Protective cellular responses to Burkholderia mallei infection

Microbes and Infection 12 (2010) 846e853 www.elsevier.com/locate/micinf

Original article

Protective cellular responses to Burkholderia mallei infection Caroline A. Rowland a,*, M. Stephen Lever a, Kate F. Griffin a, Gregory J. Bancroft b, Roman A. Lukaszewski a b

a Dstl, Biomedical Sciences, Porton Down, Salisbury, SP4 0JQ, UK London School of Hygiene and Tropical Medicine, Keppel St, London, UK

Received 5 January 2010; accepted 31 May 2010 Available online 11 June 2010

Abstract Burkholderia mallei is a Gram-negative bacillus causing the disease glanders in humans. During intraperitoneal infection, BALB/c mice develop a chronic disease characterised by abscess formation where mice normally die up to 70 days post-infection. Although cytokine responses have been investigated, cellular immune responses to B. mallei infection have not previously been characterised. Therefore, the influx and activation status of splenic neutrophils, macrophages and T cells was examined during infection. Gr-1þ neutrophils and F4/80þ macrophages infiltrated the spleen 5 h post-infection and an increase in activated macrophages, neutrophils and T cells occurred by 24 h post-infection. Mice depleted of Gr-1þ cells were acutely susceptible to B. mallei infection, succumbing to the infection 5 days post-infection. Mice depleted of both CD4 and CD8 T cells did not succumb to the infection until 14 days post-infection. Infected mMT (B cell) and CD28 knockout mice did not differ from wildtype mice whereas iNOS-2 knockout mice began to succumb to the infection 30 days post-infection. The data presented suggests that Gr-1þ cells, activated early in B. mallei infection, are essential for controlling the early, innate response to B. mallei infection and T cells or nitric oxide are important during the later stages of infection. Crown Copyright Ó 2010 Published by Elsevier Masson SAS on behalf of Institut Pasteur. All rights reserved. Keywords: Burkholderia mallei; Neutrophils; T cells; Macrophages

1. Introduction Burkholderia mallei is a Gram-negative intracellular bacillus which causes glanders, a severe disease in humans and experimental animals such as hamsters [1] and mice [2]. Glanders is a re-emerging disease [3] and is endemic in Asia, Africa [4], India [5], Pakistan [6] and Brazil [4]. Although the disease is mainly associated with horses, human cases are reported [7]. It has been eradicated from the Western world and is a notifiable disease in most countries (including the UK). Limited studies have been undertaken to study B. mallei infection, possibly due to the few natural cases now observed

* Corresponding author. Tel.: þ44 (0) 1980 614750; fax: þ 44 (0) 1980 614307. E-mail address: [email protected] (C.A. Rowland).

in the Western world, however, due to its pathogenicity, B. mallei is currently considered a bioterrorism threat [8]. We have recently described that IFN-g is important in controlling B. mallei during the early stages of infection [9], however, the role of individual cell types in the innate response to B. mallei infection is not currently known. Following intraperitoneal challenge, BALB/c mice develop an infection similar to that observed in chronic human glanders. Bacterial colonisation of the spleen occurs within 5 h postinfection [9] and abscesses develop within 2 weeks postinfection which is characteristic of both human [10] and murine disease (unpublished observation). The innate immune response is essential for protection against bacterial infections and several cell types and inflammatory molecules are involved in mediating this response. Neutrophils and macrophages infiltrate several organs during B. mallei infection [2] suggesting that they may play a role in protection against infection. Neutrophils are an essential

1286-4579/$ - see front matter Crown Copyright Ó 2010 Published by Elsevier Masson SAS on behalf of Institut Pasteur. All rights reserved. doi:10.1016/j.micinf.2010.05.012

