Journal of Medical Microbiology (2011), 60, 1570–1583
DOI 10.1099/jmm.0.032870-0
Administration of a nitric oxide donor inhibits mglA expression by intracellular Francisella tularensis and counteracts phagosomal escape and subversion of TNF-a secretion Linda Tancred,3 Maxim V. Telepnev,34 Igor Golovliov, Blanka Andersson,1 Henrik Andersson,1 Helena Lindgren and Anders Sjo¨stedt Correspondence Anders Sjo¨stedt
Department of Clinical Microbiology, Clinical Bacteriology, and Laboratory for Molecular Infection Medicine Sweden (MIMS), Umea˚ University, SE-901 85 Umea˚, Sweden
[email protected]
Received 31 March 2011 Accepted 17 June 2011
Francisella tularensis is a highly virulent intracellular bacterium capable of rapid multiplication in phagocytic cells. Previous studies have revealed that activation of F. tularensis-infected macrophages leads to control of infection and reactive nitrogen and oxygen species make important contributions to the bacterial killing. We investigated the effects of adding S-nitrosoacetyl-penicillamine (SNAP), which generates nitric oxide, or 3-morpholinosydnonimine hydrochloride, which indirectly leads to formation of peroxynitrite, to J774 murine macrophage-like cell cultures infected with F. tularensis LVS. Addition of SNAP led to significantly increased colocalization between LAMP-1 and bacteria, indicating containment of F. tularensis in the phagosome within 2 h, although no killing occurred within 4 h. A specific inhibitory effect on bacterial transcription was observed since the gene encoding the global regulator MglA was inhibited 50–100-fold. F. tularensis-infected J774 cells were incapable of secreting TNF-a in response to Escherichia coli LPS but addition of SNAP almost completely reversed the suppression. Similarly, infection with an MglA mutant did not inhibit LPS-induced TNF-a secretion of J774 cells. Strong staining of nitrotyrosine was observed in SNAP-treated bacteria, and MS identified nitration of two ribosomal 50S proteins, a CBS domain pair protein and bacterioferritin. The results demonstrated that addition of SNAP initially did not affect the viability of intracellular F. tularensis LVS but led to containment of the bacteria in the phagosome. Moreover, the treatment resulted in modification by nitration of several F. tularensis proteins.
INTRODUCTION Throughout history, many microbes have developed mechanisms to subvert the normally hostile environment encountered after phagocytosis by eukaryotic cells and thereby be able to grow in this intracellular habitat. For example, several species and subspecies of intracellular bacteria belonging to the genus Francisella have developed 3These authors contributed equally to this work. 4Present address: Department of Microbiology and Immunology, University of Texas Medical Branch, Galveston, TX 77555-1070, USA. 1Present address: Department of Clinical and Experimental Medicine, Medical Microbiology, Linko¨ping University, SE-581 85 Linko¨ping, Sweden. Abbreviations: DFO, deferoxamine; FeTPPS, 5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrinato iron (III); LDH, lactate dehydrogenase; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide; NO, nitric oxide; ONOO2, peroxynitrite; RLU, relative light units; RNS, reactive nitrogen species; SIN-1, 3-morpholinosydnonimine hydrochloride; SNAP, S-nitroso-acetyl-penicillamine.
1570
mechanisms that allow them to replicate unrestrictedly in the host cell cytoplasm, thereby becoming important human pathogens (Oyston, 2008). Since Francisella tularensis has been classified by the Centers for Disease Control and Prevention as a category A agent, it has become the focus of an intensive research effort. Despite this effort and the availability of extensive genomic information, we still have a rudimentary understanding of why it demonstrates such high virulence and is extremely contagious. For example, many intracellular bacteria depend on type III or type IV secretion systems in order to invade and survive within eukaryotic host cells but no such systems exist in F. tularensis, and the mechanisms behind its intracellular survival are not well understood (Larsson et al., 2005; Sjo¨stedt, 2006). After uptake into the host cell, the Francisella-containing phagosome evades fusion with the lysosome and bacteria rather quickly escape into the cytoplasm (Checroun et al., 2006; Clemens et al., 2004; Golovliov et al., 2003b; Santic et al., 2005). Several genes necessary for intramacrophage survival, growth
Downloaded from www.microbiologyresearch.org by 032870 G 2011 SGM IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Printed in Great Britain
Nitric oxide inhibits mglA expression by F. tularensis
within the amoeba Acanthamoeba castellanii, a putative natural reservoir of F. tularensis, and the phagosomal escape have been identified and most of them belong to the so-called Francisella pathogenicity island (FPI), which is a 34 kb genomic island (Lauriano et al., 2004; Nano & Schmerk, 2007). In particular, the roles of the four igl genes have been characterized in this regard (Bro¨ms et al., 2010). In contrast to the iglC and iglD genes, which appear to be unique to F. tularensis, homologues of iglA and iglB exist in many bacterial species, most of which are either pathogenic to animals or plants, or are plant symbionts, and the encoded proteins are thought to be involved in type VI protein secretion (Bro¨ms et al., 2009). Almost all of the proteins encoded by the FPI are conserved among the F. tularensis subspecies. The regulation of the FPI gene expression is highly complex, involving at least six regulators, MglA, SspA, FevR, MigR, Hfq and PmrA (Dai et al., 2011). Most information is available regarding MglA, SspA, PmrA and FevR, which form a complex that binds to the RNA polymerase, indicating a direct role in transcriptional regulation (Charity et al., 2007).
upregulated since it and its by-products are highly reactive and thereby potent antimicrobial agents but also have potentially deleterious effects on the host if not properly regulated (Valko et al., 2007). For example, NO can react with the superoxide anion to form the even more toxic oxidant peroxynitrite (ONOO2), which is a critical component for the development of severe pathophysiological phenomena such as shock and inflammation (Vira´g et al., 2003). At the single cell level, ONOO2 exerts potent apoptotic or necrotic effects via lipid peroxidation, inhibition of mitochondrial respiration, DNA strand breakage and depletion of intracellular energy stores (Ferrer-Sueta & Radi, 2009). Many of the outcomes are directly or indirectly related to effects on its primary targets: inactivation of metalloenzymes (Liochev & Fridovich, 1999). Moreover, it induces specific amino acid modifications that, depending on local pH and the microenvironment, result in oxidation or nitration (Liochev & Fridovich, 1999). Tyrosine nitration has been associated with several inflammatory and neurodegenerative diseases (Rubbo & Radi, 2008). Thus, ONOO2 plays a key role in mediating and regulating key physiological phenomena at both the systemic and single cell level, and also has a central role in pathological processes that lead to chronic disease, but, by affecting prokaryotic targets similar to the eukaryotic ones, is also a potent antimicrobial compound (Bogdan, 2001). Thus, NO production and ONOO2 formation need to be tightly regulated as they will potently affect both host and pathogen targets during infection.
After ingestion, microbes normally become localized to the phagolysosomal pathway leading to the exposure of a wide array of cidal mechanisms. The potency of the mechanisms is much enhanced through the exposure of the cells to various types of cytokines such as IFN-c and TNF-a (Elkins et al., 2007). Such classically activated macrophages demonstrate an intense microbicidal activity (Gordon, 2003). However, F. tularensis, like several other highly successful pathogens, possesses mechanisms to create an intracellular environment that favours its replication through subversion of the otherwise highly effective microbicidal response of phagocytic cells, and may serve as a prototype for stealth pathogens (Sjo¨stedt, 2006). Besides the evasion of phagolysosomal fusion, escape into the cytosol (Checroun et al., 2006; Clemens et al., 2004; Golovliov et al., 2003b; Santic et al., 2005), suppression of the inflammatory response upon escape into the cytosol (Rajaram et al., 2006; Telepnev et al., 2005) and abrogation of T-cell responses (Woolard et al., 2007), studies have identified additional intricate subversive mechanisms, such as inhibition of the IFN-c-induced STAT1 and iNOS expression in human and murine monocytic cells (Parsa et al., 2008; Roth et al., 2009). The concomitant reduced nitric oxide (NO) production could be directly linked to increased bacterial survival (Parsa et al., 2008). Moreover, it was demonstrated that after intraperitoneal administration of F. tularensis bacteria, the classical activation of macrophages is subverted and instead the infected cells demonstrate an alternative mode of activation driven by IL-4 and IL-13 that favours bacterial survival (Shirey et al., 2008).
