- Email: [email protected]
IgA synthesis by lymphocytes from patients with IgA nephropathy and their relatives
Kidney International, Vol. 29 (/986), pp. 1229—1233
IgA synthesis by lymphocytes from patients with IgA nephropathy and their relatives F. BRYSON WALDO, LINDA BEISCHEL, and CLARK D. WEST The Department of Pediatrics, University of Alabama at Birmingham, Birmingham, Alabama, and Children's Hospital Research Foundation, Department of Pediatrics, University of Cincinnati, Cincinnati, Ohio, USA
IgA synthesis by lymphocytes from patients with IgA nephropathy and their relatives. To investigate the hypothesis that a predisposition to IgA nephropathy (IgAN) is linked to the major histocompatibility complex
(MHC) and associated with poorly regulated IgA synthesis, we performed HLA typing and lymphocyte cultures on patients with IgAN and their relatives. Ninetten of 22 patients had elevated culture supernatant IgA concentrations (620 vs. 154 ng/2 x 106 cells, P = 0.007). Supernatent IgG and 1gM were normal. No HLA antigen occurred with increased frequency in patients. There was an increased incidence of homozygous null C4 alleles in patients (P < 0.01). In families, six of 11
mothers, six of 12 fathers, and seven of 15 siblings had elevated supernatant IgA concentrations. There was no segregation of abnormal IgA production with any HLA antigen or parental haplotype. The data confirm elevated in vitro IgA production by lymphocytes from patients with IgAN, but do not support a linkage with the MHC. The increased incidence of homozygous null C4 alleles may result from functional
lating immune complexes are frequently detected [5, 6]. Several investigators have demonstrated elevated in vitro IgA production by cultured mononuclear cells [7—9]. It appears that this elevated production is the result of both in vivo B cell activation
and a selective deficiency of IgA-specific T suppressor cells. Some investigators have identified increased numbers of IgAbearing peripheral blood and tonsillar lymphocytes [10, 11]. The disease has also been associated with different HLA antigens in different populations [12, 15]. These associations are proposed
to be related to immune response genes coded within or near the HLA-D region. Sporadic reports of familial cases of IgA nephropathy have also appeared [16—19].
elevated IgA production without an obvious inheritance pattern suggest that shared environmental factors may be important in the development of IgAN.
From these observations, we proposed that individuals (occasionally related) would be predisposed to develop IgA nephropathy if they inherited certain HLA antigens which might be associated with elevated in vivo (and in vitro) IgA production. To investigate this hypothesis, HLA typing, C4 allotyping and mononuclear cell cultures were performed on patients with IgA nephropathy and their first degree family members.
IgA deposition in the renal mesangium is associated with systemic diseases such as systemic lupus erythematosus, subacute bacterial endocarditis, chronic cirrhosis and Henoch— Schonlein purpura (HSP), but also occurs in the absence of any
Methods
differences in C4 A and B gene products. The familial clustering of
systemic symptoms. Berger and Hinglais first described a primary glomerulonephritis associated with IgA deposition in 1968 [1]. Their patients had a focal proliferative glomerulonephritis associated with diffuse mesangial immunofluorescence
for IgA, IgG and C3. Subsequent work has demonstrated diffuse mesangial, paramesangial and occasional empimembranous immune deposits [2]. Clinically these patients have recur-
rent gross, and persistent microscopic, hematuria. Although renal function is initially normal, long—term follow—up suggests
that up to 30% of patients may progress to end—stage renal disease (ESRD) over 10 to 20 years [3]. Despite many attractive
theories, the precise etiology of IgA nephropathy remains unknown, and no effective therapy has been described [4]. A variety of interesting laboratory abnormalities have been identified in patients with IgA nephropathy. Serum IgA levels are elevated in approximately 30%, and IgA containing circu-
Received for publication September 25, 1984, and in revised forms August 28 and November 7, 1985
© 1986 by the International Society of Nephrology
Patients Twenty-seven patients, followed at the Children's Hospital Medical Center, Cincinnati, Ohio, had a diagnosis of primary IgA nephropathy confirmed by renal biopsy. Secondary forms of IgA nephropathy were excluded. The patients ranged in age from four to 21 years (mean = 13.4). The patients were white except for one black male. There were 19 boys and eight girls. No patient was significantly azotemic (serum creatinine >1.5 mg/dliter) at the time of study. One patient had received a renal allograft and was receiving alternate day prednisone (approximately 1 mg/kg) at the time of study. This patient had a rapidly progressive course of IgA nephropathy and no clinical evidence of recurrence at the time of culture. Twenty—two patients had cell culture studies and HLA typing performed. C4 allotyping was performed on all 27 patients. Eleven mothers, 12 fathers, and 15 siblings (II boys and 4 girls) of the patients were studied. All relatives were well at the time of study with normal urinalyses. Writen informed consent
was obtained from parent, patients or family members as appropriate. 1229
Waldo et a!
1230
Mononuclear cell separation and culture
IgA
1,0001—
0
Heparinized whole blood was obtained sterilely and the mononuclear cell fraction separated by centrifugation on ficoll
hypaque. The cells were harvested, washed three times in Hank's balanced salt solution (HBSS, Gibco, Grand Island, New York, USA) and centrifuged through heat inactivated fetal
0
0
immunoglobulin. Cells were adjusted to a concentration of 2 x 106/mliter in RPM! 1640 (Gibco) with 15% FCS, 10,000 units!
0 0) 0 C
1,000
noglobulin production. Cell free supernatants were stored at —20°C until assayed.
0
1,000
--
E E
c'o
0
Co 100
0
C
0
5,
0
CS
C 5,
a (5 CS
U
c0 0
o
100
t O100L
Controls Patients
°
0
0 C
90% cell viability were used. Cyclohexamide l0- M was added to four cultures and caused 80 to 100% suppression of immu-
8
0
>
calf serum (FCS, Gibco) to remove any passively absorbed
mliter penicillin and 10 g/mliter streptomycin added. Cells were cultured at a density of 4 x 106 cells/2 mliter in 24 well microtiter plates (Costar, Boston, Massachusetts, USA) in a humidified 5% CO2 incubator at 37°C for 12 days. The cell viability was determined by trypan blue dye exclusion. Only supernatants from uncontaminated cultures with greater than
gM
lgG 10,0001-
(5
0 0 0 0
0
-00 0 0 0 0 8
0
0
0 0
F
0 0 10
Controls Patients
0 0
Controls Patients
Immunoglobulin measurement: IgG and IgA in the cultured
Fig. 1. Unfractionated mononuclear cell culture production of immu-
fluorescent assay, !mmunoflor (Bio—Rad, Richmond, Califor-
1gM (P = 0.05) determined. Twenty—eight healthy controls and 22 patients with IgA nephropathy were studied. Data are expressed as
supernatants were measured using a solid phase immuno- noglobulin; isolated mononuclear cells were cultured for 12 days and and nia, USA). The manufacturer's instructions were followed except for a modification to permit extremely low level !gA and IgG measurements. The standards were first diluted as directed and then further diluted 1:5 before use in the assay. The diluted standard at 0.5 mliter or 0.5 mliter undiluted culture supernatant
the supernalan! concentrations of IgA (P = 0.007), JgG (P = 0.05)
means 2 SD (illustrated by lines) for controls and as mean values for patients.