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

component of the innate response for infection with a number of intracellular pathogens [11,12] including the closely related species Burkholderia pseudomallei [13]. Macrophages are also important in immunity to bacterial infection instigating a number of bacterial killing mechanisms following infection, including the production of nitric oxide by inducible nitric oxide synthase (iNOS)-2 [14] as well as the secretion of proinflammatory cytokines [15]. Bystander T cell responses may also be involved in early responses to infection, although they were not found to be important during B. pseudomallei infection [16]. We describe here the activation status and functional role of several cell types in a mouse model of chronic B. mallei infection. We show that early infiltration and activation of neutrophils and macrophages occurs in the spleen and that Gr-1þ cells are essential for protection against B. mallei infection. We also show that T cells, B cells and iNOS are not important for survival during the early phases of B. mallei infection although T cells and iNOS were found to be important in controlling chronic B. mallei infection. The data presented in this paper provides novel, valuable information on protective cellular host immune responses to B. mallei and will assist in the generation of effective immunotherapies against this pathogen. 2. Materials and methods 2.1. B. mallei B. mallei strain ATCC 23344 was used for all challenge experiments. This strain is a highly pathogenic isolate which can cause severe disease in both humans [10] and mice [2]. All work involving live B. mallei organisms was carried out at Advisory Committee on Dangerous Pathogens (ACDP) containment level 3. Bacterial cells were grown for 48 h in nutrient broth at 37  C. Cells were harvested by centrifugation, and washed three times in PBS, before resuspending in one tenth the original volume of PBS. The number of bacteria present was calculated by spreading 0.25 ml aliquots of diluted culture onto Congo red plates and incubated for 48 h at 37  C. Congo red media was made by mixing Difco heart infusion agar (40 g/l), Congo red (0.1 g/l) and 1000 ml MilliQ water. The media mix was then sterilised (121  C; 15 min), cooled to 60  C and 20 ml/l of 10% w/v D-galactose was added and poured into agar plates. This media supported growth of B. mallei and its red colouration allowed clear visualisation of B. mallei colonies. 2.1.1. Infection of mice with B. mallei All mice were housed in a rigid wall isolator within an ACDP animal containment level 3 facility. All in vivo work was carried out under the Animal Scientific Procedures Act, 1986. Female BALB/c mice aged 6e8 weeks were purchased from Charles River (Margate, UK). Fresh food and water were provided daily and environmental enrichment was present in each cage. Wildtype BALB/c mice and BALB/c mice pretreated with neutralising antibody were challenged i.p. with a calculated dose of 1  106 colony forming units (cfu). Wildtype C57BL/6, mMT (C57BL/6) (B cell knockout (KO)) mice, CD28/ KO

847

(C57BL/6) and inducible nitric oxide synthetase (iNOS)-2 (C57BL/6) KO mice, obtained from London School of Hygiene and Tropical Medicine (LSHTM, UK) were injected i.p. (0.1 ml) with B. mallei (strain ATCC 23344) so that each mouse received a calculated dose of 4.1  104 cfu. 2.1.2. Preparation of spleen cell suspensions Infected mice were killed by cervical dislocation and spleens were removed asceptically. Spleens were dissected into sections and were passed through 70 mm nylon cell strainers into Phosphate Buffered Saline (PBS) (3 ml) in sterile 6 well plates. Spleen suspension (1 ml) was removed from each well and PBS containing 5% Foetal Calf serum (FCS) (Sigma, UK) (1 ml) was then added to each sample. Flow cytometric staining of the spleen suspension (100 ml) was performed immediately using saturating amounts of fluorochrome-labelled antibody to account for potential differences in cell numbers retrieved from infected and uninfected mice. 2.1.3. Bacteriological analysis of organs Splenic cell suspensions (250 ml) were spread onto Congo red plates and incubated (37  C; 48 h). Average bacterial counts for each organ were then calculated. 2.1.4. Flow cytometric staining of splenic cell suspensions ex vivo Aliquots of spleen suspension (100 ml) were stained with the following antibodies purchased from Pharmingen, UK: Gr-1-FITC (RB6-8C5; IgG2bk), F4/80-FITC (IgG2b), CD3FITC (AHIgG1k), CD19-PE, CD54-PE (IgG), IA/IE-PE (rat IgG2ak), CD11b-PE-Cy5 (rat IgG2b) and appropriate isotype control combinations. Samples were incubated in the dark for at least 20 min at room temperature. OptiLyse C (400 ml) was added to samples, mixed gently and incubated for 15 min to allow erythrocyte lysis to occur. Paraformaldehyde was added to samples from control and infected animals at a final concentration of 8%. Samples were refrigerated for at least 24 h prior to flow cytometric analysis to ensure bacterial killing. Samples were analysed using an EPICS XL flow cytometer (Beckman Coulter, UK) and >100,000 events were collected for each individual sample. Data was collected from each individual mouse (n ¼ 8) at each timepoint and cell populations were gated upon using FSC:SSC properties and the lineage cell marker for each cell type. Expression of activation markers was investigated on each cell population. Samples were compensated during acquisition using compensation controls. Following acquisition, data was analysed using Winlist 4.0Ô (Verity Software, Software House, Inc., USA). The positive population was determined using isotype controls and unstained samples appropriate for each stain. 2.2. Antibody-mediated depletion of Gr-1, CD4 and CD8 positive cell populations Female BALB/c mice (n ¼ 6) were injected i.p. with 500 mg of Gr-1 depleting antibodies (clone: RB6-8C5; IgG2b;

