CD1d ligation on directly signals rapid NF-␬B activation and production of bioactive IL-12

Simon C. Yue, Angela Shaulov, Ruojie Wang, Steven P. Balk, and Mark A. Exley*

Cancer Biology Program, Division of Hematology and Oncology, Beth Israel Deaconess Medical Center, Harvard Medical School, 330 Brookline Avenue, Boston, MA 02215

Edited by Peter Cresswell, Yale University School of Medicine, New Haven, CT, and approved June 14, 2005 (received for review April 22, 2005) Natural killer T cells (NKT cells) expressing a semiinvariant CD1d- operative interactions between iNKT and DCs generate a pos- reactive receptor (invariant NKT, iNKT) can be rapidly acti- itive feedback loop that can rapidly amplify IL-12-dependent vated by monocytes or immature dendritic cells (iDCs) bearing a immune responses. However, it remains unclear whether these CD1d-presented glycolipid and can in turn stimulate these iNKT–DC interactions are initiated by CD1d-presented endog- myeloid cells to mature and produce IL-12. Previous studies have enous or exogenous or by other mechanisms (19). The shown that iNKT-produced IFN␥ and CD40 ligand contribute to this objective of this study was to identify molecular interactions maturation. This study demonstrates that CD1d mediating activation of human monocytes and DCs by iNKT. ligation alone, in the absence of iNKT, could rapidly (within 24 h) stimulate production of bioactive IL-12p70 by CD1d؉ human pe- Materials and Methods ripheral blood monocytes as well as iDCs. IFN␥ alone had no effect, Cell Culture. iNKT lines (Ϸ90% pure) were obtained by purifi- but it markedly enhanced CD1d-stimulated IL-12 production. cation from healthy donor leukapak peripheral blood mononu- differentiation, as assessed by CD40 and CD1a up- clear cells by using a biotinylated mAb against the invariant TCR regulation, was also accelerated by CD1d stimulation, consistent (6B11) (21) and magnetic bead sorting with anti-biotin IgG with this representing a physiological response. CD1d ligation on beads (Miltenyi Biotec, Auburn, CA). The purified iNKT were the human monocytic cell line THP-1 similarly specifically stimu- expanded in vitro by culturing on anti-CD3-coated plates. lated IL-12 production. Biochemical studies showed that IL-12 Briefly, 0.5–1 ϫ 105 purified iNKT were mixed with an equal release correlated with rapid phosphorylation of I␬B, a critical step number of irradiated (4,000 rads) autologous peripheral blood in NF-␬B activation. Selective NF-␬B inhibition blocked this CD1d- mononuclear cells in RPMI medium 1640 with 10% FBS stimulated IL-12 production. Finally, CD1d ligation could also en- (HyClone) and 100 units͞ml IL-2 (National Biological Response hance IL-12 production in the presence of suboptimal LPS or CD40 Modifier Program, National Cancer Institute, Frederick, MD) stimulation. These findings demonstrate an innate immune signal- (21). Each 0.2 ml of cells was added to a well of a flat-bottomed ing function for CD1d and provide a mechanism for the rapid 96-well plate precoated with OKT3 mAb (American Type activation of monocytes and iDCs by CD1d-reactive T cells. Culture Collection). The cells were rested in medium with 10 units͞ml IL-2 overnight before use. antigen-presenting cell ͉ innate immunity ͉ Human monocytes were obtained from buffy coat-derived and Ficoll-Hypaque-purified peripheral blood mononuclear cells by D1d is a nonpolymorphic MHC class I-like consti- adhesion on tissue culture flasks in RPMI medium 1640 with tutively expressed by antigen-presenting cells (APC) and by 10% FBS. After extensive washing, adherent cells were released C by using PBS with 1 mM EDTA for use directly as monocytes. some epithelia, whereas CD1a–c are induced upon dendritic cell ϩ (DC) maturation. CD1d is recognized by a subpopulation of T CD14 cells represented the great majority. To generate imma- ture DCs (iDCs), adherent monocytes were cultured for 3 cells, many of which express markers typical of natural killer cells ͞ and have been termed natural killer T cells (NKT cells) (1–3). A further days with GM-CSF (100 ng ml, Immunex; or 1,000 units͞ml, Berlex Biosciences, Richmond, CA) and IL-4 (25 major fraction of CD1d-restricted T cells recognize CD1d ͞ ͞ through an invariant T cell receptor (TCR)␣ chain (V␣24-J␣18 ng ml, Endogen, Rockford, IL; or 110 units ml, Cell Sciences, in ). CD1d-reactive NKT cells using this invariant TCR Canton, MA) in RPMI medium 1640 with 10% FBS. THP-1 cells were cultured in RPMI medium 1640 with 10% FBS (HyClone, (iNKT) are largely primed in vivo for the rapid production of T ͞ ␮ helper 1 and T helper 2 (IFN␥ and IL-4) and have Logan, UT) 50 M 2-mercaptoethanol (Sigma). regulatory functions in innate and adaptive immune responses ␣ Cell Stimulations and Measurement. Monocytes or iDCs (3, 4). A glycolipid antigen isolated from marine sponge ( - ϫ 5 galactosylceramide, ␣galcer) is recognized by iNKT in the (1 10 ) were cocultured in 96-well flat-bottomed plates with ␣ iNKT. Triplicate wells were supplemented with ␣galcer (100 context of CD1d (5–8). In vivo, galcer causes activation of ͞ ͞ iNKT, rapid cytokine production, and subsequent systemic ng ml, Kirin Brewery, Gunma, Japan), LPS (25–250 ng ml for monocytes, up to 20 ␮g͞ml for iDCs, Sigma-Aldrich), or phy- activation of innate and adaptive immune cells (9, 10). tohemagglutinin (5 ␮g͞ml, Invitrogen), and supernatants were Data from many groups indicate that iNKT functions can be ϩ analyzed by ELISA. For stimulations by plate-bound Abs, 96- mediated through interactions with CD1d APC including DCs well flat-bottomed plates were coated with protein G for 4 h (20 and B cells (11–20). Murine studies have shown that iNKT activation can be enhanced by B7 ligation of CD28 and that IL-12 ␥ produced by DCs can further enhance iNKT IFN production This paper was submitted directly (Track II) to the PNAS office. (11, 12, 20). Conversely, murine and human studies have shown ␣ Abbreviations: NKT cell, natural killer T cell; iNKT, invariant NKT cell; ␣galcer, ␣-galacto- that iNKT can stimulate the maturation of galcer-pulsed DCs sylceramide; DC, dendritic cell; iDC, immature DC; LLM, N-acetyl-L-leucinyl-L-leucinyl- (11, 12, 18, 20). This DC stimulation appears to be mediated by methional; APC, antigen-presenting cell; TCR, T cell receptor; pI␬B, phospho-I␬B. CD40 ligand on the iNKT through ligation of CD40 on the DCs, *To whom correspondence should be addressed at: Harvard Institute of Medicine,

