|Home | About | Journals | Submit | Contact Us | Français|
Candida albicans is a dimorphic yeast that enters macrophages via the β-glucan receptor Dectin-1. Phagocytosis of C. albicans is characterized by actin polymerization, Syk kinase activation and rapid acquisition of phagolysosomal markers. In mice, C. albicans are able to resist the harsh environment of the phagosome and form pseudohyphae inside the phagolysosomal compartment, eventually extending from the macrophage. In this study, we investigated these unique C. albicans phagosomes and found that actin dynamically localized around the phagosomes, before disintegrating. Membrane phosphoinositides, PI(4,5)P2, PI(3,4,5)P3, PI(3,4)P2, and PI(3)P also localized to the phagosomes. This localization was not related to actin polymerization and inhibitor studies showed that polymerization of actin on the C. albicans phagosome was independent of PI3K. The ability of mature C. albicans phagosomes to stimulate actin polymerization could facilitate the escape of the growing yeast from the macrophage.
Candida albicans is an opportunistic pathogen that can cause infections of mucosal tissues and is able to invade systemically, especially in immunocompromised hosts. It possesses an array of virulence factors that regulate its adaptation to the host environment, recognition of host cells, secretion of hydrolases and phenotypic switching1. The polymorphism of C. albicans has also been suggested to contribute to pathogenesis; depending on environmental factors the yeast is able to switch between hyphal and yeast forms2.
Dendritic cells (DCs), neutrophils and macrophages (M) are able to phagocytose C. albicans via various receptors that have been shown to be involved in C. albicans recognition and regulation of subsequent immune responses 3–13. Like phagocytosis of other particles, C. albicans uptake is also accompanied by actin polymerization 14. After closure, the C. albicans phagosomes rapidly recruit markers of late endosomes and lysosomes 15. In contrast, Lamp-1 marks phagosomes 30 minutes after entry by FcR-mediated phagocytosis 16.
Mouse macrophages are not efficient in killing C. albicans. Despite the maturation of the C. albicans phagosomes, the yeasts are able to grow hyphal forms that eventually destroy the macrophage 15. Several reports have shown mechanisms used by C. albicans to escape phagocytosis and killing by macrophages. C. albicans phospholipomannan, a surface glycolipid that is shed by C. albicans, was shown to mediate escape from macrophages by inducing apoptosis 17,18. It has also been reported that a soluble factor from C. albicans suppresses nitric oxide (NO) production, but does not stimulate the production of immunosuppressive cytokines 19. Furthermore, C. albicans β-1,2-linked mannooligosaccharides, which are part of the C. albicans cell wall, have been implicated in adhesion to macrophages 20 and inhibition of TNFα production 21. Several pathogens interfere with host actin assembly, mainly by activating the Arp2/3 complex. This is used to promote uptake and to gain actin based motility to move through the cytoplasm into neighboring cells thereby spreading infection without activating immune responses 22. C. albicans has also been shown to secrete an actin rearranging factor that increases the transition of soluble actin to insoluble actin 23,24. Once inside the macrophage, C. albicans adapts to oxidative stress and starvation and induces morphological changes; at a later stage when hyphal growth enables escape from the macrophages, it activates glycolysis and downregulates stress responses 25.
Signal transitions that occur during phagosome maturation are not well characterized. Phosphoinositide (PI) are important signaling molecules in receptor-mediated signal transduction, actin remodeling and membrane trafficking 26,27. PIs, the collective name of phosphorylated derivatives of phosphatidylinositol, are membrane bound and compose less than 10% of the total cellular phospholipids. A total of seven different PIs can be produced by different combinations of phosphate groups arranged around the inositol ring. Organelle specific PI kinases and PI phosphatases mediate rapid subcellular distribution of specific PI leading to recruitment, binding and activation of effector proteins that mediate downstream signaling. Activation of phospholipid-modifying enzymes lead to the formation of phosphatidylinositol 3,4,5 trisphosphate [PI(3,4,5)P3] from phosphatidylinositol 4,5 bisphosphate [PI(4,5)P2] and phosphatidylinositol 3,4 bisphosphate [PI(3,4)P2] on the phagosomal membrane 28–32. PI(4,5)P2 and PI(3,4,5)P3 are involved in regulation of actin polymerization 28,33–38 and PI(3,4)P2 is thought to play a role in the activation of NADPH oxidase 39. Later stages of uptake associated with phagosome closure are the activation of Arf1 and Rac2 GTPases, formation of phosphatidylinositol 3 phosphate [PI(3)P], and the disappearance of polymerized actin from the phagosomal cup 16,30. PI(3)P is almost exclusively found on early endosomes and phagosomes 26,40 where it facilitates phagosome maturation 39,41–43.
