VPS33B regulates MHC class II antigen presentation in dendritic cells to drive CD4 T cell immunity Authors: Kiana Kawarizadeh, Charles N Vallez, Kathrynne A Warrick, Satomi Ando, Margaret M McDaniel, Koichi Araki, Chandrashekhar Pasare Published: 2026 ABSTRACT SummaryVacuolar protein sorting-associated protein (VPS)33B is essential for endosomal-lysosomal fusion, with its dysfunction linked to arthrogryposis, renal dysfunction, and cholestasis (ARC) syndrome. VPS33B has been implicated in mediating Toll-like receptor degradation in macrophages, a critical step to mitigate inflammation. Here, we uncover a pivotal function of VPS33B in dendritic cell (DC)-mediated antigen presentation. Loss of VPS33B in DCs impairs peptide-MHC class II presentation, leading to diminished CD4 T cell responses. Absence of VPS33B selectively compromises CD4-dependent CD8 T cell responses while not impacting CD4-independent CD8 T cell priming, indicating that VPS33B only affects the MHC class II pathway. Mechanistically, we find that VPS33B is required for optimal presentation of MHC class II-associated peptides as VPS33B-deficient DCs predominantly express MHC class II loaded with invariant chain degradation intermediates. Our work highlights VPS33B as a key regulator of antigen presentation and underscores its importance in orchestrating CD4 T cell-mediated immunity.Subject areas: Immunity Introduction The innate immune system is critical to host survival and performs a vast array of functions. Chief among these is the ability to detect and respond to microbial threats. Sensing of microbes by the innate immune system occurs largely through germ-line encoded pattern-recognition receptors (PRRs).1 PRRs, including Toll-like receptors (TLRs), C-type lectin receptors, and RIG-I-like receptors recognize conserved microbial components collectively defined as pathogen-associated molecular patterns. Activation of PRRs facilitates a signaling cascade that results in activation of the sensing cell and production of soluble mediators to facilitate clearance of the microbial threat. A critical event downstream of PRR engagement is the production of pro-inflammatory cytokines, including TNF, IL-6, and IL-1β.2,3,4 These cytokines facilitate a plethora of secondary responses (neutrophil recruitment, acute phase reactant production, and leukopoiesis) that are pivotal in controlling the infection.5,6 Despite the importance of cytokines, overproduction can result in deleterious outcomes for the host as seen during septic shock. As such, mitigating PRR signaling after initial activation is critical to host survival, ensuring early infection control without the damaging effects of hyperinflammation. A key mechanism in termination of PRR signaling is endocytosis and subsequent degradation of receptors, effectively terminating intracellular signaling events.7,8 Vacuolar protein sorting-associated protein (VPS) members were originally discovered in yeast for their role in vacuolar protein sorting, which is the process of transporting proteins to the vacuole, the yeast equivalent of the mammalian lysosome.9 Further work has demonstrated the conserved importance of VPS homologs in endosomal trafficking and cellular homeostasis across several mammalian cell types.10,11,12 These proteins are critical for their intracellular sorting and transportation of soluble proteins, and disruption of VPS members has been previously linked to specific lysosomal storage disorders.13 As a core component of the CORVET (class C core vacuole/endosome tethering) and HOPS (homotypic fusion and vacuole protein sorting) complexes, VPS33A is mainly involved in vesicle-mediated protein trafficking to lysosomes and phagosomes.14 Despite their sequence similarity, VPS33B does not integrate with the CORVET or HOPS complex and instead forms a complex with VPS16B (VIPAS39).15 This process occurs through direct interactions with SNARE proteins, thus facilitating membrane and organelle fusion.16,17,18 Interestingly, mutations in human VPS33B or VPS16B, are known to drive arthrogryposis, renal dysfunction, and cholestasis (ARC) syndrome.19,20 In addition to multisystem organ failure, ARC syndrome is characterized by persistent infections and sepsis leading to mortality within the first year of life.21 As such, the clinical implications for better understanding the role of VPS33B in the pathophysiology of these disease processes are evident. Trafficking of cellular components through the endosomal-lysosomal system plays a key role in several other processes beyond termination of TLR signaling, including cellular metabolism, adhesion, waste removal, and immunity.22,23,24 A cardinal function of the endosomal-lysosomal system is antigen presentation; wherein endogenous and/or exogenous proteins are degraded into peptide fragments via proteases and loaded onto major histocompatibility (MHC) complexes for surface presentation. Antigen presentation is a pivotal step in functional immunity as it facilitates the transition from a broad innate response into a targeted adaptive response through activation of antigen specific naive T cells. In this process, extracellular proteins are taken up by antigen-presenting cells (APCs) through the process of phagocytosis and endocytosis, leading to the formation of phagosomes and endosomes, respectively.25,26 These compartments then undergo a stepwise maturation process where they acquire a lower pH that facilitates activity of proteases necessary for proteolytic cleavage of antigens.27,28 In APCs, recycled and newly synthesized MHC II molecules accumulate in late endosomes and lysosomes, forming a specialized structure known as the MHC II compartment (MIIC).29 It is within the MIIC that a majority of peptide loading onto MHC class II molecules occurs. This is a tightly regulated process that involves the cooperation of several proteins, ultimately leading to removal of class II-associated invariant chain peptide (CLIP), or other invariant chain degradation intermediates (i.e., small leupeptin-induced proteins [SLIPs]), in exchange for exogenous antigen to be presented on the plasma membrane. Endosomal-lysosomal fusion thus plays a central role in antigen processing and presentation by regulating both the degradation of exogenous antigen and formation of, and trafficking to, the MIIC compartment. While several cell types within the mammalian immune system are capable of this function, DCs as professional APCs play a pivotal role in priming naive T cells. DCs are uniquely able to facilitate adaptive responses for several reasons: (1) DCs express many PRRs, allowing them to respond to diverse pathogenic threats;30,31 (2) DCs are highly phagocytic specifically in immature states;32 and (3) DCs can migrate from peripheral tissues to secondary lymphoid organs where they interact with and activate naive T cells.25,33 The ability to process and present exogenous antigens on MHC II, as well as MHC class I, molecules is a particularly important function of DCs and integral to their critical function of linking innate and adaptive immune systems. Here, we investigated how loss of DC-intrinsic VPS33B would alter antigen processing, presentation, and subsequent adaptive immune responses. We found that loss of DC-intrinsic VPS33B resulted in significant impairment of CD4 T cell responses to both immunization and viral infection. Interestingly, we observed no appreciable differences in CD8 T cell responses during viral infection, but CD8 T cell responses were compromised following immunization highlighting the critical role of VPS33B-dependent priming of CD4 T cell responses. Mechanistically, we found that VPS33B-deficient DCs were unable to load MHC class II molecules with peptides derived from exogenous antigens. Taken together, these results suggest that VPS33B plays a critical role in presentation of exogenous antigens on MHC class II, but not MHC class I, molecules and this study offers important insights