with IFN␥ produced by the iNKT subsequently enhancing DC 330 Brookline Avenue, Boston, MA 02215. E-mail: [email protected]. IMMUNOLOGY IL-12 production. Overall, these observations indicate that co- © 2005 by The National Academy of Sciences of the USA

www.pnas.org͞cgi͞doi͞10.1073͞pnas.0503366102 PNAS ͉ August 16, 2005 ͉ vol. 102 ͉ no. 33 ͉ 11811–11816 Downloaded by guest on September 23, 2021 Fig. 1. Reciprocal ␣galcer-dependent iNKT and monocyte͞iDC activation. (A) CD1d (thick lines) and isotype mAb control (thin lines) staining of monocytes (Left) and iDCs (Right). Viable cells were gated on CD3- and CD19-negative cells. Modest CD1d expression by both cell types is shown, with iDC expressing consistently higher levels. (B and C) IFN␥ and IL-4 release by iNKT cultured with monocytes (Left) or iDCs (Right) in the absence or presence of the indicated stimuli. Results shown are for day 1. (D) iNKT were cocultured with monocytes or iDCs with the indicated stimuli. Culture supernatants were assayed for IL-12p70. Results shown are for day 3. (E) iNKT were cocultured with monocytes in the presence of neutralizing anti-IFN␥ or isotype mAb or with exogenous IFN␥ as shown. Day 3 cultures were assayed for IL-12p70.