848

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

2.5 mg/ml; 200 ml) 24 h prior to infection. Mice were then dosed with antibody every 3 days. An isotype control group (n ¼ 6) was dosed simultaneously with 500 mg of Mac-5 (IgG2b) antibody. T helper cells (CD4þ T cells) and/or cytotoxic T cells (CD8þ T cells) were depleted prior to infection following i.p. injection with an initial loading dose of 500 mg CD4 (YTS 191) and/or CD8 (YTS 169) antibodies or control antibodies (Mac-5) to BALB/c mice (n ¼ 6/group) 3 days prior to infection. A second dose (250 mg) of appropriate antibody was administered 24 h prior to infection and then every 5 days until the end of the experiment. The efficiency of antibody-mediated depletion in the spleen at the time of infection was verified by flow cytometry with >98% Gr-1þ cells depleted with RB6-8C5, >99% depletion of CD4þ T cells with YTS 191 and >99% depletion CD8þ T cells with YTS 169. Mice were culled upon reaching a predefined humane endpoint or 28 days post-infection when the experiment was terminated.

infection ( p < 0.01) and remained elevated 24 h post-infection (Fig. 1). Splenic neutrophils returned to pre-infection levels 3 days post-infection and significantly elevated above control levels again 7 days post-infection (Fig. 1A). Splenic F4/80þ macrophages remained elevated for 7 days post-infection in comparison with uninfected control mice ( p < 0.01) (Fig. 1B). The percentage of CD3þ T cells remained at control levels 5 h, 24 h and 72 h post-infection and then decreased below control levels at days 5 and 7 post-infection ( p < 0.01) (Fig. 1C). The B cell population reduced by almost 50% within 5 h postinfection compared with uninfected mice ( p < 0.01) returning to control levels 24 h post-infection (Fig. 1D). A further reduction in the B cell population occurred 5 days postinfection ( p < 0.01) and then returned to control levels 7 days post-infection (Fig. 1D).

2.3. Statistical analysis

The activation markers CD54 (ICAM-1) and CD11b were investigated on neutrophils during B. mallei infection. Expression of CD54 on Gr-1þ neutrophils was elevated above control levels 5 h and 24 h post-infection ( p < 0.01). CD54 expression returned to control levels 3 days post-infection but increased significantly above controls at days 5 and 7 postinfection ( p < 0.01) (Fig. 2A). A significant increase in CD11b expression ( p < 0.01) occurred on Gr-1þ neutrophils 5 h post-infection (Fig. 2B) returning to levels observed in uninfected mice 24 h post-infection. CD11b expression remained constant until an increase in expression was observed again at day 7 post-infection ( p < 0.01).

Flow cytometry data obtained during infection studies was determined to be non-parametric and, therefore, data is represented as median (n ¼ 8 mice per timepoint)  99% confidence intervals. A KruskaleWallis Test was performed to determine the overall significance of the data and significance of individual data points was determined using a ManneWhitney U Test. The statistical significance of data obtained at each timepoint during the infection was compared with uninfected control animals unless otherwise stated. A predetermined p-value of less than 0.01 was deemed to indicate statistical significance (a ¼ 0.01). A log rank test was performed to compare survival data in depletion experiments and a p-value of less than 0.05 was deemed to indicate statistical significance. 3. Results Mean bacterial counts retrieved from the spleens of BALB/ c mice infected i.p. with 1  106 cfu B. mallei are shown in Fig. 1. Bacterial colonisation for individual mice is described in reference [9]. Briefly, during the first 7 days of infection, BALB/c mice infected i.p. with 1  106 cfu B. mallei did not die as a result of infection. The spleens of all mice were colonised uniformly within 5 h p.i. Variability in bacterial counts occurred from days 3e7 p.i. although the median bacterial burden in the spleen remained between 102 and 103 cfu over the first 7 days of infection.

3.2. Splenic neutrophil activation

3.3. Splenic macrophage activation A significant increase in the percentage of CD54þF4/80þ macrophages occurred 5 h and 24 h post-infection ( p < 0.01) (Fig. 2C). A decrease in CD54þF4/80þ macrophages occurred at day 3 post-infection increasing in number again at days 5 and 7 post-infection. The expression of MHC II was also investigated on macrophages by measuring expression of IA/ IE, a haplotype of MHC II expressed in BALB/c mice (Fig. 2D). The percentage of IA/IEþF4/80þ macrophages increased 5 h post-infection and was higher than control values 24 h post-infection ( p < 0.01). The percentage of IA/IEþF4/ 80þ macrophages decreased to uninfected levels 3 days postinfection but became elevated again 5 days post-infection and remained elevated 7 days post-infection ( p < 0.01).

3.1. Cellular influx to the spleen

3.4. Splenic T cell activation

During the first 7 days of infection the influx of neutrophils, macrophages and T cells into the spleens of infected BALB/c mice was investigated. Neutrophils were identified by their expression of high intensity Gr-1 and macrophages were identified by expression of F4/80. The percentage of neutrophils and macrophages within the spleen increased 5 h post-

T cell activation was measured by cell surface expression of CD25 and CD54. The percentage of CD25þCD3þ T cells decreased significantly 5 h post-infection ( p < 0.01) but returned to control levels at 24 h post-infection (Fig. 3). The CD54þCD3þ T cell population was elevated above control levels 24 h post-infection ( p < 0.01) (Fig. 3).

