Skip to main content
NCBI home page
Clinical and Experimental Immunology logoLink to Clinical and Experimental Immunology
. 1998 Dec;114(3):468–476. doi: 10.1046/j.1365-2249.1998.00745.x

Imbalances in serum proinflammatory cytokines and their soluble receptors: a putative role in the progression of idiopathic IgA nephropathy (IgAN) and Henoch–Schönlein purpura nephritis, and a potential target of immunoglobulin therapy?

G ROSTOKER *,†, J -C RYMER ‡, G BAGNARD ‡, M PETIT-PHAR †, M GRIUNCELLI †, Y PILATTE †
PMCID: PMC1905142  PMID: 9844059

Abstract

Following recent experimental data suggesting an aggravating effect of circulating proinflammatory cytokines on the histological lesions of IgAN, we studied changes in serum proinflammatory cytokines and their soluble receptors and antagonists in patients treated with polyvalent immunoglobulins (15 with severe nephropathy who had indicators of poor prognosis: heavy proteinuria, hypertension, altered renal function and Lee's histological grade III or IV; and 14 with moderate forms of IgAN who had permanent albuminuria > 300 mg/day and < 2000 mg/day, Lee's histological grade II and a glomerular filtration rate > 70 ml/min) in comparison with healthy controls (n = 20) and patients with non-IgA nephritides (n = 50). These were measured by means of specific immunometric assays before and after 9 months of immunoglobulin therapy. Total tumour necrosis factor (TNF) serum and IL-6 levels were elevated in IgAN patients before therapy, relative to controls, and normalized after immunoglobulin therapy. Levels of soluble TNF receptor of type I (sR55) and type II (sR75) increased on immunoglobulin therapy. TNF index α-55,75 used to assess biologically available TNF-α (ratio of total TNF-α divided by levels of soluble TNF receptors sR55 and sR75) was elevated before therapy and was below healthy control values after 9 months of immunoglobulin administration. Levels of serum IL-1 receptor antagonist were low prior to immunoglobulin administration in patients with severe forms of IgAN, and normalized on therapy. Serum interferon-gamma was unmodified. The histological activity index correlated with serum total TNF-α, TNF index α-55,75 and serum IL-6 levels, whereas proteinuria correlated with serum total TNF-α and TNF index α-55,75 but not with serum IL-6. These data suggest that the overproduction of proinflammatory cytokine is unbalanced by their natural antagonists in IgAN and Henoch–Schönlein syndrome. This process may play a role in the progression of the disease and be one of the targets of immunoglobulin therapy.

Keywords: IgA nephropathy, Henoch–Schönlein syndrome, immunoglobulin therapy, tumour necrosis factor-alpha, TNF soluble receptors

INTRODUCTION

IgAN is the most common form of primary glomerulonephritis world-wide and is characterized by the presence of IgA and C3 deposits in the mesangium [1]. About 25% of patients progress slowly to chronic renal failure 10 years after diagnosis, rising to 40–50% after 20 years [1]. Idiopathic Henoch–Schönlein purpura (HSP) is a severe systemic form of IgAN [1]. Although renal impairment develops slowly in most patients, a subset of patients with heavy proteinuria and high histological grade have a more rapid disease course that necessitates dialysis within 1–5 years [1]. Treatment of focal sepsis and hypertension, especially with angiotensin-converting enzyme inhibitors, is the only therapeutic measure with proven value [1]. Despite the fact that glomerular injury is immunologically mediated, IgAN and HSP do not respond to steroids or immunosuppressive agents [1]. Recently, we obtained evidence that immunoglobulin therapy might be beneficial in the active form of these nephropathies, as in many other immunologically mediated diseases poorly responsive to conventional drugs [2–4]. The efficacy of immunoglobulin therapy in immunological disorders has been tentatively attributed to the modulation of idiotypic and isotypic networks, solubilization of immune complexes, and prevention of active C3 fragment binding to target surfaces [5]. However, in recent years work has focused on modulation of monokine production [6]. The causes of IgAN are still poorly understood. Glomerular damage might be related to deposition of IgA-containing immune complexes and, possibly, IgA with various autoantibody activities within the glomerulus; this might result primarily from an abnormal IgA medullary system or from an uncontrolled mucosal response to foreign antigens [7]. Following recent evidence that circulating proinflammatory cytokines may play an aggravating role in the lesions induced by infused preformed IgA immune complexes in an experimental model of IgAN [8], and as the role of these cytokines (especially tumour necrosis factor-alpha (TNF-α)) is now strongly suspected in the induction or aggravation of renal damage in various glomerulonephritides [9], we investigated the time course of serum proinflammatory cytokines, together with their soluble receptors and the various types of IgA-containing immune complexes, in patients on immunoglobulin therapy.

SUBJECTS AND METHODS

Subjects

The subjects comprised 15 patients with severe IgAN who received high-dose immunoglobulin therapy (idiopathic IgAN n = 13; HSP n = 2; these patients had indicators of poor prognosis: high histological grade (stage III or IV of Lee's classification) [10], heavy proteinuria (> 2 g/day), decline in renal function) (2 g/kg each month of pepsin pH 4 i.v. immunoglobulins for 3 successive months, followed by intramuscular immunoglobulins (preparation content 16.5%; 0.35 ml/kg every 15 days) for another 6 months) [2]; 14 patients with moderate forms of IgAN (idiopathic IgAN n = 11, HSP n = 3; permanent albuminuria > 300 mg/day and < 2000 mg/day, and a glomerular filtration rate > 70 ml/min, and an intermediate histological grade: Lee's histological stage II) [10] who received low-dose immunoglobulin (IMIG) administered as recently reported (0.35 ml/kg of a preparation, content 16.5%, once a week for 1 month, followed by 0.35 ml/kg every 15 days for a further 8 months) [3]; 20 patients with minimal change nephropathy (INS), 20 with idiopathic membranous nephropathy (IMN) and 10 with idiopathic type I membranoproliferative glomerulonephritis (MPGN). Each of these diseases was diagnosed according to standard histological criteria, and patients with glomerulonephritis secondary to systemic or infectious diseases were excluded. All the patients were adults and were studied during an active phase of the disease, before treatment with steroids and immunosuppressive drugs (INS, IMN), angiotensin-converting-enzyme inhibitors or aspirin (IMN, MPGN), and immunoglobulin (IgAN). Samples were also taken after the completion of 9 months of immunoglobulin therapy in patients with IgAN. The control group consisted of 20 healthy volunteers from the hospital staff and medical students, matched for age and sex. All the patients gave informed consent to the study, which had received Local Ethical Committee approval.