␮g͞ml in PBS, Sigma), washed, and blocked with 2% BSA in ferred to nitrocellulose. CD1d was detected with C3D5, a mouse PBS. Wells were then washed and incubated for 4 h with 10 mAb generated against a CD1d-GST fusion protein (23). Phos- ␮g͞ml CD1d mAb 51.1 (IgG2b) or isotype control, except where pho-I␬B (pI␬B) was detected with a pI␬B-specific mAb (Cell indicated. Additional CD1d mAbs were 27.1 and 42.1 (both Signaling Technology, Beverly, MA). Secondary anti-mouse IgG1), 75.1 (IgG2a), or 59.1 (IgM) (22–24), or anti-CD40 (BD IgG-horseradish peroxidase and chemiluminescence substrate Pharmingen) in PBS. After washing, monocytes, iDC, or THP-1 were from PerkinElmer Life Sciences. cells (1 ϫ 105 per well) in RPMI medium 1640 with 10% FBS were added, with or without IFN␥ at 20 ng͞ml (R & D Systems), Results except where indicated, and culture supernatants were analyzed ␥ Reciprocal CD1d-Mediated Activation of Human iNKT, Monocytes, and for cytokines by ELISA after 1–8 days. ELISA IFN - and iDCs. Previous studies have shown that CD1d is constitutively IL-4-matched Ab pairs and standards (Endogen) were used. expressed by human monocytes as well as at somewhat elevated Matched Ab pairs and standards for IL-12p70 ELISA were from levels by DCs (23, 24). We confirmed this pattern of CD1d Endogen, and matched Ab pairs and standards for IL-12p40 ϩ expression (Fig. 1A). Interaction of CD1d APC pulsed with ELISA were fromR&DSystems. ␣galcer and iNKT stimulates them to release IL-12 and IFN␥, Parthenolide and N-acetyl-L-leucinyl-L-leucinyl-methional respectively (18, 23–25). Murine iNKT similarly induce APC to (LLM) were from Sigma. Neutralizing IFN␥ mAb (Chemicon) produce IL-12 through IFN␥- and CD40L-dependent mecha- and CD40 mAb were used where indicated. Results show the nisms (11, 12, 20). To determine whether there were functional mean of at least triplicates Ϯ SD, representative of at least three ϩ experiments with monocytes derived from multiple healthy differences between CD1d monocytes versus iDCs in their donors. stimulation of iNKT, iNKT lines were prepared from healthy donors with a mAb against complementarity-determining region Flow Cytometry and Immunoblotting. FACS was performed with 3 of the invariant TCR (21). More than 90% of these iNKT lines 1 ϫ 106 cells after blocking with 10% human serum. iNKT lines from multiple donors expressed the invariant CD1d-reactive ␣ ␣ ␤ ϩ Ϫ were analyzed with anti-V␣24-phycoerythrin (PE) (Coulter) and V 24-J 18 TCR and V 11 and were a mixture of CD4 CD8 ␤ and CD4ϪCD8Ϫ cells (data not shown). 6B11-FITC (anti-invariant TCR) (21) or anti-V 11-FITC ϩ (Coulter). Expression of APC CD1 molecules and CD40 were Rested iNKT lines were stimulated with CD1d peripheral assessed by flow cytometry. For single-color FACS, primary blood monocytes or iDCs in the presence or absence of ␣galcer mAb were used at 10 ␮g͞ml followed by anti-mouse IgG-FITC and then assessed for cytokine production as shown in Fig. 1B. (Kirkegaard & Perry Laboratories). The primary mAb for There was no significant production of IFN␥ when iNKT were CD1a–c were Vit6, 4A76, and M241, respectively (22), anti- cultured with monocytes or iDCs in the absence of ␣galcer, but CD40, and isotypes (BD Pharmingen), trace lymphocytes being addition of ␣galcer resulted in substantial IFN␥ production after gated out for CD1d FACS. Two-color FACS used lineage 1 day in both the monocyte and iDC cultures (Fig. 1B). IFN␥ markers CD1a-PE, CD1d 42.1-PE, CD14-FITC, and CD40- production was Ϸ4-fold higher in the iDC cultures (Fig. 1B), FITC (BD Pharmingen). The CD1d and mock-transfected C1R consistent with the greater amount of IL-12 produced by the cells were described in ref. 22. iDCs (see below). IL-4 production was also stimulated by For immunoblotting, cells were lysed in 1% SDS and equiv- ␣galcer-pulsed monocytes and iDCs and was greater in the iDC alent amounts of protein were run on SDS͞PAGE and trans- cultures (Fig. 1C). These results demonstrated that both mono-

11812 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0503366102 Yue et al. Downloaded by guest on September 23, 2021 Fig. 2. CD1d-dependent IL-12 production by peripheral blood monocytes. (A–E) Peripheral blood monocytes were cultured on plates coated with protein G followed by CD1d (51.1), CD40, or isotype control mAb (IgG2b or IgG1, respectively), with or without IFN␥ (20 ng͞ml), as indicated. (A) Supernatants were collected and assayed for IL-12p40 at day 8. (B) Supernatants were collected at day 3 and assayed for IL-12p70. (C) Monocytes were stimulated by culturing with CD1d mAb as above or by LPS, and IL-12p70 was measured at 1 or 3 days. (D) IFN␥ dose dependence of monocyte IL-12p70 induction by plate-bound CD1d mAb 51.1 (IgG2b) and 42.1 (IgG1). Results shown are for day 3. (E) Monocytes were cultured on a panel of plate-bound CD1d or isotype control mAb plus IFN␥ as above with LPS at 25 or 250 ng͞ml as indicated. Results shown are for day 3; supernatants were assayed for IL-12p70. (F) Monocytes were treated with plate-bound CD1d mAb or control and cultured for 3 days with IFN␥ as above. CD1a and CD40 (thick lines) or isotype mAb control (thin lines) staining. Viable cells were gated as CD3- and CD19-negative. CD1a and CD40 up-regulation was specifically accelerated by CD1d ligation.

