Functional benefits of corticosteroid and IVIG combination therapy in a coronary artery endothelial cell model of Kawasaki disease

Recombinant human TNF-α, IL-1α and IL-1β were purchased from PeproTech (Rocky Hill, NJ, USA). DEX was purchased from Sigma-Aldrich (St. Louis, MO, USA). A human immunoglobulin preparation (Venoglobulin IH; USA) was provided by Japan Blood Products Organization (JB: Tokyo, Japan).

HCAECs were purchased from Lonza (Walkersville, MD, USA) and maintained exactly as recommended by the manufacturer by using an EGM-2MV BulletKit (Lonza) at 37 °C in a humidified 5% CO2 atmosphere. We purchased two different HCAEC lots from individual donors (Lot Nos. 0000662152 and 0000626782) for this study, and all results were reproducible between these two lots. All the experiments described in this study were performed using second-passage cells.

Because the EGM-2MV BulletKit contains hydrocortisone, a corticosteroid, all the experiments described in this study were performed after hydrocortisone deprivation for at least 3 h to fairly evaluate the effects of DEX. More specifically, HCAECs were suspended in complete EGM-2MV medium and seeded into 96-well Optical-Bottom plates (Thermo Fisher Scientific; Waltham, MA, USA) at 1 × 10⁴ cells/well for apoptosis assay, 48-well culture plates (IWAKI AGC Techno Glass; Shizuoka, Japan) at 2 × 10⁴ cells/well for quantitative PCR (qPCR) and ELISA, and 24-well culture plates (IWAKI AGC Techno Glass) at 5 × 10⁴ cells/well for Western blotting (WB). The cells were grown to 90% confluency. Prior to stimulation, the medium in each well was replaced with hydrocortisone-deprived EGM-2MV for at least 3 h. The medium in each well was then replaced with hydrocortisone-deprived EGM-2MV medium containing the stimulant(s) and/or pharmacological agent(s) such as DEX and IVIG, as indicated in the figure legends.

The concentrations of HMGB1, IL-1α, IL-6 and G-CSF proteins in cell-free supernatants were measured with specific ELISA kits (R&D Systems; Minneapolis, MN, USA) in accordance with the manufacturer’s instructions.

Total RNA extraction from HCAECs, cDNA synthesis and qPCR were performed as previously described [18]. Primer sets for five genes were synthesized at Fasmac (Kanagawa, Japan): HMGB1 (sense, 5′-AGA AGT GCT CAG AGA GGT GGA-3′; antisense, 5′-CCT TTG GGA GGG ATA TAG GTT-3′), IL-1α (sense, 5′-CAA CCA GTG CTG AAG GAG-3′; antisense, 5′-TGC CGT GAG TTT CCC AGA AG-3), IL-6 (sense, 5′-CAA TAA CCA CCC CTG ACC CA-3′; antisense, 5′-GCG CAG AAT GAG ATG AGT TGT C-3′), G-CSF (sense, 5′-TGC TTA GAG CAA GTG AGG AAG ATC-3′; antisense, 5′-GCA CAC TCA CTC ACC AGC TTC T-3′), and β-actin (sense, 5′-CCC AGC CAT GTA CGT TGC TAT-3′; antisense, 5′-TCA CCG GAG TCC ATC ACG AT-3′). To determine the exact copy numbers of the five target genes, quantified concentrations of the purified PCR products of HMGB1, IL-1α, IL-6, G-CSF and β-actin were serially diluted and used as standards in each experiment. We used an aliquot of cDNA equivalent to 2 ng of each total RNA sample for each qPCR. The expression levels of mRNA were normalized to the β-actin level in each sample.

Whole cells were extracted with 200 μl of NuPAGE LDS sample buffer (Invitrogen; Carlsbad, CA, USA) containing 5% 2-mercaptethanol and then lysed by sonication. Equal amounts of whole-cell lysates were separated by SDS-PAGE (5–15% Ready Gels, Bio-Rad; Hercules, CA, USA) and transferred to PVDF membranes (Trans-Blot® Turbo™ Transfer System; Bio-Rad). WB was performed using the following antibodies (Abs) in accordance with the manufacturers’ instructions: biotinylated polyclonal goat anti-human IL-1α Ab (BAF200; R&D Systems) and polyclonal rabbit anti-heat shock protein 90 Ab (HSP90; Cell Signaling Technology; Danvers, MA, USA). ExtrAvidin-Peroxidase conjugate (Sigma-Aldrich) was used for detection of IL-1α protein, whereas peroxidase-linked anti-rabbit IgG Ab (Cell Signaling Technology) was used for detection of HSP90 protein.

The caspase 3/7 activities in HCAECs were evaluated using the Caspase-Glo® 3/7 Assay System (Promega; Madison, WI, USA) in accordance with the manufacturer’s instructions. The data were collected using a multimode microplate reader (ARVO™, Perkin Elmer; Waltham, MA, USA).

