Elevated production of interleukin-18 is associated with renal disease in patients with systemic lupus erythematosus


Professor C. W. K. Lam, Department of Chemical Pathology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, NT, Hong Kong.  E-mail: waikeilam@cuhk.edu.hk


Summary To investigate the production mechanism and proinflammatory role of the cytokine interleukin (IL-18) in lupus nephritis, we investigated the plasma concentrations of IL-18 and nitric oxide (NO) and the release of IL-18 and NO from mitogen-activated peripheral blood monomuclear cells (PBMC), in 35 SLE patients with renal disease (RSLE), 37 patients without renal disease (SLE) and 28 sex- and age-matched healthy control subjects (NC). IL-18 and NO concentrations were measured by ELISA and colourimetric non-enzymatic assay, respectively. Gene expressions of IL-18 and IL-18 receptor were analysed by RT-PCR. Plasma IL-18 and NO concentrations were significantly higher in RSLE than NC (both P < 0·01). Elevation of plasma IL-18 in RSLE correlated positively and significantly with SLE ­disease activity index and plasma NO concentration (r = 0·623, P < 0·0001 and r = 0·455, P = 0·017, respectively), and the latter also showed a positive and significant correlation with plasma creatinine (r = 0·410, P = 0·034) and urea (r = 0·685, P < 0·0001). There was no significant difference in gene expressions of IL-18 and IL-18 receptor in PBMC among RSLE, SLE and NC. Percentage increase in culture supernatant IL-18 concentration was significantly higher in RSLE than SLE and NC (both P < 0·05). The basal NO release was significantly higher in RSLE than that in SLE and NC (both P < 0·005). IL-18 is therefore suggested to play a crucial role in the inflammatory processes of renal disease in SLE.


Systemic lupus erythematosus (SLE) is a systemic autoimmune disorder characterized by the activation of T and polyclonal B lymphocytes, production of autoantibodies, and formation of immune complexes causing tissue and organ damage [1]. In some patients, skin rash and joint pain predominate, while in others glomerulonephritis is the main lesion [1]. The aetiology and pathogenetic mechanisms of this immunological disorder have not been clearly elucidated. Although abnormal production of T helper (Th) cell cytokines [2] and chemokines [3,4] have been implicated, such aberration in SLE is very complex and requires further investigation.

IL-18, formerly called interferon (IFN)-γ-inducing factor, is a proinflammatory cytokine related to the IL-1 family that is ­produced by Kupffer cells, activated macrophages, keratinocytes, intestinal epithelial cells, osteoblasts and adrenal cortex cells [5]. It plays an important role in the innate immunity and Th1 response to toxic shock and shares functional similarities with IL-12 [5]. Human IL-1 receptor protein is a functional component of the IL-18 receptor [6]. IL-18 receptors are selectively expressed on murine Th1 cells but not Th2 cells [7]. The primary functions of IL-18 include the induction of IFN-γ and TNF-α in T cells and natural killer (NK) cells [8,9], IL-8 in eosinophils [10], up-regulation of Th1 cytokines including IL-2, granulocyte macrophage colony-stimulating factor (GM-CSF) and IFN-γ[5], stimulation of the proliferation of activated T cells [5], and enhancement of Fas ligand expression in NK and cytotoxic T lymphocytes [11]. Elevated IL-18 levels have also been demonstrated in the urine of nephrotic patients [12], serum of patients with multiple sclerosis [13], adult-onset Still's disease [14], type I diabetes mellitus [15], viral infection [16,17], sepsis [18], allergic asthma [19] and inflammatory rheumatic disease [20].

