To counteract pathogen dissemination, host cells reorganize their cytoskeleton to compartmentalize bacteria and induce autophagy. Using Shigella infection of tissue culture cells, host and pathogen determinants underlying this process are identified and characterized. Using zebrafish models of Shigella infection, the role of discovered molecules and mechanisms are investigated in vivo.
Shigella flexneri is an intracellular pathogen that can escape from phagosomes to reach the cytosol, and polymerize the host actin cytoskeleton to promote its motility and dissemination. New work has shown that proteins involved in actin-based motility are also linked to autophagy, an intracellular degradation process crucial for cell autonomous immunity. Strikingly, host cells may prevent actin-based motility of S. flexneri by compartmentalizing bacteria inside ‘septin cages’ and targeting them to autophagy. These observations indicate that a more complete understanding of septins, a family of filamentous GTP-binding proteins, will provide new insights into the process of autophagy. This report describes protocols to monitor autophagy-cytoskeleton interactions caused by S. flexneri in vitro using tissue culture cells and in vivo using zebrafish larvae. These protocols enable investigation of intracellular mechanisms that control bacterial dissemination at the molecular, cellular, and whole organism level.
Shigella flexneri, a Gram-negative invasive enteropathogenic bacterium, can escape from phagosomes to the cytosol, and polymerize the host actin cytoskeleton to evade cytosolic immune responses and promote intra- and intercellular movement1,2. Despite the understanding of actin-based motility in vitro3,4, the mechanisms restricting bacterial dissemination in vivo have not been fully defined. This is critical for a more complete understanding of innate immunity and host defense.
Septins, a highly conserved family of proteins among metazoans, are guanosine triphosphate (GTP)-binding proteins that assemble into hetero-oligomeric complexes and form nonpolar filaments that associate with cellular membranes and the cytoskeleton5,6. Recent work has discovered that infected host cells can prevent Shigella actin based motility by compartmentalizing bacteria targeted to autophagy inside ‘septin cages’, revealing the first cellular mechanism that counteracts actin based motility7,8. A wide open field of investigation now lies in ‘septin biology and infection’. Septin assembly, induced by a variety of pathogens (e.g., Listeria monocytogenes7,9,10, Mycobacterium marinum7,8, Candida albicans11), may emerge as a key issue in host defense5,12.
Autophagy, a highly conserved intracellular degradation process, is viewed as a key component of cell-autonomous immunity because of its ability to deliver cytosolic bacteria to the lysosome13,14. However, the role of bacterial autophagy in vivo to restrict or promote bacterial replication remains poorly understood15,16. The zebrafish (Danio rerio) has emerged as a vertebrate model for the study of infections because it is optically accessible at the larval stages when the innate immune system is already functional17,18. Recent work has characterized the susceptibility of zebrafish larvae to S. flexneri, a paradigm for bacterial autophagy15, and has used the Shigella-zebrafish infection model to study the manipulation of autophagy for antibacterial therapy in vivo19.
This report provides new tools and assays to study S. flexneri interactions with autophagy and the cytoskeleton. In a first step, protocols to monitor autophagy-cytoskeleton interactions are described using Shigella infection of the human epithelial cell line HeLa. To assess the role of autophagy-cytoskeleton interactions on the Shigella infection process in vitro, methods to manipulate autophagy and cytoskeleton components (using siRNA or pharmacological reagents) are provided. New work has shown that by using Shigella infection of zebrafish larvae, similar assays can be applied to study the cell biology of infection in vivo. Protocols to prepare and infect zebrafish larvae are detailed, and to assess the host response to Shigella infection in vivo, protocols to determine host survival and bacterial burden of infected larvae are provided. Methods to monitor the recruitment of septin and autophagy markers to Shigella (using either fixed or living zebrafish larvae) and methods to test the role of these processes in vivo [using morpholino oligonucleotides (injected in 1-4 cell stage embryos) or pharmacological reagents (added directly to the zebrafish bath water)] are also discussed. This program of work is expected to provide insights into the mechanisms required for the control of infection by cytosolic host responses.
1. Monitoring Autophagy and the Cytoskeleton In Vitro Using Tissue Culture Cells
2. Functional Analysis of Autophagy and the Cytoskeleton In Vitro
NOTE: Both genetic and pharmacological approaches can be used to perturb autophagy in infected tissue culture cells, and the impact of these treatments on the course of infection can be monitored.