cytes and iDC could stimulate iNKT to produce IFN␥ and IL-4, was replaced by exogenous IFN␥. To minimize potential inhib- although iDCs were more potent inducers of both cytokines. itory effects mediated by Fc receptors and to optimize mAb Previous studies showed that iDCs were stimulated to produce orientation, plates were first coated with protein G followed by IL-12 when cultured with iNKT and ␣galcer (11, 12, 18). IL-12 specific or control mAb. Peripheral blood monocytes cultured on production was next assessed to determine whether iNKT could these plates for up to 8 days in the absence of IFN␥ did not stimulate IL-12 production from human monocytes as well as produce significant amounts of IL-12p40, but in the presence of iDCs. Significantly, both monocytes and iDC were rapidly (day IFN␥ CD1d ligation resulted in IL-12p40 production (Fig. 2A). 1) stimulated to produce bioactive IL-12p70 when cultured with In contrast, IFN␥ alone had no effect, and monocyte IL-12p40 iNKT and ␣galcer (Fig. 1D). The higher levels of IL-12 produced production was not stimulated by CD40 mAb. by the iDCs were consistent with relative CD1d levels (Fig. 1A) Although the transcriptional control of IL-12 is primarily as well as with the increased potency of iDC as APC in general through IL-12p40, this chain associates with IL-12p35 to gener- and in stimulating iNKT cytokine production (Fig. 1 B and C). ate the bioactive IL-12p70. Therefore, IL-12p70 production by Finally, experiments were performed to determine whether CD40- and CD1d-stimulated monocytes was also assessed. iNKT-induced human monocyte and iDC IL-12 responses were Monocytes cultured with CD1d mAb, even in the absence of dependent on iNKT IFN␥ as previously shown in mice (11, 12, IFN␥, produced substantial IL-12p70 (Fig. 2B). This CD1d- 20). Similar coincubations of iNKT with ␣galcer and monocytes stimulated IL-12p70 production was further enhanced by the were performed in the presence or absence of neutralizing mAb addition of IFN␥. CD40 mAb also stimulated IL-12p70 produc- to IFN␥. The results showed that anti-IFN␥ could substantially tion, but only in the presence of IFN␥, and it was less potent than inhibit monocyte IL-12 production (Fig. 1E). Furthermore, CD1d (Fig. 2B). additional exogenous IFN␥ could enhance iNKT-dependent Based on these results, we compared monocyte production of monocyte IL-12 production, whereas IFN␥ without ␣galcer had IL-12 in response to CD1d ligation versus LPS stimulation, the no effect (Fig. 1E). latter being a well established mediator of myeloid cell matura- Based on these data showing that CD1dϩ human monocytes tion and IL-12 production that functions through Toll-like (as well as iDCs) could rapidly stimulate and be activated by receptor 4. LPS rapidly stimulated the production of IL-12p70. iNKT, we initiated studies to further define the molecular Significantly, CD1d mAb stimulated comparable levels of IL- mechanisms mediating monocyte activation by iNKT. 12p70 production after 1 or 3 days (Fig. 2C). We also compared the activity of CD1d mAb 51.1, as used CD1d Ligation Induces IL-12 Production and Maturation of Monocytes. above, to several distinct mAb recognizing at least three CD1d Murine studies have shown that CD40L expressed by iNKT can epitopes (22–24). Both 51.1 and 42.1 recognizing distinct but stimulate DCs through CD40 ligation and that IL-12 production overlapping CD1d epitopes showed marked induction of IL-12 by the activated DCs can be enhanced by iNKT-produced IFN␥ with a dose-dependent enhancement by IFN␥, whereas IFN␥ (11, 12, 20). Stimulation of human iDCs by iNKT was similarly alone had no effect (Fig. 2D). As shown in Fig. 2E, the other found to be CD40L-dependent (18). Significantly, DC activation CD1d mAb were also able to stimulate some monocyte IL-12p70 by iNKT in these studies was also ␣galcer-dependent, suggesting production. These results showed that CD1d ligation with any of that ligation of CD1d by the invariant TCR might directly several mAb directed at multiple epitopes could alone provide contribute to DC activation (in addition to the clear role of TCR a rapid and potent stimulus for monocyte production of bioactive ligation by CD1d in activating the iNKT). The data above (Fig. IL-12p70, which could be further enhanced by IFN␥. 1E) further supported a similar IFN␥-dependent mechanism. Next, we addressed whether CD1d ligation could also influ- Therefore, the potential roles of CD1d and CD40 ligation and of ence monocyte differentiation. CD40 is minimally expressed by IFN␥ in stimulating IL-12 production by monocytes and iDC monocytes but is markedly up-regulated on iDC, and CD1a is were directly assessed. expressed only by iDC. As shown in Fig. 2F, CD40 and CD1a

CD40 and CD1d ligation by iNKT were modeled by using up-regulation by cultured monocytes was specifically enhanced IMMUNOLOGY plate-bound CD40 and CD1d mAb, and iNKT-produced IFN␥ by CD1d ligation. This effect was most pronounced at day 3. By

Yue et al. PNAS ͉ August 16, 2005 ͉ vol. 102 ͉ no. 33 ͉ 11813 Downloaded by guest on September 23, 2021 Fig. 3. CD1d ligation-induced IL-12 production by THP-1 cells. THP-1 cells were cultured on plates coated with protein G and the indicated mAb or LPS, plus IFN␥ (20 ng͞ml). Supernatants were collected on day 3 and assayed for IL-12p40 (ng͞ml). (A) mAb against CD1a–d or CD40 were used as stimuli. (B) CD1d mAb with or without heat inactivation (HI) or LPS were used as stimuli. (C) CD1d mAb were used either bound via protein G (PG) or directly applied to plates blocked with BSA or FBS as indicated.

