Sequential study of the IgA system in relapsing IgA nephropathy

Kidney International, Vol. 30 (1986), pp. 924—931

Sequential study of the IgA system in relapsing IgA nephropathy JOHN FEEHALLY, 1. JAMES BEATTIE, PAUL E.C. BRENCHLEY, BEATRICE M. COUPES, NETAR P. MALLICK, and ROBERT J. POSTLETHWAITE Departments of Renal Medicine, Manchester Royal Infirmary and Royal Manchester Children's hospital, Manchester, United Kingdom

Sequential study of the IgA system in relapsing IgA nephropathy. Cellular and immunochemical parameters of the lgA system were studied in 15 subjects with IgA nephropathy (IgAN) and 15 age— matched controls. In IgAN remission no abnormalities of the igA system were detectable by the methods used. In IgAN relapse, [macroscopic hematuria associated with upper respiratory tract infection (URTI) (N 6)1 there were rises in IgA-bearing B-lymphocytes (three

of six), T helper/suppressor cell ratio (six of six) and pokeweed mitogen—induced IgA production (four of six). Total serum and salivary

IgA were unchanged. Serum IgA profile (HPLC-ELISA) showed increases in polymer IgA (three of six). No such changes were found

single time point, using patients who may have a range of clinical manifestations at the time of study. Although some studies have related IgA immune abnormalities to disease activity, repeated studies in the same individuals during changes in clinical state have not been reported. Striking abnormalities in IgA immunity are most likely to be found when disease is very active (during bouts of macroscopic hematuria).

Furthermore, the wide variation in many of the parameters

studied, even in healthy controls, creates difficulty in interpretduring URTI in controls. These lindings support the view that an ing results when comparing groups, suggesting that sequential exaggerated IgA response to mucosal antigen challenge initiates studies of individuals at different stages of disease may be the glomerular damage and hematuria in IgAN. most useful way of identifying important disease—related phenomena. Since its first description in 1968 [1], IgA nephropathy (IgAN) has become a well recognized clinicopathological entity [2]. Its

most characteristic clinical setting is recurrent macroscopic hematuria in children and young adults, hematuria often occurring within 12 to 24 hours of the onset of upper respiratory tract or gastrointestinal infection. The coincidence of macroscopic hematuria with infection at mucosal surfaces suggests that an abnormal mucosal immune response may precipitate glomeru-

In this study, a homogeneous cohort of children and young adults with IgAN and recurrent macroscopic hematuria were studied. They were first analyzed while asymptomatic (remission) and then were investigated again at the height of macroscopic hematuria (relapse). Significant changes in immune parameters were sought by analyzing paired data for remission and relapse in the same individual. Results were also compared to age—matched healthy controls, some of whom were studied during upper respiratory tract infection. Methods

lar damage. Polymer IgA (consisting of dimer and higher polymers) is produced at mucosal surfaces in response to

Patients antigen challenge. Although only a small proportion of serum IgA is polymer, there is good evidence that glomerular IgA in Eight children and seven adults (13 male, two female) with IgAN is polymeric, since it contains J chain [3—6] and has the l)iopSy proven IgAN and recurrent episodes of macroscopic capacity to bind secretory component (SC) [3, 6, 7], although hematuria were studied. Hematuria had usually (11 of 15) been SC itself is not found in the kidney [8, 9], Animal models precipitated by infection, especially of the upper respiratory suggest that the type of mesangial injury seen in IgAN may be tract (URTI); although in some (four of 15) the precipitating induced by the deposition of circulating antigen—antibody com- event was unclear. Age range at onset (first episode of plexes containing IgA (lgA-CIC) [10] or by the accretion of IgA hematuria) was four to 28 years, and at the time of study was

eight to 33 years. Two clinical states were defined as the Recent studies in IgAN have reported increased levels of following. Remission: The patient was symptom free and was normoserum polymer IgA [12, 13] and IgA-CIC [14—19], and demonstrated possible abnormalities in the mechanisms involved in tensive. There was normal renal function (creatinine clearance

to antigen already in the mesangium [1111.

their production and clearance [20—24]. However, these studies >80 rnl/min/i .73 m2, serum creatinine <100 mol/liter) and the have been cross—sectional, investigating cohorts of patients at a only laboratory evidence of renal disease was persistent microscopic hematuria (13 of 15) and in some (seven of 15) low grade proteinuria (<1 g/24 hrs). Received for publication July 22, 1985 and in revised form March 3, 1986

Relapse: An episode of frank hernaturia. Such episodes occurred during the study in six subjects. The precipitating

