Defective removal of invariant chain peptides from MHC class II suppresses tumor antigen presentation and promotes tumor growth

SUMMARY

Tumor-draining lymph node dendritic cells (DCs) are poor stimulators of tumor antigen-specific CD4 T cells; however, the mechanism behind this defect is unclear. We now show that, in tumor-draining lymph node DCs, a large proportion of major histocompatibility complex class II (MHC-II) molecules retains the class II-associated invariant chain peptide (CLIP) fragment of the invariant chain bound to the MHC-II peptide binding groove due to reduced expression of the peptide editor H2-M and enhanced activity of the CLIP-generating proteinase cathepsin S. The net effect of this is that MHC-II molecules are unable to efficiently bind antigenic peptides. DCs in mice expressing a mutation in the invariant chain sequence that results in enhanced MHC-II-CLIP accumulation are poor stimulators of CD4 T cells and have diminished anti-tumor responses. Our data reveal a previously unknown mechanism of immune evasion in which enhanced expression of MHC-II-CLIP complexes on tumor-draining lymph node DCs limits MHC-II availability for tumor peptides.

In brief

CD4 T cell priming by dendritic cells is defective in the tumor-draining lymph node. Bandola-Simon et al. show that cDC2s in the tumor-draining lymph node retain large amounts of the CLIP “placeholder” peptide in the MHC-II peptide binding groove, thereby diminishing their ability to present antigens and impairing anti-tumor responses.

Graphical abstract

INTRODUCTION

T cell immunotherapy designed to attack and eliminate tumor cells requires that tumor-specific CD8 cytotoxic T cells receive help signals, both in the form of engaged surface costimulatory receptors on antigen-presenting cells (APCs) as well as secreted cytokines from tumor antigen-specific CD4 T helper cells.^1–3 Although tumor antigen-specific CD8 T cell priming in tumor-draining lymph nodes (tdLNs) is an efficient process, these primed CD8 T cells often do not acquire the cytotoxic function required for them to serve as efficient tumor-specific killer T cells.^4,5 CD4 T cell help is insufficient in most anti-tumor responses, and this defect has been largely tied to dysfunctional tumor antigen presentation by professional APCs in tdLNs.⁶ Among all APCs able to orchestrate anti-tumor T cell responses, conventional dendritic cells (cDCs) are particularly suited for processing and presenting tumor antigens.⁷ Based on the surface markers, transcriptional regulation, and immunological function, this heterogeneous family can be divided into two main lineages: type 1 cDCs (cDC1s) are generally (although not exclusively) considered to be specialized in cross-presentation to CD8 T cells, whereas type 2 cDCs (cDC2s) specialize in antigen presentation to CD4 T cells.^8–10

Stimulation of tumor antigen-specific CD4 T cells depends on antigen presentation on major histocompatibility complex class II (MHC-II) molecules expressed by DCs present in tdLNs.¹¹ Antigen presentation via the MHC-II pathway requires intracellular MHC-II movement to endosomal antigen processing compartments, where MHC-II binds internalized and degraded tumor antigens (reviewed in Roche and Furuta¹²). The movement of newly synthesized MHC-II to these antigen-processing compartments is mediated by its association with a chaperone protein termed the invariant chain (Ii) in the endoplasmic reticulum (ER). Ii performs a number of tasks for MHC-II: Ii assists in MHC-II folding, targets nascent MHC-II-Ii complexes to antigen-processing compartments, and, most importantly, occupies the peptide-binding cleft of MHC-II, preventing the premature binding of antigenic peptides to newly synthesized MHC-II in the ER.¹² Following the delivery of MHC-II-Ii complexes to antigen-processing compartments, MHC-II-associated Ii is sequentially degraded by the proteinase cathepsin S,¹³ leaving a short (approximately 12- to 24-amino-acid) Ii fragment in the peptide-binding cleft of the MHC-II molecule. This fragment, termed class II-associated Ii peptide (CLIP), must be removed before antigenic peptides can bind to MHC-II.^14–16 CLIP is removed from the MHC-II peptide binding cleft by H2-M, a protein that accelerates dissociation of CLIP and edits the peptide repertoire.^17–19 Different human and mouse MHC-II alleles have varying affinities for CLIP²⁰ and different proportions of surface MHC-II retaining CLIP, with some alleles (such as HLA-DR1) possessing as much as 19% of all surface MHC-II molecules containing CLIP.²¹ Genetic deletion of H2-M in mice results in the accumulation of large amounts of MHC-II-CLIP on the surface of their APCs, which makes H2-M-deficient APCs extremely poor stimulators of antigen-specific CD4 T cells.^22,23

Like pathogens, tumor cells utilize many different mechanisms to avoid recognition by the acquired immune system, including blocking antigen processing by the proteasome, modifying antigen loading onto MHC-I, and suppressing expression of MHC molecules.^24–26 In this study, we report a novel mechanism of immune evasion adopted by tumor cells; namely, suppressing CLIP removal from the peptide-binding groove of MHC-II molecules in cDCs in tdLNs, limiting the ability of those cells to present antigenic peptides to CD4 T cells and therefore promoting tumor growth.

RESULTS

CD4 T cell priming by DCs is defective in the tdLN

To examine how tumors impact T cell priming via MHC-I and MHC-II pathways in the lymph nodes, we compared the ability of cDCs to process and present the model antigen ovalbumin (OVA) to OVA-specific CD8 or CD4 T cells in the tdLN and the control contralateral non-draining lymph node (ndLN) in the same mouse. CFSE-labeled T cells were adoptively transferred into MB49 bladder carcinoma tumor-bearing mice that were subsequently immunized with OVA. The proliferation of CD8 T cells was identical in tdLNs and control ndLNs. By contrast, CD4 T cell proliferation was significantly diminished in the tdLN as compared to the ndLN (Figure 1A), revealing defective MHC-II antigen presentation. To measure the ability of cDCs in tdLNs to form peptide-MHC-II complexes in vivo, we injected 50 μg of OVA-Eα fusion protein both intratumorally (i.t.) and subcutaneously (s.c.) on the contralateral flank and analyzed the cDCs 24 h later. The ability to form MHC-II-Eα_(52–68) complexes was significantly reduced in tdLN cDCs (Figure 1B), a finding that agrees with a previous study performed with a different tumor model.⁴

Figure 1. Antigen presentation to CD4 T cells by DCs in tdLNs is defective.

