Perforin-2 Breaches the Envelope of Phagocytosed Bacteria Allowing

Home , Cathelicidin, Perforin

Perforin-2 Breaches the Envelope of Phagocytosed Bacteria Allowing Antimicrobial Effectors Access to Intracellular Targets This information is current as of October 9, 2021. Fangfang Bai, Ryan M. McCormack, Suzanne Hower, Gregory V. Plano, Mathias G. Lichtenheld and George P. Munson J Immunol published online 24 September 2018

http://www.jimmunol.org/content/early/2018/09/21/jimmun Downloaded from ol.1800365

Supplementary http://www.jimmunol.org/content/suppl/2018/09/23/jimmunol.180036 Material 5.DCSupplemental http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

by guest on October 9, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2018 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Published September 24, 2018, doi:10.4049/jimmunol.1800365 The Journal of Immunology

Perforin-2 Breaches the Envelope of Phagocytosed Bacteria Allowing Antimicrobial Effectors Access to Intracellular Targets

Fangfang Bai, Ryan M. McCormack, Suzanne Hower, Gregory V. Plano, Mathias G. Lichtenheld, and George P. Munson

Perforin-2, the product of the MPEG1 gene, limits the spread and dissemination of bacterial pathogens in vivo. It is highly expressed in murine and human phagocytes, and macrophages lacking Perforin-2 are compromised in their ability to kill phagocytosed bacteria. In this study, we used Salmonella enterica serovar Typhimurium as a model intracellular pathogen to elucidate the mechanism of Perforin-2’s bactericidal activity. In vitro Perforin-2 was found to facilitate the degradation of Ags contained within the envelope of phagocytosed bacteria. In contrast, degradation of a representative surface Ag was found to be independent of Perforin-2. Consistent with our in vitro results, a protease-sensitive, periplasmic superoxide dismutase (SodCII) Downloaded from contributed to the virulence of S. Typhimurium in Perforin-2 knockout but not wild-type mice. In aggregate, our studies indicate that Perforin-2 breaches the envelope of phagocytosed bacteria, facilitating the delivery of proteases and other antimicrobial effectors to sites within the bacterial cell. The Journal of Immunology, 2018, 201: 000–000.

acrophages and neutrophils phagocytose microorgan- Studies over the past two decades have shown that antimicrobial

isms to remove them from blood and tissues. As the peptides such as the defensins and cathelicidins attack and disrupt http://www.jimmunol.org/ M phagosome matures, the multisubunit NADPH oxidase bacterial membranes (6–9). Nuclear magnetic resonance studies of assembles on its membrane and reduces O2 to generate superoxide the cathelicidin LL-37 suggest that the peptide destabilizes the in the lumen of the phagosome (1, 2). The products of this bacterial membrane by carpeting rather than penetrating the lipid respiratory burst, superoxide and subsequently other reactive bilayer (10). Within phagolysosomes, the murine cathelicidin oxygen species (ROS), are bactericidal. The destruction and deg- CRAMP has been shown to be active against phagocytosed radation of phagocytosed microbes are further facilitated by Salmonella, most likely by disruption of the bacterium’s outer acidification of the phagosome and fusion with lysosomes that membrane (11, 12). Likewise, cathelicidin LL-37 may play a deliver oxygen-independent antimicrobial effectors such as lyso- similar role in human phagocytes (13, 14). Nearly coincident with by guest on October 9, 2021 zyme, glycosylases, proteases, and other hydrolases (3). Large the initial descriptions of LL-37 and CRAMP, the Mpeg1 gene was antimicrobials such as proteases and other hydrolases are typically identified as a potential marker of mammalian macrophages be- membrane-impermeable molecules. This property is advantageous cause of its relatively high expression in mature human and mu- in that it allows them to be confined within lysosomes and rine macrophages (6, 15, 16). Mpeg1 encodes a 73 kDa protein phagolysosomes. However, it also precludes them from reaching referred to as Perforin-2, and in their initial report, Spilsbury et al. the internal components of phagocytosed bacteria. For example, (15) noted its partial homology to the MACPF domain of Perforin, lysozyme hydrolyzes b (1, 4) glycosidic bonds of peptidoglycan, the cytolytic protein of NK cells and CTL. Unlike the carpet the primary structural component of bacterial cell walls. In Gram- mechanism of cathelicidins, Perforin is a large polypeptide that negative bacteria, peptidoglycan resides in the space between the polymerizes on target membranes. A concerted structural transi- outer and inner membranes (i.e., the periplasm). Thus, for these tion results in a pore through the lipid bilayer whose hydrophilic bacteria, a breach of the outer membrane must precede lysozyme- channel is lined with amphipathic b-strands donated by the dependent degradation of peptidoglycan (4, 5). For hydrolases MACPF domains (17, 18). It is through this channel, either at with targets in the cytosol of Gram-negative bacteria, the chal- the cell surface as originally hypothesized or within endosomal lenge is 2-fold, as their substrates are bound by both an inner and membranes as a more recent study suggests, that granzyme pro- outer membrane. teases enter tumor and virally infected cells to facilitate their

Department of Microbiology and Immunology, University of Miami Miller School of Address correspondence and reprint requests to Dr. George P. Munson, University of Medicine, Miami, FL 33136 Miami Miller School of Medicine, Department of Microbiology and Immunology, P.O. Box 016960 (R-138), Miami, FL 33136. E-mail address: [email protected] ORCIDs: 0000-0001-7432-3046 (S.H.); 0000-0002-3692-8199 (G.P.M.). The online version of this article contains supplemental material. Received for publication March 12, 2018. Accepted for publication August 28, 2018. Abbreviations used in this article: BMDM, bone marrow–derived macrophage; CI, This work was supported by National Institutes of Health/National Institute of competitive indices; DPI, diphenyleneiodonium chloride; KO, knockout; LB, Luria– Allergy and Infectious Diseases Grant AI110810. Bertani; PEM, peritoneal exudate macrophage; ROS, reactive oxygen species; SCV, Author contributions: conceptualization, F.B., R.M.M., M.G.L., and G.P.M.; methodology, Salmonella-containing vacuole; WT, wild-type. F.B., R.M.M., and G.P.M.; investigation, F.B.; validation, F.B. and G.P.M.; formal analysis, F.B.; resources, S.H., G.V.P., and R.M.M.; writing—original draft, F.B. and G.P.M.; Copyright Ó 2018 by The American Association of Immunologists, Inc. 0022-1767/18/$37.50 writing—review and editing, F.B., R.M.M., S.H., G.V.P., M.G.L., and G.P.M.; visualiza- tion, F.B. and G.P.M.; supervision, M.G.L. and G.P.M.; and funding acquisition, G.P.M. The content of this article is the sole responsibility of the authors and does not necessarily represent the ofﬁcial views of the National Institutes of Health.