© 1986 by the International Society of Nephrology

event was URTI (five of six) or gastrointestinal infection (one of

924

IgA system in relapsing IgA nephropathy

925

six). In two of six patients there was a significant transient ulation given in the light of the total lymphocyte count. The decline in renal function, serum creatinine rising from 80 to 115

mol/liter and from 100 to 240 mol/liter, respectively. Subjects were studied in remission and at the height of a subsequent relapse (within 24 hrs of onset of hematuria). No patient had hypertension and none had evidence of systemic lupus or liver cirrhosis. In no patient had there been clinical

balance of circulating immunoregulatory cells was expressed by the OKT4+/OKT8+ (helper/suppressor) cell ratio. Pokeweed mitogen—induced immuno globulin production. Af-

ter separation, mononuclear cells were washed six times in

HBSS and resuspended at 1 x l06/ml in tissue culture medium (RPM! 1640, Flow Labs) with 10% fetal calf serum, penicillin evidence of Henoch—Schonlein purpura. (100 g/ml), streptomycin (50 g/ml), and buffered by 20 mivi Hepes and sodium bicarbonate (0.85 g/liter). Two hundred jd(2 Controls x l0 cells) were incubated in round—bottomed multiwell tissue Nine age—matched children and six adults (nine male, six culture plates (Linbro). Ten d of pokeweed mitogen (PWM, female), age range seven to 33 years, were studied as controls. Gibco Laboratories, Grand Island, New York, USA) was added Adults were volunteers from laboratory and medical staff, to wells in triplicate at dilutions of 1:10, 1:25, 1:50; control wells children were attending general pediatric clinics with conditions received no mitogen. The cells were cultured for one week at having no immunological or renal component. All controls had 37°C in 5% CO2. Plates were then centrifuged at 400 g for 10 normal urine sediment and normal serum creatinine. Five minutes and 150 tl of culture supernatant aspirated from each controls (age range eight to 32 years) were initially studied well to a fresh well. Supernatants were stored at minus 20°C to during URTI and subsequently after recovery. Venous blood await analysis. and saliva were collected from each subject after an overnight Quantitation of supernatant immunoglobulin (and salivary fast. IgA and HPLC eluates) was by enzyme—linked immunosorbent assay (ELISA). Multiwell plates (Dynatech) were coated with Lymphocyte studies anti-human IgG (Dako, Copenhagen, Denmark), 1gM (Dako), Isolation of peripheral blood lymphocytes. Twenty—five to 30 or IgA (Atlantic) by incubating at 4°C for 18 hours in bicarbonml venous blood was transported in preservative free heparin. ate buffer pH 9.6. Plates were washed in PBS Tween and BSA The blood was diluted in PBS, overlaid onto Ficoll—Triosil 1 mg/ml added for one hour to prevent non-specific protein (Lymphocyte Separation Medium, Flow Labs, Helsinki, Fin- absorption. Dilutions of PWM culture supernatant (or saliva) in land) and centrifuged at 400 g for 25 minutes at room tempera- PBS-BSA-tween were added to 4°C for 18 hours. After washture. The mononuclear cells at the interface were harvested and ing, plates were incubated with 1:1000 peroxidase—conjugated washed in calcium and magnesium—free Hanks balanced salt anti-IgG, 1gM and IgA (Dako). Plates were developed by the solution (HBSS, Gibco, Grand Island, New York, USA). After addition of ABTS (Sigma Chemical Co., St. Louis, Missouri, USA) in citrate phosphate buffer pH 5.0 with 0.002% hydrogen washing, the cells were resuspended in complete HBSS. Total and differential white cell counts were performed by peroxide. Intensity of color generated was read as absorbance fluorescence microscopy after staining with acridine orange. by an automatic plate reader (Multiscan, Flow) with a 420 nm This and all other fluorescence microscopy used a Leitz Dialux filter. Standard curves used pooled IgG, 1gM and IgA myeloma proteins. Results for each immunoglobulin class were ex2OEB microscope (Leitz Inc., Heerbrugg, Switzerland). Enumeration of lymphocyte subpopulations: a) B lympho- pressed as ng produced per 2 x iO cells using the median of the cytes. B lymphocytes were estimated as surface immunoglob- triplicate of the 1:25 PWM dilution (this dose of PWM gave the ulin bearing cells. One x 106 mononuclear cells were incubated optimal response in the majority of subjects). Saliva assays. Saliva specimens were placed at 4°C immediat 4°C with fluorescein isothiocyanate conjugated (FITC) F (ab')2 goat antisera to human total immunoglobulin or IgA ately after collection, and 0.01% sodium azide was added as (Kallestad) for 30 minutes. (F (ab')2 fragments were used to soon as possible. Aliquots were then stored at minus 70°C until avoid non-specific binding to Fc receptors on some cells.) The the assays were performed. Total salivary IgA was measured by cells were washed twice in complete HBSS containing 0.01% ELISA (as described above). Salivary albumin was measured sodium azide, being kept at 4°C throughout. Cells with positive by single radial immunodiffusion. Serum immunoglobulins. Serum total IgG, IgA and 1gM were surface fluorescence were expressed as a percentage of total lymphocytes seen, and as an absolute value for each subpop- measured by laser nephelometry using a Beckman Auto ICS ulation in the light of the total lymphocyte count. nephelometer (Beckman Instruments, Fullerton, California, b) T lymphocytes. T cell subpopulations were recognized on USA). the basis of surface phenotype defined by monoclonal antibodHigh performance liquid chromatography (HPLC). HPLC ies using indirect immunofluorescence. One x 106 mononuclear was used in conjunction with ELISA to analyze IgA monomer cells were incubated at 4°C for 30 minutes with 25 pJ of mouse and polymers in serum [25]. Molecular sieving HPLC was performed using a TSK 4000 monoclonal antibody OKT3 (recognizing total T cells), OKT4 (recognizing T helper/inducer cells), or OKT8 (recognizing T SW column connected via a guard column to an HPLC pump suppressor/cytotoxic cells) (Ortho Pharmaceuticals, Raritan, (LKB). Flow rate was 0.4 ml/min using 0.1 M sodium phosphate New Jersey, USA). Cells were washed at 4°C in HBSS con- buffer pH 6.8. Serum diluted 1:20 was passed through a 1 ifilter taining 0.01% sodium azide and then incubated for 30 minutes (Millipore Corp., Bedford, Massachusetts, USA) and a 40 tl with FITC goat anti-mouse IgG (Miles Laboratories, Elkhart, sample injected on to the column. Fractions were collected Indiana, USA). After further washing, cells with positive sur- from the column and transferred to multiwell plates (Dynatech) face fluorescence were counted as a percentage of total lym- for analysis by ELISA. The ELISA plates were coated with goat anti-human IgA phocytes, and an absolute value for each circulating subpop-