Samples

Venous blood was drawn after an overnight fast into standard sterile polystyrene vacuum tubes, allowed to clot at 4°C for 2 h, then centrifuged at 4°C. Blood for plasma isolation was drawn into 5 mm EDTA in endotoxin-free tubes, stored at 4°C and centrifuged within 90 min at 4°C. Serum and plasma samples (1 ml) were aliquoted and stored at −20°C. Samples were thawed only once. The subjects were free of fever, infection and shock at the time of sampling, which took place the day before the first i.v. infusion of immunoglobulin in patients with IgAN treated with high-dose immunoglobulin, the day before the first intramuscular administration of IMIG in patients with moderate forms of the disease; further samples were taken 15 days after the last intramuscular administration of IMIG (month 9 of immunoglobulin therapy) in both groups of patients. Assays were always done on duplicate samples.

Histological activity index of IgAN

We noted proliferation of mesangial and epithelial glomerular cells blindly using a final scale of 14 points [2]. In evaluating mesangial cell proliferation, we accounted for the intensity (absent = 0; mild = 1; moderate = 2; severe = 3) and extent of the lesions (no glomeruli affected = 0; < 25% of glomeruli = 1; > 25% but < 50% = 2; > 50% but < 75% = 3; > 75% = 4). We noted epithelial cell proliferation in the same way [2]. Immunofluorescence was scored blindly from 0 to 3.

Biochemical parameters of renal disease activity

They included: 24-h albuminuria, urinary erythrocyte and leucocyte counts, plasma creatinine. Global renal function was assessed using the formula of Cockroft & Gault [11] and clearance of an inulin analogue (polyfructosan).The rate of decline in renal function was evaluated as previously described: loss in glomerular filtration rate during the period studied divided by the number of months (ml/min per month per 1.73 m2) [2,3].

Determination of plasma cytokine levels

Cytokines were measured on endotoxin-free plasma, to avoid false-positive results related to cytokine release in the device [12], by means of specific immunometric assays according to the manufacturers' recommendations, as follows: total TNF-α (TNF-α bound and unbound to soluble TNF receptors type I and type II) detection limit 5 pg/ml (Medgenix, Fleurus, Belgium); interferon-gamma (IFN-γ), detection limit 0.2 U/ml (Medgenix); IL-6, detection limit 6 pg/ml (Medgenix); and IL-1β, detection limit 1 pg/ml (Immunotech, Marseille, France). Detection limit values of the kits were obtained from the manufacturers. Results are expressed in pg/ml for TNF-α, IL-6 and IL-1β and in U/ml for IFN-γ. IL-1β and IFN-γ were not related to renal function, but TNF-α and IL-6 levels were adjusted for the glomerular filtration rate (GFR), since their levels increase in renal failure [13–16]. Results for these two cytokines were therefore calculated as follows when GFR was < 100 ml/min: cytokine concentration related to GFR = (cytokine concentration (pg/ml) × GFR determined according to Gault & Cockroft)/100.

Serum cytokine soluble receptor and antagonist assays

Levels of soluble cytokine receptors and antagonists were assayed in serum by means of specific immunometric assays according to the manufacturers' instructions, as follows: IL-1 receptor antagonist (IL-1Ra) (detection limit 6.5 pg/ml), soluble TNF receptor type I (sR 55) (detection limit 50 pg/ml) and soluble TNF receptor type II (sR 75) (detection limit 0.1 ng/ml) (Medgenix). Detection limit of the kits were obtained from the manufacturers. As cytokine soluble receptor and antagonist levels increase when renal function deteriorates [14], results (expressed in ng/ml or pg/ml) were adjusted for renal function when GFR was < 100 ml/min, as follows: cytokine soluble receptor or antagonist concentration related to GFR = soluble receptor or antagonist concentration (pg or ng/ml) × GFR determined according to Gault & Cockroft/100.

Determination of TNF-α index 55,75

Biological available TNF-α was theoretically estimated according to Dörge et al., with slight modifications [17], as follows: Index α-55,75 = TNF-α/sR 55 + sR 75 × GFR determined according to Gault & Cockroft/100.

Results of this index are expressed in arbitrary unit/ml.

Evaluation of immunoglobulin therapy

Parameters reflecting renal disease evolution, histological activity, and immunological activation were compared before and after 9 months of immunoglobulin therapy, as previously reported [2,3].

IgA immune complexes

Large and intermediate size IgA immune complexes were measured according to Imai et al. [18] after precipitation of sera with 14% (w/v) polyethylene glycol (PEG) 6000 (7% final concentration) in specific sandwich ELISAs. IgA1, IgA2 and secretory IgA-containing immune complexes were measured using sandwich ELISAs in which the specific capture antibodies were clone NI 69–11 for IgA1, clone NI 512 for IgA2, and clone NI 194, a mouse MoAb directed against bound and free secretory component (Nordic, Tilburg, The Netherlands) [19]. The specificity of these MoAbs has been determined elsewhere [20] and confirmed in our laboratory by means of indirect and sandwich ELISAs with a panel of purified immunoglobulins [19]. The revealing antibody was alkaline phosphatase-labelled affinity-purified anti-human IgA goat IgG (mouse protein-adsorbed) (Zymed, San Francisco, CA). Its specificity had been checked by indirect ELISA with purified immunoglobulins [19]. Purified monoclonal IgA1 and IgA2 and secretory IgA (human colostrum; Jackson, West Grove, PA) were used as standards in each plate.