was mixed with a 1 mliter suspension of "!mmunobeads"
HLA and C4 typing conjugated with either anti-!gA or anti-IgG. The procedure was HLA typing was performed by the clinical laboratory at then completed according to manufacturer's instructions. Lin- Hoxworth Blood Bank, Cincinnati, Ohio. C4 aflotyping was ear standard curves from 165 to 1,145 nglmliter for IgA, and 166 performed by the method of Awdeh and Alper [22]. Serum to 1,152 ng/mliter for lgG were obtained and unknown concen- immunoglobulin measurements were performed using an Auto trations determined from these. Supernatants which exceeded ICS nephelometer (Beckman, Laiolla, California, USA). the standard range were diluted and rerun. Statistics. Statistical comparisons were made using the chi 1gM was measured by an ELISA technique. Briefly, anti- squared test, analysis of variance, and multiple linear regression. human 1gM was raised in goats, the serum absorbed to monoResults specificity, and an Ig fraction obtained by octonoic acid precipitation [20]. the Ig fraction was affinity purified and biotinylated Mononuclear cell cultures fr immunoglobulin production as previously described [211. Polysytrene microtiter plates Unfractionated mononuclear cell preparations were from 30 (Limbro, McLean, Virginia, USA) were coated overnight at 4°C
with 0.1 mliter of affinity purified anti-human 1gM at 10 g/ healthy controls. Eight children (ages two to 20, mean = 11 mliter in 0.06 M pH 9 carbonate buffer. The wells were then years) and 22 adults (ages 26 to 59, mean = 35) were cultured blocked for 30 mm at 37°C with 0.2% bovine serum albumin in for 12 days in the absence of mitogenic stimulation. The mean carbonate buffer. Wells were washed with PBS. 0.01 NI EDTA, supernatant concentrations of lgA, lgG. and 1gM were 154, and 0.05% Tween 20 (wash buffer), and 0.1 mliter aliquots of 1181, and 142 ng/2 x 106 cells respectively (Fig. 1). There was undiluted culture supernatants or 1gM standards were added. A no overall relationship between age and supernatant !gA concurve was constructed from a calibrator serum of known 1gM centration (r 0.32, P > 0.05). Mean supernatant !gA in the concentration serially diluted in culture media, and incubated children was 165 ng/2 x 106 cells, and for the adults, 149 ng/2 x 106 cells (P > 0.05). Kinetic analysis of !gA production was for one hour at 37°C. The wells were washed and then incubated performed on four separate individuals. There was a plateau of for two hr at 37°C with 0.1 mliter of a 1:800 dilution of net immunoglobulin production at eight to ten days regardless biotinylated, affinity purified anti-human 1gM. After washing, 0.1 mliter of avidin—peroxidase (Sigma, St. Louis, Missouri, USA) was added for 30 mm at room temperature, and the wells
were then washed and developed with o-phenylenediamine (Sigma). A log linear curve from 50 to 1000 ng!mliter was obtained and unknowns determined from this. Values from wells coated with RPM! 1640, 15% FCS were not significantly different from those coated with wash buffer alone.
of the final value. Four individuals had repeat cultures during an
episode of upper respiratory or gastrointestinal infection and the net production of all immunoglobulin classes remained within the normal range. Mononuclear cell cultures in patients with IgA nephropathy As illustrated in Figure 1, mononuclear cells were isolated from 22 patients with IgA nephropathy. Nineteen of 22 patients
1231
Familial IgA synthesis in IgA nephropathy
with IgA nephropathy had supernatant IgA concentrations greater than two SD above the mean of control. The mean
10,000
patient supernatant IgA concentration was 620 compared to the control of 133 ng/2 x 106 cells (P = 0.007). Supernatant concentrations of IgG and 1gM were elevated in seven and three
0
patients with IgA nephropathy, respectively. However, the mean supernatant IgG and 1gM concentrations in patients were not significantly different from controls (P > 0.05). One of the IgA nephropathy patients with a normal superna-
() (0
0
tant IgA concentration had a renal allograft. The two other patients with IgA nephropathy and normal supernatant IgA concentration could not be distinguished from the entire patient group in any other way. In particular, there was no difference in their duration of disease, frequency of gross hematuria, qualitative degree of proteinuria or serum creatinine level. Serum IgA was elevated for age in II of 22 patients with IgA nephropathy. There was no correlation between the serum level of IgA and the supernatant IgA concentrations (r = 0,29, P > 0.05). Eight patients had significantly elevated supernatant IgA concentrations but normal serum IgA levels.
HLA and C4 allotyping HLA typing was performed on 27 patients with IgA nephrop-
athy, and no antigen was present in significantly increased frequency. In particular, B12, Bw35, and DR4, antigens previ-
ously associated with IgA nephropathy, were observed at frequencies of 0.23, 0,2 and 0.4, respectively, with expected
4 C
1,000
a) a) C a)
o
8 0 0 8
(a
o 8
0
cc
0 0
0
100
Controls
8
0 8
0
0
o L alrnxo
0
8 8
0 0
0
a
cc
Patients
Mothers
0
0
Fathers
Siblings
cc
0
Fig. 2. Unfractionated mononuclear cell culture production of IgA in
frequencies of 0.28, 0.15 and 0.28 (uncorrected P > 0.1). C4 is first degree relatives of patients with IgA nephropathy. coded within the HLA complex at 4 alleles (two A and two B).