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1800365 2 PERFORIN-2 BREACHES THE ENVELOPE OF PHAGOCYTOSED BACTERIA destruction and lysis (19–21). MACPF domains are also present in antibiotics were used at the following concentrations: chloramphenicol, the terminal complement proteins that form pores in the outer 10 mg/ml; ampicillin, 100 mg/ml; kanamycin, 50 mg/ml; tetracycline, 7.5 mg/ml. membranes of Gram-negative bacteria through a similar mecha- Cell preparation nism of polymerization and structural transition (22). Despite the homology of mammalian Perforin-2 to known pore- Peritoneal exudate macrophages (PEMs) were isolated as previously described (33). Briefly, mice were injected i.p. with 1 ml of 4% Brewer forming proteins, there was no further elaboration of its function thioglycollate medium, and peritoneal exudates were recovered after 4 d. A until 2013, nearly two decades after the initial report of Spilsbury modified schedule was used to collect PEMs for ROS assays (34). In brief, et al. (15, 23). In 2013, McCormack et al. (23) demonstrated mice were inoculated on days 1 and 7, and exudates were collected on day that the expression of Perforin-2 correlated with the killing of 14. Resting macrophages were collected from the peritoneal cavities of untreated mice. Peritoneal neutrophils were elicited and recovered as phagocytosed Gram-negative, -positive, and acid-fast bacteria previously described (35). Briefly, mice were injected i.p. with 1 ml of 9% in vitro. Subsequent studies with transgenic mice found that casein in PBS (2.7 mM KCl, 4.3 mM Na2HPO4, 1.47 mM KH2PO4, Perforin-2 knockout (KO) mice are unable to limit the prolifera- 137 mM NaCl, pH 7.4) containing 0.9 mM CaCl2 and 0.5 mM MgCl2.A tion and dissemination of infectious bacteria. Not surprisingly, second injection was administered 16 h after the first, and exudates were Perforin-2 KO mice succumb to infectious doses that are nonlethal collected 3 h later. Cells were maintained in IMDM (Life Technologies) supplemented with 10% FBS (Life Technologies) at 37˚C in 5% CO2. to their wild-type (WT) littermates (24–26). Moreover, this defect Bone marrow–derived macrophages (BMDM) were isolated from the fe- is not limited to a particular route of infection or pathogen. Nor is murs from the Perforin-2 WT and KO mice. The cells were cultured in it limited to mammalian Perforin-2, as similar results have been IMDM medium supplemented with 10% FBS, 20% L929 conditioned reported with zebrafish as the model organism (27). In aggregate, medium, and 10 mM L-glutamine. BMDM were cultured for 6 d before use, and the culture medium was changed every 2 d. these latter studies have demonstrated that Perforin-2 is associated with broad-spectrum bactericidal activity. In this study, we used Intracellular killing assays Downloaded from Salmonella enterica serovar Typhimurium (S. Typhimurium) as Intracellular gentamicin protection assays were performed as previously a model pathogen to elucidate the mechanism of Perforin-2– described (23, 36, 37). Briefly, 3 3 105 PEM were seeded in 24-well plates dependent killing of phagocytosed bacteria. in IMDM, 10% FBS and incubated at 37˚C in 5% CO2. Murine IFN-g (BioLegend) was added 14 h before infection at a final concentration of 50 ng/ml. Overnight, cultures of bacteria were diluted 33-fold in LB and Materials and Methods cultured aerobically at 37˚C for 3 h to midlog, at which point the optical http://www.jimmunol.org/ Mice absorbance of the culture at 600 nm was approximately 0.6. Bacteria were added at a multiplicity of infection between 20 and 50, and plates were 3 Perforin-2 KO 129 1/SvJ mice were produced at the University of Miami incubated for 45 to 60 min to allow for uptake of bacteria. Cells were then Miller School of Medicine Transgenic Core Facility as previously de- washed three times with PBS, and fresh culture medium containing 3 scribed (24). WT 129 1/SvJ and Perforin-2 KO mice of either sex were 50 mg/ml gentamicin was added to each well to kill extracellular bacteria. used at 2–6 mo of age. Mice received food and water ad libitum and were After 2 h, the concentration of gentamicin was reduced to 5 mg/ml. At housed at an ambient temperature of 23˚C on a 12 h light/dark cycle under selected time points, gentamicin was removed by PBS washes, and bac- specific pathogen-free conditions. Mice were euthanized by CO2 inhalation teria were recovered by lysis of mammalian cells in sterile water with 0.1% followed by cervical dislocation. All procedures with animals were Triton X-100. Lysates were serially diluted in PBS, and the bacteria were reviewed and approved by the University of Miami’s Institutional Animal enumerated on LB agar plates.

Care and Use Committee. by guest on October 9, 2021 ROS detection Strains and plasmids A value of 3 3 105 PEM was seeded in 96-well opaque white plates in Bacterial strains are listed in Table I. Primer sequences are listed in 200 ml of phenol red–free IMDM, 10% FBS and primed for 14 h with Table II. Strains constructed for this study are isogenic derivatives of murine IFN-g at a final concentration of 50 ng/ml. Alternatively, 3 3 105 S. Typhimurium strain LT2 (28). Plasmid pKD4 (29) was used in PCR with neutrophils were seeded in Krebs Ringer phosphate glucose buffer primer pairs sodCI-P1/sodCI-P2 or sodCII-P1/sodCII-P2 to generate (145 mM NaCl, 4.86 mM KCl, 5.7 mM sodium phosphate, kanamycin resistance cassettes bracketed by recognition sites for FLP 0.54 mM CaCl , 1.22 mM MgSO , 5.5 mM Glucose, pH 7.35). For ROS recombinase and flanking sequences targeting sodCI or sodCII. Deletions 2 4 production in neutrophils, cells were incubated with 1 mM luminol sodCI sodCII l of or were generated by Red-mediated recombination of the (Sigma-Aldrich) for 3 min before adding PMA (InvivoGen) or LPS cassettes into LT2 as described (29). Recombinants were selected on (InvivoGen) to a final concentration of 100 ng/ml. Some wells were treated Luria–Bertani (LB) agar plates with kanamycin. Recombination sites were with 10 nM diphenyleneiodonium chloride (DPI) (Sigma-Aldrich), a verified by PCR with flanking primer pairs sodCI-MfeI/sodCI-HindIII for NAPDH oxidase inhibitor, for 30 min prior to addition of luminol. The sodCI::kan and sodCII-EcoRI/sodCII-HindIII for sodCII::kan. For the same procedures were used with macrophages, except the luminol en- construction of double deletions, FLP recombinase was used to excise the hancer Diogenes (National Diagnostics) was also used (38). For the first cassette prior to insertion of the second. measurement of phagocytic ROS production, 3 mm of amino microspheres sodCII The gene was cloned from LT2 by PCR with primers 1238/1239. The polybeads (Polysciences) were treated with 8% glutaraldehyde in PBS PCR product was digested with XbaI and BamHI, then ligated into the same overnight at room temperature with agitation. Beads were washed with sites of pAH63Tc, a derivative of pAH63 (30) in which the kanamycin resis- PBS, incubated with 100 mM luminol in DMSO for 4 h at room tem- tance gene was replaced with one conferring resistance to tetracycline. To perature, and subsequently washed with 0.5 M glycine and PBS. Prepared complement sodCII mutants, the resulting plasmid, pAH63Tc-sodCII, was in- beads were stored in PBS. Perforin-2 WT or KO neutrophils were pulsed tegrated into attBl of relevant strains by a site-specific recombinase as previ- with luminol-coupled beads for 15 min and then washed with PBS. When ously described (30). The chloramphenicol-resistant strain GPM2004 was indicated, 0.5 mg/ml PMA was added before starting. Luminescence was attBl constructed by integration of pCAH63 (30) into of LT2. Plasmid recorded in an EnVision (PerkinElmer) plate reader. pTrc99aSodCII3flag was constructed by HiFi DNA Assembly (New England Biolabs) of three PCR products: one carrying sodCII amplified from pAH63Tc- Measurement of vacuolar pH sodCII with primers 1329/1330, another carrying a triple FLAG epitope amplified from pSUB11 (31) with primers 1331/1332, and the vector backbone The pH of Salmonella-containing vacuoles (SCVs) was determined as amplified from pTrc99a (32) with primers 1333/1334. Bacteria were cultured in previously described (39), with modifications. Perforin-2 WT or KO PEMs LB broth or plates. Genes encoding Perforin-2 and E2-Crimsion were cloned were seeded on eight-well slides (ibiTreat m-Slide; ibidi USA) at 50,000 into the dual expression plasmid pVitro1-neo-Lucia-SEAP (InvivoGen) by HiFi cells per well and incubated overnight. The cells were then loaded with assembly with the primer pairs 1426/1427, 1584/1585, and 1586/1587. The new dextran-FITC m.w. 10,000 at 282 mg/ml and dextran–Alexa Fluor 647 m. construction was named pMuP2–E2-Crimson. The superfold GFP-expressing w. 10,000 at 70 mg/ml in normal growth medium for 4 h, then infected with strains were constructed by integration of pCAH63-sfGFP, which was assem- TurboRFP-expressing S. Typhimurium WT LT2/pBAD-TruboRFP. After bled by HiFi with primer pairs 1258/1259 and 1257/1243 into attBl of relevant 1 h, cells were washed with HBSS containing 0.5% FBS and imaged in strains. The pBAD-TurboRFP plasmid (Addgene) was electroporated into LT2 CO2-independent medium with 10% FBS by fluorescent microscopy. The for constructing the S. Typhimurium LT2/pBAD-TurboRFP. As appropriate, ratio of the two fluorophores (488:647 nm) within SCVs was quantified The Journal of Immunology 3 with the Fiji build of ImageJ (40). To generate a standard curve, the same sterile PBS and plated in triplicate on LB agar with antibiotic selection as cells used for experimental measurement were incubated with ionophore appropriate for each strain in the initial inoculum. For each spleen and nigericin (20 mM, Sigma-Aldrich) for 5 min at room temperature to allow liver, mean CFUs were used to calculate competitive indices (CI) the free exchange of H+ ions across membranes and then with a series of according to the following formula: CI = (strain A recovered/strain B buffers of varying pH (4–7). The pH standard buffers contained 140 mM recovered)/(strain A inoculum/strain B inoculum). KCl and 5 mM glucose buffered with 15 mM Tris for pH 7, 15 mM MES for pH 6, and acetic acid for pH 4 and 5. Statistical analyses Recovery of phagocytosed bacteria and immunodetection Statistical analyses were performed with GraphPad Prism 7 software. The p values #0.05 were considered statistically significant and determined S. Typhimurium LT2/pTrc99aSodCII3flag was cultured aerobically by two-way ANOVA with Sidak multiple comparisons test or one-way overnight in LB with ampicillin at 37˚C. The bacteria were pelleted ANOVA with Tukey multiple comparisons test. and washed with sterile PBS, and aliquots of the inoculum were frozen at 280˚C for later analysis. Aliquots of the inoculum were also used to infect WT and Perforin-2 KO PEMs at a multiplicity of infection of Results ∼20–50. After 1 h incubation, the cells were washed three times with PBS, Perforin-2 limits the survival of phagocytosed bacteria then IMDM with 10% FBS and 50 mg/ml gentamicin was added to each independent of ROS well. This initial addition of gentamicin marked the 0 h time point. After 2 h, the concentration of gentamicin was decreased to 5 mg/ml. After 18 h, Because the interactions between Salmonella sp. and macro- the cells were washed with PBS three times, then PBS, 0.1% Triton X-100 phages have been extensively characterized, we chose to exploit containing a proteinase inhibitor mixture (Roche) was added to each well. After a 5-min incubation at 37˚C, the cells were manually detached with a the Salmonella/phagocyte paradigm to probe the mechanism of cell scraper, the bacterial cells were harvested by centrifugation at 4˚C at Perforin-2–dependent killing of phagocytosed bacteria (41). 3 3 10, 000 g 10 min, and the supernatant was removed. Recovered Accordingly, PEMs and neutrophils isolated from WT and Downloaded from 6 phagocytosed bacteria (approximately 10 CFU) and bacteria from the Perforin-2 KO mice were infected with S. Typhimurium. As ex- original LB culture were boiled in Laemmli loading buffer for 7 min. The pected from previous studies that have shown Salmonella survives same infection procedure was conducted with WT and Perforin-2 KO BMDM, except the phagocytosed bacteria were collected at 1, 3, and 6 h. within macrophages, the intracellular load of bacteria either The protein samples were separated on 4–20% gradient SDS-PAGE gels increased or remained nearly constant in WT phagocytes (Fig. 1A, and transferred to nitrocellulose membranes. The membranes were 1B). However, Perforin-2–deficient phagocytes had significantly blocked with 5% nonfat milk in Tris-buffered saline containing 0.1% higher intracellular loads of S. Typhimurium than WT phagocytes Tween 20 (TBST) for 2 h and then incubated at 4˚C for 16 h with primary http://www.jimmunol.org/ Abs anti-FliC (1:1000, catalog no. 629701; InvivoGen), anti-FLAG (Fig. 1A, 1B). This demonstrates that Perforin-2 limits the survival (1:5000, catalog no. F1804; Sigma-Aldrich), or anti-DnaK (1:5000, cata- and/or replication of phagocytosed S. Typhimurium and is con- log no. ab69617; Abcam) diluted in TBST with 5% nonfat milk. After sistent with previous studies that have shown Perforin-2 is a potent three washes with TBST, the membranes were incubated at 37˚C for 1 h antimicrobial effector against Salmonella as well as Gram-positive with anti-mouse HRP-labeled secondary Ab (catalog no. 115-036-062; and acid-fast bacteria (24–26, 42). Jackson ImmunoResearch Laboratories) diluted 1:5000 in TBST with 5% nonfat milk. An Odyssey Fc Imaging System (LI-COR) was used to detect One of the aforementioned studies also investigated the rela- and quantify chemiluminescence after addition of SuperSignal West Pico tionship between ROS and Perforin-2 and concluded that the Chemiluminescent Substrate (Thermo Fisher Scientific). bactericidal activity of ROS was dependent upon Perforin-2 (25).