926

Feehally ci of

(Atlantic) and incubation with the column fraction was followed by peroxidase—conjugated anti-human igA (Dako). Two plates were used for each run. On one a monomer IgA standard curve

Table 1. Peripheral blood lymphocyte populations in IgA nephropathy (remission) and controls (mean SEM)

was used with 10 d eluate per well. A second plate was optimized for polymer IgA using a polymer lgA standard curve and 20 .d eluate per well.

IgA monomer and polymer standards were prepared using sera from five patients with IgA myelomatosis (three were IgA A, two IgA ) which contained both monomer and dimer species

as shown by HPLC and ELISA. Sera were fractionated by

Control

(N OKT3+ Total T cells, % Absolute, x109/Iiter OKT4+ T Helper cells, % Absolute, x109/liter

ammonium sulphate precipitation and DEAE chromatography, OKTS + and IgA was size fractionated into monomer and dimer species I suppressor, % Absolute x109/liter by HPLC. Separate monomer and dimer standards were produced by pooling the purified components from the five sera. OKT4 + /OKT8 + Helper/suppressor The standard preparations were shown by ELISA to be free of Total B cells

IgG and 1gM and were assayed for protein content by the

Lowry method. Serum complement Components. Total hemolytic complement activity (CH5O) was measured by timed lysis of sensitized sheep red blood cells. C3, C4 and factor B were measured by

Percentage Absolute, x109/liter IgA-B cells Percentage

Absolute xlIP/ljter

15)

IgAN remission

(N= 15)

68,0

1.4

63.2

3.0

1.61

9.12

1.37

0.12

43.4

2.7 0.10

36.8

1.06

0.80

2.7 0.10

28.0 0.63

2.0 0.05

26.1 0.56

0.04

1.67

0.17

1.48

0.15

9.5 0.23

7.6

0.7

0.04

2.10 43.1

1,1

1.1

0.15

0.01

0.17

1.95

4.1

45.4

0.24 7.6

laser nephelometry and Clq by radial immunodiffusion. C3 breakdown products were measured by immunoelectrophoresis and immunofixation.

