3.35 § 2.20 >0.05

Measurement data were shown as means § SD; * p < 0.05, variant allele carriers vs wild homozygotes in Group A; ** p < 0.01, variantallele carriers vs wild homozygotes in Group AGM.

Abbreviations: F, female; M, male; rGH, recurrent gross hematuria; iGH, isolated gross hematuria; NS, nephrotic syndrome; Uab,abnormal urinalysis; HT, hypertension; AGN, acute glomerulonephritis syndrome; MP, non-nephrotic massive proteinuria.

Table III. Serum mannose binding lectin concentrations (mg/l) in healthy controls and IgAN patients with different patterns of immunedeposition

GGC/GGC GGC/GAC

n x § s n x § s p value

Normal controls 19 1711 § 177 11 293 § 63 <0.0001IgAN patients <0.0001

Group A 16 1622 § 194 12 271 § 46 <0.0001Group AGM 17 1512 § 211 10 210 § 42 <0.0001

Data are presented as means § SD.

Scand J Urol Nephrol 35

MBL gene polymorphism and IgAN 231

Scan

d J

Uro

l Nep

hrol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

UB

Kie

l on

10/2

6/14

For

pers

onal

use

onl

y.

MBL levels of most variant homozygotes were lessthan 10 mg/l and approached zero, which were totally inline with the results reported by Garred et al. (17) andSuper et al. (8) in normal population and in patientswith infectious diseases.

Based on the results of the present study, we canelucidate the in� uence of MBL gene variation on thepatterns of mesangial immune deposition in IgAN.Since the variant allele occurs much more frequently inGroup AGM than Group A, there will be more patientswith low serum MBL level and impaired � rst-linedefense. Pathogens, such as microbial, environmentalantigens, are more likely to invade circulation ofpatients in Group AGM. More patients in GroupAGM present episodes of infections prior to onset orrecurrence of IgAN. Pathogens simultaneously stimu-late the acquired immune response to generate admixedantibodies that form circulating immune complexeswhich further deposit in the mesangial area and lead toabundant immune deposition.

In 1990, Emancipator even applied the notion of“mucosal tolerance” or “oral tolerance” to explain theadmixed deposition of IgG/IgM in IgAN (18). Heascribed the generation of abundant immune com-plexes containing IgG and IgM, but mostly IgAantibodies to two routes. First, large amounts of antigenmight penetrate the mucosa prior to the development of“mucosal tolerance”. Second, the host might fail toestablish immunoglobulin isotype switching and “mu-cosal tolerance” fully. However, this explanation haslong been lacking of suf� cient convincing data, eitherimmunological or pathological, to con� rm the exis-tence of mucosal tolerance de� ciency in IgAN. In thisstudy, we found that genetic de� ciency of MBL, animportant component of the innate immunity, was alsoassociated with the concomitant deposition of IgG andIgM in IgAN as well as the clinical episodes ofinfections prior to the onset or recurrence of IgAN. We� rst unveiled the genetic background underlying thediverse patterns of glomerular immune deposition inIgAN and at the same time provided evidence that thehost � rst-line defense of the innate immunity may beinvolved in the formation of this diversity.

ACKNOWLEDGEMENTS

We thank X. Zhou and Q. Qi for assistance in preparation of themanuscript and H. Chen who helped with the examination of renalbiopsies by light microscopy and immuno� uorescence.

REFERENCES

1. Berger J. IgA glomerular deposits in renal disease.Transplant Proc 1969; 1: 939–44.

2. Liu ZH, Li LS. Polymorphism in IgA nephropathy.Nephrology 1997; 3: 63–6.

3. Thiel S, Holmskov U, Hviid L, Laursen SB, JenseniusJC. The concentration of the C-type lectin, mannan-binding protein, in human plasma increases during anacute phase response. Clin Exp Immunol 1992: 90; 31–5.

4. Kurata H, Sannoh T, Kozutsumi Y, Yokota Y, KawasakiT. Structure and function of mannan-binding proteinsisolated from human liver and serum. J Biochem 1994;115: 1148–54.

5. Holmskov U, Malhotra R, Sim RB, Jensenius JC.Collectins: collagenous C-type lectins of the innateimmune defense system. Immunol Today 1994; 15: 67–73.

6. Matsushita M. The lectin pathway of the complementsystem. Microbiol Immunol 1996; 40: 887–93.

7. Kuhlman M, Joiner K, Ezekowitz RAB. The humanmannose-binding protein functions as an opsonin. J ExpMed 1989; 169: 1733–45.

8. Super M, Thiel S, Lu J, Levinsky RJ, Turner MW.Association of low level of mannan-binding protein with acommon defect of opsonization. Lancet 1989; ii: 1236–9.

9. Summer� eld JA, Ryder S, Sumiya M, Thursz M,Gorchein A, Monteil MA, et al. Mannose-bindingprotein gene mutation associated with unusual andsevere infections in adults. Lancet 1995; 345: 886–9.

10. Sumiya M, Super M, Tabona P, Levinsky RJ, Arai T,Turner MW, et al. Molecular basis of opsonic defect inimmunode� cient children. Lancet 1991; 337: 1569–70.

11. Galla JH. IgA nephropathy. Kidney Int 1995; 47: 377–87.

12. Arai T, Tabona P, Summer� eld JA. Human mannose-binding protein is regulated by interleukins, dexametha-sone and heat shock. Q J Med 1993; 86: 575–82.

13. Valdimarsson H, Stefansson M, Vikingsdottir T, ArasonGJ, Koch C, Thiel S, et al. Reconstitution of opsonizingactivity by infusion of mannan-binding lectin(MBL) toMBL-de� cient humans. Scand J Immunol 1998; 48:116–23.

14. Lander ES, Schork NJ. Genetic dissection of complextraits. Science 1994; 265: 2037–48.

15. Turner MW. Mannose-binding lectin: the pluripotentmolecule of the innate immune system. Immunol Today1996; 17: 532–40.

16. Turner MW. Mannose-binding lectin (MBL) in healthand disease. Immunobiol 1998; 199: 327–39.

17. Garred P, Thiel S, Madsen HO, et al. Gene frequency andpartial protein characterization of an allelic variant ofmannan binding protein associated with low serumconcentration. Clin Exp Immunol 1992; 90: 517–21.

18. Emancipator SN. Immunoregulatory factors in thepathogenesis of IgA nephropathy. Kidney Int 1990; 38:1216–29.

Scand J Urol Nephrol 35

232 R. Gong et al.

Scan

d J

Uro

l Nep

hrol

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

UB

Kie

l on

10/2

6/14

For

pers

onal

use

onl

y.