Our previous studies of the roles of reactive nitrogen and oxygen species in the killing of F. tularensis revealed a complex interaction between the two species, and we concluded from studies of extra- and intracellular bacteria that NO was required but not sufficient for bacterial killing and most likely ONOO2 was the predominant effector molecule (Lindgren et al., 2005). However, using slightly different types of monocytic cells, Edwards et al. (2010) concluded that IFN-c-mediated control of F. tularensis infection was due to cytosolic mechanisms independent of reactive oxygen or nitrogen species. Herein, we studied how addition of reactive nitrogen species (RNS) to cell cultures affected localization and survival of intracellular F. tularensis and also whether RNS influenced the bacterium’s subversive effect on TNF-a secretion.
The decreased iNOS expression during F. tularensis infection may have broad implications for the host–pathogen interaction during tularaemia since NO maintains normal body homeostasis at physiological concentrations. Moreover, as part of the innate immune response, NO is
To generate killed bacterial suspensions, F. tularensis LVS was grown overnight on modified Thayer-Martin agar at 37 uC, suspended in PBS and formalin was added to a final concentration of 4 % followed by an incubation at 37 uC for 45 min. The suspension was centrifuged and the pellet containing killed bacteria was washed three times in
http://jmm.sgmjournals.org
METHODS Bacteria and mammalian cells. The live vaccine strain F. tularensis
LVS was supplied by the US Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, MD, USA. The DiglB, DiglC, DiglD and DmglA mutants and corresponding GFP-expressing bacteria have been described previously (Bro¨ms et al., 2009; Bo¨nquist et al., 2008; Golovliov et al., 2003a).
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1571
L. Tancred and others PBS and finally resuspended in PBS. Escherichia coli LPS was obtained from Sigma. The murine macrophage-like cell line J774A.1 was grown in Dulbecco’s modified Eagle’s medium with 10 % (v/v) FCS (Gibco, Invitrogen). Construction of luxAB fusions in F. tularensis LVS. To monitor
F. tularensis gene expression in vivo, the luciferase genes luxAB from Vibrio harveyi were inserted downstream of the target gene. The mglA, polA (FTL_1666) and iglC genes were amplified by PCR using primers that also contained restriction sites that enabled cloning into the suicide vector pCH257 upstream of the promoterless luxAB genes (Forsberg et al., 1994). This resulted in the transcriptional fusion of the respective target genes with the luxAB genes. An E. coli strain, S-17, carrying the respective plasmid was transferred into F. tularensis LVS by conjugation and clones were analysed by PCR to confirm correct integration. The phenotypes of the mutants were compared to that of LVS by assessing their growth in J774 cells and in C57BL/6 mice; no differences were found (data not shown). In vitro infection and determination of bacterial numbers in J774A.1 cells. One day before the start of an experiment, J774 cells
were detached, resuspended in culture medium, and seeded in 24-well plates at a density of 26105 cells per well, in 6-well plates at a density of 16106 cells per well, or in 96-well plates at a density of 66104 cells per well. After incubation overnight at 37 uC, wells were washed and reconstituted with fresh culture medium. To each well, a suspension of bacteria was added (m.o.i.5500) and bacterial uptake was allowed to occur for a 2 h period at 37 uC. The monolayer was washed twice and incubated in culture medium containing 5 mg gentamicin ml21. The uptake of F. tularensis bacteria is rather inefficient so this high m.o.i. was used to ensure that each cell was infected with at least one bacterium (Andersson et al., 2006). To some of the wells, 200 mM 5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrinato iron (III) (FeTPPS; Calbiochem, Merck), 1.0 mM S-nitroso-N-acetyl-penicillamine (SNAP) (Molecular Probes/Invitrogen) or 0.8 mM 3-morpholinosydnonimine hydrochloride (SIN-1) (Molecular Probes) was added. One millimolar SNAP generates 14 mM NO min21 and 1 mM SIN-1 generates 10 mM ONOO2 min21 (Feelisch & Kelm, 1991). The ferric chelator deferoxamine (DFO) was added at a concentration of 400 mM 30 min before addition of bacteria. The compounds were also added to the wells after subsequent washing steps and the cells were incubated for indicated periods. For determination of numbers of intracellular bacteria, J774 cells in 24-well plates were washed once, lysed with 0.1 % sodium deoxycholate, and 100 ml portions of each lysate, serially diluted in PBS, were plated on modified Thayer-Martin agar for determination of viable counts.
containing only DMEM+FBS w/o phenol red and DMEM+FBS w/o phenol red+specific treatment: SNAP, FeTPPS, SNAP+FeTPPS or SIN-1. After 2 h of incubation at 37 uC, 850 ml of the cell culture medium with MTT solution was removed from each well and 550 ml DMSO was added to dissolve the formazan deposit. After 10 min of incubation at 37 uC, the A540 was read on a Tecan Sunrise plate reader. For comparison, wells for LDH analysis were prepared in parallel. Cytotoxicity according to mitochondrial function was calculated with the following formula: cytoxicity5100–1006(A540 treated J774/A540 untreated J774). It should be noted that the absorbance of the untreated J774 cells continuously increased due to their proliferation, resulting in a doubling of the absorbance over 24 h, whereas treated cells did not proliferate. Nitrite production assay. The amount of nitrite in the supernatants from J774 cell cultures of the bacterial viability assay was measured by transferring 50 ml of cell supernatants to a 96-well plate, adding 100 ml Griess reagent and incubating for 10 min. Thereafter, the A540 was measured with the Tecan Sunrise plate reader. Absorbance values were related to a standard curve. The lower limit of detection in the assay was 2.5 mM. Levels present in cell medium alone were subtracted from the values of experimental cultures. The assay was performed in triplicate and repeated three times. Immunofluorescence assay. Cells (26105) of the J774 cell line in 100 ml DMEM+10 % FBS were seeded onto glass coverslips in 24-
well plates. The cells were infected the following day with GFPexpressing F. tularensis LVS at an m.o.i. of 500 or green-fluorescent latex beads (Sigma) at an m.o.i. of 10 for 2 h, thereafter washed and further incubated for 0, 2, 5 or 24 h. To some of the wells, SNAP and/ or FeTPPS was added together with the bacteria and maintained throughout the infection. Chloramphenicol was added to some wells at a concentration of 20 mg ml21. Also, both J774 cells and bacteria were preincubated with the antibiotic 20 min prior to infection. Immediately after infection, cells were washed and fixed in 3 % paraformaldehyde for 15 min, permeabilized with 0.15 % saponin in PBS for 10 min, blocked in 2 % BSA/0.15 % saponin in PBS for 10 min, and further incubated with primary antibody against the LAMP-1 glycoprotein (Iowa Hybridoma Bank), diluted 1 : 700 in blocking buffer and thereafter an anti-rat IgG (Molecular Probes) conjugated to Alexa 594 at a dilution of 1 : 1000 was added. Glass slides were mounted in ProLong gold mounting medium (Molecular Probes) and subsequently analysed by use of a Nikon C1 confocal microscope (Nikon Instruments Europe) for determination of the degree of colocalization of GFP-expressing F. tularensis and LAMP-1. A LAMP-1-stained phagosomal membrane around a bacterium, resulting in an intact red ring around the green fluorescent LVS, was the criterion for colocalization. In total, 100 bacteria on each coverslip were scored, each group was analysed in triplicate and experiments were repeated three times.
Lactate dehydrogenase (LDH) assay of cytotoxicity. Culture supernatants (50 ml) from the bacterial viability assay were transferred to a 96-well plate and mixed with 50 ml of a substrate mix prepared
Assay of luciferase activity. J774 cells, 66104 per well in 96-well
according to the manufacturer’s instructions (LDH assay kit; Promega). The plates were incubated in the dark, at room temperature, for 30 min, and then 50 ml of a stop solution was added. The A492 was monitored with a Tecan Sunrise plate reader (Tecan Systems) 1 h later. Uninfected J774 cells lysed in 0.1 % deoxycholate served as a positive control and this value was arbitrarily designated as representing 100 % cell lysis. Sample absorbance was expressed as a percentage of the positive control. The assay was performed with triplicate samples and repeated at least three times.
plates, were infected with LVS at an m.o.i. of 500 for 2 h. To some of the wells, SNAP was added together with the medium. After bacterial uptake, the monolayer was washed twice and incubated in cell culture medium containing 5 mg gentamicin ml21, with or without the addition of SNAP. At given time points, 50 ml n-decyl aldehyde (Sigma) diluted 1 : 10 000 in water was added to the wells and the luciferase activity was immediately measured using a PhL luminometer (Aureon Biosystems). The luciferase activity was determined as relative light units (RLU; 1 RLU51000 counts s21).