(A) Proliferation of OVA-specific T cells transferred to tumor-bearing mice in response to immunization with OVA/alum was determined by flow cytometry. Control mice were injected with PBS/alum. Three recipient mice per group were injected in each experiment, and the data shown are representative of 3 independent experiments.

(B) The ability of DCs to form MHC-II-Eα_(52–68) complexes was determined by fluorescence-activated cell sorting (FACS). Four mice were injected with OVA-Eα fusion protein in each experiment, and the data shown are representative of 2 independent experiments.

(C) DCs from tdLNs and ndLNs were cultured with T cells in the presence of OVA. The amounts of IFNγ and IL-2 released by T cells were measured by ELISA in triplicate. The data shown are representative of 3 independent experiments.

(D) DCs from tdLNs and ndLNs were incubated with OVA-Eα fusion protein together with OVA-AF594 (D) or with Eα_(52–68) peptide alone (E), and their ability to form MHC-II-Eα _(52–68) complexes was analyzed by flow cytometry. The experiments were performed in triplicate, and data shown are representative of 3 independent experiments.

Data are shown as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001; ns, not significant.

To exclude the impact of other cells, altered cDC numbers, or inability to access antigen, we purified ndLN and tdLN DCs from tumor-bearing mice and compared their capacity to stimulate OVA-specific T cells ex vivo. In agreement with our in vivo observations, we did not detect any difference in interferon (IFNγ) production by OT-I CD8 T cells primed by OVA-pulsed DCs isolated either from tdLNs or ndLNs. By contrast, interleukin-2 (IL-2) production by OT-II CD4 T cells was significantly reduced when primed by OVA-pulsed DCs isolated from tdLNs as compared to ndLNs (Figure 1C). In agreement with our in vivo findings, the ability of tdLN cDCs to generate MHC-II-Eα_(52–68) complexes was greatly diminished when the cells were incubated with OVA-Eα fusion protein ex vivo despite the fact that both ndLN DCs and tdLN cDCs internalized similar amounts of protein (Figure 1D). cDCs in the tdLN were even defective in generating MHC-II-Eα_(52–68) complexes when incubated with short Eα_(52–68) peptide (Figure 1E). These data show that, whereas the ability to stimulate CD8 T cells is not detectably altered, antigen presentation to CD4 T cells by cDCs in tdLNs is significantly reduced and that cDCs in tdLNs possess far fewer peptide-receptive MHC-II molecules available for antigen presentation.

cDC2s in tdLNs have significantly enhanced expression of MHC-II-CLIP complexes

Skin-draining LNs contain two main cDC subsets: resident cDCs and migratory cDCs.²⁷ Each of these cDC subsets is further divided into cDC1s and cDC2s, which can be distinguished by analyzing the expression of CD26, CD11c, MHC-II, XCR1, an CD172a (Sirp-α)^28–30 (Figure 2A). We observed no differences in MHC-I expression between cDC subsets in tdLNs compared to the control ndLNs (Figure 2B and S1A), a finding that is consistent with our observation that antigen presentation to CD8 T cells was not affected in the tdLNs. There was also no difference in the expression of the costimulatory molecule CD40 in any cDC subsets. We did, however, observe a reduction of surface MHC-II in migratory cDC2s (Figure 2B) but not in cDC1s (Figure S1A). This modest reduction in MHC-II expression in cDC2s was, however, far less than the reduction in peptide-receptive MHC-II observed in ex vivo peptide-pulsing experiments.

Figure 2. Tumors enhance MHC-II-CLIP expression on cDC2s in tdLNs.

(A) Gating strategy for phenotypic analysis of DC subsets in skin-draining LNs by flow cytometry

(B) MHC-I, CD40, and MHC-II expression in migratory and resident DCs in tdLNs and ndLNs was determined by flow cytometry. Four mice were analyzed in each experiment, and the results shown are representative of 3 independent experiments. Isotype control staining from both ndLN and tdLN samples was identical.

(C) MHC-II-CLIP expression in tdLN and ndLN DCs was determined by flow cytometry. The ratio of MHC-II-CLIP relative to the total MHC-II in each sample was calculated and normalized to ndLNs. Four mice were analyzed in each experiment, and the results shown are representative of 3 independent experiments. See also Figure S1.

(D) Fluorescently labeled DCs were injected into tumor-bearing mice both i.t. and s.c. on the contralateral flank. MHC-II and MHC-II-CLIP expression on transferred DCs that migrated to the LNs was analyzed by flow cytometry. The ratio of MHC-II-CLIP amount relative to the total amount of MHC-II present in each sample was calculated and normalized to ndLN DCs. Four mice were analyzed in each group, and the data shown are representative of 2 independent experiments.

Created with BioRender. Data are shown as mean ± SEM; *p < 0.05, **p < 0.01, ***p < 0.001.

MHC-II is only able to bind antigenic peptides following removal of the Ii CLIP fragment from the peptide-binding groove.^14,16 In contrast to the reduction in total MHC-II in tdLN cDC2s, we actually observed a significant increase in expression of MHC-II-CLIP complexes in both migratory and resident tdLN cDC2s as compared to their counterparts in the ndLN by using an MHC-II I-A^b-CLIP-specific monoclonal antibody (mAb) (Figure 2C). No change in MHC-II-CLIP expression was observed in cDC1s (Figure S1B). Accounting for the reduction in surface MHC-II expression, cDC2s in tdLNs have a 2-fold increase in MHC-II-CLIP as compared to ndLNs. The increase in MHC-II-CLIP in tdLN cDC2s was not insignificant, as quantitative analysis revealed that at least 13% of all surface MHC-II molecules are occupied by CLIP in cDC2s at steady state (Figure S2). Our data demonstrate that tdLN cDC2s possess less peptide-receptive MHC-II because many MHC-II molecules retain CLIP. This defect was not tumor type specific, as we observed a similar increase in MHC-II-CLIP expression on tdLN DCs in the MC38 colon adenocarcinoma and B16F10 melanoma tumor models (Figure S3).

