INFECTION AND IMMUNITY, Mar. 2005, p. 1744–1753 0019-9567/05/$08.00⫹0 doi:10.1128/IAI.73.3.1744–1753.2005 Copyright © 2005, American Society for Microbiology. All Rights Reserved.
Vol. 73, No. 3
Neutrophil Responses to Mycobacterium tuberculosis Infection in Genetically Susceptible and Resistant Mice Evgenyi B. Eruslanov, Irina V. Lyadova, Tatiana K. Kondratieva, Konstantin B. Majorov, Ilya V. Scheglov, Marianna O. Orlova, and Alexander S. Apt* Laboratory for Immunogenetics, Central Institute for Tuberculosis, Moscow, Russia Received 1 June 2004/Returned for modification 17 August 2004/Accepted 27 September 2004
The role of neutrophils in tuberculosis (TB) resistance and pathology is poorly understood. Neutrophil reactions are meant to target the offending pathogen but may lead to destruction of the host lung tissue, making the defending cells an enemy. Here, we show that mice of the I/St strain which are genetically susceptible to TB show an unusually high and prolonged neutrophil accumulation in their lungs after intratracheal infection. Compared to neutrophils from more resistant A/Sn mice, I/St neutrophils display an increased mobility and tissue influx, prolonged lifespan, low expression of the CD95 (Fas) apoptotic receptor, relative resistance to apoptosis, and an increased phagocytic capacity for mycobacteria. Segregation genetic analysis in (I/St ⴛ A/Sn)F2 hybrids indicates that the alleles of I/St origin at the chromosome 3 and 17 quantitative trait loci which are involved in the control of TB severity also determine a high level of neutrophil influx. These features, along with the poor ability of neutrophils to restrict mycobacterial growth compared to that of lung macrophages, indicate that the prevalence of neutrophils in TB inflammation contributes to the development of pathology, rather than protection of the host, and that neutrophils may play the role of a “Trojan horse” for mycobacteria. PMNs have been implicated in the control of mycobacterial infections (14, 32), but it is not known whether these cells have direct protective functions. Initial in vitro studies suggested that human neutrophils are able to kill virulent Mycobacterium tuberculosis (7, 17); however, more-recent reports did not confirm these results (9). The recruitment of PMNs to the lung has been described for acute TB (20, 41) and in experimental animals infected with mycobacteria (1, 21). Human and animal studies indicate that neutrophils, while not very helpful in mycobacterial clearance themselves (18, 42), may play an important role in the transition from innate to adaptive immune responses by producing critical cytokines and chemokines (8, 33, 39). Although this immunomodulatory function has been demonstrated for many intracellular infections, including TB (5, 34, 46), the pleotropic nature of mediators leaves unresolved the exact contribution of neutrophils to TB resistance and pathogenesis. One approach to close this gap in our knowledge was to study neutrophil reactions in strains of mice with genetically determined differences in the severity of TB, accompanied by distinct patterns of lung inflammation and pathology. Earlier, we reported that mice of the I/St and A/Sn strains differ substantially in their susceptibility to M. tuberculosis infection, the degree of lung pathology, and the dynamics of macrophage and T-cell accumulation in the lungs (11, 25, 30). This complex phenotype is under polygenic, Nramp1-independent control, which involves quantitative trait loci (QTLs) located on chromosomes 3, 9, 17, and X (22, 37). In the present work, we assessed in vivo and in vitro the capacity of neutrophils from I/St and A/Sn mice to migrate in response to mycobacteria, to engulf and kill the bacilli, to undergo apoptosis, and to produce immune response mediators. Our results suggest that, in highly susceptible I/St mice, neutrophils play the role of a “Trojan horse” for mycobacteria and contribute to the development of severe lung inflammation rather than protection of the host.
