| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Clinical Microbiology, Clinical Bacteriology, Umeå University, SE-901 85 Umeå, Sweden
2 National Research Council Canada, Institute for Biological Sciences, Ottawa, Ontario, K1A 0R6, Canada
3 Proteome Center for the Study of Intracellular Parasitism of Bacteria, Faculty of Military Health Science, University of Defence, Trebesská 1575, 500 01 Hradec Králové, Czech Republic
Correspondence
Anders Sjöstedt
Anders.Sjostedt{at}climi.umu.se
Received 1 September 2005
Accepted 28 October 2005
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The response to intracellular infections is often studied in vitro using infected, purified host cell populations, such as macrophages or dendritic cells, that are the primary targets for the infectious agents as models. However, intracellular infections elicit much broader responses from the host since they usually trigger a vigorous inflammatory response. This response profoundly affects tissues and cells adjacent to infectious foci and sometimes also results in systemic effects. Easily accessible samples to analyse local and systemic effects of infection-induced inflammation include peripheral blood cells or infected whole organ samples. By analysing transcriptional and translational profiles from such mixed cell populations at various times during the infectious process, a detailed picture of the overall effects of infection on the host can be obtained. Such an analysis will help us to understand the pathogenesis of the disease and to identify pathogen-specific inflammatory profiles.
Tularaemia is a zoonotic disease caused by the intracellular bacterial pathogen Francisella tularensis. In nature, it occurs predominantly among hares, rabbits and rodents, and spreads via direct contact with primary hosts or via arthropods or by aerosols to humans (Oyston et al., 2004). A characteristic feature of F. tularensis is its ability to withstand the antimicrobial effects of macrophages and to proliferate effectively intracellularly (Sjöstedt, 2003). Most experimental models of tularaemia have used attenuated strains such as the live vaccine strain (LVS). However, recent work has shown that there are key differences between the pathogenesis of the infection caused by LVS and virulent strains of F. tularensis (Chen et al., 2003; Conlan et al., 2003). For example, sublethal infection with LVS confers protection against aerosol challenge with LVS, but not virulent strains of the pathogen (Conlan et al., 2005). Finally, immunodeficient mice such as those unable to produce gamma interferon (IFN-
) or nitric oxide from inducible nitric oxide synthase or mice with broad immunodeficiencies such as SCID mice or neutropenic mice are rendered much more susceptible to LVS, but not to virulent strains (Chen et al., 2004). This indicates that immunocompetent mice are virtually defenceless against aerosol or intradermal challenges with even the lowest inocula of virulent F. tularensis strains. The present study was undertaken in order to better understand the extreme virulence of type A F. tularensis strains. It characterizes the transcriptional profile of the murine host response in the lung during the first 4 days of an infection with a highly virulent strain of the pathogen.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
F. tularensis and experimental infections. Stocks of type A F. tularensis strain FSC033/snMF, originally isolated from a squirrel in Georgia, USA (Johansson et al., 2000), were prepared as previously described (Conlan et al., 2003). For aerosol exposure, thawed F. tularensis stocks were diluted in Mueller–Hinton broth containing 20 % (v/v) glycerol to maintain infectivity at the high relative humidity (70–80 %) employed. Aerosols of type A F. tularensis strain FSC033/snMF were generated with a Lovelace nebulizer operating at a pressure of 40 p.s.i. (276 Pa) to produce particles in the 4–6 µm range required for inhalation and retention in the alveoli (Fitzgeorge et al., 1983). Mice were exposed to these aerosols for 7 min using a customized commercial nose-only exposure apparatus (In-tox Products, Albuquerque, NM, USA), resulting in the implantation of 10–20 organisms into the lungs (Chen et al., 2003; Conlan et al., 2003).
Bronchoalveolar lavage (BAL). Groups of six mice were killed on days 0, 1, 2 and 4 by CO2 asphyxiation. The trachea was exposed through a midline incision and cannulated with a plastic catheter. The lungs were lavaged by instillation of five 1·0 ml aliquots of PBS supplemented with 3 mM EDTA (Chen et al., 1992). The lavage fluid was centrifuged at 2450 g for 7 min and the supernatant was collected, filter-sterilized and stored at –80 °C for cytokine analysis.
Total RNA isolation and quality confirmation. Groups of six mice were sacrificed by CO2 asphyxiation on days 0, 1, 2 and 4 after aerosol exposure. Uninfected mice were sacrificed as controls. Lungs from individual mice were dissected from the thoracic cavity and immediately placed in 1·0 ml RNAlater (Qiagen) and stored at –80 °C until needed.