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

1000

* *

3

100

2 10 1 0 5

24

72

120

40 1000

*

30

100 20 10 10

0

1 0

10000

CD3+

168

1 0

5

* 1000

* *

* 2

100

* 1

10

0

Bacterial colony counts/spleen

% F4/80+ macrophages/total cells

D

10000

F4/80+

1 0

5

24

72

120

168

% CD19+ B cells / lymp ho cytes

4

3

72

120

168

Hours post-infection

Hours post-infection

B

24

70

10000

CD19+

60 1000

50 40

100

*

30

*

20

10

Bacterial co lo ny co u nts/sp leen

5

50

Bacterial colony counts/spleen

*

4

C

10000

Gr-1+

% CD3+ T cells / lymphocytes

6

Bacterial colony counts/spleen

% Gr-1+ neutrophils / total cells

A

849

10 0

1 0

Hours post-infection

5

24

72

120

168

Hours post-infection

Fig. 1. Splenic cell populations and bacterial numbers detected during the first 7 days of B. mallei infection. The median percentage of A) Gr-1Hi splenic neutrophils and B) F4/80þ macrophages in BALB/c mice (n ¼ 8) detected as a proportion of total cells collected per sample and C) CD3þ T cells and D) CD19þ B cells as a proportion of splenic lymphocytes detected in BALB/c mice (n ¼ 8) challenged i.p. with 106 cfu B. mallei. Cell type is indicated by clear bars; diamonds indicate mean bacterial numbers. Error bars indicate 99% confidence intervals for cell numbers and standard error of the mean for bacterial counts. Asterisk(s) indicate statistical significance ( p < 0.01) in comparison with uninfected mice (day 0).

3.5. The functional role of Gr-1, T and B cells, CD28 and iNOS during B. mallei infection Wildtype mice infected i.p. with B. mallei, on both a BALB/c (1  106 cfu) and C57BL/6 (4.1  104 cfu) background, did not die within 40 days following infection. Depletion of Gr-1þ cells pre-infection caused BALB/c mice to become acutely susceptible to B. mallei infection with all mice succumbing to the infection by 5 days post-infection ( p ¼ 0.0025) (Fig. 4A). BALB/c mice depleted of both CD4þ T helper cells and CD8þ cytotoxic T cells prior to infection, were more susceptible to infection than wildtype mice with 4/ 6 mice dying 14 days post-infection and a further animal dying 23 days post-infection ( p ¼ 0.004) (Fig. 4B). Depletion of the CD4þ T cell population caused 3/6 animals to be more susceptible to infection decreasing survival times to 21 days post-infection in comparison with wildtype mice ( p ¼ 0.055). Depletion of the CD8þ T cell population caused 2/6 animals to succumb to the infection by day 14 post-infection although this was not significantly different to wildtype mice ( p ¼ 0.138) (Fig. 4B). Infection of mMT (B cell) KO or CD28/ mice with B. mallei did not significantly affect survival in comparison with wildtype C57BL/6 mice up to 80 days post-infection (Fig. 5). However, iNOS-2 KO mice had

increased susceptibility to B. mallei infection in comparison to wildtype mice ( p ¼ 0.0021) where survival of iNOS-2 KO mice was significantly reduced in comparison to wildtype mice from 30 days post-infection. 4. Discussion Although cytokine responses to B. mallei infection have been previously reported, little information regarding cellular responses exists. In this study, we demonstrated that bacterial colonisation of the spleen 5h following intraperitoneal infection with 1  106 cfu B. mallei is accompanied by a rapid influx of neutrophils and macrophages. This rapid cellular recruitment to the spleen in response to B. mallei infection is supported by previous histological studies identifying neutrophil and macrophage influx to the spleen during B. mallei infection [2]. This cellular recruitment also occurs during infection with other intracellular bacteria including Salmonella typhimurium [17] and Listeria monocytogenes [12] suggesting involvement of neutrophils and macrophages in immunity to B. mallei. The role of neutrophils and macrophages in immunity to B. mallei infection was further supported by identification of rapid cellular activation in the spleen 5 h post-infection. CD54 (ICAM-1) was upregulated on neutrophils 5 h post-infection.

850

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

*

25

90

*

80

*

70 60

*

50

+

20 15

40

+

*

+

+

% CD54 Gr-1 neutrophils

C

*

30

% CD54 F4/80 macrophages

A

10 5

30 20 10 0

0 0

24

48

72

96

120

0

144

24

48

60 50

72

96

120

144

168

Hours post-infection

*

D

*

+

40

+

30 20 10

*

80

*

70 % IA/IE+ F 4/80+ macro phag es

B % CD11b Gr-1 neutrophils

Hours post-infection

60

*

50 40 30 20 10

0

0

0

24

48

72

96

120

144

168

Hours post-infection

0

24

48

72

96

120

144

168

Hours post-infection

Fig. 2. Activation status of splenic Gr-1þ neutrophils and F4/80þ macrophages during the first 7 days of B. mallei infection. Percentage of neutrophils expressing A) CD54; B) CD11b. Percentage of macrophages expressing C) CD54; D) IA/IE (MHC II). Line indicates median and individual points at each timepoint represent individual BALB/c mice (n ¼ 8) challenged i.p. with 106 cfu B. mallei. Asterisks indicate statistical significance ( p < 0.01) in comparison with uninfected mice (day 0).

