Macrophage-expressed gene 1 protein (perforin-2) is a pore-forming protein of the MACPF/CDC superfamily and the founding member of the MPEG1 family. It is constitutively expressed in macrophages, dendritic cells, neutrophils, natural killer cells and keratinocytes. The full-length isoform is a membrane protein of cytoplasmic vesicles, anchored by a C-terminal transmembrane helix, and an N-terminal MACPF domain is followed by a P2 region and an MVB12-associated beta-prism (MABP) membrane-binding domain. Vesicles carrying perforin-2 traffic to and fuse with bacterium-containing phagosomes, where the protein oligomerises into ring- and arc-shaped prepores that convert, on acidification, into giant beta-barrel transmembrane pores; the MABP domain holds the assembly on the phagolysosome membrane with the pore-forming machinery directed at the engulfed target. Perforin-2 is required for phagocyte killing of a broad range of Gram-positive, Gram-negative and acid-fast bacteria and for the bactericidal action of reactive oxygen and nitrogen species; loss-of-function variants in humans cause immunodeficiency 77, with recurrent skin and soft-tissue infection. In cross-presenting dendritic cells the same pore-forming activity permeabilises antigen-containing endocytic compartments so that internalised antigen escapes to the cytosol for proteasomal processing and presentation on MHC class I, and perforin-2-deficient mice prime CD8+ T cells poorly against cell-associated antigen. How pore formation differs between these two settings is unsettled: the protein is extensively processed by endocytic proteases, and it is disputed whether the species that perforates endocytic membranes is the released ectodomain or the full-length membrane-anchored protein, and whether acidification is required in dendritic cells, which do not acidify their phagosomes. A second, alternatively spliced isoform lacks the transmembrane domain, does not reach vesicle membranes, and is secreted.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0042742 defense response to bacterium | IBA GO_REF:0000033 | ACCEPT | Summary: MPEG1 is required for killing of intracellular bacteria in phagocytes; this is the longest-standing and best-replicated role of the protein. Reason: Core biological process. Supported by loss-of-function in mouse (systemic MRSA and Salmonella dissemination), by knockdown in human macrophages, and by a human haploinsufficiency phenotype. Nothing in the 2026 cross-presentation literature disputes the antibacterial role. Supporting Evidence: PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. PMID:33224153 We showed that haploinsufficiency of perforin-2 reduced the bactericidal capacity of human phagocytes. |
| GO:0005576 extracellular region | IEA GO_REF:0000044 | KEEP AS NON CORE | Summary: Secretion reflects the alternatively spliced isoform perforin-2b (Q2M385-2), which lacks the C-terminal transmembrane domain and consequently fails to reach vesicle membranes. Reason: Real but peripheral. The species found extracellularly is the TM-less splice isoform, not the membrane-anchored pore former, and no pore-forming or antibacterial activity has been demonstrated for the secreted form. This distinction matters more now that PMID:42618800 argues the functional pore-forming species in endocytic escape is the full-length, membrane-anchored protein rather than a released soluble fragment. Supporting Evidence: PMID:28705375 In contrast, the short isoform perforin-2b lacking the transmembrane domain failed to localize to the membrane of vesicles. PMID:28705375 On the other hand, we detected the secretion of perforin-2b in response to LPS stimulation. |
| GO:0030659 cytoplasmic vesicle membrane | IEA GO_REF:0000044 | ACCEPT | Summary: Isoform 1 (perforin-2a) is a multi-pass protein of cytoplasmic vesicle membranes, anchored by its C-terminal transmembrane domain; these vesicles later fuse with bacterium-containing phagosomes. Reason: Correct steady-state location of the pore-forming isoform, established experimentally in human macrophages. Supporting Evidence: PMID:28705375 Perforin-2a was predominantly localized to the membrane of endosome-like vesicles by a C-terminal transmembrane domain. PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0030670 phagocytic vesicle membrane | IEA GO_REF:0000044 | ACCEPT | Summary: MPEG1 acts at the phagocytic vesicle membrane, the compartment in which both antibacterial killing and antigen escape occur. Reason: Correct site of action, consistent with the phagolysosome membrane IDA and with recruitment of perforin-2 to antigen-containing compartments in dendritic cells. Supporting Evidence: PMID:37347855 identified a pore-forming protein, perforin-2 (Mpeg1), as a dedicated effector exclusive to cross-presenting cells PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0055085 transmembrane transport | IEA GO_REF:0000108 | MARK AS OVER ANNOTATED | Summary: Automatic inter-ontology inference from the wide pore channel activity annotation (GO_REF:0000108). Reason: A mechanical consequence of the GO:0022829 assignment, not an independent observation. 