SOCS3 drives proteasomal degradation of indoleamine 2,3-dioxygenase (IDO) and antagonizes IDO-dependent tolerogenesis

Abstract

Despite their common ability to activate intracellular signaling through CD80/CD86 molecules, cytotoxic T lymphocyte antigen 4 (CTLA-4)-Ig and CD28-Ig bias the downstream response in opposite directions, the latter promoting immunity, and CTLA-4-Ig tolerance, in dendritic cells (DCs) with opposite but flexible programs of antigen presentation. Nevertheless, in the absence of suppressor of cytokine signaling 3 (SOCS3), CD28-Ig-and the associated, dominant IL-6 response-become immunosuppressive and mimic the effect of CTLA-4-Ig, including a high functional expression of the tolerogenic enzyme indoleamine 2,3-dioxygenase (IDO). Here we show that forced SOCS3 expression antagonized CTLA-4-Ig activity in a proteasome-dependent fashion. Unrecognized by previous studies, IDO appeared to possess two tyrosine residues within two distinct putative immunoreceptor tyrosine-based inhibitory motifs, VPY(115)CEL and LLY(253)EGV. We found that SOCS3-known to interact with phosphotyrosine-containing peptides and be selectively induced by CD28-Ig/IL-6-would bind IDO and target the IDO/SOCS3 complex for ubiquitination and subsequent proteasomal degradation. This event accounted for the ability of CD28-Ig and IL-6 to convert otherwise tolerogenic, IDO-competent DCs into immunogenic cells. Thus onset of immunity in response to antigen within an early inflammatory context requires that IDO be degraded in tolerogenic DCs. In addition to identifying SOCS3 as a candidate signature for mouse DC subsets programmed to direct immunity, this study demonstrates that IDO undergoes regulatory proteolysis in response to immunogenic stimuli.

Fig. 1.

SOCS3 expression regulates the default functional program of CD8⁺ and CD8⁻ DCs. (A) Real-time PCR analysis of Socs3 mRNA expression. Purified CD8⁺ and CD8⁻ DCs were cultured for different times in the absence of external stimuli, and Socs3 mRNA levels were quantified by real-time PCR using Gapdh normalization. Data are presented as normalized specific gene transcript expression in the samples relative to normalized transcript expression in the respective control cultures—that is, freshly harvested CD8⁺ or CD8⁻ DCs (fold change = 1; dotted line). Data are means ± SD from four experiments. (Inset) Socs3 expression was evaluated in freshly harvested DC subsets (indicated) by PCR, using Gapdh expression as a control. Overexpression of Socs3 subverts the basal tolerogenic phenotype of CD8⁺ DCs (B) and increases the immunogenic potential of CD8⁻ DCs (C). Splenic DCs were fractionated according to CD8 expression, pulsed with the P815AB peptide, and transferred into recipient mice to be assayed at 2 wk for skin test reactivity to the eliciting peptide. CD8⁺ cells were used either alone (B) or as a minority fraction (3%) in combination with CD8⁻ DCs (C). Both subsets were injected either as such or after transfection with control or Socs3 mRNA. The asterisk (P < 0.01−0.001; experimental vs. control footpads) indicates the occurrence of a positive skin test reaction as a result of unopposed immunogenic presentation of the peptide by the DCs. Data are mean values ± SD of three experiments.

Fig. 2.

SOCS3 expression regulates the acquisition of an immunogenic vs. tolerogenic function in DC subsets in response to environmental stimuli. P815AB-pulsed CD8⁻ DCs (majority population) and CD8⁺ DCs were transferred into recipient mice to be assayed for skin test reactivity at 2 wk. The IDO inhibitor 1-MT was added to selected cultures at the final concentration of 4 μM. When used in combination with CD8⁻ DCs, the minority CD8⁺ DC fraction was used as such or after treatment with CD28-Ig (A) or rIL-6 (B), with or without concomitant Socs3 gene silencing by siRNA. Untreated cells and/or cells transfected with control siRNA were also used. In C, minority fractions of peptide-pulsed CD8⁺ or CD8⁻ DC subsets (indicated) were injected either as such or after transfection with control or Socs3 mRNA and subsequent conditioning by CTLA-4-Ig. In both A and C, Ig-Cγ3 was used as a control for both fusion proteins. *, P < 0.005, experimental vs. control footpads (n = 3).

Fig. 3.