and controls. IgA production (stimulated or unstimulated) did not increase significantly in the group in IgAN relapse as a Statistical methods whole, although there were marked increases in two subjects Parametric data are tabulated as mean SEM, and non- and individuals with IgAN were more likely to have increased parametric data as median with ranges. Comparison of means PWM-stimulated IgA production (four of six) compared to used Student's t-test (for parametric data) and Mann—Whitney controls during URTI (two of five) (Fig. 3). There were no significant changes in IgG or 1gM production. test (for non-parametric data). Serum imrnunoglobulins. Total serum lgG and 1gM were normal in all study groups. In IgAN remission total serum IgA Results was 2.5 (1.5—5.8) g/liter. Adult normal range in the laboratory Lymphocyte subpopulalions performing the assays was 0.6—3.3 g/liter. However, IgA levels In remission IgAN subjects did not differ from controls in are low in childhood and only slowly increase to adult levels percentage or absolute numbers of T helper, T suppressor, total during the second decade of life. Eight of the 15 IgAN subjects had serum IgA levels above the age—corrected normal range. B, or lgA-B cells (Table 1). Paired data for changes in lymphocyte subpopulations in During IgAN relapse or control URTI levels did not change IgAN relapse and control URTI are shown in Table 2. significantly in any individual. T lymphocyte subpopuiations. There was a significant inSerum complement components. C3, C4, Clq and Factor B were normal in IgAN remission and controls, and remained so crease in helper/suppressor cell ratio in IgAN relapse (2.27 0.45) compared to remission (1.52 0.23) (P < 0.01), paired in relapse and control URTI, giving no evidence of widespread complement activation. C3 breakdown products were not def-test), while no such change was seen in control IJRTI (1.60 0.1 vs. 1 .86 0.39). These changes in helper/suppressor cell tected. ratio are shown in Figure 1. The change in helper/suppressor Serum IgA profile. Using HPLC-ELISA analysis the majority cell ratio is predominantly due to a decrease in the T suppressor of IgA (>90%) in normal serum appears as monomer preceded subpopulation, although no changes in T cell parameters during by much smaller peaks of dimer and higher polymers, which in some subjects may be undetectable. Figure 4 shows the polyIgAN relapse are statistically significant. B lymphocyte subpopuiations. Total B cell numbers did not mer IgA profile in fasting serum in one control subject and in change in IgAN relapse. There were, however, increases in three subjects with IgAN during remission and relapse. In the percentage and absolute IgA B cell numbers in IgAN relapse, control and in remission in one IgAN subject, dimer and while during control URTI there were decreases in the same polymer peaks were barely detectable, although they were parameters. However, variations were wide and these shifts clearly seen in the other two IgAN subjects in remission. In relapse in all three IgAN subjects there were striking increases were not statistically significant. in dimer and higher IgA polymers. Such increases in polymer Changes in percentage IgA B cells are shown in Figure 2. PWM-induced immunoglobulin synthesis. Values for lgG, IgA were not seen in three other IgAN relapses or during IgA, and 1gM in vitro synthesis are shown in Tables 3 and 4. control URTI. Two of the three subjects with detectable Results are given for unstirnulated cultures and with the addi- increases in polymer TgA were those whose relapses were tion of PWM (1:25), the dose (of three used) which gave the severe enough to cause a transient decline in renal function. Salivary IgA. Total salivary IgA was corrected for dilution optimum response under the experimental conditions used. There were no significant differences between IgAN remission variation by expressing results as a ratio IgA/albumin. This

927

IgA system in relapsing IgA nephropathy Table 2. Changes in peripheral blood lymphocyte populations during IgAN relapse and control URTI (paired data

Control (N =

lgAN (N = 6)

5)

URTI

Healthy

OKT3 + (Total T cell)

% Absolute, x 10°/liter OKT4 + (T helper cell)

%

Absolute, x 10°/liter Absolute, x 10°/liter OKT4 +/OKT8 + Helper/suppressor Total B cells % Absolute X 10°/liter lgA-B cells % Absolute x 10°/liter a Mean + SEM ° P < 0.01 (paired

6.2

0.18

44.5 0.89

5.8 0.09

1.6 0.07

22.4

0.61

0.46

3.5 0.08

0.19

1.52

0.23

2.27

0.45°

1.5

8.9 0.18

1.2

0.05

0.03

6.6 0.23

0.8 0.08

1.31

0.29

2.27

0.48

27.8

11

3.27 71.8

24

1.8

70.6

0.33

1.48

2.4 0.41

66.7 1.46

43.4 0.96

3.9 0.16

45.0 0.89

5.1 0.13

40.7 0.93

4.8

26.3

3.9 0.24

29.3 0.62

3.7

27.6

0.66

0.20

1.86

0.39

1.60

6.9 0.12

1.0

0.03

9.4 0.19

0.21

8.9

2.24 50.0

Relapse 65.1 1.47

66.1 1.52

OKT8 + (T suppressor cell)

%

Remission

53

5.4

0.21

14

0.1

1.0

t-test)

Controls

gA nephropathy

4.0 -

in patients with IgAN predisposes to mesangial IgA deposition and glomerular hematuria. This hypothesis is endorsed by the finding that mesangial IgA in IgAN is polymeric [3—7], since it is

known that mucosal IgA is polymeric, whereas the majority of serum IgA is monomeric. Previous studies have sought abnormalities in parameters of the IgA immune system in IgAN, and although findings have not always been consistent, a number of abnormalities have been described.

3.5 —

3.0—

2.5

Increased numbers of circulating IgA-B cells have been reported in IgAN [24] but this could not be confirmed when more rigorous methodology used F(ab')2 anti-IgA antibodies

—

I—

0

2.0-

(eliminating non-specific Fc binding) to define IgA positive cells

0 1.5

[26, 27]. Recently, enumeration of plasma cells in lymphoid tissue (using tonsillectomy specimens) showed an increase in

—

IgA producing cells in IgAN [28, 29]. Circulating T cell subpop-

ulations may also be abnormal in IgAN with increased 1.0

helper/suppressor ratio [20, 30, 311. The possibility that this altered T cell ratio influences IgA production is supported by

0.5

0

Healthy

URTI

Remission

Relapse

predominantly polymeric 33]) has been variably reported as

Fig. 1. OKT4/OKT8 ratios in control URTI and IgA nephropathy.

ratio was not significantly different in IgAN (11.1

findings of increased T alpha cells and abnormal IgA-specific T cell help and suppression in IgAN [20, 23, 32], However, IgA production from lymphocyte culture using pokeweed mitogen (PWM) (a T cell—dependent stimulus producing IgA which is

2.6)

normal [27] or increased [20, 21] in IgAN; limited evidence has not associated changes in in vitro IgA production with disease activity in IgAN [27].