into the pathophysiology of VPS33B-associated conditions. Results VPS33B-deficient DCs are defective in their ability to prime CD4 T cells in vitro Given the previously established role in endosomal-lysosomal trafficking of microbially associated cargo,8 we hypothesized that VPS33B is required for MHC class II-restricted presentation of exogenous proteins. To assess if loss of VPS33B in DCs affects CD4 T cell priming, we cultured CD11c+ splenic DCs (sDCs) from either mice expressing floxed allele of VPS33B (VPS33Bfl/fL or VPS33BWT) (Figure S1A) or from mice lacking VPS33B in the myeloid compartment (VPS33BΔCSF1R) with naive, ovalbumin (OVA) 323-339-specific, OT-II T cells. Successful deletion of VPS33B in splenic DCs in VPS33BΔCSF1R was confirmed by western blotting (Figure S1B). VPS33BΔCSF1R sDCs, cultured with endotoxin-free OVA protein and stimulated with the TLR4 ligand lipopolysaccharide (LPS), displayed defective priming of OT-II T cells as measured by dilution of cell trace violet (CTV) dye (Figure 1A). Consistent with impaired proliferation, OT-II T cells primed using VPS33BΔCSF1R sDCs exhibited lower CD44 expression and produced significantly less interferon-gamma (IFNγ) (Figures 1B and 1C), suggesting that both CD4 T cell priming and differentiation were negatively impacted in the absence of DC-intrinsic VPS33B. VPS33B-deficient DCs are defective in their ability to prime CD4 T cells in vitro (A) Representative flow plots of CTV-stained OT-II T cells (pre-gated on live, CD4+, CD8−) co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence or absence of OVA (10 and 100 μg/mL) and LPS (100 ng/mL) for 72 h. Graphical quantification of % proliferating (CTV-diluted) OT-II T cells to the right. (B) Representative flow plots of CD44 expression on OT-II T cells (pre-gated on live, CD4+, CD8−) co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs and stimulated with OVA (10 μg/mL) and LPS (100 ng/mL) for 72 h. Graphical quantification of % CD44+ OT-II T cells to the right. (C) IFNγ was quantified by ELISA from supernatants taken from OT-II T cells co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence of OVA (10 and 100 μg/mL) and LPS (100 ng/mL) for 72 h. (D) Representative flow plots of CTV-stained OT-I T cells (pre-gated on live, CD4−, CD8+) co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence or absence of OVA (10 and 100 μg/mL) and LPS (100 ng/mL) for 72 h. Graphical quantification of % proliferating (CTV-diluted) OT-I T cells to the right. (E) Representative flow plots of CD44 expression on OT-I T cells (pre-gated on live, CD4−, CD8+) co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs and stimulated with OVA (10 μg/mL) and LPS (100 ng/mL) for 72 h. Graphical quantification of % CD44+ OT-I T cells to the right. (F) IFNγ was quantified by ELISA from supernatants taken from OT-I T cells co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence of OVA (10 and 100 μg/mL) and LPS (100 ng/mL) for 72 h. NP = no protein (OVA); error bars shown as mean ± SEM. (A–F) n = 3–5 independent experiments. Statistical analysis was performed by two-way ANOVA. ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, n.s. = not significant. Our data demonstrate that absence of VPS33B impairs the ability of DCs to prime CD4 T cell responses. While DCs are the professional APCs of the immune system, other cell types, such as B cells and macrophages, have the ability to present peptides from exogenous antigens on MHC class II molecules. To determine the function of VPS33B in antigen presentation by other APCs, we performed priming experiments using B cells isolated from the spleens of VPS33BWT and VPS33Bfl/fL mice crossed to Vav iCre mice (VPS33BΔVav), which deletes VPS33B in all nucleated hematopoietic cells. We found that OT-II T cells primed with VPS33BWT and VPS33BΔVav B cells demonstrated similar proliferation and activation (Figures S1C and S1D), confirming a specific role for DC-intrinsic VPS33B in MHC class II antigen presentation. Additionally, to exclude a possible TLR4-specific role for VPS33B in DCs, we performed CD4 T cell priming experiments using CpG-B, a TLR9 agonist. We again found that VPS33BΔCSF1R were impaired in their ability to prime OT-II T cells, suggesting that the defect in priming ability of VPS33B deficient DCs is not restricted to just TLR4 responses. (Figures S2A–S2C). While endosomal-lysosomal fusion is critical for MHC class II peptide presentation, MHC class I-peptide presentation is more reliant on proteasomal degradation and transport of cytosolic peptides through the transporter associated with antigen processing and Tapasin.34,35 Our sDC preparation was a mix of cDC1s and cDC2s and the former are capable of cross-presenting exogenous antigens on MHC class I molecules.36 To understand if VPS33B regulates the trafficking of cargo for cross presentation, we examined the responses of OT-I T cells to DCs in the presence of OVA protein and either LPS or CpG-B. Our data show that CD8 T cells from OT-I transgenic mice co-cultured with VPS33BΔCSF1R sDC, in the presence of OVA and LPS or OVA and CpG-B, exhibited similar proliferation and activation profiles to those cultured with VPS33BWT control sDCs (Figures 1D and 1E, S2D and S2E). In line with this, we found no differences in secretion of IFNγ by CD8 T cells, whether they were primed by VPS33B-sufficient or -deficient DCs (Figures 1F and S2F). Together, our data indicate that DC-intrinsic VPS33B is required for priming of naive CD4, but not CD8 T cell responses in vitro. Mice lacking VPS33B in the myeloid compartment have defective CD4 and CD8 T cell priming in response to protein immunization To further investigate the role of VPS33B in T cell priming in vivo, VPS33BWT and VPS33BΔCSF1R mice were immunized with a formulation of OVA and LPS in incomplete Freund’s adjuvant (IFA). The endogenous OVA-specific CD4 and CD8 T cells were collected from draining lymph nodes and assessed via recall with OVA protein (Figure 2A). CD4 T cells from VPS33BΔCSF1R mice had reduced CD44 expression (Figure 2B), lower Tbet expression (Figure 2C), and produced less IFNγ (Figure 2D) upon re-stimulation with OVA, corroborating our in vitro studies. These data suggest that absence of VPS33B in DCs negatively impacted their ability to prime antigen-specific CD4 T cells in vivo. Mice lacking VPS33B in the myeloid compartment have defective CD4 and CD8 T cell priming in response to protein immunization (A) Schematic of experimental set-up for in vivo immunizations. Mice were immunized with 50 μg of OVA and 5 μg of LPS in IFA (100 μL/mouse; hock injection). Draining Ig LNs were collected, processed, and cultured with OVA (100 μg/mL) for 72 h. (B) Representative flow plots of CD44 expression on CD4 T cells (pre-gated on live, CD90.2+, CD4+) from immunized VPS33BWT and VPS33BΔCSF1R IgLNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Graphical quantification of % CD44+ OT-II T cells to the right. (C) Representative flow plots of Tbet expression in CD4 T cells (pre-gated on live, CD90.2+, CD4+) from immunized VPS33BWT and VPS33BΔCSF1R IgLNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Graphical quantification of % Tbet+ OT-II T cells to the right. (D) Representative flow plots of IFNγ expression in CD4 T cells (pre-gated on live, CD90.2+, CD4+) from immunized VPS33BWT and VPS33BΔCSF1R Ig LNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Cells were rested with IL-2 for 48 h and re-stimulated with Phorbol 12-myristate 13-acetate (PMA) + Ionomycin for 6 h. Graphical quantification of % IFNγ+ OT-II T cells to the right. (E) Representative flow plots of CD44 expression on CD8 T cells (pre-gated on live, CD90.2+, CD8+) from immunized VPS33BWT and VPS33BΔCSF1R Ig LNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Graphical quantification of % CD44+ OT-I T cells below. (F) Representative flow plots of Tbet expression in CD8 T cells (pre-gated on live, CD90.2+, CD8+) from immunized