Migration through the tumor microenvironment enhances MHC-II-CLIP expression on DCs

We observed the greatest increase in MHC-II-CLIP in migratory DCs present in tdLNs, and we therefore asked whether migration through the tumor microenvironment induced these changes. We injected fluorescently labeled DCs into a tumor-bearing recipient, with half of the cells being injected i.t. and the other half injected s.c. in the contralateral flank. Despite the reduced expression of total MHC-II, the absolute amount of MHC-II-CLIP on transferred DCs that migrated to tdLNs was significantly increased in comparison to ndLNs. When normalized for total MHC-II, DCs that migrated to tdLNs had a 4-fold increase in MHC-II-CLIP as compared to DCs that migrated from the non-tumor site (Figure 2D). These data demonstrate that migration of DCs through the tumor microenvironment en route to tdLNs reduces total MHC-II expression and increases the proportion of these MHC-II molecules retaining CLIP in the peptide-binding groove.

cDC2s in tdLNs have reduced expression of H2-M and enhanced cathepsin S activity

The peptide editor H2-M removes CLIP from the MHC-II peptide binding cleft,¹⁷ and for this reason, we examined the intracellular expression of H2-M in DCs from tumor-bearing mice. Our analysis revealed significantly reduced amounts of H2-M in both migratory and resident cDC2s in tdLNs (Figure 3A), with only a modest reduction in migratory cDC1s (Figure S4) in comparison to ndLNs. A similar reduction in H2-M expression was also observed in the MC38 and B16F10 transplantable tumor models (Figure S3). To directly examine whether reduced H2-M expression can impair CLIP removal from MHC-II, we examined MHC-II-CLIP expression on H2-M heterozygous mice. DCs isolated from H2-M heterozygous mice express normal amounts of total MHC-II (Figure S5) but have a 2-fold increase in MHC-II-CLIP as compared to wild-type mice (Figure 3B). These data demonstrate that reduced expression of H2-M, as observed in tdLN cDC2s, leads to defective removal of CLIP from MHC-II.

Figure 3. tdLN cDC2s have reduced H2-M expression and enhanced cathepsin S activity.

(A) The expression of H2-M in cDC subsets in tdLNs and ndLNs was determined by flow cytometry. Four mice were analyzed in each experiment, and the results shown are representative of 3 independent experiments.

(B) Surface expression of MHC-II-CLIP in DC subsets in H2-M^+/+ or H2-M^+/− mice was analyzed by flow cytometry. Three mice from each strain were analyzed in each experiment, and the results shown are representative of 3 independent experiments.

(C) Gating strategy used for the purification of DC subsets from tdLNs and ndLNs by flow cytometry.

(D) The enzymatic activity of cathepsin S in cell lysates of FACS-sorted DCs from tdLNs and ndLNs was measured in a fluorometric assay. The results shown are from 4 independent experiments combined.

(E) The expression of cystatin C in FACS-sorted DCs from tdLNs and ndLNs was determined by immunoblot analysis. The results shown are from 3 independent experiments combined.

Data are shown as mean ± SEM; *p < 0.05, **p < 0.01. See also Figures S4 and S6.

Cathepsin S is the proteinase responsible for cleaving MHC-II-associated Ii to generate CLIP fragments,¹³ and for this reason we compared cathepsin S activity in cDC1s and cDC2s purified from tdLNs and ndLNs (Figure 3C). We observed a 2-fold increase in cathepsin S activity in lysates of cDC2s (Figure 3D) but not cDC1s (Figure S6A) isolated from tdLNs as compared to ndLNs. Cystatin C is a known endo/lysosomal inhibitor of cathepsin S³¹; therefore, we wondered whether the enhanced cathepsin S activity in tdLN cDCs stems from the loss of its inhibition by cystatin C. As predicted, immunoblot analysis revealed that the expression of cystatin C is significantly reduced in tdLN cDC2s (Figure 3E) but not in cDC1s (Figure S6B) in comparison to ndLNs. Since cathepsin S generates MHC-II-CLIP complexes, and H2-M removes CLIP from these complexes, our data reveal a novel tumor-induced mechanism of immune evasion in which enhanced activity of cathepsin S, together with reduced expression of H2-M, leads to the accumulation of MHC-II-CLIP complexes on the surface of cDCs.

Increased expression of MHC-II-CLIP reduces the amount of peptide-receptive MHC-II

To ask whether enhanced expression of MHC-II-CLIP directly suppresses antigen presentation and promotes tumor growth, we generated the M98A Ii mouse strain with increased binding affinity of CLIP to MHC-II. To achieve this, endogenous Ii was modified such that the amino acid methionine at position 98 of the CLIP fragment of Ii was mutated to alanine (Figure 4A), as shown in Ito et al.³² By crossing these mice onto an H2^b genetic background, we were able to monitor MHC-II I-A^b-CLIP complex expression on the surface of DCs. Whereas total MHC-II expression was not altered in M98A Ii mice (Figure 4B), MHC-II-CLIP surface expression was significantly increased on all cDC subsets as compared to control wild-type (WT) Ii mice.

Figure 4. Increased CLIP retention on MHC-II impairs antigen presentation to CD4 T cells.

(A) The CLIP peptide sequence in WT Ii and M98A Ii mice.

(B) Surface expression of MHC-II and MHC-II-CLIP complexes on WT Ii or M98A Ii DC subsets was determined by flow cytometry. Four mice were analyzed in each experiment, and the results shown are representative of 3 independent experiments.

(C) The ability of WT Ii and M98A Ii DCs to form MHC-II-Eα_(52–68) and MHC-I-OVA_(257–264) complexes was measured by flow cytometry. Four mice were analyzed in each experiment, and the results shown are representative of 3 independent experiments.

(D) The proliferation of OVA-specific T cells adoptively transferred to WT Ii or M98A Ii mice following immunization with OVA/alum was determined by flow cytometry. Three mice of each strain were injected in each experiment, and the data shown are representative of 3 independent experiments.

(E) WT Ii or M98A Ii DCs were cultured with OT-II T cells either in the presence of OVA protein or OVA_(323–339) peptide in triplicate. The amount of IL-2 released by proliferating T cells was measured by ELISA. The data shown are representative of 3 independent experiments.

(F) WT Ii or M98A Ii DCs were cultured with OT-II T cells in the presence of OVA protein under Th2 skewing conditions in triplicate. T cell proliferation and IL-4 production by T cells were analyzed by FACS. The data shown are representative of 3 independent experiments.

Data are shown as mean ± SEM; *p < 0.05, **p < 0.01, ****p < 0.0001.

We asked whether the increased affinity of CLIP to MHC-II on M98A Ii cDCs impacts their function. Purified cDCs were incubated with peptides that bind specifically to either MHC-II or to MHC-I. Incubation with both peptides simultaneously allowed us to measure the formation of MHC-II-Eα_(52–68) complexes as well as MHC-I-OVA_(257–264) complexes. MHC-II on cDC from M98A Ii mice bound far fewer Eα_(52–68) peptides as compared to their WT Ii counterparts (Figure 4C), whereas both WT Ii and M98A Ii cDCs bound comparable amounts of OVA_(257–264) peptides to MHC-I. These data demonstrate that, although WT Ii and M98A Ii cDCs express identical amounts of total MHC-II molecules, the MHC-II on M98A Ii DCs is much less peptide receptive due to increased CLIP retention.