Total RNAs were isolated from the lungs using an RNeasy isolation kit (Qiagen) and stored at –70 °C until use. RNA samples with a ratio of A260 to A280 between 2·0 and 2·1 were used for experiments. The quality of the RNA was also checked by running 300 ng of each sample on a Lab-On-A-Chip (Caliper Technologies Corp.) which was evaluated on a bioanalyser (Agilent Technologies); only non-degraded RNA samples were used for microarray and quantitative real-time PCR (Q-PCR) experiments.
Microarray construction. In-house-produced cDNA arrays consisting of 20 600 mouse clones derived from two different clone sets were used. These include a 15 000 mouse cDNA set from the National Institute of Aging (Kargul et al., 2001; http://lgsun.grc.nia.nih.gov/cDNA/15k.html) and a 5400 cDNA clone set obtained from Research Genetics. Universal Scorecard (Amersham), five plant clones (Pinus sylvestris) and 18 housekeeping genes (cloned in-house) were included as controls. The clones were amplified by PCR and the purified PCR products were dissolved in 50 % DMSO and then spotted on microscope slides using a Microgrid II arrayer (Biorobotics).
Labelling and hybridization. RNA concentration was measured with Nanodrop (NanoDrop Technologies) and adjusted to approximately 25 µg per sample. Synthesis and labelling of cDNA was performed according to a protocol available at http://www.umu.se/climi/bact/Microarray/index.html. Briefly, aminoallyl-dUTP (Amersham) was incorporated into cDNA during reverse transcription using Superscript II enzyme (Gibco-BRL). cDNA was then indirectly labelled with Cy3 or Cy5 fluorescent dyes (Amersham). A test sample from one mouse and a reference sample from a pool of untreated mice were labelled with Cy3 and Cy5 dye, respectively. The fluorophores were reversed on every other array to compensate for dye bias. Labelled cDNA from test and reference samples was mixed and purified. cDNA was then dissolved in DIG Easy Hyb solution (Roche) supplemented with tRNA (Sigma) and fish sperm DNA (Sigma). Hybridization was performed at the temperature recommended for the DIG Easy Hyb, 37 °C, overnight in a Genetac Hybridization Station (Genomic Solutions). Slides were washed sequentially with 0·1x SSC, 0·1 % SDS followed by 0·1x SSC.
Statistical analysis. Between nine and 11 replicated arrays were performed for each of the three time points examined. Arrays were scanned using a Scanarray 4000XL (Perkin Elmer) at three different intensity settings. The images were analysed with the software ScanArray Express (Perkin Elmer). Median signals were obtained and five or six replicated arrays of good quality were used for statistical analysis. Background correction was performed by subtracting the local background signals from the foreground signals. The background-corrected signals from the three scans of each array were combined using a method similar to that of Dudley et al. (2002). Data from the two channels on each array were normalized using print-tip MA-loess (Yang et al., 2002). The normalized data were analysed using the B statistic suggested by Lonnstedt & Speed (2002). Differentially expressed genes were selected according to ratio and B values.
Q-PCR. RNA from the same control and test samples used in the microarray experiments was reverse transcribed into cDNA. Specific primers for selected genes were designed using the Primer3 program (http://frodo.wi.mit.edu/cgi-bin/primer3/primer3_www.cgi).
Q-PCR was performed using SYBR green I PCR kit (Applied Biosystems) in an ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Each reaction contained 12·5 µl SYBR green I PCR kit, 250 nM reverse and forward primers and 5 µl cDNA. The total volume of the reaction was adjusted with water to 25 µl. PCR was performed in MicroAmp 96-well plates (Applied Biosystems) capped with MicroAmp optical adhesive seal. The reactions were incubated at 50 °C for 2 min, 10 min at 95 °C, followed by 45 cycles of 15 s at 95 °C and 1 min at 60 °C. The PCRs were subjected to a heat dissociation protocol present in the ABI SDS 2.0 software (Applied Biosystems).
BAL tumour necrosis factor alpha (TNF-
) and IFN- measurements.