The hallmark of pulmonary tuberculosis (TB) is recruitment to the lung tissue and activation of inflammatory cells, resulting in granuloma formation (40). This response restricts mycobacterial dissemination and provides defense of the lung against TB. However, it can lead to extensive tissue damage, resulting both in the reduced lung function and in the aerosol transmission of mycobacteria released from disintegrating granulomas (36). An inflammatory tuberculous lesion consists of a central area of mycobacterium-infected macrophages surrounded by noninfected phagocytes and lymphocytes found at the periphery (19). Whereas the contributions of macrophages and lymphocytes to TB resistance, including effector functions, antigen presentation, and cytokine production, have been well characterized (12, 31, 44), the role of polymorphonuclear cells (PMNs) has received little attention. In infectious inflammation, PMNs (principally neutrophils) are the first phagocytes to arrive from the circulation and attempt to eliminate invading pathogens via oxygen-dependent and oxygen-independent mechanisms. The former mechanism results from the generation of reactive oxygen species (27), whereas the latter mechanism reflects the capacity of PMNs to degranulate and release preformed oxidants and proteolytic enzymes from granules (3, 48). Neutrophil degranulation is meant to target the offending pathogen; however, these powerful substances may cause the destruction of neighboring cells and dissolution of tissue (13, 23). This neutrophil-dependent tissue damage is known as the neutrophil paradox, where the defending cells become an enemy (47). Thus, a strict regulation of neutrophil influx and their turnover in infected tissues is essential for minimizing tissue damage.
* Corresponding author. Mailing address: Laboratory for Immunogenetics, Central Institute for Tuberculosis, Yauza Alley 2, Moscow 107564, Russia. Phone: (7095) 268 78 10. Fax: (7095) 963 80 00. E-mail:
[email protected]. 1744
VOL. 73, 2005
NEUTROPHIL RESPONSE TO MYCOBACTERIA AND TB SEVERITY MATERIALS AND METHODS
Animals. Inbred mice of the I/StYCit (I/St) and A/JSnYCit (A/Sn) strains and their (A/Sn ⫻ I/St)F2 hybrids were bred under conventional conditions at the Animal Facilities of the Central Institute for Tuberculosis (Moscow, Russia) in accordance with guidelines from the Russian Ministry of Health (no. 755) and National Institutes of Health Office of Laboratory Animal Welfare (assurance no. A5502-01). Water and food were provided ad libitum. Male mice 2 to 4 months of age were used. All experimental procedures were approved by the institutional animal care committee. Bacteria, infection, and neutrophil counts. To induce experimental TB, male mice 2 to 4 months of age were infected intratracheally (i.t.) with 103 CFU of M. tuberculosis strain H37Rv/mouse. Establishment and storage of mycobacteria (30), as well as the protocol for i.t. challenge (11), were as previously described. To study acute neutrophil influx into the peritoneal cavities and lungs, mice were injected intraperitoneally (i.p.) or i.t. with 108 H37Rv CFU/mouse and sacrificed 2 to 3 h postchallenge. Peritoneal cavities were washed with phosphate-buffered saline (PBS)–EDTA, and lung cell suspensions were obtained by enzymatic disruption (25). In Giemsa-stained cytospin preparations, the neutrophil content was estimated by counting neutrophils in at least 40 microscopic fields per spot (⬎300 cells). Purification of peritoneal neutrophils. Neutrophils were isolated from peritonal cavities 18 h postinjection of 2 ml of 3% peptone in saline. Peritoneal exudates were collected, and cells were washed and separated on a two-step Percoll gradient (1.073- and 1.083-g/ml layers) by centrifugation at 450 ⫻ g and 20°C. The neutrophil-rich cell bands were collected, washed in PBS, stained with Giemsa stain, and examined microscopically. Neutrophil purity was consistently ⬎95%, with a viability of ⬎99% by trypan blue exclusion. Cytological examination of lung prints. The right lungs from infected mice were cut, pressed once against filter paper, and serially printed onto glass microscope slides. Prints were dried, fixed with 96% ethanol, and stained for the presence of acid-fast bacilli by using the Ziehl-Neelsen method with Giemsa counterstaining. Mycobacterial