VPS33BWT and VPS33BΔCSF1R Ig LNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Graphical quantification of % Tbet+ OT-I T cells below. (G) Representative flow plots of IFNγ expression in CD8 T cells (pre-gated on live, CD90.2+, CD8+) from immunized VPS33BWT and VPS33BΔCSF1R IgLNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. Cells were rested with IL-2 for 48 h and re-stimulated with Phorbol 12-myristate 13-acetate (PMA) + Ionomycin for 6 h. Graphical quantification of % IFNγ+ OT-I T cells below. (H) Graphical quantification of % granzyme B (GrzB)+ expression in CD8 T cells (pre-gated on live, CD90.2+, CD8+) from VPS33BWT and VPS33BΔCSF1R IgLNs in the presence or absence of OVA (100 μg/mL) for 72 h. (I) IFNγ was quantified by ELISA from supernatants taken from total immunized LNs cultured in the presence or absence of OVA (100 μg/mL) for 72 h. NP = no protein (OVA); error bars shown as mean ± SEM. n = 3 biological replicates for one independent experiment. Statistical analysis was performed by two-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001, n.s. = not significant. Interestingly, unlike our findings in vitro, immunization of VPS33BΔCSF1R mice revealed a markedly reduced CD8 T cell priming and effector function (granzyme B and IFNγ production) as compared to VPS33BWT mice (Figures 2E–2I). Previous studies have established that CD8 T cell activation in vivo relies heavily on help from CD4 T cells, specifically through CD40−CD40L interactions.36,37,38 Given the observation of defective CD4 priming in VPS33BΔCSF1R mice, it is plausible that this deficiency in CD4 T cell help is a primary reason for the impaired CD8 T cell priming witnessed in vivo. Impaired CD4 and CD8 T cell priming is not due to inherent T cell-intrinsic defects in mice lacking VPS33B in DCs and other myeloid cells Tonic TCR signaling provided by DCs is critical to maintain T cell responsiveness to foreign antigens.39 Thus, it is reasonable to suspect that absence of VPS33B in DCs and other myeloid cells has an impact on the endogenous T cell repertoire and may alter the precursor frequency of OVA-specific CD4 and CD8 T cells. To address this possibility, we performed immunization experiments after adoptive transfer of CD45.1+ OT-II and OT-I T cells and tracked the expansion of these cells in the draining lymph nodes of CD45.2 recipient mice (Figure 3A). At 7 days post-immunization, levels of CD45.1+ OVA-specific CD4 and CD8 T cells in VPS33BΔCSF1R mice were nearly undetectable, suggesting a lack of initial priming and subsequent expansion (Figure 3B). Similar findings were seen upon immunization with CpG-B (Figure 3C). These findings demonstrate that DC-intrinsic VPS33B plays a critical role in in vivo priming of CD4 T cell responses. Consequently, loss of VPS33B in DCs also results in impaired CD8 T cell priming in vivo which requires CD4 T cell help. Impaired CD4 and CD8 T cell priming is not due to inherent T cell-intrinsic defects in mice lacking VPS33B in DCs and other myeloid cells (A) Schematic of experimental set-up for adoptive T cell transfer and protein immunization. CD45.1+ OT-I and OT-II T cells were isolated and transferred into VPS33BWT and VPS33BΔCSF1R mice. 24 h later, mice were immunized with 50 μg of OVA and 5 μg of LPS or 10 μg of CpG-B in IFA (100 μL/mouse; hock injection). 7 days later, IgLNs were collected and subject to flow cytometry. (B) Representative flow plots gating for transferred OT-I (live, CD90.2+, Vα2+, CD8+, CD45.1+) or transferred OT-II (live, CD90.2+, Vα2+, CD8−, CD45.1+) T cells. Pre-gated on live, CD90.2+, Vα2+ cells. Quantification of % CD45.1+, Vα2+ OT-II and OT-I cells from immunized LNs to the right. (C) Representative flow plots gating for transferred OT-I (live, CD90.2+, Vα2+, CD8+, CD45.1+) or transferred OT-II (live, CD90.2+, Vα2+, CD8−, CD45.1+) T cells. Pre-gated on live, CD90.2+, Vα2+ cells. Quantification of % CD45.1+, Vα2+ OT-I and OT-II cells from immunized LNs to the right. Error bars shown as mean ± SEM. n = 4–5 biological replicates for one independent experiment. Statistical analysis was performed by unpaired t test. ∗p < .05, ∗∗p < 0.01, ∗∗∗p < 0.0001, n.s. = not significant. DC-intrinsic VPS33B is required for CD4 T cell priming yet dispensable for the activation of CD8 T cell responses during viral infection To determine the role of DC-intrinsic VPS33B in priming CD4 and CD8 T cells against pathogens, we employed an acute lymphocytic choriomeningitis virus (LCMV) infection model (LCMV-Armstrong). Importantly, the highly inflammatory nature of LCMV-Armstrong infection has been shown to elicit CD8 T cell priming and early activation independent of CD4 help.40,41 Based on this, we hypothesized that the lack of DC-intrinsic VPS33B would impair CD4 T cell priming with no appreciable effects on early CD8 T cell responses. To test this, we sorted naive CD4 T cells and CD8 T cells from SMARTA and P14 T-cell receptor (TCR) transgenic mice42 (Figures S3A–S3D), which are specific to MHC class II and MHC class I-restricted viral peptides derived from LCMV Armstrong, respectively. Viral-specific T cells were transferred into VPS33BWT and VPS33BΔCSF1R mice 24 h prior to LCMV-Armstrong infection (Figure 4A). We found that CD4 T cells (SMARTA) transferred into VPS33BΔCSF1R mice failed to expand when compared to CD4 T cells transferred into VPS33BWT mice, as seen in the spleen and inguinal lymph nodes (IgLNs) (Figures 4B–4D). Interestingly, CD8 T cell expansion was unaffected after LCMV-Armstrong infection of VPS33BΔCSF1R mice (Figures 4E–4G). Together, these data indicate that DC-intrinsic VPS33B is critical for priming of viral-specific CD4 T cell responses but is dispensable for activation of CD8 T cell responses not requiring CD4 T cell help. DC-intrinsic VPS33B is required for CD4 T cell priming yet dispensable for the activation of CD8 T cell responses during viral infection (A) Schematic of experimental set-up for adoptive T cell transfer and LCMV Armstrong infection. Naive SMARTA (50,000) or P14 (5,000) T cells were isolated and transferred into VPS33BWT and VPS33BΔCSF1R mice. 24 h later, mice were infected with LCMV Armstrong (2 × 105 PFU; i.p). IgLNs and spleens were collected 7 days later and subject to flow cytometry. (B) Representative flow plots gating for transferred SMARTA T cells (live, CD90.2+, CD4+, CD45.1+, Vα2+) in the Ig LN (top) and spleen (bottom) of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. Pre-gated on live, CD90.2+, CD4+, CD8− cells. (C) Graphical quantification of total numbers of expanded SMARTA T cells and % expanded SMARTA T cells (% of all live, CD90.2+, CD4+ cells) in the Ig LN of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. (D) Graphical quantification of total numbers of expanded SMARTA T cells and % expanded SMARTA T cells (% of all live, CD90.2+, CD4+ cells) in the spleens of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. (E) Representative flow plots gating for transferred P14 T cells (live, CD90.2+, CD8+, CD45.1+, Vα2+) in the Ig LN (top) and spleen (bottom) of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. Pre-gated on live, CD90.2+, CD8+, CD4− cells. (F) Graphical quantification of total numbers of expanded P14 T cells and % expanded P14 T cells (% of all live, CD90.2+, CD8+ cells) in the Ig LN of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. (G) Graphical quantification of total numbers of expanded P14 T cells and % expanded P14 T cells (% of all live, CD90.2+, CD8+ cells) in the spleens of LCMV-infected VPS33BWT and VPS33BΔCSF1R mice. Error bars shown as mean ± SEM. n = 3–5 biological replicates for one independent experiment. Statistical analysis was performed by unpaired t test. ∗∗p < 0.01, ∗∗∗∗p < 0.0001, n.s. = not significant. Loss of VPS33B does not affect DC cytokine production, costimulatory molecule expression, or migratory capacity Naive T cell priming is a stringent process that requires several critical signals relayed