The levels of TNF- and IFN- in undiluted BAL supernatants were measured by the Beadlyte mouse cytokine detection kit (Upstate Biotechnology) on a Luminex 100IS system (Luminex Corp.) according to the manufacturers' instructions. The lower limit of detection was 4·5 pg ml–1 for TNF- and 3·1 pg ml–1 for IFN-.
|
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
20 c.f.u.) aerosol model of type A F. tularensis infection (Conlan et al., 2003), all mice developed clinical signs of illness including weight loss at day 4 and died of the infection by day 5 (Fig. 1
a). Consistent with our previous observations (Conlan et al., 2003), histopathological examination of the lung from mice killed on day 4 of infection revealed acute suppurative and necrotic bronchopneumonia, which involved one or more lung lobes (Fig. 1b, c), and acute to subacute vasculitis and perivasculitis.
|
We categorized a total of 167 differentially expressed genes with clear identities and characterized functions. We grouped the differentially expressed genes into the following categories: (i) metabolism (44 %), (ii) signal transduction (6 %), (iii) growth (12 %), (iv) response to pathogens (5 %), (v) ion homeostasis (5 %) and (vi) miscellaneous functions (28 %).
Cluster analysis of differentially expressed genes
Differentially expressed genes with ratios >2 or <0·5 and B values greater than zero were selected for cluster analysis. Replicated clones and clones with missing values were excluded. Genes were categorized as upregulated (1), downregulated (–1) or unchanged (0), and then subjected to SOM (self-organising map) cluster analysis. The results fitted into 22 out of 27 possible clusters (3x3x3). Six clusters contained more than 10 genes, constituting 84 % of the genes (Fig. 2). Genes included in cluster 1 participate in host defence responses to infection and they were activated extensively on day 4 but hardly or not at all on days 1 and 2. Genes in cluster 2 are involved in cellular killing mechanisms. These genes were downregulated on day 1 and unchanged on days 2 and 4. We hypothesize that this is due to a bacteria-triggered downregulation to avoid killing mechanisms of the host cell. In keeping with this, we have previously demonstrated that macrophages infected in vitro with LVS rapidly downregulate intracellular signalling and cytokine secretion in macrophages (Telepnev et al., 2003, 2005). Cluster 7 contains genes that participate in regulation of transcription and metabolism. They were upregulated on day 1 and unchanged on days 2 and 4. This may reflect the stimulation of cell expression by bacteria to obtain resources necessary for their proliferation.
|
). The correlation between microarray data and Q-PCR results in terms of the magnitude and direction of gene expression patterns for all genes validated was significant (r2=0·6795, P<0·0001), indicating good agreement between the two assays for the identification of differentially regulated genes (Table 1 and Fig. 3). Q-PCR analyses were performed for 27 genes and three time points. All of these genes had been found to be significantly differentially expressed at least for one time point by the microarray analysis with the exception of TNF- and IFN-, which were expressed at such low levels that they were below the detection limit. They were included because they are known to be important for the development of protective Th1 immune responses. All of these genes were confirmed as differentially expressed at least for one time point (P<0·05) by Q-PCR and, of these, 21 were significant after a Bonferroni correction for mass significance. Of the remaining samples for these genes that were not identified as differentially expressed by microarray analysis, 18 were found to be significantly regulated (P<0·05) by Q-PCR using the Bonferroni correction.
|
|
|
and IFN- mRNA by quantification of these cytokines in the lung. In support of the gene expression data, both TNF- and IFN- proteins were significantly upregulated in the BAL fluid on day 4 after aerosol exposure to type A F. tularensis (Fig. 4).
|
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous work has demonstrated that mice that inhale a low dose of type A F. tularensis strain FSC033 develop clinical signs of disease by day 3 and die within 5 days (Conlan et al., 2003). Around day 3 of infection, in conjunction with the onset of clinical symptoms, very large numbers of bacteria (
108 c.f.u. per organ) are present in lung, liver and spleen (Conlan et al., 2003). The present data on gene regulation and cytokine levels in the lungs of the infected mice correlate well with the clinical presentation. During the first 2 days of infection, very few inflammatory changes were observed in the lungs whereas, by day 4, histology revealed localized massive, acute inflammation and intense vasculitis. Also, levels of IFN- and TNF- in BAL rose dramatically from day 2 to day 4.
The failure to detect TNF- and IFN- transcripts by the microarray demonstrates an obvious limitation of this technique. When constitutive levels of a given gene are very low, even a marked upregulation may not be detected by the DNA microarray since the increased levels are still below the detection limit. Both cytokine genes are known to be expressed at low constitutive levels (Fink et al., 1998; Whelan et al., 2003).
