Sorting nexin 5 deficiency impairs MHC class II antigen presentation and inflammatory responses during tuberculosis Authors: Beatriz R S Dias, Kubra F Naqvi, Victoria A Ektnitphong, Samuel Alvarez-Arguedas, Kathryn C Rahlwes, Priscila C Campos, Michael U Shiloh Published: 2026 ABSTRACT AbstractTuberculosis (TB) remains one of the leading causes of death from a single infectious agent worldwide, yet the host pathways that regulate antigen presentation and lung inflammation during Mycobacterium tuberculosis (Mtb) infection are incompletely defined. Sorting nexin 5 (SNX5) is a protein implicated in endosomal trafficking, antigen processing, and antiviral host defense, but its contribution to antibacterial immunity is unknown. Here, we show that SNX5-deficient mice exhibit increased mortality following low-dose aerosol Mtb infection despite unchanged pulmonary bacterial burden compared with wild-type mice. Snx5−/− mice developed exacerbated lung inflammation without major alterations in immune cell recruitment. In macrophages, Snx5 did not affect phagocytosis, vacuolar maturation, intracellular bacterial control, or global transcriptional responses to Mtb but was required for efficient major histocompatibility complex (MHC) class II antigen presentation. Snx5 deficiency was associated with reduced endolysosomal proteolysis and impaired MHC class II antigen presentation in vitro, resulting in reduced activation of antigen-specific CD4+ T cells without altering surface MHC class II abundance or costimulatory molecule expression. Together, these findings identify SNX5 as a regulator of MHC class II antigen presentation that influences inflammatory outcomes during pulmonary Mtb infection.Keywords: antigen presentation, host-pathogen interaction, Mycobacterium tuberculosis, sorting nexin 5, tuberculosis Introduction Tuberculosis (TB) remains a significant global health challenge. Although TB-associated mortality declined to 1.23 million people in 2024, TB continued to be the leading cause of infectious disease-related deaths.1 Incidence increased to 8.3 million cases in 2024, the highest number recorded since global TB monitoring began in 1995.1 Despite the identification of Mycobacterium tuberculosis (Mtb) nearly 150 years ago, an effective vaccine that reliably prevents adult pulmonary TB is still lacking.2,3 Treatment requires prolonged multidrug regimens, and drug-resistant strains continue to emerge.4–6 These challenges highlight the need to understand how Mtb interacts with host cells, how early immune responses shape pulmonary inflammation, and how these processes influence disease outcome. Innate immune recognition of Mtb involves pattern recognition receptors that detect bacterial cell wall lipids and nucleic acids. After uptake by macrophages, Mtb resides within membrane-bound phagosomes where it can persist. Macrophage activation facilitates phagosomal maturation via fusion with host endolysosomal compartments and acidification to pH ≤5.7,8 However, phagosome maturation and fusion with lysosomes are often disrupted during infection, limiting antimicrobial activity.9–13 These alterations influence antigen processing and the timing of downstream adaptive immune responses.14–16 Because protective immunity depends on appropriately coordinated CD4+ T cell responses,17–21 the pathways that govern phagosomal maturation in infected phagocytes are critical determinants of host control. In addition to their essential role in innate cell-autonomous immunity, Mtb-infected macrophages and dendritic cells (DCs) activate adaptive immunity through antigen presentation. Efficient antigen presentation requires lysosomal processing of internalized bacteria, followed by loading of peptide antigens onto major histocompatibility complex (MHC) class II molecules and transport of peptide-MHC complexes to the cell surface for recognition by T-cell receptors.22 In humans and mice, initiation of adaptive immunity to Mtb infection is delayed relative to many other infections.23,24 Although Mtb is an airway and lung pathogen, antigen-specific CD4+ T cell responses are initiated in lung-draining mediastinal lymph nodes, where DCs present processed Mtb antigens to naïve CD4+ T cells, leading to their activation, proliferation, and migration to the site of infection.25,26 Mtb has evolved multiple strategies to limit its exposure to degradative compartments including interference with membrane trafficking and phagosome maturation,10,11,27–32 and these mechanisms can reduce antigen processing and prevent efficient antigen presentation.14,33 The balance between these bacterial evasion strategies and host endolysosomal processing pathways is a key determinant of whether infection is contained or progresses toward persistent inflammation.34 Sorting nexins are a family of phosphoinositide-binding proteins that regulate endosomal sorting, membrane remodeling, and intracellular cargo trafficking.35–37 Sorting nexin 5 (SNX5) contains a PX domain, which allows binding to different phosphoinositides, and a BAR domain that senses and induces membrane curvature and tubulation.38–41 SNX5 functions as a component of the retromer complex, where it captures endosomal cargo for retrograde trafficking to the Golgi network, thereby influencing the localization and trafficking of multiple cellular components.42 In addition to its role in the retromer complex, SNX5 associates with SNX4 and SNX17 to form the recycler complex, which mediates the retrieval of autophagy-related proteins, including autophagy-related gene 9 and syntaxin 17, from the autophagosome surface following autophagosome- lysosome fusion.43 Through these activities, SNX5 is positioned to regulate endolysosomal trafficking events that intersect with both autophagy and antigen processing pathways. SNX5 has also been implicated in immune cell function. It is required for efficient macropinocytosis by macrophages,44 including the participation in the uptake of M. bovis by IFN-γ–activated macrophages,45 and has been linked to antigen processing.44 These observations suggest that SNX5 may influence how phagocytes acquire and process microbial antigens, although its role in host defense against intracellular bacterial pathogens has not been defined. Recently, an image-based genome-wide small interfering RNA screen identified SNX5 as a required factor for virus-induced autophagy but not for other forms of stress-induced autophagy.46 Deletion of Snx5 increases cellular and murine susceptibility to numerous viruses including Sindbis virus, herpes simplex virus type 1, West Nile virus, and chikungunya virus, resulting in enhanced viral replication.46 Mechanistically, SNX5 is recruited to virus-containing endosomes, where it promotes the generation of phosphatidylinositol-3-phosphate by the class III phosphatidylinositol-3-kinase complex, thereby activating autophagy-dependent antiviral immune responses.46 These findings establish SNX5 as a regulator of endosomal signaling events that couple membrane trafficking to immune defense. Although SNX5 has no previously established role in Mtb infection, Mtb similarly traffics through the endolysosomal system following internalization. This parallel raised the possibility that SNX5-dependent trafficking pathways might also influence host responses to intracellular bacterial infection. Here, we investigated the role of