Enhanced MHC-II-CLIP expression in cDCs decreases antigen presentation to CD4 T cells and skews T cell responses toward Th2

To determine the functional impact of MHC-II-CLIP accumulation, we next assessed the APC function of M98A Ii cDCs both in vivo and ex vivo. The proliferation of adoptively transferred OT-II T cells was dramatically reduced in M98A Ii mice immunized with OVA as compared to WT Ii mice (Figure 4D). Furthermore, cDCs from M98A Ii mice had reduced ability to stimulate OVA-specific CD4 T cells ex vivo in comparison to WT Ii cDCs when incubated with either full-length OVA protein or with OVA_(323–339) peptide (Figure 4E), demonstrating that antigen presentation of intact antigen and of pre-processed antigenic peptides was reduced when cDCs expressed enhanced MHC-II-CLIP.

Generally, CD4 T cells polarized toward the Th1 phenotype support anti-tumor immunity, whereas Th2 polarization is protumorigenic.³³ It has been suggested that upregulation of CLIP during antigen presentation shifts CD4 T cell polarization toward Th2.^34–36 We therefore measured the production of a signature Th2 cytokine, IL-4, by proliferating OT-II T cells primed by either WT Ii or M98A Ii cDCs. Although M98A Ii cDCs induced less CD4 T cell proliferation, those T cells that did proliferate expressed significantly more IL-4 as compared to T cells primed by WT Ii cDCs (Figure 4F), showing that higher levels of surface MHC-II-CLIP on cDCs skews CD4 T cells toward the pro-tumorigenic Th2 phenotype.

High MHC-II-CLIP expression on cDCs diminishes anti-tumor responses

Effective elimination of tumors requires CD4 T helper cells to prime antigen-specific anti-tumor CD8 T cells.^4,33 To determine whether decreased antigen presentation to CD4 T cells due to increased MHC-II-CLIP on cDCs can affect tumor growth, we inoculated WT Ii or M98A Ii mice with OVA-expressing B16F10 melanoma cells. S.c. B16F10-OVA tumors grew more rapidly in M98A Ii mice in comparison to WT Ii mice (Figure 5A), demonstrating that accumulation of MHC-II-CLIP in cDCs diminishes anti-tumor responses. Even though OVA-bearing tumors were larger in M98A Ii mice relative to WT Ii mice, suggesting higher antigen expression, M98A Ii cDCs were poor stimulators of OVA-specific CD4 T cell proliferation ex vivo and were unable to stimulate OVA-specific CD4 T cells to secrete IL-2 (Figure 5B). In addition to s.c. tumor implantation, we also monitored the growth of intravenously injected B16F10 cells in a model of metastatic melanoma. The anti-tumor response in the lungs of M98A Ii mice was diminished as compared to control mice, leading to a far greater tumor burden in those mice as compared to WT Ii mice (Figure 5C). Taken together, our data show that enhanced expression of MHC-II-CLIP complexes in cDC2s, whether as a consequence of exposure to tumor-derived factors in tumor-bearing mice or as a consequence of point mutation in the Ii sequence in M98A Ii mice, suppresses antigen presentation by DCs, diminishes anti-tumor T cell responses, skews T helper differentiation toward the pro-tumorigenic Th2 type, and, consequently, enhances tumor growth.

Figure 5. Accumulation of MHC-II-CLIP in DCs diminishes anti-tumor responses.

(A) S.c. B16F10-OVA tumor growth in WT Ii and M98A Ii mice was measured twice a week. Six mice of each strain were injected in each experiment, and the data shown are representative of three independent experiments. The statistical significance of differences in tumor growth between WT Ii and M98A Ii mice at each time point is indicated.

(B) Proliferation of OT-II T cells cultured with DCs purified from tdLNs of B16F10-OVA tumor-bearing WT Ii or M98A Ii mice was measured by flow cytometry. In parallel experiments, IL-2 released into the culture medium by proliferating T cells was measured by ELISA in triplicate. The data shown are representative of 3 independent experiments.

(C) Lung B16F10 melanoma metastasis was induced in WT Ii and M98A Ii mice, and metastatic lung foci were counted on the lung sections using the multipoint tool in the ImageJ software. Four mice of each strain were used in each experiment. The data shown are representative of 3 independent experiments. Representative images of H&E-stained lung sections for each mouse strain are shown.

Data are shown as mean ± SEM; *p < 0.05, **p < 0.01.

DISCUSSION

Anti-tumor T cells are primed in tdLNs by DCs that have acquired antigens from the tumor site and present them as peptide-MHC-I or MHC-II surface complexes,^27,37 and for this reason studies directed toward identifying and correcting DC dysfunction in cancer are of high importance. The crucial role of cDC1s in cross-presentation of tumor antigens through MHC-I to prime tumor-specific CD8 T cells has been well established.³⁸ It is becoming increasingly clear that CD4 T cell help is instrumental to mount effective anti-tumor immune responses,^4,39,40 and therefore recent studies have focused on cDC2s, which are more efficient in MHC-II antigen presentation. They are essential for generating CD4 T cell responses,^41,42 and recent reports show that defects in cDC2 migration and differentiation impair anti-tumor immunity.^41–43 Here, we identify a novel defect in the MHC-II antigen presentation pathway that mostly impacts cDC2s and leads to impaired anti-tumor CD4 T cell responses.

Efficient antigen presentation requires the removal of the MHC-II peptide binding-groove placeholder CLIP by H2-M before antigenic peptides can bind to MHC-II molecules.^17,44 We now show that exposure to the tumor microenvironment reduces expression of MHC-II and the peptide editor H2-M in cDCs while simultaneously increasing the activity of the CLIP-generating endosomal proteinase cathepsin S. Together, these changes lead to enhanced expression of MHC-II-CLIP complexes and reduced ability of MHC-II to bind antigenic peptides. We estimated that, at steady state, at least 13% of MHC-II on the surface of cDC2s was occupied with CLIP, a value that increases to nearly 30% on cDC2s in tdLNs. Our estimate of surface MHC-II-CLIP occupancy in mouse cDCs agrees with a previous study in human B cells demonstrating that 19% of HLA-DR1 molecules are occupied by CLIP.⁴⁵ Curiously, cDC1s possess proportionally less MHC-II-CLIP than cDC2s. Our finding that MHC-II-CLIP dysregulation was more dramatic in cDC2s than in cDC1s highlights the importance of this cell subset for CD4 T cell priming. As shown in our report, the net effect of inefficient CLIP peptide removal is diminished tumor antigen presentation by MHC-II on DCs present in the tdLNs that leads to enhanced tumor growth.