Confocal microscopy This was based on two principle findings. First, chemical inhibi- by guest on October 9, 2021 tion of ROS production significantly enhanced the survival of RAW264.7 cells were transfected with pMuP2–E2-Crimson plasmid with phagocytosed S. Typhimurium relative to mock-treated cells in Lipofectamine LTX (Invitrogen) following the manufacturer’s instructions. The transfected cells were seeded on 25-mm coverslips, and the cells were WT but not Perforin-2–deficient PEMs. Second, WT PEMs killed infected with superfolder GFP-expressing S. Typhimurium WT strain a ΔsodCI strain of S. Typhimurium much more efficiently than GPM2010 or DsodCI sodCII strain GPM2012 as described. The cells were WT S. Typhimurium. SodCI is a periplasmic superoxide dis- counterstained with DAPI and were mounted with ProLong Diamond mutase that neutralizes ROS and thus promotes the survival of Antifade Mountant (Life Technologies). Images were collected on a Leica S. Typhimurium within phagosomes (25, 43). However, the sodCI SP5 inverted confocal microscope with a motorized stage and analyzed with the Fiji software package (40). mutant was found to be no less fit than the WT strain when Perforin-2 KO PEMs were used. Thus, both a chemical and ge- Murine infections netic analysis suggest that ROS is not a significant bactericidal Bacteria were cultured overnight in LB medium at 37˚C and diluted in effector when Perforin-2 is absent. sterile PBS. For competition assays, selected strains of S. Typhimurium As with the previous study, we observed similar effects with were mixed at a 1:1 ratio. Perforin-2 WT and KO 129 3 1/SvJ mice were both PEMs and neutrophils. In either case, the S. Typhimurium inoculated by i.p. injection. The CFU of each inoculum was quantified by plating and ranged from 500 to 1000 total CFUs. Spleens and livers were sodCI mutant was efficiently killed by Peforin-2–proficient but collected 4 d after inoculation and then homogenized (Omni International not –deficient phagocytes (Fig. 1C, 1D). The source of phagocytic Tulsa) in 500 ml of PBS, 0.1% Tween 20. Homogenates were diluted in ROS is the multisubunit NADPH oxidase NOX2 (44). Assembly

Table I. Bacteria strains

Strain Description Reference LT2 WT S. Typhimurium (28) LT2b LT2 derivative that spontaneously lost pSLT virulence plasmid This study GPM2004 LT2 attBl::pCAH63, chloramphenicol resistant This study GPM2004b LT2b attBl::pCAH63, chloramphenicol resistant This study ST188 LT2 DsodCI::kan This study ST188b LT2b DsodCI::kan This study ST189 LT2 DsodCI DsodCII::kan This study GPM2008 LT2 DsodCI DsodCII::kan attBl::pAH63Tc-sodCII This study GPM2010 LT2 attBl::pCAH63, sfGFP, and chloramphenicol resistant This study GPM2012 LT2 DsodCI DsodCII::kan attBl::pCAH63, sfGFP, and chloramphenicol resistant This study 4 PERFORIN-2 BREACHES THE ENVELOPE OF PHAGOCYTOSED BACTERIA