MTT assay of mitochondrial activity. MTT [3-(4,5-dimethylthia-
Quantitative real-time PCR. The total RNA was isolated from 16106 J744 cells per well that had been infected at an m.o.i. of 500 with F. tularensis LVS and treated with SNAP and/or FeTPPS, using the guanidinium isothiocyanate/phenol-based monophasic extraction solution, TRIzol (Gibco), according to the recommendations of the
zol-2-yl)-2,5-diphenyl tetrazolium bromide] (Sigma) is reduced to insoluble purple formazan by the mitochondrial activity of viable cells. One hundred microlitres of MTT in PBS (5 mg ml21) was added to J774 cell cultures in 24-well plates and to control wells 1572
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis manufacturer. DNA contaminations were removed with DNA-free DNase Treatment and Removal Reagent (Ambion/Invitrogen), according to the supplied protocol. Reverse transcription (RT) of DNA-free RNA was performed in a total volume of 20 ml, containing 5 mg RNA, using a SuperScript II Reverse Transcriptase kit (Invitrogen), as suggested in the protocol. Specific primers for selected genes were designed using the Primer3 program available at http://frodo.wi.mit.edu/primer3/. Quantitative real-time PCR was performed using a SYBR Green I PCR kit (Applied Biosystems) in an ABI Prism 7900HT Sequence Detection System (Applied Biosystems). For amplification by PCR, specific primers for iglC, mglA and tul4 were synthesized by Eurofins MWG Operon (Ebersberg) and the sequences were as follows: F-tul4, GTGCCATGATACAAGCTTCC; Rtul4, GCTGTCCACTTACCGCTTCA; F-mglA, TTGCAGTGTATAGGCTTAGTGTGA; R-mglA, ATATTCTTGCATTAGCTCGCTGT; F-iglC, TTTCATATCTGTAGCACTTGCTTG; R-iglC, CCAGGCTCTATAAATCCAACAATA. Each reaction contained 12.5 ml SYBR Green I PCR kit, 250 nM each of reverse and forward primer and 1 ml cDNA. The total volume of the reaction was adjusted with water to 25 ml. PCR was performed in MicroAmp 96-well plates (Applied Biosystems) capped with MicroAmp optical adhesive seal. The reactions were incubated at 50 uC for 2 min, 10 min at 95 uC, followed by 45 cycles of 15 s at 95 uC, and 1 min at 60 uC. The PCRs were subjected to a heat dissociation protocol present in the ABI SDS 2.0 software (Applied Biosystems). We observed that expression of the housekeeping gene tul4 (Lauriano et al., 2004) was markedly downregulated by the SNAP treatment, therefore Ct values were normalized to the expression of the host cell housekeeping gene bactin for each treatment group and time point since b-actin expression was not affected by any of the treatments in our experiments (data not shown). Importantly, in numerous experiments, intracellular bacterial c.f.u. were identical (P.0.2) between the treatment groups at the time for RNA isolation. The DDCt method was employed to calculate the relative levels of Francisella gene expression (Schmittgen & Livak, 2008). Statistical comparisons were performed on normalized Ct values. Assay of TNF-a secreted in supernatants. J774 cells, 26105 per
well in 24-well plates, were infected with LVS, mglA or igl mutants, or heat-killed LVS at an m.o.i. of 500 for 2 h. After washing, the cells were further incubated in cell culture medium in the presence or absence of 5 mg E. coli LPS ml21, and with or without the addition of SNAP. The supernatants were analysed for the presence of TNF-a by utilization of the OptEIA Set (BD Biosciences Pharmingen). All procedures for the ELISA followed the instructions of the manufacturers. The lower level of detection was 15 pg ml21.
peptide mass fingerprinting and partial peptide sequencing (MALDITOF/TOF) (Alphalyse A/S). To confirm the nitration of proteins in these bands, a search for nitro-modifications of tryptophan and tyrosine was also included in the MS analysis. The identified proteins had Mascot scores between 94 and 464, which was within the 95 % confidence level of the analysis. Statistical analysis. All values were expressed as means±SEM. One-
way analysis of variance (ANOVA), followed by post hoc testing (Bonferroni) and two-tailed Student’s t-test, were used to identify differences between groups. All statistical analyses were carried out using the SPSS software, version 18.0.
RESULTS Effect of SNAP and SIN-1 on the intracellular survival and localization of LVS Our previous findings showed that the bactericidal effect of IFN-c-activated macrophages was critically dependent on NO (Lindgren et al., 2005). It is a highly reactive molecule that quickly forms ONOO2 in the presence of superoxide, and results from our previous study also indicated that ONOO2 was required for the IFN-c-activated macrophages to eradicate LVS (Lindgren et al., 2005). However, the exact roles of NO and ONOO2 in the killing are still unclear (Edwards et al., 2010). To this end, we investigated whether the addition of the NO donor SNAP or the ONOO2 donor SIN-1 affected the intracellular survival of F. tularensis during a 4 h period. None of the treatments affected the viability of the intracellular bacteria during this time (Table 1; data not shown for SIN-1). These results are in contrast to the effects of SIN-1 on extracellular bacteria, which were killed within at most 2 h (Lindgren et al., 2005). Thus, the intracellular bacteria are not rapidly killed simply by the addition of potent microbicidal compounds to the cell cultures as they are when exposed in PBS. We further investigated whether any effects of SNAP were observed over time and followed intracellular growth of LVS up to 24 h after infection, a time point when LVS in
Immunoblot analysis. LVS was exposed to ONOO2 in vitro and
assessed for protein nitration. Bacteria were grown overnight on MC plates and then incubated at OD51 in PBS with or without addition of SIN-1 (0.8 mM) at 37 uC for 2 h. The bacteria were then pelleted, lysed in sterile water and Laemmli sample buffer and thereafter subjected to electrophoresis on a 20 cm 15 % SDS-PAGE gel. Each lane was loaded with lysate prepared from 26109 bacteria. Proteins were transferred onto a nitrocellulose membrane using a semi-dry blotter (EBU 4000; C.B.S. Scientific). Membranes were then incubated for 1 h at room temperature with primary anti-nitrotyrosine rabbit IgG diluted 1 : 10 000 (Sigma). After washing, a secondary horseradish peroxidase-conjugated anti-rabbit antibody was applied at 1 : 2000 for 1 h, room temperature (Amersham/GE Healthcare). After addition of ECL reagent (Amersham), nitrated proteins were visualized on photographic film.
Table 1. Effect of SNAP treatment on viability of F. tularensis LVS in J774 cells F. tularensis was allowed to infect J774 cells at an m.o.i. of 500 for 2 h with or without the addition of 1.0 mM SNAP and incubated for the indicated periods. Data represent the mean log10 c.f.u.±SD of three cultures from one representative experiment of three performed. There were no significant differences in bacterial numbers of the SNAP-treated group compared to the untreated LVS-infected group, within each time point (Student’s t-test). Treatment
F. tularensis LVS (log10 c.f.u.) 0h
1h
2h
4h
4.9±0.2 4.8±0.2
5.0±0.1 5.0±0.2
5.2±0.2 5.0±0.1
5.1±0.2 5.0±0.2
MS analysis of nitrated proteins. To identify the nitrated proteins
from the Western immunoblot analysis, corresponding Coomassiestained SDS-PAGE gel protein bands were analysed by matrix-assisted laser desorption/ionization time of flight (MALDI-TOF) MS using http://jmm.sgmjournals.org
No SNAP 1.0 mM
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1573
L. Tancred and others
untreated J774 cells reaches maximal numbers. After 24 h, LVS numbers in untreated macrophages increased by 1.9 log10, while, in contrast, the SNAP-treated group declined by 1.5 log10 (Table 2). The combination of SNAP and the ONOO2 decomposition catalyst FeTPPS completely restored the growth defect caused by SNAP alone (Table 2), indicating that ONOO2 serves as an effector of the SNAP-mediated killing of intracellular F. tularensis. FeTPPS inhibits the biological effects of ONOO2 by catalytically decomposing it to nitrate, and the concentration of FeTPPS used, 200 mM, has been found to effectively decompose even the highest concentrations of ONOO2 in vitro (Misko et al., 1998). As ONOO2 generated from SNAP was likely to mediate the intracellular bacterial growth inhibition, we determined the levels of nitrite in culture supernatants, since nitrite is a stable end metabolite of ONOO2 and NO. Nitrite was detected at all time points between 0 and 24 h and ranged from 200 to 400 mM (Table 3). In comparison, this was markedly higher than levels resulting after IFN-c stimulation of peritoneal exudate cells, ,100 mM (Lindgren et al., 2005). After 24 h of infection, the integrity and monolayer of the J774 cells was intact, although there were some morphological signs of cytopathogenicity, whereas detachment of cells and cellular destruction was much evident at later time points (data not shown). For quantification, the cytopathogenic effects of the treatments on LVS-infected J774 cells were monitored by an LDH release assay. Between 0 and 9 h, infected, untreated cells displayed low levels of toxicity, at most 10 %, while at 24 h, the host cells were severely affected by the infection and toxicity levels reached 100 % (Table 4). For the FeTPPS group, LDH
Table 3. Effect of SNAP and/or FeTPPS on nitrite levels in J774 cells infected with F. tularensis LVS F. tularensis was allowed to infect J774 cells at an m.o.i. of 500 for 2 h. After addition of the respective treatment, SNAP (1.0 mM) and/or FeTPPS (200 mM), the cells were incubated for the indicated periods. Data represent mean nitrite levels (mM)±SEM of three cultures from one representative experiment of two performed. The detection limit was 2.5 mM. Without addition of SNAP, nitrite could not be detected in the supernatants. Group
Nitrite levels (mM)
LVS LVS/SNAP LVS/FeTPPS LVS/SNAP/FeTPPS
0h
7h
,2.5 166±3 ,2.5 381±2*
,2.5 226±3 ,2.5 377±2*
9h
24 h
,2.5 ,2.5 243±3 249±4 ,2.5 ,2.5 395±1* 387±1*
*Significantly different (P,0.001; Student’s t-test) from the level observed in infected cultures treated with SNAP only. Comparisons between SNAP-treated groups only.