3. In Vivo Imaging of S. flexneri Interactions with Autophagy and the Cytoskeleton
NOTE: The zebrafish model of Shigella infection can be used to investigate septin caging and autophagy in vivo19.
4. Functional Analysis of Autophagy and the Cytoskeleton In Vivo
NOTE: The impact of pharmacological and genetic perturbations of autophagy on the course of infection can be monitored at the whole-animal level, and at the level of the single cell.
Upon infection of tissue culture cells in vitro, S. flexneri can escape from the phagosome and invade the cytosol. In the cytosol, host cells can prevent the actin-based motility of Shigella by compartmentalizing bacteria inside septin cages (Figure 1A). Bacteria entrapped by septin cages can also be labeled by autophagy markers p62 (Figure 1B) and LC3 (Figure 1C). These observations highlight a novel mechanism of host defense that restricts dissemination of invasive pathogens, and also reveal new links between autophagy and the cytoskeleton. Strikingly, the depletion of autophagy markers significantly reduces septin caging of bacteria (Figure 2A), and work has also shown that the depletion of septin caging significantly reduces recruitment of autophagy markers8. Thus, at least in the case of Shigella, septin cage assembly and autophagosome formation can be viewed as interdependent processes. Other cellular requirements for compartmentalization of Shigella by septin cages include actin polymerization and actomyosin activity (Figure 2B).
There is no natural mouse model of shigellosis, and investigation of Shigella pathogenesis, septin biology and bacterial autophagy in vivo can benefit from a new animal model of infection, the zebrafish larvae19. It is possible to infect zebrafish larvae by injecting bacteria in various anatomical sites such as caudal intravenous injections for survival experiments, and tail muscle injections for in vivo microscopy (Figure 3A). Depending on the dose, S. flexneri injected in zebrafish larvae can either be cleared within 48 hr post-infection, or may result in a progressive and ultimately fatal infection (Figures 3B–3D). Shigella virulence factors are expressed at 28 ° C, the optimal growth temperature of zebrafish, and zebrafish infection by Shigella is strictly dependent upon its type III secretion system (T3SS)19, an essential virulence determinant in human disease. Taken together, these observations indicate that the zebrafish larva represents a valuable new host for in vivo analysis of Shigella infection.
The optical accessibility of zebrafish larvae enables visualization of septin caging in vivo (Figure 4A), an achievement that has never before been accomplished using mammalian host models. To complement evidence that septin cages entrap bacteria targeted to autophagy in vivo, one can infect transgenic zebrafish larvae expressing GFP-Lc3 and observe autophagy marker recruitment to Shigella (Figure 4B). For ultrastrucutral analysis of Shigella autophagosomes in vivo, electron microscopy can be used to clearly show the cytosolic sequestration of bacteria by double membrane vacuoles19. Autophagy is viewed as a key component of cell-autonomous immunity and a crucial defense mechanism against intracytosolic bacteria14-16. To characterize autophagy function in vivo, p62 morpholino-treated zebrafish larvae can be used. Unlike the core autophagy machinery [e.g., the 36 autophagy related proteins (ATGs)26], p62 is not essential for vertebrate development27 and thus zebrafish larvae can develop normally prior to infection. Strikingly, p62-depleted larvae inoculated with S. flexneri result in significantly increased mortality and increased bacterial burden19. In agreement with in vitro work showing that septin cage assembly is interdependent with autophagosome formation7,8, septin recruitment to Shigella is clearly reduced in p62-depleted larvae (Figure 4C). These data demonstrate that zebrafish survival depends on p62-mediated autophagy to control intracellular bacterial infection.
Figure 1. The septin cage in vitro. (A) HeLa cells were infected with S. flexneri for 4 hr 40 min, fixed, labeled with antibodies to SEPT9 and phalloidin, and imaged by confocal microcopy. Scale bar, 1 µm. (B) HeLa cells were infected with S. flexneri for 4 hr 40 min, fixed, labeled with antibodies to p62 and SEPT2, and imaged by fluorescent light microscopy. Scale bar, 1 µm. (C) HeLa cells were transfected with GFP-ATG8/LC3, infected with S. flexneri for 4 hr 40 min, fixed, labeled with antibodies to SEPT2, and imaged by fluorescent light microscopy. Scale bar, 1 µm. These figures have been modified from Mostowy et al7.