Awdeh et al have defined "extended" HLA haplotypes by allotyping C4, C2, factor B and glyoxalase [23]. Patients with
IgA nephropathy have been reported to have an increased
production whereas in the W family, IgA production was
elevated only in the parents. Thus, no correlation between an incidence of C4 null genes at either the A or B locus [24]. No HLA antigen or parental haplotype and increased IgA producextended haplotype has been reported in increased incidence in tion could be made. IgA nephropathy [25]. Homozygous null C4B genes were found in five of 27 patients. This differs significantly from the observed
Discussion
parent affected, three had both, and four had neither parent affected. HLA typing was performed on three complete fami-
between IgA nephropathy and HLA Bw35 in France and Australia, HLA DR4 in France and Japan, and HLA B12 in
lies. The H family had two affected members, one with typical IgA nephropathy and the other with typical HSP. In this family, elevated IgA was present in the father and several siblings but not in the mother. The segregation of elevated immunoglobulin production in siblings followed no HLA antigen or parental
Florida [12—15]. No association with any HLA antigen, how-
incidence in the Cincinnati area of 3 out of 75 (x2 = 5.47, P < Studies of in vitro immunoglobulin production by mononu0.01) [26]. Homozygous null B genes have reported to occur in clear cells from patients with IgA nephropathy have suggested the extended haplotype B12, DR4, B5, C2 C, C4 A3, C4BqO. abnormal immunoregulation of IgA synthesis. Studies by others Based on HLA and C4 typing alone, two of our patients may using slightly different techniques have yielded similar, but not have had this haplotype. This incidence of two out of 27 identical, results to those of Cosio and Beale [9, 27, 28]. Beale patients would not differ from the observed frequency in the et al studied patients with HSP, an IgA related renal disease, Boston area of 10 out of 295 (x2 = I.!). and found increased spontaneous (non-mitogen stimulated) production of IgA by unfractionated mononuclear cells [7]. B Family studies cells isolated from these patients, in the absence of T cells Immunoglobulin production by mononuclear cells from first produced significant amounts of IgA, suggesting that in vivo B degree relatives of patients with IgA nephropathy was mea- cell activation had occurred. Allogenic mixing experiments sured. As illustrated in Figure 2, there was elevated supernatant suggested that defective T suppressor cell activity was also IgA concentrations in six of 11 mothers, six of 12 fathers and present. Cosio et al studied patients with IgA nephropathy and seven of 15 siblings. No significant increase in the concentra- demonstrated increased spontaneous production of IgA by tions of IgG or 1gM in supernatants was observed (not shown). unfractionated mononuclear cells [8]. Elevated parental IgA production varied; four families had one Previous studies have also identified significant associations
ever, was identified in a group of patients from the United Kingdom [29]. An increased incidence of homozygous null A or
B genes for the fourth component of complement, which lies within the major histocompatibility complex (MHC), has also haplotype. In the L family, all members had elevated IgA been reported in patients with IgA nephropathy, or HSP [24].
1232
Waldo et a!
Furthermore, a number of families in which more than one pursue this possibility. Mononuclear cells obtained from apmember have IgA nephropathy have been reported in one of proximately half of the parents and siblings of patients with IgA these, the asymptomatic brother of a patient with the disease nephropathy produced abnormally large amounts of IgA in was found to have glomerular IgA deposits [12, 16—19]. Prelim- culture. However, no clear Mendelian pattern of inheritance inary evidence of elevated IgA synthesis by cultured lympho- was observed. Families were observed with one, both or cytes from family members has also appeared [30]. Recently a neither parent involved. Three kinships had complete HLA large kindred from Western Kentucky has been reported [31]. typing, and elevated IgA production did not segregate with any This kindred had an unusually high incidence of IgA nephrop- measured HLA haplotype. The data clearly indicate familial athy, was not excessively inbred, and pedigree analysis sug- clustering of elevated IgA production, but point to no simple gested an autosomal dominant inheritance mode with poor pattern of inheritance. Although elevated IgA production is associated with IgA penetrance. From these observations, we proposed that a gene in or near nephropathy, its pathogenetic importance remains unclear. the MHC conferred a predisposition to IgA nephropathy, and Elevated IgA production alone would seem inadequate to was associated with abnormal immunoregulation and elevated produce disease since half the family members studied had this in vitro IgA production. Any HLA allotype in linkage dis- phenotype while reported familial cases of IgA nephropathy equilibrium with this proposed gene would occur with increased were relatively rare. It is possible that healthy family members frequency in patients with IgA nephropathy. Furthermore, with elevated IgA production actually have occult IgA nephrophealthy first degree relatives having this HLA allotype would athy and would have positive mesangial immunofluorescence also have the abnormal immunoregulatory gene and elevated in for IgA if biopsied. Such a case of occult IgA nephropathy has been previously described [12]. Until a reliable, non-invasive vitro IgA production. Unfractionated mononuclear cells from our patients with IgA marker of IgA nephropathy is established this question will nephropathy cultured without mitogenic stimulation produced remain unanswered.
Since the data do not establish a genetic pattern for the significantly more IgA than controls, confirming the presence of abnormal immunoregulation in patients with IgA nephropathy. observed familial clustering of elevated IgA production, the Interestingly, no significant correlation was found between the possibility of environmental factors must be considered. Viral
serum level of IgA and the quantity of IgA produced in cell agents, such as Epstein—Barr virus and other members of the culture. The serum concentration of IgA is determined not only by its rate of synthesis, but by its rate of clearance, which could be elevated in patients with IgA nephropathy. Furthermore, the rate of in vitro synthesis does not necessarily reflect that which occurs in vivo. In our patients with IgA nephropathy, contrary to previous results, no HLA antigen was observed in increased frequency. However, there was an increased incidence of homozygous null
herpes family, are well known B cell mitogens and could served as shared environmental exposures. Dietary substances such as bacterial lipopolysaccharide could also act as local mitogens for
B genes for C4. The reason for disease associations with the MHC are unclear. Awdeh have established extended haplotypes within the MHC defined by HLA typing, combined with C2, C4, B and glyoxalase typing [23]. These extended haplotypes are conserved in a population by crossover suppression and male transmission bias. It has been proposed that disease susceptibile genes within this complex are also conserved and help maintain linkage between diseases and the HLA antigens. However, the extended haplotype containing homozygous null
tion in patients and their family members.
T-helper/inducer cells in the gut associated lymphoid tissue. Studies of unrelated individuals living in the same household as
those patients with IgA nephropathy could help clarify this issue. Finally, it will be important to determine if the same cellular mechanisms are responsible for elevated IgA produc-
Acknowledgments This work was supported in part by National Research Service Award AM 07051 (to F. B. Waldo), and by grant RR 00123 from the
General Clinical Research Centers Branch, both from the National Institutes of Health. Thanks to Ms. Karen Kilgo and Kathie Thomas for secretarial assistance.