Mixed IgA1-IgG, IgA2-IgG, IgA1-IgM and IgA2-IgM complexes were measured by sandwich ELISAs with capture antibodies consisting of clone NI 69–11 (IgA1) or clone NI 512 (IgA2); the revealing antibodies were, respectively, alkaline phosphatase-labelled affinity-purified anti-human IgG goat IgG and anti-human IgM goat IgG (mouse protein-adsorbed) (Jackson). The specificity of these revealing antibodies was checked by indirect ELISAs with a panel of purified immunoglobulins. Results for these mixed IgA-containing immune complexes are expressed in optical density (OD) units.

To allow comparison, positive and negative precipitates and a pooled precipitate from healthy donors with known OD values were run in each plate for each type of IgA immune complex.

IgA–fibronectin complexes were assayed according to the manufacturer's recommendations with the Euro-Diagnostica ELISA kit (Malmö, Sweden) modified from the assay initially designed by Cederholm et al. [21]. Results were expressed in OD units.

Serum IgA was measured by means of nephelometry with a specific polyclonal antiserum (Behring Institute, Marburg, Germany).

Statistical analysis

As values were not normally distributed, we performed non-parametric analysis of variance with the Kruskal–Wallis test (GraphPad Instat, San Diego, CA). In cases of significant difference, immunological parameters in the other types of glomerulonephritides and in IgAN before and after therapy were compared with those of healthy controls, by using the non-parametric Mann–Whitney test (GraphPad Instat). We analysed paired variables in IgAN patients on immunoglobulin therapy using the Wilcoxon non-parametric test (GraphPad Instat). Variables are expressed as medians and range. Correlations between variables were analysed using the non-parametric Spearman test (GraphPad Instat). Differences with P values < 0.05 were considered significant.

RESULTS

Clinical effects of immunoglobulin therapy

Patients with severe form on high-dose immunoglobulins

There was a significant decrease in proteinuria (median before 3.54 g/day; after 1.21 g/day; P < 0.0005), haematuria (before 56 470/min; after 15 405/min; P < 0.005) and leukocyturia (before 8578/min; after 3477/min; P < 0.05). The decline in the glomerular filtration rate was greatly slowed down or stopped (median rate of decline in glomerular filtration before intervention: −1.5 ml/min per month; after 0/ml per min per month; P < 0.005). There was also a significant decrease in the histological index of activity (median before 5; after 2.5; P < 0.001) and a fall in the staining intensity of glomerular IgA and C3 deposits (median IgA before 2; after 1; P < 0.01) (median C3 before 2; after 0.5; P < 0.001). Systemic symptoms in the patients with HSP disappeared within 2 months of starting IgIV therapy.

Patients with moderate form on low-dose immunoglobulins

IMIG therapy induced a significant decrease in albuminuria (median before 766 mg/24 h; after 171; P < 0.01), but no change in glomerular filtration rate or urinary erythrocyte and leucocyte counts. Systemic symptoms in the three patients with HSP disappeared within 2 months of starting IMIG therapy. IMIG therapy also induced a significant decrease in the histological index of activity (median before 4; after 0; P < 0.005), but no change in the histological sclerosis index. No change in the intensity of IgA or C3 deposits was observed.

Serum TNF-α and soluble receptors

Total TNF-α serum levels were clearly elevated in the IgAN patients compared with the controls, and normalized during immunoglobulin administration (Table 1 and Fig. 1). Values in severe forms of the disease were much higher than in moderate forms. Patients with INS and IMN also had elevated total serum TNF-α levels, but less so than patients with IgAN (Table 1 and Fig. 1).

Table 1.

Serum tumour necrosis factor-alpha (TNF-α) and soluble receptors in IgAN on immunoglobulin therapy

graphic file with name cei0114-0468-t1.jpg

Open in a new tab

Fig. 1.

Fig. 1

Open in a new tab

Total serum tumour necrosis factor-alpha (TNF-α) in IgAN on immunoglobulin therapy. Values are expressed in pg/ml in box plots.

Levels of soluble TNF receptor type I (sR 55) were higher than control values only in patients with moderate forms of IgAN before therapy and remained high in this group on immunoglobulin therapy (Table 1). Both type I receptor (sR 55) and type II receptor (sR 75) increased in severe IgAN during high-dose immunoglobulin therapy (Table 1). Both types of TNF soluble receptor were elevated in INS, IMN and MPGN (Table 1). Theoretical evaluation of biologically available TNF-α by the index α-55,75 showed elevated values before therapy in both severe and moderate forms of IgAN, and a return to levels lower than healthy control values after 9 months of immunoglobulin therapy (Table 1 and Fig. 2). This index was much higher in severe forms of IgAN than in moderate forms (Table 1 and Fig. 2). The index α-55,75 was significantly lower in the other histological types of glomerulonephritides than in healthy controls (Table 1 and Fig. 2).

Fig. 2.

Fig. 2

Open in a new tab

Tumour necrosis factor (TNF) index α-55–75 in IgAN on immunoglobulin therapy. Values are expressed in arbitrary units/ml in box plots.

Serum IL-1 and IL-1Ra

Serum IL-1β levels did not differ from those of healthy controls in patients with IgAN, before or during therapy, as in the other nephritides (Table 2). Levels of serum IL-1 receptor antagonist were low in patients with severe forms of IgAN prior to immunoglobulin administration, and normalized after 9 months of therapy; low values of IL-1Ra were also observed in MPGN (Table 2).

Table 2.

Serum IFN-γ, IL-6, IL-1 and IL-1Ra (IL-1 receptor antagonist) in IgAN and other primary glomerulonephritis

graphic file with name cei0114-0468-t2.jpg

Open in a new tab

Serum IL-6 and IFN-γ

Serum IL-6 levels were elevated in both groups of IgAN patients before therapy and normalized during immunoglobulin administration (Table 2 and Fig. 3). High serum IL-6 levels were also seen in all types of glomerulonephritides (Table 2 and Fig. 3). There was no difference in serum IFN levels among the different patients with nephritides and the healthy controls (Table 2). Furthermore immunoglobulin therapy did not modify IFN-γ levels in IgAN patients (Table 2).