A number of genes that were strongly upregulated on day 4 of infection are known to be dependent on IFN- for their transcription. These include ß2 microglobulin, a glycoprotein constituting a substantial part of the major histocompatibility complex class I (MHC I) molecule. It is also involved in iron-uptake metabolism (Waheed et al., 2002). Its transcription is in part dependent on nuclear factor kappa-B (NF-
B) and interferon regulatory factors (IRFs) and is upregulated by cytokines such as TNF-, IFN-, IFN-ß and IFN-. Guanylate nucleotide-binding protein 2 (GBP-2) was also found to be strongly upregulated on day 4 of infection. It is a member of the p65 GBP family, one of four families of IFN-inducible GTPases. Its expression in macrophages is induced by IFN- and lipopolysaccharide (LPS). It was found to be strongly induced in the livers of mice infected with another intracellular bacterial pathogen, Listeria monocytogenes, in an IRF-1-dependent manner (Boehm et al., 1998).
Additionally, IFN--induced GTPase was slightly downregulated on days 1 and 2 but, together with T-cell-specific GTPase, strongly upregulated on day 4. Both are members of the p47 group of interferon-induced GTPases. They are active in phagocytic and secretory pathways and have been shown to be induced during infection by the intracellular pathogens Toxoplasma gondii (Collazo et al., 2002) and L. monocytogenes (Macmicking, 2005; McCaffrey et al., 2004). Pre-B-cell colony-enhancing factor was upregulated on day 4 of infection. Originally, it was found to act as a growth factor for B cells but was later identified as an inflammatory cytokine, the expression of which is enhanced by IFN-. Spi2a was downregulated on day 1 but strongly upregulated on day 4. It belongs to a family of intracellular and extracellular serine and cysteine proteinase inhibitors called serpins. Their target enzymes regulate very diverse processes, including apoptosis, inflammation and neoplasia (Gettins, 2000). Its expression has been shown to be strongly induced by IFN- and in vivo by intracellular bacteria such as Mycobacterium bovis bacille Calmette–Guérin (BCG), L. monocytogenes and Salmonella typhimurium, and by bacterial products like LPS or peptidoglycan (Hamerman et al., 2002).
Some genes upregulated on day 4 of infection are known to be induced by TNF-. For instance, LPS-induced TN factor (Litaf, TBX-1) is a transcription factor, responsible for TNF- expression after LPS stimulation (Myokai et al., 1999). It was upregulated at all three time points during infection. The same type of upregulation was also observed for tissue plasminogen activator (Plat), which is a serine protease that converts plasminogen to plasmin. Upregulation of host plasminogen activators has also been observed during other bacterial infections. For example, urokinase (one of two mammalian plasminogen activators) was enhanced during infection with Staphylococcus aureus and Borrelia burgdorferi (Lähteenmaki et al., 2005). Induction of plasminogen activators is regulated by a number of cytokines such as TNF- and interleukin 1 and 2.
Tissue inhibitors of metalloproteinases (Timps) are natural inhibitors of proteolytic activity of matrix metalloproteinases (Lambert et al., 2004). Timp3 was upregulated at all three time points investigated. It is located in the extracellular matrix and can control levels of TNF- by inhibition of the TNF--converting enzyme (TACE) (Black, 2004).
The S100 calcium-binding protein A9 (S100A9) is one of three calcium-binding proteins from the S100 family, which constitutes a group of proinflammatory proteins released from phagocytic cells, especially neutrophils (Roth et al., 2003). Expression of S100A9 is induced by LPS and oxygen species and induces release of neutrophils from bone marrow and recruitment of leukocytes (Vandal et al., 2003). Elevated expression of S100A9 was found in severe acute respiratory syndrome patients (Reghunathan et al., 2005). In our inhalation tularaemia model, it was downregulated on day 1 but upregulated on days 2 and 4.
Transglutaminase 2 belongs to the thiol- and Ca2+-dependent acyl transferases. When it is aberrantly activated in tissues and cells, dysregulated inflammation results and this may relate to a variety of diseases, including neurodegenerative and autoimmune diseases. It was upregulated at all three time points during infection. This enzyme has a mitochondrial location and causes selective depletion of mitochondrial glutathione, leading to a decreased ability to resist oxidative stress (Newsholme & Calder, 1997). Downregulation of peroxisome proliferative activated receptor gamma coactivator 1 beta (Ppargc1b), which was observed on days 2 and 4, may also contribute to a diminished ability to withstand reactive oxygen species because it is critically required in mitochondrial metabolism and respiration (St-Pierre et al., 2003). It is one of many co-activators of nuclear receptors and interacts with and enhances the activity of hepatic nuclear factor-4, peroxisome proliferative activated receptor- and the glucocorticoid receptor. Tropomyosin was downregulated at all three time points. It is part of the cytoskeleton, but it has also been identified as an oxygen sensor and its downregulation may contribute to a decreased cellular adaptation to oxidative stress (Thorne et al., 2004).