There is evidence that circulating total and polymeric IgA may be abnormal in IgAN. An increase in total serum IgA is reported in 30 to 60% of patients with IgAN 116, 27, 34—36]. However 90% of circulating IgA is monomeric, and changes in Discussion polymer IgA may not be reflected in total IgA levels, increased Recurrent episodes of macroscopic hematuria are character- polymer IgA in absolute levels and also as a percentage of total istic of IgAN. Such hematuria is often precipitated by mucosal serum IgA has been reported [12, 13] but not confirmed by (respiratory or gastrointestinal) infection [21. Since IgA has a others [15, 37—39]. Some workers [12] measured IgA by commajor role in mucosal immunity, it has been proposed that an petitive inhibition radioimmunoassay in fractions from gel filabnormal IgA mucosal immune response during such infections tration or sucrose density ultracentrifugation. Unfortunately, compared to controls (7.9 2.4), and did not change in IgAN relapse (12.7 4.2). In control URTI the ratio was lower (2.8 0.6), but the data is from four subjects only.

928

Feehally ci a! Controls

IgA nephropathy

Controls

IgA nephropathy

. (230)

100 -

S

6.0

90

5.5

-

80 — 70

5.0

(0 (0

U

60

—

-

4.5 50

4.0

U,

40

—

-

30 —

3,5

0)

U

4

20

3.0

—

10—

0

2.5 2.0

S

URTI

Fig. 3. PWM siimulated IgA production in control URTI and IgA

nephropathy.

1.5

Table 3. in vitro immunoglobulin synthesis by peripheral blood lymphocytes in IgA nephropathy (remission) and controls, median (range) ng/2 X iO cells

1.0

0.5

0—

Healthy

S

Healthy

URTI

Remission

Fig. 2. Percentage IgA-B cells in control URTI and IgA nephropathy.

this type of radioimmunoassay is known to overestimate poiymer IgA [40]. Cross—sectional studies in IgAN show increased IgA-CIC by a number of assays, including the conglutinin binding assay [14, 15], Raji cell assay [15—18], and anti-IgA inhibition assay [17, 191. However the conglutinin assay has also been found negative [361. Some groups report an association between the level

Control

IgA nephropathy

10.1 (2.0—38.5) 15.9 (1,5—48.3)

8.3 (3.5—38.7) 15.2 (4.0—239)

26.2 (12.6—37.7) 35.5 (16.4—123)

22.1 (1.0—87.4) 34.2 (1.0—203)

12.3 (5.1—46) 27.0 (1.0—221)

10.8 (1.0—28.1) 34.6 (2.0—200)

N=15

Relapse

lgA unstimulated PWM1:25 lgG unstimulated PWM1:25

N=14

1gM

unstimulated PWMI:25

of CIC and disease activity, exemplified by macroscopic hematuria [14, 17, 371, others do not [15]. CIC, particularly those containing IgA, increase after food in healthy subjects [41] and this response may he exaggerated in patients with glomerular disease, including IgAN [42]. Of all studies of CIC in

IgAN, only one [17] states that CIC were assayed in fasting serum samples, so that findings of raised lgA-CIC levels must be interpreted with caution. Inconsistencies in the IgA immune system abnormalities found in IgAN may have some basis in methodological varia-

normalities and disease activity in IgAN have been reported but only on the basis of cross—sectional studies where individuals have been studied once, Immune parameters have wide biological variation even in healthy populations so that demonstrating

significant differences with disease activity in such cross— sectional analyses may be difficult.

The present study was designed to overcome these diffi-

culties. A homogeneous cohort of 1gAN subjects with typical recurrent hematuria was selected. Patients with hypertension, tions described above. However, variability in patient selection heavy proteinuria or renal impairment were excluded, as were and study design may also contribute. IgAN is a clinically patients with systemic diseases associated with mesangial lgA heterogeneous disease; some patients have hypertension, renal deposition (particularly SLE and cirrhosis). These may modify impairment and proteinuria, and are not prone to recurrent immunity, and it is possible in these entities that IgA deposition macroscopic hematuria. Such patients may be at a later stage of in the kidney occurs through different immunopathogenic disease evolution when immune mechanisms are less relevant mechanisms. Cases of Henoch—SchOnlein purpura were also to the persistence and progression of their renal disease. If IgA excluded from the study. When pathogenic mechanisms are immune system abnormalities have major importance in IgAN, better defined, Henoch—Schönlein purpura and IgAN may such abnormalities may best be demonstrated during episodes prove to be part of the same spectrum of disease; but until such of macroscopic hematuria. Associations between immune ab- mechanisms are known, differences in age of onset, systemic

929

IgA system in relapsing IgA nephropathy

Table 4. In vitro immunoglobulin synthesis by peripheral blood lymphocytes in IgA nephropathy relapse and control URTI (paired data), median (range) ng/2 X 10 cells Control N