B alleles was not observed at an increased incidence in patients, Reprint requests to Dr. F. B. Waldo, The Children's Hospital, /600 Seventh Avenue South, Birmingham, Alabama 35233, USA. making this explanation unlikely in the present series. The association between C4 B null alleles and IgA nephropathy may result instead from the difference in hemolytic activity References of C4, A and B gene products. Most of the hemolytic C4 activity
resides in C4, B gene products. Patients who are null at both C4, B loci have relatively low hemolytic C4 levels [26]. The classical pathway of complement should be important in the processing and clearance of IgA-lgG immune complexes. A relative deficiency in C4 hemolytic activity could, thus, be one of several predisposing factors in the development of IgAN. The above data confirm the association between elevated in
vitro production of IgA and nephropathy, but provide no evidence to link elevated in vitro IgA production with a particular HLA antigen. Abnormal IgA immunoregulation could, however, be inherited at a locus not linked to the MHC and, thus, occur within families. Family studies were done to
1. BERGER J, HINGLAIS N: Les depots intercapillaires d'IgA-IgG. J Urol Nephrol 74:694—695, 1968
2. CLARKSON AR, SEYMOUR AE, THOMPSON AJ, HAYNES WDG, CIIAN YL, JACKSON B: IgA nephropathy: A syndrome of uniform
morphology, diverse clinical features and uncertain prognosis. Clin Nephrol 8:459—471, 1977
3. DROZ D: Natural history of primary glomerulonephritis with mesangial deposits of IgA. Contr Nephrol 8:335—340. 1977
4. Eco J, SANCHO J, BLASCO R, RIvE F, HERNANDO L: Immunopathogenetic aspects of IgA nephropathy. Adv Nephrol 12:103—137, 1983
5. WOODROFFE AJ, GORMLY AA, MCKENZIE PE, W00TON AM, THOMPSON AJ, SEYMOUR AE, CLARKSON AR: Immunologic studies in IgA nephropathy. Kidney mt 18:366—374, 1980
Familial IgA synthesis in IgA nephropathy 6. LESAVRE PH, DIGEON M, BACH iF: Analysis of circulating IgA and
detection of immune complexes in primary IgA nephropathy. C/in
Exp Immunol 48:6l—69, 1982 7. BEALE MG, NASH GS, BERTOVICH J, MACDERMOTT RP: Similar
disturbances in B cell activity and regulatory T cell function in Henoch—Schonlein purpura and systemic lupus erythematosus. J Immunol 128:486—491, 1982
AD, CONLEY ME, MICHAEL AF: Immune regulation of immunoglobulin production in IgA nephrop-
8. Costo FG, LANN 5,
FOLAMI
athy. C/in Immunol Immunopatho/ 23:430—436, 1982 9. Ecwo J, BLASCO R, SANCHO J: Increased rates of polymeric IgA synthesis by circulating lymphoid cells in IgA mesangial glomerulonephritis. C/in Exp Immunol (in press) 10. NOMOTO Y, SAKAI H, ARIMORI S: Increase of IgA bearing lymphocytes in peripheral blood from patient with IgA nephropathy. Am J C/in Path 71:158—160, 1979 11. BENE MC, FAURE G, HURAULT DE LIGNY B, KESSLER M, DUHEILLE J:
Immunoglobulin A nephropathy quantitative im-
1233
the possible role of HLA antigens. Transplantation 29:505—506, 1980 20. STEINBUCK M, AUDRAN R: The isolation of IgG from mammalian
sera with the aid of caproic acid. Arch Biochem Biophys
134:279—284, 1969 21. BAYER EA, SKUTELSKY E, WILCHEK M: The avidin—biotin com-
plex in affinity cytochemistry. Methods in Enzymes 62:308—315, 1979 22. AWDEH ZL, ALPER CA: Inherited structural polymorphism of the
fourth component of human complement. Proc Nat/ Acad Sci USA
77:3576—3580, 1980 23. AWDEH ZL, RAUM D, YuRRI5 EJ, ALPER CA: Extended HLA/
complement allele haplotypes: Evidence for T/t like complex in
man. Proc Nail Acad Sci USA 80:259—261, 1983 24. MCLEAN RH, WYATT RJ, JuLIAN BA: Inherited deficiency of the fourth component of complement (C4) in IgA nephropathy, Henoch—
Schonlein purpura and recurrent hematuria. (abstract) Kidney Int 21:202, 1982
inversion of the immunoglobulin A versus immunoglobulin G
25. ARNAIZ—VLLLENA A, GONZALO A, REGUERRO JR, VICARI0 JL, ORTUNO J: Extended HLA haplotypes in Berger's disease. C/in
secreting cell balance. J C/in Invest 71:1342—1347, 1983 AV, MAHONEY JJ, FULLER TJ: Higher prevalence of
Nephrology 22:320—321, 1984 26. WELCH TR, BIESCHEL L, BERRY A, FORRISTAL J, WEST CD: The
munohistomorphometry of the tonsillar plasma cells evidences an 12. RICHMAN
HLA-B12 in patients with IgA nephropathy. (abstract) Ann Intern effect of null C4 alleles on complement function. C/in Immuno/ Med 90:201, 1979 Immunopathol 34:316—325, 1985 13. NOEL H, DESCAMPS B, JUNGERS P: HLA antigens in three types of 27. ROTHSCHILD E, CHATENOUD L: T cell subset modulation of glomerulonephritis. C/in Immuno/ lmmunopatho/ 10:19—23, 1978 immunoglobulin production in IgA nephropathy and membranous 14. FAUCHET R, LE POGAMP P, GENETET B, CHEVET D, GUEGUEN M, glomerulonephritis. Kidney mt 25:557—564, 1984 28. EGID0 J, BLASCO R, SANCHO J, LOZANO L: T cell dysfunctions in SIMON P, RAMEE MP, CARTIER F: HLA-DR4 antigen and IgA nephropathy. Tissue Antigens 16:405—410, 1980 IgA nephropathy: specific abnormalities in the regulation of IgA 15. KASHWABARA H, SHISHIDO Y, YOKOYAMA T, MIGAJUINA T: HLA synthesis. C/in Immunol Immunopathol 25:201—212, 1983 29. RASHID Mt, PAPIHA SS, AGRUYANNIS B, MORLEY AR, WARD in IgA nephropathy. Tissue Antigens 16:411-412, 1980 16. M0NT0LIu J, DARNELL A, TORRAS A, ERCILLA G, VALLES M, MK, ROBERTS DF, KERR DNS: The associations of HLA and the REVERT L: IgA nephropathy. Arch Intern Med 140:1374—1375, 1980 genetic markers with glomerulonephritis. Hum Genet 62:38—44, 17. TOLKOFF—RUBIN NE, CO5INI AB, FULLER T, RUBIN RH, COLVIN 1983 RB: IgA nephropathy in HLA-identical siblings. Transplantation 30. EGIDO J, BLASCO R, SANCHO J, LOZANO L, GUTIERREZ—MILLET V: Immunological studies in a familial IgA nephropathy. C/in Exp 26:430—433, 1978 18. SABATIER JC, GENIN C, ASSENOT H, COLON S, DUCRET F, Immuno/ 54:532—538, 1983 BERTHOUX FC: IgA glomerulonephritis in HLA identical brothers. 31. JULIAN BA, QUIGGINS PA, THOMPSON iS, WOODFORD SY, GLEASON K, WYATT Ri: Familial IgA nephropathy evidence of an C/in Nephro/ 11:35—38, 1979 19. KATZ A, KARANICOLAS 5, FALK J: Family studies in IgA nephritis: inherited mechanism of disease. N Engi J Med 312:202—208, 1985
IgA synthesis by lymphocytes from patients with IgA nephropathy and their relatives F. BRYSON WALDO, LINDA BEISCHEL, and CLARK D. WEST The Department of Pediatrics, University of Alabama at Birmingham, Birmingham, Alabama, and Children's Hospital Research Foundation, Department of Pediatrics, University of Cincinnati, Cincinnati, Ohio, USA
IgA synthesis by lymphocytes from patients with IgA nephropathy and their relatives. To investigate the hypothesis that a predisposition to IgA nephropathy (IgAN) is linked to the major histocompatibility complex