'Transmembrane transport' denotes regulated translocation of solutes across a membrane; an MPEG1 pore is a lytic breach whose biological output is killing of an engulfed bacterium or wholesale leakage of endocytic contents into the cytosol. Even the antigen-translocation activity described in PMID:42618800 is escape through a breach rather than transport in the GO sense. The inference is also only as sound as the GO:0022829 assignment, whose definition fits gap junctions and porins rather than lytic pore formers. |
| GO:0005576 extracellular region | EXP PMID:28705375 Regulation of expression and trafficking of perforin-2 by LP... | KEEP AS NON CORE | Summary: Secretion reflects the alternatively spliced isoform perforin-2b (Q2M385-2), which lacks the C-terminal transmembrane domain and consequently fails to reach vesicle membranes. Reason: Real but peripheral. The species found extracellularly is the TM-less splice isoform, not the membrane-anchored pore former, and no pore-forming or antibacterial activity has been demonstrated for the secreted form. This distinction matters more now that PMID:42618800 argues the functional pore-forming species in endocytic escape is the full-length, membrane-anchored protein rather than a released soluble fragment. Supporting Evidence: PMID:28705375 In contrast, the short isoform perforin-2b lacking the transmembrane domain failed to localize to the membrane of vesicles. PMID:28705375 On the other hand, we detected the secretion of perforin-2b in response to LPS stimulation. |
| GO:0030659 cytoplasmic vesicle membrane | EXP PMID:28705375 Regulation of expression and trafficking of perforin-2 by LP... | ACCEPT | Summary: Isoform 1 (perforin-2a) is a multi-pass protein of cytoplasmic vesicle membranes, anchored by its C-terminal transmembrane domain; these vesicles later fuse with bacterium-containing phagosomes. Reason: Correct steady-state location of the pore-forming isoform, established experimentally in human macrophages. Supporting Evidence: PMID:28705375 Perforin-2a was predominantly localized to the membrane of endosome-like vesicles by a C-terminal transmembrane domain. PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0030659 cytoplasmic vesicle membrane | EXP PMID:31537793 The cryo-EM structure of the acid activatable pore-forming i... | ACCEPT | Summary: Isoform 1 (perforin-2a) is a multi-pass protein of cytoplasmic vesicle membranes, anchored by its C-terminal transmembrane domain; these vesicles later fuse with bacterium-containing phagosomes. Reason: Correct steady-state location of the pore-forming isoform, established experimentally in human macrophages. Supporting Evidence: PMID:28705375 Perforin-2a was predominantly localized to the membrane of endosome-like vesicles by a C-terminal transmembrane domain. PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0030670 phagocytic vesicle membrane | ISS PMID:37347855 Perforin-2 is a pore-forming effector of endocytic escape in... | ACCEPT | Summary: MPEG1 acts at the phagocytic vesicle membrane, the compartment in which both antibacterial killing and antigen escape occur. Reason: Correct site of action, consistent with the phagolysosome membrane IDA and with recruitment of perforin-2 to antigen-containing compartments in dendritic cells. Supporting Evidence: PMID:37347855 identified a pore-forming protein, perforin-2 (Mpeg1), as a dedicated effector exclusive to cross-presenting cells PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0042590 antigen processing and presentation of exogenous peptide antigen via MHC class I | ISS PMID:37347855 Perforin-2 is a pore-forming effector of endocytic escape in... | ACCEPT | Summary: MPEG1 permeabilises antigen-containing endocytic compartments so that internalised antigen reaches the cytosol for proteasomal processing and loading onto MHC class I. Reason: Core biological process, and now supported by three independent lines of work: the original genetic screen and