Increased CD54 expression occurs on neutrophils during inflammation and infection [18] and may be involved in aggregation of neutrophils during inflammatory responses [19] suggesting a potential function of CD54 expression on neutrophils during B. mallei infection. An increase in the proportion of CD11bþ neutrophils was also observed 5 h postinfection, corresponding with neutrophil migration into the tissues, suggesting that CD11b may be involved in neutrophil migration from the blood vessels into the spleen [20]. The role of neutrophils in B. mallei-mediated immunity was further investigated by antibody depletion. The antibody RB6-8C5 (anti-Gr-1) has been used during infection studies to demonstrate the role of neutrophils in protection against infection with intracellular pathogens including Legionella pneumophila [21], Fransicella tularensis [22] and most recently B. pseudomallei [13]. Therefore, these antibodies were used in this study to determine the role of neutrophils during B. mallei infection. An essential role for Gr-1þ cells in the innate response to B. mallei infection was evident, demonstrating that these cells were important in innate immunity to infection. Gr1þ neutrophils comprise the majority of Gr-1þ cells in the spleen (data not shown), however, Gr-1 is also expressed on

a subset of memory CD8þ T cells, Gr-1þ monocytes and plasmacytoid dendritic cells which can also be depleted by RB6-8C5 treatment [23]. Consequently, these Gr-1þ populations may also be involved in the Gr-1-mediated protective response to B. mallei. Depletion of CD8þ T cells in this study did not affect survival during the early phases of B. mallei infection eliminating the role of Gr-1þCD8þ cells in Gr-1mediated immunity. However, the role of other Gr-1þ cell types was not assessed in our experiments. A monoclonal antibody (1A8) which does not deplete non-neutrophil Gr-1þ cells has been recently described [24] and future studies should employ the use of this antibody to further determine the role of neutrophils in B. mallei infection. Macrophage activation and influx to the spleen during infection suggests that macrophages may play a role in B. mallei-mediated immunity. CD54 upregulation on macrophages 5e24 h post-infection demonstrated cellular activation in response to B. mallei. Upregulation of CD54 on macrophages [25] occurs during inflammation and infection and may be involved in infiltration of macrophages into the spleen [26] or in antigen presentation. IFN-g is a potent macrophage activator that specifically initiates gene expression of MHC II

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853 25

CD25

A

CD54

20

% CD3+ T cells

6 5

Nu mber of BALB/c mice

*

15

851

anti-Gr-1 (RB6-8C5) control (Mac 5)

4 3

P=0.0025

2 1

10

0 0

*

5

5

24

Hours post-infection

Fig. 3. Activation status of splenic CD3þ T cells during the first 24 h of B. mallei infection. Percentage of splenic CD3þ T cells expressing CD25 or CD54. Line indicates median and individual points at each timepoint represent individual BALB/c mice (n ¼ 8) challenged i.p. with 106 cfu B. mallei. Asterisks indicate statistical significance ( p < 0.01) in comparison with uninfected mice (0 h).

Number of BALB/c mice

0

2

3

4

5 6 7 8 9 10 11 12 13 14 Days post infection

6

B 0

1

ns

5 4

P=0.004

ns

anti-CD4 anti-CD8

3

anti-CD4+CD8 control

2 1 0 0

5

10

15

20

25

30

Days post-infection

Fig. 4. Survival of BALB/c mice following depletion of Gr-1þ cells and T cell populations during B. mallei infection. A) Gr-1þ cells were depleted in BALB/ c mice (n ¼ 6) 24 h prior to i.p. infection with 106 cfu B. mallei using 500 mg RB6-8C5 specific antibodies or isotype-matched control (Mac-5) antibodies. B) CD4þ T cells (YTS169), CD8þ T cells (YTS 191) or both subsets (YTS169 þ YTS 191) were depleted in BALB/c mice (n ¼ 6) 24 h prior to infection using specific antibodies.