We have recently reported significantly elevated plasma IL-18 concentration in SLE patients compared to controls [21], and the elevation of IL-18 to IL-4 ratio was positively correlated to SLE disease activity index (SLEDAI) [22]. However, the functional role of IL-18 is not well understood. Earlier studies have suggested that increased nitric oxide (NO) and IL-12 production could be important in the pathogenesis of glomerulonephritis in autoimmunity [23]. In an attempt to elucidate the inflammatory role of IL-18 in SLE, we studied the production of IL-18 and NO, and the mRNA expression of IL-18 and IL-18 receptor (R) α, β-chain in patients with or without renal disease. Prompted by the work of Liu and Jones (1998) [24] on the impaired production of IL-12 by peripheral blood mononuclear cells (PBMC), we also investigated the mechanism of in vitro production of IL-18 by PBMC upon their activation with different lymphocyte mitogens including T cell mitogen, phytohaemagglutinin (PHA) and B cell mitogen, lipopolysaccharide (LPS) for understanding the pathogenic role of IL-18.

Materials and methods

SLE patients, control subjects and blood samples

Seventy-two Chinese SLE patients were recruited at the Rheumatology Out-Patient Clinic of the Prince of Wales Hospital, Hong Kong. Diagnosis of SLE was established according to the 1982 revised American Rheumatism Association criteria (ARA) [25], and disease activity evaluated by the SLEDAI [26]. The SLE patients were divided into two groups: 35 SLE patients with renal disease (RSLE group) and 37 SLE patients without renal disease (SLE group). Twenty-eight sex- and age-matched healthy Chinese volunteers were recruited as controls (NC). Twenty ml of heparinized venous peripheral blood were collected from each patient and control subject. The above protocol was approved by the Clinical  Research  Ethics  Committee  of  the  Chinese  University of Hong Kong and informed consent was obtained from all ­participants.

Whole blood assay (WBA)

The method of Viallard et al. [27] was adopted. After a maximum storage period of 1 h of collected heparinized blood at room temperature, blood was diluted 1 : 1 with RPMI 1640 (Gibco Laboratories, NY, USA), and 1 ml aliquots were deposited in each well of a 24-well plate (Nalge Nunc International, IL, USA). The blood culture was then incubated with or without phytohaemagglutinin (PHA) (Sigma Co., MO, USA) at 5 µg/ml and lipopolysaccharide (LPS) at 25 µg/ml (Sigma) for 24 h at 37°C in a 5% CO2 atmosphere. After incubation, the cell-free supernatant was harvested and stored at −70°C until enzyme-linked immunosorbent assay (ELISA). The percentage (%) increase of IL-18 release  from  PBMC  of  each  group  after  the  incubation  with PHA and LPS was calculated as follows:

[(treatment – control)/control]× 100%.

IL-18 assays

Plasma and culture supernatant bioactive IL-18 concentrations of SLE patients and control subjects were measured by ELISA using human IL-18 ELISA kit of Medical & Biological Laboratories Co Ltd, Nagoya, Japan.

Assay for NO

Plasma NO in SLE patients and control subjects were measured in terms of nitrite concentrations as NO is rapidly converted to nitrites and nitrates. A colourimetric non-enzymatic assay kit (Oxford Biomedical Research Inc., MI, USA) was used for measuring the concentrations of total nitrite in plasma and culture supernatant. Briefly, samples were deproteinized by precipitation in zinc sulphate. After centrifugation, the supernatants and the nitrite standards were mixed overnight with 0·5 g granular cadmium beads for reduction of nitrate to nitrite. Following removal of cadmium beads, the supernatants were mixed with the Griess reagent in a 96-well flat-bottomed microtitre plate and the absorbance of developed colour was measured at 540 nm with a microtiter plate reader.