IgA unstimulated PWM 1:25 IgG unstimulated PWM 1:25

=

IgAN N

15

=6

Healthy

URTI

Remission

Relapse

13.7 (2.5—38.5) 17.4 (1.5—48.3)

12.3 (6.4—74.3) 13.0 (5.5—61.9)

10.1 (3.5—33.9) 12.3 (4.0—38.2)

12.0 (1.0—57) 18.0 (1.0—84.2)

27.8 (22.3—37.7) 35.9 (22.8—66.4)

19.3 (13.5—44.7) 26.0 (16.7—146)

26.1 (3.0—66.2) 23.2 (1.0—79.8)

22.9 (11.1—50.1) 26.1 (7.5—102)

19.1 (7.4—23.1) 37.2 (30.8—77.6)

9.0 (1.0—67.7) 21.2 (0.5—203)

13.7 (6.3—46.0) 21.4 (1.0—94.1)

13.5 (7.6—47.6) 28.5 (9.3—201)

1gM

unstimulated PWM 1:25

25

p

d

m

p

I

I

I

I

dm I I

II d

p

mp

d

m

I

I

15 5

1

2

3

4

Fig. 4. HPLC-ELISA serum IgA profile in one control (1) and three subjects with IgAN (2-4) in remission (—) and relapse (---). Dimer and higher polymer IgA peaks are seen in two IgAN subjects (2, 4) in remission but are barely detectable in the control (1) and one IgAN remission (3). Increases in polymer IgA are seen in all three IgAN relapses (2—4).

involvement, and association with intercurrent infection justify controls with URTI also showed some increase in IgA productheir separation. tion. A larger study would be required to seek confirmation of Patients were studied in remission and compared to any possible difference between IgAN and controls. The preage—matched controls, and some IgAN subjects were subse- sent studies were performed using fetal calf serum, and do not quently studied in relapse at the height of macroscopic exclude the possibility that a serum factor in IgAN may modify hematuria, when maximum IgA immune abnormality might be IgA synthesis in vivo. In the present study, total serum IgA was increased in IgAN expected. The significance of any changes was sought by comparison with remission data in the same patient by paired but did not rise further in relapse. Using HPLC-ELISA, three analysis. The possibility that any changes found were a conse- of six subjects studied showed clear increases in dimer and quence of the intercurrent infection which precipitated the higher polymers of IgA in relapse compared to remission (Fig. hematuria, rather than the relapse itself, was investigated by 4), although three others had no detectable change in IgA studying some controls during URTI and making similar com- profile. The HPLC-ELISA technique used does not define with parisons of paired data. No such sequential study of IgAN has certainty whether the polymer IgA is aggregated IgA or represents antigen—antibody complexes in which the antibody moeity previously been reported. In remission in IgAN we found no significant abnormalities in is IgA. Our studies were performed on fasting serum, so that IgA system parameters compared to controls. In relapse there food—mediated rises in IgA-CIC, which might modify the serum was a significant increase in helper/suppressor T cell ratio (P < IgA profile, were excluded. 0.01) and increases were also found in percentage and absolute Studies of salivary IgA were undertaken to see if changes in numbers of IgA-B cells, which were not statistically significant. circulating IgA were paralleled by changes in the IgA mucosal In vitro production of IgG and 1gM from lymphocytes stimu- system. However, there were no changes in salivary IgA lated by PWM was not changed in IgAN relapse and control (corrected for dilution factors using salivary albumin levels) in

URTI. However, in IgAN relapse four of six individuals IgA during remission or relapse compared to controls. The showed rises in IgA production, and this was particularly great majority of salivary IgA will be secretory IgA (dimer IgA

striking in two (Fig. 3). But this was not the case in the two + SC), but the assay used in this study does not distinguish other IgAN relapses studied, and furthermore two of five secretory IgA from dimer IgA (without attached SC). It is

930

Feehally ci a!

interesting to speculate that excess of polymer IgA in the serum may be a reflection of impaired interaction between dimer IgA

13.

produced by mucosal lymphoid tissue in response to antigen challenge and the SC dependent mechanism which transports IgA to the mucosal surface, allowing dimer IgA to be freed into the serum. Such an abnormality might be indicated by abnormal amounts of free SC and dimer IgA in saliva. These sequential data support the contention that episodes of macroscopic hematuria in IgAN are associated with abnormal

activation of the IgA system, while no abnormalities were detected in the same subjects studied during remission. Such activation (indicated by increased helper/suppressor T cell ratio, increased igA-B cells, and increased PWM driven IgA production) may lead to the increases in circulating polymer IgA. Recent studies suggest that mesangial IgA in IgAN is polymeric 3—7], so deposition of such circulating polymer IgA

may be an important mediator of glomerular damage and hematuria. Acknowledgments Feehally and B.M. Coupes received financial support from the

J.

Manchester and North West Region Kidney Research Association. We are grateful to Professor lB. Houston and Dr. R. Gokal for allowing us to study patients tinder their care. Aspects of this work were presented at the lXth International Congress of Nephrology, Los Angeles, 1984.