In addition to decreased expression of H2-M, we observed increased activity of cathepsin S as well as reduced expression of its endogenous inhibitor cystatin C in tdLN cDC2s. Expression of cystatin C is suppressed by IL-6 signaling,⁴⁶ and tumor-associated DCs upregulate the genes controlling IL-6 signaling,⁴⁷ potentially providing a link between IL-6 signaling, cystatin C expression, and cathepsin S activity in tdLN cDCs. Since cathepsin S is the proteinase responsible for converting the Ii p10 fragment to CLIP,¹³ enhanced activity of cathepsin S increases the generation of MHC-II-CLIP complexes. It is important to note that enhanced cathepsin S proteinase activity could also potentially degrade antigenic peptides in antigen-processing compartments,⁴⁸ thereby reducing the pool of available tumor peptides. Curiously, myeloid cells in the tumor microenvironment express large amounts of both cell-associated and secreted cathepsin S,⁴⁹ and cathepsin S contributes to tumor growth by switching myeloid cell polarization from an anti-tumor M1 type to a tolerogenic M2-type response.⁵⁰ It is therefore likely that dysregulation of cathepsin S activity in cancer is detrimental not only by affecting MHC-II-CLIP expression in tdLN cDC2 (as shown in this report) but also by skewing myeloid cell polarization toward a pro-tumorigenic phenotype. Cancer treatments using cathepsin S inhibitors have been proposed,⁵¹ and it is an intriguing possibility that this therapy could potentially suppress tumor growth by specifically targeting cathepsin S in cDC2s to suppress MHC-II-CLIP accumulation.

In patients with acute myeloid leukemia, expression of large amounts of MHC-II-CLIP on the surface of myeloid leukemic blasts correlates with poor clinical outcome.⁵² Direct demonstration of a role of enhanced MHC-II-CLIP expression in promoting tumor growth came from our studies using M98A Ii mice with increased affinity of the CLIP peptide for MHC-II.³² As we observed in tdLNs, cDCs that possess large amounts of MHC-II-CLIP have a reduced ability to effectively bind antigenic peptides and efficiently activate CD4 T cells. Because of this, solid tumors transplanted s.c. as well as intravenously administered melanoma cells grow more rapidly in M98A Ii mice than they do in WT Ii mice.

Tumors have been shown to suppress the activity of the MHC-II transcriptional regulator CIITA in host APCs,^53,54 and it has been assumed that reduced expression of MHC-II protein is responsible for reduced anti-tumor CD4 T cell responses. However, H2-M is also a CIITA target gene,⁵⁵ and our data show that it is not only reduced MHC-II expression but also reduced expression of H2-M (together with enhanced activity of cathepsin S) that provides a one-two punch resulting in the accumulation of MHC-II-CLIP complexes, therefore reducing the pool of MHC-II available for antigenic peptides.

In addition to MHC-II-CLIP accumulation, we observed poor CD4 T helper cell priming and increased Th2 polarization in M98A Ii mice. Skewing toward Th2 promotes tumor growth and is affected by the limiting dose of antigen that stimulates a given T cell.^56–58 It is therefore possible that diminished antigen presentation by cDCs that retain more CLIP in the MHC-II peptide-binding groove leads to excessive Th2 responses that foster a pro-tumorigenic environment in the tdLN. It has been shown that IL-4, a signature cytokine produced by Th2 cells, induces cathepsin activity in tumor-associated macrophages.⁴⁹ It is therefore likely that the increase in Th2 that we observe as a consequence of MHC-II-CLIP accumulation in cDC2s also contributes to increased cathepsin S activity, further elevating MHC-II-CLIP levels.

Tumor cells developed many different mechanisms designed to evade detection by cells of the immune system. Some of these include dysregulation of MHC class I transcription and protein expression,^59,60 downregulation of costimulatory molecule expression,^61,62 and reduced expression of the MHC-II transactivator CIITA.⁵⁴ Our study highlights a previously unrecognized mechanism of immune evasion; namely, disrupting anti-tumor CD4 T cell responses by reducing the amount of peptide-receptive MHC-II on cDC2s through inefficient removal of the placeholder CLIP peptide from MHC-II. The ability of tumor cells to promote a microenvironment that enhances MHC-II-CLIP expression represents yet another mechanism by which tumor cells have modified the MHC antigen presentation pathway to avoid recognition by the acquired immune system.

Limitations of the study

While our study clearly shows that increased expression of MHC-II-CLIP in murine tdLN DCs suppresses anti-tumor immune responses, it remains to be determined whether these changes are significant in cancer in humans. Additionally, given that different MHC-II alleles have different affinities for CLIP, whether MHC-II alleles with low affinity for CLIP also accumulate MHC-II complexes has not been investigated in this study. These issues will need to be addressed to assess the importance of this mechanism of immune evasion by tumor cells on different genetic backgrounds.

RESOURCE AVAILABILITY

Lead contact

Requests for further information, resources, and reagents should be directed to and will be fulfilled by the lead contact, Paul Roche (Paul.Roche@nih.gov).

Materials availability

All reagents generated in this study are available from the lead contact upon request.

Data and code availability

• Data reported in this paper will be shared by the lead contact upon request.

• This paper does not report original code.

• Any additional information required to reanalyze the data reported in this paper is available from the lead contact upon request.

STAR★METHODS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Animals

C57BL/6Crl mice were purchased from Charles River. OT-II and OT-I mice were purchased from the Jackson Laboratories. Male eight-to twelve-week-old male mice were used for all in vivo experiments. For all in vitro experiments age- and sex-matched mice were used. Mice with the M98A point mutation on the H-2^b background were generated by crossing previously described by Ito et al.³² NOD WT mice (to generate WT Ii strain) or NOD M98A mice (to generate M98A Ii strain) to the NOD.B10Sn-H2bIJ congenic strain, backcrossed at least 5 times, and maintained as homozygotes. For genotyping purposes, PCR was performed with a custom primer pair spanning the mutation site in the Cd74 gene: 5′-AGGCCACCACTGCTTACTTC-3′ and 5′-TTTCCTTCCTGCCGCCTTCCTTAC-3′, using C1000 Thermal Cycler (Biorad) with the following conditions: 95°C 10 min, followed by 40 cycles of 98°C 10 s, 64°C 30 s, and 72°C 30 s, followed by a final 5 min extension at 72°C. The PCR product was submitted to Genewiz (Azenta Life Sciences) for purification and Sanger sequencing. Mice were bred and maintained in-house at the National Cancer Institute at Frederick animal facility. Animals were housed with littermates, up to 5 total mice/cage and given ad libitum access to a standard diet and water.