Taken together these observations indicate that C. albicans is a complex pathogen able to utilize an array of signaling mechanisms to evade the host defense response. In this study we followed the dynamics of maturing C. albicans phagosomes in macrophages using real time microscopy. We assessed the localization of macrophage actin and found that phagosomes containing growing C. albicans are accompanied by large amounts of actin. We also assessed the localization of several phosphoinositides known to be involved in phagocytosis and found that they were increased on phagosomes containing live C. albicans. This is suggestive of extensive signaling processes being employed around the phagosomes.
Mice used in this study (BALB/c) were from the Sir William Dunn School of Pathology (University of Oxford) breeding colonies and between 8 and 12 weeks of age at the time of study. Animals were kept and handled in accordance with institutional guidelines. Femurs were dissected from sacrificed mice and marrow washed out of the femoral cavities. Cells were then pelleted at 500g for 5 minutes prior to resuspension in fresh RPMI containing 10% heat inactivated fetal bovine serum, 4mM L-glutamine 20U/ml penicillin, 20 μg/mL streptomycin and 20 % L-cell conditioned medium (source of M-CSF). Cells were then transferred to petri dishes and allowed to adhere. Fresh medium was added on day 3 and day 5 after isolation. Cells were plated for experiments at day 6 or 7 after isolation. Bone marrow derived M were used at 1×105 per well in 24 well plates.
RAW264.7, a murine macrophage-like cell line (RAW cells) from the American Type Culture, (Manassas, VA) were cultured in Advanced-DMEM with 2% heat inactivated (h.i.) Fetal Bovine Serum, 4 mM L-glutamine, 20 U/mL penicillin, and using Invitrogen cell culture reagents (Carlsbad, CA) at 37°C with 5% CO2. RAW cells were prepared for ratiometric microscopy by plating ~2.5×105 RAWs cells per coverslip the day before imaging. After the cells had attached to the coverslip (3 hours), they were transfected with plasmids encoding the fluorescent chimeras, using Roche FuGene-6 as described in the manufacturer’s protocol (Roche Diagnostics, Indianapolis, IN). This gave a transfection efficiency of 10–20%.
Plasmids encoding monomeric versions (A207K) of cyan fluorescent protein (CFP) and citrine (YFP) were used where indicated 44. β actin from pEYFP.actin.C1 (Clontech, Mountainview, CA) was transferred into the clontech C1 vector expressing either monomeric CFP or monomeric Cherry (mCherry) (Roger Tsien, University of California San Diego, San Diego, CA). The PH domains of human PLCδ1 and Bruton’s tyrosine kinase (Btk) were a gift from Tamas Balla 45,46. The PLCδ1PH construct was PCR amplified adding Xho1 and BamH1 restriction sites, then subcloned into pmCitrine-N1 (Clontech, Mountain View, CA). The BtkPH construct was PCR amplified and subcloned into the pmCitrine-N1 vector between Xho1 and HindIII. The AktPH construct, a gift from Tobias Meyer (Stanford, Palo Alto, CA), was subcloned into pmCitrine-C1 between BamH1 and Xba1 (Clontech, Mountain View, CA). Human Tapp1 constructs were obtained from the MRC Protein Phosphorylation Unit, Dundee, Scotland and the C-terminal PH domain was subcloned into pmCitrine-C1 at the EcoR1-BamH1 site. GFP-2xFYVE, the tandem FYVE finger domains from hepatocyte growth factor-regulated tyrosine kinase substrate (Hrs) was donated by Harold Stenmark (Norwegian Radium Hospital, Oslo, Norway) and cloned into pmCitrine.C1 between HindIII and Kpn1 sites (Clontech, Mountainview, CA). All DNA sequences were confirmed at the University of Michigan DNA Sequencing Core. pEGFP-Actin vector was from BD Biosciences (Oxford, UK).