Reverse transcription-polymerase chain reaction (RT-PCR)

PBMC were isolated from heparinzed venous blood for each subject by Ficoll-Paque density gradient centrifugation (Amersham Pharmacia Biotech Ltd, Uppsala, Sweden). Total RNA from PBMC was extracted using RNeasy Mini Kit (QiagenGmbH, Hilden, Germany) according to the manufacturer's instructions. Extracted RNA was reverse transcribed into first-stand complementary DNA using first-strand cDNA Synthesis Kit (Amersham and Pharmacia Biotech). PCR was performed in a reaction mixture containing 3 mm MgCl2, 250 µm dNTPs, 2 units of AmpliTaq Gold DNA polymerase (Perkin Elmer, CA., USA), 50 pmol of 5′ and 3′ primers (Gibco) in PCR reaction buffer for 33 cycles (94°C for 1 min, 60°C for 1 min, and 72°C for 1 min) after an initial 12 min of denaturation at 94°C. All RT-PCR were performed in the linear range of the PCR reaction according to the preliminary experiment.   PCR   Primers   were   as   the   following:   IL-18   sense, 5′-AGCTCGGGATCCATGTACTTTGGCAAGCTTGAATCT AAATTATCA-3′ and antisense, 5′-ACTGAATTCCTAGTCT TCGTTTTGAACAGTGAACATTATAGA-3′, yielding a 494-bp product [28]; IL-18 Rα sense, 5′-CCCAACGATAAAGAA GAACGCC-3′ and antisense, 5′-TGTCTGTGCCTCCCGTG C­TGGC-3′,     yielding     a     419-bp     product     [9];     IL-18     Rβ    sense, 5′-AACACAACCCAGTCCGTCCAA-3′ and antisense, 5′-AACATCAGGAAATAGGCTCAG-3′, yielding a 291-bp product [9]; β-actin sense, 5′-AGCGGGAAATCGTGCGTG-3′ and antisense, 5′-CAGGGTACATGGTGGTGCC-3′, yielding a 300-bp product [29]. After the amplification reaction using PTC-200 DNA EngineTM (MJ Research, Inc., MA, USA), PCR products were electrophoresed on 2% agarose gel in TAE buffer (pH 8·0) and stained with ethidium bromide. The electrophorectic bands were documented with Gene Genius Gel Documentation System (Syngene Inc., Cambridge, UK) and intensities of PCR band quantified using Bio-Rad Quantity OneTM software (Bio-Rad Laboratories, CA, USA). The ratio between intensities of PCR bands   of   IL-18,   IL-18   Rα,   β  and   their   corresponding   β-actin was calculated and expressed as relative intensity of mRNA ­expression.

Statistical analysis

Because IL-18 and NO concentrations were not in a Gaussian distribution, the Mann–Whitney rank sum test was used to assess their differences between patient and control groups. The Spearman's rank correlation test was used to ascertain the correlations between plasma IL-18, NO and SLEDAI. Results were expressed as median (interquartile range) or mean ± standard deviation (s.d.) as appropriate. All analyses were performed using the statistical package for the Social Sciences (SPSS) statistical software for Windows, Version 9·0 (SPSS Inc., IL, USA). A probability (P) < 0·05 was considered as significantly different.


SLE patients and control subjects

The age, sex, SLEDAI score, duration of diagnosis, serum creatinine   and   urea   and   drug   treatment   of   the   study   populations are summarized in Table 1. Thirty-five SLE patients with renal diseasse (RSLE: 34 females and one male, mean ± s.d. age of 39·1 ± 10·1 years, range 20–59) and 37 SLE patients (SLE: 37 females, 39·3 ± 10·8 years, range 20–67) were recruited. The mean duration of the diagnosis of SLE at the time when patients were evaluated for this study was 12·4 ± 6·3 years (range 1·7–26·6) and 9·0 ± 6·8 years (range 0·3–25·6 years) for RSLE and SLE patients, respectively. The SLEDAI scores of RSLE and SLE patients were 7·9 ± 5·9 (range 0–20) and 2·8 ± 5·6 (range 0–32), respectively. The mean ± s.d. serum creatinine concentrations of RSLE and SLE patients were 105·2 ± 73·0 and 68·2 ± 11·7 µmol/l (normal range: 44–107), P < 0·05, and corresponding serum urea concentrations 8·5 ± 5·7 and 4·7 ± 1·3 mmol/l (normal range 3·4–8·9), P < 0·0001. Twenty-eight healthy control subjects (NC: 27 females and one male, aged 38·5 ± 7·9 years, range 22–51) were recruited. There was no significant difference among the ages of the RSLE or SLE patients and NC subjects (all P > 0·05), and all the three groups were sex-matched.