SNX5 during Mtb infection in vivo and in macrophages. We show that in vivo Snx5 deficiency increased susceptibility to Mtb infection without altering bacterial burden, instead leading to exacerbated lung inflammation. At the cellular level, deletion of Snx5 did not impair macrophage cell-autonomous antimicrobial immunity but markedly reduced MHC class II–dependent antigen presentation to CD4+ T cells. Together, these findings identify SNX5 as a regulator of antigen presentation that shapes inflammatory outcomes during Mtb infection and highlight a previously unappreciated role for an endosomal sorting component in host immunity to tuberculosis. Methods Mice Snx5 −/− mice, backcrossed to the C57BL/6J genetic background for more than 10 generations, were obtained from Beth Levine46 and bred to C57BL/6J mice (The Jackson Laboratory) to obtain Snx5+/− heterozygotes. Littermate wild-type (WT) control mice produced by heterozygous crosses were used for all experiments. All experimental protocols involving the use of mice were approved by the Institutional Animal Care and Use Committee of the University of Texas (UT) Southwestern Medical Center (protocol number 2017-101977). Mtb strains Mtb Erdman WT and mCherry-expressing Mtb (mCherry-Mtb) strains were cultivated in Middlebrook 7H9 (BD Biosciences; 271310) medium supplemented with 0.5% v/v glycerol (Fisher Scientific; G33-4), 0.05% v/v Tween 80 (Fisher Scientific; BP338-500), and 10% v/v Middlebrook OADC (oleic acid, albumin, dextrose, catalase) (BBL; 212351). Aerosol infection Equal numbers of littermates female and male Snx5−/− and WT mice were infected via aerosol with a low-dose inoculum of Mtb, as previously described.47–51 Briefly, mid-log-phase Mtb was washed 3 times with phosphate-buffered saline (PBS) and sonicated to disperse clumps. Then, the bacteria were resuspended in PBS (OD600 of 0.1). This suspension was introduced to the nebulizer of a GlassCol aerosol chamber, calibrated to infect mice with approximately 200 bacteria per mouse. To determine the initial inoculum, on the day of infection (day 0), the entire lung from 6 mice (3 WT and 3 Snx5−/−) was collected, homogenized, and plated on 7H11 agar plates supplemented with 10% OADC (BBL; 212351), 0.5% v/v glycerol (Fisher Scientific; G30-4), and 50 µg/mL cycloheximide (Thermo Fisher Scientific; 357420050). At various time points postinfection (0, 7, 21, 42, 90, 120, and 150 d), the left lung, half of the spleen, and the left lobe of the liver were collected, homogenized, and plated on 7H11 plates for quantitation of colony-forming units (CFUs). For cytokine detection, the remaining homogenate was centrifuged and filtered. IFN-γ (Invitrogen; 88-7314-88), TNF-α (R&D Systems; MTA00B-1), IL12p40 (R&D Systems; MP400), and IL-2 (Invitrogen; 88-7024-88) were measured by ELISA according to the manufacturer’s instructions. For histology and immunohistochemistry analysis, the superior right lung lobe was inflated and fixed in 10% buffered formalin. For survival studies, infected mice were euthanized upon reaching a loss of 15% of their maximum body weight.47–51 Histopathology Lungs from infected mice were collected, fixed with 10% buffered formalin, transferred to 70% ethanol, and paraffin embedded. Then, paraffin-embedded lungs were sectioned and stained with hematoxylin and eosin by the UT Southwestern Medical Center Histo-Pathology core facility. The slides containing lung sections were scanned using a NanoZoomer S60 digital scanner. Next, the inflammation area in each lung was quantified in a blinded analysis using ImageJ software (version 1.54g) (National Institutes of Health). Flow cytometric assessment of lung cell population Lungs from WT and Snx5−/− mice uninfected or infected with Mtb were minced and incubated in PBS with 0.5% bovine serum albumin (BSA), 0.2 mg/mL DNase I (Sigma-Aldrich; DN25), and 5 mg/mL collagenase type IV (Gibco; 17102-015) on an orbital shaker at 100 rpm at 37 °C for 30 min. After digestion, the lung tissues were processed through a 70 μm cell strainer on a 50 mL conical tube. Then, the filters were washed with BSA buffer (PBS + 0.5% BSA), and the single-cell suspension was centrifuged at 350 x g for 8 min at 4 °C, and then resuspended in 1 mL of ACK lysis buffer (Gibco; A10492-01). After 1 min, BSA buffer was added to stop the reaction, and cells were centrifuged at 350 x g for 7 min at 4 °C. Cells were then transferred to a 96-well plate. After centrifugation, cells were resuspended in 50 μL of Fc-blocking solution (Invitrogen; 14-0161, 1:100) for 20 min at 4 °C. Cells were washed with staining buffer (PBS + 2.5% fetal bovine serum [FBS]) twice and centrifuged at 350 g for 5 min at 4 °C. Finally, samples were incubated with surface antibody cocktail resuspended in brilliant stain buffer (Invitrogen; 00-4409-75) for 40 min at 4 °C and fixed with 1% paraformaldehyde (PFA) overnight. Samples were acquired on a BD LSR II at the UT Southwestern Flow Cytometry Facility. The compensation matrix was calculated using UltraComp ebeads (Invitrogen; 01-3333-42) for the antibodies and tetramers and ArC Amine Reactive Compensation Bead Kit (Invitrogen; A10628) for the LIVE/DEAD Fixable dead cell stain kits. Data analysis was performed using FlowJo 10 software (TreeStar). The antibody cocktail included CD4 (Super Bright 600; Invitrogen; 63-0042-82, 0.625 μL per test), CD45 (eFluor 450; Invitrogen; 48-0451-82, 2.5 μL per test), CD3 (APC-eFluor 480; Invitrogen; 47-0032-82, 5 μL per test), CD11b (Super Bright 780; Invitrogen; 78-0112-82, 0.15 μL per test), Ly-6G (Super Bright 702; Invitrogen; 67-9668-82; Invitrogen; 0.15 μL per test), F4/80 (Alexa Fluor 488; Invitrogen; 1 μL per test), CD8 (PE-Cyanine 7; Invitrogen; 25-0081-82, 0.15 μL per test), CD11c (PE-Cyanine 5; Invitrogen; 0.3 μL per test), Ly-6C (Alexa Fluor 700; BioLegend; 128024, 0.5 μL per test), and Live/Dead fixable Aqua Blue (Invitrogen; L34957, 1:500). T cell functionality Spleen and digested lungs from WT and Snx5−/− mice uninfected or infected with Mtb were processed through a 70 μm cell strainer on a 50 mL conical tube and washed twice. Cells were resuspended in 1 mL of RPMI with 0.5 mg/mL of anti-CD28 (BioLegend; 102102) and Golgi Plug (1:1000, BD Biosciences, 51-230KZ) and transferred to a 6-well plate, precoated with 1 μg/mL of anti-CD3 (BioLegend; 100302) for 5 h at 37 °C. After 5 h of incubation, the cells were collected and washed 2 times with PBS containing 1% FBS. Cells were incubated in 1 mL of ACK lysis buffer for 1 min at room temperature and washed 2 times with PBS + 1% FBS. After centrifugation, cells were resuspended in 50 μL of Fc-blocking solution (Invitrogen; 14-0161, 1:100) for 15 min at 4 °C. Cells were washed and incubated with surface antibody cocktail (CD4 (Super Bright 600; Invitrogen; 63-0042-82, 0.625 μL per test), CD45 (eFluor 450; Invitrogen; 48-0451-82, 2.5 μL per test), CD3 (APC-eFluor 480; Invitrogen; 47-0032-82, 5 μL per test), and Live/dead fixable Aqua Blue (Invitrogen; L34957, 1:500)) resuspended in brilliant stain buffer (Invitrogen; 00-4409-75) for 30 min at 4 °C. Cells were then washed twice and permeabilized and fixed with BD Cytofix/Cytoperm (BD Biosciences, 554722) according to the manufacturer’s instructions. Cells were washed twice with BD Perm/Wash Buffer (BD Biosciences, 554723) and incubated with intracellular antibody cocktail (IFN-γ (Alexa