Cell lines and cell culture

The B16F10, B16F10-OVA melanoma (generous gift from Dr. Edith Lord, University of Rochester),⁶³ MC38 colorectal adenocarcinoma and MB49 bladder carcinoma cell lines (generous gifts from Dr. Tim Greten, National Cancer Institute) were cultured in 5% CO₂ in complete DMEM media containing 10% fetal bovine serum, 2 mM L-glutamine, 1% penicillin/streptavidin. All cell lines tested negative for mycoplasma contamination. For co-culture experiments, purified DCs (5 × 10⁴ cells) were cultured with 5 × 10⁵ OVA-specific CD4 or CD8 T cells in RPMI-1640 media (Gibco) supplemented with 10% FBS, 2 mM L-glutamine, 100 μM β-mercaptoethanol, 25 mM HEPES and 50 μg/mL gentamicin.

Study approval

All mice were cared for in accordance with National Institutes of Health guidelines with the approval of the National Cancer Institute Animal Care and Use Committee under protocol numbers 23–445, 23–447, EIB-076, and EIB-098.

METHOD DETAILS

Tumor models

Tumor cells were split 24 h before inoculation, and 0.5 × 10⁶ cells in PBS were injected subcutaneously into one flank of mice using a 27G × ½ in needle. Tumor growth was monitored for 2–3 weeks and measured using calipers. Mice were euthanized when the tumors reached the size limit of 1 cm³. For lung metastasis assays, 3 × 10⁵ B16F10 cells were injected intravenously and after 10 days the mice were euthanized. Lungs were perfused and fixed in 10% formalin for 48 h, preserved in 70% ethanol, and embedded in paraffin. The paraffin blocks were cut into 10 μm longitudinal sections and stained with hematoxylin and eosin (H&E) by Histoserv Inc. (Germantown, MD). Images of the lung sections were acquired by microscopy and the metastatic foci were counted using the multi-point tool in ImageJ software.

Tissue digestion and flow cytometry staining

Spleens and LNs were dissected and digested for 30 min at 37°C in RPMI containing collagenase D (2.5 mg/mL) and DNase I (1 mg/mL) (Millipore Sigma). The cells were washed and resuspended in PBS containing 0.5% BSA and 2mM EDTA, filtered through a 70 μM nylon mesh, and counted before staining for flow cytometry. For surface staining, up to 2 × 10⁶ cells were stained at 4°C with Fc receptor blocking mAb 2.4G2 (Leinco) and fluorochrome-conjugated primary mAb or isotype-control Ab in 50 μL FACS buffer (PBS containing 0.5% BSA and 2 mM EDTA). Propidium iodide (MilliporeSigma) was used to access cell viability. For intracellular staining, cells were stained with either Live/Dead Fixable Dead Cell Stain reagent (ThermoFisher) or Ghost Dye Violet 510 (Cytek Biosciences), followed by staining with mAbs for surface molecules. Cells were then fixed and permeabilized using BD Cytofix/Cytoperm kit (BD Biosciences) according to the manufacturer’s instructions, stained with mAb for intracellular molecules, and washed extensively with permeabilization buffer. Flow cytometric analyses were performed with LSR Fortessa (BD Biosciences), and data were analyzed using FlowJo_v10.9.0 software (BD Biosciences). To estimate MHC-II occupancy by CLIP, streptavidin-AlexaFluor594 binding to DCs stained with saturating amounts of biotinylated primary mAb was measured by FACS. Anti-MHC-II I-A^b-CLIP mAb (clone 15G4) and pan anti-pMHC-II mAb (clone Y3P) were biotinylated using an EZ-Link Sulfo-NHS-LC-Biotinylation Kit (ThermoFisher) using 20-fold biotin molar excess, according to manufacturer instructions. Both mAb were biotinylated with similar efficiency. The MFI of samples labeled with MHC-II-CLIP mAb was determined and expressed as the percentage of total peptide-MHC-II. Cell sorting was performed using a BD FACSymphony S6 Cell Sorter. To sort the cDC1 and cDC2 subsets, ndLN or tdLN from 20 from tumor-bearing mice were pooled and a single cell suspension was prepared as described above. DCs were first enriched using a PanDC Isolation Kit (Miltenyi Biotec), followed by staining with anti-B220, anti-CD11c, anti-CD26, anti-XCR1, and anti-CD172a antibodies. The cDC1 (B220⁻CD11c⁺CD26⁺XCR1⁺CD172a⁻) and cDC2 (B220⁻CD11c⁺CD26⁺XCR1⁻CD172a⁺) subsets were sorted by gating on the respective populations. For co-culture experiments, DCs were enriched from LN using a PanDC isolation kit (Miltenyi Biotec) and the T cells were isolated using a Mouse CD4 T cell isolation kit or Mouse CD8 T cell isolation kits (Miltenyi Biotec). T cells were labeled with either 1 μM CFSE (ThermoFisher) or 0.5 μM Cell Trace Violet (ThermoFisher) according to manufacturer’s instructions and T cell proliferation was analyzed by flow cytometry.