Mpeg1-/- priming defect, an independent 2026 study reproducing impaired cross-presentation and weakened CD8+ responses in P2-deficient dendritic cells, and the mechanistic follow-up in PMID:42618800. The three papers disagree about the proximal mechanism - released ectodomain versus anchored full-length protein, and escape versus antigen uptake and control of acidification - but agree that MPEG1 is required for the process. Supporting Evidence: PMID:37347855 Mpeg1-/- mice failed to efficiently prime CD8+ T cells to cell-associated antigens, revealing an important role for perforin-2 in cytosolic entry of antigens during cross-presentation. PMID:41951589 loss of P2 markedly impairs DC-mediated cross-presentation of both soluble and particulate antigens, leading to weakened antigen-specific CD8+ T cell responses PMID:42618800 we find that endocytic escape is mediated by full-length, membrane-anchored perforin-2 rather than by the proteolytically released ectodomain |
| GO:0022829 wide pore channel activity | IDA PMID:31537793 The cryo-EM structure of the acid activatable pore-forming i... | ACCEPT | Summary: Pore formation is the single molecular function of MPEG1: MACPF prepore oligomers convert to a giant transmembrane beta-barrel. The IDA rests on cryo-EM structures of both the soluble prepore and the membrane-bound state plus AFM of perforation. Reason: Core molecular function, and the only MF this gene carries. Kept as a single activity even though PMID:42618800 (EMBO J 2026) argues that the pore-forming species differs between contexts - full-length membrane-anchored protein for endocytic escape rather than the proteolytically released ectodomain - and that antigen translocation is pH-independent, in contrast to the acid-triggered prepore-to-pore transition characterised in PMID:31537793. Those are differences of regulation, topology and target membrane, not of chemistry: in both settings MPEG1 oligomerises and inserts a beta-barrel. GO's molecular-function branch does not distinguish them, so the annotation is unchanged and the divergence is recorded here and in suggested_questions. Separately, note a definitional mismatch that is a family-wide ontology issue rather than an error in this annotation: GO:0022829 is defined in terms of energy-independent facilitated diffusion through an un-gated channel, with gap junctions and porins as examples, which does not describe an acid-gated lytic MACPF pore. GO:0140911 pore-forming activity is a better textual fit and is already carried by PRF1 via InterPro; adding it to MPEG1 would be an improvement. GO:0022829 is retained because it is the term GO applies consistently across this protein class (PRF1, GSDMD, MPEG1), and changing it for MPEG1 alone would fragment the family. Supporting Evidence: PMID:31537793 We further discover that MPEG1 prepore-like assemblies can be induced to perforate membranes through acidification PMID:26402460 Subsequently, Perforin-2 polymerizes and forms large clusters of 100 Γ
pores in the bacterial surface with Perforin-2 cleavage products present in bacteria. PMID:42618800 we find that endocytic escape is mediated by full-length, membrane-anchored perforin-2 rather than by the proteolytically released ectodomain PMID:42618800 Moreover, we show that perforin-2-mediated antigen translocation does not require low pH, explaining how perforin-2 can form pores in cross-presenting dendritic cells which do not acidify their phagosomes. |
| GO:0061474 phagolysosome membrane | IDA PMID:31537793 The cryo-EM structure of the acid activatable pore-forming i... | ACCEPT | Summary: MPEG1 is active in the phagolysosome membrane, where it remains anchored while its pore-forming machinery faces the engulfed target. Reason: Correct and specific site of action. PMID:42618800 reinforces rather than undermines this annotation: if endocytic escape is mediated by full-length, membrane-anchored perforin-2, then the active species is by definition resident in the compartment membrane. What that paper does contest is the accompanying interpretation from PMID:31537793 that the MABP-mediated orientation exists in order to spare the host membrane, since in cross-presenting dendritic cells the host compartment membrane is the one that gets perforated. Supporting Evidence: PMID:31537793 This unexpected mechanism of membrane interaction suggests that MPEG1 remains bound to the phagolysosome membrane while simultaneously forming pores in engulfed bacterial targets. PMID:31537793 These data reveal that a multi-vesicular body of 12 kDa (MVB12)-associated Ξ²-prism (MABP) domain binds membranes such that the pore-forming machinery of MPEG1 is oriented away from the bound membrane. PMID:42618800 we find that endocytic escape is mediated by full-length, membrane-anchored perforin-2 rather than by the proteolytically released ectodomain |