Overall, the findings in the current study demonstrated that a vigorous cellular inflammatory response ensues in Francisella-infected mouse lungs as evidenced by a marked increase of IFN- and TNF- in BAL fluid and concomitant activation of pathways known to be regulated by these cytokines. Thus, experimental inhalation tularaemia appears to result in an innate Th1 type immune response. Despite this apparently appropriate host immune response, type A F. tularensis infection is uniformly lethal in inbred and outbred mice regardless of their genetic background (Conlan et al., 2003). One notable observation is that these events appear to be initiated after the second day of infection and it is known that the target organs harbour more than 108 c.f.u. of bacteria a day later. Hence, it is possible that bacterial replication is so rapid that it overwhelms even the prominently activated antimicrobial host immune mechanisms observed on day 4 of infection. It has been observed that immune activation mediated in vitro via toll-like receptor 2 and 4 is inhibited in F. tularensis-infected human and mouse monocytic cells (Telepnev et al., 2003, 2005). The relevance of these findings to the situation in vivo is not known, however, and the findings on day 4 indicate clearly that there is no general suppression of immune responses in the infected mice. This does not exclude the possibility that the effects observed in vitro are still relevant to the situation in an individual cell in vivo. Thus, one of the reasons for the extremely rapid replication of type A F. tularensis in vivo may be unrestricted growth in macrophages that are unable to activate effective antibacterial defence mechanisms or that virulent strains of F. tularensis are able to modulate the host response by delaying the generation and secretion of IFN- and TNF-.
The fact that the mouse model of inhalation tularaemia results in 100 % mortality even at low dose might be considered aberrant compared with the human disease. However, this is also the case in other models of type A tularaemia such as the rabbit or the monkey, in which the bacterium causes an extremely aggressive infection and the lethal doses are very low (Baskerville & Hambleton, 1976; Baskerville et al., 1978). Even in humans, inhaled inocula of <100 c.f.u. cause severe clinical disease (Saslaw et al., 1961) with mortality of 30–60 % in the absence of effective chemotherapy. Therefore, the findings of the present study might well mimic the pathogenesis of tularaemia in other host species.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
ß T cell- and interferon gamma-dependent mechanism. Vaccine 23, 2477–2485.[CrossRef][Medline]This article has been cited by other articles:
|
|
 |
|
  P. E. Carlson Jr., J. Horzempa, D. M. O'Dee, C. M. Robinson, P. Neophytou, A. Labrinidis, and G. J. Nau Global Transcriptional Response to Spermine, a Component of the Intramacrophage Environment, Reveals Regulation of Francisella Gene Expression through Insertion Sequence Elements J. Bacteriol., November 15, 2009; 191(22): 6855 - 6864. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. C. Cowley Editorial: Proinflammatory cytokines in pneumonic tularemia: too much too late? J. Leukoc. Biol., September 1, 2009; 86(3): 469 - 470. [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Schmerk, B. N. Duplantis, P. L. Howard, and F. E. Nano A Francisella novicida pdpA mutant exhibits limited intracellular replication and remains associated with the lysosomal marker LAMP-1 Microbiology, May 1, 2009; 155(5): 1498 - 1504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Miura and Y. Rikihisa Liver Transcriptome Profiles Associated with Strain-Specific Ehrlichia chaffeensis-Induced Hepatitis in SCID Mice Infect. Immun., January 1, 2009; 77(1): 245 - 254. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Mares, S. S. Ojeda, E. G. Morris, Q. Li, and J. M. Teale Initial Delay in the Immune Response to Francisella tularensis Is Followed by Hypercytokinemia Characteristic of Severe Sepsis and Correlating with Upregulation and Release of Damage-Associated Molecular Patterns Infect. Immun., July 1, 2008; 76(7): 3001 - 3010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. KuoLee, X. Zhao, J. Austin, G. Harris, J. W. Conlan, and W. Chen Mouse Model of Oral Infection with Virulent Type A Francisella tularensis Infect. Immun., April 1, 2007; 75(4): 1651 - 1660. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Bosio, H. Bielefeldt-Ohmann, and J. T. Belisle Active Suppression of the Pulmonary Immune Response by Francisella tularensis Schu4 J. Immunol., April 1, 2007; 178(7): 4538 - 4547. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
, solid line) and body weight loss (
, dashed line) of C57BL/6 mice after aerosol infection with 
