Increased mesangial cell hyaluronan expression in lupus nephritis is mediated by anti-DNA antibody-induced IL-1β

original article

http://www.kidney-international.org & 2006 International Society of Nephrology

Increased mesangial cell hyaluronan expression in lupus nephritis is mediated by anti-DNA antibody-induced IL-1b S Yung1,2, RCW Tsang1, JKH Leung1 and TM Chan1,2 1

Department of Medicine, Queen Mary Hospital, University of Hong Kong, Hong Kong, China

The mechanism by which anti-DNA antibodies contribute to the pathogenesis of lupus nephritis (LN) remains to be fully elucidated. Hyaluronan (HA) is an important extracellular matrix constituent that accumulates during tissue injury, and participates in lymphocyte recruitment to sites of inflammation. The role of HA in the pathogenesis of LN has not been defined. We investigated the expression of HA in renal biopsies and circulating HA levels in patients with diffuse proliferative LN, and the effect of human anti-DNA antibodies on HA synthesis in cultured human mesangial cells (HMC). HA expression was increased in the mesangium, and in the periglomerular and tubular distribution in LN kidney biopsies. LN patients showed increased levels of circulating HA, especially during active disease, which correlated with anti-DNA antibody titers (r ¼ 0.35, P ¼ 0.0234). Anti-DNA antibodies isolated during active LN but not remission increased de novo synthesis of 3H-labeled HA, which was accompanied by induction of HA synthase (HAS) II transcription, and enhanced IL-1b, IL-6, and tumor necrosis factor-a secretion in HMC (Po0.001 for all). Only anti-DNA antibody induction of IL-1b enhanced HA synthesis, which was abrogated by inhibitors of de novo mRNA or protein synthesis. Our findings demonstrate that HA expression is significantly increased within the mesangium in diffuse proliferative LN mediated through anti-DNA antibody-induced IL-1b. Given that HA plays a pivotal role during inflammatory responses, influences cellular behavior and assists in the recruitment of lymphocytes to sites of injury, it is likely that HA contributes to the pathogenesis of LN. Kidney International (2006) 69, 272–280. doi:10.1038/sj.ki.5000042 KEYWORDS: anti-DNA antibodies; hyaluronan; IL-1b; human mesangial cells

Correspondence: TM Chan and S Yung, Department of Medicine, Queen Mary Hospital, University of Hong Kong, Pokfulam Road, Hong Kong SAR, China. E-mail: [email protected] 2

Both these authors contributed equally to this work.

Received 7 June 2005; revised 30 July 2005; accepted 11 August 2005 272

Lupus nephritis (LN) is a common complication in systemic lupus erythematosus (SLE), and an important cause of renal failure. Anti-DNA antibodies represent a hallmark of SLE, and are implicated in the pathogenesis.1–4 They can be eluted from renal tissues in LN, and have been demonstrated to bind to resident renal cells.2–4 The pathogenetic mechanisms involving anti-DNA antibodies have not been fully elucidated. An important step in the initiation and propagation of inflammatory tissue damage is the recruitment of neutrophils and mononuclear cells to sites of injury.5 In addition, progression from acute to chronic inflammation is accompanied by increased synthesis and degradation of extracellular matrix (ECM) components. Degradation of matrix components generates fragments at sites of inflammation, which may exhibit different biological activities from their parent molecules.6,7 Hyaluronan (HA) is an important ECM component that undergoes dynamic regulation during inflammatory processes.8,9 HA is a non-sulfated glycosaminoglycan comprising repeating disaccharide units of D-glucuronic acid and N-acetyl-D-glucosamine. In its native form, HA occurs as a macromolecule of 1–6  106 Da.10 HA is normally widely distributed within the ECM and on cell surfaces, and is particularly enriched in pericellular matrices surrounding migrating and proliferating cells during embryonic development, tissue repair and tumorigenesis.11–13 Independent investigators have demonstrated increased renal cortical HA expression in non-lupus crescentic glomerulonephritis, allograft rejection and tubulo-interstitial diseases.14–17 Studies in MRL/lpr mice have demonstrated increased periglomerular and peritubular HA expression.18 Mesangial cells and the surrounding matrix are strategically located juxtaposed to adjacent capillary loops, and they interact with inflammatory mediators, immune deposits and complement components. Mesangial cell proliferation, mesangial deposition of anti-DNA antibodies and increased expression of proinflammatory or profibrotic mediators are prominent features in LN. The objective of this study was to investigate the role of HA in the pathogenesis of LN. In vivo studies were conducted to investigate circulating HA levels and HA expression in renal biopsies in patients with LN. Human polyclonal anti-DNA antibodies were isolated from these patients serially during active disease and remission, Kidney International (2006) 69, 272–280

original article

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

and their effects on de novo HA synthesis in cultured human mesangial cells (HMC) were investigated.

a

b

c

d

e

f

RESULTS HA expression in renal biopsies and serum HA levels in patients with LN

Healthy controls showed no glomerular HA expression and minimal tubular HA staining (Figure 1d and g). LN renal biopsies showed significant HA expression within the mesangium, periglomerular area and tubules, and the intensity of HA expression correlated with that of IgG deposition in the glomerulus (Figure 1k–m). Three of the 12 biopsies of active diffuse proliferative LN also showed increased HA staining associated with glomerular lymphocytic infiltrates (Figure 1f and i). Circulating HA levels were elevated in patients with LN, and serum HA levels were higher during active disease compared with remission specimens (Figure 2a and Table 1). Circulating HA levels correlated with the titer of anti-DNA antibodies (Figure 2b).