levels were similar to those for the untreated group, but were slightly lower at 24 h, 73 %. The SNAP and SNAP+FeTPPS groups showed similar LDH release up to 7 h, but thereafter cells treated with the combination showed higher LDH release than cells treated with SNAP alone (Table 4), likely due to rapid replication of LVS in
Table 4. Effect of SNAP and FeTPPS treatment on cytopathogenicity of J774 cells during F. tularensis LVS infection Table 2. Effect of SNAP and/or FeTPPS treatment on viability of F. tularensis LVS in J774 cells F. tularensis was allowed to infect J774 cells at an m.o.i. of 500 for 2 h, and after addition of the respective treatment, SNAP (1.0 mM) or FeTPPS (200 mM), cells were incubated for the indicated periods. Data represent mean log10 c.f.u.±SEM of three cultures from one representative experiment of at least two performed. Significant differences between the groups were determined by one-way ANOVA within each time point. Group
LVS LVS/SNAP LVS/FeTPPS LVS/SNAP/FeTPPS
Group
% LDH release 0h
F. tularensis LVS (log10 c.f.u.) 0h
7h
9h
24 h
5.2±0.2 5.3±0.1 4.5±0.2D 5.3±0.1
5.9±0.1 4.4±0.1* 5.6±0.2 5.1±0.2D
6.4±0.0 4.9±0.0* 6.4±0.1 5.6±0.1*
7.1±0.2 3.8±0.1* 7.0±0.1 7.0±0.0
*Significantly different bacterial numbers ANOVA) compared to untreated LVS-infected DSignificantly different bacterial numbers ANOVA) compared to untreated LVS-infected 1574
F. tularensis was allowed to infect J774 cells at an m.o.i. of 500 for 2 h, and after addition of the respective treatment, SNAP (1.0 mM) or FeTPPS (200 mM), cells were incubated for the indicated periods. Data represent the mean LDH release±SEM of three cultures from one representative experiment of at least two performed. Significant differences between the groups were determined by one-way ANOVA within each time point.
(P,0.001; one-way cultures. (P,0.01; one-way cultures.
LVS LVS/SNAP LVS/FeTPPS LVS/SNAP/FeTPPS
7h
9h
24 h
5.0±0.7 9.5±0.9 10.5±0.5 102.9±3.5 6.9±0.7 19.6±1.3* 26.4±2.1D 48.0±1.4D 5.4±0.2 10.5±1.1 18.9±1.6d 73.2±0.9D 5.9±0.2 25.8±3.3D 40.1±0.5D 74.5±1.3D
*Significantly different cytopathogenicity level (P,0.05; one-way ANOVA) compared to that of untreated LVS-infected cultures. DSignificantly different cytopathogenicity level (P,0.001; one-way ANOVA) compared to that of untreated LVS-infected cultures. dSignificantly different cytopathogenicity level (P,0.01; one-way ANOVA) compared to that of untreated LVS-infected cultures.
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis
the combination group (Table 2). Since LVS multiplied as well in the presence of SNAP+FeTPPS as in untreated cells, the effect of SNAP on the intracellular bacteria after 24 h appeared to be bactericidal and not indirect, due to compromised host cell viability. To corroborate the LDH results of infected J774 cells, an MTT assay assessing toxic effects on mitochondrial function was also performed. Since the MTT substrate can be cleaved by both eukaryotic and prokaryotic dehydrogenases, only the effects on uninfected J774 cells were assessed. In parallel, LDH was also measured for uninfected cells. We found that the two assays gave very similar results for the 2 and 24 h samples, but the MTT assay showed somewhat higher values for the 5 and 7 h time points (Fig. 1a, b). However, mitochondrial activity did not differ significantly between the treated groups at any time point (P.0.1, data not shown). Despite the lack of early cidal effects of the RNS, the addition of SNAP could still modulate the intracellular habitat and thereby affect the bacteria. Previously, we and others have demonstrated that F. tularensis rapidly escapes from the phagosome of human and mouse monocytic cells (Chong et al., 2008; Clemens et al., 2004; Golovliov et al., 2003b; Wehrly et al., 2009). F. tularensis LVS reaches the cytosol of THP-1 and J774 cells within at most 2 h after infection (Golovliov et al., 2003b; Lindgren et al., 2004) and the LAMP-1 colocalization decreases concomitantly (Bo¨nquist et al., 2008; Golovliov et al., 2003b). To examine whether SNAP addition affected the phagosomal escape, we followed colocalization of bacteria with LAMP-1, which is a marker for late endosomes/lysosmes. Without treatment, at 0 and 2 h post-infection, 21 % and 8 % of bacteria, respectively, had colocalized with LAMP-1, therefore, presumably a minority of the F. tularensis bacteria were localized inside the phagosome. In cultures treated with SNAP, significantly higher (P,0.05) numbers
of bacteria, 38 % at 0 h and 47 % at 2 h, colocalized with LAMP-1-positive phagosomes (Figs 2 and 3a). We also asked whether addition of the ONOO2 decomposition catalyst FeTPPS influenced the observed effects of SNAP. Treatment with SNAP and FeTPPS led to a partial reversal (P,0.05) of the effect of SNAP since the colocalization was 28 % at 1 and 2 h but also differed significantly from that of untreated cultures (P,0.01) (Fig. 2). From 5 h and onwards, FeTPPS completely reversed (P.0.05 vs untreated LVS) the increased colocalization of bacteria and LAMP-1 conferred by SNAP (Fig. 2). Of note, LVS multiplied unrestrictedly after SNAP+FeTPPS treatment as evidenced by the confocal micrographs (Fig. 3b) and growth curves (Table 2). Thus, the results indicated that ONOO2 played a very important, but not exclusive, role in containing LVS in the phagosome. Next, we asked whether the effect of SNAP mimicked the effect of a general inhibitor of bacterial protein synthesis. To this end, we added a high concentration, 20 mg ml21, of chloramphenicol to the cell cultures. As expected, it conferred a static effect and no bacterial replication was observed over 24 h (data not shown) but the bacterial colocalization with LAMP-1 was not affected (Fig. 2). Thus, bacterial protein synthesis was not necessary for the phagosomal escape. Together, the results showed that SNAP did not significantly affect the viability of intracellular F. tularensis during the first 4 h of infection but conferred significant growth inhibition at 24 h. Despite the lack of effect on the number of intracellular bacteria during the first 4 h, addition of SNAP rapidly led to partial inhibition of the rapid escape of F. tularensis from the phagosome, which normally occurs within 2 h. Moreover, the effect of SNAP was distinct from that of an inhibitor of protein synthesis since the latter did not prevent the phagosomal escape.