| GO:0002468 dendritic cell antigen processing and presentation | ISS GO_REF:0000024 | ACCEPT | Summary: Dendritic-cell antigen processing and presentation, transferred from the mouse ortholog. Reason: Correct cell-type context: perforin-2 was identified as an effector exclusive to cross-presenting dendritic cells, and Mpeg1-null dendritic cells are defective. Supporting Evidence: PMID:37347855 identified a pore-forming protein, perforin-2 (Mpeg1), as a dedicated effector exclusive to cross-presenting cells PMID:41951589 loss of P2 markedly impairs DC-mediated cross-presentation of both soluble and particulate antigens, leading to weakened antigen-specific CD8+ T cell responses |
| GO:0002478 antigen processing and presentation of exogenous peptide antigen | ISS GO_REF:0000024 | ACCEPT | Summary: General exogenous-antigen presentation term, parent of the MHC class I child that this gene also carries. Reason: Correct but less informative than GO:0042590; the evidence is specifically for the MHC class I cross-presentation route. |
| GO:0030670 phagocytic vesicle membrane | ISS GO_REF:0000024 | ACCEPT | Summary: MPEG1 acts at the phagocytic vesicle membrane, the compartment in which both antibacterial killing and antigen escape occur. Reason: Correct site of action, consistent with the phagolysosome membrane IDA and with recruitment of perforin-2 to antigen-containing compartments in dendritic cells. Supporting Evidence: PMID:37347855 identified a pore-forming protein, perforin-2 (Mpeg1), as a dedicated effector exclusive to cross-presenting cells PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0042590 antigen processing and presentation of exogenous peptide antigen via MHC class I | ISS GO_REF:0000024 | ACCEPT | Summary: MPEG1 permeabilises antigen-containing endocytic compartments so that internalised antigen reaches the cytosol for proteasomal processing and loading onto MHC class I. Reason: Core biological process, and now supported by three independent lines of work: the original genetic screen and Mpeg1-/- priming defect, an independent 2026 study reproducing impaired cross-presentation and weakened CD8+ responses in P2-deficient dendritic cells, and the mechanistic follow-up in PMID:42618800. The three papers disagree about the proximal mechanism - released ectodomain versus anchored full-length protein, and escape versus antigen uptake and control of acidification - but agree that MPEG1 is required for the process. Supporting Evidence: PMID:37347855 Mpeg1-/- mice failed to efficiently prime CD8+ T cells to cell-associated antigens, revealing an important role for perforin-2 in cytosolic entry of antigens during cross-presentation. PMID:41951589 loss of P2 markedly impairs DC-mediated cross-presentation of both soluble and particulate antigens, leading to weakened antigen-specific CD8+ T cell responses PMID:42618800 we find that endocytic escape is mediated by full-length, membrane-anchored perforin-2 rather than by the proteolytically released ectodomain |
| GO:0140367 antibacterial innate immune response | ISS GO_REF:0000024 | ACCEPT | Summary: Antibacterial innate immune response is the umbrella process for the phagolysosomal killing activity. Reason: Core biological process, orthology-transferred from the mouse ortholog and matched by direct human evidence. Supporting Evidence: PMID:33224153 Functional studies revealed that the truncation variant resulted in significantly reduced capacity of the patient's phagocytes to kill intracellular bacteria. |
| GO:0042742 defense response to bacterium | IMP PMID:33224153 MPEG1/Perforin-2 Haploinsufficiency Associated Polymicrobial... | ACCEPT | Summary: MPEG1 is required for killing of intracellular bacteria in phagocytes; this is the longest-standing and best-replicated role of the protein. Reason: Core biological process. Supported by loss-of-function in mouse (systemic MRSA and Salmonella dissemination), by knockdown in human macrophages, and by a human haploinsufficiency phenotype. Nothing in the 2026 cross-presentation literature disputes the antibacterial role. Supporting Evidence: PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. PMID:33224153 We showed that haploinsufficiency of perforin-2 reduced the bactericidal capacity of human phagocytes. |
