Consecutive patients aged >18 years who met the European League against Rheumatism/American College of Rheumatology classification criteria for SLE (21) were recruited from the Sapienza Lupus Clinic of Rome. Patients were recorded as having LN if they fulfilled one of the following renal criteria: persistent proteinuria greater than 0.5 grams per day or greater than 3+ if quantitation not performed or cellular casts (22). The LN patients were divided into two groups according to the presence (a-LN) or absence (r-LN) of LN activity based on Systemic Lupus Erythematosus Disease Activity Index SLEDAI score (23). Patients who underwent a biopsy for the evaluation of kidney involvement as a standard of care management were classified in accordance with the ISN/RPS classification criteria (24). For each patient the following variables were recorded at study entry: age, sex, ethnicity, disease duration, hypertension, smoking habits and drug therapy.

All patients and control subjects included in this study provided written informed consent. Data were used anonymously in accordance with the latest version of the Helsinki Declaration of human research ethics. The study protocol was approved by the local ethics committees of the Policlinico Umberto I, Rome, Italy.

A 5-ml peripheral venous blood sample was drawn from each study participant; sera were obtained by centrifugation and stored at −20°C until the examination. A fresh urine sample (midstream) from each patient was collected in a sterile container; the urine was centrifuged to remove sediments then frozen in aliquots at −80°C for later test.

Renal tissues were obtained from LN patients and, as control, from HC at the time of kidney donation.

Serum and urinary IL-32 evaluations were assessed using commercially available IL-32 DuoSet^® ELISA KIT (R&D systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. The same kit was used for the evaluation of IL-32 in urine samples. This sandwich ELISA kit is validated for serum samples, but like other ELISA kits or multiplex immunoassays kit, specific for the dosage of cytokines, it is also used for other biological fluids such as urine, after centrifugation and urinary sediment discard (25–27).

Paraffin-embedded sections of 5-μm thickness were stained with haematoxylin and eosin (H&E) for the histological evaluation. Immunohistochemistry was performed on 5-μm-thick paraffin-embedded sections from SG, as previously described (12). The primary antibody mouse anti-human IL-32 (Novus Biologicals, Littleton, CA) was added and incubated for 1 h at room temperature (RT). Isotype-matched irrelevant antibody was used as a negative control (Abcam, Cambridge, UK). The number of positive cells was determined by counting the reactive cells on microphotographs obtained from three randomly selected high-power microscopic fields (original magnification × 400).

Human embryonic kidney 293 (HEK293) cells, stably transfected with TLR3 gene (HEK293T3) (InvivoGen, San Diego, CA, USA) were maintained in Dulbecco’s Modified Eagle Medium (DMEM, Sigma-Aldrich, Milan, Italy), containing 50 U/ml penicillin, 50 mg/ml streptomycin, 100 mg/ml Normocin™, 2 mM L-glutamine, 10 μg/ml Blasticidin as selective antibiotic, 10% fetal calf serum (FCS, Sigma-Aldrich).

In some experiments control untransfected HEK293 were used. Cell cultures were grown under standard conditions at 37°C in a humified atmosphere containing 5% CO₂.

Ig fractions were isolated from sera of a-LN patients or healthy donors using a slightly modified protocol involving precipitation with, as previously described (28). HEK293T3 cells were incubated, at 37°C for different incubation times, with purified IgG from LN patients (LN-IgG, 200 μg/ml), normal human serum IgG (NHS- IgG, 200 μg/ml) or 10 μg/ml Poly(I:C) (InvivoGen), a synthetic analog of dsRNA considered a potent activator of TLR3 and can therefore be used as a positive control. To investigate TLR3 involvement, cells were also treated with a TLR3 inhibitor (27 μM, TLR3/dsRNA Complex Inhibitor, Merck Millipore, Milano, Italy) alone or in combination with LN-IgG.

We preliminary determined the optimal IgG concentration and incubation time on the basis of concentration/time curve; all the experiments were shown at the best concentration and incubation time.

To prepare whole protein extracts, HEK293T3 cells, untreated or treated with LN-IgG, NHS-IgG, Poly(I:C), or TLR3 inhibitor, for 45 min, were resuspended in RIPA lysis buffer (50 mM Tris-HCl pH 7.4, 0.5% Triton X-100, 0.25% Nadeoxycholate, 0.1% SDS, 150 mM NaCl, 1mM EDTA and 5mM MgCl₂, including proteases and phosphatases inhibitors). Protein content was determined by Bradford assay, using bovine serum albumin (BSA) as standard (Bio-Rad, Richmond, CA, USA). Equal amounts of whole protein extracts were subjected to 10% sodium dodecyl sulfate (SDS-PAGE) and then blotted onto polyvinylidene difluoride (PVDF) membranes (Bio-Rad). After blocking with 5% defatted dried milk in Tris-buffered saline (TBS), containing 0.05% Tween-20 membranes were probed with rabbit polyclonal anti-phospho-TBK1/NAK or anti-phospho-NF-κB p65 antibodies (Cell Signaling, Inc Danvers, MA, USA). Bound antibodies were visualized with HRP-conjugated anti-rabbit IgG antibodies (Sigma-Aldrich) and immunoreactivity was assessed by the enhanced chemiluminescence (ECL) Western blotting system (Amersham Pharmacia Biotech, Piscataway, NJ, USA). To control nonspecific reactivity and verify sample loading, PVDF membranes were stripped and reprobed with anti-rabbit IgG and anti-β-actin Abs respectively (Sigma-Aldrich). Densitometric scanning analysis was performed by Mac OS X (Apple Computer International), using NIH Image 1.62 software. The density of each band in the same gel was analysed, and the densitometric TBK1/β-actin or p65/β-actin ratios are shown.