be important in the differentiation of naı¨ve CD4þ T cells into Th1 cells in response to IL-27 [33]. Splenic expression of IL-27 has been shown to be upregulated 24 h following B. mallei infection [9] suggesting a link between CD54 expression and IL27 in the development of type 1 immunity to B. mallei infection. The number of T cells and B cells in the spleen diminished 5e7 days following infection. Lymphoid depletion of the splenic white pulp, composed mainly of T and B cells, from days 3e7 100 90

ns

80

ns

70 % Survival

through signalling via the IFN-g receptor on the macrophage surface [27]. An increase in MHC IIþ macrophages 24 h postinfection corresponds with a peak in IFN-g production during B. mallei infection [9] suggesting a link between IFN-g and MHC II expression. MHC II and CD54 are involved in antigen presentation during bacterial infection [28,29] and an increase in MHC II and CD54 expression on macrophages 5e7 days post-infection suggests that B. mallei infected macrophages may present antigen to stimulate cell-mediated immunity. Nitric oxide is one of a number of mechanisms of macrophage-mediated pathogen killing, therefore, its role in B. mallei infection was investigated. Infection of inducible nitric oxide synthase (iNOS) KO mice revealed that nitric oxide was not important for survival during early B. mallei infection. However, recent in vitro studies have described the importance of iNOS in clearance of B. mallei from infected mouse macrophages [30,31]. We found that although iNOS was not important early in infection, it was important for controlling chronic B. mallei infection. The number of intracellular bacteria within macrophages appears to be important for stimulating NO production [30]. It is possible that in the initial stages of disease macrophage bacterial loads are not high enough to stimulate NO production. This may contrast with the later stages of the disease when bacterial loads are likely to have increased to a level where iNOS is required for further control of infection. In addition to this, other immune mechanisms, e.g. TNF-a, may be able to compensate for the lack of iNOS until the chronic stages of infection. Further work should address other mechanisms of macrophage-mediated killing which may be important in B. mallei infection. Splenic T cell activation (CD54) occurred in the early phase of the response to B. mallei. CD54 is upregulated on T cells during infection with Mycobacterium tuberculosis [32] and may

60

P=0.0021

50 40

P2NOS2

30

µMT CD28KO

20

Wild Type 10 0 0

10

20

30

40

50

60

70

80

90

Days post-infection

Fig. 5. Survival of wildtype C57BL/6 mice (n ¼ 10), iNOS-2 (P2NOS2; C57BL/6) knockout (n ¼ 15), CD28 (C57BL/6) knockout (n ¼ 9) or B cell (C57BL/6) knockout (mMT) mice (n ¼ 13) during i.p. infection with a calculated dose of 4  104 cfu B. mallei.

852

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

post-infection occurs during B. mallei infection of BALB/c mice [2]. It is, therefore, possible that destruction of the splenic architecture is responsible for the decrease in lymphocyte numbers observed in this study. Neither CD4þ T helper cells nor CD8þ cytotoxic T cells alone were important for host control of B. mallei infection and similarly were not involved in vaccinemediated protection during B. mallei infection [34]. However, the importance of both subsets in host control of B. mallei infection is highlighted by the decreased survival time in the absence of both subsets 2e3 weeks following infection. This is similar to that described in a murine model of B. pseudomallei infection [16] although the clinical relevance of this is unclear as HIV infection does not appear to be a risk factor for human melioidosis [35]. CD28, a T cell costimulatory molecule involved in T cell activation, was not essential for T cellmediated immunity 2e3 weeks post-infection as CD28/ KO mice did not have an increased susceptibility to the disease. Generation of antigen-specific CD8þ T cells in CD28/ mice occurred during infection with the intracellular pathogen L. monocytogenes [36] suggesting that adaptive responses may have been formed in the absence of CD28 during B. mallei infection. B cells were also found not to be important during early or chronic B. mallei infection which was also observed in a mouse model of B. pseudomallei infection [16]. However, differences exist in the importance of induction of antibody and B cell responses prior to subsequent B. mallei infection between murine and non-human primate models. B cells are important in mediating protective responses in a murine vaccination model suggesting that vaccine-induced antibody responses could be beneficial in protecting against B. mallei challenge [34]. This is in contrast to effects observed in a non-human primate model of B. mallei infection where monkeys with high titres of antibody were not protected against subsequent challenge with B. mallei [37] suggesting that B cells and antibody-mediated immunity are not essential for control of B. mallei infection. The data presented in this paper demonstrates that the early, innate, cellular response to B. mallei is mediated by Gr-1þ cells and that T cells and iNOS are involved in control of the later stages of the disease. This information builds on previous data showing type 1 response are important in immunity to B. mallei infection providing novel data on protective responses to this pathogen which may be used to develop immunotherapies targeting this pathogen. Acknowledgements The authors would like to thank Bob Gwyther and Tom Laws, Dstl, Porton Down, for performing statistical analysis of data presented in this manuscript. We would also like to gratefully acknowledge the animal technicians at Dstl for their assistance with this study. This research was funded by the UK MoD. References [1] D.L. Fritz, P. Vogel, D.R. Brown, D.M. Waag, The hamster model of intraperitoneal Burkholderia mallei (Glanders), Vet. Pathol. 36 (1999) 276e291.