Fluor 488; Invitrogen; 53-7311-82, 1 μL per test), TNF-α (Brilliant Violet 786; Invitrogen; 417-7321-82, 1.25 μL per test), and IL-2 (PE-Cyanine 7; Invitrogen; 25-7021-82, 5 μL per test)) diluted in BD Perm/Wash Buffer for 30 min at 4 °C. fixed with 1% PFA overnight at 4 °C. Samples were acquired and data analysis were performed as described above in Flow cytometric assessment of lung cell population. After all washed steps, cells were centrifuged at 350 x g for 5 min at 4 °C. Cells and cell culture Bone marrow cells from Snx5−/− or WT mice femurs and tibia were flushed and used to generate bone marrow–derived macrophages (BMDMs) and bone marrow–derived dendritic cells (BMDCs). To obtain BMDMs, bone marrow cells were cultivated in DMEM (Gibco; 11965-092) containing 20% heat-inactivated endotoxin-free FBS (Gibco; 1600-044), 30% L929 cell-conditioned media, GlutaMAX (Gibco; 35050-061), 1 mM sodium pyruvate (Gibco; 11360-070), 1 mM HEPES (Cytiva; SH30237.01), and 50 mM β-mercaptoethanol (Gibco; 21985-023). Media was replaced after 3 d of culture. After 7 d, adherent BMDMs were detached from Petri dishes (100 × 15mm) with 1 mM EDTA in PBS and cultured in DMEM supplemented with 10% FBS and 5% L929 cell-conditioned media. To obtain BMDCs, bone marrow cells were cultured in RPMI medium (Gibco; 11875-093) supplemented with 10% FBS (Gibco), GlutaMAX (Gibco; 35050-061), 1 mM HEPES (Cytiva; SH30237.01), and 20 ng/mL mouse recombinant granulocyte-macrophage colony-stimulating factor (GM-CSF) (PeproTech; 315-03). After 3 d of culture, additional media containing recombinant GM-CSF was added, and after 6 d, 50% of the supernatant was replaced by fresh media containing recombinant GM-CSF. Unattached cells were harvested on day 8 and used for experiments. RAW 264.7 murine macrophage cell line was cultured in DMEM supplemented with 10% FBS and 1 mM HEPES. THP-1, a human leukemia monocytic cell line, was cultivated using RPMI 1640 media supplemented with 10% FBS, GlutaMAX, 1 mM HEPES, and 1 mM sodium pyruvate. Cell viability was determined with a trypan-blue exclusion test, and cells were incubated overnight at 37 °C in a 5% CO2 incubator before use in experiments. Lentiviral transduction for knockdown For Snx5 knockdown in RAW 264.7 cells, a shRNA specific for Snx5 (Sigma-Aldrich; Mission small interfering RNA, TRCN0000381718 - GTACCGGTGCAGAGCTGCATCGACTTATCTCGAGATAAGTCGATGCAGCTCTGCATTTTTTG) and a control scrambled sequence were obtained from Sigma-Aldrich. To generate lentivirus particles, HEK293T cells were cotransfected with the lentiviral vectors and psPAX2 and pMD2 (packing plasmids) at a mass ratio of 4:3:1. After 72 h, HEK293T cell supernatant was collected and purified by centrifugation at 500 x g for 5 min and then filtered using a 0.45-μm filter. Then, RAW 264.7 cells were incubated with the virus-containing supernatant and 8 μg/mL polybrene (Milipore; TR-1003-G) for 48 h, followed by puromycin selection (3 μg/mL) (Fisher Scientific; BP2956-100). Western blot Cells were collected in MPER buffer (Thermo Fisher Scientific; 78501) supplemented with Protease Inhibitor (Roche, 11836170001) and incubated for 30 min at 4 °C. Next, cellular extracts were quantified using BCA assay (Thermo Fisher Scientific; 23225). A total of 60 μg of protein from each extract was loaded onto 4% to 20% SDS-PAGE for electrophoresis. Proteins were then transferred to a nitrocellulose membrane using a Bio-Rad Trans-blot Turbo apparatus for 7 min at 25 V, 1.3A. The membrane was then blocked with 5% milk powder (Lab Scientific bioKEMIX, 978-907-4243) in TBST (1× TBS + 0.1% Tween 20) for 1 h at room temperature. Membranes were then incubated overnight at 4 °C with anti-SNX5 (1:400, sc-515215; Santa Cruz), anti-SNX6 (1:400, sc-365965; Santa Cruz), anti-SNX32 (1:1000, 25763-1-AP; Proteintech), or anti-actin (1:5000, sc-47778; Santa Cruz) diluted in TBST + 1% milk solution. Each membrane was subsequently incubated with horseradish peroxidase–coupled anti-rabbit (Jackson; 111-035-003) or anti-mouse (Jackson: 115-035-174) IgG secondary antibodies. Blots were then developed with an ECL Chemiluminescence Kit (Bio-Rad; 170-5060) and detected using the ChemiDoc MP Imaging System. Densitometry quantification of the bands was performed using ImageJ software. Macrophage infection WT and Snx5−/− BMDMs were seeded and, after 3 h, where indicated, primed with 100 UI/mL IFN-γ for 24 h. For Mtb infection, the mid-log-phase bacterial culture was washed 3 times with PBS. The bacterial suspension was centrifuged at 300 x g to remove bacterial aggregates and then, to generate a single-cell suspension, sonicated using Branson digital sonifier (3 times for 7 s and 90% of amplitude). Cells were inoculated with Mtb, centrifuged at 1,500 rpm for 10 min, and then incubated at 37 °C, 5% CO2, and 95% humidity. After 30 min, cells were washed twice with PBS to remove noninternalized bacteria. Fresh media with or without IFN-γ was added, and the cells were incubated at 37 °C for different time points. Cell density and multiplicity of infection (MOI) are described in the following sections. Measurement of bacterial growth in macrophages WT and Snx5−/− BMDMs were seeded in 48-well plates (2 × 105), with or without IFN-γ (100 UI/mL) priming (R&D Systems; 485-MI), and infected with Mtb (MOI of 1). Cells were lysed on the same day of infection (day 0) and 1, 3, and 5 d postinfection with 0.5% Triton X-100 (Sigma-Aldrich; T8787) in water. Serial dilutions were plated on 7H11 plates, and CFUs were counted after 3 wk of incubation. Binding and phagocytosis assay The effect of Snx5 knockout on the binding and phagocytosis of Mtb was evaluated by CFU and by microscopic quantification. For the CFU experiment, BMDMs were plated in 48-well plates (2 × 105), while for microscopy cells were seeded in 24-well plates (2 × 105) containing coverslips. Initially, the cells, with or without IFN-γ (100 UI/mL) priming, were incubated for 10 min at 4 °C. Then, they were incubated with mCherry-Mtb (MOI of 1), centrifuged at 4 °C for 10 min at 500 x g, and incubated at 4 °C for another 30 min; this step allows binding without bacterial phagocytosis. After this period, the cells were washed to remove noninternalized Mtb. For the binding study, the cells were fixed with 1% PFA overnight at 4 °C or lysed with 0.5% Triton X-100, as described in Measurement of Bacterial Growth in Macrophages. For the phagocytosis study, the cells were reincubated with Dulbecco’s Modified Eagle Medium at 37 °C. After 1 h, the cells were washed and fixed with 1% PFA overnight at 4 °C or lysed with 0.5% Triton X-100. Fixed cells were labeled with Alexa Fluor 488–wheat germ agglutinin (1:1,000) (Invitrogen; W11261) for 1 h, then mounted using ProLong Diamond Antifade kit containing DAPI (Invitrogen; P36962), protected from light, and cured for 24 h at room temperature. Then, images were acquired on a Zeiss AxioImager Z2. For quantification, triplicate samples were analyzed for each condition, and a minimum of 100 cells were counted per sample using Imaris (7.4.0 version; Bitplane). Immunofluorescence WT and Snx5−/− BMDMs were plated in 24-well plates (2 × 105) containing coverslips, with or without IFN-γ (100 UI/mL) priming, and infected with mCherry-Mtb (MOI of 1). After 24 h, infected cells were fixed with 1% PFA overnight at 4 °C, and then transferred to PBS. Cells were permeabilized