*

g

*

h

i

*

*

* j

k

l

Effect of anti-DNA antibodies on HA synthesis in HMC

Kidney International (2006) 69, 272–280

m Glomerular HA expression (AU)

The recovery of anti-DNA antibodies from serum samples after affinity isolation was 75.174.1 and 76.376.0% for remission and active samples, respectively, and IgG anti-DNA antibodies accounted for 88.774.0 and 85.776.5% of serum anti-DNA activity in remission and active samples, respectively. Measurement of IgG1, IgG2, IgG3, and IgG4 levels in anti-DNA antibody preparations showed that IgG1 was the predominant isotype, and that the relative proportion of IgG4 was significantly higher during active disease compared with samples obtained during remission (Table 2). In accordance with our previous report, the results of cellular enzyme-linked immunosorbent assay (ELISA) demonstrated that anti-DNA antibodies from patients with LN showed significant binding to HMC compared with control IgG (Figure 3a). Serum samples depleted of anti-DNA antibodies (non-anti-DNA immunoglobulins) did not show significant binding to HMC (Figure 3a). HA synthesis in cultured HMC was not altered by nonanti-DNA immunoglobulins, nor by Remission Ab samples (anti-DNA antibodies isolated during remission), when compared with normal human IgG (Figure 3b). In contrast, anti-DNA antibodies isolated during active LN (Active Ab) stimulated de novo synthesis of 3H-labeled HA in a timedependent manner, and HA was significantly increased in both the conditioned medium (CM) and cell extract (CE) fractions after 12 h of stimulation by Active Ab (Figure 4a). At 24 h, Active Ab increased total 3H-labeled HA by 1.88-fold compared with control IgG (Po0.001). The hydrodynamic size of 3H-labeled HA synthesized de novo was next investigated (Figure 4b and c). HA synthesized by HMC in the presence of control IgG, similar to that from cells exposed to medium alone (data not shown), consisted of high molecular weight (HMW) HA that eluted in the Vo of the column (P1). The 3H-labeled HA thus

3

r = 0.680 P = 0.0149

2

1

0 0

1 2 3 Glomerular IgG deposition (AU)

4

Figure 1 | Representative renal HA expression. Paraffin sections of normal tissue obtained from patients who underwent nephrectomy for tumor (a, d, g, and j) or renal biopsies from patients with diffuse proliferative LN (b, c, e, f, h, i, k, and l) were subjected to staining with eosin and hematoxylin (a–c). Sections were also used for cytochemical staining for HA expression using the peroxidase–antiperoxidase method (d–i). Immunoreactive products were visualized using diaminobenzidine as chromagen and counterstained with hematoxylin. Constitutive HA expression was observed in the periglomerular tissue and tubules but not within the glomerulus in control specimens (d). Specimens obtained from LN patients demonstrated increased HA expression in the glomerulus (asterisk), in addition to increased periglomerular and tubular staining (e and f). In 25% of LN biopsies, increased HA expression was also observed in infiltrating cells in the glomerulus (f, arrow). Part of the glomerulus (T) shown in d–f is magnified to facilitate comparison of HA expression in control and LN renal tissues (g–i). Cryosections were used to determine IgG deposition within the glomeruli (j–l). Original magnification  200 (a–f) and (j–l). Original magnification  1600 (g–i). The expression of IgG deposition and HA in the glomerulus were scored without knowledge of other clinical or laboratory data (m).

synthesized accounted for 427.575.3 and 331.8731.7 dpm/ mg cellular protein in the CM and CE fractions, respectively, and the distribution in the two compartments was 56% and 44% respectively. Remission Ab did not alter the hydrodynamic size or distribution of HA synthesized by HMC 273

original article

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

600 400 200 0

b 600 Anti-DNA antibody levels (IU/ml)

Serum HA levels (ng/ml)

P<0.001

r = 0.3578 P = 0.0234 400 200 0

Control Inactive LN Active LN

0

200 400 600 Serum HA levels (ng/ml)

800

Figure 2 | Serum levels of HA in healthy controls and patients with diffuse proliferative LN. (a) HA levels were measured in serum samples obtained from control subjects and from patients with LN during active disease and remission sequentially using a commercial ELISA specific to HA (n ¼ 20 for each group). Horizontal lines represent the respective mean value of each group. Po0.001, active LN vs LN in remission or controls. (b) Serum HA levels from active (K) or inactive (J) LN patients correlated with the titers of anti-DNA antibodies.

Percentage of total IgG during remission

Percentage of total IgG during active disease

P-value

60.2778.95 26.30712.26 11.1676.07 2.2571.74

52.3378.45 35.4879.53 7.5573.31 4.6572.20

0.101 0.137 0.366 0.035

*

4

Active Ab

Non-anti-DNA Ig

Inactive Ab

Non-anti-DNA Ig

0

3

2

60 40 20 0

Inactive LN Active LN

Inactive LN

Anti-DNA Ab

6

80

Non-anti-DNA Ig

8

100

Anti-DNA Ab

*

Non-anti-DNA Ig

b 10

[ H]-glucosamine labeled HA (dpm/
g cellular protein)

a

Control IgG

IgG1 IgG2 IgG3 IgG4

HMC-IgG binding (
g IgG/
g cellular protein)

mRNA expression for the three HA synthase (HAS) isoforms was investigated by reverse transcriptase-polymerase chain reaction (RT-PCR). HAS II and III (but not HAS I) mRNAs were constitutively expressed in the presence of serum-free medium (data not shown), and were not altered by control IgG (Figure 5). Anti-DNA antibodies induced HAS II transcription, and the induction was more prominent with Active Ab (Po0.05 and Po0.001 for Remission Ab and Active Ab vs control IgG, respectively, and Po0.001 for Remission Ab vs Active Ab). HAS III mRNA expression was unaffected by anti-DNA antibodies. The stimulatory effect of Active Ab on HA synthesis was dependent on de novo mRNA and protein synthesis (Figure 6), and induction of HA synthesis was abrogated when the anti-DNA antibodies were pre-incubated with exogenous DNA before their addition to HMC (Figure 4b and c).

800

Table 2 | Relative proportions of different IgG isotypes in anti-DNA antibody preparations IgG isotype

Mechanisms by which anti-DNA antibodies induced HA synthesis in HMC

a

Previous studies have demonstrated that IL-1b, IL-6 and tumor necrosis factor-a (TNF-a) can stimulate HA synthesis in other cell types.19–21 As these cytokines are upregulated in active LN,22 we investigated their potential role in modulating HA synthesis in HMC. Prior cytokine determination

Control IgG

compared to control IgG (Figure 4b). In comparison, 64.675.6% of the HA synthesized in HMC upon stimulation with Active Ab was secreted into the CM and 35.475.0% was cell-associated. Active Ab not only increased HMW HA in both fractions, but also induced the generation of low molecular weight (LMW, P2) HA (904.0712.4 and 305.2724.1 dpm/mg cellular protein for P1 and P2, respectively, in CM fraction and 645.7747.1 and 34.577.8 dpm/mg cellular protein for P1 and P2, respectively, in CE fraction).