(MHC) and associated with poorly regulated IgA synthesis, we performed HLA typing and lymphocyte cultures on patients with IgAN and their relatives. Ninetten of 22 patients had elevated culture supernatant IgA concentrations (620 vs. 154 ng/2 x 106 cells, P = 0.007). Supernatent IgG and 1gM were normal. No HLA antigen occurred with increased frequency in patients. There was an increased incidence of homozygous null C4 alleles in patients (P < 0.01). In families, six of 11
mothers, six of 12 fathers, and seven of 15 siblings had elevated supernatant IgA concentrations. There was no segregation of abnormal IgA production with any HLA antigen or parental haplotype. The data confirm elevated in vitro IgA production by lymphocytes from patients with IgAN, but do not support a linkage with the MHC. The increased incidence of homozygous null C4 alleles may result from functional
lating immune complexes are frequently detected [5, 6]. Several investigators have demonstrated elevated in vitro IgA production by cultured mononuclear cells [7—9]. It appears that this elevated production is the result of both in vivo B cell activation
and a selective deficiency of IgA-specific T suppressor cells. Some investigators have identified increased numbers of IgAbearing peripheral blood and tonsillar lymphocytes [10, 11]. The disease has also been associated with different HLA antigens in different populations [12, 15]. These associations are proposed
to be related to immune response genes coded within or near the HLA-D region. Sporadic reports of familial cases of IgA nephropathy have also appeared [16—19].
elevated IgA production without an obvious inheritance pattern suggest that shared environmental factors may be important in the development of IgAN.
From these observations, we proposed that individuals (occasionally related) would be predisposed to develop IgA nephropathy if they inherited certain HLA antigens which might be associated with elevated in vivo (and in vitro) IgA production. To investigate this hypothesis, HLA typing, C4 allotyping and mononuclear cell cultures were performed on patients with IgA nephropathy and their first degree family members.
IgA deposition in the renal mesangium is associated with systemic diseases such as systemic lupus erythematosus, subacute bacterial endocarditis, chronic cirrhosis and Henoch— Schonlein purpura (HSP), but also occurs in the absence of any
Methods
differences in C4 A and B gene products. The familial clustering of
systemic symptoms. Berger and Hinglais first described a primary glomerulonephritis associated with IgA deposition in 1968 [1]. Their patients had a focal proliferative glomerulonephritis associated with diffuse mesangial immunofluorescence
for IgA, IgG and C3. Subsequent work has demonstrated diffuse mesangial, paramesangial and occasional empimembranous immune deposits [2]. Clinically these patients have recur-
rent gross, and persistent microscopic, hematuria. Although renal function is initially normal, long—term follow—up suggests
that up to 30% of patients may progress to end—stage renal disease (ESRD) over 10 to 20 years [3]. Despite many attractive
theories, the precise etiology of IgA nephropathy remains unknown, and no effective therapy has been described [4]. A variety of interesting laboratory abnormalities have been identified in patients with IgA nephropathy. Serum IgA levels are elevated in approximately 30%, and IgA containing circu-
Received for publication September 25, 1984, and in revised forms August 28 and November 7, 1985
© 1986 by the International Society of Nephrology
Patients Twenty-seven patients, followed at the Children's Hospital Medical Center, Cincinnati, Ohio, had a diagnosis of primary IgA nephropathy confirmed by renal biopsy. Secondary forms of IgA nephropathy were excluded. The patients ranged in age from four to 21 years (mean = 13.4). The patients were white except for one black male. There were 19 boys and eight girls. No patient was significantly azotemic (serum creatinine >1.5 mg/dliter) at the time of study. One patient had received a renal allograft and was receiving alternate day prednisone (approximately 1 mg/kg) at the time of study. This patient had a rapidly progressive course of IgA nephropathy and no clinical evidence of recurrence at the time of culture. Twenty—two patients had cell culture studies and HLA typing performed. C4 allotyping was performed on all 27 patients. Eleven mothers, 12 fathers, and 15 siblings (II boys and 4 girls) of the patients were studied. All relatives were well at the time of study with normal urinalyses. Writen informed consent
was obtained from parent, patients or family members as appropriate. 1229
Waldo et a!
1230
Mononuclear cell separation and culture
IgA
1,0001—
0
Heparinized whole blood was obtained sterilely and the mononuclear cell fraction separated by centrifugation on ficoll
hypaque. The cells were harvested, washed three times in Hank's balanced salt solution (HBSS, Gibco, Grand Island, New York, USA) and centrifuged through heat inactivated fetal
0
0
immunoglobulin. Cells were adjusted to a concentration of 2 x 106/mliter in RPM! 1640 (Gibco) with 15% FCS, 10,000 units!
0 0) 0 C
1,000
noglobulin production. Cell free supernatants were stored at —20°C until assayed.
0
1,000
--
E E
c'o
0
Co 100
0
C
0
5,
0
CS
C 5,
a (5 CS
U
c0 0
o
100
t O100L
Controls Patients
°
0
0 C
90% cell viability were used. Cyclohexamide l0- M was added to four cultures and caused 80 to 100% suppression of immu-
8
0
>
calf serum (FCS, Gibco) to remove any passively absorbed
mliter penicillin and 10 g/mliter streptomycin added. Cells were cultured at a density of 4 x 106 cells/2 mliter in 24 well microtiter plates (Costar, Boston, Massachusetts, USA) in a humidified 5% CO2 incubator at 37°C for 12 days. The cell viability was determined by trypan blue dye exclusion. Only supernatants from uncontaminated cultures with greater than
gM
lgG 10,0001-
(5
0 0 0 0
0
-00 0 0 0 0 8
0
0
0 0
F
0 0 10
Controls Patients
0 0
Controls Patients
Immunoglobulin measurement: IgG and IgA in the cultured
Fig. 1. Unfractionated mononuclear cell culture production of immu-
fluorescent assay, !mmunoflor (Bio—Rad, Richmond, Califor-
1gM (P = 0.05) determined. Twenty—eight healthy controls and 22 patients with IgA nephropathy were studied. Data are expressed as
supernatants were measured using a solid phase immuno- noglobulin; isolated mononuclear cells were cultured for 12 days and and nia, USA). The manufacturer's instructions were followed except for a modification to permit extremely low level !gA and IgG measurements. The standards were first diluted as directed and then further diluted 1:5 before use in the assay. The diluted standard at 0.5 mliter or 0.5 mliter undiluted culture supernatant
the supernalan! concentrations of IgA (P = 0.007), JgG (P = 0.05)
means 2 SD (illustrated by lines) for controls and as mean values for patients.