from DCs to the interacting T cells.43 As outlined by the “three signal paradigm,” productive T cell priming requires the following: (1) antigen processing and presentation on MHC molecules; (2) upregulation of costimulatory molecules (CD80/86); and (3) secretion of pro-inflammatory cytokines.2,30 This tight regulation ensures the activation of adaptive responses only when PRRs are activated by microbial signals and prevents the potential activation of T cells when self-antigens are presented on MHC molecules in the absence of signal 2 and 3. To investigate the underlying mechanisms of impaired T cell priming in the absence of VPS33B, we sought to characterize generation of all three signals by VPS33B-deficient DCs following TLR stimulation. We found that VPS33BΔCSF1R BMDCs expressed similar surface expression of MHC class II (signal 1) or the co-stimulatory molecule CD86 (signal 2) at baseline and levels of both MHC class II and CD86 were equally upregulated upon TLR signaling (Figures 5A and 5B). We also observed similar steady-state expression of both MHC class I and MHC class II on sDCs from VPS33BWT and VPS33BΔCSF1R (Figures S4A and S4B). These data suggest a defect in T cell priming is not due to an inherent impairment in MHC or costimulatory molecule expression. Loss of VPS33B does not affect DC cytokine production, costimulatory molecule expression, or migratory capacity (A) Representative histograms depicting surface MHC II expression on VPS33BWT and VPS33BΔCSF1R BMDCs left unstimulated (US) or stimulated with LPS (100 ng/mL) for 6 h in vitro. Plots were pre-gated on live, CD11b+, CD11c+ cells. Fluorescence minus one (FMO) depicted by unfilled black line. Graphical quantification of MHC MFI (median fluorescence intensity) (x103) to the right. (B) Representative histograms of surface CD86 expression on VPS33BWT and VPS33BΔCSF1R BMDCs left US or stimulated with LPS (100 ng/mL) for 6 h in vitro. Plots were pre-gated on live, CD11b+, CD11c+ cells. FMO depicted by unfilled black line. Graphical quantification of CD86 MFI (median fluorescence intensity) (x103) to the right. (C) Representative flow plots of intracellular TNFα production by VPS33BWT and VPS33BΔCSF1R BMDCs left US or stimulated with LPS (100 ng/mL) for 60 min in vitro. Plots were pre-gated on live, CD11b+, CD11c+ cells. Graphical quantification of % TNFα+ CD11b+ CD11c+ cells stimulated with LPS (100 ng/mL) for the indicated time points below. (D) Representative flow plots of intracellular IL-12 production by VPS33BWT and VPS33BΔCSF1R BMDCs left US or stimulated with LPS (100 ng/mL) for 180 min in vitro. Plots were pre-gated on live, CD11b+, CD11c+ cells. Graphical quantification of % IL-12+ CD11b+ CD11c+ cells stimulated with LPS (100 ng/mL) for the indicated time points below. (E) Graphical quantification of % TNFα+ VPS33BWT and VPS33BΔCSF1R CD11c+ sDCs stimulated with LPS (100 ng/mL) for the indicated time points in vitro. (F) Graphical quantification of % IL-12+ VPS33BWT and VPS33BΔCSF1R CD11c+ sDCs stimulated with LPS (100 ng/mL) for the indicated time points in vitro. (G) Representative flow plots assessing FITC+ DCs in the ipsilateral LNs and contralateral LNs of VPS33BWT and VPS33BΔCSF1R. FITC (2 mg/mL) in acetone/dibutly phthalate (1:1) was applied on the flank of VPS33BWT and VPS33BΔCSF1R mice for 18 h. Ipsilateral LNs and contralateral LNs (negative controls) were collected and subject to flow cytometry. Pre-gated on live, CD11c+, MHCIIhi before FITC expression was assessed. Graphical quantification of %FITC+ DCs to the right. (H) Graphical quantification of CD86 MFI (mean fluorescence intensity) (x103) on DCs from the Ipsilateral LNs and contralateral LNs of VPS33BWT and VPS33BΔCSF1R mice 18 h post FITC painting. Pre-gated on live, CD11c+, MHCIIhi. (A–F) n = 2–4 independent experiments. (G and H) n = 3 biological replicates from one independent experiment. Error bars shown as mean ± SEM. Statistical analysis was performed by two-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, n.s. = not significant. Furthermore, we found that BMDCs from VPS33BWT and VPS33BΔCSF1R displayed similar levels of pro-inflammatory cytokine production upon both LPS and CpG-B stimulation (Figures 5C, 5D, S4C, and S4D). We found a similar phenotype for sDCs taken from VPS33BWT and VPS33BΔCSF1R mice (Figures 5E and 5F). Because DCs must not only process antigen but also migrate to secondary lymphoid organs to initiate T cell responses, we next evaluated whether DC trafficking was affected by VPS33B deficiency. We performed an FITC painting assay to assess FITC+ DCs within draining lymph nodes. We found that the number of FITC+ DCs within the ipsilateral IgLN was no different between VPS33BWT and VPS33BΔCSF1R mice (Figure 5G), demonstrating that the defects observed in T cell priming were not due to impaired migration of DCs to secondary lymphoid organs. Additionally, we found comparable upregulation of CD86 on ipsilateral Ig LN DCs from VPS33BWT and VPS33BΔCSF1R mice (Figure 5H). Importantly, we also observed no differences in DC subset frequencies in the draining lymph node 12 h after subcutaneous LPS injection (Figure S4E). Together, these findings demonstrate that VPS33B-deficient DCs are not impaired in MHC classes I or II expression, co-stimulatory molecule upregulation, pro-inflammatory cytokine production, or lymph node trafficking. Therefore, the observed defect in CD4 T cell priming is likely due to a specific impairment in antigen processing or peptide loading rather than generalized dysfunction in DCs lacking VPS33B. VPS33B-deficient DCs are impaired in the presentation of MHC class II-associated peptides derived from exogenous antigens We next sought to investigate if the stark defects in CD4 T cell priming could be attributed to impaired antigen processing and presentation. The loading of exogenous peptides onto MHC class II requires the trafficking of endocytosed proteins to the MIIC compartment, wherein cathepsins generate peptide fragments that can be loaded onto MHC class II molecules.44 Newly synthesized MHC class II molecules are initially stabilized through binding of the invariant chain (Ii/CD74), which is then cleaved to generate CLIP/SLIP-bound MHC class II.45,46 As such, cathepsin-generated peptides must displace invariant chain degradation intermediates in order to be loaded onto MHC class II for presentation on the plasma membrane. We next examined the levels of CLIP/SLIP-bound MHC class II, hypothesizing that VPS33BΔCSF1R DCs would be unable to generate exogenous peptides necessary to displace CLIP/SLIP within the MIIC. Indeed, VPS33BΔCSF1R DCs demonstrated significantly higher surface staining with 15G4, an antibody specific for both CLIP and SLIP-bound MHC class II, at baseline (Figure 6A).47 Moreover, in VPS33BWT DCs, TLR stimulation led to a marked decrease in 15G4 staining, reflecting efficient peptide exchange, presumably with proteins present in the serum. In contrast, VPS33BΔCSF1R DCs failed to downregulate surface 15G4 following TLR activation, suggesting impaired peptide exchange and persistent CLIP/SLIP association with MHC class II on the cell surface (Figure 6A). VPS33B-deficient DCs are impaired in the presentation of MHC class II associated peptides derived from exogenous antigens (A) Representative flow plots of surface 15G4 expression on VPS33BWT and VPS33BΔCSF1R BMDCs left US or stimulated with LPS (100 ng/mL) for the indicated time points. Plots were pre-gated on live, CD11b+, CD11c+ cells. Graphical quantification of % 15G4+ BMDCs and 15G4 MFI (mean fluorescence intensity) to the right. (B) Representative flow plots of surface YAe expression on VPS33BWT and VPS33BΔCSF1R BMDCs left US or pulsed with I-Ed tetramer (10 μg/mL) for 1 h before being chased for the indicated time points. Plots were pre-gated on live, CD11b+, CD11c+ cells. Graphical quantification of % YAe+ BMDCs to the right. (C) Representative flow plots of CTV-stained OT-II T cells (pre-gated on live, CD4+, CD8−) co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence or absence of OVA323-339 (0.01 and 0.1 μM) and LPS (100 ng/mL) for 72 h. Graphical quantification of % proliferating (CTV-diluted) OT-II T cells to the right. (D) Representative flow plots of % CD44+ OT-II T cells co-cultured with VPS33BWT and VPS33BΔCSF1R sDCs in the presence or absence of OVA323-339 (0.01 and 0.1 μM) and LPS (100 ng/mL) for 72 h. Graphical quantification of % CD44+ OT-II T cells to the right. NP = no peptide (OVA323-339). Error bars shown as mean ± SEM. (A–D) n = 2–4 independent experiments. Statistical analysis was performed by two-way ANOVA. ∗p < 0.01, ∗∗p < 0.01, ∗∗∗p < 0.01, ∗∗∗∗p < 0.0001, n.s. = not significant. We next aimed to demonstrate that degradation and presentation of exogenously administered protein was defective in DCs lacking VPS33B. To assess this, we utilized I-Ed-MHC-II tetramers with YAe antibody staining tools. I-E molecules are not expressed by B6 mice48 but uptake and subsequent degradation of I-Ed results in the generation of the immunodominant peptide, Eα (I-E 52–68) loaded on I-Ab.49 The monoclonal antibody YAe specifically recognizes the Eα peptide-I-Ab MHC complex, allowing quantification of exogenous antigen processing and presentation.50 We fed VPS33BWT and VPS33BΔCSF1R BMDCs with a low dose of tetramer (obtained from the NIH tetramer core) as a source of exogenous antigen for 1 h before the tetramer was washed off and cells were left in the incubator for the indicated time points to chase for Eα peptide loading. While the YAe antibody staining was measurable and increased over time in VPS33BWT, we observed a significant defect in YAe staining in VPS33BΔCSF1R BMDCs at all time points (Figure 6B). These data indicate that the processing and/or presentation of peptides onto MHC class II molecules are impaired in the absence of VPS33B. This impaired exogenous antigen presentation in VPS33B-deficient DCs could be responsible for defective CD4 T cell responses seen in vitro and in vivo against OVA and LCMV antigens. Given that our results indicate that VPS33B is required for invariant chain degradation intermediate displacement and peptide loading onto MHC class II molecules, we next sought to rule out any inherent defect in protease activity in the absence of VPS33B. To this end, we utilized DQ-OVA to investigate whether protease-mediated degradation of exogenous protein was impaired. VPS33BWT and VPS33BΔCSF1R BMDCs were pulsed with DQ-OVA and chased for the given timepoints. We found that degradation of DQ-OVA and generation of fluorescently labeled peptides was unaltered in VPS33BΔCSF1R BMDCs as compared to VPS33BWT BMDCs (Figure S5A), suggesting that protease activity within the endolysosomal pathway is unaffected by absence of VPS33B and likely does not contribute to the observed defect in antigen presentation. Lastly, we sought to determine if the defects in the ability of VPS33B-deficient DCs to prime CD4 T cells could be overcome by providing peptides that bypass the need for antigen processing. To do this, we set up similar in vitro priming co-cultures as Figure 1; however, DCs were pre-incubated with the already processed OVA peptide (OVA323-339), instead of OVA protein. MHC class II peptide exchange can occur at the cell surface eliminating intermediate steps required for antigen processing.51 We found that OT-II T cells cultured with peptide-loaded VPS33BWT and VPS33BΔCSF1R sDCs demonstrated no differences in proliferation as assessed by CTV dilution (Figure 6C). We also found expression of CD44 was similar in OT-II T cells primed by VPS33B sufficient and VPS33B deficient DCs. (Figure 6D). Taken together, these results demonstrate that DC-intrinsic VPS33B is required for optimal presentation of MHC class II-associated peptides derived from exogenous antigens. Discussion Loss-of-function mutations in VPS33B and its partner VPS16B lead to clinical syndromes with devastating outcomes for children. Despite the known importance of VPS33B in regulating vesicular trafficking and inflammatory responses, whether it plays a role in regulating cargo handling that affects generation of adaptive immune responses has not been examined. Here, we demonstrate that loss of DC-intrinsic VPS33B leads to impaired presentation of MHC class II-associated peptides. Consequently, this results in abrogated CD4 T cell responses during immunization and acute viral infection. Given the known helper role of CD4 T cells in eliciting CD8 T cell responses, we also observed impaired CD8 T cell responses during CD4-dependent responses (immunization), but not during CD4-independent responses (LCMV Armstrong and in vitro priming). These results suggest that loss of VPS33B predominantly affects MHC class II peptide loading, while the cytosolic route of MHC class I peptide generation and presentation appear to be largely unaffected. Our findings thus implicate VPS33B as a critical mediator of the antigen processing machinery required for presentation of MHC class II peptides and CD4 T cell responses. Previous work has demonstrated a key role for VPS33B in macrophages in termination of Toll-like receptor (TLR) signaling by directing endocytosed TLRs to the lysosome and facilitating endosomal-lysosomal fusion and degradation.8 Drosophila mutants lacking VPS33B succumbed to non-lethal infections due to excessive proinflammatory responses. Additionally, murine macrophages deficient in VPS33B failed to terminate TLR signaling in response to microbial ligands.8 Interestingly, we found that the absence of VPS33B in DCs resulted in minor differences in inflammatory responses including cytokine production and costimulatory molecule expression. Additionally, in vivo migration and maturation of conventional DCs were unaffected in the absence of VPS33B. Macrophages and DCs have many inherent differences including expression of proteases as well as magnitude of inflammatory responses following TLR activation.52,53 Collectively, these data suggest that VPS33B plays a dominant role in degrading receptors and cargo. Defects in this endosomal pathway appear to have macrophage- and DC-specific manifestations including increased magnitude of innate inflammatory responses and impaired antigen presentation, respectively. Delineating the differences between DCs and macrophages lacking VPS33B requires further investigation. During synthesis and early trafficking, MHC class II molecules are associated with invariant chain and subsequent degradation intermediates (CLIP/SLIP), which is required for stabilization of these molecules. Higher affinity peptides produced during cathepsin-mediated processing are then able to displace CLIP/SLIP, with the assistance of H2-M (HLA-DM in humans).54 H2-M plays a critical role in this process by stabilizing empty MHC class II molecules, allowing for loading of newly generated peptides.55,56 Interestingly, we have demonstrated that VPS33BΔCSF1R mice phenocopy H2M knockout mice, with both displaying increased surface MHC II-CLIP/SLIP complexes and markedly defective CD4 T cell responses50,57 due to inability to process exogenous proteins. However, H2M knockout mice maintain normal responses to peptide immunization58 in agreement with our in vitro priming experiments with OVA 323 peptide. Thus, VPS33B, like H-2M, plays a critical role in linking antigen uptake to CD4 T cell responses, albeit through distinct mechanisms. Intact CD8 T cell priming in vitro, however, might be the result of unaffected vesicular transportation of endocytosed antigens59 or even increased cytosolic leak of antigens as a result of halted endosomal-lysosomal fusion. However, our data unequivocally establish the critical role of CD4 T cells in determining priming of antigen-specific CD8 T cells in vivo. While generation of MHC class I peptides seems to be unaffected by the absence of VPS33B in DCs, lack of CD4 T cell help completely abrogates generation of CD8 