T cell proliferation assays

CFSE-labeled T cells (2 × 10⁶ cells/mouse) were injected into the tail vein of healthy or tumor-bearing mice. 24 h after T cell transfer, recipient mice were immunized by intraperitoneal injection of 1:1 OVA/alum (200 μg OVA protein/mouse), and control mice received 1:1 PBS/alum. After 72 h T cell proliferation in LN or spleen was analyzed by flow cytometry. Purified DCs (5 × 10⁴ cells) were cultured with 5 × 10⁵ OVA-specific CD4 or CD8 T cells in the presence or absence of full-length OVA protein (Worthington), OVA_(257–264) peptide (GenScript), or OVA_(326–339) peptide (InvivoGen) in RPMI-1640 media (Gibco) supplemented with 10% FBS, 2 mM L-glutamine, 100 μM β-mercaptoethanol, 25 mM HEPES and 50 μg/mL gentamicin. Culture supernatants were collected after 24 h and IL-2 and IFNγ levels were measured by ELISA (R&D). DCs isolated from tdLN of B16F10-OVA-tumor-bearing mice were co-cultured with CFSE-labeled OT-II CD4 T cells and T cell proliferation was assessed by flow cytometry as well as by measuring IL-2 levels in co-culture supernatant after 72 h. For Th2 skewing, DCs isolated from WT Ii or M98A Ii mice were co-cultured with CTV-labeled OT-II T cells in the presence of OVA protein (100 μg/mL), rhIL-2 (200 U/ml), rmIL-4 (10 ng/mL) and anti-IFNγ mAb. Three days later, T cells were re-stimulated with 50 ng/mL phorbol 12-myristate 13-acetate (PMA) (MilliporeSigma) and 1μM ionomycin (MilliporeSigma) for 2 h to stimulate the production of cytokines, followed by additional 2-h stimulation in the presence of 1 μL/mL protein transport inhibitor GolgiStop (containing Monensin) (BD Biosciences). After the stimulation, T cell proliferation and intracellular levels of IL-4 were analyzed by FACS.

DC migration assay

DCs isolated from healthy C57BL/6Crl donor mice were labeled with green fluorescence PKH67 cell linker (Millipore Sigma) according to manufacturer’s instructions. 1 × 10⁶ DCs were injected into recipient tumor-bearing mice intratumorally as well as subcutaneously on the contralateral non-tumor flank in the same mouse as a control. Mice were sacrificed 3 days later and both tdLN and control ndLN were analyzed by flow cytometry.

Cathepsin S activity assay

FACS-sorted cDC1 and cDC2 from tdLN and ndLN were lysed on ice for 10 min and centrifuged for 5 min at 1200 rpm at 4°C to remove cell debris. Cathepsin S activity at pH 6.0 was measured using a Cathepsin S Activity kit (Millipore Sigma) according to the manufacturer’s instructions. Briefly, 50 μL of each cell lysate, with or without the addition of cathepsin S inhibitor, was mixed with 50 μL of reaction buffer and 2 μL of cathepsin S substrate labeled with amino-4-trifluoromethyl coumarin (AFC). After incubation for 2 h at 37°C the fluorescence was measured at λ_Ex = 400nm/λ_Em = 505nm using SpectraMax iD3 plate reader (Molecular Devices).

Immunoblotting

Equal numbers of FACS-sorted cDC1 or cDC2 from tdLN and ndLN were lysed on ice for 30 min in 10 mM Tris/150 mM NaCl buffer pH 7.5 containing 1% Triton X-100, 100 μg/mL BSA, and protease inhibitor cocktail (ThermoFisher). Cell lysates were mixed with 2x Laemmli sample buffer (Biorad), heated at 90°C for 5 min, and analyzed by SDS-PAGE using the Mini-PROTEAN Tetra cell system (Biorad). Proteins were transferred to PVDF membranes using the Trans-Blot Turbo Transfer system (Biorad). Membranes were incubated with 5% nonfat dry milk in PBS and incubated with appropriate primary antibodies, HRP-conjugated secondary antibodies, and ECL reagent (PerkinElmer). Detection was performed using the KwikQuant Imager (Kindle Biosciences) and analyzed and quantified with KwikQuant Image analyzer v.5.9 (Kindle Biosciences).

Peptide loading assay

Conjugates of Eα peptide and maleimide activated OVA (ThermoFisher) were prepared according to manufacturer instructions. Briefly, 2 mg of Imject maleimide-activated OVA was mixed with 2 mg of N-terminal cysteine-modified Eα_(39–68) peptide in conjugation buffer and incubated for 2 h at room temperature. The conjugates were then desalted by centrifugation using Zeba Spin Desalting Columns, 7K MWCO (ThermoFisher). For in vivo assays, 50 μg of OVA-Eα conjugate was injected into tumor-bearing mice both i.t. and s.c. on the contralateral flank. Control tumor-bearing mice were injected with PBS. For in vitro assays, purified DCs were incubated for 2 h with 100 μg/mL OVA-Eα conjugate, 2 mM Eα _(52–68) peptide, or 2 mM OVA_(257–264) (SIINFEKL) peptide, either on ice or at 37 °C, washed twice, and stained for flow cytometric analysis. Surface complexes of MHC-II I-A^b-Eα_(52–68) were detected using specific mAb (clone YAe) (ThermoFisher) and MHC-I H2^b-OVA_(257–264) complexes were detected using specific mAb (clone 25-D1.16) (ThermoFisher) and quantitated by flow cytometry. The MFI values for each sample incubated on ice was subtracted from the MFI for the sample incubated at 37 °C and the values shown are net MFI.

QUANTIFICATION AND STATISTICAL ANALYSIS

Unless specifically noted, all data are representative of >3 separate experiments. Experimental group assignment was determined by genotype or, if all wild-type mice, by random designation. All statistical analyses were performed using GraphPad Prism v8 Software (Dotmatics). Error bars represent mean ± standard error of the mean (SEM) calculated using Prism; dots represent individual measurements. Statistical differences between the groups were determined using the Student’s unpaired or pair t test when applicable and p values < 0.05 were considered statistically significant. The statistical significance of differences between tumor growth was assessed by Multiple t-tests (one unpaired t test per row). The exact value of n and what it represents is indicated in each figure legend. The tumor measurements were done blinded to group assignment by the technical staff.

KEY RESOURCES TABLE.