with cold methanol for 20 min at −20 °C, washed 3 times in PBS, and blocked with 3% BSA for 1 h at room temperature. Next, the cells were stained with primary antibodies in 3% BSA overnight at 4 °C. The primary antibodies used were anti-EEA1 (Abcam; ab2900, 1:250), and anti-LAMP (Novus Biologicals; NB120-19294, 1:50). Alexa Fluor 488 or 647–conjugated donkey anti-rabbit (1:1,000 or 1:500, respectively) was used as a secondary antibody. Finally, the coverslips were mounted using ProLong Gold Antifade kit containing DAPI (Invitrogen; P36962), protected from light, and cured for 24 h at room temperature. Z-stack images were acquired with a Nikon CSU-W1 SoRa spinning disk confocal microscope using a 100×/1.4 objective. Triplicate samples were quantified for each condition in a blinded analysis, and a minimum of 100 bacteria were counted per sample using Imaris (7.4.0 version; Bitplane). Cytokine quantification To assess cytokine production in Mtb-infected cells, WT and Snx5−/− BMDMs (106) were seeded in 12-well plates (106), primed or not with IFN-γ (100 UI/mL), and infected with Mtb (MOI of 5). After 24 h, the supernatant was collected and filtered (Pall Corporation; 8582). MCP-1 (Bio-Rad; 171G5019M), IL-6 (R&D Systems; M6000b), TNF-α (R&D Systems; MTA00B), and IL-12p40 (R&D Systems; MP400) were measured by ELISA according to the manufacturer’s instructions. Endolysosomal proteolytic assay Proteolytic activity in BMDMs was evaluated using DQ-BSA (DQ Green BSA, 2540586), a modified BSA that fluoresces upon cleavage in hydrolytic compartments, such as lysosomes. WT and Snx5−/− BMDM were plated in 24-well plates (2 × 105) containing coverslips, primed or not with IFN-γ (100 UI/mL), and infected with mCherry-Mtb (MOI of 1). After 24 h, cells were incubated with DQ-BSA (10 μg/mL) for 30 min. The cells were fixed with 1% PFA overnight at 4 °C, and then transferred to PBS. Coverslips were mounted, and images were acquired as described in Immunofluorescence. The intensity of DQ-BSA per pixel for at least 50 cells was analyzed using Imaris (7.4.0 version; Bitplane). Antigen presentation assay BMDMs, DCs, or RAW 264.7 cells were plated in 96-well (5 × 104) round-bottom plates (Thermo Fisher Scientific; 163320) with or without IFN-γ (100 UI/mL) priming. After overnight incubation, the cells were fed with vehicle or ovalbumin 323–339 (OVA323–339) (the epitope of OVA presented by I-Ab, 5 mg/mL) (InvivoGen; 5825-43-01), Ag85B (10 ng/mL) (Ray-Biotech, 228-11030 (discontinued) or BEI, NR5326), or heat-killed Mtb (MOI of 10) (Table 1). Then, cells were washed with PBS and cocultured with antigen-specific MHC class II–restricted T cell hybridomas (105) (Table 1) in a medium supplemented with 1 mM aminoguanidine hemisulfate (BioSynth FA170452; and Santa Cruz sc-202930). Do-11-10 cells (85082301) were from the European Collection of Authenticated Cell Cultures and BB7 hybridoma cells were a generous gift of Dr. Cliff Harding (Case Western Reserve University). After co-incubation, the supernatants were collected, and T cell activation was measured using IL-2 ELISA (Invitrogen; 88-7024-88). Table 1.Antigen presentation assay experimental setup.Antigen-presenting cellsAntigenIncubation timeT cell hybridomaTime of antigen-presenting cell–T cell coculture RAW 264.7 (I-Ad) OVA323–334 (5 mg/mL)4 hDO11.1024 h BMDMs (I-Ab) Ag85B (10 ng/mL)Heat-killed Mtb (MOI: 10) 18 hBB772 h DCs (I-Ab) Ag85B (10 ng/mL)Heat-killed Mtb (MOI: 10) 18 hBB772 h Antigen presentation assay experimental setup. Ag85B (10 ng/mL) Ag85B (10 ng/mL) Heat-killed Mtb (MOI: 10) Heat-killed Mtb (MOI: 10) Ag85B (10 ng/mL) Ag85B (10 ng/mL) Heat-killed Mtb (MOI: 10) Heat-killed Mtb (MOI: 10) Flow cytometry BMDMs (106), primed or not with IFN-γ (100 UI/mL), were washed twice with PBS and incubated with 1% BSA at 4 °C for 30 min to reduce nonspecific binding. Then, the cells were labeled with surface markers specific to MHC class II allophycocyanin (APC) (R&D Systems; FAB6118A, 10 µL per test), CD80 (FITC) (BD Biosciences; 553768, 0.25 µg per test), or CD86 (PE) (BD Biosciences; 553692, 0.06 µg per test) at 4 °C for 30 min. Subsequently, the cells were washed again with PBS and fixed with 1%PFA. 20,000 events were acquired on the BD FACSCalibur. Data analysis was performed using FlowJo 10 software. Click chemistry To compare the amount of peptide conjugated to MHC class II on the cell surface of BMDMs, we adapted a published protocol.52 Of note, this protocol uses an extended influenza A virus hemagglutinin (HA) peptide (HA318–338), which must be cleaved and loaded onto MHC class II for presentation on the cell surface. This peptide was designed based on the crystal structure of the HA322–334 association with the human HLA-DR1 MHC class II. Due to technical limitations in using PMA-activated THP-1 cells (human macrophage–like cells), we used BMDMs in our experiments. WT or Snx5−/− cells (106) were plated in 35 × 10 mm Petri dishes, primed with IFN-γ (100 UI/mL) for 24 h, and incubated with 50 mM of influenza HA318–338 in FBS-free media. Control cells were incubated with nonmodified peptide (YGACPKYVKQNTLKLATGMRN-nonclickable), and testing cells were incubated with a clickable version of the HA peptide, in which the lysine (K326) was converted into a propargylglycine (pra), a non-naturalized amino acid compatible with bio-orthogonal labeling using CalFluor488 (YGACPKYV(pra)QNTLKLATGMRN-clickable) (synthesized by GenScript). After 18 h, cells were gently detached by flushing with PBS and stained with Live/Dead fixable Aqua Blue (Invitrogen; L34957, 1:500) to assess cell viability. Cells were then fixed with 4% PFA for 15 min and incubated with quenching buffer (100 mM glycine and 100 mM NH4Cl in PBS) for 20 min to reduce autofluorescence. Next, the surface presentation of HA was labeled with 100 mM of CalFluor488 (Broadpharm; BP-24164) (diluted in a buffer containing 1 mM CuSO4, 0.5 mM THPTA, 5 mM sodium L-ascorbate, 0.01% BSA, and aminoguanidine 2.2%) for 1 h. Surface MHC class II were subsequently labeled with APC–conjugated anti-MHC class II (R&D Systems; FAB6118A, 10 µL per test). Samples were acquired on a BD FACSCanto, and data analysis was performed using FlowJo 10 software. RNA isolation for RNA sequencing WT and Snx5−/− BMDMs from 3 different animals were plated in 6-well plates (106) and infected or not with Mtb (MOI of 5). After 4 or 24 h of infection, samples were lysed using TRIzol (Invitrogen; 15596018) for 5 min, incubated with chloroform, and centrifuged at 12,000 x g for 5 min at 4 °C. Then, the aqueous phase was subject to RNA extraction using the RNeasy Mini Kit (Qiagen; 74104) according to the manufacturer’s protocol. RNA sequencing data acquisition and analysis RNA samples were transferred to McDermott Center Next Generation Sequencing Core at UT Southwestern Medical Center for RNA quality control evaluation, complementary DNA library preparation, and sequencing. The RNA quality was confirmed using Agilent TapeStation 4200 (RNA integrity number ≥8.0), and the initial RNA concentration was determined using Qubit. A total of 1 mg of DNase-treated RNA was used for library preparation using Illumina’s TruSeq Stranded mRNA Library Prep Kit. Sequencing was performed using Illumina NextSeq 2000 with a sequencing depth of 25 to 35 million reads and 75 bp single-end reads. Then, RNA sequencing reads were aligned to the