was mixed with a 1 mliter suspension of "!mmunobeads"
HLA and C4 typing conjugated with either anti-!gA or anti-IgG. The procedure was HLA typing was performed by the clinical laboratory at then completed according to manufacturer's instructions. Lin- Hoxworth Blood Bank, Cincinnati, Ohio. C4 aflotyping was ear standard curves from 165 to 1,145 nglmliter for IgA, and 166 performed by the method of Awdeh and Alper [22]. Serum to 1,152 ng/mliter for lgG were obtained and unknown concen- immunoglobulin measurements were performed using an Auto trations determined from these. Supernatants which exceeded ICS nephelometer (Beckman, Laiolla, California, USA). the standard range were diluted and rerun. Statistics. Statistical comparisons were made using the chi 1gM was measured by an ELISA technique. Briefly, anti- squared test, analysis of variance, and multiple linear regression. human 1gM was raised in goats, the serum absorbed to monoResults specificity, and an Ig fraction obtained by octonoic acid precipitation [20]. the Ig fraction was affinity purified and biotinylated Mononuclear cell cultures fr immunoglobulin production as previously described [211. Polysytrene microtiter plates Unfractionated mononuclear cell preparations were from 30 (Limbro, McLean, Virginia, USA) were coated overnight at 4°C
with 0.1 mliter of affinity purified anti-human 1gM at 10 g/ healthy controls. Eight children (ages two to 20, mean = 11 mliter in 0.06 M pH 9 carbonate buffer. The wells were then years) and 22 adults (ages 26 to 59, mean = 35) were cultured blocked for 30 mm at 37°C with 0.2% bovine serum albumin in for 12 days in the absence of mitogenic stimulation. The mean carbonate buffer. Wells were washed with PBS. 0.01 NI EDTA, supernatant concentrations of lgA, lgG. and 1gM were 154, and 0.05% Tween 20 (wash buffer), and 0.1 mliter aliquots of 1181, and 142 ng/2 x 106 cells respectively (Fig. 1). There was undiluted culture supernatants or 1gM standards were added. A no overall relationship between age and supernatant !gA concurve was constructed from a calibrator serum of known 1gM centration (r 0.32, P > 0.05). Mean supernatant !gA in the concentration serially diluted in culture media, and incubated children was 165 ng/2 x 106 cells, and for the adults, 149 ng/2 x 106 cells (P > 0.05). Kinetic analysis of !gA production was for one hour at 37°C. The wells were washed and then incubated performed on four separate individuals. There was a plateau of for two hr at 37°C with 0.1 mliter of a 1:800 dilution of net immunoglobulin production at eight to ten days regardless biotinylated, affinity purified anti-human 1gM. After washing, 0.1 mliter of avidin—peroxidase (Sigma, St. Louis, Missouri, USA) was added for 30 mm at room temperature, and the wells
were then washed and developed with o-phenylenediamine (Sigma). A log linear curve from 50 to 1000 ng!mliter was obtained and unknowns determined from this. Values from wells coated with RPM! 1640, 15% FCS were not significantly different from those coated with wash buffer alone.
of the final value. Four individuals had repeat cultures during an
episode of upper respiratory or gastrointestinal infection and the net production of all immunoglobulin classes remained within the normal range. Mononuclear cell cultures in patients with IgA nephropathy As illustrated in Figure 1, mononuclear cells were isolated from 22 patients with IgA nephropathy. Nineteen of 22 patients
1231
Familial IgA synthesis in IgA nephropathy
with IgA nephropathy had supernatant IgA concentrations greater than two SD above the mean of control. The mean
10,000
patient supernatant IgA concentration was 620 compared to the control of 133 ng/2 x 106 cells (P = 0.007). Supernatant concentrations of IgG and 1gM were elevated in seven and three
0
patients with IgA nephropathy, respectively. However, the mean supernatant IgG and 1gM concentrations in patients were not significantly different from controls (P > 0.05). One of the IgA nephropathy patients with a normal superna-
() (0
0
tant IgA concentration had a renal allograft. The two other patients with IgA nephropathy and normal supernatant IgA concentration could not be distinguished from the entire patient group in any other way. In particular, there was no difference in their duration of disease, frequency of gross hematuria, qualitative degree of proteinuria or serum creatinine level. Serum IgA was elevated for age in II of 22 patients with IgA nephropathy. There was no correlation between the serum level of IgA and the supernatant IgA concentrations (r = 0,29, P > 0.05). Eight patients had significantly elevated supernatant IgA concentrations but normal serum IgA levels.
HLA and C4 allotyping HLA typing was performed on 27 patients with IgA nephrop-
athy, and no antigen was present in significantly increased frequency. In particular, B12, Bw35, and DR4, antigens previ-
ously associated with IgA nephropathy, were observed at frequencies of 0.23, 0,2 and 0.4, respectively, with expected
4 C
1,000
a) a) C a)
o
8 0 0 8
(a
o 8
0
cc
0 0
0
100
Controls
8
0 8
0
0
o L alrnxo
0
8 8
0 0
0
a
cc
Patients
Mothers
0
0
Fathers
Siblings
cc
0
Fig. 2. Unfractionated mononuclear cell culture production of IgA in
frequencies of 0.28, 0.15 and 0.28 (uncorrected P > 0.1). C4 is first degree relatives of patients with IgA nephropathy. coded within the HLA complex at 4 alleles (two A and two B).