Stock cultures of a laboratory strain of C. albicans (ATCC 18804) were maintained on Sabouraud’s dextrose agar (Difco Laboratories, Detroit, MI) at 4°C. For experiments, C. albicans was grown in 10 ml Sabouraud’s dextrose broth (Difco Laboratories, Detroit, MI) in a shaking incubator at 30°C for 24 hours, to obtain a stationary phase culture. The yeast cultures were washed in PBS and then added to cells at a ratio of 5:1. The cells were incubated for 30 minutes at 4°C to allow attachment, then washed and incubated for 2 hours at 37°C. Infection efficiency was around 30%. To inhibit PI3K activation, RAW cells were infected as described above, but after a 1.5 hour incubation, cells were washed and incubated with a 50 μM solution of LY294002 for 30 minutes at 37°C (EMD Biosciences, San Diego, CA), to avoid inhibition of phagocytosis. Sixty phagosomes were measured for the control group and 71 phagosomes for the LY-treated population. Other particles used in this study include zymosan (Invitrogen), 6.0 micron polystyrene polybead ® (Polysciences Inc., Warrington, Pennsylvania USA) and heat killed C. albicans (killed for 1 hour at 100°C). These particles were washed 3 times with medium before adding to the macrophages at a ratio of 5:1, as described above.
For microscopy, coverslips were placed in Leiden chambers (Harvard Apparatus, Holliston, MA) at 37°C in Ringer’s buffer (155 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1 mM MgCl2, 2 mMNaH2PO4, 10 mM glucose, 10 mM HEPES at pH 7.2). Transiently transfected macrophages containing internalized C. albicans were imaged.
M on coverslips were cooled to 4°C and washed three times with pre-chilled medium. Particles were added to the M at 5:1 ratio, incubated for 1 hour at 4°C to synchronise uptake. Unbound particles were removed by washing 3 times with pre-chilled medium. The cells were then incubated at 37°C for 2 hours. Cells were fixed with 4% paraformaldehyde (Sigma) in PBS for 20 minutes at 4°C. Cells were permeabilised in buffer containing 0.25% Saponin (Sigma), 1% BSA (Sigma), 1% heat inactivated Goat serum (Sigma) and 1% heat inactivated rabbit serum (Sigma) in PBS (permeabilisation buffer) at room temperature for 30 minutes. 1D4B rat anti-mouse LAMP1 antibody (eBioscience) and FITC-phalloidin (Sigma) were added at a concentration of 10 μg/ml in permeabilisation buffer for 1 hour at room temperature. Cells were washed three times prior to incubation with Alexa-488 goat anti rat antibody (Invitrogen) in the same buffer. After two washes with buffer and two washes with PBS, coverslips were mounted on glass microscopy slides with Dako Cytomation Fluorescent Mounting Medium (Dako Cytomation). Slides were analysed using a Radiance 2000 (Biorad) confocal microscope as described below. To quantificatify actin positive phagosomes, one hundred phagosomes were scored.
Images were acquired on an inverted fluorescence microscope (Nikon TE300) with a 60× 1.4 Planapo objective, a mercury arc lamp as the source of epifluorescent light and a cooled digital CCD camera (Quantix; Photometrics). The microscope was equipped with trans and epifluorescence shutters, a temperature-controlled stage, filter wheels for both excitation and emission filters, and dichroic mirrors that allow detection of multiple fluorophores. All images were acquired and processed using Metamorph 6.2r6 (Universal Imaging, Malvern, PA). Fluorescence excitation and emission wavelengths were selected via a JP4v2 filter set (Chroma Technology, Rockingham, VT) and Lambda 10-2 filter wheel controller (Sutter Instruments).
Phagosomes which contained growing C. albicans were imaged in RAW cells expressing fluorescent proteins. YFP (excitation 505 nm, emission 540 nm), CFP (excitation 435 nm, emission 490 nm) and phase-contrast images were recorded every thirty seconds. Macrophages expressing actin-mCherry (exciation 572 nm, emission 630 nm) were imaged using the Texas Red/FITC filter set (Omega Optical, Brattleboro, VT).