the seventh day of culture, CD40 and CD1a were up-regulated results confirmed that THP-1 stimulation was directly mediated even without CD1d ligation. CD80 up-regulation was so rapid by CD1d mAb. that CD1d ligation had little augmenting effect (data not shown). THP-1 Stimulation by CD1d Ligation Is Mediated Through NF-␬B. CD1d Ligation Stimulates IL-12 Production by the THP-1 Monocytic Cell THP-1 cells were next used to investigate the mechanism of Line. To investigate the mechanism of CD1d-stimulated IL-12 CD1d-mediated activation. Because NF-␬B plays a major role in production, we screened human cell lines and found that THP-1, regulating myeloid expression of IL-12p40 in response to acti- a human monocytic leukemia cell line, expressed CD1d at levels vation by LPS and other Toll-like receptors, we determined similar to those observed on peripheral blood monocytes (Ϸ1% whether the NF-␬B pathway was activated in response to CD1d of the level of C1R.CD1d cells), confirmed by immunoblotting ligation on THP-1 cells. An obligate initial step in NF-␬B (data not shown). In addition to CD1d, the THP-1 cells ex- activation is phosphorylation of I␬BbyI␬B kinase, with subse- pressed low levels of CD1a but did not express detectable CD1b, quent ubiquitination and proteosome-mediated degradation of CD1c, or CD40 (data not shown). pI␬B and release of NF-␬B. To assess this activation step, we We next assessed IL-12 production in response to CD1d carried out immunoblotting for pI␬B. THP-1 cells were prein- ligation on the THP-1 cells. There was no detectable production cubated with LLM (10 ␮M), a proteosome inhibitor to block of the bioactive IL-12p70 under any conditions, including LPS, pI␬B degradation, and then added to CD1d mAb or control- and no significant stimulation of IL-12p40 in the absence of coated wells. An increase in pI␬B was observed within 15 min exogenous IFN␥ (data not shown). However, with the addition specifically in wells coated with protein G and CD1d mAb (Fig. of IFN␥, the plate-bound CD1d mAb specifically stimulated 4A). A further increase was seen at 45 min, comparable to the IL-12p40 production at levels that were comparable to LPS activation observed with LPS at 15 min. stimulation (Fig. 3A). Consistent with the lack of detectable We next carried out inhibitor studies to determine whether the CD40 expression on the THP-1 cells, there was no stimulation production of IL-12p40 in response to CD1d ligation depended by CD40 mAb. To further determine whether the IL-12 stimu- on NF-␬B activation. CD1d-stimulated production of IL-12 lation was specific to CD1d, the effect of CD1a ligation on could be blocked by a number of relatively nonspecific antago- THP-1 cells was also assessed. In contrast to CD1d, the plate- nists of NF-␬B, including proteosome inhibitors and pyrrolidine bound CD1a mAb did not stimulate IL-12p40 release (Fig. 3A). dithiocarbamate (not shown). Parthenolide is a relatively specific Plate-bound CD1b and CD1c mAb similarly had no effect. NF-␬B antagonist that inhibits activity of the I␬B kinase complex Additional controls confirming that crosslinking by the plate- and phosphorylation of I␬B (26). To enhance specificity, we first bound CD1d mAb was mediating the THP-1 cell stimulation carried out parthenolide dose–response studies to determine the included heat inactivation, which abrogated the ability of the minimal concentration of drug needed to suppress LPS- CD1d mAb to stimulate IL-12 production (Fig. 3B). CD1d mediated phosphorylation of I␬B in THP-1 cells. As shown in stimulation also depended on protein G coating (Fig. 3C). These Fig. 4B, parthenolide at a concentration of 10 ␮M (in the

Fig. 4. CD1d-mediated activation of NF-␬B in THP-1 cells. (A) THP-1 cells were preincubated for 15 min with LLM (10 ␮M) and then cultured on wells coated with protein G (PG) and 51.1 CD1d or control mAb, as indicated. Cells were lysed at 15 or 45 min, and lysates were immunoblotted with anti-pI␬B Ab. Densitometry was as shown. (B) THP-1 cells were pretreated for 1 h with LLM (10 ␮M) and parthenolide at the indicated concentrations and then stimulated with LPS for 15 min. Cell lysates were then immunoblotted with anti-pI␬B. (C) THP-1 cells were cultured on wells coated with protein G and the indicated mAb in medium with IFN␥ (20 ng͞ml), without or with parthenolide (10 ␮M, preincubated for 1 h before adding to wells). Supernatants were collected on day 3 and assayed for IL-12p40. The data are presented as % LPS control (without parthenolide).

11814 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0503366102 Yue et al. Downloaded by guest on September 23, 2021 Fig. 5. CD1d-dependent IL-12 production by iDCs. Peripheral blood-derived iDC were incubated on plates coated with protein G followed by CD1d (51.1) and CD40, isotype control mAb (IgG2b and IgG1, respectively), or LPS (0–20 ␮g͞ml) plus IFN␥. Supernatants were assayed for IL-12p70. (A) Day 1 iDC dose–response curves for CD1d mAb with or without 1 ␮g͞ml CD40 mAb. (B) Day 3 iDC dose–response curves for CD1d mAb with or without 1 ␮g͞ml CD40 mAb. (C) Day 3 iDC dose–response curve for LPS with or without 1 ␮g͞ml CD1d mAb.