Active LN

Figure 3 | Comparison of HMC-binding activity between anti-DNA antibodies and serum samples depleted of anti-DNA antibodies (non-anti-DNA immunoglobulins) obtained from lupus patients during inactive and active disease (a). Horizontal line represents the mean value of each group. *Po0.001, Active Ab or inactive Ab vs control IgG. De novo synthesis of 3H-labeled HA by HMC after incubation with control IgG, non-anti-DNA immunoglobulins (nonanti-DNA Ig) or anti-DNA antibodies (10 mg/ml IgG final concentration for all) for 24 h (b). Samples of CM were dialyzed into 8 M urea buffer, pH 6.0, and loaded onto DEAE-Sephacel columns to remove contaminating proteins and glycoproteins. Samples were further purified by Uno Q anion-exchange chromatography and samples eluting with 0.45 M NaCl were collected, digested with papain and subjected to gel filtration chromatography on Sepharose CL-4B under dissociative conditions to separate 3H-labeled HA from digested material. 3H-labeled HA eluted in the void volume of the column and fractions with radioactivity were pooled, and an aliquot subjected to beta scintillation counting to determine total 3H-labeled HA synthesized under control and experiment conditions. Horizontal lines represent the respective mean values for de novo synthesis of 3 H-labeled HA in each group.

Table 1 | Comparison of total IgG concentration, anti-DNA antibody titer, HMC-binding IgG level, and HA concentration in serum samples obtained from patients with LN obtained during active disease or remission

IgG concentration (mg/dl) Anti-DNA antibodies (IU/ml) HMC-binding IgG (mg of HMC-bound IgG/mg cellular protein) HA (ng/ml) 274

Remission (mean7s.d., n=20)

Active LN (mean7s.d., n=20)

P-value

854.47214.1 95.4754.2 2.670.5 67.05756.98

1471.67524.3 305.8798.5 4.371.6 225.407194.80

0.001 0.001 0.001 0.001

Kidney International (2006) 69, 272–280

original article

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

CM

*

500 400 *

300 200 100 0 CE 500 400

*

300 200

*

100 0

0

3

6

12

24

Vo 140 120 100 80 60 40 20 0

P1

c

Vt P2

Preincubation with DNA

120 100 80 60 40 20 0 20

30

Time (h)

40 50 60 70 Fraction number

80

90

[3H]-glucosamine labeled HA (dpm/
g cellular protein)

b 600

[3H]-glucosamine labeled HA (dpm/
g cellular protein)

[3H]-glucosamine labeled HA (dpm/
g cellular protein)

a

Vo 140

P1

Vt P2

120 100 80 60 40 20 0 Preincubation with DNA

120 100 80 60 40 20 0 20

30

40 50 60 70 Fraction number

80

90

Figure 4 | Total 3H-labeled HA synthesized by HMC in relation to the duration of incubation with control IgG (&), Remission Ab ( ) or Active Ab (’) in the presence of [3H]-glucosamine for periods up to 24 h (a). At select time points, the CM was decanted and the cells extracted with 4 M GuHCl/4% CHAPS. 3H-labeled HA was purified by a combination of dialysis and anion-exchange chromatography. The results are mean7s.d. of three separate experiments. *Po0.05 compared to control IgG. Gel filtration chromatography of 3H-labeled HA synthesized de novo in HMC. Equal aliquots (100 ml) of (b) CM or (c) CE obtained from cells stimulated with control IgG (J), Remission Ab ( ) or Active Ab (K) were subjected to gel filtration chromatography under dissociative condition. The increase in 3H-labeled HA synthesis upon stimulation with Active Ab (upper panels) was abrogated when anti-DNA antibodies were preincubated with exogenous DNA (5 mg/ml) before addition to HMC (bottom panel). P1 and P2 represent HMW and LMW 3H-labeled HA respectively. Vo and Vt represent the void and total volume of the column, respectively.

b

HAS III

-actin (204 bp) Control IgG: + – Remission Ab: – + Active Ab: – –

– – +

+ – –

– + –

– – +

800 2.5 2.0 1.5 1.0

** *

0.5 0.0

HAS II

HAS III

Figure 5 | RT-PCR for HAS II and HAS III mRNA after stimulation with anti-DNA antibodies or control IgG. Total RNA was extracted from HMC using TriReagent according to the manufacturer’s instructions and 1 mg of total RNA reverse transcribed to cDNA before PCR amplification for HAS I, II, and III transcripts. (a) PCR products were separated by flat bed electrophoresis on a 2% agarose gel. HAS I was not expressed by HMC. a-Actin was used as the housekeeping gene. (b) Primer to a-actin mRNA densitometric ratios are shown for control IgG (&), Remission Ab ( ) and Active Ab (’). *Po0.05, control vs Remission Ab; **Po0.001, Active Ab vs control IgG or Remission Ab.

showed no cytokine activity to IL-1b, IL-6, and TNF-a in all anti-DNA antibody preparations (data not shown). Remission Ab did not induce these cytokines in HMC (Figure 7). Active Ab induced the secretion of IL-1b, IL-6, and TNF-a in HMC by 2.20  , 2.28  , and 1.75  , respectively (26.876.4, 24.177.4, and 17.276.0 pg/mg cellular protein for IL-1b, IL-6, and TNF-a, respectively; Po0.001 for all when compared to control IgG or Remission Ab). Neutralizing antibodies to IL-6 or TNF-a had no effect on the induction of HA synthesis by Active Ab (Figure 8a). In contrast, the stimulatory effect of Active Ab on total HA synthesis was abrogated by IL-1 receptor antagonist in a dose-dependent manner (Figure 8a). Maximum inhibition was observed with 1 mg/ml IL-1 receptor antagonist, which abrogated both HMW and LMW 3H-labeled HA synthesis (Figure 8b). Kidney International (2006) 69, 272–280

[3H]-glucosamine labeled HA (dpm/
g cellular protein)

HAS II

Ratio of primer/-actin(AU)

a

* 600

400

200

0

Control IgG

Remission Ab

Active Ab

Figure 6 | Effect of actinomycin D or cycloheximide on the induction of 3H-labeled HA synthesis in HMC by anti-DNA antibodies. Confluent growth-arrested HMC were incubated in the absence (’) or presence of actinomycin D (5 mg/ml, ) or cycloheximide (5 mg/ml, &) for 1 h before stimulation of cells with control IgG or anti-DNA antibodies for another 24 h. The results show that HA induction by anti-DNA antibodies is dependent on de novo gene transcription and translation. Data represent mean values and s.d. of three individual experiments. *Po0.01, with or without drug treatment, or compared with control IgG.