Fig. 1. LDH and MTT assays. Cytopathogenicity of uninfected J774 macrophages following treatment with SNAP (1 mM), FeTPPS (200 mM), SNAP+FeTPPS or SIN-1 (0.8 mM). J774 cells were incubated with the respective treatment for the indicated time and cytopathogenicity was followed as either LDH release or MTT reduction. Data represent the mean±SEM of three cultures from one representative experiment of at least two performed. Asterisks indicate a cytopathogenicity level significantly different to that of SNAP-treated J774 cells within each time point. *P,0.05, ** P,0.01, *** P,0.001, according to one-way ANOVA. http://jmm.sgmjournals.org
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1575
L. Tancred and others
Fig. 2. Quantification of LAMP-1 colocalization with F. tularensis LVS following treatment with SNAP (1 mM), FeTPPS (200 mM), SNAP+FeTPPS or chloramphenicol (CL; 20 mg ml”1). J774 cells were infected for 2 h with F. tularensis LVS at an m.o.i. of 500 or latex beads at an m.o.i. of 10 and after washing were further incubated for up to 24 h. Fixed samples were labelled for the late endosomal/lysosomal marker LAMP-1. Percentages represent the fraction of F. tularensis- or latex beadcontaining phagosomes stained for the late endosomal/lysosomal marker LAMP-1. Results are expressed as mean values±SEM from three separate experiments where 100 bacteria on triplicate coverslips were counted in each experiment. Asterisks indicate a colocalization level significantly different from that of LVS within each time point. **P,0.01, ***P,0.001, according to one-way ANOVA.
Effect of SNAP on expression of iglC and mglA The results showed that SNAP blocked the phagosomal escape and replication of LVS. In view of this and the previously published data demonstrating that DiglC and DmglA mutants are defective in intracellular replication (Bo¨nquist et al., 2008; Telepnev et al., 2003), we investigated whether the SNAP treatment affected expression of the two genes. Recombinant strains harbouring a fusion between either iglC, mglA or the polA gene and a gene encoding luciferase were used to infect the J774 cells. polA has previously been demonstrated to be constitutively expressed (Sullivan et al., 2006), a finding corroborated by us (data not shown). The luciferase activity from all constructs decreased markedly after addition of SNAP to the cultures (Fig. 4). The effect of SNAP was much more marked on the mglA reporter (P,0.001 at 0 and 1 h; P,0.01 at 2 h) than on the polA and iglC reporters. The effect of FeTPPS treatment could not be evaluated as the dark-coloured substance quenched luminescence. To further assess how the effect of SNAP was related to bacterial gene regulation, a real-time RT-PCR analysis was performed with RNA prepared from LVS-infected J774 cells that were untreated, treated with SNAP, or treated with SNAP+FeTPPS (Table 5). The samples were normalized for any differences in host cell numbers by using expression of the housekeeping gene b-actin since it was not affected by the treatments. The tul4 gene, encoding a 17 kDa membrane lipoprotein, has been previously described as constitutively expressed (Lauriano et al., 2004) and was included as a control. The addition of SNAP to the cell cultures reduced the expression of mglA approximately 50-fold at 1 h and 100-fold at 2 h (P,0.001 for both) relative to non-treated cultures, whereas the expression of iglC and tul4 was downregulated 4- and 10fold (P,0.001) at 1 and 2 h, respectively (Table 5). In the 1576
presence of SNAP+FeTPPS, the expression of DmglA increased relative to SNAP-treated cultures, but was still threefold lower (P,0.001) than in non-treated cultures. The downregulation of both iglC and tul4 was restored by addition of FeTPPS (Table 5). Both the luciferase gene reporter assay and the real-time PCR showed that SNAP treatment of LVS-infected cell cultures very significantly decreased the expression of mglA whereas the expression of iglC and tul4 was much less affected. This reduced expression of mglA when the macrophage culture was treated with SNAP may explain the inability of the bacteria to replicate. Effect of SNAP on intracellular signalling It is well known that intracellular signalling can be powerfully modulated by RNS (Radi, 2004). Thus, the effect of SNAP on the phagosomal escape of LVS may be due to modulation of the signalling in the infected J774 cells. To test this, we followed TNF-a secretion from the macrophages after addition of E. coli LPS. Previously we found that LPS-induced TNF-a secretion was virtually abolished by the F. tularensis infection (Telepnev et al., 2003) and this was confirmed here (Fig. 5). After 120 min, LVS-infected, LPS+SNAP-treated cells had secreted 1890 pg TNF-a ml21, and this was approximately 30-fold higher than that secreted from LVS-infected cells without SNAP, 70 pg ml21, and similar to the 2150 pg ml21 secreted from uninfected LPS-treated cells (Fig. 5). Thus, addition of SNAP almost completely reversed the F. tularensis-mediated inhibition. Notably, SNAP treatment of uninfected cultures did not stimulate TNF-a secretion and in fact conferred some inhibitory effect on the LPSstimulated TNF-a secretion (Fig. 5). This finding suggested that the effect of SNAP on TNF-a secretion was not directly related to the cell signalling. It should be remarked that NO
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis
Fig. 3. Colocalization of GFP-expressing F. tularensis LVS and the late endosomal marker LAMP-1. J774 cells were infected for 2 h with F. tularensis LVS at an m.o.i. of 500 or latex beads at an m.o.i. of 10 and further incubated for 2 h (a) or 24 h (b). In the representative confocal images, the green channel shows bacteria or latex particles and the red channel shows LAMP-1 staining for the indicated treatment. Bar, 5 mm. Confocal images were acquired with a Nikon C1 confocal microscope and assembled using Adobe Photoshop CS4 (Adobe Systems).
http://jmm.sgmjournals.org
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1577
L. Tancred and others
LPS-induced TNF-a secretion relied on iron as previously demonstrated (Xiong et al., 2004).
Fig. 4. Effect of SNAP treatment on the expression of iglC, mglA and a control gene, polA. Values are expressed as percentages of untreated cells. Asterisks indicate an RLU ratio significantly lower than that of LVSpolAluxAB. **P,0.01, ***P,0.001. Values are means from three experiments, performed with five replicates for each group.
per se in fact may effectuate inhibition of TNF-a secretion (Garba´n & Bonavida, 2001; Xiong et al., 2004). Treatment of both infected and uninfected cells with SNAP+FeTPPS completely blocked the LPS-induced TNF-a secretion (Fig. 5). Thus, the cell signalling is dependent on ONOO2 in both infected and uninfected cells. This is in agreement with previous studies on the mechanisms of LPS signalling in monocytic cells that demonstrated that a rapid, transient rise in the intracellular level of iron preceded NF-kB activation in response to LPS, and concluded that ONOO2 was the likely effector molecule for the cytosolic release of iron (Xiong et al., 2003). To test whether iron was also a messenger in our experimental model, the iron chelator DFO was supplied to the cultures. Addition of DFO to uninfected or infected cell cultures together with LPS+SNAP significantly lowered (P,0.05) the TNF-a levels compared to those in cultures treated with LPS+SNAP only (Fig. 5). Thus, optimal
Altogether, the results suggest that SNAP affects the bacteria, which thereby are incapable of inhibiting the LPS-induced TNF-a secretion, whereas FeTPPS, in contrast, likely inhibits the ONOO2-dependent cell signalling required for LPS-induced TNF-a secretion. These findings, together with the finding that SNAP treatment led to reduced escape of the bacteria from the phagosome, indicate that the phagosomal confinement of bacteria explains their inability to block the cell signalling leading to TNF-a secretion. MglA, IglB, IglC and IglD are required by LVS to inhibit LPS-induced cell signalling We also analysed the effects of infection with an MglA mutant or Igl mutants on cell signalling. J774 cells were infected with heat-killed bacteria, DmglA, or mutants of the MglA-regulated genes DiglB, DiglC and DiglD, and the effects on LPS-induced signalling were analysed. In all of the cell cultures infected with mutant strains or heat-killed bacteria, the level of secretion of TNF-a in response to LPS stimulation was similar to levels in uninfected, LPSstimulated cell cultures, whereas there were no detectable levels in F. tularensis LVS-infected cultures (Table 6). Thus, expression of each of the factors IglB, IglC, IglD and MglA in live bacteria is needed for the execution of the inhibition of TLR-induced signalling. We previously demonstrated that expression of the Igl proteins was below the detection level in the MglA mutant (Bro¨ms et al., 2009; Bo¨nquist et al., 2008), so either the levels of the Igl proteins per se are not sufficient for executing the inhibition of TNF-a secretion, or other MglA-regulated proteins, besides the Igl proteins, also contribute to the inhibition. In summary, an MglA-dependent regulation leading to a cytosolic location appears to be necessary for the inhibition of TNF-a secretion.