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Rat anti-mouse CD45R/B220 PE (clone RA3–6B2)	BD Biosciences	Cat # 553090; RRID: AB_394620
Rat anti-mouse CD45R/B220 FITC (clone RA3–6B2)	ThermoFisher	Cat # 11-0452-82; RRID: AB_465054
Rat anti-mouse CD45R/B220 Alexa Fluor 594 (clone RA3-6B2)	Biolegend	Cat # 103254; RRID: AB_2563229
Rat anti-mouse CD45 BrilliantViolet510 (clone 30-F11)	Biolegend	Cat # 103138; RRID: AB_2563061
Armenian hamster anti-mouse CD11c PE Cyanine7 (clone N418)	ThermoFisher	Cat # 25-0114-82; RRID: AB_469590
Rat anti-mouse CD26 PE (clone H194-112)	Biolegend	Cat # 137804; RRID: AB_2293047
Rat anti-mouse MHC-I (H-2Kb) FITC (clone AF6-88.5.5.3)	ThermoFisher	Cat # 11-5958-82; RRID: AB_11149502
Rat anti-mouse MHC-II (I-A/I-E) eFluor450 (clone M5/114.15.2)	ThermoFisher	Cat # 48-5321-82; RRID: AB_1272204
Armenian hamster anti-mouse CD40 APC (clone 3/23)	Biolegend	Cat # 124612; RRID: AB_1134072
Mouse anti-mouse XCR1 FITC (clone ZET)	Biolegend	Cat # 148210; RRID: AB_2564366
Rat anti-mouse CD172a (SIRP alpha) APC (clone P84)	ThermoFisher	Cat # 17-1721-82
Rat anti-mouse CD172a (SIRP alpha) Alexa Fluor 594 (clone P84)	Biolegend	Cat # 144020; RRID: AB_2629588
Mouse anti-mouse MHC-II-CLIP biotin (clone 15G4)	Lisa Denzin, Rutgers	N/A
Mouse anti-mouse MHC-II-Eα(52-68) FITC (clone YAe)	ThermoFisher	Cat # 11-5741-82; RRID: AB_996692
Mouse anti-mouse MHC-I-OVA(257-264) APC (clone 25-D1.16)	ThermoFisher	Cat # 17-5743-82; RRID: AB_1311286
Rat anti-mouse CD4 PE Cyanine7 (clone GK1.5)	ThermoFisher	Cat # 25-0041-82; RRID: AB_469576
Mouse anti-mouse CD45.1 APC (clone A20)	ThermoFisher	Cat # 17-0453-82; RRID: AB_469398
Rat anti-mouse H2-M Alexa Fluor 647 (clone 2C3A)	Lisa Denzin, Rutgers	N/A
Rat anti-mouse IL-4 APC (clone 11B11)	ThermoFisher	Cat # 17-7041-82; RRID: AB_469494
Rat anti-mouse IFNγ PE (clone XMG1.2)	ThermoFisher	Cat # 12-7311-82; RRID: AB_466193
Rat anti-mouse CD197 (CCR7) eFluor450 (clone 4B12)	ThermoFisher	Cat # 48-1971-82; RRID: AB_1944351
Rat anti-Mouse CD32/CD16 purified (clone 2.4G2)	Leinco Technologies	Cat #C381; RRID: AB_2737484
Rabbit anti-mouse cystatin C purified (clone JJ09-16)	ThermoFisher	Cat # MA5-32475; RRID: AB_2809752
Anti-rabbit IgG (Rabbit TrueBlot) HRP (clone eB182)	Rockland	Cat # 18-8816-31; RRID: AB_469528
Mouse anti-mouse beta Actin HRP (clone BA3R)	ThermoFisher	Cat # MA5-15739-HRP; RRID: AB_2537667
Biological samples
Mouse tissue samples (LN, spleen, lung, tumor)	National Cancer Institute	IACUC: 23-445; 23-447; EIB-076; EIB-098
Chemicals, peptides, and recombinant proteins
Collagenase D	MilliporeSigma	Cat # 11088866001
DNase I	MilliporeSigma	Cat # 10104159001
Streptavidin Alexa Fluor 594	ThermoFisher	Cat #S11227
Streptavidin Alexa Fluor 647	ThermoFisher	Cat #S32357
Live/dead fixable viability dye UV450	ThermoFisher	Cat #L34961
Ghost Dye fixable viability dye Violet510	Cytek	Cat # SKU13-0870-T100
propidium iodide	MilliporeSigma	Cat #P4170-25MG
Ovalbumin	Worthington	Cat # CAS:9006-59-1
OVA peptide (323-339)	InvivoGen	Cat # vac-isq
OVA peptide (257-264)	GenScript	Cat # RP10611
PMA	MilliporeSigma	Cat #P8139-5MG
lonomycin	MilliporeSigma	Cat # 407950-5MG
BD GolgiStop	BD Biosciences	Cat # 554724
Halt protease inhibitor cocktail (100x)	ThermoFisher	Cat #78438
Eα(39-68) peptide (ASFEAQGALANIAVDKA)	GenScript	Order ID: 430169-2
Extended Eα peptide (CTIWRLEEFAKF-Eα peptide)	GenScript	Order ID: 630195-1
Western Lightning Plus-ECL reagent	Perkin Elmer	Cat # NEL105001EA
Imject Maleimide-Activated Ovalbumin	ThermoFisher	Cat # 77126
Imject Alum	ThermoFisher	Cat # 77161
Critical commercial assays
BD Cytofix/Cytoperm	BD Biosciences	Cat # 554714
PanDC Isolation kit	Miltenyi Biotec	Cat # 130-100-875
Mouse CD4 T cell isolation kit	Miltenyi Biotec	Cat # 130-104-454
Mouse CD8α T cell isolation kit	Miltenyi Biotec	Cat # 130-104-075
Experimental models: Cell lines
B16F10	Edith Lord, Rochester	Brown et al.63
B16F10-OVA	Edith Lord, Rochester	Brown et al.63
MB49	Tim Greten, NCI	N/A
MC38	Tim Greten, NCI	N/A
Experimental models: Organisms/strains
Mouse: C57BL/6Crl	Charles River	Strain #: 027
Mouse: OT-II (B6.Cg-Tg(TcraTcrb)425Cbn/J)	The Jackson Laboratory	Strain #: 004194
Mouse: OT-I (C57BL/6-Tg(TcraTcrb)1100Mjb/J)	The Jackson Laboratory	Strain #: 003831
Mouse: NOD WT/WT	Kai Wucherpfennig, DFCI	Ito et al.32
Mouse: NOD M98A/M98A	Kai Wucherpfennig, DFCI	Ito et al.32
Mouse: NOD.B10Sn-H2bIJ	The Jackson Laboratory	Strain #: 002591
Mouse: WT Ii	This paper	N/A
Mouse: M98A Ii	This paper	N/A
Software and algorithms
FlowJo	Becton Dickinson	https://www.flowjo.com/
GraphPad Prism	Dotmatics	https://www.graphpad.com/
ImageJ	ImageJ	https://imagej.net/ij/

Highlights.

• Tumors promote accumulation of MHC-II-CLIP in tumor-draining lymph node cDC2s

• cDC2s in tumor-draining lymph nodes are non-receptive to peptides

• Elevated MHC-II-CLIP expression in lymph node cDC2s impairs anti-tumor responses