Table 1.  Characteristics of RSLE, SLE patients and control subjects
  1. n.a. = not applicable; n.d. = not determined. *P < 0·05 and **P < 0·0001 compared with SLE patients (Mann–Whitney rank sum test).

Sex (female/male)34/137/027/1
Age, year (mean ± s.d., range)39·1 ± 10·139·3 ± 10·838·5 ± 7·9
SLEDAI score (mean ± s.d., range)7·9 ± 5·92·8 ± 5·6n.a.
Duration of diagnosis, years (mean ± s.d., range)12·4 ± 6·39·0 ± 6·8n.a.
Serum creatinine, µmol/l (mean ± s.d., range)105·2 ± 73·0*68·2 ± 11·7n.d.
Serum urea, mmol/l (mean ± s.d., range)8·5 ± 5·7**4·7 ± 1·3n.d.
Treatment with prednisolone
 Patient no. (%)35 (100·0)19 (51·4)n.a.
 Daily dose, mg6·7 ± 5·13·2 ± 6·6 
Treatment with hydroxychloroquine
 Patient no. (%)14 (40·0)23 (62·2)n.a.
 Daily dose, mg87·5 ± 99·0112·0 ± 108·0 
Treatment with azathioprine
 Patient no. (%)13 (37·1)5 (13·5)n.a.
 Daily dose, mg20·0 ± 30·09·0 ± 30·0 
Treatment with cyclosporin A
 Patient no. (%)9 (25·7)0 (0·0)n.a.
 Daily dose, mg20·9 ± 43·50 

Plasma concentrations of IL-18 and NO

Plasma concentrations of cytokine IL-18 was significantly higher in RSLE and SLE than NC [median (interquartile range): RSLE 203·8 (143·7–392·3) and SLE 207·9 (119·8–271·1) versus NC 127·0 (82·6–195·2) pg/ml, both P < 0·01]. There was significant and positive correlation between IL-18 concentration and SLEDAI score in RSLE patients (r = 0·623, P < 0·0001) (Fig. 1a). However, no significant  correlation  was  observed  between  IL-18  concentration and SLEDAI score in SLE patients (r = 0·251, P = 0·135) (Fig. 1b). Plasma NO concentration was significantly higher in RSLE patients than NC [291·2 (183·4–465·4) versus 172·7 (105·1–249·7) µm, P = 0·009], but no significant difference was found between SLE and NC (P = 0·075). In RSLE, plasma IL-18 concentration  correlated  significantly  with  plasma  NO  concentration (r = 0·455, P = 0·017), and the latter also showed a positive and significant correlation with plasma creatinine (r = 0·410, P = 0·034) and urea (r = 0·685, P < 0·0001).

Figure 1.

Correlation between plasma IL-18 concentration and SLEDAI of (a) RSLE and (b) SLE patients. (a) r = 0·623; P < 0·001. (b) r = 0·251; P = 0·135.

In vitro IL-18 and NO production using WBA

In the absence of PHA and LPS stimulation, the spontaneous in vitro production of IL-18 from PBMC was similar in groups of RSLE, SLE patients and NC (P > 0·05). After 24 h incubation with PHA and LPS, culture supernatant IL-18 concentrations was significantly elevated in RSLE (P = 0·03), SLE (P = 0·004) and NC (P = 0·027) compared to medium control (Fig. 2a). The percentage increase in culture supernatant IL-18 level of RSLE was significantly higher than that of SLE and NC [RSLE 93·6 (45·5–171·1) versus SLE 39·2 (15·3–106·5), P = 0·031; and NC 43·8 (27·2–88·6) pg/ml, P = 0·013] (Fig. 2b). As shown in Fig. 3, basal NO release of PBMC in RSLE was significantly higher than that of SLE and NC [RSLE 425·6 (412·8–443·7) versus SLE 386·6 (363·3–423·8), P = 0·005; RSLE 425·6 (412·8–443·7) versus NC 385·6 (355·5–409·8) µm, P = 0·0002]. However, PHA and LPS did not show any significant effect on NO release from all three groups (data not shown).