Table 5. Quantitative PCR analysis of the effects of SNAP and/or FeTTPS on gene regulation of intracellular F. tularensis LVS J774 cells were infected with F. tularensis (m.o.i.5500) and further incubated for 1 h or 2 h in the presence or absence of SNAP (1 mM) and/or FeTPPS (200 mM). Values represent fold expression level changes for each gene and treatment in relation to that of LVS in untreated cell cultures. Group
Fold expression relative to LVS 1h
LVS LVS+FeTPPS LVS+SNAP LVS+SNAP+FeTPPS
2h
mglA
iglC
tul4
mglA
iglC
tul4
1.0 1.5 246.3* 23.6*
1.0 21.5 24.1* 21.4
1.0 1.2 23.8* 1.1
1.0 1.3 2118.6* 23.3*
1.0 21.4 27.5* 21.8
1.0 21.2 29.4* 21.9
*Indicates a significantly different (P,0.001) expression level of the specified LVS gene compared to that of LVS in the untreated group within each time point. 1578
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis
Table 6. TNF-a secretion in E. coli LPS-stimulated J774 cells infected with F. tularensis LVS heat-killed bacteria or various mutant strains F. tularensis-infected (m.o.i.5500) or uninfected (2) J774 cells were incubated in the presence (+) or absence (2) of E. coli LPS (5 mg ml21). The results are from one representative experiment out of three performed. Data represent the mean TNF-a level (pg ml21) of three cultures±SEM. The detection limit was 15 pg ml21. After washing (0 min), all cytokine levels were below the limit of detection (data not shown). Strain
Fig. 5. TNF-a secretion (pg ml”1) in J774 cells after 2 h of LVS infection in the presence or absence of E. coli-derived LPS (5 mg ml”1), SNAP (1 mM), FeTPPS (200 mM) or DFO (400 mM). Black bars represent values from uninfected cultures and white bars represent values from LVS-infected cultures. Three asterisks indicate significant differences (P¡0.001) in TNF-a levels between uninfected and LVS-infected cultures. The experiment was repeated three times with triplicate wells.
Effect of SNAP on nitration of F. tularensis proteins As we observed effects of SNAP on intracellular viability of LVS and its localization and gene regulation, and each of the effects appeared to be at least partially dependent on ONOO2, we assessed the levels of nitrotyrosine-modified proteins in LVS bacteria after exposure to ONOO2 in vitro. Nitration is a low-abundance oxidative protein modification that can form from the reaction of tyrosine or tryptophan with RNS. Two bands containing intense nitrotyrosine-reactive material were reproducibly observed by Western blot analysis in repeated experiments (Fig. 6). Also, a third band of higher molecular mass, approximately 50 kDa, was sometimes observed. The two major bands were subjected to protein analysis and 2 and 5 proteins, respectively, were identified. The molecular masses of the proteins from each band varied at most by 1.5 kDa. Nitration had occurred in 1 and 3 of the protein bands, respectively, as determined by MS. Specific nitration of one or two tyrosines or tryptophans was present in each of the following four proteins: 50S ribosomal protein L16, 50S ribosomal protein L9, bacterioferritin and a CBS domain pair protein of unknown function (Table 7).
DISCUSSION Our data revealed the somewhat unexpected finding that intracellular F. tularensis bacteria were not killed within a 4 h period by the direct or indirect addition of RNS. This is in contrast to the rapid killing of extracellular bacteria mediated by SIN-1 (Lindgren et al., 2005), and a possible explanation may be that scavenging and detoxifying http://jmm.sgmjournals.org
2 2 LVS LVS DiglB DiglB DiglC DiglC DiglD DiglD DmglA DmglA HKD HKD
LPS
2 + 2 + 2 + 2 + 2 + 2 + 2 +
Time (h) 1h
2h
,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 ,15 33±5
,15* 1020±10 ,15* ,15* ,15* 812±6* ,15* 950±23 ,15* 1127±20 ,15* 1083±53 ,15* 1010±16
*Significantly different (P,0.001; one-way ANOVA) from the level observed in uninfected LPS-treated cell cultures. Comparisons for 2 h time point only. DHK, Heat-killed F. tularensis LVS bacteria.
intracellular mechanisms partially degrade the RNS and ROS. In fact, the mechanisms mediating killing of intracellular F. tularensis may much depend on the host cells. A recent study using murine and human monocytic cells identified a critical role for IFN-c, in agreement with many previous studies, but failed to identify the direct effectors since killing did not coincide with a defect in phagosomal escape and restriction of the cytosolic replication was independent of reactive oxygen or nitrogen species (Edwards et al., 2010). In agreement, we previously observed that IFN-c treatment of infected J774 cells did not affect the phagosomal escape of F. tularensis LVS (Bo¨nquist et al., 2008), but also, partly in contrast, that murine peritoneal exudate cells effectuated killing through a complex interaction between oxygen and nitrogen species and most likely ONOO2 was the predominant effector molecule (Lindgren et al., 2005). One cause for this discrepancy may be the cell type used since peritoneal exudate cells may exert more potent cidal activities than other monocytic cells, likely due to higher production of NO by the former. In the present study, we observed nitrite levels that were three to six times higher in the SNAPtreated cultures than what we previously observed in
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1579
L. Tancred and others
Fig. 6. Nitration of LVS proteins following exposure to ONOO” in vitro. LVS was incubated in PBS±0.8 mM SIN-1 at 37 6C for 2 h. Bacterial lysates corresponding to 2¾109 bacteria were separated on a 20 cm 15 % SDS-PAGE gel and blotted onto nitrocellulose. Nitration was reproducibly detected as two intense bands (arrows) following incubation with an anti-nitrotyrosine antibody.
activated peritoneal exudate cell cultures (Lindgren et al., 2005). Thus, the confinement of bacteria to the phagosome may be an effect of the high levels of NO. To predict how the addition of NO affects intracellular conditions is very difficult, since its propensity to autooxidize or react with other oxidizing and nitrosating molecules is not easily predicted, but, generally, these reactions are considered too slow to be biologically relevant (Reiter, 2006). However, it has been found that membranes and various proteins can accelerate these processes (Mo¨ller et al., 2007). Even so, NO is highly diffusible and it appears likely that significant intracellular concentrations will result from the SNAP treatment of the F. tularensis-infected J774 cells. At the same time, the formation of, for example, ONOO2 is also a likely event of the treatment. SNAP
treatment has been found to protect cells against oxidative stress by induction of certain heat-shock proteins but also to potently decrease mitochondrial respiration (Szabo´ & Salzman, 1995). If ONOO2 formation occurs, then the adverse effects on the cells will be more pronounced since it not only inhibits mitochondrial respiration, but also causes DNA breakage, consumption of ATP and NAD, and thereby cell damage (Vira´g et al., 2003). Our results indicated that ONOO2 may not be the only effector molecule of SNAP treatment during the first few hours of infection since addition of FeTPPS only in part reversed the effects of SNAP on the intracellular localization of the bacteria and their gene expression. However, the longterm, cidal effect of SNAP observed after 24 h was completely reversed by the addition of FeTPPS, strongly implying that it was mediated by ONOO2. Thus, the effects of SNAP may differ over time, and the initial effect, confinement of bacteria, may be dependent on both NO and ONOO2 whereas, over time, an ONOO2 effect becomes dominant and leads to bactericidi. Similar to our findings of a particularly strong effect of SNAP on the global regulator MglA, a study on the regulation of the Salmonella pathogenicity island-2 (SPI2) revealed that NO generated by IFN-activated macrophages inhibited expression of SPI2 effectors, and specifically that the expression of a sensor kinase, SsrA, was strongly reduced (McCollister et al., 2005), the expression of which is critical to the overall SPI2-dependent transcription. Altogether, this resulted in more efficient interaction between Salmonella-containing vacuoles and the late endosomal/lysosomal system, leading to decreased bacterial survival in IFN-c-activated phagocytes (McCollister et al., 2005). Thus, the continuously evolving interactions between host cells and intracellular pathogens may have resulted in the development of host mechanisms that specifically target bacterial virulence factors critical to their intracellular survival, such as SsrA and MglA. Since we observed that the chloramphenicol-mediated inhibition of
Table 7. Identification of nitrated LVS proteins Proteins in the two corresponding Coomassie-stained gel bands shown in Fig. 6 were identified by MS as: I, bacterioferritin and two 50S subunit ribosomal proteins; and II, a CBS domain pair protein of unknown function. Nitration of tyrosine or tryptophan of the identified proteins was confirmed by MS analysis. NA, Not applicable. Gel band I. I.