Mus musculus reference genome. Finally, downstream analyses were performed using iDEP (integrated Differential Expression and Pathway analysis) online tools.53 Briefly, the data were preprocessed to remove low level expressed genes, and differential expression analysis between groups was performed using the DESeq2 R package with a false discovery rate threshold of 0.1. Genes with a fold change >2 found by DESeq2 were considered significant. Statistical analysis Statistical analysis and graphing were performed using GraphPad Prism software (version 10). After verifying data normality, an unpaired Student’s t test or Mann-Whitney test was used to compare two groups. Comparisons among 3 or more groups were assessed with 1-way analysis of variance for Gaussian distributions or the Kruskal-Wallis nonparametric test. Differences among groups were considered statistically significant when P < 0.05. Results Snx5 −/− mice succumb earlier than WT mice to Mtb infection despite similar organ bacterial burden To first test if Snx5 is required for host immune defense during TB, we used an established in vivo model of Mtb infection in which littermate Snx5−/− and congenic C57BL/6 mice were infected with a low dose (∼ 200 CFU) of virulent Mtb via aerosol (Fig. 1A).47–51 Compared with WT mice, Snx5−/− mice were more susceptible to Mtb infection, succumbing earlier to disease (Fig. 1B). This phenotype was not sex dependent because both male and female Snx5−/− mice showed increased susceptibility to infection (Fig. 1B). Surprisingly, lung, liver, and spleen CFUs were similar when comparing WT and Snx5−/− mice during both early (21 and 42 d postinfection) and chronic stages (90, 120, and 150 d postinfection) of infection (Fig. 1C). These results suggest that the absence of SNX5 leads to increased mouse mortality during Mtb infection independent of increased bacterial burden. Snx5 −/− mice succumb to disease earlier than WT mice without affecting bacterial replication. (A) WT and Snx5−/− mice were infected with approximately 200 CFUs of Mtb Erdman. (B) Kaplan-Meier curve of survival of Mtb-infected mice. (C) WT and Snx5−/− mice were infected with approximately 200 CFUs of Mtb Erdman, and lung, spleen, and liver CFUs were determined at 7, 21, 42, 90, 120, and 150 d. (D) Bacterial CFUs in lungs, liver, and spleens of WT or Snx5−/− mice at the indicated time point after infection with Mtb. Each symbol corresponds to 1 animal (Mann-Whitney test). Mtb-induced pulmonary inflammation is increased by Snx5 deletion without increasing the influx of immune cells Because mouse mortality can be impacted by gene deletion in the absence of differences in Mtb burden,54 a phenotype frequently related to inflammatory immunopathology,55 we analyzed inflammation in lung sections from Mtb-infected WT and Snx5−/− mice. Using quantification of hematoxylin and eosin–stained lung sections from WT and Snx5−/− mice (Fig. 2A; Fig. S1), we observed that the percentage of inflamed lungs reached a plateau after 90 d postinfection in WT mice, while inflammatory lung regions in Mtb-infected Snx5−/− mice increased over time. At 150 d, lung inflammation was significantly increased in the lungs of Snx5−/− mice compared with WT littermate control mice (Fig. 2B). Mtb-induced pulmonary inflammation is increased by Snx5 deletion without increasing the influx of immune cells. (A) Representative photomicrographs of hematoxylin and eosin–stained lungs at 42, 90, 120, and 150 d postinfection with Mtb. Scale bars = 1,000 μm. (B) Quantitation of inflammatory areas in lungs of Mtb-infected mice at the indicated time point after infection with Mtb. Measurement was performed with ImageJ. Results are the mean ± SD for 5 to 10 animals per group. *P < 0.05 (for indicated comparison; Mann-Whitney test). (C–I) Percentage of (C) neutrophils (Ly6G+ CD11b+), (D) monocytes (CD11b+ Ly6C+), (E) macrophages (F4/80+), (F) DCs (CD11c+), (G) T cells (CD3+), (H) CD4+ T cells (CD3+ CD4+), and (I) CD8+ T cells (CD3+ CD8+) in the lung of mice during low-dose Mtb infection. Each circle corresponds to 1 animal. Bars are the mean ± SD for 3 to 5 animals per group per time point. Statistical analysis was by 1-way analysis of variance. To investigate whether the increased inflammation in the lung of Snx5−/− mice was due to increased lung-infiltrating populations during Mtb infection, lung cells from WT and Snx5−/− mice infected or not with Mtb were analyzed by flow cytometry using myeloid and lymphocyte markers. There was no difference in the influx of neutrophils (Ly6G+ CD11b+), monocytes (CD11b+ Ly6C+), macrophages (F4/80+), DCs (CD11c+), or T cells (CD3+) into the lungs of WT and Snx5−/− mice (Fig. 2C–I; Fig. S2). Because we did not observe a difference in the influx of CD3+ T cells to the lungs, we next investigated whether Snx5 deletion impacted the ability of immune cells and T cells to produce proinflammatory cytokines. We first measured cytokine levels in lung homogenates of mice. No difference was observed in the production of IFN-γ, TNF-α, or IL-2 (Fig. 3A–C). However, Snx5 deficiency was associated with decreased production of IL-12p40 in the lungs of Mtb-infected mice (Fig. 3D). Next, to determine if production of cytokines differed in explanted T cells, ex vivo lung cells and splenocytes activated with CD3 and CD28 were processed for intracellular staining and analyzed by flow cytometry to characterize cytokine production by CD4+ T cells (Fig. S3). As expected, the production of IFN-γ and TNF-α was increased in lung CD4+ T cells from mice infected with Mtb (Fig. 3E, F). However, production of IFN-γ, TNF-α, and IL-2 was comparable between lung CD4+ T cells from WT and Snx5−/− mice (Fig. 3E–G). Similar results were observed in ex vivo splenocyte derived CD4+ T cells from WT and Snx5−/− mice (extended data, Fig. 4). Together, these findings indicate that Snx5 deletion is associated with increased lung inflammation during Mtb infection without altering immune cell recruitment or CD4+ T cell cytokine production, suggesting a role for SNX5 in limiting inflammation-driven immunopathology. Snx5 deficiency decreases the production of IL-12p40 in lung homogenates of Mtb-infected mice. (A) IFN-γ, (B) TNF-α, (C) IL-2, and (D) IL-12p40 were assessed in supernatants of homogenized lungs by ELISA. **P < 0.01 (for indicated comparison; Mann-Whitney test). (E–G) Lung explanted CD4+ T cells were characterized for cytokine expression, including (E) IFN-γ, (F) TNF-α, and (G) IL-2 using surface and intracellular cytokine staining for flow cytometry. Each circle corresponds to 1 animal. Bars are the mean ± SD for 3 to 5 animals per group per time point. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (for indicated comparison; 1-way analysis of variance). Snx5 deletion does not impact macrophage cell-autonomous activity in controlling Mtb infection. (A) Representative Western blot to detect SNX5 levels in BMDMs from WT and Snx5−/− mice. (B) Description of the binding and phagocytosis experimental procedures. (C–H) WT and Snx5−/− BMDMs either primed or not with IFN-γ, were infected with mCherry-Mtb (red). (C–E) Binding and (F–H) phagocytosis were analyzed by confocal microscopy with quantification of bacterial number and by quantification of bacterial CFUs. For confocal microscopy, cell nuclei were labeled with DAPI (blue). Scale bars = 20 µm. Bars are the mean ± SEM. from 3 independent experiments. The circles shown correspond to each experiment. (I) Mycobacterial growth in WT or Snx5−/− BMDMs infected with Mtb, either primed or not with IFN-γ. Bars are the mean ± SEM from 3 independent experiments. The circles shown correspond to each experiment. *P < 0.05 (for indicated comparison; 2-way analysis of variance). ACTB, actin. Snx5 deletion does not impact macrophage cell-autonomous immunity during Mtb infection Because the role of SNX5 during Mtb infection has not been previously explored, we next assessed whether Mtb modulates SNX5 expression in BMDMs. At 4 and 24 h postinfection, no significant changes in SNX5 expression were observed in either resting or IFN-γ-activated macrophages (Fig. S5A, B). We also investigated whether Snx5 deletion alters the expression of SNX6 and SNX32, closely related members of the SNX-BAR family known for their high homology with SNX5 and interchangeable roles.56 We observed that in BMDM from Snx5−/− mice (Fig. 4A), Snx5 deletion did not impact the expression of SNX6 or SNX32 at 24 h postinfection (Fig. S5C–E). To test whether Snx5 has a cell-autonomous role in controlling Mtb infection, we infected BMDMs from WT or Snx5−/− mice and quantified mycobacterial binding, phagocytosis, and intracellular growth. To investigate the impact of Snx5 deletion on Mtb binding, BMDMs were kept at 4 °C, allowing bacteria-cell interaction without bacterial phagocytosis (Fig. 4B). Mtb binding and phagocytosis were evaluated by microscopic quantification (Fig. 4C, D, F, and G) and by CFUs (Fig. 4E, H). No difference in Mtb binding was observed between BMDMs from WT or Snx5−/− mice, either primed or not primed with IFN-γ (Fig. 4C–E). Similarly, Snx5 deletion did not impact Mtb phagocytosis by BMDMs (Fig. 4F–H). Additionally, after 5 d of infection, IFN-γ-activated BMDMs from WT or Snx5−/− mice controlled Mtb replication to a similar extent (Fig. 4I). Due to the role of SNX5 in membrane trafficking,57 we next evaluated the impact of Snx5 deletion on the maturation of the Mtb-containing vacuole. We analyzed colocalization of Mtb with markers of early endosomes (EEA1) and lysosomes (LAMP1) using immunofluorescence microscopy of fixed cells. Specifically, we infected WT and Snx5−/− macrophages with mCherry-Mtb and fixed cells 24 h after infection. As shown in Fig. 5, the recruitment of EEA1 (Fig. 5A, B) and LAMP1 (Fig. 5C, D) to Mtb-containing vacuoles was similar in Snx5−/− and WT macrophages, despite priming with IFN-γ. These results suggest that deficiency of Snx5 does not impact macrophage cell-autonomous immunity during Mtb infection. Snx5 deletion does not alter the recruitment of EEA1 and LAMP1 to the Mtb-containing vacuole. WT and Snx5−/− BMDMs, either primed or not with IFN-γ, were infected with mCherry-Mtb (red). After 24 h, cells were fixed and stained for (A, B) EEA1 or (C, D) LAMP1 (green). For confocal microscopy, cell nuclei were labeled with DAPI (blue). Representative photomicrographs (A, C) and quantitation (B, D) of the colocalization of mCherry-Mtb and EEA1 (A, B) or LAMP1 (C, D). For each coverslip, at least 100 bacteria were evaluated. Scale bars = 10 μm. Bar graphs are the mean ± SD from 1 experiment performed in quadruplicate. The circles shown correspond to each replica. Statistical analysis was by t test. Snx5 deletion does not result in broad transcriptional changes in macrophages during Mtb infection To understand how SNX5 might impact mouse susceptibility to Mtb infection, we used bulk RNA sequencing to investigate its impact on macrophage transcriptional responses to Mtb infection. To that end, we compared the transcriptomes of WT and Snx5−/− BMDMs, uninfected or Mtb-infected, at 4 and 24 h postinfection. Principal component analysis clearly separated infected and uninfected cells. However, it does not distinguish between WT and Snx5−/− cells (Fig. S6A). Genome mapping identified 152 and 187 significantly upregulated genes and 4 and 41 significantly downregulated genes in the WT-infected cells compared with uninfected WT cells after 4 and 24 h of infection, respectively (Fig. S6B). In Mtb-infected Snx5−/− cells, 208 and 300 genes were upregulated, while only 2 and 55 were downregulated after 4 and 24 h of infection, respectively, compared with uninfected Snx5−/− cells (Fig. S6B). As expected, in both WT and Snx5−/− murine BMDMs infected with Mtb, we observed upregulation of classical immunologically significant genes. These included several chemokines (e.g. Ccl2, Ccl3, Ccl4, Ccl5), cytokines (e.g. Il1A, Tnf), signaling molecules (e.g. Ifh1, Nod2, Sqstm1 [p62]), IFN-stimulated genes (e.g. Ifit1, Isg20, Igtp, Iigp1), and transcription factors (e.g. Irf1, Irf7, Stat2) (Fig. S6C). Gene Ontology biological process enrichment analysis further revealed that among upregulated genes were genes with described roles in responses to cytokines, defense responses, and responses to external biotic stimuli (Fig. S6D). However, consistent with the principal component analysis, macrophage transcriptomes of Snx5−/− and WT BMDMs, either infected or not with Mtb, did not demonstrate any meaningful differences between the two genotypes (Fig. S6D). Only 1 differentially expressed gene was identified between Snx5−/− and WT BMDMs at 4 h (Cryab, downregulated in Snx5−/− uninfected cells) and 24 h (Capn11, upregulated in Snx5−/−–infected cells) after infection (Fig. 6C; Fig. S6B). Thus, we conclude that Snx5 deletion does not result in broad transcriptional changes in macrophages during Mtb infection. Lack of SNX5 modestly affects proinflammatory cytokine production by BMDMs. (A) TNF-α, (B) IL-12p40, (C) MCP-1, and (D) IL-6 were assessed in cell supernatants of infected macrophages, either primed or not with IFN-γ (100 UI/mL). Bars are the mean ± SD from 1 experiment performed in quadruplicate. The circles shown correspond to each replica. *P < 0.05, ****P < 0.0001 (for indicated comparison; 1-way analysis of variance). Lack of Snx5 modestly affects proinflammatory cytokine production by BMDMs Because we observed that Snx5 deletion decreased IL12p40 production in lung homogenates of Mtb-infected mice, we next assessed if the production of proinflammatory cytokines by Mtb-infected macrophages in vitro is affected by Snx5 deletion (Fig. 6). In IFN-γ–activated, but not resting, macrophages, Snx5 deletion resulted in increased TNF-α secretion (Fig. 6A) and, consistent with our in vivo study, decreased IL-12p40 secretion (Fig. 6B). No differences were observed in the production of MCP-1 or IL-6 (Fig. 6C, D). Snx5 deficiency impairs MHC class II–restricted antigen presentation by macrophages and DCs Because Snx5 has previously been shown to impact antigen processing44 and antigen presentation in B cells,58 we hypothesized that Snx5 is required for antigen presentation during Mtb infection. We first assessed whether Snx5 deficiency impacts antigen processing in our model. To this end, resting or IFN-γ-activated WT and Snx5 −/− BMDMs were infected or not with Mtb and incubated with DQ-BSA, a marker of degradative compartments. Consistent with a prior report,44 Snx5 deletion reduced DQ-BSA intensity in both Mtb-infected and uninfected BMDMs, regardless of IFN-γ priming (Fig. 7A, B). However, no difference was observed in the colocalization of Mtb and DQ-BSA between WT and Snx5 −/−–infected cells (Fig. 7C). Together, these