Mpeg1 is not essential for antibacterial or antiviral immunity, but is implicated in antigen presentation

Abstract

To control infections phagocytes can directly kill invading microbes. Macrophage-expressed gene 1 (Mpeg1), a pore-forming protein sometimes known as perforin-2, is reported to be essential for bacterial killing following phagocytosis. Mice homozygous for the mutant allele Mpeg1^tm1Pod succumb to bacterial infection and exhibit deficiencies in bacterial killing in vitro. Here we describe a new Mpeg mutant allele Mpeg1^tm1.1Pib on the C57BL/6J background. Mice homozygous for the new allele are not abnormally susceptible to bacterial or viral infection, and irrespective of genetic background show no perturbation in bacterial killing in vitro. Potential reasons for these conflicting findings are discussed. In further work, we show that cytokine responses to inflammatory mediators, as well as antibody generation, are also normal in Mpeg1^{tm1.1Pib/tm1.1Pib} mice. We also show that Mpeg1 is localized to a CD68-positive endolysosomal compartment, and that it exists predominantly as a processed, two-chain disulfide-linked molecule. It is abundant in conventional dendritic cells 1, and mice lacking Mpeg1 do not present the model antigen ovalbumin efficiently. We conclude that Mpeg1 is not essential for innate antibacterial protection or antiviral immunity, but may play a focused role early in the adaptive immune response.

Keywords: dendritic cells; inflammation; macrophage expressed gene; monocytes and macrophages; mpeg1; perforin 2.

Figure 1

Gene targeting of Mpeg1. (a) The Mpeg1 locus, Mpeg1 structure and predicted allele structure after insertion of loxP into the 5′ UTR and the selectable marker (neo) and loxP into the 3′ UTR [Mpeg ^tm1Pib (targ)]. Subsequent removal of the Mpeg1 coding sequence and neomycin (neo) cassette by cre‐mediated recombination yields the knockout allele [Mpeg1 ^tm1.1Pib (flox)]. Sequences cloned and used as 5′ or 3′ flanking probes (pr) are indicated. Coordinates are from the mouse genomic reference sequence GRCm38.p3 C57/BL6. (b) Validation of ES cell and F1 mouse genotypes by Southern blotting. The predicted sizes of genomic fragments generated following EcoRV or AflII treatment and recognized by the indicated 5′ and 3′ probes are shown in a. (c) Validation of cre‐mediated Mpeg1 deletion in a floxed mouse. Shown is the relevant DNA sequence of the 428‐bp flox/flox PCR product amplified using the primers shown as closed arrowheads in the Mpeg1 ^tm1.1Pib allele shown in a. These primers yield a 2800‐bp product from WT mouse genomic DNA (upper panel). (d) Protein extracts from 10⁶ BMDMs derived from WT or knockout (flox/flox) mice were separated by 10% SDS–PAGE and immunoblotted for Mpeg1 (1:2000 rabbit anti‐Mpeg), followed by immunoblotting for GAPDH (1:5000 anti‐GAPDH). These were compared with extracts from COS‐1 cells transiently expressing Mpeg1, and from the MutuDC line (10⁴ cells). Image shown represents three independent experiments. (e) Total RNA from unstimulated or IFNγ‐ and LPS‐stimulated BMDMs derived from WT or knockout (flox/flox) mice was assessed via RT‐PCR for the presence of Mpeg1, Pfpl, Dtx4 and housekeeping Gapdh transcripts. ES cell, embryonic stem cell; BMDM, bone marrow‐derived macrophage; IFN, interferon; LPS, lipopolysaccharide; Mpeg1, macrophage‐expressed gene 1; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; RT‐PCR, reverse‐transcriptase PCR; UTR, untranslated region; WT, wild type. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 2