Fig. 3.

Fig. 3

Open in a new tab

Serum IL-6 in IgAN on immunoglobulin therapy. Values are expressed in pg/ml in box plots.

Correlations between proinflammatory cytokines, their soluble receptors and renal parameters

In IgAN the histological activity index correlated with serum total TNF-α (r = 0.46; P < 0.001), TNF index α-55,75 (r = 0.44; P < 0.005), and serum IL-6 (r = 0.28; P < 0.05), whereas proteinuria correlated with serum total TNF-α (r = 0.27; P < 0.05), TNF index α-55,75 (r = 0.27; P < 0.05) but not with serum IL-6. Total TNF-α, index α-55,75 and IL-6 were strongly intercorrelated in these patients (TNF-α/index α-55,75: r = 0.91 (P < 0.005); TNF-α/IL-6, r = 0.65 (P < 0.001)). In IgAN, serum IL-1Ra and IL-1β correlated negatively only after therapy (r = − 0.48; P < 0.01), whereas TNF-sR 55 correlated strongly with TNF-sR 75 both before (r = 0.85; P < 0.001) and after immunoglobulin therapy (r = 0.78; P < 0.001). Haematuria, leucocyturia and decline in glomerular filtration rate did not correlate with serum cytokines and their soluble receptors in IgAN.

TNF-sR 55 and TNF-sR 75 correlated in the other glomerulonephritides (IMN r = 0.57, P < 0.05; INS r = 0.80, P < 0.005; MPGN r = 0.35, P < 0.5). TNF-sR 55 and TNF-sR 75 also correlated in the healthy controls (r = 0.67, P < 0.005). Total TNF-α correlated strongly with index α-55,75 (IMN r = 0.94, P < 0.001; MPGN r = 0.76, P < 0.01; INS r = 0.64, P < 0.05; healthy controls r = 0.69, P < 0.01). The index α-55,75 did not correlate with sR 55 or sR 75 in any of these nephritides. Proteinuria did not correlate with serum IL-1β, and IL-6, total serum TNF-α, or index α-55,75 in any of these non-IgA nephritides.

IgA immune complexes

IgA1-containing immune complexes (IC) were elevated only in moderate IgAN prior to immunoglobulin therapy (P < 0.05 compared with healthy controls; Mann–Whitney test), and returned to normal during treatment (P < 0.05 compared with pretherapeutic values; Wilcoxon test) (IgA1-IC, median (range) in ng/ml: severe IgAN, before therapy 1500 (25–4800), after 1400 (25–3500); moderate IgAN, before therapy 2550 (125–4300), after 1400 (0–4100); healthy controls, 1900 (1250–4200)). We found no significant difference between IgAN patients before and after therapy or between the same patients before therapy and the healthy controls with respect to IgA1-IgG, secretory IgA or IgA2-IgG immune complexes (data not shown). IgA2, IgA1-IgM and IgA2-IgM IC values were lower than in the controls in the two groups of IgAN before therapy (P < 0.05, Mann–Whitney test) (median IgA2 in ng/ml: severe IgAN, 32; moderate IgAN, 62; healthy controls, 500) (IgA1-IgM, median OD unit × 100/ml: severe IgAN, 12; moderate, 16; healthy controls, 41.5) (IgA2-IgM, median OD unit × 100/ml: severe IgAN, 17.5; moderate, 15; healthy controls, 27). IgA1-IgM and IgA2-IgM complexes increased during immunoglobulin administration, suggesting a modulation of the idiotypic network (P < 0.05, Wilcoxon test) (IgA1-IgM, median OD unit × 100/ml: severe IgAN, 20; moderate IgAN, 20) (IgA2-IgM, median OD unit × 100/ml: severe IgAN, 31; moderate IgAN, 19). Patients with other glomerulonephritides had normal or low levels of IgA1, IgA1-IgG, IgA1-IgM, secretory IgA, IgA2, IgA2-IgG and IgA2-IgM IC relative to the healthy controls (data not shown).

IgA–fibronectin aggregates in serum were elevated in the patients with severe IgAN before immunoglobulin therapy (P < 0.05 compared with healthy controls; Mann–Whitney test) and normalized during treatment (P < 0.05 compared with pretherapeutic values; Wilcoxon test) (median (range) in OD unit × 100: severe IgAGN, before therapy 35.75 (1.5–141), after 21.8 (2–103); moderate IgAN, before therapy 15.7 (3.1–80), after 11.5 (1.1–24.10); healthy controls, 24.10 (15–37.5)). Patients with other glomerulonephritides had normal or low levels of IgA–fibronectin aggregates in serum relative to the healthy controls (data not shown).

In the patients with IgAN, none of the various IgA IC or IgA–fibronectin aggregates correlated with each other or with serum IL-6, TNF-α, TNF index α-55,75, histological activity, proteinuria, leukocyturia or the decline in renal function. Serum IgA correlated positively with IgA1-containing IC (r = 0.50, P < 0.0001) but not with IgA–fibronectin aggregates or with other types of IgA immune complexes.

DISCUSSION

There is now increasing evidence that dysregulated cytokine production plays a crucial role in human autoimmune and inflammatory processes [21]. Many cytokines are released in inflamed glomeruli and interstitium by infiltrating leucocytes and resident renal cells, and can therefore mediate glomerular and interstitial injury [9]. The potential pathological role of circulating substances released by immune cells in glomerular diseases has recently been borne out by the fact that immunoadsorption on immobilized protein A reduces the frequency of proteinuric recurrences of segmental glomerulosclerosis on grafts, and the ability of a 100-kD protein eluted from this protein A column to induce albuminuria when administered intravenously and intraperitoneally to rats [22].