results indicate that Snx5 deficiency impairs lysosomal degradative capacity without altering Mtb delivery to degradative compartments. Snx5 deletion mildly reduced DQ-BSA intensity in macrophages. (A–C) WT and Snx5−/− BMDMs, either primed or not with IFN-γ, were infected or not with mCherry-Mtb (red). After 24 h, cells were incubated with DQ-BSA (green) for 30 min and then fixed. For confocal microscopy, cell nuclei were labeled with DAPI (blue). (A) Representative photomicrographs. Scale bars = 10 μm. (B) Quantitation of DQ-BSA intensity per pixel. For each coverslip, at least 50 cells were evaluated. Lines represent the median and violin plot quartiles (25% and 75%) *P < 0.05, ****P < 0.0001 (for indicated comparison; Mann-Whitney test). (C) Quantification of the colocalization of mCherry-Mtb and DQ-BSA. For each coverslip, at least 100 bacteria were evaluated. Bars are the mean ± SD from 1 of 3 experiments performed in triplicate. The circles shown correspond to each replica. Statistical analysis was by Mann-Whitney test. We next evaluated the functional consequences of Snx5 deletion on antigen presentation. We incubated macrophages with the known T cell epitope OVA323–33959,60 the full-length Mtb protein Ag85B,15 or heat-killed Mtb and then cocultured them with CD4+ T cell hybridoma cells with specificity against either OVA323–339 (DO11.10)59,60 or Ag85B (BB7).15 We then quantified IL-2 accumulation in the supernatant as a measure of CD4+ T cell activation59,61 (Fig. 8A). Deletion of Snx5 in BMDMs resulted in reduced IL-2 production by MHC class II–restricted CD4+ T cells when IFN-γ-stimulated macrophages were incubated with full-length Ag85B (Fig. 8B; Fig. S7) and heat-killed Mtb (Fig. 8C; Fig. S7). Snx5 deficiency impairs antigen presentation by macrophages and DCs. (A) Timeline of antigen presentation assay. (B, C) BMDMs, (D) RAW 264.7, or (E, F) DCs, either primed or not with IFN-γ, were fed with OVA 323–339, Mtb Ag85B, or heat-killed Mtb as depicted on the graphs. After (D) 4 h or (B, C, E, and F) 18 h, the cells were incubated with antigen-specific MHC class II–restricted T cell hybridomas. After 24 h, supernatants were assessed for IL-2 secretion by T cells. Bars are the mean ± SD from 1 of 3 experiments performed at least in triplicate. The circles shown correspond to each replica. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (for indicated comparison; t test). Importantly, reduced T cell activation was also observed when RAW 264.7 cells were provided with the mature MHC class II peptide OVA323–339 (Fig. 8D; Fig. S7), suggesting that the defect is not solely attributable to impaired antigen processing. Because DCs are the most efficient antigen-presenting cells, we next investigated whether Snx5 was also required for antigen presentation by DCs. As expected, DCs did not require IFN-γ to induce efficient antigen presentation, and like macrophages, Snx5 deletion reduced antigen presentation by approximately 40% (Fig. 8E, F; Fig. S7). Snx5 deficiency reduces peptide loading onto MHC class II without altering surface expression of antigen presentation machinery Expression of peptide-MHC complexes on the surface of antigen-presenting cells is essential for efficient antigen presentation to T cells. To determine whether SNX5 deletion affects the MHC class II antigen presentation machinery, we assessed the expression of MHC class II and the costimulatory molecules, CD80 and CD86, on the surface of Snx5−/− and WT BMDMs. No differences were observed in the cell surface expression of MHC class II (Fig. 9A; Fig. S8A), CD80 (Fig. 9B; Fig. S8B), and CD86 (Fig. 9C; Fig. S8C) comparing Snx5−/− and WT BMDMs. Snx5 deletion reduces virus HA peptide loading onto MHC class II without affecting surface expression of MHC class II and costimulatory molecules. (A–C) Expression of (A) MHC class II, (B) CD80, and (C) CD86 molecules on the surface of WT or Snx5−/− BMDMs, either primed or not with IFN-γ (100 UI/mL), analyzed by flow cytometry. Bars are the mean ± SD from 3 independent experiments. The circles shown correspond to each experiment. (D-F) IFN-γ–activated BMDMs were incubated with HA peptides for 18 h and then labeled with CalFluor 488. (D) Representative figure of the click chemistry assay to detect HA-peptide loaded on MHC class II, adapted from Baranov et al.52 BMDMs were incubated with the extended influenza A virus HA peptide (HA318–338) containing a non-naturalized amino acid (a “clickable” peptide) for 18 h. This peptide must be cleaved and loaded onto MHC class II for presentation on the cell surface, where it can be detected by CalFluor 488. (E) Representative quantification of HA-CalFluor-488 quantification (geometric mean fluorescence intensity [gMFI]). Bars are the mean ± SD from 1 of 5 independent experiments. (F) Quantification of HA-CalFluor-488 (gMFI) normalized to control WT cells. Circles represent individual experiments. **P < 0.01 (for indicated comparison; t test). We therefore hypothesized that the reduced T cell activation observed in Snx5−/− BMDMs was associated with impaired peptide loading to MHC class II. To test this, we adapted a protocol whereby cells are incubated with an HA-peptide containing an unnatural amino acid, propargylglycine, that permits detection of surface-exposed peptide-MHC class II complexes via click chemistry to a fluorescent dye (Fig. 9D).52 In WT control cells incubated with the nonclickable HA318–338 peptide (lacking propargylglycine), the CalFluor 488 signal was approximately 2-fold lower compared with the cells incubated with the clickable peptide (Fig. 9E; Fig. S8D). Notably, CalFluor background labeling did not differ between WT and Snx5−/− BMDMs. Among cells incubated with the clickable peptide, we observed a reduction of approximately 30% in CalFluor 488 signal in Snx5−/− BMDMs compared with WT control mice (Fig. 9E, F; Fig. S8D). Taken together, these data indicate that Snx5 deletion impairs peptide loading onto MHC class II molecules, resulting in reduced levels of peptide-loaded MHC class II at the cell surface without altering total surface expression of MHC class II or costimulatory molecules. Discussion In this study, we identify SNX5 as an important regulator of host defense during Mtb infection. Although Snx5 deficiency did not alter bacterial burden, Snx5−/− mice were more susceptible to Mtb infection, a phenotype associated with exacerbated pulmonary inflammation. At the cellular level, we show that Snx5 deletion impairs antigen presentation to MHC class II–restricted CD4+ T cells by both macrophages and DCs, without affecting total MHC class II or costimulatory molecule surface expression, but instead reducing the amount of peptide-loaded MHC class II on the surface of antigen-presenting cells. Together, our findings reveal a previously unappreciated role for SNX5 in regulating Mtb immune responses and inflammation-driven immunopathology. The increased mortality of Snx5−/− mice in the absence of higher bacterial loads suggests that factors beyond bacterial replication contribute to disease progression. Similar phenotypes have been observed in other models of Mtb infection, where loss of disease tolerance rather than pathogen burden drives mortality.54,62,63 For example, deletion of Cybb (gp91), a subunit of the PHOX multiprotein complex, leads to increased mouse mortality due to excessive lung