Figure 2. Cellular requirements for Shigella-septin cage formation. (A) HeLa cells were treated with control (CTRL), p62, ATG5, ATG6 or ATG7 siRNA. Whole-cell lysates of siRNA-treated cells were immunoblotted for GAPDH, p62, ATG5, ATG6, or ATG7 to show the efficiency of siRNA depletion (top). siRNA-treated cells were infected with S. flexneri for 4 hr 40 min, fixed, and labeled for quantitative microscopy. Graphs (bottom) represent the mean % ± SEM of Shigella inside SEPT2 cages from n ≥3 experiments per treatment. (B) HeLa cells were infected with S. flexneri, treated with DMSO, cytochalasin D (CytD), latrunculin B (LatB), nocodazole (Noco), or blebbistatin (Bleb) and after 4 hr 40 min were fixed and labeled for quantitative microscopy. Graphs represent the mean % ± SEM of Shigella inside SEPT2 cages from two independent experiments per treatment. These figures have been modified from Mostowy et al7.
Figure 3. The zebrafish model of Shigella infection. (A) Images to illustrate orientation of the zebrafish larva under the stereomicroscope. (Left panel) Zebrafish larvae 72 hr post fertilization were positioned laterally in the injection plate with their dorsal side facing the injection needle. (Middle panel) Bloodstream infection was performed by injecting the bacteria (red solution) in the caudal vein, posterior to the urogenital opening. (Right panel) Infection in the tail muscle was performed by injecting the bacteria (red solution) over a somite. (B) Survival curves of 72 hr post fertilization larvae injected with various doses of S. flexneri and incubated at 28 °C for 48 hr post infection. The effective inoculum was classified as low (<103 CFU, open circles), medium (~4 x 103 CFU, open triangles) or high (~104 CFU, open squares). Mean % ± SEM (horizontal bars) from n ≥3 experiments per inoculum class. (C) Enumeration of live bacteria in homogenates from individual larvae at various times post infection measured by plating onto LB. Note, only larvae having survived the infection are included in enumeration analysis. Mean ± SEM (horizontal bars) also shown. (D) Distribution of GFP-Shigella determined by live imaging using a fluorescent stereomicroscope at various times post infection using a low, medium, or high dose inoculum (caudal intravenous injections). Overlay of transmission image (grey) and GFP fluorescence (green). (B)–(D)These figures have been modified from Mostowy et al19.
Figure 4. The cell biology of Shigella infection in vivo. (A) Zebrafish larvae were infected in the tail muscle with GFP-Shigella (low dose) for 24 hr, fixed, labeled with antibodies against SEPT7 (red) and to GFP (green), and imaged by confocal microscopy. Scale bar, 5 µm. (B) GFP-Lc3 zebrafish larvae were infected with mCherry-Shigella (medium dose) for 4 hr, fixed, labeled with antibodies against mCherry (red) and to GFP (green), and imaged by confocal microscopy. Scale bar, 1.5 µm. (C) Zebrafish larvae treated with either control (CTRL; left image) or p62 (right image) morpholinos were infected with GFP-Shigella for 4 hr (medium dose), fixed, labeled with antibodies against SEPT7 (red) and to GFP (green), and imaged by confocal microscopy. Arrows highlight some examples of Shigella entrapped in septin cages (CTRL) or not (p62 depleted) a 4 hr post infection. Scale bar, 5 µm. These figures have been modified from Mostowy et al19.
When monitoring autophagy and the cytoskeleton in vitro using tissue culture cells, the protocols described in sections 1 and 2 can be applied to a wide variety of tissue culture cell types. Moreover, to follow autophagy (e.g., ATG8/LC3+ve autophagosomes) and cytoskeleton (e.g., actin tails, septin cages) dynamics in real-time during Shigella infection using live imaging, tissue culture cells can be transiently transfected using GFP-, RFP- or CFP-tagged constructs as previously described7,8. To increase the percentage of cells infected by Shigella (i.e., generally desirable for real-time analysis considering that Shigella can invade 5-30% of HeLa cells at 100:1 MOI), directly add 400 µl of Shigella (OD600 = 0.3-0.6) to cells in 2 ml MEM (serum-starved) and wait at least 1.5 hr post infection for sufficient bacterial entry, escape from the phagosome, replication, autophagy recognition and septin caging. Alternatively, one may use the Shigella M90T AfaI strain that expresses the adhesin AfaE and have much higher invasion abilities in epithelial cells compared to the M90T strain28. Of note, the M90T AfaI strain has not yet been tested in vivo using zebrafish. Plates of Shigella colonies can be kept at 4 °C for 2-3 days and used for several experiments. However, over time, colonies of Shigella that have lost the virulence plasmid can also absorb the Congo Red and appear to have retained their virulence plasmid. For this reason we recommend to use fresh bacterial stocks when possible.