I. II.
1580
Protein name FTL
Molecular mass (kDa)
Number of Tyr, Trp
Nitration MS exp. 1
Nitration MS exp. 2
Nitration MS exp. 3
Bacterioferritin FTL0617 50S ribosomal protein L16 FTL0243 50S ribosomal protein L9 FTL1026 CBS domain pair protein FTL1075
16.8
Tyr: 6 Trp: 1 Tyr: 3 Trp: 2
Tyr123
NA
Tyr123
15.7
16.1 22.5
Tyr: 3 Trp: 0 Tyr: 11 Trp: 0
Trp93 Tyr92,104
2
Trp93 Tyr92,104
Tyr77
Tyr135
Tyr77
NA
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Tyr181
or 184
Tyr181
or 184
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis
protein synthesis did not affect escape from the phagosome, it appears likely that the inhibitory effect on mglA transcription per se did not affect the escape although it may effectively prevent bacterial multiplication if bacteria were localized in the cytosol. Besides the effect on MglA, we also observed that specific nitration occurred in LVS bacteria exposed to SNAP. The targets appeared to be highly specific since we repeatedly observed similar patterns of two strongly stained bands, and MS analyses revealed that the same amino acids were modified in different experiments. Nitration is often detrimental to protein function and is an irreversible modification. Even at low levels it can lead to toxic effects such as pro-oxidant, pro-apoptotic or pro-aggregant properties (Ferrer-Sueta & Radi, 2009). The fact that two ribosomal proteins were found to be nitrated may have contributed to a global effect of decreased de novo protein synthesis. Since such an effect may resemble that of chloramphenicol treatment, it presumably did not affect the phagosomal escape. Tyrosine nitration of bacterioferritin can affect its function as an iron storage protein but also as ROS detoxifier, thus such modification is likely to render LVS more susceptible to oxidative stress. A tyrosine chain near the di-iron centre has previously been hypothesized to constitute an electron transfer path during uptake and release of iron by bacterioferritins (Carrondo, 2003). Thus, it is likely that the identified modifications of the F. tularensis proteins lead to detrimental effects that potently affect the capacity of the bacterium to escape from the phagosome. There are two possible scenarios whereby addition of SNAP restores TNF-a secretion from F. tularensis-infected cells. One is via formation of ONOO2 and an agonistic effect on TNF-a-inducing pathways (Xiong et al., 2003); however, considering that we observed that addition of SNAP actually lowered the levels of LPS-induced TNF-a secretion, it appears to be a less likely explanation for the effects of SNAP, although we cannot completely exclude the possibility since it is difficult to exactly predict the interactions in the multiply treated, infected cells. An alternative explanation that appears more likely is that the phagosomal containment conferred by SNAP was directly linked to the lack of inhibitory effects on TNF-a secretion. The four mutants analysed, which did not repress TNF-a secretion, are all incapable of escaping from the phagosome, as demonstrated by us and others (Bo¨nquist et al., 2008; Lauriano et al., 2004; Santic et al., 2005). Thus, similar to bacteria in the cells that had been treated with SNAP, phagosomally located bacteria are incapable of inhibiting TNF-a secretion, as we previously demonstrated for the DiglC mutant (Telepnev et al., 2003). The findings collectively suggest that lack of a cytosolic location is a likely explanation for the absence of the inhibitory effects on TNF-a secretion. F. tularensis, similar to other successful intracellular pathogens, appears to rely on pathogenic mechanisms that http://jmm.sgmjournals.org
allow it to manipulate the intracellular environment through subversion of the normally highly potent killing mechanisms of monocytic cells. Accordingly, it has been viewed as a prototypic stealth pathogen (Sjo¨stedt, 2006). In support of this hypothesis, a recent study demonstrated that F. tularensis inhibits IFN-c-mediated signalling (Parsa et al., 2008). However, the inhibition occurred even when bacteria had an extracellular location, unlike the effects we observed, so there appears to be no direct link between the mechanisms. Another recent study demonstrated that infection with F. tularensis initially resulted in the upregulation of proinflammatory genes, but subsequently the inflammation was mitigated and production of IL-4 and IL-13 ensued, which redirected the macrophage phenotype and resulted in increased bacterial survival (Shirey et al., 2008). Our findings of inhibition of TNF-a secretion may be part of this redirected activation pattern, and similar to our findings, Shirey et al. (2008) demonstrated that the redirected macrophage differentiation required a cytosolic location of the bacteria. The present findings help to identify mechanisms exerted by macrophages to control the replication of F. tularensis and they are an illustration of how host mechanisms target bacterial factors essential for intracellular survival and replication.
ACKNOWLEDGEMENTS We thank Moa Lavander for construction of the mglAluxAB and polAluxAB strains. Grant support was obtained from the Swedish Medical Research Council (2010-9485) and the Medical Faculty, Umea˚ University, Umea˚, Sweden. The work was performed in part at the Umea˚ Centre for Microbial Research (UCMR).
REFERENCES Andersson, H., Hartmanova´, B., Ryde´n, P., Noppa, L., Na¨slund, L. & Sjo¨stedt, A. (2006). A microarray analysis of the murine macrophage
response to infection with Francisella tularensis LVS. J Med Microbiol 55, 1023–1033. Bogdan, C. (2001). Nitric oxide and the immune response. Nat
Immunol 2, 907–916. Bo¨nquist, L., Lindgren, H., Golovliov, I., Guina, T. & Sjo¨stedt, A. (2008). MglA and Igl proteins contribute to the modulation of
Francisella tularensis live vaccine strain-containing phagosomes in murine macrophages. Infect Immun 76, 3502–3510. Bro¨ms, J. E., Lavander, M. & Sjo¨stedt, A. (2009). A conserved alpha-
helix essential for a type VI secretion-like system of Francisella tularensis. J Bacteriol 191, 2431–2446. Bro¨ms, J. E., Sjo¨stedt, A. & Lavander, M. (2010). The role of the
Francisella tularensis pathogenicity island in type VI secretion, intracellular survival, and modulation of host cell signaling. Front Microbiol 1, 136. Carrondo, M. A. (2003). Ferritins, iron uptake and storage from the
bacterioferritin viewpoint. EMBO J 22, 1959–1968. Charity, J. C., Costante-Hamm, M. M., Balon, E. L., Boyd, D. H., Rubin, E. J. & Dove, S. L. (2007). Twin RNA polymerase-associated proteins
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1581
L. Tancred and others control virulence gene expression in Francisella tularensis. PLoS Pathog 3, e84. Checroun, C., Wehrly, T. D., Fischer, E. R., Hayes, S. F. & Celli, J. (2006). Autophagy-mediated reentry of Francisella tularensis into the
McCollister, B. D., Bourret, T. J., Gill, R., Jones-Carson, J. & Va´zquezTorres, A. (2005). Repression of SPI2 transcription by nitric oxideproducing, IFNc-activated macrophages promotes maturation of
Salmonella phagosomes. J Exp Med 202, 625–635.
endocytic compartment after cytoplasmic replication. Proc Natl Acad Sci U S A 103, 14578–14583.
Misko, T. P., Highkin, M. K., Veenhuizen, A. W., Manning, P. T., Stern, M. K., Currie, M. G. & Salvemini, D. (1998). Characterization of the
Chong, A., Wehrly, T. D., Nair, V., Fischer, E. R., Barker, J. R., Klose, K. E. & Celli, J. (2008). The early phagosomal stage of Francisella
cytoprotective action of peroxynitrite decomposition catalysts. J Biol Chem 273, 15646–15653.
tularensis determines optimal phagosomal escape and Francisella pathogenicity island protein expression. Infect Immun 76, 5488– 5499.
Mo¨ller, M. N., Li, Q., Lancaster, J. R., Jr & Denicola, A. (2007).
Clemens, D. L., Lee, B. Y. & Horwitz, M. A. (2004). Virulent and
Nano, F. E. & Schmerk, C. (2007). The Francisella pathogenicity island. Ann N Y Acad Sci 1105, 122–137.
avirulent strains of Francisella tularensis prevent acidification and maturation of their phagosomes and escape into the cytoplasm in human macrophages. Infect Immun 72, 3204–3217. Dai, S., Mohapatra, N. P., Schlesinger, L. S. & Gunn, J. S. (2011).
Regulation of Francisella tularensis virulence. Front Microbiol 1, 144. Edwards, J. A., Rockx-Brouwer, D., Nair, V. & Celli, J. (2010).