Awdeh et al have defined "extended" HLA haplotypes by allotyping C4, C2, factor B and glyoxalase [23]. Patients with
IgA nephropathy have been reported to have an increased
production whereas in the W family, IgA production was
elevated only in the parents. Thus, no correlation between an incidence of C4 null genes at either the A or B locus [24]. No HLA antigen or parental haplotype and increased IgA producextended haplotype has been reported in increased incidence in tion could be made. IgA nephropathy [25]. Homozygous null C4B genes were found in five of 27 patients. This differs significantly from the observed
Discussion
parent affected, three had both, and four had neither parent affected. HLA typing was performed on three complete fami-
between IgA nephropathy and HLA Bw35 in France and Australia, HLA DR4 in France and Japan, and HLA B12 in
lies. The H family had two affected members, one with typical IgA nephropathy and the other with typical HSP. In this family, elevated IgA was present in the father and several siblings but not in the mother. The segregation of elevated immunoglobulin production in siblings followed no HLA antigen or parental
Florida [12—15]. No association with any HLA antigen, how-
incidence in the Cincinnati area of 3 out of 75 (x2 = 5.47, P < Studies of in vitro immunoglobulin production by mononu0.01) [26]. Homozygous null B genes have reported to occur in clear cells from patients with IgA nephropathy have suggested the extended haplotype B12, DR4, B5, C2 C, C4 A3, C4BqO. abnormal immunoregulation of IgA synthesis. Studies by others Based on HLA and C4 typing alone, two of our patients may using slightly different techniques have yielded similar, but not have had this haplotype. This incidence of two out of 27 identical, results to those of Cosio and Beale [9, 27, 28]. Beale patients would not differ from the observed frequency in the et al studied patients with HSP, an IgA related renal disease, Boston area of 10 out of 295 (x2 = I.!). and found increased spontaneous (non-mitogen stimulated) production of IgA by unfractionated mononuclear cells [7]. B Family studies cells isolated from these patients, in the absence of T cells Immunoglobulin production by mononuclear cells from first produced significant amounts of IgA, suggesting that in vivo B degree relatives of patients with IgA nephropathy was mea- cell activation had occurred. Allogenic mixing experiments sured. As illustrated in Figure 2, there was elevated supernatant suggested that defective T suppressor cell activity was also IgA concentrations in six of 11 mothers, six of 12 fathers and present. Cosio et al studied patients with IgA nephropathy and seven of 15 siblings. No significant increase in the concentra- demonstrated increased spontaneous production of IgA by tions of IgG or 1gM in supernatants was observed (not shown). unfractionated mononuclear cells [8]. Elevated parental IgA production varied; four families had one Previous studies have also identified significant associations
ever, was identified in a group of patients from the United Kingdom [29]. An increased incidence of homozygous null A or
B genes for the fourth component of complement, which lies within the major histocompatibility complex (MHC), has also haplotype. In the L family, all members had elevated IgA been reported in patients with IgA nephropathy, or HSP [24].
1232
Waldo et a!
Furthermore, a number of families in which more than one pursue this possibility. Mononuclear cells obtained from apmember have IgA nephropathy have been reported in one of proximately half of the parents and siblings of patients with IgA these, the asymptomatic brother of a patient with the disease nephropathy produced abnormally large amounts of IgA in was found to have glomerular IgA deposits [12, 16—19]. Prelim- culture. However, no clear Mendelian pattern of inheritance inary evidence of elevated IgA synthesis by cultured lympho- was observed. Families were observed with one, both or cytes from family members has also appeared [30]. Recently a neither parent involved. Three kinships had complete HLA large kindred from Western Kentucky has been reported [31]. typing, and elevated IgA production did not segregate with any This kindred had an unusually high incidence of IgA nephrop- measured HLA haplotype. The data clearly indicate familial athy, was not excessively inbred, and pedigree analysis sug- clustering of elevated IgA production, but point to no simple gested an autosomal dominant inheritance mode with poor pattern of inheritance. Although elevated IgA production is associated with IgA penetrance. From these observations, we proposed that a gene in or near nephropathy, its pathogenetic importance remains unclear. the MHC conferred a predisposition to IgA nephropathy, and Elevated IgA production alone would seem inadequate to was associated with abnormal immunoregulation and elevated produce disease since half the family members studied had this in vitro IgA production. Any HLA allotype in linkage dis- phenotype while reported familial cases of IgA nephropathy equilibrium with this proposed gene would occur with increased were relatively rare. It is possible that healthy family members frequency in patients with IgA nephropathy. Furthermore, with elevated IgA production actually have occult IgA nephrophealthy first degree relatives having this HLA allotype would athy and would have positive mesangial immunofluorescence also have the abnormal immunoregulatory gene and elevated in for IgA if biopsied. Such a case of occult IgA nephropathy has been previously described [12]. Until a reliable, non-invasive vitro IgA production. Unfractionated mononuclear cells from our patients with IgA marker of IgA nephropathy is established this question will nephropathy cultured without mitogenic stimulation produced remain unanswered.
Since the data do not establish a genetic pattern for the significantly more IgA than controls, confirming the presence of abnormal immunoregulation in patients with IgA nephropathy. observed familial clustering of elevated IgA production, the Interestingly, no significant correlation was found between the possibility of environmental factors must be considered. Viral
serum level of IgA and the quantity of IgA produced in cell agents, such as Epstein—Barr virus and other members of the culture. The serum concentration of IgA is determined not only by its rate of synthesis, but by its rate of clearance, which could be elevated in patients with IgA nephropathy. Furthermore, the rate of in vitro synthesis does not necessarily reflect that which occurs in vivo. In our patients with IgA nephropathy, contrary to previous results, no HLA antigen was observed in increased frequency. However, there was an increased incidence of homozygous null
herpes family, are well known B cell mitogens and could served as shared environmental exposures. Dietary substances such as bacterial lipopolysaccharide could also act as local mitogens for
B genes for C4. The reason for disease associations with the MHC are unclear. Awdeh have established extended haplotypes within the MHC defined by HLA typing, combined with C2, C4, B and glyoxalase typing [23]. These extended haplotypes are conserved in a population by crossover suppression and male transmission bias. It has been proposed that disease susceptibile genes within this complex are also conserved and help maintain linkage between diseases and the HLA antigens. However, the extended haplotype containing homozygous null
tion in patients and their family members.
T-helper/inducer cells in the gut associated lymphoid tissue. Studies of unrelated individuals living in the same household as
those patients with IgA nephropathy could help clarify this issue. Finally, it will be important to determine if the same cellular mechanisms are responsible for elevated IgA produc-
Acknowledgments This work was supported in part by National Research Service Award AM 07051 (to F. B. Waldo), and by grant RR 00123 from the
General Clinical Research Centers Branch, both from the National Institutes of Health. Thanks to Ms. Karen Kilgo and Kathie Thomas for secretarial assistance.