Table II. Oligonucleotide primers

Primer Sequence (59→39) sodCI-P1 59-TACACAATATTGTCGCTGGTAGCTGGTGCGCTCATCAGTTGTGTAGGCTGGAGCTGCTTC-39 sodCI-P2 59-ATTGTCACCGCCTTTATGGATCATCAATGAGTGACCTTTCATATGAATATCCTCCTTAGT-39 sodCI-MfeI 59-TTTCAATTGATTAATGGTATTTACGATACAACC-39 sodCI-HindIII 59-TTTAAGCTTATGGCTATGTTGCTGTTATTTCTC-39 sodCII-P1 59-GCAGGCCGCCAGCGAGAAAGTAGAGATGAATCTGGTGACTGTGTAGGCTGGAGCTGCTTC-39 sodCII-P2 59-CGCCGCCGCCGAGCGGTTTCGGCTGATCGGACATGTTATCATATGAATATCCTCCTTAGT-39 sodCII-EcoRI 59-TTTGAATTCAACAGGCGACCACATGTAACGGAG-39 sodCII-HindIII 59-TTTAAGCTTCACTGGCTCCGGGTTATTTAATGA-39 1238 59-GCGTCTAGAGGGTTATGACGTGCCGTAATC-39 1239 59-GCGGGATCCTCTTCACTTGTCGTCATCGTCCTTGTAGTCTTTAATGACGCCGCAGGCGTAAC-39 1329 59-ACCCGGGGATCCTTTATGACGTGCCGTAATCGC-39 1330 59-TGTAGTCGAATTCTTTAATGACGCCGCAGGCGTAA-39 1331 59-CGGCGTCATTAAAGAATTCGACTACAAAGACCATGACG-39 1332 59-CAAAACAGCCAAGGGAACTTCGAAGCAGCTCCAG 1333 59-GCTTCGAAGTTCCCTTGGCTGTTTTGGCGGATGA-39 1334 59-CGGCACGTCATAAAGGATCCCCGGGTACCGAG-39 1584 59-CCAAGTCCAGCTATGGATAGCACTGAGAACGTC-39 1585 59-ATCGTAAGGGTACTGGAACAGGTGGTGGCG-39 1586 59-CACCTGTTCCAGTACCCTTACGATGTACCGG-39 1587 59-AGTGCTATCCATAGCTGGACTTGGATCCTCTTC-39 Downloaded from 1426 59-CCTTACGATGTACCGGATTACGCATGAAGCTAGCTGGCCAGACATGATAAG-39 1427 59-GATTGCTTTGAATTAGCGGTGGTTTTCAC-39 1258 59-TGAATTAATGGCGATGACGCATC-39 1259 59-ATTATACGAGCCGGATGATTAATTGTCAACGCATGCAAGCTTGGCAC-39 1257 59-TCGTGAGGATGCGTCATCGCCATTAATTCAGCCAAGCTTCGAATTC-39 1243 59-GTTGACAATTAATCATCCGGCTCGTATAATCTACTGTTTCTCCATACCCG-39

Underlined nucleotides indicate primer/template mismatches. http://www.jimmunol.org/ of the enzymatic complex involves the translocation of cytosolic SodCI and SodCII are functionally redundant in proteins to the endosomal membrane to form the active complex Perforin-2 KO phagocytes that generates the respiratory burst (44). Likewise, Perforin-2, a Having excluded the possibility that Perforin-2 deficiency results transmembrane protein of cytosolic vesicles, dynamically trans- in impaired ROS production, we next considered the possibility locates to and fuses with phagocytic vesicles containing bacteria that S. Typhimurium sodCI mutants are resistant to ROS in (24, 25). This raised the possibility that Perforin-2 is involved in Perforin-2–deficient but not –proficient phagocytes. As the genome of by guest on October 9, 2021 the assembly and/or activation of the NADPH oxidase. If true, the S. Typhimurium encodes a second periplasmic superoxide dismutase respiratory burst would be deficient in Perforin-2 KO phagocytes (SodCII), we considered the possibility that it provides resistance to and would account for the survival of S. Typhimurium sodCI ROS in Perforin-2 KO phagocytes, although previous studies have mutants in Perforin-2–deficient phagocytes. concluded that SodCII does not attenuate ROS toxicity in WT cells To determine whether Perforin-2 is required for ROS production, and animals (11, 43). To determine whether SodCI and SodCII are a luminol-based chemiluminescence assay was used to quantify functionally redundant, we infected Perforin-2–proficient and –defi- ROS productions in IFN-g–primed PEMs, peritoneal macrophages cient phagocytes with a sodCI sodCII double mutant. Unlike the isolated without thioglycollate stimulation, and neutrophils from ΔsodCI sodCII+ strain that was killed by WT but not Perforin-2 KO WT and Perforin-2 KO mice. We found that total ROS production phagocytes (Fig. 1C, 1D), the ΔsodCI sodCII::kan strain was killed by was equally robust in WT and Perforin-2 KO macrophages that both Perforin-2 WT and KO phagocytes (Fig. 3A, 3B). Complemen- were stimulated with PMA or LPS (Fig. 2A, 2B). The kinetics of tation of the double mutant with sodCII allowed the complemented ROS production were also similar, as macrophages of both ge- strain to proliferate in Perforin-2 KO but not WT phagocytes (Fig. 3C, notypes exhibited peak ROS production at 60 min. To confirm that 3D). Consistent with our findings that ROS production is similar in the chemiluminescent signal was due to ROS production by a both WT and Perforin-2–deficient phagocytes (Fig. 2), DPI inhibition NADPH oxidase, phagocytes were pretreated with DPI, an in- of NADPH oxidase allowed the Salmonella double mutant to prolif- hibitor of NAPDH oxidases. As expected, DPI-treated cells pro- erate in both Perforin-2 WT and KO PEMs (Fig. 4C). Likewise, both duced negligible amounts of ROS (Fig. 2). Likewise, ROS WT Salmonella and the sodCI mutant reached higher cellular loads in production was also negligible in unstimulated macrophages (data DPI-treated PEMs than in untreated PEMs (Fig. 4). Thus, in agree- not shown). As with macrophages, we found few statistically ment with previous studies (43, 45), we conclude that SodCII does not significant differences between total ROS production in WT and protect against ROS in WT phagocytes. However, the nullification of Perforin-2 KO neutrophils (Fig. 2C, 2D). As was the case for total SodCII is clearly dependent upon Perforin-2 because SodCII is able to ROS, phagocytic ROS production was similar in both WT and protect phagocytosed S. Typhimurium from the bactericidal effects of Perforin-2 KO neutrophils (Fig. 2E). Although the signal was too ROS in Perforin-2 KO phagocytes. low for us to quantify phagocytic ROS production in macrophages, we note that bacterial superoxide dismutases promote the survival of phagocytosed S. typhimurium in WT and Peforin-2 KO Perforin-2 facilitates the degradation of Ags enclosed within PEMs (Figs. 1, 3). This indicates that both types of PEMs produce the bacterial envelope bactericidal levels of phagocytic ROS. Thus, we conclude that Although SodCI and SodCII have similar enzymatic properties, impaired ROS production cannot account for the survival of only SodCI provides resistance to endosomal ROS in WT S. Typhimurium DsodCI mutants in Perforin-2 KO phagocytes. phagocytes (43). Studies by the Slauch laboratory have shown that The Journal of Immunology 5

FIGURE 1. Perforin-2 limits the survival of phagocytosed bacteria. As indicated, PEM or peritoneal neutrophils were isolated from WT and Perforin-2 KO mice and stimulated with IFN-g for 14 h prior to infection with (A and B)WTS. Typhimurium strain GPM2004 or (C and D)strainST188(sodCI::kan). Gentamicin was Downloaded from used to eliminate extracellular bacteria, and intracellular bacteria were enumerated by plating cellular lysates. Mean and SD are shown (n = 3). Two-way ANOVAwith Sidak multiple comparisons test. *p # 0.05. http://www.jimmunol.org/ by guest on October 9, 2021