DISCUSSION

Severe proliferative LN remains an important cause of acute or chronic renal failure. Although the deposition of anti-DNA antibodies and inflammatory mediators within the kidney parenchyma are well demonstrated in LN, the pathogenetic mechanisms pertaining to inflammatory and fibrotic processes have not been fully elucidated. There is accumulating evidence that HA is involved in immunemediated renal injury.15,23–28 Expression of HA in normal kidneys is limited to the medullary and papillary interstitium, where it helps to maintain the oncotic pressure for urinary concentration.23,24 Increased renal HA expression occurs in animal models of autoimmune crescentic glomerulonephritis, renal ischemia and kidney allograft rejection, and 275

original article

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

P<0.001

50 40 30 20 10 0

Control Remission Active IgG Ab Ab

c

50 P<0.001

40 30 20 10 0

Control Remission Active IgG Ab Ab

TNF- secretion (pg/
g cellular protein)

b

60

IL-6 secretion (pg/
g cellular protein)

IL-1 secretion (pg/
g cellular protein)

a

50 40

P<0.001

30 20 10 0

Control Remission Active IgG Ab Ab

Figure 7 | Effect of anti-DNA antibodies on IL-1b, IL-6, and TNF-a secretion in HMC. Confluent growth-arrested HMC were stimulated with control IgG or anti-DNA antibodies for 24 h. The culture medium was decanted, centrifuged to remove cell debris and the levels of (a) IL-1b, (b) IL-6, and (c) TNF-a determined using commercial ELISAs. Horizontal bars represent the mean value of each group. Po0.001, compared with control IgG or Remission Ab.

[3H]-labeled HA synthesis (percentage compared to basal levels)

a 120 100 80 60 40 20 0

*

* * *

*

*

0 0.01 0.1 0.5 1 10 25 Concentration of neutralizing antibody/receptor antagonist (
g/ml)

[3H]-glucosamine labeled HA (dpm/
g cellular protein)

b

Vt 140 CM Vo 120 100 80 60 40 20 0 120 CE 100 80 60 40 20 0 20 30 40 50 60 70 80 90 Fraction no.

Figure 8 | (a) Effect of neutralizing antibodies to IL-6 and TNF-a, and IL-1 receptor antagonist on de novo 3H-labeled HA synthesis. HMC were incubated with varying concentrations of neutralizing antibodies to IL-6 and TNF-a, and IL-1 receptor antagonist (0–25 mg/ ml) for 1 h before the addition of Active Ab for a further 24 h. The CM was assessed for 3H-labeled HA. *Po0.01, compared to samples not incubated with IL-1 receptor antagonist. (b) Gel filtration chromatography of 3H-labeled HA synthesis. To determine the modulatory effect of IL-1 receptor antagonist on 3H-labeled HA synthesis by Active Ab, HMC were incubated with control IgG in the presence (&) or absence (’) of IL-1 receptor antagonist (1 mg/ml), or Active Ab in the presence (J) or absence (K) of IL-1 receptor antagonist. Aliquots of CM (upper panel) and CE (lower panel) were chromatographed on Sephacryl S-1000 under dissociative conditions. Vo and Vt represent the void and total volume of the column, respectively.

increased renal cortical HA expression has been observed in various human acute or chronic inflammatory renal diseases.16,25–28 Feusi et al.18 demonstrated the expression of HA in the tubulo-interstitium and crescents of mice with LN, but not within the mesangium. Our findings from renal tissue of patients with diffuse proliferative LN show increased HA expression in a periglomerular, mesangial, and tubulointerstitial distribution. The discrepancy with regard to mesangial HA expression could be related to differences between species or staining procedures. HA normally binds to various matrix molecules such as aggrecan, versican or CD44,29 and such binding may interfere with the binding by exogenous HA-binding protein, thereby leading to falsenegative staining. In this respect, all samples in the present study were treated with papain to remove proteins bound to HA in order to prevent false-negative staining results. Approximately 65% of the HA synthesized by HMC in the presence of Active Ab was secreted into the CM. This could explain the prominent periglomerular (in addition to 276

intraglomerular) HA expression in LN renal tissues. As HA can bind or induce chemokine/cytokine synthesis and upregulate the expression of adhesion molecules, HA assumes a pivotal role in the recruitment and retention of lymphocytes at sites of inflammatory tissue injury.8,30–32 Lymphocytes can thus bind, via abundant CD44 on their cell surface, to secreted HA molecules or HA molecules still associated with mesangial cells. Other investigators have also shown that HA can modulate monocyte infiltration and transmigration through their binding to CD44.33 In this regard, 25% of LN renal biopsy samples showed significant HA staining on infiltrating cells within the glomerulus. The association between mesangial HA expression and IgG deposition also suggests the possibility that anti-DNA antibodies might influence the synthesis of HA in HMC. The correlation between the increased circulating HA levels and anti-DNA antibody titers in patients with LN lends support to the hypothesis that anti-DNA antibodies might be involved in the induction of HA. HA subpopulations of different molecular weights can exert different effects on T-lymphocyte-mediated liver injury in mice.34 Chemical insult such as hyperglycemia can also stimulate HA synthesis in rat mesangial cells and lead to increased monocyte adhesion.35 Anti-DNA antibodies have been demonstrated to crossreact with proteoglycans and HA.36,37 Given that HA can also bind to complement,38 the interaction between anti-DNA antibodies, complement components, resident or infiltrating cells in the renal parenchyma could occur via HA and result in the propagation of local inflammatory processes. Data from the in vitro studies show that anti-DNA immunoglobulins obtained from patients during active LN could induce de novo HA synthesis in HMC. Polymerization of HA occurs at the plasma membrane through the action of one or more of three HA synthase isomers.39 These enzymes orchestrate the polymerization and translocation of HA out of the cell into the pericellular or ECM, and the three isoforms have distinct tissue and temporal expressions that dictate their respective roles in the regulation of HA biosynthesis.39 Previous studies have shown that only HAS II is essential for tissue development, whereas the roles of HAS I and III have not been fully elucidated. In this study, we have shown that HMC constitutively expressed HAS II and Kidney International (2006) 69, 272–280