phagocytosis in vivo and in vitro. Mice were injected i.p. with 3% peptone followed 18 h later by either 108 mycobacterial CFU or 108 zymosan particles (Sigma, St. Louis, Mo.) per mouse and then sacrificed 2.5 h postchallenge. The peritoneal cavity cells were washed three times to remove extracellular mycobacteria and particles. Individual cytospin preparations were stained exactly as described above for lung prints and evaluated microscopically. These procedures resulted in ⬎95% pure neutrophil preparations containing exclusively intracellular mycobacteria (Fig. 4). For the assessment of mycobacterial phagocytosis in vitro, peritoneal neutrophils from noninfected mice (3 ⫻ 105 cells/well of a flat-bottomed 96-well plate) were incubated at 37°C with M. tuberculosis at a multiplicity of infection (MOI) of 10:1 in antibiotic-free, supplemented Dulbecco’s modified Eagle medium. To evaluate the role of opsonization, 10% of either freshly isolated or heat-inactivated (HI) autologous mouse serum or HI fetal calf serum (FCS) was added. After a 2-h incubation, neutrophils were detached from the plastic with PBSEDTA and washed to remove extracellular mycobacteria. The proportion of infected cells was determined by a microsopic evaluation of ⬎200 ZiehlNeelsen–Giemsa-stained neutrophils in cytospin preparations. Transmigration assay. Chemotaxis was evaluated in 24-well transwell plates with 3-m-pore-size filters (Costar-Corning, Badhoevedorp, The Netherlands). Lower chambers were filled with either culture medium (RPMI 1640) alone, medium containing 20 g of M. tuberculosis H37Rv sonicate/ml, noninfected interstitial lung macrophages (3.5 ⫻ 105 cells/well), or lung macrophages infected with mycobacteria at an MOI of 10:1. Isolation and infection of interstitial lung macrophages were done as previously described (26). A total of 106 peritoneal exudate cells in 100 l of medium were added to the upper chambers and allowed to migrate through the membrane for 1 h at 37°C. The transmigrated cells were collected totally from the lower chamber, stained with anti-Ly-6G⫹ monoclonal antibodies (MAbs) (clone RB6-8C5; Caltag Laboratories, Burlingame, Calif.), and evaluated using FACSCalibur (Becton Dickinson, San Jose, Calif.). Fluorescence-activated cell sorter analysis. The following directly conjugated MAbs were used in different combinations following blocking of Fc receptors with anti-CD16/CD32 MAbs: PE–anti-CD62L, PE–anti-CD18, PE–anti-CD11b, PE–anti-CD95, and PE–anti-CD95L. Apoptosis was evaluated by Annexin Vpropidium iodide staining. All fluorochrome-labeled reagents were purchased from BD-PharMingen (San Diego, Calif.). Cells were analyzed by flow cytometry using MultiGraph and FlowJo software. Soluble mediators. Lung cells (2 ⫻ 106/ml) were cultured in the presence or absence of 10 g of H37Rv sonicate/ml. Interleukin-6 (IL-6) and MIP-2 were
1745
measured in 48-h culture supernatants using enzyme-linked immunosorbent assay kits from PharMingen. To measure active transforming growth factor 1 (TGF-1), lung cells were cultured in the medium containing 1% mouse serum since FCS contains endogenous TGF-1. Samples were analyzed using a TGF-1 Emax immunoassay (Promega, Madison, Wis.). Antimycobacterial activity of neutrophils. The growth of mycobacteria in neutrophils in vitro was estimated using [3H]uracil uptake (26). Neutrophils were recovered from peritoneal cavities and plated at 3 ⫻ 105 cells/well in supplemented RPMI 1640 medium containing 10% fresh autologous mouse serum. Cells were allowed to adhere for 1 h, and mycobacteria were added at different MOIs. Mycobacteria cultured without neutrophils and neutrophils cultured in the absence of mycobacteria served as controls. Cultures were pulsed with 1 Ci of [3H]uracil/well for the last 18 h of 24-h incubation, and [3H]uracil uptake was measured in a liquid scintillation counter (Wallac, Turku, Finland). Genotyping of F2 mice. Genotypes of the SSLPs D3Mit215, D9Mit89, and D17Mit175 were determined in (I/St ⫻ A/Sn)F2 mice by using PCR performed with isolated tail DNA as previously described (22, 37). Statistical analysis. Significance of the differences was estimated using an unpaired Student’s t test and the Mann-Whitney U test. Data were expressed as means ⫾ standard errors of the means; P ⬍ 0.05 was considered statistically significant.