Processing of Mpeg1 during biosynthesis. (a) MutuDCs were pulse labeled with 50 μCi ³⁵S‐met for 30 min, chased for the indicated times and then lysed in 0.5% (w/v) SDS. Samples were immunoprecipitated with protein G–Sepharose and rabbit anti‐Mpeg1, then resuspended in LSB/0.1 M DTT and separated by 10% SDS–PAGE. The experiment was performed three times. (b) Immunoprecipitated Mpeg1 was analyzed via proteomics. Scaled domain structure of Mpeg1 follows Pang et al. Asterisks indicate N‐glycosylation site; Nontryptic cleavages identified are indicated beneath. The height of the mark reflects the number of times each cleavage was identified across all runs. Also shown are predicted molecular weights based on the major cleavage sites. (c) BMDMs were pulse labeled with 100 μCi ³⁵S‐met for 60 min and chased for the indicated times. At each point, the medium was removed and separately analyzed for Mpeg1. Samples were immunoprecipitated using guinea pig Gp202 antiserum (antibody) and protein A–Sepharose. Before SDS–PAGE each sample was divided into two. LSB/DTT was added to one, and LSB to the other. The experiment was performed two times. (d) BMDMs were stained with 1:400 Gp202 and 1:800 anti‐Gp‐AF568 and 1:400 anti‐CD‐68‐AF647. Images are single optical slices representing five independent experiments. Ab, antibody; BMDM, bone marrow‐derived macrophage; DTT, dithiothreitol; LSB, Laemmli sample buffer; MABP, multivesicular body of 12‐kDa‐associated β‐prism; MACPF, membrane attack complex/perforin domain; L, L‐domain; Mpeg1, macrophage‐expressed gene 1; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; SP, signal peptide; TM, transmembrane domain. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 3

Mpeg1 is not essential for bacterial killing by BMDMs in vitro. BMDMs were infected with (a) Mycobacterium smegmatis (MOI = ~10–15), (b) Legionella pneumophila (MOI = ~10), (c) Staphylococcus aureus (MOI = ~5–10) and (d) Escherichia coli K12 (MOI = 30–50). Macrophages infected with M. smegmatis were permeabilized and stained for Mpeg1 and bacteria (see Supplementary figure 6a for details). At different times after infection, intracellular bacteria were released by lysing the cells in water and enumerated. Statistical significance was assessed using the Student's unpaired t‐test and showed no significant differences between groups. Data are mean ± standard deviation of the percentage of survival at each time point obtained from two (Legionella) or at a minimum of three independent experiments. BMDM, bone marrow‐derived macrophage; DAPI, 4′,6‐diamidino‐2‐phenylindole; KO, knockout; MOI, multiplicity of infection; WT, wild type. [Colour figure can be viewed at wileyonlinelibrary.com]

Figure 4

Increased full‐length Mpeg1 after BMDM activation. Day 9 BMDMs were activated overnight with LPS (10 ng mL⁻¹) and/or IFNγ (100 ng mL⁻¹). Extracts from 1 × 10⁶ cells were resolved by 10% SDS–PAGE and transferred to nitrocellulose. (a) The membranes were sequentially probed with anti‐MPEG1 antibody and rabbit anti‐GAPDH. The figure shown is representative of two independent experiments. (b, c) The fold increase of Mpeg1 on activation was evaluated via densitometry. The Student's t‐test was used to compare outcomes. *P < 0.05. BMDM, bone marrow‐derived macrophage; FL, full length; IFN, interferon; LPS, lipopolysaccharide; Mpeg1, macrophage‐expressed gene 1; P, processed form; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis.

Figure 5

Mpeg1 is not required for inflammasome activation. BMDMs were treated with the indicated activators and cytokine release was measured by ELISA. Pyroptosis was evaluated via measurement of LDH release. B. cer., Bacillus cereus; BMDM, bone marrow‐derived macrophage; C. rod., Citrobacter rodentium; E. coli., Escherichia coli; Flu, influenza virus; IAV, influenza A virus; IL, interleukin; LDH, lactate dehydrogenase; LPS, lipopolysaccharide; LPS trans., transfected LPS; Med., medium; Mpeg1, macrophage‐expressed gene 1; Nig, nigericin; TNF, tumor necrosis factor. The experiment was carried out three times.