| GO:0031410 cytoplasmic vesicle | IDA PMID:26402460 Perforin-2 is essential for intracellular defense of parench... | ACCEPT | Summary: Cytoplasmic vesicle localisation observed directly in human cells. Reason: Correct, though the cytoplasmic vesicle membrane annotation carried by this gene is more informative, since MPEG1 is an integral membrane protein of those vesicles rather than a luminal cargo. Supporting Evidence: PMID:26402460 Perforin-2 is a transmembrane protein of cytosolic vesicles -derived from multiple organelles- that translocate to and fuse with bacterium containing vesicles. |
| GO:0042742 defense response to bacterium | IMP PMID:26402460 Perforin-2 is essential for intracellular defense of parench... | ACCEPT | Summary: MPEG1 is required for killing of intracellular bacteria in phagocytes; this is the longest-standing and best-replicated role of the protein. Reason: Core biological process. Supported by loss-of-function in mouse (systemic MRSA and Salmonella dissemination), by knockdown in human macrophages, and by a human haploinsufficiency phenotype. Nothing in the 2026 cross-presentation literature disputes the antibacterial role. Supporting Evidence: PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. PMID:33224153 We showed that haploinsufficiency of perforin-2 reduced the bactericidal capacity of human phagocytes. |
| GO:0045335 phagocytic vesicle | ISS GO_REF:0000024 | ACCEPT | Summary: Phagocytic vesicle localisation, transferred from the mouse ortholog. Reason: Correct, though the phagocytic vesicle membrane annotation carried by this gene is the more informative form of the same statement. |
| GO:0050829 defense response to Gram-negative bacterium | ISS GO_REF:0000024 | ACCEPT | Summary: MPEG1 restricts Gram-negative intracellular bacteria, demonstrated for Chlamydia trachomatis and Salmonella typhimurium. Reason: Correct and appropriately specific; a subtype of the core antibacterial process. Supporting Evidence: PMID:23753625 Knockdown of perforin-2 in macrophages did not alter the invasion of host cells but did result in chlamydial growth that closely mirrored that detected in HeLa cells. PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. |
| GO:0050829 defense response to Gram-negative bacterium | IMP PMID:23753625 Perforin-2 restricts growth of Chlamydia trachomatis in macr... | ACCEPT | Summary: MPEG1 restricts Gram-negative intracellular bacteria, demonstrated for Chlamydia trachomatis and Salmonella typhimurium. Reason: Correct and appropriately specific; a subtype of the core antibacterial process. Supporting Evidence: PMID:23753625 Knockdown of perforin-2 in macrophages did not alter the invasion of host cells but did result in chlamydial growth that closely mirrored that detected in HeLa cells. PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. |
| GO:0050830 defense response to Gram-positive bacterium | ISS GO_REF:0000024 | ACCEPT | Summary: MPEG1 restricts Gram-positive intracellular bacteria, demonstrated for Staphylococcus aureus in human skin cells and in mouse. Reason: Correct and appropriately specific; a subtype of the core antibacterial process. Supporting Evidence: PMID:30609079 P-2 overexpression resulted in a reduction of intracellular S. aureus PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. |
| GO:0050830 defense response to Gram-positive bacterium | IDA PMID:30609079 Single cell analyses reveal specific distribution of anti-ba... | ACCEPT | Summary: MPEG1 restricts Gram-positive intracellular bacteria, demonstrated for Staphylococcus aureus in human skin cells and in mouse. Reason: Correct and appropriately specific; a subtype of the core antibacterial process. Supporting Evidence: PMID:30609079 P-2 overexpression resulted in a reduction of intracellular S. aureus PMID:26402460 Perforin-2 knockout mice are unable to control the systemic dissemination of methicillin-resistant Staphylococcus aureus (MRSA) or Salmonella typhimurium and perish shortly after epicutaneous or orogastric infection respectively. |
Loading supporting contentβ¦
Download this section (compressed HTML)Q: Does MPEG1 have one pore-forming molecular function deployed in two settings, or two mechanistically distinct pore-forming activities? PMID:42618800 reports that endocytic escape uses full-length membrane-anchored perforin-2 and does not require low pH, whereas the antibacterial pore characterised structurally in PMID:31537793 is acid-triggered. Is the difference one of regulation and topology only, or of the assembly pathway itself?