TNF-α is a primary mediator in the pathogenesis of infection and inflammation, and plays a central role in host defences and tissue homeostasis [16]. TNF-α interacts with two high-affinity transmembrane receptors of 55 and 75 kD found on numerous cell types, triggering its biological effects [16]. The extracellular domain of TNF membrane receptors can be shed by proteolytic cleavage, giving rise to TNF-binding proteins known as soluble receptors sR 55 and sR 75 [16]. There is now considerable evidence that these TNF-binding proteins act as a ‘buffer’, protecting against harmful local and systemic effects of TNF. TNF-binding proteins act by competing with cell surface receptors for TNF-α and -β; most of the biological effects of TNF are blocked both in vitro and in vivo by sR 55 and sR 75 [16]. TNF-binding proteins are found at high concentrations in the plasma of healthy individuals and function as physiological anti-cytokines by forming high-affinity complexes with TNF [16].

TNF-α shares a wide range of biological activities with IL-1, in that it plays a pivotal role in septic shock and cachexia, and may contribute to the pathogenesis and progression of rheumatoid arthritis, graft-versus- host disease and allograft rejection [16]. An imbalance between TNF-α and its soluble inhibitory binding proteins in serum has recently been forwarded as playing a central pathophysiological role in severe meningococcaemia, cerebral malaria and polyneuropathy, organomegaly, endocrinopathy, M protein and skin changes (POEMS) [16,24,25]. Administered intravenously to rabbits, human recombinant TNF-α induces endothelial damage, and leucocyte and fibrin accumulation in the glomerular capillary lumen [26]. In vitro, TNF-α stimulates the growth of epithelial glomerular cells in culture [27]. TNF-α may be cytolytic for mesangial cells at high concentration and exerts significant growth-stimulatory effects at lower concentrations [28]. Systemic administration of TNF-α aggravates a passive model of anti-glomerular basement membrane antibody-mediated nephritis in rats [29] and acute IC glomerulonephritis in rabbits [30], while anti-TNF antiserum and soluble TNF receptors ameliorate glomerular lesions and proteinuria in an autologous form of nephrotoxic serum nephritis [31,32]. In this study we found increased serum TNF-α levels in IgAN patients before immunoglobulin administration, and their normalization on therapy. Calculation of TNF index -55,75 suggests that this increase might be related to free, biologically available TNF-α. This remains to be demonstrated by tests measuring TNF-α biological activity. Its magnitude was clearly higher in severe forms, while other glomerulonephritides showed a slight increase in total TNF, in keeping with previous reports [9]. Synthesis of soluble proteins sR 55 and sR 75 was enhanced and matched the increase in TNF in non IgA-nephritides and was adapted to this biological situation. Patients with IgAN did not overproduce these natural TNF inhibitors, pointing to a defect in this physiological mechanism counterbalancing TNF-α overproduction. As both groups of IgAN patients had increased levels of TNF-α, and as macrophage infiltrates of renal tissue expressing TNF are generally seen only in severe nephropathies [1,33], it is tempting to postulate that a large proportion of serum TNF originates from circulating mononuclear cells, which synthesize increased amounts of TNF-α in IgAN [34]. Interestingly, TNF-α and index α-55,75 correlated with the histological activity index, which took into account mesangial and epithelial proliferation. Our results are in keeping with the recent reports on the increase of serum TNF in HSP and detection of high levels of urinary TNF-α in IgAGN [35,36]. Unlike other proinflammatory cytokines, data are unavailable on the possible potentiation of glomerular lesions by infused TNF-α in the passive model of IgAN studied by Montinaro et al. [8], but the similarities between IL-1β and TNF-α in terms of both their biological properties [16] and their ability to enhance glomerular damage in various experimental models of non IgA nephritides [9,29,30] suggest that systemic TNF may also increase the nephritogenicity of IgA IC and worsen glomerular lesions in this nephropathy. Our hypothesis is consistent with the recently described relationship between serum TNF-α and glomerular expression of the cellular adhesion molecule ICAM-1 in human glomerulonephritides [37]. The normalization of serum TNF levels on therapy should be interpreted in the light of reports that i.v. immunoglobulins can down-regulate TNF-α production by mononuclear cells in vitro and in vivo [6,38]. The increases in TNF soluble receptors sR 55 and sR 75 after 9 months of therapy strongly suggest that long-term immunoglobulin administration can induce persistent release of TNF binding proteins, as previously shown in hypogammaglobulinaemic patients who received a single 0.4 g/kg infusion of i.v. immunoglobulin [39].

IL-1α and IL-1β share many of the biological properties of TNF-α, whereas IL-1Ra is a pure receptor antagonist with no agonistic activity in vitro or in vivo [16]. Recently, a local or systemic imbalance between IL-1 and its natural inhibitor IL-1Ra has been incriminated in the onset and progression of POEMS, rheumatoid arthritis and inflammatory bowel disease [16,25,40]. IL-1 stimulates human mesangial cell growth [28]. Administration of recombinant IL-1 causes an exacerbation in a passive model of IgAN, and passive anti-glomerular basement membrane (GBM) nephritis in rabbits [8,29], whereas infusion of IL-1Ra improves experimental anti-GBM nephritis by suppressing mesangial cell proliferation and reducing macrophage accumulation and crescent formation [41]. In human IgAN, circulating monocytes produce an excess of IL-1β [42]. In this study, whereas levels of IL-1β, which has a very short half-life in serum [12], were similar in IgAN to values in controls and other patients with primary glomerulonephritis, patients with severe forms had low levels of the natural IL-1 antagonist IL-1Ra prior to immunoglobulin therapy, pointing to an imbalance between IL-1 and IL-1Ra in this subset of IgAN which may play a role in the progression of renal lesions. Levels of IL-1Ra were normalized in these patients during therapy, raising the possibility that long-term administration of i.v. immunoglobulins (and possibly IMIG) can not only normalize IL-1 production by monocytes [6] but also promotes persistent synthesis of IL-1Ra [39].