Using real time confocal microscopy, samples were prepared as described above and images collected on a Radiance 2000 (Biorad) confocal microscope, fitted with a four-line argon laser (457, 477, 488, 514 nm), a green HeNe laser (543 nm), a red diode laser (638 nm) and a TiS multiphoton laser (infra red). The microscope was an inverted Nikon TE300, with a 60X water immersion objective. The software controlling the camera was Bio-Rad Lasersharp 2000 (v. 5.0). Images were collected with the optimal iris aperture and the minimum laser power that fitted the whole grey scale. Macrophages and C. albicans remained viable troughout the course of these experiments.
To measure the amount of actin or PH domain chimera recruited to the phagosome, a region was drawn around the internalized C. albicans phagosome, and the average fluorescence intensity of the phagosome, IP, was divided by the average fluorescence intensity for the cell, IC. As a control, the average intensity of the nucleus, IN, was divided by the IC. Recruitment of the fluorescent chimera to the yeast phagosome was defined as IP/IC greater than 1.0. Images were further processed for figures using Adobe Photoshop version 6. Imaris software was used for 3 dimensional reconstruction.
The early events of C. albicans phagocytosis have been well examined. In contrast, there is presently little known about the later stages of C. albicans infection, when the yeast is fully internalized and forming pseudohyphae inside the macrophages. In the present study, we examined actin dynamics on mature C. albicans phagosomes (2+ hours after infection) in macrophages (M). During static experiments, cells were stained for polymerized actin with FITC-phalloidin two hours after C. albicans yeast phagocytosis, when the C. albicans had formed (pseudo-) hyphae. A number of (pseudo-) hyphae were marked by polymerized actin around their phagosomes. This could be seen in both RAW264.7 M and primary mouse M (Fig. 1A-B). Furthermore, C. albicans (pseudo)-hyphae breaking out of the macrophage also exhibited increased actin localization around the hyphae (Fig. 1C and Movie 1A-B). In general, more C. albicans (pseudo) hyphae were associated with actin compared to heat killed C. albicans phagosomes (data not shown). Large concentrations of actin could be seen around C. albicans (pseudo-) hyphae, when these (pseudo-) hyphae were growing from one M into adjacent M (Fig. 1D). A 3-dimensional reconstruction of a Z-series taken at this stage, shows that a solid tube of actin was formed along the (pseudo-) hyphae (Fig. 1D).
While synchronized experiments give reasonable similarity in the timing and dynamics of acquired proteins at short incubation periods, certain signaling molecules already show differences in kinetics as soon as 10 minutes after synchronized uptake (Henry et al. 2004). The dynamics of actin polymerisation were assessed by real time confocal microscopy. In these experiments RAW264.7 M expressing GFP-actin were allowed to phagocytose C. albicans yeast and two hours after phagocytosis, when C. albicans was growing (pseudo-) hyphae inside the M, the actin dynamics were followed. Figure 2 shows a sequence of images taken from a timecourse experiment of a M with three phagosomes, two with C. albicans (pseudo-) hyphae and one with yeast (See also Movie 2A-B). At the start of the experiment one (pseudo-) hyphal phagosome had a high concentration of surrounding actin while the other (pseudo-) hyphal phagosome did not show surrounding actin. In time the phagosome without actin started to acquire actin. Actin moved around the phagosome and sometimes left the phagosome completely, indicating that actin localization around phagosomes is an active process.
Since actin polymerization occurs during the formation of a phagocytic cup, there may be other signals associated with the early stages of phagocytosis that are also detectable along with actin on the mature C. albicans phagosomes. We examined the dynamics of PI(4,5)P2, PI(3,4,5)P3, PI(3,4)P2 and PI(3)P using YFP chimeras of binding domains specific for each phosphoinositide.