this phenomenon is not the result of differential expression in Perforin-2 KO but not WT phagocytes and animals (Fig. 3 and (43). Rather, it is due to differential degradation of the two su- below). Moreover, the apparent Perforin-2–dependent destruction of peroxide dismutases; SodCI is protease-resistant, whereas SodCII SodCII was not due to differences in bacterial load because the dif- is protease-sensitive (11, 43, 45). Thus, SodCII does not protect ference in SodCII abundance was statistically significant when ex- phagocytosed S. Typhimurium because it is proteolytically de- perimental replicates were quantified and normalized to DnaK of the graded. However, we have found that SodCII is functional in bacterial cytoplasm (Fig. 5B). A further indication that the bacterial Perforin-2 KO phagocytes (Fig. 3). Therefore, we considered it loads were similar in these experiments is the fact that the difference possible that the degradation of SodCII is Perforin-2–dependent. in the amount of DnaK normalized to host b-actin was insignificant, To determine whether this is the case, we infected WT and KO whereas the difference in the amount of SodCII normalized to b-actin PEMs with a strain of S. Typhimurium expressing SodCII-FLAG was significant (Fig. 5B). so that we could track the persistence of SodCII in the periplasm We also examined the degradation of the three bacterial Ags at with a mAb against the FLAG epitope. We also used Abs against earlier time points. Because these experiments required signif- DnaK and flagellin to monitor the abundance of cytoplasmic and icantly more phagocytes, we used BMDM, which can be ob- surface Ags, respectively. Western blots of intracellular bacteria tained at higher yields than PEMs. As with PEMs, extracellular recovered 18 h postinfection revealed that extracellular flagellin, flagellin was efficiently degraded in both WT and Perforin-2 KO which was abundant on the surface of the bacteria prior to BMDM (Fig. 5C). There was also apparent degradation of both phagocytosis, was efficiently degraded in both WT and KO PEMs SodCII and DnaK in both types of cells. Nevertheless, there was (Fig. 5A). In contrast, there was a clear difference in the abun- less SodCII in bacteria recovered from WT than KO phagocytes. dance of SodCII in bacteria recovered from Perforin-2 KO and This difference was statistically significant at 3 and 6 h when WT PEMs (Fig. 5A). We acknowledge that the lack of suitable normalized to DnaK, although concurrent degradation of DnaK Abs against SodCII prevents us from determining by Western blot likely results in an underestimation of the difference (Fig. 5D). if the loss of the SodCII signal is the result of the complete de- The degradation of DnaK also appears to lag that of SodCII, struction of SodCII or merely clipping of the FLAG epitope. How- although it is unclear if this is due to differences in protease ever, our in vitro data are consistent with Perforin-2–dependent accessibility and/or susceptibility. Differences in expression destruction of SodCII because SodCII confers a protective phenotype may also contribute to the apparent differences in degradation 6 PERFORIN-2 BREACHES THE ENVELOPE OF PHAGOCYTOSED BACTERIA Downloaded from http://www.jimmunol.org/ by guest on October 9, 2021

FIGURE 2. Perforin-2 is not required for ROS production in phagocytes. As indicated, WT and Perforin-2 KO peritoneal (A and B) macrophages and (C and D) neutrophils were stimulated with PMA or LPS to elicit ROS production, which was detected by luminol-based chemiluminescence. (E) Phagocytic ROS production of Perforin-2 WT or KO neutrophils induced with PMA was detected after phagocytosis of luminol-coupled beads. An enhancer was added to amplify chemiluminescence when macrophages were used. As indicated, some cells were also treated with DPI, an inhibitor of the phagocytic NAPDH oxidase. ROS activity is reported as mean relative light units (RLUs) 6 SD (n = 3). Data were analyzed by two-way ANOVAwith Sidak multiple comparisons test. *p # 0.05. rates of SodCII compared with DnaK. Relative to b-actin, the pathogen-associated molecular patterns and an indication of differences in SodCII in WT compared with Perforin-2 KO Perforin-2 trafficking (24), was similar when RAW264.7 cells BMDM was significant even at the 1 h time point. This cannot expressing Perforin-2–E2-Crimson were infected with either WT be due to higher loads of the bacteria in Perforin-2 KO S. Typhimurium or sodCI sodCII double mutant (Supplemental phagocytes because the amount of DnaK normalized to b-actin Fig. 2). Although there was no a priori reason to suspect that the also decreased over time, even in KO cells. Additionally, ex- presence or absence of bacterial superoxide dismutases would trapolation of our PEMs bactericidal assays suggests the dif- affect the intracellular trafficking of Perforin-2, the latter results ferences in bacterial loads are likely to be negligible at early further support a model in which Perforin-2 attacks the envelope time points, especially at 1 h. of phagocytosed bacteria. In aggregate, our results demonstrate that Perforin-2 facilitates the degradation of internal but not extracellular Ags of phago- Perforin-2 negates SodCII in vivo cytosed bacteria. This is not due to differences in the acidification Having established that Perforin-2 facilitates the degradation of of SCVs, which could affect protease activity, because the SodCII in vitro, the relevance of our observations was evaluated acidification profiles of SCVs in WT and Perforin-2 KO PEMs in a murine infection model. In brief, WT and Perforin-2 KO were found to be similar (Supplemental Fig. 1). Rather, it most mice were inoculated i.p. with a mixture of WT S. Typhimurium likely involves a direct interaction between Perforin-2 and the and a sodCI mutant at a 1:1 ratio. Four days postinfection, liver bacterial envelope because as a type I transmembrane protein, and spleen homogenates were plated on selective media to the putative pore-forming MACPF domain of Perforin-2 would enumerate the load of each strain. Consistent with previous be located inside the vacuole with the bacteria. Moreover, studies, fewer sodCI::kan bacteria were recovered than WT Perforin-2 has been shown to colocalize with phagocytosed bacteria recovered from Perforin-2 WT mice. The derived CI WT S. Typhimurium (25). The rapid formation of intracellular were accordingly low and demonstrate that the sodCI mutant is punctate bodies containing Perforin-2, which is triggered by significantly attenuated in Perforin-2–proficient mice (Fig. 6A). The Journal of Immunology 7 Downloaded from http://www.jimmunol.org/

FIGURE 3. SodCII is functional in the absence of Perforin-2. As indicated, PEM or peritoneal neutrophils were isolated from WT and Perforin-2 KO mice and stimulated with IFN-g for 14 h prior to infection with S. Typhimurium strain (A and B)ST189(ΔsodCI sodCII::kan)or (C and D) GPM2008 (ΔsodCI sodCII::kan sodCII+). Gentamicin was used by guest on October 9, 2021 to eliminate extracellular bacteria, and intracellular bacteria were enumerated by plating cellular lysates. Mean and SD are shown (n = 3). Data were analyzed by two-way ANOVA with Sidak multiple comparisons test. *p # 0.05.

In contrast, there was little to no attenuation of the sodCI::kan strain in Perforin-2 KO mice as indicated by CI near or at 1.0 (Fig. 6A). Similar results were obtained with a strain of FIGURE 4. Inhibition of ROS production allows Salmonella to pro- S. Typhimurium that had spontaneously lost the pSLT virulence liferate intracellularly. PEMs from WT and Perforin-2 KO mice were plasmid (Fig. 6B) (46). Thus, SodCI is not essential in the stimulated with IFN-g 14 h prior to infection. As indicated, some cells absence of Perforin-2. were also treated with DPI 30 min before infection with (A)WT Based on our in vitro studies, the persistence of SodCII was the S. Typhimurium strain GPM2004, (B)strainST188(sodCI::kan), or (C) most likely explanation for the lack of attenuation of the sodCI S. Typhimurium strain ST189 (DsodCI sodCII::kan). Gentamicin was mutant in Perforin-2 KO mice. Indeed, this was found to be the used to eliminate extracellular bacteria, and intracellular bacteria were case because a sodCI sodCII double mutant was found to be enumerated by plating cellular lysates. Mean and SD are shown (n =3). signiﬁcantly attenuated relative to WT bacteria in Perforin-2 KO Data were analyzed by two-way ANOVA with Sidak multiple compari- mice (Fig. 6C). Furthermore, complementation of the sodCI sons test. *p # 0.05, comparison of DPI-treated Perforin-2 WT and KO cells, **p # 0.05, comparison of Perforin-2 WT and KO cells without sodCII double mutant with sodCII resulted in a strain that was as DPI treatment. virulent as WT S. Typhimurium in Perforin-2 KO mice (Fig. 6D). In fact, there appeared to be a slight competitive advantage of the complemented strain over the WT strain as indicated by CI .1. This could be the result of higher expression levels of sodCII Discussion from a heterologous promoter in our construct. In contrast, Perforin-2 is a type I transmembrane protein that we have previ- complementation with sodCII failed to rescue the double mutant ously shown localizes to endosomal vesicles as well as the in WT mice (Fig. 6D). In aggregate, our in vivo studies dem- endoplasmic reticulum, Golgi, and plasma membrane (25). A onstrate that SodCII confers a protective advantage in Perforin-2 separate study that focused on the proteome of endocytic vesicles KO but not WT mice. As such, they are consistent with our also found that Perforin-2 is present in endosomes following in vitro ﬁnding that Perforin-2 facilitates the degradation of phagocytosis of latex beads by J774 macrophages (47). The same SodCII (Fig. 7). study also found that Perforin-2 was more abundant in late 8 PERFORIN-2 BREACHES THE ENVELOPE OF PHAGOCYTOSED BACTERIA