presence of LLM) was required to substantially inhibit the ulation of IL-12 production was confirmed in the human THP-1 response to LPS. monocytic leukemia cell line and was found to be mediated by Based on this dose–response, THP-1 cells were pretreated rapid activation of the NF-␬B signal transduction pathway. with 10 ␮M parthenolide (in the absence of LLM) and then Finally, we found that both CD40 and LPS, two known physi- stimulated with CD1d or isotype control mAb. Parthenolide ological stimuli of iDC and monocytes, respectively, could act completely blocked the basal and CD1d mAb-stimulated pro- synergistically with CD1d to stimulate IL-12p70 production. duction of IL-12p40 (Fig. 4C). LPS-stimulated IL-12 production Taken together, these studies support a direct role for CD1d was only partially blocked, either reflecting more robust NF-␬B ligation in the rapid activation of the innate immune response by activation by LPS or an NF-␬B independent pathway to IL-12 CD1d-reactive NKT cells. production. Taken together, these data indicated that CD1d A major function of conventional activated CD4ϩ T cells is to ligation stimulated IL-12 production through an NF-␬B- interact with iDCs bearing cognate MHC class II-presented mediated pathway. antigens and stimulate their further maturation through CD40L ligation of CD40. Autoreactive T cell clones specific for group 1 Synergistic Effects of CD1d with Either LPS or CD40 Ligation on APC CD1 (CD1a, b, or c) were similarly found to stimulate Activation. The studies above indicate that CD1d ligation can the maturation of human iDCs, but distinct mechanisms includ- provide a rapid signal that is presumably integrated with other ing TNF␣ appeared to be involved and IL-12 production was physiological stimuli for DC activation. To test this hypothesis, observed only when iDCs were also stimulated with LPS (18). In costimulation by CD1d and either CD40 ligation or LPS was contrast to previous studies, the data reported here demonstrate assessed. As shown in Fig. 5A, CD40 mAb (1 ␮g͞ml) alone was that CD1d ligation can directly mediate monocytic cell activa- a very weak early stimulus of iDC at 1 day, but it significantly tion. The ability to directly stimulate monocytes through CD1d enhanced the stimulation by CD1d. Synergy was again observed ligation may be a unique property of CD1d-reactive NKT cells at day 3, with CD40 enhancing stimulation by lower concentra- and reflect an important physiological role for iNKT–monocyte tions of CD1d mAb (Fig. 5B). At lower concentrations of CD40 interactions in the initiation of innate immune responses. The mAb, similar synergy was seen, whereas at 10 ␮g͞ml CD1d (Fig. hypothesis that monocytes are physiologically activated by iNKT 5 A and B) or CD40 mAb (data not shown) the stronger iDC through CD1d ligation is consistent with the constitutive expres- responses precluded detection of CD1d synergy. sion of CD1d on monocytes as well as DCs (23, 24), versus the Similarly, low concentrations of LPS produced significantly regulated expression of proteins mediating interactions with less monocyte IL-12, which could be specifically enhanced in the conventional T cells (MHC class II, CD1a–c, CD40, CD80, and presence of CD1d mAb (Fig. 5C). A second CD1d mAb, 42.1, CD86). which alone could also induce IL-12 (Fig. 2 D and E), could also CD1d has not been shown to associate with adaptor proteins act synergistically with CD40 mAb (data not shown). Therefore, that might link it to NF-␬B, and previous biochemical studies of two distinct known physiologically important stimuli of APC CD1d have focused on endosomal targeting and its role in IL-12 production (LPS and CD40) could act in synergy with antigen presentation (27, 28). However, there are data suggest- ligation by different CD1d mAb. ing that the CD1d cytoplasmic tail may mediate other interac- tions. An early study found that human T cell lines could be Discussion stimulated to release intracellular Ca2ϩ by Abs to CD1b or CD1c This study addressed the molecular interactions mediating ac- but not by Abs to CD1a (29), which uniquely lacks a cytoplasmic tivation of human CD1dϩ monocytic lineage cells by CD1d- tail tyrosine. More recently it was found that crosslinking of reactive NKT cells. We initially showed that cultured human CD1d, but not a cytoplasmic tail deletion mutant, on intestinal iNKT could stimulate the rapid production of IL-12 from both epithelial cells could stimulate IL-10 production (30). ␣galcer-pulsed CD1dϩ monocytes and iDCs. Ab blocking im- In other studies CD1d was shown to bind to the MHC class plicated IFN␥ in the IL-12 response of monocytes, consistent II–invariant chain complex in the endoplasmic reticulum and with murine data (11, 12). Possible mechanisms mediating this associate with this complex on the cell surface, interactions that stimulation, including CD40L ligation of APC CD40, TCR were independent of the CD1d cytoplasmic tail (31). This MHC ligation of the CD1d-␣galcer complex, and iNKT-produced class II association is intriguing, because TCR-mediated ligation IFN␥, were then further dissected. Remarkably, CD1d ligation of MHC class II proteins can provide a coactivation signal in B alone, through any of multiple independent mAb against at least cells (32). However, this B cell signal is transduced through MHC three CD1d epitopes, could rapidly stimulate CD1dϩ monocytes class II association with the B cell receptor and its signal as well as iDC to produce bioactive IL-12p70. Furthermore, transduction complex, which are not present in monocytes. One CD1d ligation enhanced monocyte differentiation, consistent study has shown that human monocytes can be activated by MHC with this representing a physiological response. Notably, IFN␥ by class II crosslinking through mitogen-activated protein kinase

itself had no effect, but it markedly increased the CD1d- pathways (33). Although we have not observed consistent mi- IMMUNOLOGY mediated stimulation of IL-12 production. CD1d-mediated stim- togen-activated protein kinase pathway activation in response to