Sample of HEK293T3 and HEK293 cells untreated or treated, for 16 h, with LN-IgG, NHS-IgG, Poly(I:C), or TLR3 inhibitor were lysed as described above, separated in 10% SDS-PAGE and then analysed by Western blot as previously described. Membranes were incubated with rabbit polyclonal anti-IL-32 (Abcam) followed by HRP-conjugated anti-rabbit IgG antibodies (Sigma-Aldrich). To control nonspecific reactivity and verify sample loading, PVDF membranes were stripped and reprobed with anti-rabbit IgG and anti-β-actin Abs respectively (Sigma-Aldrich. Densitometric scanning analysis was performed by Mac OS X (Apple Computer International), using NIH Image 1.62 software. The density of each band in the same gel was analysed, and the densitometric IL-32/β-actin ratios are shown.

All the statistical analyses were performed by GraphPad Prism software Inc., (San Diego, CA, USA). Data were expressed as mean ± standard deviation (SD) or median [interquartile range (IQR)], according to the distribution. Kolmogorov-Smirnov test was used to assess the normal distribution of the data. Differences between numerical variables were evaluated with Kruskal-Wallis and Mann-Whitney tests. Correlation was tested with Pearson’s correlation coefficient in normally distributed variables or Spearman’s rank order in variables which are not normally distributed. For comparison of categorical variables or percentages we used Fisher’s exact and X² tests when appropriate. For Western blot experiments, statistical analysis was performed using Student’s t test. P-values less than 0.05 were considered as significant.

Sixty SLE patients with LN (51 females and 9 males) with a mean age of 37 years (SD 11.65), 50 SLE patients without renal involvement (43 females and 7 males) with a mean age of 45 years (SD 14.5) and 30 healthy control samples recruited from the donor transfusion center of our hospital, with age and sex carefully matched with the patients groups to ensure no significant differences. All the patients and controls were Caucasian. Table 1 shows demographic and clinical features of SLE patients with and without renal involvement. Forty (66.7%) LN patients had an active renal disease [a-LN, median SLEDAI-2K 11.5 (3-5)] and 20 patients (33.3%) were in renal remission [r-LN, median SLEDAI-2K 0 (0-8)]. Table 2 summarizes clinical and demographic data of LN patients.

IL-32 serum levels were significantly higher in patients with r-LN [median 1368 (3910)] compared to SLE patients without renal involvement [median 203 (662.8)] (p=0.03) ( Figure 1A ). Urinary levels of IL-32 were analysed in 40 LN patients, 20 SLE patients without renal involvement and 20 HC. We didn’t find any significant difference in urinary level of IL-32 through the groups.

We observed a direct correlation between IL-32 serum levels and disease duration in LN patients (p=0.02; r 0.2978) ( Figure 1B ).

To evaluate the tissue expression of IL-32 in LN patients, immunohistochemical analysis was performed on renal biopsies of 20 a-LN and 8 HC. Histologically, normal kidney did not exhibit significant immunostaining for IL-32; on the contrary, IL-32 was strongly expressed in renal samples of LN patients, especially in those with class III and IV LN, compared to HC group ( Figure 2 ). Intense IL-32 expression was observed in epithelial cells of proximal and distal tubular segments and within mesangial cells; IL-32-expressing lymphocytes were also observed infiltrating the peritubular interstitium ( Figures 2A–D ). The semiquantitative scores for IL-32 in epithelial cells were highly correlated with those for microscopic inflammation on H&E sections (R = 0.57 and p < 0.0001)

Analysis, by Western blot, of cell lysates from HEK293/T3 showed that LN-IgG, as well as Poly(I:C), a TLR3 agonist used as positive control, induced TBK1 phosphorylation ( Figure 3A ) and NF-κB activation ( Figure 3B ), as revealed, respectively, by anti-phospho-TBK1/NAK and anti-phospho-NF-κB p65 antibody reactivity. Virtually, no increase in this activation was observed in untreated cells or following NHS-IgG treatment. On the contrary, TBK1 and NF-κB p65 activation significantly decreased when cells were treated with TLR3 inhibitor in combination with LN-IgG. As a control, no significant activation was evident when cells were treated with TLR3 inhibitor alone. The specificity of the antibody reaction was also verified using anti-rabbit IgG and in western blot results no band appeared ( Figure 3 ).

To determine whether LN-IgG were able to trigger IL-32 expression in a TLR3-dependent manner, we analyzed, by Western blot, HEK293/T3 and control untransfected HEK293 cells. Results showed that IL-32 expression was significantly increased in HEK293/T3 cells stimulated with LN-IgG compared to untreated cells or treated with NHS-IgG ( Figure 4A ). IL-32 expression induced by LN-IgG was similar to that one triggered by Poly(I:C), a TLR3 agonist used as positive control. The TLR3 inhibitor used in combination with LN-IgG induced a significant reduction of IL-32 expression. The specificity of the antibody reaction was also verified using anti-rabbit IgG and no band was evident observing obtained results ( Figure 4A ). Interestingly, results obtained analyzing untransfected HEK293 cell samples showed a significant lower IL-32 expression ( Figure 4B ). Therefore, these data further support the hypothesis of TLR3 involvement in LN-IgG-induced IL-32 production through activation of TBK1 and NF-κB.