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

III but not HAS I mRNA. Anti-DNA antibodies obtained from active LN stimulated de novo HA synthesis in HMC through induction of HAS II gene expression. Although HAS II mRNA was also induced by anti-DNA antibodies during remission, this appears insufficient to result in significant de novo HA synthesis. As identical concentrations of antiDNA antibodies were used in these experiments, the data highlight the functional heterogeneity of anti-DNA antibodies at different phases of disease, and imply that antiDNA antibodies still detectable during clinical remission might be ‘less pathogenic’. The results on IgG isotypes confirmed the predominance of IgG1 in anti-DNA antibody preparations.40,41 Some investigators have suggested an association between IgG1 and IgG3 anti-DNA antibodies with renal disease, whereas others have observed an association between IgG2 anti-DNA antibodies and nephritic flare.40,41 It is of interest to note the increase in the relative abundance of IgG4 isotype during active disease. This isotype association with disease activity requires confirmation with more clinical samples, and the pathogenetic implications remain to be investigated. Multiple proinflammatory cytokines are induced during active LN.22 Our results show that IL-1b, IL-6, and TNF-a were induced in HMC by ‘Active’ but not ‘Remission’ antiDNA antibodies, and that IL-1b is involved, at least in part, in the induction of de novo HA synthesis in HMC by anti-DNA antibodies. HA has been implicated in the amplification of inflammatory and/or remodeling processes in kidney tissues.28,42 In general, HMW HA is primarily involved in normal physiologic processes whereas the generation of LMW HA occurs under pathological circumstances. Independent researchers have demonstrated that an ECM enriched with HMW HA favors smooth muscle cell proliferation but inhibits vascular endothelial cell proliferation and migration.43–45 HMW HA can also form intercellular cables that act as an inflammatory matrix for the retention of leukocytes at sites of injury. Intermediate-size HA (Mr 0.2  106 Da) can induce multiple chemokines in alveolar macrophages,8 and HA oligosaccharides have been demonstrated to induce collagen type I and III in rat lung fibroblasts.46 Although it has been suggested that the accumulation of HA molecules in injured tissues could lead to an excessive fibrotic response,46 recent studies have shown that HMW HA (2  106 Da) could antagonize the induction of collagen types III and IV by TGF-b1 in human proximal tubular epithelial cells, thereby limiting progressive fibrosis and facilitating reparative processes following injury.47 Our results show that both LMW and HMW HA molecules are induced in HMC by anti-DNA antibodies during active LN. We speculate that these HA molecules are involved in distinct proinflammatory or reparative processes, the balance of which could influence fibrotic and healing processes following an episode of acute LN. In conclusion, we have demonstrated that during the acute phase of diffuse proliferative LN patients have high Kidney International (2006) 69, 272–280

original article

circulating levels of HA, which correlate with anti-DNA antibody titers, and that anti-DNA antibodies could stimulate the synthesis of both HMW and LMW HA in mesangial cells through the induction of HAS II and IL-1b. The HA thus synthesized could play a significant pathogenetic role in inflammatory, fibrotic and healing processes, by virtue of their interaction with cytokines, adhesion molecules and mononuclear inflammatory cells. MATERIALS AND METHODS Serum samples from patients with LN Sequential serum samples that demonstrated high IgG binding to HMC were collected from 20 patients with biopsy-proven diffuse proliferative LN (15 women and five men, mean age 42.577.5 and 43.778.9 years, respectively). Each patient contributed two samples, one during active disease and another during remission (defined by clinical manifestations together with SLE Disease Activity Indices (SLEDAI) of X10 and o4, respectively).48 Immune complexes were removed by polyethylene glycol precipitation.2 Sera from 20 sex- and aged-matched healthy individuals were used as controls. Studies on renal biopsies A total of 12 renal biopsies showing active diffuse proliferative LN (Class IV) were included. Normal kidney tissue from five patients who underwent nephrectomy for tumor was included as control. Unless otherwise stated, all incubations were for 1 h at 371C. Paraffin sections were stained with eosin and hematoxylin (Sigma, Tin Hang Technology Ltd, Hong Kong) according to standard procedures. To detect HA expression, paraffin sections were digested with papain (15 IU/ml) in 200 mM sodium acetate, pH 6.0, for 1 h at 601C before incubation with biotinylated HA binding protein (1:50) overnight at room temperature.49 Samples were then washed in phosphatebuffered saline (PBS) and incubated with streptavidin-conjugated horseradish peroxidase secondary antibody. Signal detection was by the peroxidase–anti-peroxidase method (Dako, Gene Company, Hong Kong) with hematoxylin counterstaining. Glomerular staining was graded as previously described (0: no staining; 1 þ : faint staining; 2 þ : moderate staining; 3 þ : intense staining).50 To detect IgG deposition, cryosections were incubated with fluorescein isothiocyanate-conjugated goat anti-human IgG (dilution 1:100). After washing with PBS, sections were mounted with fluorescence mountant (DAKO, Gene Company, Hong Kong) and epifluorescence was viewed using an Axiovert 135 inverted microscope (Zeiss, Gold Pacific Ltd, Hong Kong). All samples were coded and evaluated without the knowledge of the clinical information. Four separate fields of each specimen were taken. The extent and intensity of staining was scored semiquantitatively (0: absent; 1 þ : mild; 2 þ : moderate; 3 þ : marked; 4 þ ; severe), and a mean score was calculated for each sample.50 In vitro studies on HMC Primary cultures of HMC were established from nephrectomized kidneys and maintained in RPMI 1640 medium supplemented with 15% fetal bovine serum (FBS).2 All experiments were performed on cells of the 5–7th passages that had been growth-arrested for 72 h, in the absence of serum and complement components. Cells were stimulated with either human IgG anti-DNA antibodies or normal human IgG at a final concentration of 10 mg/ml for periods up to 24 h. This dose was chosen as preliminary studies demonstrated a dose-dependent response of IgG binding to HMC up to 10 mg/ml. 277