[2] D.L. Fritz, P. Vogel, D.R. Brown, D. Deshazer, D.M. Waag, Mouse model of sublethal and lethal intraperitoneal glanders (Burkholderia mallei), Vet. Pathol. 37 (2000) 626e636. [3] M.B. Wittig, P. Wohlsein, R.M. Hagen, S. Al Dahouk, H. Tomaso, H.C. Scholz, K. Nikolaou, R. Wernery, U. Wernery, J. Kinne, M. Elschner, H. Neubauer, Glanders e a comprehensive review, Dtsch Tierarztl Wochenschr 113 (9) (2000) 323e330. [4] D.A. Dance, Melioidosis: the tip of the iceberg? Clin. Microb. Rev. 4 (1) (1991) 52e60. [5] R.D. Verma, Glanders in India with special reference to incidence and epidemiology, Indian Vet. J. 58 (1981) 177e183. [6] G. Muhammad, M.Z. Khan, M. Athar, Clinico-microbiological and therapeutic aspects of glanders in Equines, J. Equine. Vet. Sci. 9 (3) (1988) 93e96. [7] S. Arun, H. Neubauer, A. Gurel, G. Ayyildiz, B. Kuscu, T. Yesildere, H. Meyer, W. Hermanns, Equine glanders in Turkey, Vet. Rec. 144 (1999) 255e258. [8] P. Bossi, F. Van Loock, A. Tegnell, G. Gouvras, Bichat clinical guidelines for bioterrorist agents, Euro. Surveill. 9 (12) (2004) E1eE2. [9] C.A. Rowland, G. Lertmemongkolchai, A. Bancroft, A. Haque, M.S. Lever, K.F. Griffin, M.C. Jackson, M. Nelson, A. O’Garra, R. Grencis, G. J. Bancroft, R.A. Lukaszewski, The critical role of Type 1 cytokines in controlling initial infection with Burkholderia mallei, Infect. Immun. 74 (9) (2006) 5333e5340. [10] A. Srinivasan, C.N. Kraus, D. Deshazer, P.M. Becker, J.D. Dick, L. Spacek, J.G. Bartlett, W.R. Byrne, D.L. Thomas, Glanders in a military research microbiologist, N. Engl. J. Med. 345 (4) (2001) 256e258. [11] J.W. Conlan, R.J. North, Neutrophils are essential for early anti-listeria defense in the liver, but not in the spleen or peritoneal cavity, as revealed by a granulocyte-depleting monoclonal antibody, J. Exp. Med. 179 (1) (1994) 259e268. [12] J.W. Conlan, Critical roles of neutrophils in host defense against experimental systemic infections of mice by Listeria monocytogenes, Salmonella typhimurium and Yersinia enterolitica, Infect. Immun. 65 (2) (1997) 630e635. [13] A. Easton, A. Haque, K. Chu, R.A. Lukaszewski, G.J. Bancroft, A Critical role for neutrophils in resistance to experimental infection with Burkholderia pseudomallei, J. Infect. Dis. 195 (2007) 99e107. [14] K.P. Beckerman, H.W. Rogers, J.A. Corbett, R.D. Schreiber, M.L. McDaniel, E.R. Unanue, Release of nitric oxide during the T Cellindependent pathway of macrophage activation, J. Immunol. 150 (3) (1993) 888e895. [15] A. Demuth, W. Goebel, H.U. Beuscher, M. Kuhn, Differential regulation of cytokine and cytokine rceptor mRNA expression upon infection of bone marrow-derived macrophages with Listeria monocytogenes, Infect. Immun. 64 (9) (1996) 3475e3483. [16] A. Haque, A. Easton, D. Smith, A. O’Garra, N. Van Rooijen, G. Lertmemongkolchai, R.W. Titball, G.J. Bancroft, The role of T cells in innate and adaptive immunity against murine Burkholderia Pseudomallei, J. Infect. Dis. 193 (2006) 370e379. [17] A.C. Kirby, U. Yrlid, M.J. Wick, The innate immune response differs in primary and secondary salmonella Infection, J. Immunol. 169 (2002) 4450e4459. [18] J.H. Wang, D.M. Sexton, H.P. Redmond, R.W. Watson, D.T. Croke, D. Bouchier-Hayes, Intercellular adhesion molecule-1 (ICAM-1) is expressed on human neutrophils and is essential for neutrophil adherence and aggregation, Shock 8 (5) (1997) 357e361. [19] S.Z. Wang, P.K. Smith, M. Lovejoy, J.J. Bowden, J.H. Alpers, K.D. Forsyth, Shedding of L-selectin and PECAM-1 and upregulation of Mac1 and ICAM-1 on neutrophils in RSV bronchiolitis, Am. J. Physiol. 275 (1998) L983eL989. [20] W.W. Agace, M. Patarroyo, M. Svensson, E. Carlemalm, C. Svanborg, Esherichia coli induces transuroepithelial neutrophil migration by an intercellular adhesion molecule-1-dependent mechanism, Infect. Immun. 63 (10) (1995) 4054e4062. [21] K. Tateda, T.A. Moore, J.C. Deng, M.W. Newstead, X. Zeng, A. Matsukawa, M.S. Swanson, K. Yamaguchi, T.J. Standiford, Early recruitment of neutrophils determines subsequent T1/T2 host responses in a murine