When monitoring the cell biology of infection in vivo, protocols described in sections 3 and 4 use wildtype AB line zebrafish. To monitor Shigella-leukocyte interactions, transgenic zebrafish lines can be used, e.g., mpx:GFP or lyz:DsRed to visualize neutrophils19,29,30 or mpeg1:mcherry to visualize macrophages19,31. To visualize autophagy in vivo, the GFP-Lc3 zebrafish transgenic line19,24 can be used as described in section 3.8.
To perturb autophagy in vivo, the effective morpholino oligonucleotide dose has to be assessed experimentally based on its efficiency to inhibit transcript splicing or protein translation. It is advisable to perform a titration experiment and to confirm the depletion by RT-PCR (for splice morpholino oligonucleotide) or by SDS-PAGE (for translational morpholino oligonucleotide)32. RNA isolation from zebrafish embryos or larvae can be performed using guanidinium thiocyanate-phenol-chloroform extraction. To extract protein from zebrafish larvae (8 to 15 larvae/tube), mechanically homogenize using a pestle in 200 µl lysis buffer (1 M Tris, 5 M NaCl, 0.5 M EDTA, 0.01% octylphenol ethylene oxide condensate, and protease inhibitor). Centrifuge tubes at 19,000 x g at 4 ° C for 15 min and transfer the supernatant to a new tube. Add Laemmli buffer and heat the sample at 95 ° C for 15 min. Lysates can be stored at -80 ° C until needed, and can be evaluated by Western blotting as described in section 2.3.
The zebrafish is an excellent model for in vivo drug application. Analysis using morpholino oligonucleotides can be complemented with established drugs to manipulate autophagy (e.g., rapamycin and bafilomycin). Uninfected and/or infected larvae can be treated with rapamycin (1.5 µM) or bafilomycin (80 nM) diluted in E2 and autophagic flux can be evaluated by Western blotting as described in 25,33. The consequence of autophagy manipulation on the outcome of the infection and survival of the infected larvae can be evaluated as described in section 3.5.
In addition to studying host cell determinants, in vitro and in vivo protocols can be applied to assess bacterial determinants required for autophagy recognition, using bacterial mutant strains that are differentially recognized by autophagy, e.g., ShigellaΔicsA (the Shigella protein IcsA recruits N-WASP and then Arp2/3 for actin tail and septin cage formation; in its absence there can be no actin tails, no septin cages) and ShigellaΔicsB (Shigella avoids the autophagic response via the bacterial effector protein IcsB, which prevents the recruitment of autophagy machinery to IcsA; in its absence there can be more septin cages, more autophagy)7,8.
Shigella is not a natural pathogen of zebrafish and grows optimally at 37 °C. However, work has shown that virulence factors required for Shigella invasion, escape from the phagocytic vacuole and replication in the cytosol can be expressed and are functional in zebrafish larvae at 28 °C19. 28 °C is the most commonly used temperature for zebrafish rearing and standard temperature to ensure normal zebrafish development23. Strikingly, the major pathogenic events that lead to shigellosis in humans (i.e., macrophage cell death, invasion and multiplication within epithelial cells, cell-to-cell spread, inflammatory destruction of the host epithelium) are faithfully reproduced in the zebrafish model of Shigella infection19.
Autophagy and cytoskeleton genes are ubiquitously expressed and have a wide range of biological functions. Mouse studies have shown that knockout of essential autophagy26 or septin genes5 are embryonic lethal, and it is likely that some of these genes will also be essential for zebrafish development (although this problem may be reduced by the fact that zebrafish have multiple paralogous genes33). If so, there are several alternatives to overcome this issue, including (i) the use of pharmacological reagents to regulate autophagy and the cytoskeleton, (ii) morpholinos can be titrated down, (iii) knockout of genes can be designed for only specific cell types, and/or (iv) genes involved in autophagic recognition that are not essential for animal development (e.g., p62) may be targeted.