original article

Addition of higher concentrations of IgG did not result in further IgG binding to HMC. To determine whether the effect of anti-DNA antibodies on HA synthesis in HMC was dependent on de novo mRNA or protein synthesis of HAS I, II, and III, parallel experiments were performed in the presence of actinomycin-D (5 mg/ml) or cycloheximide (5 mg/ml), respectively. Preliminary experiments showed maximum inhibition of gene transcription and mRNA translation at these doses, without significant cytotoxic effect. To investigate the mechanisms by which anti-DNA antibodies modulated HA synthesis in HMC, cells were incubated with neutralizing antibodies to IL-6 or TNF-a, or with IL-1 receptor antagonist for 1 h before incubation with anti-DNA antibodies. Measurement of total IgG or IgG isotypes in serum samples or anti-DNA antibody preparations IgG concentration in serum samples or anti-DNA antibody preparations was measured by ELISA as described previously.2 The concentrations of IgG1, IgG2, IgG3, and IgG4 in anti-DNA antibody preparations were determined using a commercial human IgG subclass profile ELISA kit according to the manufacturer’s instructions (Zymed Laboratories, Onwon Trading Limited, Hong Kong). The detection ranges were 0.43–13.72, 0.17–5.32 , 0.042–1.34, and 0.024–0.76 mg/ml, respectively, for IgG1, IgG2, IgG3, and IgG4. Cellular ELISA to determine immunoglobulin binding to HMC HMC were seeded into 96-well tissue culture plates at a density of 10 000 cells/cm2 in RPMI 1640 containing 15% FBS until 90% confluent. The cells were then depleted of serum for 72 h before determining the amount of IgG bound to HMC.2 In some experiments, anti-DNA antibodies were premixed with DNA at concentrations of 0–10 mg/ml for 2 h at 371C before incubation with HMC to investigate the specificity of the anti-DNA antibodies. The amount of IgG bound to HMC was expressed as mg HMC-bound IgG/mg cellular protein and seropositivity for HMC binding was defined by readings exceeding mean þ 3s.d. of control.2 To determine total cellular protein, HMC cultured in 96-well plates were lysed with 4 M urea buffer, 20 mM sodium acetate, pH 6.0, containing 1% Triton X-100 (50 ml). Protein content was then measured using a modified Lowry assay (BioRad, Hong Kong). Isolation of human polyclonal anti-DNA antibodies from patients with LN Polyclonal IgG anti-double-stranded DNA antibodies were isolated from sera of LN patients using sequential affinity chromatography, and the IgG level was determined, as previously described.2 Their purity was confirmed by 10% SDS-PAGE and ability to bind exogenous DNA. Anti-DNA activity in serum samples and antiDNA antibody preparations was determined using a commercial ELISA (Microplate autoimmune anti-DNA quantitative ELISA, BioRad, Hong Kong). The detection range was 20–1000 IU/ml, and readings of 460 IU/ml indicated positive results. HAS I, II, and III mRNA expression by RT-PCR HMC RNA was extracted with TriReagent (Onwon Trading Limited, Hong Kong) according to the manufacturer’s instructions. A 1 mg portion of total RNA was reverse transcribed to cDNA with M-MLV reverse transcriptase using the random hexamers method and amplified using PCR in a total volume of 20 ml, as previously described.49,51 After amplification, 10 ml of both test and control PCR reaction mix was separated by electrophoresis on a flat-bed agarose gel (2% w/v). Images of the gels were captured using the 278

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

ChemiGenius gel documentation system, and the density of bands determined with the ChemiGenius analysis software (Syngene, Gene Company, Hong Kong). Results were expressed as the ratio of HAS I, II, and III to a-actin expression. The primer sequences used are as follows:

Primer

Sequence

bp

HAS I forward HAS I reverse

50 -AGCAGGACGCGCCCAAGCCCACTC-30 50 -TCCTCCGCCTCCACCTCCCGATAG-30

562

HAS II forward HAS II reverse

50 -TCCCGGTGAGACAGATGAGT-30 50 -ACCCGGTTCGTGAGATGC-30

495

HAS III forward HAS III reverse

50 -AGTGCAGCTTCGGGGATGA-30 50 -TGATGGTAGCAATGGCAAAGAT-30

453

a-Actin forward a-Actin reverse

50 -GGAGCAATGATCTTGATCTT-30 50 -TCCTGAGGTACGGGTCCTTCC-30

204

Measurement of cytokine concentration in culture supernatant HMC were incubated with anti-DNA antibodies or control IgG for 24 h. The supernatant was decanted, centrifuged at 2000 g for 10 min to remove cell debris and the levels of IL-1b, IL-6, and TNF-a were measured using respective commercial ELISA kits according to the manufacturer’s instructions (PharMingen, Bio-Gene, Hong Kong). The detection ranges were 5–300 pg/ml for IL-6 and IL-1b, and 15–1000 pg/ml for TNF-a. Measurement of HA in serum samples HA levels were determined in serum samples obtained from control subjects or patients with diffuse proliferative LN using a commercial HA ELISA according to the manufacturer’s instructions (Corgenix, Denver, CO, USA). Detection range for HA was 10–1000 ng/ml and the ELISA showed no crossreactivity to other glycosaminoglycans. Measurement of de novo HA synthesis by metabolic labeling and extraction of 3H-labeled HA Confluent growth-arrested HMC were radiolabeled with D-6-3Hglucosamine hydrochloride (25 mCi/ml; Amersham Biosciences, Hong Kong) for 24 h as previously described,19,52 and the CM decanted. Cell-associated 3H-labeled macromolecules were extracted with 4 M guanidine-HCl and 4% CHAPS.19 Samples were buffered, exchanged into 8 M urea, 20 mM Bis-Tris, pH 6.0, containing 0.1% CHAPS and applied to DEAE-Sephacel columns equilibrated in the same buffer. Unbound material was washed with 8 M urea buffer containing 0.15 M NaCl to remove 3H-labeled glycoproteins. 3Hlabeled HA was eluted with the above buffer containing 0.3 M NaCl, dialyzed into 8 M urea buffer containing 0.15 M NaCl and further purified on a Uno Q column equilibrated in the same buffer, interfaced with a Biologic Workstation (BioRad, Hong Kong). 3 H-labeled HA was washed with (a) 0.1 M NaCl (10 ml) and (b) 0.3 M NaCl (10 ml) before elution with a linear NaCl gradient (15 ml). In all, 0.5 ml fractions were collected and monitored for radioactivity by scintillation beta counting. Total eluted 3H-labeled HA was further concentrated by ethanolic potassium acetate precipitation in the presence of carrier HA (50 mg/ml), washed with 95% ethanol, dried under nitrogen and stored at 201C until analysis of its Kidney International (2006) 69, 272–280

original article

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

hydrodynamic size by Sephacryl S-1000 gel filtration chromatography under dissociative conditions.19,52

7.

8.