T cell responses against protein antigens. Curiously, absence of VPS33B did not affect priming and clonal expansion of LCMV-specific CD8 T cells. It is important to note that in this study we evaluated acute expansion and activation of CD8 T cells but prior work has demonstrated the requirement for CD4 T cell help in generation of memory CD8 T cell responses;60 therefore, it is reasonable to speculate that loss of DC-intrinsic VPS33B would result in defective recall CD8 T cell responses following a secondary challenge with LCMV. A previous study found that loss of DC-intrinsic VPS33B exacerbated house dust mite (HDM)-induced allergic asthma by promoting Th2 generation and expansion.61 While our study did not assess Th2 responses, this appears contradictory to our findings. Interestingly, there is evidence that varying levels of antigen can influence T helper differentiation. Previous work has demonstrated that priming with low doses of antigen or using peptides with an altered avidity for TCR (altered peptide ligands [APL]), can generate predominant Th2 responses.62 In contrast, high levels of antigen, or unmodified antigen, lead to a strong Th1 response with minimal Th2 cells present.62 While the precise mechanisms underlying preferential Th subset generation are incompletely understood, differences in adapter molecule phosphorylation and calcium flux have been observed.63 In this study, we observed impaired, but not completely defective, YAe staining in VPS33BΔCSF1R BMDCs, suggesting some level of MHC class II peptide generation. Therefore, we posit that loss of DC-intrinsic VPS33B results in sub-optimal levels of MHC class II-peptide complexes that might be insufficient to generate strong TCR signaling associated with Th1 responses. However, low levels of MHC class II-peptide complexes may skew T cell responses toward a Th2 phenotype and promote worsened Th2-mediated diseases as previously observed. Future experiments will indeed examine generation of Th2 responses in mice lacking VPS33B in DCs by using allergens and other antigens known to drive type II Immunity. Clinically, mutations in VPS33B or its binding partner VIPAR (VPS16, encoded by VIPAS39) cause ARC syndrome, a rare multisystem disorder characterized by defective cell-mediated immunity and recurrent, often severe, infections.8 The link between VPS33B dysfunction and immune compromise underscores the importance of this protein complex in host defense. Supporting this, pathogens have evolved virulence factors that exploit VPS33B or related components of the VPS/HOPS complex to facilitate enhanced pathogenicity.64,65,66 For example, Mycobacterium tuberculosis (Mtb) encodes the phosphatase PtpA, which dephosphorylates VPS33B, thereby blocking phagosome-lysosome fusion and enhancing intracellular bacterial survival.67 Loss of PtpA activity attenuates Mtb growth, highlighting the critical role of VPS33B-mediated vesicular trafficking in pathogen clearance. Beyond its role in intracellular pathogen control, effective CD4 T cell responses are essential for controlling Mtb infection. Recognition of infected macrophages by CD4 T cells promotes protection against pulmonary pathology, emphasizing the need for functional antigen presentation pathways.68,69 Our findings suggest that impaired MHC class II antigen due to VPS33B deficiency may undermine these protective T cell responses, contributing to susceptibility to infection observed in ARC syndrome. Similarly, targeting of VPS33B by Mtb might result in impaired processing and presentation of Mtb-derived peptides leading to a skewed CD4 T cell response in infected patients. While BCG immunization leads to robust priming of Mtb-specific CD4 T cells, these could be futile responses since Mtb infected macrophages and DCs might not express matching Mtb derived epitopes that could be recognized by primed CD4 T cells. This might explain the failure of multiple vaccines against Tuberculosis. Future studies are warranted to elucidate upstream signaling events regulating VPS33B phosphorylation and its role in regulating antigen presentation during immune activation. These findings provide crucial insights into the immunodeficiency observed in ARC syndrome patients and provide avenues for potential therapeutic intervention in addressing some of the symptoms of the disease. Limitations of the study We chose to utilize CSF1R Cre for our VPS33B DC deletion. We recognize that Cre expressed under the control of CSF1R promoter will delete VPS33B in all myeloid cell compartments in vivo, including monocytes and macrophages. Further studies require the use of a more dendritic cell targeted deletion of VPS33B, such as using Cre transgenes such as Clec9a Cre, CD11c Cre, or XCR1 Cre. Here, we identify VPS33B as a key molecular player in exogenous antigen handling and presentation onto the MHC II complex. Deep mechanistic profiling of how VPS33B is able to regulate this process and examination of colocalization of VPS33B and mediators of this process (MIIC and H2M) by high resolution microscopy is needed to fully understand the role of VPS33B in antigen processing and presentation by DCs. Resource availability Lead contact Requests for further information and resources should be directed to and will be fulfilled by the lead contact, Chandrashekhar Pasare (chandrashekhar.pasare@cchmc.org). Materials availability VPS33Bfl/fL mice used in this study will be available upon request and following fulfillment of institutional material transfer agreement requirements. Data and code availability •Original western blot images will be shared by the lead contact upon reasonable request. Original western blot images will be shared by the lead contact upon reasonable request. •This study did not report original code. This study did not report original code. •Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request. Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request. Acknowledgments We thank all the Pasare lab members for generous help and useful discussions. This study was supported by a grant from the National Institutes of Health (AI155426) to C.P. C.P. is also supported by AI123176 from the National Institutes of Health. K.K. was supported by Pathways to Cancer Therapeutics T32 under grant no. CA 117846. C.N.V. is supported by National Institutes of Health F30 grant (AI188599). K.A.W. is supported by the Pre-doctoral Fellowship by the American Heart Association grant no. 24PRE1198808. We thank the NIH Tetramer Core Facility (contract number 75N93020D00005) for providing I-Ed tetramers. Author contributions K.K. and C.N.V. conceptualized the study, designed and performed the experiments, interpreted the data, participated in the writing – original draft, review, and editing. K.A.W. performed the experiments and writing – editing. S.A. performed LCMV infections and writing – editing. M.M.M. and K.A. conceptualized the study and writing – editing. C.P. conceptualized the study, designed the experiments, interpreted all the data, and participated in writing – review, and editing, funding acquisition, resources, and supervision. Declaration of interests The authors declare no competing interests. STAR★Methods Key resources table REAGENT or RESOURCESOURCEIDENTIFIERAntibodiesFc shield CD16/CD32Tonbo biosciencesCat#70-0161-U500; RRID:AB_2621487anti-CD90.2 biotinBioLegend105304; RRID:AB_313175anti-NK1.1 biotinBioLegendCat#108703; RRID:AB_313390anti-Ly6G biotinBioLegendCat#127603; RRID:AB_1186105anti-Ter119 biotinBioLegendCat#116203; RRID:AB_313704anti-F4/80 biotinBioLegendCat#123105; RRID:AB_893499anti-CD317 biotinBioLegendCat#127006; RRID:AB_2028466anti-CD19 biotinBioLegendCat#115503; RRID:AB_313638anti-CD11b (BV785)BioLegendCat#101243; RRID:AB_2561373anti-CD11c (BV650)BioLegendCat#117339; RRID:AB_2562414anti-CD45.1(PE)BioLegendCat#110707; RRID:AB_313496anti-CD45.2 (AF700)BioLegendCat#109821; RRID:AB_493730anti-CD90.2 (BV785)BioLegendCat#105331; RRID:AB_2562900anti-CD8α (APC/Cyanine7)BioLegendCat#100713; RRID:AB_312752anti-CD4 (PE)BioLegendCat#100511; RRID:AB_312714anti-15G4 (PE)Santa CruzCat#sc-53946 PE; RRID:AB_831549anti-CD44 (APC)BioLegendCat#103011; RRID:AB_312962anti-MHCII (AF700)BioLegendCat#107622; RRID:493727anti-MHCI (FITC)BioLegendCat#114605; RRID:AB_313596anti-CD86 (PE)BioLegendCat#105106; RRID:AB_313159anti-XCR1 (BV421)BioLegendCat#148216; RRID:AB_2565230anti-SIRPα (PE/Cyanine7)InvitrogenCat#251721-82; RRID: AB_2762640anti-Tbet (PE/Cyanine7)BioLegendCat#644824; RRID:AB_2561761anti-GranzymeB (FITC)BioLegendCat#515403; RRID:AB_2114575anti-TCR Vα2BioLegndCat#127810; RRID:AB_1089250anti-TNFα (APC)BioLegendCat#506307; RRID:AB_315428Anti-IFNγ (PE/Cyanine7)BioLegendCat#505825; RRID:AB_1595591anti-IL-12/IL-23 p40 (AF700)BioLegendCat#505213; RRID:AB_2894642anti-H3Cell SignalingCat#5192S; RRID:AB_1950410anti-VPS33BGift from Helmut Kramer UTSW Medical Center–anti-IFNγ capture antibodyBioLegendCat#505701; RRID:AB_315389Biotinylated anti-IFNγ detection antibodyBioLegend505803; RRID:AB_315397Biotinylated anti-CD19 antibodyBioLegendCat#115504; RRID:313639YAe biotinInvitrogenCat#13-5741-82; RRID:AB_657821Streptavidin (BV421)BioLegendCat#405226; RRID:AB_3662382Streptavidin nanobeadsBioLegendCat#480015Anti-biotin MicroBeadsMiltenyi BiotecCat#130-097-046Bacterial and virus strainsLCMV ArmstrongGift from Koichi Araki Lab CCHMCAraki et al.70Chemicals, peptides, and recombinant proteinsZombie YellowBioLegendCat#423104RIPA bufferThermoFisherCat#89901Protease inhibitor Cocktail TabletsRocheCat#1169748001Phosphatase inhibitor Cocktail TabletsRocheCat#4906845001Dynabeads Protein GInvitrogenCat#10003DNuPAGE LDS bufferInvitrogenCat#NP0007Fetal calf serum (FCS)Sigma-AldrichCat#F0926DNAse ISigma-AldrichCat#D8515Recombinant GM-CSFBioLegendCat#5763062-Mercaptoethanol (BME)Sigma-AldrichCat#M6250Paraformaldehyde aqueous solutionThermoFischerCat#15710Fluorescein isothiocyanate (FITC) isomer ISigma-AldrichCat#F7250Dibutyl PhthalateSigma-AldrichCat#524980AcetoneSigma-AldrichCat#179124PBSThermoFisherCat#14190250Hyclone RPMI-1640 with L-GlutamineCytivaCat#SH3002701Bovine Serum Albumin (BSA)Sigma-AldrichCat#A7906Phosphatase inhibitorRocheCat#4906845001Collagenase IVWorthingtonCat#LS004188DNase IRocheCat#10104159001TRIzol ReagentInvitrogenCat#15596026DMEM, high glucoseThermoFisherCat#11965118Gibco HEPESThermoFisherCat#15630080OPDSigma-AldrichCat#P1526EDTASigma-AldrichCat#E9884RBC lysis bufferSigma-AldrichCat#R7757IL-2BioLegendCat#575402PMAInvivogenCat#tlrl-pmaIonomycinInvivogenCat#inh-ionBrefeldin ABioLegendCat#420601CD4 T cell isolation kitBioLegendCat#480039CD8 T cell isolation kitBioLegendCat#480043Cell Trace VioletInvitrogenCat#C34557EndoFitTM OVAInvivogenCat#vac-povaDQ-OVAInvitrogenCat#D12053OVA 323-339InvivogenCat#vac-isqLipopolysaccharideSigma-AldrichCat#L288CpG-BInvivogenCat#Tlrl-1668Experimental models: Organisms/strainsfLSanger Institute Mouse Genetics projectRRID:IMSR_EM:07849CSF1R CreJackson LabRRID:IMSR_JAX:029206Vav CreJackson LabRRID:IMSR_JAX:008610OT-IIJackson LabRRID:IMSR_JAX:004194OTIJackson LabRRID:IMSR_JAX:003831SMARTAJackson LabRRID:IMSR_JAX:030450P14Jackson LabRRID:MMRRC_037394-JAXSoftware and algorithmsFlowJo 10https://www.flowjo.comN/AGraphPad Prism v10https://www.graphpad.comN/AAdobe Illustrator CC 2024https://www.adobe.com/N/AOtherNovocyte 2001AgilentN/AAutoMacsMiltenyi Biotec–Polyvinylidene fluoride (PVDF) sheetBioRadCat#1620177HyBlot Autoradiography FilmThomasSciCat#1141J52EasySep EasyEight MagnetStemcellCat#NC0651008Bolt 4–12% Bis-Tris 1.0 mm Protein GelsThermoCat#NW04122BOXKodak x-OMAT 2000 DeveloperKodakN/A Experimental model and study participant details Mice To generate VPS33BΔCSF1R mice, mice with floxed allele of VPS33B (Vps33Bfl/fL(https://www.mousephenotype.org/data/alleles/MGI:2446237/tm1a(EUCOMM)Wtsi). Exon 2 and 3 of the VPS33b allele are flanked by loxP sites in these mice (Figure S1A) and crossing them to CSF1R Cre mice obtained from Jackson (C57BL/6-Tg (Csf1r-cre)1Mnz/J) led to generation of VPS33BΔCSF1R mice and they had complete deletion of VPS33B protein in dendritic cells (Figure S1B). VPS33BΔVav mice were generated by crossing VPS33Bfl/fLCg-Commd10Tg(Vav1-icre)A2Kio/J). VPS33Bfl/fL12 weeks of age and were bred and housed in a specific pathogen-free facility at Cincinnati Children’s Hospital Center. For all infectious experiments, mice were transferred to a specific facility prior to infection. No influence of sex was reported for all conducted experiments. Mice were group-housed in environmentally enriched cages with humidity and temperature-controlled rooms with a 12-h light dark cycle. Mouse genotypes were determined using PCR with transgene specific probes (Integrated DNA Technologies). All mouse experiments were done as per protocol approved by Institutional Animal Care and use Committee (IACUC) guidelines at Cincinnati Children’s Hospital Medical Center (#2024-0012 and #2024-0015). Method details T cell isolation Single cell suspensions were obtained from the spleens of OT-II (Jackson Laboratory; B6. Cg-Tg (TcraTcrb)425Cbn/J) or OTI (Jackson Lab; C57BL/6-Tg (TcraTcrb)1100Mjb/J) transgenic mice. Naive CD4 T cells were isolated using a mouse naive CD4 T cell isolation kit (BioLegend; 480039) and naive CD8 T cells were isolated using a mouse naive CD8 T cell isolation kit (BioLegend; 480043). Naive CD4 and CD8 T cells were then labeled with CTV according to the manufacturer’s instructions (Invitrogen; C34571) before co-culturing with splenic DCs as described below. In vivo protein immunizations VPS33BWT and VPS33BΔCSF1R mice were immunized with 50 μg of OVA and 5 μg of LPS emulsified in Incomplete Freund’s Adjuvant (IFA) (100 μL/mouse; hock injection). The draining Ig LN was collected 7 days later, and a single cell suspension was made. Total cells were left unstimulated or re-stimulated with EndoFit OVA protein (Invivogen; vac-pova) (10 and 100 μg/mL) and left at 37° for 72 h. Cells were spun down and supernatants were collected for IFNγ ELISAs. Half of the cells were subject to CD44, Tbet, and GrzB flow cytometry staining while the other half was resuspended in low dose (10 U/mL) IL-2 (BioLegend; 575402) for an additional 48-h. Rested cells were then stimulated with PMA (Invivogen; tlrl-pma) and Ionomycin (Invivogen; inh-ion) with Brefeldin A (BioLegend; 420601) for 6 h and intracellular IFNγ was measured via flow cytometry. In vivo T cell transfer Where indicated, naive OT-II and OTI T cells were isolated as described above. Cells were washed with PBS 3x before OT-II (50,000) and OTI (5,000) T cells were transferred via retro-orbital injection into VPS33BWT and VPS33BΔCSF1R mice. 24-h post transfer, mice were immunized with 50 μg of OVA and 5 μg of LPS in IFA (100 μL/mouse; hock injection). Ig LNs were collected 7 days later, and cells were isolated for flow cytometry analysis. LCMV- Armstrong infection Single cell suspensions were generated from the spleens of SMARTA (Jackson Laboratory; B6.Cg-Ptprca Pepcb Tg (TcrLCMV)1Aox/PpmJ) and P14 (Jackson Laboratory; B6.Cg-Tcratm1Mom Tg(TcrLCMV)327Sdz/TacMmjax) transgenic mice and naive CD4 and CD8 T cells were isolated as described above. Cells were washed 3x with PBS and 50,000 SMARTA or 5,000 P14 T cells were retro-orbitally transferred into VPS33BWT and VPS33BΔCSF1R mice where indicated. 