Interleukin 4 is inactivated via selective disulfide-bond reduction by extracellular thioredoxin

Extracellular TRX Inhibition Enhances Cytokine Outputs of M2 Macrophages Derived from THP-1 Cells.

Because we noticed that macrophage morphology was sensitive to TRX inactivation, we performed a cytokine screen to characterize the effect of TRX on macrophages. We first evaluated an established cellular model involving the THP-1 human monocytic cell line (16). Specifically, THP-1 cells were differentiated into macrophages by exposure to phorbol 12-myristate 13-acetate (PMA); cells thus treated are commonly referred to as unpolarized macrophages in the “M0 state” (16). M0 macrophages can then be polarized into M1 macrophages with IFN-γ and LPS or into M2 macrophages with IL-4.

To test the effect of endogenous extracellular TRX on macrophage polarization, we added NP161, a small-molecule inhibitor of extracellular TRX, to cultures of either M1 or M2 macrophages. For this exploratory study, we looked for the most profound changes in the concentrations of secreted cytokines. Among the 62 cytokines screened, changes of at least fivefold were identified for 11 cytokines (Fig. 2). Notably, addition of NP161 reduces secretion of these cytokines in M1 macrophages, whereas it increases cytokine levels in M2 macrophages.

The above observations led us to suspect a role for extracellular TRX in suppressing the M2 state of macrophages. Because IL-13 can elicit the same immune response as IL-4, our observed effect could also be mediated through IL-13. Therefore, we hypothesized that extracellular TRX influenced this differential polarization of macrophages by directly inactivating either IL-4 or IL-13, two cytokines whose activities are known to require the presence of disulfide bonds (24, 25). This hypothesis was directly tested through biochemical studies, as described below.

TRX Preferentially Reduces IL-4 over IL-13.

The cytokines IL-4 and IL-13 are homologous helical proteins that signal through a shared receptor, IL-4Rα (Fig. 3). Both cytokines harbor three disulfide bonds. To directly test the hypothesis that TRX regulates IL-4 and/or IL-13 function by reduction of an allosteric disulfide bond, we first needed to produce sufficient quantities of both recombinant proteins. Genes encoding the mature human IL-4 and IL-13 were expressed in Escherichia coli, and the proteins were isolated as inclusion bodies. As detailed in Materials and Methods, each cytokine was refolded, purified, and demonstrated to have comparable biological activity to authentic standards in a TF-1 cell-proliferation assay. The ED₅₀ values of IL-4 and IL-13 in this proliferation assay were 1.7 and 0.5 µg/mL, respectively (SI Appendix, Fig. S1).

To test whether recombinant human TRX was able to recognize and react with recombinant human IL-4 or IL-13, we took advantage of the active-site C35S mutant (Fig. 1B) that has been previously used to trap mixed disulfide adducts between TRX and its substrates (6, 15). This mutant protein (13 kDa) was purified and incubated with IL-4 (17 kDa) or IL-13 (15 kDa), and the protein mixtures were analyzed via nonreducing SDS/PAGE. A prominent ∼30-kDa adduct was observed when mutant TRX was incubated with IL-4; under identical conditions, only a small amount of the putative adduct was observed between the C35S TRX mutant and IL-13 (Fig. 4A).

To quantify the specificity of TRX for IL-4 versus IL-13, we adopted an established kinetic assay for TRX activity, using insulin as a reference substrate (11). Steady-state kinetic analysis revealed that TRX had significantly higher specificity toward IL-4 than IL-13 (Fig. 4B and Table 1). Kinetic parameters for insulin recognition by TRX were in agreement with previously reported data (11, 26). Notably, to our knowledge, IL-4 appears to be the most preferred extracellular substrate of TRX identified to date.

To verify the preference of TRX for IL-4 over IL-13, the TRX-promoted rate of IL-4 deactivation was compared in the presence or absence of an initially equal concentration of IL-13. To simulate physiological conditions, substantially lower cytokine concentrations were employed in these assays than those used for the estimation of kinetic parameters. As predicted, addition of IL-13 had negligible effect on the rate of oxidative deactivation of IL-4 (Fig. 4C). By way of confirmation, the extent of TRX reduction of the two cytokines was directly quantified at 5 min (Fig. 4D), and was found to be in line with the specificity of TRX for IL-4 over IL-13, as measured above.