C.A. Rowland et al. / Microbes and Infection 12 (2010) 846e853

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

model of Legionella pneumophila pneumonia, J. Immunol. 166 (2001) 3355e3361. A. Sjostedt, R.J. North, J.W. Conlan, The requirement of tumour necrosis factor-a and interferon-g for the expression of protective immunity to secondary murine tularaemia depends on the size of the challenge inoculum, Microbiology 142 (1996) 1369e1374. J. Matsuzaki, T. Tsuji, K. Chamoto, T. Takeshima, F. Sendo, T. Nishimura, Successful elimination of memory-type CD8þ T cell subsets by the administration of anti-Gr-1 monoclonal antibody in vivo, Cell. Immunol. 224 (2) (2003) 98e105. J.M. Daley, A.A. Thomay, M.D. Connolly, J.S. Reichner, J.E. Albina, Use of Ly6G-specific monoclonal antibody to deplete neutrophils in mice, J. Leukoc. Biol. 83 (2008) 64e70. S. Ghosh, R.K. Saxena, Early effect of Mycobacterium tuberculosis infection on Mac-1 and ICAM-1 expression on mouse peritoneal macrophages, Exp. Mol. Med. 36 (5) (2004) 387e395. F.C. Gibson III, A.B. Onderdonk, D.L. Kasper, A.O. Tzianabos, Cellular mechanism of intraabdominal absecess formation by Bacteroides fragilis, J. Immunol. 160 (1998) 5000e5006. K. Schroder, P.J. Hertzog, T. Ravasi, D.A. Hume, Interferon-g: an overview of signals, mechanisms and functions, J. Leukoc. Biol. 75 (2004) 163e189. T. Lebedeva, M.L. Dustin, Y. Sykulev, ICAM-1 co-stimulates target cells to facilitate antigen presentation, Curr. Opin. Immunol. 17 (2005) 251e258. A. Neild, T. Murata, C.R. Roy, Processing and major histocompatibility complex class II presentation of Legionella pneumophila antigens by infected macrophages, Infect. Immun. 73 (4) (2005) 2336e2343.

853

[30] P.J. Brett, M.N. Burtnick, H. Su, V. Nair, F.C. Gherardini, iNOS activity is critical for clearance of Burkholderia mallei from infected RAW 264.7 murine macrophages, Cell Microbiol. 10 (2) (2008) 487e498. [31] J. Jones-Carson, J. Laughlin, M.A. Hamad, A.L. Stewart, M.I. Voskuil, A. Vazquez-Torres, Inactivation of [FeeS] metalloproteins mediates nitric oxide-dependent killing of Burkholderia mallei, PLoS One 3 (4) (2008) e1976. [32] D.K. Mitra, S.K. Sharma, A.K. Dinda, M.S. Bindra, B. Madan, B. Ghosh, Polarized helper T cells in tubercular pleural effusion: phenotypic identity and selective recruitment, Eur. J. Immunol. 35 (8) (2005) 2367e2375. [33] T. Owaki, M. Asakawa, N. Morishima, K. Hata, F. Fukai, M. Matsui, J. Mizuguchi, T. Yoshimoto, A role for IL-27 in early regulation of Th1 differentiation, J. Immunol. 175 (4) (2005) 2191e2200. [34] G.C. Whitlock, R.A. Lukaszewski, B.M. Judy, S. Paessler, A.G. Torres, D.M. Estes, Host immunity in the protective response to vaccination with heat-killed Burkholderia mallei, BMC Immunol. 9 (2008) 55. [35] W. Chierakul, A. Rajanuwong, V. Wuthiekanun, N. Teerawattanasook, M. Gasiprong, A. Simpson, W. Chaowagul, N.J. White, The changing pattern of bloodstream infections associated with the rise in HIV prevalence in northeastern Thailand, T Roy Soc. Trop. Med. Hyg. 98 (2004) 678e686. [36] W. H-Mittrucker, M. Kursar, A. Kohler, R. Hurwitz, S.H.E. Kaufmann, Role of CD28 for the generation and expansion of antigen-specific CD8þ T lymphocytes during infection with Listeria monocytogenes, J. Immunol. 167 (2001) 5620e5627. [37] I.N. Manzeniuk, I.N. Khomiakov, G.M. Titareva, E.A. Ganina, A. Buziun, D.V. Naumov, E.A. Svetoch, Homeostatic changes in monkeys in a model of Glanders, Antibiot. Khimioter. 42 (5) (1997) 29e34.