Suggested experts: GO molecular-function curators, MACPF/CDC structural biologists
Q: If the full-length, transmembrane-anchored protein makes the pore during cross-presentation, what is the function of the extensive proteolytic processing by endocytic proteases and of the released ectodomain? Is the ectodomain the antibacterial effector, an inactivation product, or something else?
Q: If acidification is not the trigger in cross-presenting dendritic cells, what is? Candidate triggers include lipid composition of the target membrane, redox state, a protease-generated conformational change, or local concentration reached on oligomerisation.
Q: How does the cell avoid catastrophic self-perforation if the pore-forming machinery is deployed from an anchor in the host compartment membrane? The MABP-orientation model of PMID:31537793 was the previous answer; PMID:42618800 argues that in dendritic cells the host membrane is in fact the perforated one.
Q: Should MPEG1 also carry GO:0140911 pore-forming activity? GO:0022829 wide pore channel activity is defined in terms of un-gated facilitated diffusion through gap junctions and porins, which does not describe a gated lytic MACPF beta-barrel, yet it is the term used consistently across PRF1, GSDMD and MPEG1.
Suggested experts: GO ontology developers
Q: Do the mechanisms proposed by PMID:42618800 and PMID:41951589 describe the same cell biology? One places the critical event at the endocytic membrane (anchored pore, pH-independent escape) and the other at the plasma membrane and at endosomal pH control (macropinocytic uptake, restraint of acidification). Both agree perforin-2 is required for cross-presentation.
Experiment: Reconstitute Mpeg1-null dendritic cells and macrophages with (i) wild-type perforin-2, (ii) a protease-resistant full-length variant that cannot be cleaved at the juxtamembrane site, and (iii) the ectodomain alone, then score cytosolic antigen delivery, CD8+ T cell priming and killing of intracellular bacteria in parallel in the same cells. A double dissociation would settle whether the two deployments use different species of the protein.
Hypothesis: Endocytic escape requires the membrane anchor, whereas bacterial killing requires the released ectodomain.
Type: structure-function complementation
Experiment: Use liposomes and supported bilayers of defined composition with purified full-length (detergent- or nanodisc-reconstituted) and ectodomain-only perforin-2, and test prepore-to-pore conversion by AFM and release assays as a function of pH, lipid composition, redox state and protease exposure. Compare directly against the acid-triggered conversion reported in PMID:31537793.
Hypothesis: Pore formation in cross-presenting dendritic cells is triggered by something other than luminal acidification.
Type: in vitro reconstitution
Experiment: Apply cryo-electron tomography to antigen-containing compartments of wild-type and Mpeg1-null cross-presenting dendritic cells to visualise perforin-2 assemblies in situ and determine on which leaflet and in which membrane the barrels sit.
Hypothesis: The pore that mediates antigen escape is built in the host compartment membrane, not in an engulfed target.
Type: in situ structural imaging
Experiment: Screen for MPEG1 point mutants that abolish killing of intracellular bacteria while preserving cross-presentation, and vice versa, and map them onto the MACPF, P2 and MABP domains and the juxtamembrane processing sites.
Hypothesis: The two deployments are separable genetically.
Type: genetic separation-of-function screen
Loading supporting contentβ¦
Download this section (compressed HTML)Loading supporting contentβ¦
Download this section (compressed HTML)