A possible pathogenic role of circulating IL-6 in glomerulonephritis was suggested by high IL-6 serum level and mesangioproliferative nephritis in IL-6 transgenic mice [43]. IL-6 is also an autocrine growth factor for mesangial cells [44]. Increases in urinary IL-6 levels have been reported in active phases of IgAN and in chronically damaged kidneys [44,45]. IL-6 failed to potentiate histological lesions when administered alone intraperitoneally in a passive model of IgAN [8]. In contrast, co-administration with IL-1 increased the nephrotogenicity of IgA IC in this model, and induced severe MPGN [8]. Interestingly, circulating T cells and monocytes from patients with IgAN express more IL-6 mRNA and synthesize more IL-6 in vitro than cells from healthy controls [46]. The increases in serum IL-6 in both IgA and non-IgA nephropathies in this study point to a non-specific amplifying role of this cytokine, as suggested by the findings of Gordon and co-workers [45]. The normalization of IL-6 serum levels on immunoglobulin therapy suggests that IL-6 production is modulated in vivo, as evidenced in vitro [6], although commercial human immunoglobulin contains natural anti-IL-6 antibodies which may have accounted for this effect [47].

In a model of passive experimental IgAGN, intraperitoneally administered IFN-γ potentiates glomerular histological changes [8]. Yokoyama and co-workers recently found that serum levels of IFN-γ were elevated in IgAN patients with acute exacerbations triggered by infections; this was associated with over-expression of MHC class II antigens on renal biopsy specimens [48]. Our patients, who were free of infections, had similar serum IFN levels to healthy controls, which persisted during immunoglobulin therapy. These results suggest that increases in serum IFN-γ may play a role in infection-related clinical and histological flares-ups of the disease, but do not appear to be involved in the pathogenesis of glomerular injury in patients with permanent proteinuria.

Experimental models have shown that mesangial deposition of IgA IC is not sufficient to elicit glomerular damage, and that other factors such as extrarenal proinflammatory cytokines may modulate the renal response to nephritogenic antibodies and play a role in the onset of histological lesions [1,7,8]. Moreover, TNF-α-mediated proteinuria induced by adriamycin was also recently shown to promote mesangial deposition of preformed IgA IC in BALB/c female mice [49]. IgA1-containing IC were found to be increased only in moderate forms of IgAN, and normalized on low-dose immunoglobulin therapy. We observed an increase in IgA–fibronectin complexes in severe forms of IgAN which normalize during immunoglobulin therapy. We found no anomalies in the other types of IgA IC. IgA–fibronectin complexes explain the binding of IgA to collagens I, II and IV [21], and are strongly associated with IgAN but not with other glomerular diseases [50]. In experimental glomerulonephritis, IC containing fibronectin are preferentially located in the mesangium, leading some authors to suggest their role in mesangial deposition of IgA [51]. The decrease in IgA–fibronectin aggregates in severe forms of IgAN during immunoglobulin therapy might be due to the normalization of proinflammatory cytokines, as fibronectin is produced during inflammation and appears to have modulatory effects on leucocytes and inflammatory mediator production [52]. It is likely that the analysis of the properties of IgA IC might shed more light than single quantification for the understanding of their nephritogenicity and their potential modulation by immunoglobulin therapy.

In conclusion, our data suggest that the production of proinflammatory cytokines and their natural inhibitory proteins may be disrupted in IgAN, play a role in disease progression, and be one target of immunoglobulin therapy. The abnormalities may differ among subsets of patients, both in their magnitude and in counterbalancing mechanisms (soluble receptors and antagonists). If these hypotheses are confirmed, then new therapies aimed at rapidly neutralizing proinflammatory cytokines, e.g. IL-1Ra and anti-TNF strategies [53,54], may attenuate glomerular damage and fibrotic sequelae, especially in severe forms of IgAN where natural mechanisms aimed at buffering the noxious effects of TNF and IL-1 may be deficient.