HCAECs were stimulated with 100 ng/ml of TNF-α, or 10 ng/ml of IL-1α or IL-1β for 24 h in the presence and absence of 10 mg/ml IgG and 1000 nM DEX, alone or in combination. The concentrations of cytokines and DEX were determined based on the results of preliminary experiments (Additional files 1 and 2: Figs. S1 and S2). Damage to the HCAECs due to the inflammatory stimuli was evaluated by their release of HMGB1 protein. It is of note that HMGB1 protein release by the HCAECs was not suppressed at all by IgG treatment, whereas it was significantly suppressed by DEX treatment, for each of the tested cytokine stimuli (Fig. 1a). The HMGB1 mRNA expression level in the HCAECs showed no significant changes under any conditions (Fig. 1b). Furthermore, the cytokine-enhanced caspase 3/7 activities in the HCAECs were inhibited by DEX treatment, but not by IgG treatment (Fig. 1c).

Both IL-1α and IL-1β are cytokines belonging to the IL-1 family, bind to the same receptors and induce the same biological functions, but their production mechanisms by cells are different [19]. Briefly, secretion of activated IL-1β requires Nod-like receptor (NLR) family, pyrin domain-containing 3 (NALP3)/caspase 1-dependent inflammasome activation [20], whereas IL-1α is a DAMP, like HMGB1 [19]. We next examined the effects of IgG and DEX on cytokine-induced release of IL-1α. Both TNF-α and IL-1β stimulation induced significant release of IL-1α into the HCAEC culture supernatants, but DEX treatment effectively inhibited that release (Fig. 2a). DEX treatment also significantly inhibited both mRNA (Fig. 2b) and intracellular protein expression of IL-1α (Fig. 2c). Curiously, IL-1α protein could hardly be detected in the IgG-treated HCAEC culture supernatants (Fig. 2a), but IgG treatment was less effective than DEX in inhibiting both mRNA (Fig. 2b) and intracellular protein expression (Fig. 2c) of IL-1α induced by inflammatory stimuli. We posited that high-dose IgG might interfere with detection of IL-1α protein in this ELISA system. To confirm that, we added recombinant IL-1α protein equivalent to the maximum IL-1α level released by HCAECs in this study (approximately 100 pg/ml; see Additional file 1: Fig. S1, middle right graph) to culture medium samples, to which IgG was then added at concentrations of 0 to 10 mg/ml. As posited, ELISA examination of those samples found that the detectability of IL-1α protein decreased with increasing IgG concentration, and it could not be detected at all at 10 mg/ml of IgG (Additional files 3: Figs. S3). These findings are in agreement with an earlier finding that anti-IL-1α antibody is contained in human immunoglobulin preparations [21].

We next examined the effects of high-dose IgG and DEX on cytokine-induced expression of IL-6 and G-CSF, which are crucially involved in the pathogenesis of KD [16, 17]. We previously demonstrated that both IL-6 and G-CSF mRNA expressions were decreased the most by IVIG treatment among all TNF-α-inducible genes in HCAECs [15]. Consistent with that report, both IgG and DEX effectively inhibited TNF-α-induced production of IL-6 and G-CSF by HCAECs (Fig. 3a, upper graphs). On the other hand, similar to our previous finding that IL-1β stimulation led to IVIG-resistant production of IL-6 and G-CSF by HCAECs [15], we newly found that IL-1α induced exactly the same IVIG-resistant production of those cytokines (Fig. 3a, lower graphs). DEX treatment partially inhibited IL-1 s-induced IL-6 production by HCAECs (Fig. 3a, lower left graph), but its effect was much weaker than that of TNF-α (Fig. 3a, upper left graph). In addition, G-CSF production induced by IL-1 s was hardly inhibited by DEX treatment, similar to the case of IVIG treatment (Fig. 3a, lower right graph). The mRNA expression patterns for both IL-6 and G-CSF (Fig. 3b) were in line with their protein production patterns (Fig. 3a).

We next investigated the inhibitory effects of high-dose IgG and DEX at various time points after cytokine stimulation. HCAECs were stimulated with TNF-α or IL-1β alone, followed by treatment with DEX and/or IgG at 0, 12 and 24 h later (Fig. 4). The cell supernatants were collected at 72 h after the cytokine stimulation. The effects of DEX and IgG on TNF-α-induced production of IL-6 and G-CSF (anti-inflammatory effects) were strongest when the drugs were administered at the start of stimulation (0 h). In addition, the combination of DEX and IgG showed the strongest anti-inflammatory effects, even though they were added later (Fig. 4, lower graphs vs. middle and upper graphs). DEX alone also effectively inhibited IL-1α release induced by TNF-α or IL-1β stimulation, but its effects were stronger when it was added earlier (Fig. 4, upper right graph).