FIGURE 5. Perforin-2 facilitates the degradation of Ags enclosed within the bacterial envelope. (A) PEMs isolated from Perforin-2 WT and KO mice were infected with a strain of S. Typhimurium expressing SodCII-FLAG. After 18 h, the phagocytosed bacteria were recovered, and the indicated Ags were detected by Western blot. (B) Quantiﬁcation of Western blot data from three separate experiments in PEMs. (C) Downloaded from Similar infection experiments were conducted with BMDM, except the phagocytosed bacteria were recovered at 1, 3, and 6 h postphagocytosis. (D) Quan- tiﬁcation of Western blot data from three separate experiments in BMDM. Mean and SD are shown (n = 3). Data were analyzed by two-way ANOVA with

Sidak multiple comparisons test. *p # 0.05. http://www.jimmunol.org/ by guest on October 9, 2021

endosomes and lysosomes than early endosomes. Perforin-2 is ROS has been confirmed by several studies that have shown also coincident with subunits of the phagocytic NAPDH oxidase, that sodCI mutants are more susceptible to ROS killing in vitro proton transporters, and many other antimicrobial effectors of and less virulent than WT Salmonella in vivo (43, 45, 54, 55). phagosomes and/or the phagolysosomes (47, 48). LPS stimulation However, we have found that a sodCI null mutant is able to of BMDM has also been shown to increase the abundance of proliferate in Perforin-2–deficient phagocytes. Moreover, the Perforin-2 in endolysosomes compared with untreated cells (48). sodCI null mutant is as virulent as WT S. Typhimurium in This is consistent with our own studies in which we reported that Perforin-2–deficient but not –proficient mice. Because we have LPS results in the accumulation of Perforin-2 in vesicular struc- found that the production of phagocytic ROS is independent tures and that Perforin-2 colocalizes with phagocytosed bacteria of Perforin-2, the survival of the sodCI mutant is not because such as Escherichia coli and S. Typhimurium (24, 25). In aggre- of differences in ROS production. Rather it is because of the gate, these studies demonstrate that the subcellular distribution of persistence of SodCII, a second periplasmic superoxide dis- Perforin-2 is consistent with its ability to facilitate the destruction mutase in Perforin-2–deficient phagocytes. Consistent with our of phagocytosed bacteria. in vitro results, we have also found that SodCI and SodCII are Although most phagocytosed bacteria are rapidly killed, functionally redundant in Perforin-2 KO mice. some are able to resist phagocytic antimicrobials and even The persistence of SodCII was unexpected because previous survive within professional phagocytes. The latter includes studies have shown that it is normally degraded by proteases of S. Typhimurium, which must survive the respiratory burst and the phagolysosome (43, 45). In contrast, SodCI is resistant to other antimicrobial assaults prior to the formation of SCVs: proteolytic degradation and thus provides resistance to ROS in specialized niches within macrophages that afford the pathogen the phagolysosome (43). Our results in Perforin-2 WT cells a more favorable environment than phagosomes or phag- and animals are consistent with these latter studies because olysosomes (49). A central player in the survival of the respi- sodCII is unable to complement a sodCI sodCII double mutant. ratory burst is SodCI, a periplasmic superoxide dismutase that However, sodCII is able to complement the double mutant in converts superoxide to hydrogen peroxide, which is subse- Perforin-2 KO animals and isolated phagocytes. Thus, the quently detoxified by bacterial catalases and peroxidases (50– presence of Perforin-2 is associated with the inactivation of 53). The pivotal role of SodCI in protecting Salmonella from SodCII. The Journal of Immunology 9 Downloaded from

FIGURE 6. SodCII is functional in Perforin-2 KO but not WT mice. WT and Perforin-2 KO mice were inoculated by i.p. injection with S. Typhimurium

WT strain GPM2004 and (A) DsodCI::kan strain ST188, (B) DsodCI::kan strain ST188b, (C) DsodCI sodCII::kan strain ST189, or (D) DsodCI sodCII::kan http://www.jimmunol.org/ sodCII+ strain GPM2008 at a 1:1 ratio. Organs were harvested 4 d postinfection, and strains were enumerated on selective media. CI are derived from the ratios of mutant strains to WT strains, with compensation for any differences in inocula. Medians and SD are indicated by horizontal bars (n = 5–11). Data were analyzed by two-way ANOVA with Sidak multiple comparisons test. *p # 0.05.

How does Perforin-2 inactivate SodCII? Consistent with pre- (Fig. 7). This model is consistent with our experimental ob- vious studies, we have shown that SodCII is proteolytically servations with S. Typhimurium because the protease that degraded in the phagosome and/or phagolysosome in Perforin-2– degrades SodCII would enter the periplasmic space through by guest on October 9, 2021 proficient cells (43, 45). However, the degradation of SodCII is poly–Perforin-2 pores. Because SodCII is not anchored in the significantly attenuated in phagocytes lacking Perforin-2. Similar periplasm, it may also diffuse through the pore and be degraded results were observed with cytoplasmic DnaK in BMDM. This in the lumen of the phagosome (11, 43, 45). In either case, was not due to a general defect in protease activity in the SodCII is able to persist in the periplasm and protect the bac- phagolysosome because the surface Ag flagellin was degraded terium from the bactericidal effects of ROS when Perforin-2 is whether Perforin-2 was present or not. Thus, Perforin-2 facil- absent. itates the degradation of Ags contained within the envelope of In addition to Perforin-2, there is evidence that the cath- phagocytosed bacteria. However, it is unlikely that Perforin-2 elicidins CRAMP and LL-37 also play a role in disrupting the is itself a protease because it lacks significant homology to a envelope of phagocytosed Gram-negative bacteria in murine and known protease or protease motif. What Perforin-2 does have is human macrophages, respectively (11–14). Of particular rele- a MACPF domain (Fig. 7) (56–58). This suggests that Perforin- vance to this study are previous studies that have shown that 2 is a pore-forming protein, and putative Perforin-2 pores have CRAMP is active against phagocytosed S. Typhimurium (11, been imaged by transmission electron microscopy (25). How- 12). In one, it was shown that CRAMP inhibits the division of ever, to date, it has not been possible to confirm that the imaged S. Typhimurium. The result was filamentous bacteria, and it structures contain Perforin-2 because of the absence of suitable was further shown that filamentation was protease-dependent Abs. Nor has it been confirmed that the structures are in fact (12). In another study, it was shown that CRAMP is associated pores. However, other MACPF-containing proteins such as with the loss of SodCII from the periplasm of S. Typhimurium complement protein C9 and Perforin have been shown to po- in vitro (11). Moreover, SodCII promoted the survival of lymerize and form pores in lipid membranes (17, 20, 22, 59, S. Typhimurium in CRAMP-deficient but not -proficient mice. 60). For example, 22 monomers of C9 polymerize to form a However, in the latter study, the authors also noted that the loss pore in the outer membrane of Gram-negative bacteria with an of CRAMP failed to fully abolish the degradation of SodCII. inner diameter of 120 A˚ (22, 59, 61). Because Perforin-2 is a Our study suggests that this is most likely due to the activity of type I transmembrane protein, with its membrane spanning Perforin-2. Although it is clear that Perforin-2 and cathelicidins a-helix near its carboxy terminus, the MACPF domain of can act independently of one another, it remains to be deter- Perforin-2 would reside in the lumen of endosomes and phag- mined if they also act synergistically. In the case of Gram-negative osomes. In this orientation, its MACPF domain would reside in bacteria, a particularly intriguing model is the possibility that the same compartment as phagocytosed bacteria. Thus, we Perforin-2 forms a conduit in the outer membrane through which propose that Perforin-2 polymerizes (perhaps after cleavage cathelicidin or other antimicrobial peptides transit to reach the from its transmembrane domain by a lysosomal protease) bacterial inner membrane. Alternatively, the deployment of in- and forms pores in the envelope of phagocytosed bacteria dependent mechanisms to disrupt the envelope of phagocytosed 10 PERFORIN-2 BREACHES THE ENVELOPE OF PHAGOCYTOSED BACTERIA