Yue et al. PNAS ͉ August 16, 2005 ͉ vol. 102 ͉ no. 33 ͉ 11815 Downloaded by guest on September 23, 2021 CD1d ligation (data not shown), a link between MHC class II provide a further mechanism for CD1d clustering through CD4 and CD1d signaling in monocytic cells remains possible. The binding of MHC class II͞CD1d complexes in the case of CD4ϩ established role of NF-␬B in activating IL-12p40 transcription in iNKT. In any case, even with a relatively high-affinity CD1d response to other stimuli suggests that CD1d ligation functions ligand it is not clear whether the invariant TCR can mediate the similarly upstream of NF-␬B. Further studies are underway to level of CD1d crosslinking achieved with CD1d Abs. Indeed, it determine whether CD1d is linked to NF-␬B by AP-2 or MHC seems likely that the signal generated in vivo by CD1d ligation class II proteins or by a distinct mechanism. will in many cases not be strong enough to by itself trigger Recently, physiologically relevant endogenous and pathogen- monocyte activation and will instead be integrated with other derived relatively high-affinity CD1d-presented antigens recog- signals. In addition to CD40 ligation, this includes at least one nized by the invariant TCR have been defined (34–36). This also Toll-like receptor-mediated signal (LPS) functioning through is significant with respect to CD1d-mediated signal transduction, NF-␬B, which would provide a molecular mechanism for iNKT because the invariant TCR may require a high-affinity antigen to modulation of innate immune responses. achieve a threshold level of CD1d crosslinking. However, an alternative to a high-affinity antigen is a lower-affinity multiva- We thank colleagues, especially Dr. S. Porcelli (Albert Einstein College of Medicine, Bronx, NY) for CD1d mAb; Drs. M. Brenner, S. Porcelli, lent glycolipid antigen. The invariant TCR could function to ␣ stabilize or enhance the clustering of CD1d that is bound to and S. B. Wilson for advice; and Kirin Brewery for galcer. This work was supported by the National Institutes of Health (Grant R01 low-affinity multivalent antigen on the monocyte cell surface. In DK066917 to M.A.E. and Grant R01 AI42955 to S.P.B.), the Dana this context, CD1d is known to accumulate in lipid rafts, and Farber͞Harvard Cancer Center Melanoma Specialized Program of their disruption can inhibit iNKT recognition (37). An interac- Research Excellence (Grant P50 CA93683), and the Hershey Family tion between CD1d and MHC class II on the cell surface may Prostate Cancer Research Fund.