Reprint requests to Dr. J. Feehally, Department of Leicesier

IVephrology,

General Hospital, Gwendolen Road, Leicester LE5 4PW,

lgA2 in circulating immune complexes and in renal deposits of Berger's and Henoch—SchOnlein glomerulonephritis. Proc EDTA 19:648—654, 1982

14. CoPo R, BASOLO B, MARTINA U, R0LLIN0 C, DE MARCI-II M, GIACCHINO F, MAZZUCCO 0, MESSINA M, PICCOLI U: Circulating

immune complexes containing IgA, IgG and 1gM in patients with primary lgA nephropathy and with Henoch—SchOnicin nephritis. C/in Nephrol 18:230—239, 1982 15. LESAVRE P, DIGEON M, BACH JF: Analysis of circulating IgA and

detection of immune complexes in primary IgA nephropathy. Clin Exp Immunol 48:61—69, 1982 16. STACHURA I, SINGH 0, WHITESIDE TL: Immune abnormalities in IgA nephropathy. C/in J,nmuno/ linmunopaihol 20:373—388, 1981 17. EGIDO J, SANCHO J, RIVERA F, HERNANDO L: The role of IgA and

lgG immune complexes in IgA nephropathy. IVephron 18.

BERGER J, HINGLAIS N: Les depOts intercapillaires d'IgA-IgG. J Urol Nephro/ 74:694—695, 1968 2. CLARKSON AR, WOODROFFE AJ, SEYMOUR AE: IgA nephropathy, in Recent Advances in Renal Medicine 2, edited by PETERS DK, JONES NF, London, Churchill Livingstone, 1982, pp. 167—175 3. Bi MC, FAUR U. DUHEILLE J: IgA nephropathy: CharacterisaI.

tion of the polymeric nature of mesangial deposits by in vitro binding of free secretory component. C/in Exp Irnrriunol 47: 527—534, 1982 4. Dorsirn V, CASANOVA C, MALETTA G, ZUCCHELLI P: The presence

of J chain in mesangial immunodeposits of IgA nephropathy. Proc EDIA 19:655—662, 1982 5. KOMATSU N, NAGURA H, WATANABE K, OMOTO Y, KOBAYASHI

K: Mesangial deposition of J

chain—linked

polymeric IgA in IgA

nephropathy, Nephron 33:61—64, 1983 6. TOMINO Y, SAKAI H, MIURA M, ENDOH M, NOM0TU Y: Detection

Med 74:375—381, 1983 20. EGIDO J, BiAsco R, SANCHO J, HERNANDO L: Abnormalities of

immune regulation in patients with igA nephropathy. Proc EDTA 17:660—667, 1982

21. LINNE T, WASSERMAN J: Lymphocyte subpopulations and ig production in vitro in IgA nephropathy. (abstract) mt j I'ediatr Nephro/ 3:146, 1982 22. ROCCATELLO D, Coo R, BASOLO B, MARTINA G, ROLLINO C, CORDONNIER D, BUSQUEI G, PIccI0Tro G, SENA LM, PICCOLI U:

Interaction between the macrophage system and IgA immune complexes in IgA nephropathy. Proc EDTA 20:610—616, 1983 23. SARA! H, ENDOH M, T0MIN0 Y, NOMOTO Y: Increase of IgA specific helper 1' alpha cells in patients with IgA nephropathy. Clin Exp Immunol 50:77—82, 1982 24. ENDOFI M, SARA! H, T0MIN0 Y, NOMOTO Y: Increase of periph-

eral blood B cells with Fe receptor for IgA in patients with IgA nephropathy. Scand J irnrnunol 17:437—441, 1983 25. SANN G, SCHNEIDER U, LOEKE S, DOERR HW: Rapid fractionation

of serum immunoglobulins by high pressure liquid gel permeation

chromatography. Application to routine serologic procedures. J ImmunoiMeth 59:121—127, 1983 26. ALNO—NAKHAL M, SAINT ANDRE JP, HoussIN A, HUREZ D: Increase or not in the number of B lymphocyte carriers of IgA in Berger's disease. (letter) Nouve Presse Med 10:2583, 1981 27. Cosio FG, LAM S, FOLAMI AO, CONLEY ME, MICHAEL AF: immune regulation of immunoglobulin production in lgA ncphropathy. C/in Immunol Immunopathol 23:430—436, 1982 28. BENE MC, FAURE 0, HURAULT DE, LIGNY B, KESSLER M, DUHEILLE i: IgA nephropathy.Quantitative immunohistomorphom-

etry of the tonsillar plasma cells evidences an inversion of IgA

of polymeric IgA in glomeruli from patients with tgA nephropathy. C/in Exp Immunol 49:419—425, 1982 7. EGIDO J, SANCHO J, MAMPASO F, LOPEL—TRASCASA M, SANCHEZ

CRESPO M, BLASCO R, HERNANDO L: A possible common patho-

versus IgG secreting cell balance. J C/in Invest 71:1342—1347, 1983 29. EGIDO J, BLASCO R, LOZANO L, SANCHO I, UARCIA—HOYO R: Immunological abnormalities in the tonsils of patients with igA

genesis of the mesangial IgA glotnerulonephritis in patients with Berger's disease and SchOnlein—Henoch syndrome. Proc EDTA