While the zebrafish is an ideal model system to investigate autophagy and the cytoskeleton during Shigella infection, molecular tools are currently lacking. The field needs to generate new tools and drive cell specific expression of the proteins of interest. To knock down expression of autophagy/cytoskeleton genes, new morpholino sequences are required, and novel methods for genome engineering (e.g., TALEN, CRISPR/Cas9) can also be used. In the meantime, several tools previously generated for human or mouse studies may equally work for zebrafish.
The intracellular bacteria S. flexneri has emerged as an exceptional model organism to address key issues in biology, including the ability of bacteria to be recognized by the immune system1,2. The host cell employs septins to restrict the motility of S. flexneri and target them to autophagy, a critical component of cell autonomous immunity7,8. These observations suggest a new molecular framework to study autophagy and its ability to degrade cytosolic bacteria. A major issue is now to fully decipher the underlying molecular and cellular events, and to validate these events analyzed in vitro during bacterial infection in vivo using relevant animal models. To this end, the zebrafish has been established as a valuable new host for the analysis of S. flexneri infection19. Interactions between bacteria and host cells can be imaged at high resolution, and the zebrafish model should prove useful for understanding the cell biology of Shigella infection in vivo. Zebrafish larvae can be used to investigate the role of bacterial autophagy in host defense, and work has shown that that the perturbation of autophagy can adversely affect host survival in response to Shigella infection19.
The observations generated from study of Shigella, septin caging and autophagy in vitro using tissue culture cells and in vivo using zebrafish larvae might provide fundamental advances in understanding host defense. They could also suggest the development of new strategies aimed at combating infectious diseases.
A critical aim of this report is to make sense of the molecular and cellular events analyzed in vitro (i.e., autophagy, actin tails, septin caging) during bacterial infection in vivo in the context of an entire organism, using zebrafish larvae. If not familiar with zebrafish biology and handling, one may refer to in depth protocols for proper zebrafish husbandry23 and in vivo analysis of zebrafish infection19,35.
The authors have nothing to disclose.
Work in the SM laboratory is supported by a Wellcome Trust Research Career Development Fellowship [WT097411MA].
Name of Material/ Equipment | Company | Catalog Number |
Bafilomycin A1 | Sigma-Aldrich | B1793 |
Blebbistatin | Sigma-Aldrich | B0560 |
Borosilicate glass microcapillars | Harvard Apparatus | 30038 |
Coarse manual manipulator | Narishige | M-152 |
Cytochalasin D | Sigma-Aldrich | C6762 |
4',6-diamidino-2-phenylindole (DAPI) | Molecular Probes | D1306 |
Dumont #5 fine tweezers | Fine Science Tools | 11254-20 |
Forchlorfeneuron | Sigma-Aldrich | 32974 |
Goat anti-mouse IgG (H+L) antibody | Molecular Probes | N/A |
Goat ant-rabbit IgG (H+L) antibody | Molecular Probes | N/A |
JetPEI transfection reagent | Polyplus transfection | 101-01N |
Latrunculin B | Sigma-Aldrich | L5288 |
LC3 antibody | Novus Biologicals | NB100-2220 |
Low melting agarose | Promega | V2111 |
MatTek glass bottom dish | MatTek corporation | P35G-1.0-14 |
MEM plus L-alanyl-L-glutamine | GIBCO | 41090028 |
MEM non-essential amino acids solution | GIBCO | 11140-035 |
Microinjector | Narishige | IM-300 |
Micropipette puller device | Sutter Instrument Co., Novato, | P-87 |
Mineral oil | Sigma-Aldrich | P35G-1.0-14 |
Monoclonal anti-tubulin, acetylated antibody | Sigma-Aldrich | T6793 |
Nocodazole | Sigma-Aldrich | M1404 |
N-phenylthiourea | Sigma-Aldrich | P7629 |
Paraformaldehyde | Sigma-Aldrich | 158127 |
Phalloidin | Molecular Probes | A12379 |
Phenol red solution | Sigma-Aldrich | P0290 |
Protease inhibitor cocktail | Roche | 4693116001 |
p62/SQSTM1 antibody | Cliniscience | PM045 |
Rapamycin | Sigma-Aldrich | R8781 |
Sodium pyruvate | GIBCO | 11360039 |
Transfection reagent | Life Technologies | 12252-011 |
Vectashield hard set mounting medium containing DAPI | Vector Laboratories | H-1500 |