Qualitative assessment of HA by gel filtration chromatography To determine the hydrodynamic size of 3H-labeled HA, equal volumes of the CM and CE from HMC cultured under control or experimental conditions were treated with papain (15 IU/ml, 100 ml) at 601C for 1 h. Following papain inactivation with 4 M guanidine HCl buffer (400 ml), the material was chromatographed on an analytical Sephacryl S-1000 column (0.05  1.2 m). To confirm the presence of 3H-labeled HA, aliquots of papain-treated samples were further incubated with Streptomyces hyaluronidase (50 mIU/ml) before gel filtration chromatography.19,52 Fractions (0.5 ml) were collected at a rate of 2.4 ml/h, and the radioactivity was monitored by scintillation beta counting. The void (Vo) and total (Vt) volume of each column were calibrated with purified 3H-labeled HA from HMC cultured with 15% FBS and unincorporated [3H]-glucosamine, respectively. The hydrodynamic size of 3H-labeled HA synthesized de novo was calculated using the equation [(Ve–Vo)/ (Vt–Vo)], where Ve represents the fraction number of the eluted sample/peak. Statistics All experiments were repeated three times. Results are expressed as mean7s.d. Statistical analysis was performed using GraphPad Prism version 3.00 for Windows (GraphPad Software, San Diego, CA, USA). Results with ‘Active Ab’, ‘Remission Ab’ and controls were compared using the Mann–Whitney U-test. Correlations between IgG deposition and HA expression in renal biopsies, and between serum HA and anti-DNA levels were examined using the Spearman’s method. Two-tailed Po0.05 was considered statistically significant.

9.

10. 11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

ACKNOWLEDGMENTS

This work was supported by the Hong Kong Research Grants Council Earmarked Research Grants (HKU7215/01M and HKU7374/03M), CRCG Grant (10203346) and the Wai Hung Charity Foundation. We are grateful to Drs PC Tam and SM Chu and their surgical team for the collection of renal tissue, and to Dr KW Chan from the Department of Pathology for retrieving the renal biopsy specimens. Part of this work was presented at the 36th Annual Meeting of the American Society of Nephrology, San Diego, USA (J Am Soc Nephrol 2003; 14: 594A).

22. 23.

24.

25.

26.

REFERENCES 1. Tan EM. Autoantibodies to nuclear antigens (ANA): their immunobiology and medicine. Adv Immunol 1982; 33: 167–240. 2. Chan TM, Leung JKH, Ho KN, Yung S. Mesangial cell-binding anti-DNA antibodies in patients with systemic lupus erythematosus. J Am Soc Nephrol 2000; 13: 1219–1229. 3. Winfield JB, Faiferman I, Koffler D. Avidity of anti-DNA antibodies in serum and IgG glomerular eluates from patients with systemic lupus erythematosus. J Clin Invest 1997; 59: 90–96. 4. Termaat RM, Assmann KJM, von Son JPHF et al. Antigen-specificity of antibodies bound to glomeruli of mice with systemic lupus erythematosus-like syndromes. Lab Invest 1993; 68: 164–173. 5. DeGrendele H, Estess P, Siegelman MH. Requirement for CD44 in activated T cell extravasation into an inflammatory site. Science 1997; 278: 672–675. 6. Clark R, Wikner N, Doherty D, Norris D. Cryptic chemotactic activity of fibronectin for human monocytes resides in the 120-kDa fibroblastic cell-binding fragment. J Biol Chem 1988; 263: 12115–12123. Kidney International (2006) 69, 272–280

27.

28.

29. 30.

31.

Laskin DL, Soltys RA, Berg RA, Riley DA. Activation of alveolar macrophages by native and synthetic collagen-like peptides. Am J Respir Cell Biol 1994; 10: 58–64. McKee CM, Penno MB, Cowman M et al. Hyaluronan (HA) fragments induce chemokine gene expression in alveolar macrophages. The role of HA size and CD44. J Clin Invest 1996; 98: 2403–2413. Sampson PM, Rochester CL, Freundlich B, Elias JA. Cytokine regulation of human lung fibroblast hyaluronan (hyaluronic acid) production. Evidence for cytokine-regulated hyaluronan (hyaluronic acid) degradation and human lung fibroblast-derived hyaluronidase. J Clin Invest 1992; 90: 1492–1503. Laurent TC, Fraser RE. Hyaluronan. FASEB J 1992; 6: 2397–2404. Oksala O, Salo T, Tammi R et al. Expression of proteoglycans and hyaluronan during wound healing. J Histochem Cytochem 1995; 43: 125–135. Underhill CB, Nguyen HA, Shizari M, Cutly M. CD44 positive macrophages take up hyaluronan during lung development. Dev Biol 1993; 155: 324–336. Bartolazzi A, Peach R, Aruffo A, Stamenkovic I. Interaction between CD44 and hyaluronate is directly implicated in the regulation of tumour development. J Exp Med 1994; 180: 53–66. Lewington AJ, Padanilam BJ, Martin DR, Hammeramn MR. Expression of CD44 in kidney after acute ischemic injury in rats. Am J Physiol 2000; 278: R247–R254. Wells F, Larsson E, Tengblad A et al. The localization of hyaluronan in normal and rejected human kidneys. Transplantation 1990; 50: 240–243. Nishikawa K, Andres G, Bhan AK et al. Hyaluronate is a component of crescents in rat autoimmune glomerulonephritis. Lab Invest 1993; 68: 146–153. Sibalic V, Fan X, Loffing J, Wu¨thrich RP. Upregulated renal tubular CD44, hyaluronan and osteopontin in kdkd mice with interstitial nephritis. Nephrol Dial Transplant 1997; 12: 1344–1353. Feusi E, Sun L, Sibalic A et al. Enhanced hyaluronan synthesis in the MRL-Faslpr kidney: role of cytokines. Nephron 1999; 83: 66–73. Yung S, Coles GA, Davies M. IL-1b, a major stimulator of hyaluronan synthesis in vitro of human peritoneal mesothelial cells: relevance to peritonitis in CAPD. Kidney Int 1996; 50: 1337–1343. Jones S, Jones S, Phillips AO. Regulation of renal proximal tubular epithelial cell hyaluronan generation: implication for diabetic nephropathy. Kidney Int 2001; 59: 1739–1749. Wilkinson TS, Potter-Perigo S, Tsoi C et al. Pro- and anti-inflammatory factors cooperate to control hyaluronan synthesis in lung fibroblasts. Am J Respir Cell Mol Biol 2004; 31: 92–99. Lemay S, Mao C, Singh AK. Cytokine gene expression in the MRL/lpr model of lupus nephritis. Kidney Int 1996; 50: 85–93. MacPhee PJ. Estimating rat renal medullary interstitial oncotic pressures and the driving force for fluid uptake into ascending vasa recta. J Physiol (London) 1998; 506: 529–538. Ginetzinsky AG. Role of hyaluronidase in the reabsorption of water in renal tubules: the mechanism of action of antidiuretic hormone. Nature 1958; 182: 1218–1219. Turney JH, Davison AM, Forbes MA, Cooper EH. Hyaluronic acid in end-stage renal failure treated by haemodialysis: clinical correlates and implications. Nephrol Dial Transplant 1991; 6: 566–570. Ha¨llgren R, Engstrom-Laurent A, Nisbeth U. Circulating hyaluronate. A potential marker of altered metabolism of the connective tissue uremia. Nephron 1987; 46: 150–154. Johnsson C, Tufveson G, Wahlberg J, Ha¨llgren R. Experimentally-induced warm renal ischemia induces cortical accumulation of hyaluronan in the kidney. Kidney Int 1996; 50: 1224–1229. Ha¨llgren R, Gerdin B, Tufveson G. Hyaluronic acid accumulation and redistribution in rejecting rat kidney graft. Relationship to the transplantation edema. J Exp Med 1990; 171: 2063–2096. Toole BP. Hyaluronan: from extracellular glue to pericellular cue. Nat Rev Cancer 2004; 4: 528–539. Beck-Schimmer B, Oertli B, Pasch T, Wu¨thrich RP. Hyaluronan induces monocyte chemoattractant protein-1 expression in renal tubular epithelial cells. J Am Soc Nephrol 1998; 9: 2283–2290. Oertli B, Beck-Schimmer B, Fan X, Wu¨thrich RP. Mechanisms of hyaluronan-induced up-regulation of ICAM-1 and VCAM-1 expression by murine kidney tubular epithelial cells: hyaluronan triggers cell adhesion molecule expression through a mechanism involving activation of nuclear factor-kB and activating protein-1. J Immunol 1998; 161: 3431–3437.