RESULTS Neutrophil inflammation, recruitment, and mobility in TBsusceptible and -resistant mice. We first determined whether or not susceptible I/St mice differ from resistant A/Sn mice regarding the dynamics of lung neutrophil infiltration after infection. A pronounced accumulation of Ly-6G-positive cells (neutrophils) was observed exclusively in the lungs of I/St mice starting at week 1 and peaking between weeks 2 and 3 postinfection (Fig. 1a and b). Microscopic examination of the lung prints at week 3 postinfection demonstrated that the proportions of cells with typical polysegmented neutrophil nuclei were 19.2% ⫾ 2.0% and 3.7% ⫾ 0.5% (P ⬍ 0.008) in I/St and A/Sn mice, respectively. Ziehl-Neelsen staining revealed numerous mycobacteria inside I/St neutrophils (Fig. 1c), whereas no bacilli were detected in A/Sn neutrophils (results not shown). We asked whether the difference in neutrophil accumulation in the lungs between the two strains of mice was due to a different capacity to migrate to the site of M. tuberculosis infection. First, we challenged I/St and A/Sn mice i.t. or i.p. with live M. tuberculosis and estimated the proportion of Ly-6G⫹ cells in lung and peritoneal exudate cells. The proportion of neutrophils recruited to the two different sites of infection was significantly (P ⬍ 0.01) higher in I/St mice than that in A/Sn mice (Fig. 1d). There was no detectable neutrophil influx in PBS-injected control mice. We next compared the migration capacity of I/St and A/Sn neutrophils by an in vitro transwell assay. Peritoneal exudate cells recovered from both I/St and A/Sn mice at day 5 after peptone injection contained ⬃60% PMNs and ⬃40% mononuclear cells (Fig. 1e). These cells were allowed to migrate toward mycobacterium-infected lung macrophages, mycobacteria alone, mycobacterial sonicate, or medium alone. Irrespective of mouse strain, only PMNs but not monocytes migrated to the bottom chamber (Fig. 1f), and the number of migrating cells was ⬃6-fold higher (P ⬍ 0.01) in I/St cultures than in A/Sn cultures (Fig. 1g and h). Interestingly, the number of migrating neutrophils did not depend upon the presence of an infectious stimulus (Fig. 1h), suggesting that I/St neutrophils have an intrinsically higher migration activity. We also evaluated whether activation following migration to extravascular sites is impaired in A/Sn neutrophils. CD62L
1746
ERUSLANOV ET AL.
INFECT. IMMUN.
FIG. 1. Distinct patterns of neutrophil inflammation following TB challenge and neutrophil migration in I/St and A/Sn mice. Accumulation of Lyt-6G-positive cells following infection, as enumerated by fluorescence-activated cell sorter (a and b), and the presence of neutrophils containing mycobacteria in I/St lungs (c) are shown. Neutrophil influx to the lungs and peritoneal cavities shortly after a high-dose (108 CFU/mouse) mycobacterial challenge was significantly higher in I/St mice than in A/Sn mice (d). In a transmigration assay of peptone-attracted peritoneal cavity cells (e), neutrophils selectively migrated through a microporous filter in response to mycobacterial sonicate (f and g). There was a sixfold interstrain difference in the migration capacity of neutrophils (g and h), which did not depend upon migration stimulus (h). A P of ⬍0.01 between I/St and A/Sn mice in panels a, b, d, g, and h is given for three to six independent experiments (total n ⫽ 12 to 24), and error bars represent the standard deviations from a Student’s t test.