References

  • 1.Galla J. IgA nephropathy. Kidney Int. 1995;47:377–87. doi: 10.1038/ki.1995.50. [DOI] [PubMed] [Google Scholar]
  • 2.Rostoker G, Desvaux-Belghiti D, Pilatte Y, et al. High-dose immunoglobulin therapy for severe IgA nephropathy and Henoch–Schönlein purpura. Ann Intern Med. 1994;120:476–84. doi: 10.7326/0003-4819-120-6-199403150-00005. [DOI] [PubMed] [Google Scholar]
  • 3.Rostoker G, Desvaux-Belghiti D, Pilatte Y, et al. Immunomodulation with low-dose immunoglobulins for moderate IgA nephropathy and Henoch–Schönlein purpura: preliminary results of a prospective uncontrolled trial. Nephron. 1995;69:327–34. doi: 10.1159/000188480. [DOI] [PubMed] [Google Scholar]
  • 4.Hammarströmm L, Lundkvist I, Petterson A, Edvard-Smith C. The use of IVIG in immunological disorders other than cytopenias. In: Yap PL, editor. Clinical applications of intravenous immunoglobulin therapy. Edinburgh: Churchill Livingstone; 1992. pp. 117–37. [Google Scholar]
  • 5.Nydegger U. Hypothetic and established action mechanisms of therapy with immunoglobulin G. In: Imbach P, editor. Immunotherapy with intravenous immunoglobulins. London: Academic Press; 1991. pp. 27–36. [Google Scholar]
  • 6.Andersson JP, Andersson UG. Human intravenous immunoglobulin modulates monokine production in vitro. Immunology. 1990;71:372–6. [PMC free article] [PubMed] [Google Scholar]
  • 7.Oliviera D. IgA related glomerulopathies. In: Oliviera D, editor. Immunological aspects of renal disease. Cambridge: University Press; 1992. pp. 87–113. [Google Scholar]
  • 8.Montinaro V, Hevey K, Aventaggiato L, et al. Extrarenal cytokines modulate the glomerular response to IgA immune complexes. Kidney Int. 1992;42:341–53. doi: 10.1038/ki.1992.295. [DOI] [PubMed] [Google Scholar]
  • 9.Ortiz A, Gomez-Chiarri M, Alonso J, et al. The potential role of inflammatory and fibrogenic cytokines in the glomerular diseases. J Lipid Med Cell Signalling. 1994;9:55–74. [PubMed] [Google Scholar]
  • 10.Lee K, Rao V, Franklin W, et al. IgA nephropathy: morphologic predictors of progressive renal disease. Hum Pathol. 1982;13:314–22. doi: 10.1016/s0046-8177(82)80221-9. [DOI] [PubMed] [Google Scholar]
  • 11.Cockroft D, Gault M. Prediction of creatinine clearance from the serum creatinine. Nephron. 1976;16:31–41. doi: 10.1159/000180580. [DOI] [PubMed] [Google Scholar]
  • 12.Whiteside T. Cytokine measurements and interpretation of cytokine assays in human disease. J Clin Immunol. 1994;14:327–39. doi: 10.1007/BF01546317. [DOI] [PubMed] [Google Scholar]
  • 13.Trinchieri G, Perussia B. Immune interferon: a pleiotropic lymphokine with multiple effects. Immunol Today. 1985;6:131–6. doi: 10.1016/0167-5699(85)90080-5. [DOI] [PubMed] [Google Scholar]
  • 14.Descamps-Latscha B, Herbelin A, Thu Nguyen A, et al. Balance between IL1β, TNFα and their specific inhibitors in chronic renal failure and maintenance dialysis. J Immunol. 1995;154:882–92. [PubMed] [Google Scholar]
  • 15.Herbelin A, Urena P, Thu Nguyen A, et al. Elevated circulating levels of interleukin-6 in patients with chronic renal failure. Kidney Int. 1991;39:954–60. doi: 10.1038/ki.1991.120. [DOI] [PubMed] [Google Scholar]
  • 16.Dayer JM, Burger D. Interleukin 1, tumor necrosis factor and their specific inhibitors. Eur Cytokine Netw. 1994;5:563–71. [PubMed] [Google Scholar]
  • 17.Dörge SE, Roux-Lombard P, Daher JM, et al. Plasma levels of tumor necrosis factor and soluble TNF receptors in kidney transplant recipients. Transplantation. 1994;58:1000–8. doi: 10.1097/00007890-199411150-00005. [DOI] [PubMed] [Google Scholar]
  • 18.Imai H, Chen A, Wyatt R, et al. Composition of IgA immune complexes precipitated with polyethylene glycol. A method for isolation and analysis of immune complexes. J Immunol Methods. 1987;103:239–45. doi: 10.1016/0022-1759(87)90295-x. [DOI] [PubMed] [Google Scholar]
  • 19.Rostoker G, Terzidis H, Petit-Phar M, et al. Secretory IgA are elevated in both saliva and serum of patients with various types of primary glomerulonephritis. Clin Exp Immunol. 1992;90:305–11. doi: 10.1111/j.1365-2249.1992.tb07947.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Reimer C, Philipps D, Aloisio C, et al. Specificity and association constants of 33 monoclonal antibodies to human IgA epitopes. Immunol Letters. 1989;21:209–16. doi: 10.1016/0165-2478(89)90106-5. [DOI] [PubMed] [Google Scholar]
  • 21.Cederholm B, Wieslander J, Bygren Heinegard D. Circulating complexes containing IgA fibronectin in patients with primary IgA nephropathy. Proc Natl Acad Sci USA. 1988;85:4865–8. doi: 10.1073/pnas.85.13.4865. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Cavallo M, Pozzili P, Thorpe R. Cytokines and autoimmunity. Clin Exp Immunol. 1994;96:1–7. doi: 10.1111/j.1365-2249.1994.tb06220.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Dantal J, Bigot E, Bogers W, et al. Effect of plasma protein absorption on protein excretion in kidney-transplant recipients with recurrent nephrotic syndrome. N Engl J Med. 1994;330:7–14. doi: 10.1056/NEJM199401063300102. [DOI] [PubMed] [Google Scholar]
  • 24.Girardin E, Roux-Lombard P, Grau G, et al. Imbalance between tumor necrosis factor-alpha and soluble TNF receptor concentrations in severe meningococcaemia. Immunology. 1992;76:20–23. [PMC free article] [PubMed] [Google Scholar]
  • 25.Gherardi R, Belec L, Soubrier M, et al. Overproduction of proinflammatory cytokines imbalanced by their antagonists in POEMS syndrome. Blood. 1996;87:1458–65. [PubMed] [Google Scholar]
  • 26.Bertani T, Abbate M, Zoja C, et al. Tumor necrosis factor induces glomerular damage in rabbits. Am J Pathol. 1989;134:419–30. [PMC free article] [PubMed] [Google Scholar]
  • 27.Yanagisawa M, Imai H, Fukushima Y, et al. Effects of tumor necrosis α and interleukin 1β on the proliferation of cultured glomerular epithelial cells. Virchows Arch. 1994;424:581–6. doi: 10.1007/BF00195770. [DOI] [PubMed] [Google Scholar]