The proteasome inhibitor MG132 confers IDO-dependent, immunosuppressive properties on CD28-Ig in CD8⁺ DCs. CD8⁺ DCs were conditioned by overnight incubation with CD28-Ig or IL-6. Ig-Cγ3 was used as a stimulation control for CD28-Ig. The proteasome inhibitor, MG132, was added at 10 μM for 1 h before addition of the stimuli. The IDO inhibitor, 1-MT, was added to selective cultures at 4 μM. (A) Conditioned CD8⁺ DCs were pulsed with the P815AB peptide and injected, in combination with a majority fraction of CD8⁻ DCs, into recipients hosts that were assayed for the development of P815AB-specific skin test reactivity at 2 wk after cell transfer. *, P < 0.005, experimental vs. control footpads. (B) IDO activity was evaluated in terms of kynurenine production in culture supernatants from CD8⁺ DCs. In both A and B, results are mean values ± SD from three experiments.

Fig. 4.

The proteasome inhibitor MG132 antagonizes SOCS3-dependent immunostimulatory effects in CD8⁺ DCs used as a minority population in combination with CD8⁻ DCs. CD8⁺ DCs, used either as such or after transfection with control or Socs3 mRNA, were conditioned by overnight incubation with CTLA-4-Ig. Ig-Cγ3 was used as control. MG132 and 1-MT were added to selective cultures as in Fig. 3. (A) The development of P815AB-specific skin test reactivity was assessed at 2 wk after cell transfer, as in Fig. 3. Data are means (± SD) from three experiments. *, P < 0.001, experimental vs. control footpads. (B) IDO activity was evaluated in terms of kynurenine production in supernatants from cultured CD8⁺ DCs. Results are mean values (± SD) of three experiments.

Fig. 5.

SOCS3 accelerates IDO turnover in DCs by means of ubiquitination and proteasomal degradation. (A) Unfractionated DCs were transfected with IDO-Flag mRNA and treated with CD28-Ig in the presence or absence of MG132. Cells were lysed and IDO-Flag was immunoprecipitated (IP) with anti-Flag. Sequential immunoblotting was conducted using anti-phosphotyrosine (PY), anti-ubiquitin, and anti-IDO. One-tenth aliquots of whole cell lysates (WCL) from parallel samples were blotted with SOCS3- and β-tubulin-specific antibodies. H, heavy chain of the anti-Flag antibody (55 kDa). One experiment is shown representative of several. (B) Unfractionated DCs were transfected with control or Socs3 mRNA and IDO protein expression was monitored over time (indicated) in WCL by means of Western blot using an IDO-specific monoclonal antibody. SOCS3 expression was also assayed. One experiment is shown representative of three.

Fig. 6.

IDO contains ITIM sequences that are necessary for SOCS3-mediated ubiquitination and proteasomal degradation. (A) Lysates from P1 cells stably transfected with Socs3 (P1.SOCS3) and CD28-Ig-treated DCs (DCs) were pulled down with unphosphorylated (ITIM1 and ITIM2) or phosphorylated (pITIM1 and pITIM2) IDO peptides and immunoblotted with anti-SOCS3 antibodies, which were also used in parallel Western blot analyses of WCL. (B) Lysates from P1 cells, transfected with Flag-tagged wild-type IDO (wtIDO) or the mutant IDOY₁₁₅F/Y₂₅₃F, were immunoprecipitated with anti-Flag and then sequentially blotted with anti-SOCS3 antibodies, which were also used in parallel Western blot analyses of WCL. Empty vector (EV) was used as a control. β-tubulin expression was evaluated as a loading control. H, heavy chain of the anti-Flag antibody. One experiment is representative of three. (C) Unfractionated DCs were transfected with wtIDO-Flag or IDOY₁₁₅F/Y₂₅₃F-Flag mRNA and treated with CD28-Ig in the presence of MG132. Cells were lysed and wtIDO-Flag and IDOY₁₁₅F/Y₂₅₃F-Flag proteins were immunoprecipitated (IP) with anti-Flag. Sequential immunoblotting was conducted using anti-IDO and anti-ubiquitin. Arrowhead indicates the major ubiquitinated form of IDO. One-tenth aliquots of WCL from parallel samples were blotted with β-tubulin-specific antibodies. One experiment is shown representative of two.

Fig. 7.

Reverse signaling is instrumental in T cell conditioning of antigen-presenting DCs to fully meet the needs of flexibility and redundancy and tip the balance in favor of immunity or tolerance. On engagement of intracellularly signaling CD80/CD86 molecules by CD28 or CTLA-4, SOCS3 could be a major discriminator of function, by affecting IDO lifespan in the DC and thus sustaining or subverting the basic functional program of the DC. Preponderant conditioning by CD28/IL-6 would uniformly ensure immunogenic presentation, whereas the action of CTLA-4/IFN-γ would help establish and spread tolerance.