Figure 2.

IL-18 release from PBMC of RSLE, SLE patients and NC. (a) Supernatant IL-18 concentrations in PBMC culture without or with PHA (5 µg/ml) and LPS (25 µg/ml) for 24 h; (b) % increase of IL-18 release from PBMC after the incubation with PHA (5 µg/ml) and LPS (25 µg/ml) for 24 h.

Figure 3.

NO release from PBMC of RSLE, SLE patients and NC after 24 h culture.

RT-PCR of IL-18 and IL-18 receptor a and b

PBMC from all subjects were found positive for the expressions of IL-18 and IL-18Rα, β. As shown in Figs 4 and 5, there was no significant difference of gene expressions for IL-18 and IL-18 Rα, β in PBMC among groups of RSLE, SLE and NC (all P > 0·05).

Figure 4.

Representative results for the mRNA expression of IL-18, IL-18 Rα, β in PBMC from RSLE, SLE and NC using semiquantitative RT-PCR. PBMC was purified from heparinized blood from subjects and total RNA was extracted, reverse transcribed and analysed by PCR as described. (a) IL-18; (b) IL-18 Rα and (c) IL-18 Rβ. M: 100 base pair ladder molecular weight marker.

Figure 5.

Expression of mRNA of IL-18, IL-18 Rα, β in PBMC from RSLE, SLE and NC using semiquantitative RT-PCR. PBMC was purified from heparinized blood from subjects and total RNA was extracted, reverse transcribed and analysed by PCR as described. (a) relative band intensity of IL-18; (b) relative band intensity of IL-18 Rα and (c) relative band intensity of IL-18 Rβ. The results are expressed as mean ± s.d. The difference between NC, SLE, RSLE were assessed by Mann–Whitney rank sum tests, all P > 0·05.


It has been suggested that SLE is a Th2-polarized disease because of its production of autoantibodies specific for self-antigens [30]. However, other studies have demonstrated that serum cytokines for Th1 response including IL-12 [31], TNF-α[32] and IFN-γ[33] were also significantly higher in SLE patients. Our previous studies have demonstrated that SLE patients exhibited significantly higher plasma concentrations of proinflammatory cytokine IL-12, IL-17 and IL-18, and Th2 cytokine IL-4 [21,22]. Using MRL/lpr mice with spontaneous lupus-like autoimmune disease, it was shown that daily injection of IL-18 or IL-18 plus IL-12 resulted in accelerated protienuria, glomerulonephritis and raised levels of proinflammatory cytokines in MRL/lpr mice [34]. Therefore, IL-18 is suggested to be an important mediator of lupus-like disease including lupus nephritis. SLE patients with diffuse proliferative lupus nephritis also showed the predominance of Th1 immune response and the peripheral blood Th1 to Th2 ratio could be useful as a parameter that reflects the renal histological activity [35].

Our present results demonstrated that the elevation of plasma IL-18 level correlated positively with NO and SLEDAI in SLE patients with renal disease, but not in patients without renal ­disease. These further indicate that IL-18 played an inflammatory role in glomerulonephritis of SLE patients. As also observed by us, the plasma IL-18 concentrations in RSLE and SLE patients did not show any correlations with the dosages of prednisolone, hydroxychloroquine, azathioprine and cyclosporin A. The mechanistic investigation of IL-18 production by PBMC was studied using WBA, which can preserve better the natural environment and constitute an appropriate milieu for studying production of in vitro cytokines [27]. It also preserves the natural intercellular interactions and circulating stimulatory and inhibitory mediators, including soluble receptors that are present at their physiological concentrations. Our finding of significant increase in IL-18 release from culture supernatant of RSLE but not SLE and NC points to a heightened IL-18 secretory capability of PBMC of RSLE patients. This further strengthens that IL-18 is an important inflammatory mediator in SLE with renal disease.