Figure 6

Mpeg1 is not essential for bacterial clearance. WT or Mpeg1 ^−/− mice were infected with the indicated pathogens. Tissue homogenates were serially diluted and aliquots plated on appropriate media. Colonies were counted and reported as CFU/gram of tissue. (a) Six mice of each genotype infected by aerosolized Mycobacterium tuberculosis (~100–200 CFU) were killed 8 weeks after the infection and CFU in lungs and spleen assessed. Weight loss was evaluated at the indicated times after the infection. The experiment was carried out once. (b) Ten mice of each genotype were infected with 1 × 10⁷–5 × 10⁷ CFU of Staphylococcus aureus USA 300 via intranasal inhalation. At the indicated times, CFU in lungs were evaluated, and the proportions of neutrophils (Gr⁺) and macrophages (CD11b⁺, F4/80⁺) in lungs compared. The experiment was carried out three times. (c) Eight mice of each genotype were inoculated intranasally with 4 × 10⁴ CFU of Legionella longbeachae. CFU in lungs was assessed 3 days after the infection. The experiment was carried out once: each data point represents a single mouse. Statistical significance was assessed using the Student's t‐test. ns = P > 0.05. CFU, colony forming units; KO, knockout; Mpeg1, macrophage‐expressed gene 1; WT, wild type.

Figure 7

(a) Mpeg1 is not involved in the control of LCMV infection. Nine mice of each genotype were infected with 2 × 10⁶ FFU of LCMV Docile strain intravenously and viral titers in tissues assessed at day 8 or 28. Each point indicates a single mouse. The experiment was carried out two times. (b) Mpeg1 does not influence the antibody response. Six mice of each genotype were injected with 10 μg ovalbumin–adjuvant mixture subcutaneously on day 0. In the experiment shown in the lower panel, mice received a boost at week 6. Sera were collected and checked for antibody level at the indicated times using an indirect ELISA assay. Data represent two independent experiments: each point indicates a single mouse. Statistical significance was assessed using the Student's t‐test with Holm–Šídák correction for multiple comparisons. ns = P > 0.05. FFU, focus‐forming unit; KO, knockout; LCMV, lymphocytic choriomeningitis virus; Mpeg1, macrophage‐expressed gene 1; WT, wild type.

Figure 8

Role of Mpeg 1 in antigen presentation. (a) H‐2K ^bm1 OCSs were intravenously injected into WT or Mpeg1 ^−/− mice, followed by intravenous injection of equal numbers of CTV^high cells labeled with ovalbumin_257–264 peptide and unlabeled CTV^low cells after 6–8 days. CTV^high and CTV^low cell numbers in recipients were determined 36–48 h later by flow cytometry. (b) Equal numbers of CTV‐labeled OT‐I T cells were intravenously injected into WT or Mpeg1 ^−/− mice, followed by intravenous injection of 20 × 10⁶ OCSs or 25 μg ovalbumin after 24 h. Dividing OT‐I T cells in spleen were enumerated 72 h later by flow cytometry. (c) Equal numbers of CTV‐labeled OT‐II T cells were intravenously injected into WT or Mpeg1 ^−/− mice, followed by intravenous injection of 20 × 10⁶ IA ^b−/− OCS mice or 50 μg ovalbumin 24 h later. About 72 h later the number of OT‐II dividing cells in spleen was assessed by flow cytometry. The experiment was performed two times. Each data point represents a single mouse. Statistical significance was assessed using the Student's t‐test. ns = P > 0.05; *P < 0.05; **P < 0.01. CTL, cytotoxic T lymphocyte; CTV, Cell Trace Violet; MHC, major histocompatibility complex; Mpeg1, macrophage‐expressed gene 1; OCS, ova‐coated splenocyte; WT, wild type.