TRX Reduces the C46–C99 Disulfide Bond in IL-4.

IL-4 contains three disulfide bonds. The high specificity of TRX for IL-4 and the observation of a single adduct between C35S TRX and IL-4 suggested that a unique disulfide of IL-4 was targeted by TRX. Mass spectrometric analysis was therefore used to identify this recognition site (SI Appendix, Scheme S1). As summarized in Table 2, the disulfide bond between C46 and C99 of IL-4 was exclusively reduced by TRX. Given that this disulfide bond is essential for cytokine function (24), our findings suggest that TRX recognition of IL-4 has the potential to be a biologically relevant regulatory mechanism.

TRX Selectively Inactivates the Cytokine Activity of IL-4.

A cellular model was used to assess the biological relevance of the observed protein–protein recognition between IL-4 and TRX. Proliferation of the TF-1 erythroleukemia cell line was evaluated in the presence of TRX. Growth of this cell line requires IL-4, IL-13, or GM-CSF in the culture medium (27). The IC₅₀ of TRX was 50 nM in cultures containing IL-4 whereas, under equivalent conditions, the IC₅₀ of TRX was 2.2 µM in cultures containing IL-13. TRX had no effect on TF-1 cell growth in the presence of GM-CSF (Fig. 5).

TRX Mitigates IL-4–Driven Pathological Conditions in Chronic Pancreatitis.

In light of the above findings in vitro, we sought to assess the pathophysiological relevance of TRX-mediated IL-4 inactivation in vivo. To do so, we took advantage of a recent study highlighting the role of IL-4 signaling in a mouse model of chronic pancreatitis (28).

Chronic pancreatitis is characterized by progressive irreversible damage to the pancreas (29). Some of the key histological features include inflammation, fibrosis, and acinar cell death (30), which in part are promoted by M2 macrophages. Because pharmacological inhibition of the IL-4 receptor decreases these pathological phenotypes and halts chronic pancreatitis progression (28), we used these pathological features as evidence for IL-4 inactivation in vivo. To that end, we first induced chronic pancreatitis in C57BL/6J mice by repetitive injection with cerulein over 4 wk (six injections per d, 3 d/wk). At the beginning of week 3, we initiated dosing of recombinant TRX to diseased mice. Compared with the control group, TRX-treated mice showed a marked reduction in the amount of active, oxidized IL-4 in pancreatic tissue (Fig. 6B). Moreover, TRX treatment limited pancreatic fibrosis, as shown by increased pancreas weight, histology, and decreased fibrosis-associated gene expression (Fig. 6 A and C–H), providing further support for the ability of TRX to inactivate IL-4 in vivo.

Chemicals and Other Reagents.

Unless otherwise specified, reagents were from Sigma-Aldrich. DTT was from Invitrogen, SDS/polyacrylamide gradient gels (4 to 20%) were from Bio-Rad, Ni-NTA resin was from Qiagen, the HiTrap Q anion-exchange column was from GE Healthcare, and 7-kDa molecular mass cutoff spin columns were from Pierce. Cell-culture medium, FBS, antibiotics, and sterile PBS were from Invitrogen. Glutamine was from Lonza.

Macrophage Polarization.

The human THP-1 monocytic cell line was maintained in RPMI 1640 culture medium containing 10% heat-inactivated FBS and penicillin/streptomycin. Cells were seeded at 10⁶ cells per mL and differentiated into M0 macrophages by incubation for 48 h in the presence of 100 ng/mL phorbol 12-myristate 13-acetate (P8139; Sigma). These M0 cells were then maintained in the same state for an additional 72 h in RPMI medium. THP-1–derived M0 macrophages were polarized into M1 macrophages by incubation for 36 h with 20 ng/mL IFN-γ (285-IF; R&D Systems) and 1 ng/mL LPS (tlrl-pb5lps; InvivoGen). Alternatively, macrophages were polarized into the M2 state by incubation for 36 h with 20 ng/mL interleukin 4 (I4269; Sigma) or 20 ng/mL interleukin 13 (50-813-223; Fisher). When needed, 33 µM NP161 (a small-molecule inhibitor of TRX) was added.

Luminex Assay.