  • 28.Striker L, Peten E, Elliot S, et al. Mesangial cell turnover: effect of heparin and peptide growth factors. Lab Invest. 1991;64:446–56. [PubMed] [Google Scholar]
  • 29.Tomosugi N, Cashman S, Pusey C, et al. Modulation of antibody mediated glomerular injury in vivo by bacterial liposaccharide, tumor necrosis factor and IL1. J Immunol. 1989;142:3083–90. [PubMed] [Google Scholar]
  • 30.Camussi G, Tetta C, Bussolino F, et al. Effect of leukocyte stimulation on rabbit immune complex glomerulonephritis. Kidney Int. 1990;38:1047–55. doi: 10.1038/ki.1990.311. [DOI] [PubMed] [Google Scholar]
  • 31.Hruby Z, Shirota K, Jothy S, Lowry R. Antiserum against tumor necrosis factor-alpha and a protease inhibitor reduce immune glomerular injury. Kidney Int. 1991;40:43–51. doi: 10.1038/ki.1991.177. [DOI] [PubMed] [Google Scholar]
  • 32.Karkar A, Tam F, Steinkasserer A, et al. Modulation of antibody-mediated glomerular injury in vivo by IL1-ra, soluble IL1 receptor and soluble TNF receptor. Kidney Int. 1995;48:1738–46. doi: 10.1038/ki.1995.472. [DOI] [PubMed] [Google Scholar]
  • 33.Yoshioka K, Takemura T, Murakami K, et al. In situ expression of cytokines in IgA nephritis. Kidney Int. 1993;44:825–33. doi: 10.1038/ki.1993.317. [DOI] [PubMed] [Google Scholar]
  • 34.Matsumoto K. Increased release of tumor necrosis factor α by monocytes from patients with glomerulonephritis. Clin Nephrol. 1993;40:148–54. [PubMed] [Google Scholar]
  • 35.Furukawa S, Matsubara T, Yone K, et al. Kawasaki disease differs from anaphylactoid purpura and measles with regard to tumour necrosis factor alpha and interleukin 6 in serum. Eur J Pediatr. 1992;151:44–47. doi: 10.1007/BF02073890. [DOI] [PubMed] [Google Scholar]
  • 36.Tsai-Hung W, Shiu-Chuan W, Tung-Po H, et al. Increased excretion of tumor necrosis factor alpha and interleukin 1 beta in urine from patients with IgA nephropathy and Schönlein–Henoch purpura. Nephron. 1996;74:79–88. doi: 10.1159/000189285. [DOI] [PubMed] [Google Scholar]
  • 37.Yokoyama H, Takaeda M, Wada T, et al. Glomerular ICAM-1 expression related to circulating TNF-α in human glomerulonephritis. Nephron. 1997;76:425–33. doi: 10.1159/000190225. [DOI] [PubMed] [Google Scholar]
  • 38.Achiron A, Margalit R, Hershkovitz R, et al. Intravenous immunoglobulin treatment of experimental T cell-mediated autoimmune disease: up regulation of T cell proliferation and down-regulation of tumor necrosis α secretion. J Clin Invest. 1994;93:600–5. doi: 10.1172/JCI117012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Aukrust P, Froloand S, Liabakk N, et al. Release of cytokines, soluble cytokine receptors and interleukin-1 receptor antagonist after intravenous immunoglobulin administration in vivo. Blood. 1994;84:2136–43. [PubMed] [Google Scholar]
  • 40.Chomarat P, Vannier E, Dechanet J, et al. Balance of IL1 receptor antagonist/IL1β in rheumatoid synovium and its regulation by IL4 and IL10. J Immunol. 1995;154:1432–9. [PubMed] [Google Scholar]
  • 41.Lan H, Nikolic-Paterson D, Zarama M, et al. Suppression of experimental crescentic glomerulonephritis by the interleukin 1 receptor antagonist. Kidney Int. 1993;43:479–85. doi: 10.1038/ki.1993.70. [DOI] [PubMed] [Google Scholar]
  • 42.Matsumoto K. Interleukin-1 production by monocytes from patients with glomerulonephritis after stimulation in vitro with soluble immune complexes. Clin Nephrol. 1991;36:267–73. [PubMed] [Google Scholar]
  • 43.Suematsu S, Matsuda T, Aozasa K, et al. IgG1 plasmocytosis in IL6 transgenic mice. Proc Natl Acad Sci USA. 1989;86:7547–51. doi: 10.1073/pnas.86.19.7547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Horii Y, Muraguchi A, Iwano M, et al. Involvement of IL6 in mesangial proliferative glomerulonephritis. J Immunol. 1989;143:3949–55. [PubMed] [Google Scholar]
  • 45.Gordon C, Richards N, Howie A, et al. Urinary IL6: a marker for mesangial proliferative glomerulonephritis? Clin Exp Immunol. 1991;86:145–9. doi: 10.1111/j.1365-2249.1991.tb05787.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Nakamura T, Ebihara I, Takahashi T, et al. Increased interleukin 6 mRNA expression by peripheral blood T cells from patients with IgA nephropathy. Autoimmunity. 1993;15:171–9. doi: 10.3109/08916939309019924. [DOI] [PubMed] [Google Scholar]
  • 47.Svenson M, Bagge Hansen M, Bendtzen K. Binding of cytokines to pharmaceutically prepared human immunoglobulin. J Clin Invest. 1993;92:2533–9. doi: 10.1172/JCI116862. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Yokoyama H, Takaeda M, Wada T, et al. Intraglomerular expression of MHC class II and KI-67 antigens and serum γ interferon levels in IgA nephropathy. Nephron. 1992;62:169–75. doi: 10.1159/000187028. [DOI] [PubMed] [Google Scholar]
  • 49.Chen A, Ding SL, Sheu LF, et al. Experimental IgA nephropathy: enhanced deposition of glomerular IgA immune complex in proteinuric state. Lab Invest. 1994;70:639–47. [PubMed] [Google Scholar]
  • 50.Jeannette C, Wieslander J, Tuttle, et al. Serum IgA–fibronectin aggregates in patients with IgA nephropathy and Henoch–Schönlein purpura: diagnostic value and pathogenic implications. Am J Kidney Dis. 1991;18:466–71. doi: 10.1016/s0272-6386(12)80115-6. [DOI] [PubMed] [Google Scholar]
  • 51.Cosio F, Bakaletz A. The role of fibronectin on the clearance and tissue uptake of antigens and immune complexes in rats. J Clin Invest. 1987;80:1270–9. doi: 10.1172/JCI113202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Wahl S, Allen J, Hines H, et al. Synthetic fibronectin peptides suppress arthritis in rats by interrupting leucocyte adhesion and recruitment. J Clin Invest. 1994;94:655–62. doi: 10.1172/JCI117382. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Wardle EN. Neutralizing inflammatory cytokines in nephritis. Nephron. 1995;69:223–7. doi: 10.1159/000188460. [DOI] [PubMed] [Google Scholar]
  • 54.Ortiz A, Egido J. Is there a role for specific anti-TNF strategies in glomerular diseases? Nephrol Dial Transplant. 1995;10:309–11. [PubMed] [Google Scholar]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology

RESOURCES