24-h post T cell transfer, mice were infected with LCMV Armstrong (2 × 105 PFU; i.p). Ig LN and spleens were collected 7 days later, and T cell expansion was assessed via flow cytometry. Splenic DC (sDC) and BMDC isolation Spleens were collected and single cell suspensions were generated. Cells were brought to a concentration of 100 x 106 cells/mL in staining buffer (PBS supplemented with 2 mM EDTA and 2% FCS). Cells were then blocked with FC Shield (1:500) (anti-mouse CD16/32, Tonbo) on ice for 15 min before the following biotinylated antibodies were added (1:400): CD90.2 (BioLegend; 105304), NK1.1(BioLegend; 108703), Ly6G (BioLegend; 127603), F4/80 (BioLegend; 123105), Ter119 (BioLegend; 116203), CD317 (BioLegend; 127006), and CD19 (BioLegend; 115503). Cells were incubated on ice for 15 min before streptavidin nanobeads (BioLegend; 480015) were added. Cells were incubated on ice for another 15 min prior to magnetic separation. The negative fraction was collected for in vitro co-cultures described below. Bone marrow progenitors were isolated from femurs and tibias of VPS33BWT and VPS33BΔCSF1R mice. Following RBC lysis, cells were plated at 0.75 x 106/mL in BMDC media (5% RPMI supplemented with GM-CSF). Media was replaced every two days, and cells were harvested for experiments on day 5 by gently pipette mixing. B cell isolation Spleens were collected and single cell suspensions were generated. Cells were brought to a concentration of 100 x 106 cells/mL in staining buffer (PBS supplemented with 2 mM EDTA and 2% FCS). Cells were then blocked with FC Shield (1:500) (anti-mouse CD16/32, Tonbo) on ice for 15 min, washed, and incubated with biotinylated CD19 antibody (BioLegend; 115504) at 1:400 for 20 min on ice. Excess antibody was then washed off, and cells were incubated with anti-biotin microbeads (Miltenyi Biotec; 130-097-046) for 20 min prior to washing and positive selection of CD19+ cells (AUTOMACS, Miltenyi Biotec). APC -T cell co-cultures for in vitro priming sDCs were co-cultured with CTV labeled T cells at a 1:4 DC: T cell ratio in a 96-well tissue culture-treated round bottom plate in 10% RPMI. Cells were left unstimulated or stimulated with EndoFit OVA (10 and 100 μg/mL), OVA323-339 (0.01 and 0.1 μM), LPS (100 ng/mL), or CpG-B (1 μM) where indicated and left at 37° for 72 h prior to being subject to flow cytometry as outlined below. B cells were co-cultured with CTV labeled T cells at a 2:1 B:T cell ratio in a 96-well tissue culture-treated round bottom plate in 10% RPMI. Cells were left unstimulated or stimulated with EndoFit OVA (300 μg/mL) with LPS (100 ng/mL) where indicated and left 37° for 72 h prior to being subject to flow cytometry as outlined below. Flow cytometry For surface markers, cells were washed 2x with PBS then stained with Zombie-Yellow (1:1000) for 10 min at room temperature. Cells were then blocked with FC Shield (1:500) (anti-mouse CD16/32, Tonbo) on ice for 15 min, washed, and incubated on ice with antibodies of interest for 30 min. The following antibodies were used at 1:400: CD11b (BioLegend; 101243), CD11c (BioLegend; 117339), CD45.1 (Biolegend; 110707), CD45.2 (BioLegend; 109821), CD90.2 (BioLegend; 105331), CD8α (BioLegend; 100713), CD4 (BioLegend, 100511), 15G4 (SantaCruz; sc-53946 PE), CD44 (BioLegend; 103011). CTV staining (Invitrogen; C34557) was performed at 1:1000 in PBS for 20 min at 37o C. For intracellular staining, cells were fixed in 1% PFA and permeabilized (Invitrogen; 00-5523-00) prior to staining. The following antibodies were used for intracellular staining at 1:250: IFNγ (BioLegend; 505825), Tbet (BioLegend; 644824), GranzymeB (BioLegend; 515403), TNFα (BioLegend, 506307), IL-12 (BioLegend; 505213). Cells were washed with FACS buffer (PBS, 2% FCS, and 2 mM EDTA) 2x before being analyzed using Novocyte 3001 (ACEA Biosciences). Data were analyzed using FlowJo software (BD). All gates were set at unstimulated control samples or fluorescence minus one (FMO). Western blotting sDCs were lysed in 1x RIPA buffer (ThermoFisher Scientific; 37576) supplemented with protease (#1169748001; Roche) and phosphatase (#4906845001; Roche) inhibitors. Protein was quantified using Pierce BCA Protein Assay Kit (Pierce Coomassie Plus Assay Reagent, ThermoFisher). Cell lysates were boiled in a 1X Laemmli buffer at 70°C for 10 min. Lysates were then separated by 4–12% SDS–PAGE and transferred onto 0.45 μM PVDF membranes (Merck Millipore). Membranes were blocked for 30 min in 5% BSA in TBST prior to overnight incubation with antibodies against the following proteins: VPS33B (Ab was a gift from Helmut Kramer UTSW Medical Center 1:1000) and H3 (Cell Signaling; 5192S, 1:10,000). Membranes were washed with TBST before being incubated with their respective HRP-conjugated secondary antibodies. Western blots were developed using SuperSignal West Pico PLUS Chemiluminescent Substrate (Thermo Fisher; 34577) and ECL signal was recorded on X-ray Films using a developer (Kodak). ELISA IFNγ capture antibody (BioLegend; 505701) was diluted in a bicarbonate coating buffer and used to coat 96-well flat bottom plates overnight at 4°. Plates were blocked in 1% BSA for 2 h at room temperature before loaded with supernatant samples overnight at 4°. Biotinylated IFNγ detection antibody (BioLegend; 505803) was added to wells for 2 h at room temperature, washed, and incubated with streptavidin before protein concentration was quantified by OPD detection according to the manufacturer’s instructions. FITC painting assays Fluorescein isothiocyanate (FITC) isomer I was purchased from Sigma-Aldrich (SKU F7250). Mice were anesthetized and shaved on their left flank. 24h later, mice were again anesthetized and 40 mL of FITC (2 mg/mL) resuspended in a 1:1 ratio of acetone: dibutyl phthalate was applied directly onto the skin by pipetting. 18h after FITC application, mice were euthanized and ipsilateral and contralateral inguinal lymph nodes (IgLN) were harvested. Lymph nodes were finely minced and digested in DMEM supplemented with collagenase IV (1 mg/mL) (Worthington; LS004188) and DNase I (40 μg/mL) (Roche; 10104159001) for 40 min at 37o C. Suspensions were then filtered and subjected to flow cytometry as outlined above. DQ-OVA uptake and degradation assays BMDCs from VPS33BWT and VPS33BΔCSF1R mice were isolated as described above. BMDCs were pulsed with DQ-OVA (Invitrogen; D12053) at 10 μg/mL in 1 mL of PBS for 20 min at 37o C. Cells were then washed, re-suspended in 5% RPMI, transferred to individual Eppendorf tubes, and chased for the given timepoints. At the given timepoints, cells were placed on ice to prevent further degradation. As a control, following 20 min DQ-OVA pulse, cells were placed directly on ice (0 min chase). Samples were then processed on ice for flow cytometry to assess for DQ-OVA degradation and fluorescence. YAe staining BMDCs from VPS33BWT and VPS33BΔCSF1R mice were isolated as described above. BMDC were pulsed for 1 h at 37° with 10 μg/mL of I-Ed tetramer (I-Ed 52–68 SFERFEIFPKE) generated from the NIH Tetramer Core Facility. After 1 h, cells were spun down and washed with 10% RPMI before being placed back at 37° for 30, 60, or 120 min. At the indicated time points, cells were immediately placed on ice to prevent further protein processing. Once all-time points were collected, cells were subject to flow cytometry using YAe biotin (Invitrogen; 13-5741-82) followed by Streptavidin BV421 (BioLegend; 405226) antibodies. Quantification and statistical analysis Number of biological replicates and independent experiments as well as statistical tests performed are reported in each figure legend. All statistical tests were performed using GraphPad Prism 10. Error bars shown as ± SEM. Statistical significance was assessed by the test depicted in the respective figure legend (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 and ∗∗∗∗p < 0.0001; n.s., not significant). Footnotes Supplementary data related to this article can be found online at https://doi.org/10.1016/j.isci.2026.116052. Supplemental information Document S1. Figures S1–S5 References Associated Data Supplementary Materials Document S1. Figures S1–S5 Data Availability Statement •Original western blot images will be shared by the lead contact upon reasonable request. Original western blot images will be shared by the lead contact upon reasonable request. •This study did not report original code. This study did not report original code. •Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request. Any additional information required to reanalyze the data reported in this study is available from the lead contact upon request.