This multiplexed assay for secreted proteins was performed at the Human Immune Monitoring Center at Stanford University. Human 62-plex was from Affymetrix, and used according to the manufacturer’s recommendations with modifications, as detailed elsewhere (41). Briefly, antibody-linked beads were added to a 96-well plate and washed in a BioTek ELx405 washer. Culture supernatants were added to these wells and incubated at room temperature for 1 h followed by overnight incubation at 4 °C with shaking (500 to 600 rpm on an orbital shaker). Plates were washed in a BioTek ELx405 washer, and the biotinylated detection antibody was added for 75 min at room temperature with shaking. Plates were washed again as above, and streptavidin-phycoerythrin was added. After incubation for 30 min at room temperature, plates were washed once more and reading buffer was added to the wells. Plates were read on a Luminex 200 instrument with a lower bound of 50 beads per sample per cytokine. Each sample was measured in duplicate. Control beads (Radix BioSolutions) were added to all wells.

Cross-Linking of C35S TRX and IL-4 or IL-13.

IL-4 (10 µM) or IL-13 (10 µM) was incubated for 30 min at room temperature with C35S TRX (10 µM) in 20 mM Tris⋅HCl, 1 mM EDTA (pH 7.6). Samples were diluted with 2× Laemmli sample buffer (Bio-Rad) and applied to a nonreducing 4 to 20% SDS/polyacrylamide gel (Bio-Rad).

Thioredoxin Activity Assay.

Before use, recombinant human TRX was freshly reduced with a 10-fold molar excess of DTT on ice. The excess DTT was removed by passing the solution through a 7-kDa molecular mass cutoff spin column. TRX concentration was determined by A₂₈₀ (ε = 7,570 M⁻¹⋅cm⁻¹), and the protein was freshly used within 2 h. Steady-state kinetic analysis of TRX-mediated reduction of insulin and IL-4 was performed via a coupled assay containing 6 µM TrxR, 10 nM TRX, and 0.3 mM NADPH in a buffer containing 50 mM Tris⋅HCl and 2 mM EDTA (pH 7.5). The reaction rate was calculated from the slope of the absorbance curve at 340 nm, using the extinction coefficient of NADPH (6,220 M⁻¹⋅cm⁻¹). Michaelis–Menten parameters were determined by fitting the kinetic data using GraphPad Prism 6.

IL-4 and IL-13 Competition Assay.

Before use, recombinant human TRX was freshly reduced with a 10-fold molar excess of DTT on ice. The excess DTT was removed by passing the solution through a 7-kDa molecular mass cutoff spin column. TRX concentration was determined by A₂₈₀ (ε = 7,570 M⁻¹⋅cm⁻¹), and the protein was freshly used within 2 h. Human recombinant IL-4 (200-04) and recombinant IL-13 (200-13) were from PeproTech. Kinetic analysis of TRX-mediated reduction of IL-4 was performed via a coupled assay containing 1 µM TrxR, 5 nM TRX, 60 µM NADPH, 50 nM IL-4, and 50 nM IL-13 in PBS buffer (pH 7.4). The rate of oxidative inactivation of IL-4 was compared with that in the presence of equimolar IL-13 as a competitive substrate. At specific time points, aliquots were withdrawn and diluted 5,000-fold into cold PBS (4 °C). Cytokine concentrations were determined via ELISAs.

Mass Spectrometric Determination of the Target Disulfide Bond in IL-4.

A 60-µL solution containing IL-4 (34 µg, 20 µM) and C35S TRX (26 µg, 20 µM) was incubated for 30 min at room temperature in 20 mM Tris⋅HCl, 1 mM EDTA (pH 7.6). The solution was then diluted with 7.5 µL of 8 M urea in 100 mM NH₄HCO₃ and incubated with iodoacetic acid (320 µg, 25 mM) for 1 h at room temperature in the dark. The reaction was buffer-exchanged three times on a Zeba desalting column (Thermo Scientific) into 20 mM Tris, 1 mM EDTA, 1 M urea, 12.5 mM NH₄HCO₃ and then incubated with DTT (93 µg, 10 mM) for 30 min at room temperature. Iodoacetamide (320 µg, 25 mM) was added to the samples and allowed to incubate at room temperature in the dark for 30 min. Reconstituted trypsin solution (20 mg/mL in resuspension buffer; Promega) was added to a final concentration of 6 µg/mL. The samples were digested for 4 h in a 37 °C water bath, after which the digestion was quenched by adding formic acid to a final concentration of 7.5% (vol/vol). Peptides were desalted using C18 StageTips (42), lyophilized overnight, resuspended in 5% formic acid in water, and analyzed by mass spectrometry on an Orbitrap Elite Mass Spectrometer (Thermo Scientific) in data-dependent acquisition mode, where the top 10 peaks per acquisition cycle were selected for collision-induced fragmentation. Peptides were identified by searching the spectra against the human proteome using the SEQUEST algorithm (43), with a false discovery rate cutoff of 1% at the peptide level. The relative amounts of peptides containing Cys3, Cys24, Cys46, Cys65, and Cys99 were determined by integration of their precursor (MS1) peak intensities.