12. Rosenberger, C. M., R. L. Gallo, and B. B. Finlay. 2004. Interplay between antibacterial effectors: a macrophage antimicrobial peptide impairs intracellular Salmonella replication. Proc. Natl. Acad. Sci. USA 101: 2422–2427. 13. Sonawane, A., J. C. Santos, B. B. Mishra, P. Jena, C. Progida, O. E. Sorensen, R. Gallo, R. Appelberg, and G. Griffiths. 2011. Cathelicidin is involved in the intracellular killing of mycobacteria in macrophages. Cell. Microbiol. 13: 1601–1617. 14. Stephan, A., M. Batinica, J. Steiger, P. Hartmann, F. Zaucke, W. Bloch, and M. Fabri. 2016. LL37:DNA complexes provide antimicrobial activity against intracellular bacteria in human macrophages. Immunology 148: 420–432. 15. Spilsbury, K., M. A. O’Mara, W. M. Wu, P. B. Rowe, G. Symonds, and Y. Takayama. 1995. Isolation of a novel macrophage-specific gene by differential cDNA analysis. Blood 85: 1620–1629. 16. Gudmundsson, G. H., B. Agerberth, J. Odeberg, T. Bergman, B. Olsson, and R. Salcedo. 1996. The human gene FALL39 and processing of the cathelin precursor to the antibacterial peptide LL-37 in granulocytes. Eur. J. Biochem. 238: 325–332. 17. Law, R. H., N. Lukoyanova, I. Voskoboinik, T. T. Caradoc-Davies, K. Baran, M. A. Dunstone, M. E. D’Angelo, E. V. Orlova, F. Coulibaly, S. Verschoor, et al. 2010. The structural basis for membrane binding and pore formation by lymphocyte perforin. Nature 468: 447–451. 18. Voskoboinik, I., J. C. Whisstock, and J. A. Trapani. 2015. Perforin and granzymes: function, dysfunction and human pathology. Nat. Rev. Immunol. 15: 388–400. 19. Lichtenheld, M. G., K. J. Olsen, P. Lu, D. M. Lowrey, A. Hameed, H. Hengartner, and E. R. Podack. 1988. Structure and function of human perforin. Nature 335: 448–451. 20. Podack, E. R., K. J. Olsen, D. M. Lowrey, and M. Lichtenheld. 1989. Structure and function of perforin. Curr. Top. Microbiol. Immunol. 140: 11–17. 21. Thiery, J., D. Keefe, S. Boulant, E. Boucrot, M. Walch, D. Martinvalet, Downloaded from FIGURE 7. Model for Perforin-2–dependent degradation of SodCII. I. S. Goping, R. C. Bleackley, T. Kirchhausen, and J. Lieberman. 2011. Perforin (A) Domain organization of Perforin-2. MACPF domains are present in other pores in the endosomal membrane trigger the release of endocytosed granzyme proteins of the innate immune system, and in the case of complement protein C9 B into the cytosol of target cells. Nat. Immunol. 12: 770–777. and Perforin-1 have been shown to polymerize and form pores in lipid mem- 22. Dudkina, N. V., B. A. Spicer, C. F. Reboul, P. J. Conroy, N. Lukoyanova, H. Elmlund, R. H. Law, S. M. Ekkel, S. C. Kondos, R. J. Goode, et al. 2016. branes. The P2 domain is a domain of unknown function that is conserved in Structure of the poly-C9 component of the complement membrane attack orthologs of Perforin-2. As a type I transmembrane protein, the MACPF domain complex. Nat. Commun. 7: 10588. of Perforin-2 would reside within the lumen of endosomes. (B) Upon activation 23. McCormack, R., L. R. de Armas, M. Shiratsuchi, J. E. Ramos, and E. R. Podack. http://www.jimmunol.org/ and likely cleavage from its transmembrane domain, Perforin-2 polymerizes on 2013. Inhibition of intracellular bacterial replication in fibroblasts is dependent on the perforin-like protein (perforin-2) encoded by macrophage-expressed gene the outer membrane of S. Typhimurium. Subsequent pore formation allows 1. J. Innate Immun. 5: 185–194. proteases and other antimicrobial effectors to enter the periplasmic space and 24. McCormack, R. M., K. Lyapichev, M. L. Olsson, E. R. Podack, and degrade SodCII. Alternatively, SodCII may diffuse through the poly–Perforin-2 G. P. Munson. 2015. Enteric pathogens deploy cell cycle inhibiting factors to pore and be destroyed in the lumen of the phagolysosome. Pores do not lead to block the bactericidal activity of Perforin-2. Elife 4: e06505. 25. McCormack, R. M., L. R. de Armas, M. Shiratsuchi, D. G. Fiorentino, the inactivation of SodCI because it is protease-resistant and tethered within the M. L. Olsson, M. G. Lichtenheld, A. Morales, K. Lyapichev, L. E. Gonzalez, periplasm. Wavy lines represent Ags on the outer surface of the bacterial N. Strbo, et al. 2015. Perforin-2 is essential for intracellular defense of paren- membrane. Enzymatic reactions are represented by straight lines. chymal cells and phagocytes against pathogenic bacteria. Elife 4: e06508. 26. McCormack, R., W. Bahnan, N. Shrestha, J. Boucher, M. Barreto, C. M. Barrera,