1. Bendelac, A., Rivera, M. N., Park, S. H. & Roark, J. H. (1997) Annu. Rev. 20. Hayakawa, Y., Takeda, K., Yagita, H., Van Kaer, L., Saiki, I. & Okumura, K. Immunol. 15, 535–562. (2001) J. Immunol. 166, 6012–6018. 2. Porcelli, S. A. & Modlin, R. L. (1999) Annu. Rev. Immunol. 17, 297–329. 21. Tahir, S. M., Cheng, O., Shaulov, A., Koezuka, Y., Bubley, G. J., Wilson, S. B., 3. Benlagha, K. & Bendelac, A. (2000) Semin. Immunol. 12, 537–542. Balk, S. P. & Exley, M. A. (2001) J. Immunol. 167, 4046–4050. 4. Stein-Streilein, J. (2003) J. Exp. Med. 198, 1779–1783. 22. Exley, M., Garcia, J., Balk, S. P. & Porcelli, S. (1997) J. Exp. Med. 186, 109–120. 5. Kawano, T., Cui, J., Koezuka, Y., Toura, I., Kaneko, Y., Motoki, Ueno, H., 23. Exley, M., Garcia, J., Wilson, S. B., Spada, F., Gerdes, D., Tahir, S. M., Patton, Nakagawa, R., Sato, H., Kondo, E., et al. (1997) Science 278, 1626–1629. K. T., Blumberg, R. S., Porcelli, S., Chott, A. & Balk, S. P. (2000) Immunology 6. Brossay, L., Chioda, M., Burdin, N., Koezuka, Y., Casorati, G., Dellabona, P. 100, 37–47. & Kronenberg, M. (1998) J. Exp. Med. 188, 1521–1528. 24. Spada, F. M., Borriello, F., Sugita, M., Watts, G. F., Koezuka, Y. & Porcelli, 7. Burdin, N., Brossay, L., Koezuka, Y., Smiley, S. T., Grusby, M. J., Gui, M., S. A. (2000) Eur. J. Immunol. 30, 3468–3477. Taniguchi, M., Hayakawa, K. & Kronenberg, M. (1998) J. Immunol. 161, 25. Nieda, M., Nicol, A., Koezuka, Y., Kikuchi, A., Takahashi, T., Nakamura, H., 3271–3281. Furukawa, H., Yabe, T., Ishikawa, Y., Tadokoro, K., et al. (1999) Hum. 8. Spada, F. M., Koezuka, Y. & Porcelli, S. A. (1998) J. Exp. Med. 188, 1529–1534. Immunol. 60, 10–19. 9. Carnaud, C., Lee, D., Donnars, O., Park, S. H., Beavis, A., Koezuka, Y. & 26. Hehner, S. P., Hofmann, T. G., Droge, W. & Schmitz, M. L. (1999) J. Immunol. Bendelac, A. (1999) J. Immunol. 163, 4647–4650. 163, 5617–5623. 30, 10. Eberl, G. & MacDonald, H. R. (2000) Eur. J. Immunol. 985–992. 27. Roberts, T. J., Sriram, V., Spence, P. M., Gui, M., Hayakawa, K., Bacik, I., 11. Tomura, M., Yu, W. G., Ahn, H. J., Yamashita, M., Yang, Y. F., Ono, S., Bennink, J. R., Yewdell, J. W. & Brutkiewicz, R. R. (2002) J. Immunol. 168, Hamaoka, T., Kawano, T., Taniguchi, M., Koezuka, Y. & Fujiwara, H. (1999) 5409–5414. J. Immunol. 163, 93–101. 28. Cernadas, M., Sugita, M., van der, W., Cao, X., Gumperz, J. E., Maltsev, S., 12. Kitamura, H., Iwakabe, K., Yahata, T., Nishimura, S., Ohta, A., Ohmi, Y., Sato, Besra, G. S., Behar, S. M., Peters, P. J. & Brenner, M. B. (2003) J. Immunol. M., Takeda, K., Okumura, K., Van Kaer, L., et al. (1999) J. Exp. Med. 189, 171, 4149–4155. 1121–1128. 29. Theodorou, I. D., Boumsell, L., Calvo, C. F., Gouy, H., Beral, H. M. & Debre, 13. Wilson, S. B., Kent, S. C., Horton, H. F., Hill, A. A., Bollyky, P. L., Hafler, J. Immunol. 144, D. A., Strominger, J. L. & Byrne, M. C. (2000) Proc. Natl. Acad. Sci. USA 97, P. (1990) 2518–2523. 7411–7416. 30. Colgan, S. P., Hershberg, R. M., Furuta, G. T. & Blumberg, R. S. (1999) Proc. 14. Yang, O. O., Racke, F. K., Nguyen, P. T., Gausling, R., Severino, M. E., Horton, Natl. Acad. Sci. USA 96, 13938–13943. H. F., Byrne, M. C., Strominger, J. L. & Wilson, S. B. (2000) J. Immunol. 165, 31. Kang, S. J. & Cresswell, P. (2002) EMBO J. 21, 1650–1660. 3756–3762. 32. Lang, P., Stolpa, J. C., Freiberg, B. A., Crawford, F., Kappler, J., Kupfer, A. 15. Nishimura, T., Kitamura, H., Iwakabe, K., Yahata, T., Ohta, A., Sato, M., & Cambier, J. C. (2001) Science 291, 1537–1540. Takeda, K., Okumura, K., Van Kaer, L., Kawano, T., et al. (2000) Int. Immunol. 33. Matsuoka, T., Tabata, H. & Matsushita, S. (2001) J. Immunol. 166, 2202–2208. 12, 987–994. 34. Zhou, D., Mattner, J., Cantu, C., Schrantz, N., Yin, N., Gao, Y., Sagiv, Y., 16. Naumov, Y. N., Bahjat, K. S., Gausling, R., Abraham, R., Exley, M. A., Hudspeth, K., Wu, Y. P., Yamashita, T., et al. (2004) Science 306, 1786–1789. Koezuka, Y., Balk, S. B., Strominger, J. L., Clare-Salzer, M. & Wilson, S. B. 35. Wu, D., Xing, G. W., Poles, M. A., Horowitz, A., Kinjo, Y., Sullivan, B., (2001) Proc. Natl. Acad. Sci. USA 98, 13838–13843. Bodmer-Narkevitch, V., Plettenburg, O., Kronenberg, M., Tsuji, M., et al. 17. Faunce, D. E. & Stein-Streilein, J. (2002) J. Immunol. 169, 31–38. (2005) Proc. Natl. Acad. Sci. USA 102, 1351–1356. 18. Vincent, M. S., Leslie, D. S., Gumperz, J. E., Xiong, X., Grant, E. P. & Brenner, 36. Mattner, J., DeBord, K. L., Ismail, N., Goff, R. D., Cantu, C., III, Zhou, D., M. B. (2002) Nat. Immunol. 3, 1163–1168. Saint-Mezard, P., Wang, V., Gao, Y., Yin, N., et al. (2005) Nature 434, 525–529. 19. Brigl, M., Bry, L., Kent, S. C., Gumperz, J. E. & Brenner, M. B. (2003) Nat. 37. Lang, G. A., Maltsev, S. D., Besra, G. S. & Lang, M. L. (2004) Immunology 112, Immunol. 4, 1230–1237. 386–396.

11816 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0503366102 Yue et al. Downloaded by guest on September 23, 2021