inflammation caused by increased caspase-1 activation and neutrophil infiltration.63 Likewise, Ppif (cyclophilin D) knockout mice exhibit increased T cell proliferation associated with enhanced T cell metabolic activity.54 In our study, histological analysis revealed that Snx5−/− mice exhibited more extensive lung inflammation than WT control mice, consistent with an inability to balance protective and pathological immune responses. Notably, the exacerbated inflammation was not associated with increased influx of immune cells into the lung, and we observed only mild effects on proinflammatory cytokine production in lung homogenates. Among the cytokines analyzed, IL-12p40 levels were reduced in the lungs of Mtb-infected Snx5−/− mice and in IFN-γ-activated Snx5−/− macrophages. Because IL-12p40 is shared by both IL-12 and IL-23 and can also exist as a monomer or homodimer, our data do not distinguish which p40-containing cytokine pathway is affected. Nevertheless, IL-12p40 plays important roles in host defense against Mtb, including promoting type 1 immune responses and coordinating adaptive immunity through DC migration and CD4+ T cell priming.64–67 Whether altered production of IL-12p40 contributes to the increased susceptibility and pulmonary inflammation observed in vivo remains an important open question. Thus, further studies will be required to elucidate how Snx5 regulates lung inflammation during Mtb infection. Because we used whole-body knockout mice, it is also possible that nonimmune cells contributed to the phenotype reported here. Our flow cytometric analysis did not include natural killer cell–specific markers; therefore, we cannot exclude a contribution of natural killer cells to the inflammatory phenotype observed in Snx5-deficient mice. At the cellular level, we found that Snx5 deletion has no impact on macrophage cell-autonomous response against Mtb, including bacterial uptake. Previous studies have reported that SNX5 is required for efficient macropinocytosis of soluble albumin44 and contributes to M. bovis uptake in IFN-γ–activated macrophages.45 However, the latter study did not directly assess the effect of genetic Snx5 deficiency on bacterial uptake. In contrast, using Snx5-deficient macrophages, we observed no effect on Mtb binding or uptake. Thus, the findings are not directly comparable and may also reflect differences in experimental approaches and in the mycobacterial species studied. Functional redundancy among SNX family members in genetically deficient cells and pathogen-specific differences in the utilization of endocytic pathways may also contribute to the observed differences. More broadly, the role of SNX5 during infection appears to vary across pathogens and cell types. For example, depletion of SNX5/6 increases Chlamydia trachomatis progeny in epithelial cells,68 whereas SNX5 restricts infection by several viruses, including Sindbis virus, herpes simplex virus type 1, West Nile virus, and chikungunya virus, in epithelial cells and fibroblasts through an autophagy-dependent mechanism.46 Together, these findings suggest that SNX5’s role during infection is both pathogen- and cell type-specific, likely reflecting differential engagement of membrane trafficking and autophagic pathways by distinct intracellular pathogens. Our finding that Snx5 deletion reduces antigen presentation adds to growing evidence implicating SNX family members in endomembrane trafficking and immune signaling.69–73 Previous studies showed that SNX5 supports antigen processing in macrophages44 and participates in antigen presentation in B cells by coordinating actin remodeling and lysosomal dynamics.58 Consistent with these reports, we observed that Snx5 deficiency mildly reduced degradation of DQ-BSA, suggesting impaired endolysosomal proteolysis, and decreased peptide-loaded MHC class II at the cell surface. Importantly, the lack of changes in the total MHC class II and costimulatory molecule expression suggests that Snx5 deletion affects the generation and/or trafficking of peptide-MHC class II complexes, rather than their transcriptional regulation. This interpretation is further supported by our transcriptomic analysis, which revealed no differences in gene expression between Snx5-knockout and WT BMDM control mice. Successful antigen presentation depends on multiple coordinated events, ranging from antigen acquisition to the display of peptide-MHC complexes on the cell surface.74 Notably, the antigen presentation defect persisted when cells were provided with a preprocessed MHC class II peptide, suggesting that SNX5 regulates steps beyond antigen processing alone. Based on our data, SNX5 may facilitate antigen presentation through multiple, non–mutually exclusive pathways. First, SNX5 may impact peptide loading within the MHC class II compartment, potentially by regulating endosomal-lysosomal dynamics required for optimal antigen presentation, consistent with the reduction in proteolytic activity observed in Snx5-deficient cells.44 Second, SNX5 may affect postloading trafficking of peptide-MHC class II complexes to the plasma membrane, as SNX5 has been shown to participate in the recycling and retrograde trafficking of several membrane receptors.38,39,75,76 Whether SNX5’s effect on antigen presentation depends on its association with the retromer complex or reflects a retromer-independent role remains an open question. As previously noted, we observed increased susceptibility of Snx5−/− mice to Mtb infection, associated with higher pulmonary inflammation that was not accompanied by increased influx of neutrophils, macrophages, DCs, monocytes, or T cells. A recent study demonstrated that the spatial organization of inflammatory lesions, rather than cell numbers, can dramatically influence disease outcomes during Mtb infection.77 In this context, combining our in vivo findings with the observed reduction in antigen presentation in Snx5−/− cells raises the possibility that SNX5 may influence the spatial organization or local interactions of immune cells in the lung, thereby shaping the quality and localization of inflammatory response against Mtb. A limitation of our study is that we did not establish the molecular mechanism by which SNX5 regulates peptide loading and presentation. Although the clickable-peptide assay indicates reduced surface peptide-MHC class II complexes, it remains unclear whether this defect arises from peptide exchange dynamics, changes in vesicular trafficking, or both. In addition, because this assay used an influenza A–derived peptide, developing Mtb-specific clickable peptides will be necessary to determine whether SNX5 selectively affects the presentation of particular mycobacterial epitopes or more broadly impacts antigen presentation. Furthermore, we did not directly assess whether SNX5 impacts T cell priming in vivo or how this relates to the increased mortality in Snx5-deficient mice, which remain important questions for future investigation. In conclusion, we demonstrate that SNX5 is a key regulator of the host immune response to Mtb, promoting efficient antigen presentation and controlling inflammation without affecting bacterial burden. These findings extend the known functions of Snx5 beyond endosomal trafficking and antiviral defense to include regulation of antibacterial immunity and immunopathology. Supplementary Material