279

original article

32.

33.

34.

35.

36. 37.

38.

39.

40.

41.

42.

280

Tanaka Y, Adams DH, Shaw S. Proteoglycans on endothelial cells present adhesion-inducing cytokines to leukocytes. Immunol Today 1993; 14: 111–115. Levesque MC, Haynes BF. Activated T lymphocytes regulate hyaluronan binding to monocyte CD44 via production of IL-2 and IFN-g. J Immunol 2001; 166: 188–196. Nakamura K, Yokohama S, Yoneda M et al. High, but not low, molecular weight hyaluronan prevents T-cell-mediated liver injury by reducing proinflammatory cytokines in mice. J Gastroenterol 2004; 39: 346–354. Wang A, Hascall VC. Hyaluronan structures synthesized by rat mesangial cells in response to hyperglycemia induce monocyte adhesion. J Biol Chem 2004; 279: 10279–10285. Faaber P, Capel PJ, Rijke GP et al. Cross-reactivity of anti-DNA antibodies with proteoglycans. Clin Exp Immunol 1984; 55: 502–508. Kashihara N, Hirakawa S, Mino Y et al. A study on cross-reactivity of anti-DNA antibody with glycosaminoglycans. Acta Med Okayama 1993; 47: 255–259. Prehm P. Synovial hyaluronate in rheumatoid arthritis binds C1q and is covalently bound to antibodies: a model for chronicity. Ann Rheum Dis 1995; 54: 408–412. Itano N, Sawai T, Yoshida M et al. Three isoforms of mammalian hyaluronan synthases have distinct enzymatic properties. J Biol Chem 1999; 274: 25085–25092. Devey ME, Lee SR, Le Page S et al. Serial studies of the IgG subclass and functional affinity of DNA antibodies in systemic lupus erythematosus. J Autoimmun 1988; 1: 483–494. Biji M, Dijstelbloem HM, Oost WW et al. IgG subclass distribution of autoantibodies differs between renal and extra-renal relapses in patients with systemic lupus erythematosus. Rheumatology 2000; 41: 62–67. Go¨ransson V, Johnsson C, Jacobson A et al. Renal hyaluronan accumulation and hyaluronan synthase expression after

S Yung et al.: Anti-DNA antibodies induce HA synthesis in HMC

43.

44. 45. 46.

47.

48.

49.

50.

51.

52.

ischaemia–reperfusion injury in the rat. Nephrol Dial Transplant 2004; 19: 823–830. Slevin M, Kumar S, Gaffney J. Angiogenic oligosaccharides of hyaluronan induce multiple signaling pathways affecting vascular endothelial cell mitogenic and wound healing responses. J Biol Chem 2002; 277: 41046–41059. Wight TN, Kinsella MG, Qwarnstrom EE. The role of proteoglycans in cell adhesion, migration and proliferation. Curr Opin Cell Biol 1992; 4: 793–801. West DC, Kumar S. Hyaluronan and angiogenesis. Ciba Found Symp 1989; 143: 187–201. Li Y, Rahmanian M, Widstrom C et al. Irradiation-induced expression of hyaluronan (HA) synthase 2 and hyaluronidase 2 genes in rat lung tissue accompanies active turnover of HA and induction of types I and III collagen gene expression. Am J Respir Cell Mol Biol 2000; 23: 411–418. Ito T, Williams JD, Fraser D, Phillips AO. Hyaluronan attenuates transforming growth factor-b1-mediated signaling in renal proximal tubular epithelial cells. Am J Pathol 2004; 164: 1979–1988. Bombardier C, Gladman DD, Urowitz MB et al. Derivative of the SLEDAI. A disease activity index for lupus patients. The Committee of prognosis Studies in SLE. Arthritis Rheum 1992; 53: 630–640. Yung S, Thomas GJ, Davies M. Induction of hyaluronan metabolism after mechanical injury of human peritoneal mesothelial cells in vitro. Kidney Int 2000; 58: 1953–1962. Peters H, Border WA, Ru¨ckert M et al. -Arginine supplementation accelerates renal fibrosis and shortens life span in experimental lupus nephritis. Kidney Int 2003; 63: 1382–1392. Yung S, Woods A, Chan TM et al. Syndecan-4 up-regulation in proliferative renal disease is related to microfilament organization. FASEB J 2001 published online 10.1096/fj.00-0794fje. Yung S, Coles GA, Williams JD, Davies M. The source and possible significance of hyaluronan in the peritoneal cavity. Kidney Int 1994; 46: 527–533.

Kidney International (2006) 69, 272–280