Thioredoxin-Mediated Inhibition Assay.

TF-1 cells (American Type Culture Collection) were cultured in RPMI medium 1640 (Gibco) supplemented with 5% (vol/vol) FBS, 10 mM Hepes, 1 mM sodium pyruvate, and penicillin/streptomycin. TF-1 cells were washed with PBS and seeded at 3 × 10⁵ cells per mL in 96-well plates (0.2 mL per well) in medium supplemented with 8 ng/mL recombinant IL-4, IL-13, or GM-CSF. Cells were incubated with TRX concentrations ranging from 0 nM to 10 µM at 37 °C and 5% CO₂ for 48 h. Cells were counted on a BD Accuri C6 Cytometer (BD Biosciences). Data were collected in triplicate and fit to a log(inhibitor) vs. response (three-parameter) equation using GraphPad Prism 6.

Pancreatitis Model and Treatment.

Chronic pancreatitis was induced by repetitive cerulein injections (44). In brief, mice were given six hourly i.p. injections of 50 μg⋅kg⁻¹ body weight of cerulein (Sigma-Aldrich) 3 d/wk for a total of 4 wk. Mice were then killed 3 d after the last cerulein injection, and pancreatic tissues were analyzed. For the TRX therapy, all mice were given cerulein injections 3 d/wk for a total of 4 wk as above, and 3 wk following start of the cerulein injections, mice were either given vehicle control (PBS) or TRX (i.p., 250 mg/kg, two times per d, 3 d/wk for 1 wk) until being killed at the fourth week. The Stanford Institutional Animal Care and Use Committee (IACUC) approved all animal studies, and animals were housed in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-accredited facility.

Quantitative RT-PCR.

Total RNA was isolated from pancreatic tissue using TRIzol reagent (Invitrogen) according to the manufacturer’s instructions. In brief, cDNA was generated using the GoScript Reverse-Transcription System (Promega). Quantitative PCR was performed with an ABI 7900 Sequence Detection System (Applied Biosystems) using designed specific TaqMan probes and primers as follows: αSMA (forward, 5′-CTCCCTGGAGAAGAGCTACG-3′; reverse, 5′-TGACTCCATCCCAATGAAAG-3′; probe, 5′-AAACGAACGCTTCCGCTGCC-3′); collagen 1A1 (forward, 5′-AGAAGGCCAGTCTGGAGAAA-3′; reverse, 5′-GAGCCCTTGAGACCTCTGAC-3′; probe, 5′-TGCCCTGGGTCCTCCTGGTC-3′); fibronectin (forward, 5′-TGGTGGCCACTAAATACGAA-3′; reverse, 5′-GGAGGGCTAACATTCTCCAG-3′; probe, 5′-CAAGCAGACCAGCCCAGGGA-3′); TGF-β (forward, 5′-CCCTATATTTGGAGCCTGGA-3′; reverse, 5′-CTTGCGACCCACGTAGTAGA-3′; probe, 5′-CCGCAGGCTTTGGAGCCACT-3′); and GAPDH (forward, 5′-TGTGTCCGTCGTGGATCTGA-3′; reverse, 5′-CCTGCTTCACCACCTTCTTGA-3′; probe, 5′-CCGCCTGGAGAAACCTGCCAAGTATG-3′). Samples were normalized to GAPDH and displayed as fold induction over untreated controls, unless otherwise stated.

Statistical Analysis.

Unpaired Student’s t test was used to determine statistical significance between two groups. Values are expressed as mean ± SEM (Prism 7; GraphPad Software).