The PH domain of Plcδ1 binds to PI(4,5)P2 47. Plcδ1PH-YFP and actin-CFP were expressed in macrophages. The macrophages were infected with C. albicans and after two hours the fluorescent cells were examined for CFP and YFP chimera recruitment to the phagosomes. Detectable increases of Plcδ1PH-YFP and actin-CFP were observed on the C. albicans phagosomes (Fig. 3). Of the 8 phagosomes measured, 7 phagosomes exhibited Plcδ1PH-YFP recruitment to the phagosome, where the intensity in the phagosome IP, was greater than the intensity in the cell, IC (IP/IC > 1) (Fig. 6). Thus, PI(4,5)P2 was found on mature C. albicans phagosomes along with actin, suggesting it may play a role in this process.
To assess the localization of PI(3,4,5)P3 and PI(3,4)P2, macrophages were transfected to express AktPH-YFP together with actin-CFP to measure and compare the increase of 3′ phosphoinositides on the growing yeast phagosome to the increase in actin. The PH domain of Akt binds to both PI(3,4,5)P3 and PI(3,4)P2 48. C. albicans phagosomes that exhibited actin polymerization also exhibited a marked recruitment of AktPH-YFP (Fig. 4). The localization of the AktPH-YFP chimera to the phagosome was observed in all 8 phagosomes measured (IP/IC > 1) (Fig. 6). When the individual dynamics of PI(3,4,5)P3 and PI(3,4)P2 were measured using the BtkPH domain and the Tapp1PH domain, respectively, similar patterns were observed (data not shown). Therefore, PI(3,4,5)P3 and PI(3,4)P2 were present on C. albicans phagosomes that also contained actin.
The 2xFYVE domain from Hrs (FYVE) binds to PI(3)P 49. The dynamics of PI(3)P on the growing C. albicans phagosome were traced in macrophages expressing 2xFYVE-YFP and actin-CFP. YFP fluorescence increased on those phagosomes that also had actin polymerized around them, indicating the recruitment of 2xFYVE-YFP to the phagosome (Fig. 5). All 5 phagosomes measured exhibited 2XFYVE-YFP recruitment to the phagosome (IP/IC > 1) (Fig. 6). Thus, PI(3)P was also localized to mature C. albicans phagosomes, however its localization seemed to be unrelated to the dynamics of the actin polymerization.
During phagocytosis, PI3K-dependent signals facilitate actin polymerization 50. Similar signals could drive the actin polymerization on C. albicans phagosomes. To examine if the actin on C. albicans phagosomes was dependent upon PI3K, macrophages were treated with the PI3K-inhibitor LY294002. Since inhibition of PI3K blocks phagocytosis 51, the macrophages were only treated with LY294002 after being allowed to undergo initial phagocytosis of C. albicans. Thus, the actin polymerization during phagosome maturation would only be affected if dependent upon continued PI3K activation. In control and LY294002-treated macrophages transfected with actin-mCherry, the intensity of actin-mCherry on the C. albicans phagosome (IP) was measured, and divided by the overall mCherry intensity in the total cell (IC). If actin did polymerize on the phagosome, IP/IC would be greater than the baseline level of 1.0. In LY294002 treated cells there were comparable levels of actin polymerized on the yeast phagosomes (control = 1.18 +/− 0.38 n=66, LY trt = 1.16 +/− 0.42 n=71) (Fig. 7). This suggests that even though different 3′ phosphoinositides localized to the C. albicans phagosomes, these phosphoinositides are not necessary for actin polymerization around the C. albicans phagosomes.
This is the first study to examine the dynamics of C. albicans phagosomes 2 hours after initial infection. Our findings are summarized in Figure 8. Actin was polymerized on the phagosome and, through live cell imaging, could be observed cycling on and off the phagosome. Membrane phosphoinositides that are important in FcR-mediated phagocytosis were increased on the mature C. albicans phagosome (Fig. 8). Strikingly, inhibition of PI3K did not affect the actin polymerization observed on the yeast phagosomes. This implied that the yeast phagosome was triggering a PI3K-independent pathway leading to actin polymerization.