E. A. Dauer, N. E. Freitag, W. N. Khan, E. R. Podack, and K. Schesser. 2016. by guest on October 9, 2021 bacteria may be an insurance strategy against pathogen resis- Perforin-2 protects host cells and mice by restricting the vacuole to cytosol transitioning of a bacterial pathogen. Infect. Immun. 84: 1083–1091. tance to any particular mechanism. 27.Benard,E.L.,P.I.Racz,J.Rougeot,A.E.Nezhinsky,F.J.Verbeek, H. P. Spaink, and A. H. Meijer. 2015. Macrophage-expressed perforins mpeg1 and mpeg1.2 have an anti-bacterial function in zebrafish. J. Innate Immun. 7: Disclosures 136–152. The authors have no financial conflicts of interest. 28. Nikaido, H., M. Levinthal, K. Nikaido, and K. Nakane. 1967. Extended deletions in the histidine-rough-B region of the Salmonella chromosome. Proc. Natl. Acad. Sci. USA 57: 1825–1832. 29. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal References genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 1. Karimi, G., C. Houeé Levin, M. C. Dagher, L. Baciou, and T. Bizouarn. 2014. 97: 6640–6645. Assembly of phagocyte NADPH oxidase: a concerted binding process? Biochim. 30. Haldimann, A., and B. L. Wanner. 2001. Conditional-replication, integration, Biophys. Acta 1840: 3277–3283. excision, and retrieval plasmid-host systems for gene structure-function studies 2. Nauseef, W. M. 2004. Assembly of the phagocyte NADPH oxidase. Histochem. of bacteria. J. Bacteriol. 183: 6384–6393. Cell Biol. 122: 277–291. 31. Uzzau, S., N. Figueroa-Bossi, S. Rubino, and L. Bossi. 2001. Epitope tagging of 3. Cederlund, A., G. H. Gudmundsson, and B. Agerberth. 2011. Antimicrobial chromosomal genes in Salmonella. Proc. Natl. Acad. Sci. USA 98: 15264–15269. peptides important in innate immunity. FEBS J. 278: 3942–3951. 32.Amann,E.,B.Ochs,andK.J.Abel.1988.Tightlyregulatedtacpromoter 4. Ellison, R. T. , III, and T. J. Giehl. 1991. Killing of gram-negative bacteria by vectors useful for the expression of unfused and fused proteins in Escherichia lactoferrin and lysozyme. J. Clin. Invest. 88: 1080–1091. coli. Gene 69: 301–315. 5. Martinez, R. J., and S. F. Carroll. 1980. Sequential metabolic expressions of the 33. Zhang, X., R. Goncalves, and D. M. Mosser. 2008. The isolation and charac- lethal process in human serum-treated Escherichia coli: role of lysozyme. Infect. terization of murine macrophages. Curr. Protoc. Immuno. Chapter 14: Unit 14 1. Immun. 28: 735–745. 34. Nathan, C. F., and R. K. Root. 1977. Hydrogen peroxide release from mouse 6. Gallo, R. L., K. J. Kim, M. Bernfield, C. A. Kozak, M. Zanetti, L. Merluzzi, and peritoneal macrophages: dependence on sequential activation and triggering. R. Gennaro. 1997. Identification of CRAMP, a cathelin-related antimicrobial J. Exp. Med. 146: 1648–1662. peptide expressed in the embryonic and adult mouse. J. Biol. Chem. 272: 13088– 35. Luo, Y., and M. E. Dorf. 2001. Isolation of mouse neutrophils. Curr. Protoc. 13093. Immunol. Chapter 3: Unit 3 20. 7. Turner, J., Y. Cho, N. N. Dinh, A. J. Waring, and R. I. Lehrer. 1998. Activities of LL-37, a cathelin-associated antimicrobial peptide of human neutrophils. Anti- 36. Lutwyche, P., C. Cordeiro, D. J. Wiseman, M. St-Louis, M. Uh, M. J. Hope, microb. Agents Chemother. 42: 2206–2214. M. S. Webb, and B. B. Finlay. 1998. Intracellular delivery and antibacterial 8. Wiesner, J., and A. Vilcinskas. 2010. Antimicrobial peptides: the ancient arm of activity of gentamicin encapsulated in pH-sensitive liposomes. Antimicrob. the human immune system. Virulence 1: 440–464. Agents Chemother. 42: 2511–2520. 9. Zanetti, M. 2004. Cathelicidins, multifunctional peptides of the innate immunity. 37. Laroux, F. S., X. Romero, L. Wetzler, P. Engel, and C. Terhorst. 2005. Cutting J. Leukoc. Biol. 75: 39–48. edge: MyD88 controls phagocyte NADPH oxidase function and killing of gram- 10. Henzler Wildman, K. A., D. K. Lee, and A. Ramamoorthy. 2003. Mechanism of negative bacteria. J. Immunol. 175: 5596–5600. lipid bilayer disruption by the human antimicrobial peptide, LL-37. Biochemistry 38. Yamazaki, T., C. Kawai, A. Yamauchi, and F. Kuribayashi. 2011. A highly 42: 6545–6558. sensitive chemiluminescence assay for superoxide detection and chronic 11. Kim, B., S. M. Richards, J. S. Gunn, and J. M. Slauch. 2010. Protecting against granulomatous disease diagnosis. Trop. Med. Health 39: 41–45. antimicrobial effectors in the phagosome allows SodCII to contribute to viru- 39. Drecktrah, D., L. A. Knodler, R. Ireland, and O. Steele-Mortimer. 2006. The lence in Salmonella enterica serovar Typhimurium. J. Bacteriol. 192: 2140– mechanism of Salmonella entry determines the vacuolar environment and in- 2149. tracellular gene expression. Traffic 7: 39–51. The Journal of Immunology 11

40. Schindelin, J., I. Arganda-Carreras, E. Frise, V. Kaynig, M. Longair, T. Pietzsch, 50. Aussel, L., W. Zhao, M. He´brard, A. A. Guilhon, J. P. Viala, S. Henri, L. Chasson, S. Preibisch, C. Rueden, S. Saalfeld, B. Schmid, et al. 2012. Fiji: an open-source J. P. Gorvel, F. Barras, and S. Me´resse. 2011. Salmonella detoxifying enzymes are platform for biological-image analysis. Nat. Methods 9: 676–682. sufficient to cope with the host oxidative burst. Mol. Microbiol. 80: 628–640. 41. Steele-Mortimer, O. 2008. The Salmonella-containing vacuole: moving with the 51. He´brard, M., J. P. Viala, S. Me´resse, F. Barras, and L. Aussel. 2009. Redundant times. Curr. Opin. Microbiol. 11: 38–45. hydrogen peroxide scavengers contribute to Salmonella virulence and oxidative 42. Fields, K. A., R. McCormack, L. R. de Armas, and E. R. Podack. 2013. Perforin- stress resistance. J. Bacteriol. 191: 4605–4614. 2 restricts growth of Chlamydia trachomatis in macrophages. Infect. Immun. 81: 52. Slauch, J. M. 2011. How does the oxidative burst of macrophages kill bacteria? 3045–3054. Still an open question. Mol. Microbiol. 80: 580–583. 43. Krishnakumar, R., M. Craig, J. A. Imlay, and J. M. Slauch. 2004. Differences in 53. Storz, G., and J. A. Imlay. 1999. Oxidative stress. Curr. Opin. Microbiol. 2: 188–194. enzymatic properties allow SodCI but not SodCII to contribute to virulence 54. Craig, M., and J. M. Slauch. 2009. Phagocytic superoxide specifically damages in Salmonella enterica serovar Typhimurium strain 14028. J. Bacteriol. 186: an extracytoplasmic target to inhibit or kill Salmonella. PLoS One 4: e4975. 5230–5238. 55. Fang, F. C., M. A. DeGroote, J. W. Foster, A. J. Baümler, U. Ochsner, 44. Panday, A., M. K. Sahoo, D. Osorio, and S. Batra. 2015. NADPH oxidases: an T. Testerman, S. Bearson, J. C. Gia´rd, Y. Xu, G. Campbell, and T. Laessig. 1999. overview from structure to innate immunity-associated pathologies. Cell. Mol. Virulent Salmonella typhimurium has two periplasmic Cu, Zn-superoxide dis- Immunol. 12: 5–23. mutases. Proc. Natl. Acad. Sci. USA 96: 7502–7507. 45. Krishnakumar, R., B. Kim, E. A. Mollo, J. A. Imlay, and J. M. Slauch. 56. McCormack, R., and E. R. Podack. 2015. Perforin-2/Mpeg1 and other pore- 2007. Structural properties of periplasmic SodCI that correlate with viru- forming proteins throughout evolution. J. Leukoc. Biol. 98: 761–768. lence in Salmonella enterica serovar Typhimurium. J. Bacteriol. 189: 57. Podack, E. R., and G. P. Munson. 2016. Killing of microbes and cancer by the 4343–4352. immune system with three mammalian pore-forming killer proteins. Front. 46. McClelland, M., K. E. Sanderson, J. Spieth, S. W. Clifton, P. Latreille, L. Courtney, Immunol. 7: 464. S. Porwollik, J. Ali, M. Dante, F. Du, et al. 2001. Complete genome sequence of 58. Ni, T., and R. J. C. Gilbert. 2017. Repurposing a pore: highly conserved perforin- Salmonella enterica serovar Typhimurium LT2. Nature 413: 852–856. like proteins with alternative mechanisms. Philos. Trans. R. Soc. Lond. B Biol. 47. Duclos, S., G. Clavarino, G. Rousserie, G. Goyette, J. Boulais, V. Camossetto, Sci. 372: 20160212. E. Gatti, S. LaBoissie`re, P. Pierre, and M. Desjardins. 2011. The endosomal 59. Podack, E. R., J. Tschoop, and H. J. Mu¨ller-Eberhard. 1982. Molecular orga- proteome of macrophage and dendritic cells. Proteomics 11: 854–864. nization of C9 within the membrane attack complex of complement. Induction of 48. Nakamura, N., J. R. Lill, Q. Phung, Z. Jiang, C. Bakalarski, A. de Mazie`re, circular C9 polymerization by the C5b-8 assembly. J. Exp. Med. 156: 268–282. J. Klumperman, M. Schlatter, L. Delamarre, and I. Mellman. 2014. Endosomes 60. Podack, E. R., and J. Tschopp. 1982. Polymerization of the ninth component of Downloaded from are specialized platforms for bacterial sensing and NOD2 signalling. Nature 509: complement (C9): formation of poly(C9) with a tubular ultrastructure resembling the 240–244. membrane attack complex of complement. Proc. Natl. Acad. Sci. USA 79: 574–578. 49. Anderson, C. J., and M. M. Kendall. 2017. Salmonella enterica serovar typhi- 61. Tschopp, J., and E. R. Podack. 1981. Membranolysis by the ninth component of murium strategies for host adaptation. Front. Microbiol. 8: 1983. human complement. Biochem. Biophys. Res. Commun. 100: 1409–1414. http://www.jimmunol.org/ by guest on October 9, 2021