Restricted cytosolic growth of Francisella tularensis subsp. tularensis by IFN-c activation of macrophages. Microbiology 156, 327–339.
Acceleration of nitric oxide autoxidation and nitrosation by membranes. IUBMB Life 59, 243–248.
Oyston, P. C. (2008). Francisella tularensis: unravelling the secrets of
an intracellular pathogen. J Med Microbiol 57, 921–930. Parsa, K. V., Butchar, J. P., Rajaram, M. V., Cremer, T. J., Gunn, J. S., Schlesinger, L. S. & Tridandapani, S. (2008). Francisella gains a
survival advantage within mononuclear phagocytes by suppressing the host IFNc response. Mol Immunol 45, 3428–3437. Radi, R. (2004). Nitric oxide, oxidants, and protein tyrosine nitration.
Elkins, K. L., Cowley, S. C. & Bosio, C. M. (2007). Innate and adaptive
Proc Natl Acad Sci U S A 101, 4003–4008.
immunity to Francisella. Ann N Y Acad Sci 1105, 284–324. Feelisch, M. & Kelm, M. (1991). Biotransformation of organic nitrates
Rajaram, M. V., Ganesan, L. P., Parsa, K. V., Butchar, J. P., Gunn, J. S. & Tridandapani, S. (2006). Akt/Protein kinase B modulates macro-
to nitric oxide by vascular smooth muscle and endothelial cells. Biochem Biophys Res Commun 180, 286–293.
phage inflammatory response to Francisella infection and confers a survival advantage in mice. J Immunol 177, 6317–6324.
Ferrer-Sueta, G. & Radi, R. (2009). Chemical biology of peroxynitrite: kinetics, diffusion, and radicals. ACS Chem Biol 4, 161–177.
Reiter, T. A. (2006). NO* chemistry: a diversity of targets in the cell. Redox Rep 11, 194–206.
Forsberg, A. J., Pavitt, G. D. & Higgins, C. F. (1994). Use of
Roth, K. M., Gunn, J. S., Lafuse, W. & Satoskar, A. R. (2009).
transcriptional fusions to monitor gene expression: a cautionary tale. J Bacteriol 176, 2128–2132.
Francisella inhibits STAT1-mediated signaling in macrophages and prevents activation of antigen-specific T cells. Int Immunol 21, 19–28.
Garba´n, H. J. & Bonavida, B. (2001). Nitric oxide disrupts H2O2dependent activation of nuclear factor kB. Role in sensitization of human tumor cells to tumor necrosis factor-a-induced cytotoxicity.
Rubbo, H. & Radi, R. (2008). Protein and lipid nitration: role in redox signaling and injury. Biochim Biophys Acta 1780, 1318–1324.
J Biol Chem 276, 8918–8923.
The Francisella tularensis pathogenicity island protein IglC and its regulator MglA are essential for modulating phagosome biogenesis and subsequent bacterial escape into the cytoplasm. Cell Microbiol 7, 969–979.
Golovliov, I., Sjo¨stedt, A., Mokrievich, A. & Pavlov, V. (2003a). A
method for allelic replacement in Francisella tularensis. FEMS Microbiol Lett 222, 273–280.
Santic, M., Molmeret, M., Klose, K. E., Jones, S. & Kwaik, Y. A. (2005).
Golovliov, I., Baranov, V., Krocova, Z., Kovarova, H. & Sjo¨stedt, A. (2003b). An attenuated strain of the facultative intracellular
Schmittgen, T. D. & Livak, K. J. (2008). Analyzing real-time PCR data
bacterium Francisella tularensis can escape the phagosome of monocytic cells. Infect Immun 71, 5940–5950.
Shirey, K. A., Cole, L. E., Keegan, A. D. & Vogel, S. N. (2008).
Gordon, S. (2003). Alternative activation of macrophages. Nat Rev
Immunol 3, 23–35. Larsson, P., Oyston, P. C. F., Chain, P., Chu, M. C., Duffield, M., Fuxelius, H.-H., Garcia, E., Ha¨lltorp, G., Johansson, D. & other authors (2005). The complete genome sequence of Francisella
by the comparative C(T) method. Nat Protoc 3, 1101–1108. Francisella tularensis live vaccine strain induces macrophage alternative activation as a survival mechanism. J Immunol 181, 4159– 4167. Sjo¨stedt, A. (2006). Intracellular survival mechanisms of Francisella tularensis, a stealth pathogen. Microbes Infect 8, 561–567.
tularensis, the causative agent of tularemia. Nat Genet 37, 153–159.
Sullivan, J. T., Jeffery, E. F., Shannon, J. D. & Ramakrishnan, G. (2006). Characterization of the siderophore of Francisella tularensis
Lauriano, C. M., Barker, J. R., Yoon, S. S., Nano, F. E., Arulanandam, B. P., Hassett, D. J. & Klose, K. E. (2004). MglA regulates
and role of fslA in siderophore production. J Bacteriol 188, 3785– 3795.
transcription of virulence factors necessary for Francisella tularensis intraamoebae and intramacrophage survival. Proc Natl Acad Sci U S A 101, 4246–4249.
Szabo´, C. & Salzman, A. L. (1995). Endogenous peroxynitrite is
Lindgren, H., Golovliov, I., Baranov, V., Ernst, R. K., Telepnev, M. & Sjo¨stedt, A. (2004). Factors affecting the escape of Francisella
tularensis from the phagolysosome. J Med Microbiol 53, 953–958. Lindgren, H., Stenman, L., Ta¨rnvik, A. & Sjo¨stedt, A. (2005). The
contribution of reactive nitrogen and oxygen species to the killing of Francisella tularensis LVS by murine macrophages. Microbes Infect 7, 467–475. Liochev, S. I. & Fridovich, I. (1999). Superoxide and iron: partners in
crime. IUBMB Life 48, 157–161. 1582
involved in the inhibition of mitochondrial respiration in immunostimulated J774.2 macrophages. Biochem Biophys Res Commun 209, 739–743. Telepnev, M., Golovliov, I., Grundstro¨m, T., Ta¨rnvik, A. & Sjo¨stedt, A. (2003). Francisella tularensis inhibits Toll-like receptor-mediated activation of intracellular signalling and secretion of TNF-a and IL-
1 from murine macrophages. Cell Microbiol 5, 41–51. Telepnev, M., Golovliov, I. & Sjo¨stedt, A. (2005). Francisella tularensis
LVS initially activates but subsequently down-regulates intracellular signaling and cytokine secretion in mouse monocytic and human peripheral blood mononuclear cells. Microb Pathog 38, 239–247.
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
Journal of Medical Microbiology 60
Nitric oxide inhibits mglA expression by F. tularensis
Valko, M., Leibfritz, D., Moncol, J., Cronin, M. T., Mazur, M. & Telser, J. (2007). Free radicals and antioxidants in normal physiological
Woolard, M. D., Wilson, J. E., Hensley, L. L., Jania, L. A., Kawula, T. H., Drake, J. R. & Frelinger, J. A. (2007). Francisella tularensis-infected
functions and human disease. Int J Biochem Cell Biol 39, 44–84.
macrophages release prostaglandin E2 that blocks T cell proliferation and promotes a Th2-like response. J Immunol 178, 2065–2074.
Vira´g, L., Szabo´, E., Gergely, P. & Szabo´, C. (2003). Peroxynitrite-
induced cytotoxicity: mechanism and opportunities for intervention. Toxicol Lett 140–141, 113–124. Wehrly, T. D., Chong, A., Virtaneva, K., Sturdevant, D. E., Child, R., Edwards, J. A., Brouwer, D., Nair, V., Fischer, E. R. & other authors (2009). Intracellular biology and virulence determinants of Francisella
tularensis revealed by transcriptional profiling inside macrophages. Cell Microbiol 11, 1128–1150.
http://jmm.sgmjournals.org
Xiong, S., She, H., Takeuchi, H., Han, B., Engelhardt, J. F., Barton, C. H., Zandi, E., Giulivi, C. & Tsukamoto, H. (2003). Signaling role of intracellular iron in NF-kB activation. J Biol Chem 278, 17646–17654. Xiong, H., Zhu, C., Li, F., Hegazi, R., He, K., Babyatsky, M., Bauer, A. J. & Plevy, S. E. (2004). Inhibition of interleukin-12 p40 transcription and NF-kB activation by nitric oxide in murine macrophages and
dendritic cells. J Biol Chem 279, 10776–10783.
Downloaded from www.microbiologyresearch.org by IP: 54.242.161.225 On: Sat, 19 Mar 2016 08:03:36
1583