Actin polymerization around latex bead phagosomes has been studied extensively showing that polymerization has a role in fusion between phagosomes and late endocytic organelles 52–54. The increased number of actin-positive live C. albicans phagosomes compared to phagosomes containing heat killed C. albicans may be explained by the continuous growth of the (pseudo-) hyphae, which required higher levels of organelle fusion to accommodate the demand for membrane. Further analysis is needed to confirm this hypothesis. Several pathogens are known to be able to interfere with host actin assembly, mainly by activating the Arp2/3 complex. This complex is used to promote uptake and to gain actin-based motility to move through the cytoplasm into neighbouring cells, thereby spreading infection without activating humoral immune responses 22. C. albicans has also been shown to secrete an actin rearranging factor that increases the transition of soluble actin to insoluble actin 23,24. It is possible that the actin polymerisation on the yeast phagosome was mediated by C. albicans interference with the actin cytoskeleton. The presence of PI(4,5)P2 could also be facilitating the actin polymerization 55. However, in an experiment where C. albicans blastoconidia were phagocytosed followed by inhibition of actin polymerisation with cytochalasin D, C. albicans was still able to grow (pseudo-) hyphae and escape from the M (R.P. da Silva, unpublished data), suggesting that actin polymerization is not necessary for C. albicans survival inside M.
It is interesting that the mature phagosomes possess membrane markers associated with the earlier stages of phagocytosis. Even PI(3)P, which marks the endocytic pathway, appears on FcR-mediated phagosomes within the first 10 minutes of phagocytosis 16. The persistence of these early markers could indicate disregulation in the maturation process of the C. albicans phagosome, providing a means for C. albicans to escape the killing mechanisms of mature phagosomes. Mycobacterium tuberculosis employs a similar mechanism to avoid phagocytic killing in macrophages. By altering levels of calcium inside the macrophage, the bacterium persists indefinitely in an early endosome-like compartment 56,57. It is possible that C. albicans also secrete factors that interrupt phagosome maturation, giving it additional time to develop into filaments.
PI(3)P is known to be excluded from the plasma membrane and is seen in vesicles of the endocytic pathway 16,26. It participates in phagosomal maturation and NADPH oxidase activation 39,41–43. This PI was observed around the C. albicans phagosomes. PI(3,4)P2 was also present around the C. albicans phagosomes. The presence of PI(3)P and PI(3,4)P2, that also activate the oxidase, suggests that the NADPH oxidase is activated on C. albicans phagosomes 39. Upon phagocytosis the NADPH oxidase is assembled and activated leading to production of reactive oxygen intermediates that mediate killing and degradation of ingested particles. Further work is needed to assess if NADPH oxidase production takes place at these phagosomes.
Notably, not all C. albicans observed inside M were marked by polymerized actin. This may be due to the active process of polymerization occurring at specific points during phagosomal maturation. The live C. albicans phagosome is unique because of (pseudo-) hyphal growth, which forces the M to accomodate this growth in order to retain the C. albicans inside a phagosome. The mechanism of C. albicans escape from macrophages remains to be determined, but the findings described here indicate C. albicans evasion of normal phagosome maturation and highlights intriguing new directions for future exploration.
Movie 1. RAW264.7 macrophages were transfected with GFP-actin and labeled with lysotracker. Macrophages were allowed to phagocytose C. albicans yeast and unbound yeast was washed away. Cells were incubated at 37°C with 5% CO2 for one hour. Images were collected and the movie represents 60 minutes of real-time microscopy. A) Transmission images. B) Lysotracker in red and actin in green. Large amounts of actin can be seen around the C. albicans (pseudo)–hyphae that is pushing out of the macrophage.
Movie 2. RAW264.7 macrophages were transfected with GFP-actin and were challenged with C. albicans yeast. Unbound yeast was washed away and cells were incubated at 37°C with 5% CO2. After 2 hours image collection was started. The movie represents 125 minutes of real-time microscopy. A) Transmission images. B) GFP-actin shows the dynamic movement of actin around the C. albicans (pseudo)-hyphae.
This work was funded by NIH Grants AI35950, AI64668 to J.S., The Wellcome Trust, The Medical Research Council UK, EPA Cephalosporin Scholarship and Henry Goodger Scholarship. Gordon D. Brown is a Wellcome Trust International Senior Research Fellow in Biomedical Science in South Africa. Philip R. Taylor is a Wellcome Trust Research Career Development Fellow (grant 070579).