Inversion in the ratio of IgA: lgG bearing lymphocytes and increased polymeric IgA synthesis. C/in Li-p Iinnunol nephropathy:

17:660—667, 1980

8. DOBRIN RS, KNUDSON FE, MICHAEL AF: The secretory immune system and renal disease. C/in Exp Iminunol 21:318—328, 1975 9. WHITWORTH JA, LEIBOWITZ S, KENNEDY MC, CAMERON JS, CHANTLER C: IgA and glomcrular disease, C/in Nephrol 5:34—37, 1976

57:101—106, 1984

30. CHATENOUD L, BACH MA: Abnormalities of 'I' cell subsets in glomerulonephritis and systemic lupus erythematosus. Kidney Int 20:267—274, 1981 3!. FORNASIERI A, SINICO R, FIORINI U, GOLDANIGA D, COLASANTI

U, VENDEMIA F, GIBELIl A, D'AMICO U: T lymphocyte subsets in

10. COUSER WG, SALANT IJJ: In situ immune complex formation. Kidney

mt

primary and secondary glomerulonephritis. Proc EDTA 19:635—

17:l-l3, 1980

Ii. MAUER SM, SUTHERLAND Ri, HOWARD AJ, FISH AJ, NAJARIAN iS, MICHAEL AF: The glomerular mesangium. 111: Acute mesangial injury: A new model of glomerulonephritis. J Exp Med 137: 553—570, 1973

12. LOPEZ—TRASCASA M, EGIDO J, SANCHO i, HERNANDO L: IgA

glomeruloncphritis (Berger's disease): Evidence of high serum

36:52—59,

1984 HALL RP, STACHURA I, CASON J, WHITESIDE TL, LAWLEY Ti:

IgA-containing circulating immune complexes in patients with tgA nephropathy. Am J Med 74:56—63, 1983 19. VALENTIJN RA: Presence of circulating macromolceular IgA in patients with haematuria due to primary IgA nephropathy. Am J

United Kingdom.

References

levels of polymeric IgA. C/in Exp Irnrnuno/ 42:247—254, 1980 Coo R, BAsoLo B, MAZZUCCO U, BULZONI MR, ROCCATELLO D, MESSINA M, MARTINA G, ROLLINO C, Piccou U: IgAl and

641, 1982

32. SARA! H, N0MOlo Y, ARIMORI S: Decrease of IgA specific

suppressor T cell activity in patients Exp Immunol 38:243—248, 1979 33.

with igA nephropathy. C/in

Tissue origins of human polymeric and monomeric IgA. J I,nmunol 128:990—995, 1982

KUTTEH Wi-I, PRINCE SJ, MESTECKY J:

931

IgA system in relapsing IgA nephropathy 34. THOMPSON AJ, CHANY C, W000ROFFE AJ, CLARKSON AR, SEY-

MOUR AE: Vascular IgA deposits in clinically normal skin of patients with renal disease. Pathology 12:407—413, 1980

35. LAGRUE G, HURBEC G, FOURNEL M, 1NTRArOR L: Glomerulone-

phrites a depots d'IgA etudes des immunoglobulines sériques. J Urol Nephrol 80:385—388, 1974 36. MUSTONEN J, PASTERNACK A, HELIN H, RILVA A, PENTTINEN K,

WAGER 0, HARMOINEN A: Circulating immune complexes, the concentration of serum IgA and the distribution of HLA antigens in IgA nephropathy. Nephron 29:170—175, 1981 37. WOODROFFE AJ, GORMLEY AA, MCKENZIE PE, WOOTTON AM, THOMPSON AJ, SEYMOUR AE, CLARKSON AR: Immunological

studies in IgA nephropathy. Kidney mt 18:366—374, 1980 38. NEWKIRK MM, KLEIN MH, KATZ A, FISHER MM, UNDERDOWN BJ: Estimation of polymeric IgA in human serum. J Immunol

130:1176—1181, 1983

39. DELACROIX DL, ELKON KB, GEVBEL AP, H0DG50N HF, DIVE C, VAERMAN JP: Changes in size, subclass and metabolic properties of

serum immunoglobulin A in liver disease and other disease with high serum immunoglobulin A. J Clin Invest 71:358—367, 1983 40. DELACROIX DL, DEHENNIN JP, VAERMAN JP:

Influence of molecII Solid

ular size of IgA on its immunoassay by various techniques.

phase radioimmunoassays. J

41. PAGANELLI containing

Immunol Meth 48:327—337, 1982

R, LEVINSKY RJ, BRO5TOFF J: Immune complexes

food proteins in normal and atopic subjects after oral

challenge and effect of sodium cromoglycate on antigen absorption. Lancet i: 1270—1272, 1979 42. CAIRNS SA, LONDON RA, MALLICK NP: Circulating immune complexes following food: Delayed clearance in idiopathic glomerulonephritis. J Clin Lab Med 21:507—512, 1982