B alleles was not observed at an increased incidence in patients, Reprint requests to Dr. F. B. Waldo, The Children's Hospital, /600 Seventh Avenue South, Birmingham, Alabama 35233, USA. making this explanation unlikely in the present series. The association between C4 B null alleles and IgA nephropathy may result instead from the difference in hemolytic activity References of C4, A and B gene products. Most of the hemolytic C4 activity
resides in C4, B gene products. Patients who are null at both C4, B loci have relatively low hemolytic C4 levels [26]. The classical pathway of complement should be important in the processing and clearance of IgA-lgG immune complexes. A relative deficiency in C4 hemolytic activity could, thus, be one of several predisposing factors in the development of IgAN. The above data confirm the association between elevated in
vitro production of IgA and nephropathy, but provide no evidence to link elevated in vitro IgA production with a particular HLA antigen. Abnormal IgA immunoregulation could, however, be inherited at a locus not linked to the MHC and, thus, occur within families. Family studies were done to
1. BERGER J, HINGLAIS N: Les depots intercapillaires d'IgA-IgG. J Urol Nephrol 74:694—695, 1968
2. CLARKSON AR, SEYMOUR AE, THOMPSON AJ, HAYNES WDG, CIIAN YL, JACKSON B: IgA nephropathy: A syndrome of uniform
morphology, diverse clinical features and uncertain prognosis. Clin Nephrol 8:459—471, 1977
3. DROZ D: Natural history of primary glomerulonephritis with mesangial deposits of IgA. Contr Nephrol 8:335—340. 1977
4. Eco J, SANCHO J, BLASCO R, RIvE F, HERNANDO L: Immunopathogenetic aspects of IgA nephropathy. Adv Nephrol 12:103—137, 1983
5. WOODROFFE AJ, GORMLY AA, MCKENZIE PE, W00TON AM, THOMPSON AJ, SEYMOUR AE, CLARKSON AR: Immunologic studies in IgA nephropathy. Kidney mt 18:366—374, 1980
Familial IgA synthesis in IgA nephropathy 6. LESAVRE PH, DIGEON M, BACH iF: Analysis of circulating IgA and
detection of immune complexes in primary IgA nephropathy. C/in
Exp Immunol 48:6l—69, 1982 7. BEALE MG, NASH GS, BERTOVICH J, MACDERMOTT RP: Similar
disturbances in B cell activity and regulatory T cell function in Henoch—Schonlein purpura and systemic lupus erythematosus. J Immunol 128:486—491, 1982
AD, CONLEY ME, MICHAEL AF: Immune regulation of immunoglobulin production in IgA nephrop-
8. Costo FG, LANN 5,
FOLAMI
athy. C/in Immunol Immunopatho/ 23:430—436, 1982 9. Ecwo J, BLASCO R, SANCHO J: Increased rates of polymeric IgA synthesis by circulating lymphoid cells in IgA mesangial glomerulonephritis. C/in Exp Immunol (in press) 10. NOMOTO Y, SAKAI H, ARIMORI S: Increase of IgA bearing lymphocytes in peripheral blood from patient with IgA nephropathy. Am J C/in Path 71:158—160, 1979 11. BENE MC, FAURE G, HURAULT DE LIGNY B, KESSLER M, DUHEILLE J:
Immunoglobulin A nephropathy quantitative im-
1233
the possible role of HLA antigens. Transplantation 29:505—506, 1980 20. STEINBUCK M, AUDRAN R: The isolation of IgG from mammalian
sera with the aid of caproic acid. Arch Biochem Biophys
134:279—284, 1969 21. BAYER EA, SKUTELSKY E, WILCHEK M: The avidin—biotin com-
plex in affinity cytochemistry. Methods in Enzymes 62:308—315, 1979 22. AWDEH ZL, ALPER CA: Inherited structural polymorphism of the
fourth component of human complement. Proc Nat/ Acad Sci USA
77:3576—3580, 1980 23. AWDEH ZL, RAUM D, YuRRI5 EJ, ALPER CA: Extended HLA/
complement allele haplotypes: Evidence for T/t like complex in
man. Proc Nail Acad Sci USA 80:259—261, 1983 24. MCLEAN RH, WYATT RJ, JuLIAN BA: Inherited deficiency of the fourth component of complement (C4) in IgA nephropathy, Henoch—
Schonlein purpura and recurrent hematuria. (abstract) Kidney Int 21:202, 1982
inversion of the immunoglobulin A versus immunoglobulin G
25. ARNAIZ—VLLLENA A, GONZALO A, REGUERRO JR, VICARI0 JL, ORTUNO J: Extended HLA haplotypes in Berger's disease. C/in
secreting cell balance. J C/in Invest 71:1342—1347, 1983 AV, MAHONEY JJ, FULLER TJ: Higher prevalence of
Nephrology 22:320—321, 1984 26. WELCH TR, BIESCHEL L, BERRY A, FORRISTAL J, WEST CD: The
munohistomorphometry of the tonsillar plasma cells evidences an 12. RICHMAN
HLA-B12 in patients with IgA nephropathy. (abstract) Ann Intern effect of null C4 alleles on complement function. C/in Immuno/ Med 90:201, 1979 Immunopathol 34:316—325, 1985 13. NOEL H, DESCAMPS B, JUNGERS P: HLA antigens in three types of 27. ROTHSCHILD E, CHATENOUD L: T cell subset modulation of glomerulonephritis. C/in Immuno/ lmmunopatho/ 10:19—23, 1978 immunoglobulin production in IgA nephropathy and membranous 14. FAUCHET R, LE POGAMP P, GENETET B, CHEVET D, GUEGUEN M, glomerulonephritis. Kidney mt 25:557—564, 1984 28. EGID0 J, BLASCO R, SANCHO J, LOZANO L: T cell dysfunctions in SIMON P, RAMEE MP, CARTIER F: HLA-DR4 antigen and IgA nephropathy. Tissue Antigens 16:405—410, 1980 IgA nephropathy: specific abnormalities in the regulation of IgA 15. KASHWABARA H, SHISHIDO Y, YOKOYAMA T, MIGAJUINA T: HLA synthesis. C/in Immunol Immunopathol 25:201—212, 1983 29. RASHID Mt, PAPIHA SS, AGRUYANNIS B, MORLEY AR, WARD in IgA nephropathy. Tissue Antigens 16:411-412, 1980 16. M0NT0LIu J, DARNELL A, TORRAS A, ERCILLA G, VALLES M, MK, ROBERTS DF, KERR DNS: The associations of HLA and the REVERT L: IgA nephropathy. Arch Intern Med 140:1374—1375, 1980 genetic markers with glomerulonephritis. Hum Genet 62:38—44, 17. TOLKOFF—RUBIN NE, CO5INI AB, FULLER T, RUBIN RH, COLVIN 1983 RB: IgA nephropathy in HLA-identical siblings. Transplantation 30. EGIDO J, BLASCO R, SANCHO J, LOZANO L, GUTIERREZ—MILLET V: Immunological studies in a familial IgA nephropathy. C/in Exp 26:430—433, 1978 18. SABATIER JC, GENIN C, ASSENOT H, COLON S, DUCRET F, Immuno/ 54:532—538, 1983 BERTHOUX FC: IgA glomerulonephritis in HLA identical brothers. 31. JULIAN BA, QUIGGINS PA, THOMPSON iS, WOODFORD SY, GLEASON K, WYATT Ri: Familial IgA nephropathy evidence of an C/in Nephro/ 11:35—38, 1979 19. KATZ A, KARANICOLAS 5, FALK J: Family studies in IgA nephritis: inherited mechanism of disease. N Engi J Med 312:202—208, 1985