Using RT-PCR, all PBMC of RSLE, SLE and NC groups showed similar levels for the expressions of IL-18 and IL-18 Rα and β (Figs 4 and 5). Although the gene expression of IL-18 Rβ but not IL-18Rα in lymph node cells of MRL lpr/lpr mice has been shown to be increased comparing with non-lupus mice [36], the discrepancy between the animal study and ours may be due to the different characteristics of murine lymph node cells and human PBMC in these two studies. Moreover, the SLE patients in our study exhibited a wide range of SLEDAI while MRL lpr/lpr mice may have a more homogeneous disease activity. Nevertheless, continued expression of IL-18Rα, β in PBMC (Figs 4 and 5) suggested that the Th1 cells and NK cells could be activated by the elevated IL-18 to trigger inflammatory response in SLE patients. Although plasma IL-18 level was significantly higher in SLE and RSLE than NC, gene expression of IL-18 was similar between patients and NC (Figs 4 and 5). As our ELISA measured only the active form of IL-18 but not the inactive pro-IL-18, the discrepancy between results of ELISA and RT-PCR of IL-18 may be due to  the  similar  level  of  gene  expression  of  inactive  pro-IL-18  in all three groups but up-regulated post-translational cleavage of inactive pro-IL-18 by caspase-1 in PBMC of SLE and RSLE. Therefore, the intracellular activity of caspase-1 should be investigated further in PBMC from control and SLE patients with or without renal disease. In our study of patients with rheumatoid arthritis, we also could not detect any significant change in expressions of IL-18 and IL-18 receptors in PBMC compared with normal control using RT-PCR and cDNA expression array (data not shown). However, PBMC is a mixture of various types of T cells, monocytes, macrophages and granulocytes. The present RT-PCR analysis of IL-18 and IL-18 receptors cannot illustrate the change of gene expression of individual cell types. Therefore, further experiments using purified cell types (e.g. T helper cells, macrophages, etc.) are required.

A previous study has showed that IL-18 accelerates spontaneous autoimmune disease with characteristic glomerulonephritis and vasculitis [34]. The inflammation is exacerbated further by the synergistic action of elevated IL-12 and IL-18 in SLE patients [21,22]. Our present data also revealed that the plasma NO concentration was significantly higher in SLE patients with renal disease. Moreover, we observed that plasma concentrations of IL-18 correlated significantly with plasma NO and the latter showed a positive correlation with plasma creatinine and urea in RSLE. Excessive NO production may play a pivotal role in the pathogenesis of renal disease in SLE. In murine models, NO was found to be related in the pathogenesis of arthritis and glomerulonephritis [37,38]. Elevated production of NO has also been documented in rheumatoid arthritis [39,40] and SLE patients [41,42]. Up-regulated expression of inducible NO synthase (iNOS) was found in the kidneys of patients with active glomerulonephritis, including those with lupus [41]. It is therefore possible that pathogenesis may first involve IL-18 and IL-12 promoting Th1 cell activation, which augments Th1 cytokine IFN-γ production, thereby inducing the expression of NO synthease and the production of NO that mediate glomerulonephritis and vasculitis [43]. Accordingly, our present study should also suggest that IL-18 may serve as a potential target for treatment of autoimmune diseases, including SLE.


This study was supported by a Direct Grant for Research of The Chinese University of Hong Kong and a donation from Zindart (De Zhen) Foundation Ltd, Hong Kong.