(L-selectin), the molecule mediating neutrophil-endothelial interaction, is down-regulated after neutrophil tissue homing and activation (6). In contrast, the expression of CD11b, an ␣M subunit of the Mac-1 integrin, increases upon neutrophil activation (29). We compared the expression of these molecules in blood and lung neutrophils at week 3 post-TB challenge. Both A/Sn and I/St neutrophils lost their
resting CD62LhiCD11blo phenotype and acquired an activated CD62LloCD11bhi surface phenotype after leaving circulation (Fig. 2a and b). Interestingly, the second 2 (CD18) subunit of Mac-1 was almost identically expressed in blood and lung neutrophils (Fig. 2c), indicating that CD11b-specific rather than broader anti-Mac-1 antibodies should be used to assess cell activation.
VOL. 73, 2005
NEUTROPHIL RESPONSE TO MYCOBACTERIA AND TB SEVERITY
FIG. 2. I/St and A/Sn neutrophils acquired the identical activation phenotype during extravasation at week 3 postinfection. Down-regulation of CD62L (a) and up-regulation of CD11b (b) in the Lyt-6G⫹ cells after their transition from blood to the site of inflammation is shown. The expression of CD18 remained stable (c). The results of one out of two similar experiments performed with a mixture of cells obtained from three individual mice and gated for Lyt-6G are displayed.
I/St neutrophils do not express CD95, are resistant to apoptosis, and have a longer lifespan. Although extravascular neutrophils live longer than circulating neutrophils, they still are short-lived cells and their lifespan is an important factor which determines the size of the neutrophil population at the inflamed site. Fas-mediated interactions play a fundamental role in the regulation of neutrophil apoptosis (24), and differences in Fas (CD95) expression may explain the difference in neutrophil survival. To address this issue, we compared the levels of Fas and Fas-ligand expression on circulating and lung neutrophils in I/St and A/Sn mice before and after TB infection. In naive I/St mice, neither blood nor lung neutrophils expressed CD95 and the proportion of CD95⫹ neutrophils increased only slightly 3 weeks postchallenge (Fig. 3A). In naive A/Sn mice, the proportion of CD95⫹ cells among circulating neutrophils was also low; however, in the lungs, it exceeded 20%. Moreover, after TB challenge, the proportion of CD95⫹ neutrophils significantly increased in A/Sn mice, reaching ⬃75% in circulation and ⬃40% in the lung tissue (Fig. 3A). Neither blood nor lung neutrophils expressed CD95L. We then compared I/St and A/Sn neutrophil apoptotic death and survival time. Given that both peritoneal and lung neutrophils display an activated, tissue-associated phenotype, we extracted neutrophils from peritoneal cavities after 18-h peptone stimulation, incubated them in vitro, and examined the proportion of apoptotic cells. As early as 8 h into culturing, a significantly (P ⬍ 0.01) high proportion of A/Sn neutrophils expressed the Annexin V⫹ propidium iodide⫹ late-apoptoticstage phenotype compared to that for I/St neutrophils (Fig. 3B). After 20 h of culture, ⬃70% of I/St neutrophils looked alive whereas no live A/Sn neutrophils remained in cultures (Fig. 3C; confirmed also by trypan blue exclusion). Thus, there was a substantial difference in the lifespans of I/St and A/Sn neutrophils. Taken together, these results suggest that I/St neutrophils live longer and are more resistant to CD95-mediated apopto-
1747
sis. Their A/Sn counterparts have a shorter lifespan and readily acquire the CD95-positive phenotype following TB challenge. Interstrain differences in neutrophil mycobacterial phagocytosis. To establish the possible role of neutrophils in TB defense and/or pathology, we examined whether neutrophils from the two mouse strains differ in their ability to ingest mycobacteria. Neutrophils were attracted to peritoneal cavities by 18-h peptone stimulus, mycobacteria were injected at a high dose, and cytospin cell preparations were obtained 2.5 h postchallenge and examined microscopically. Neutrophil infiltration was substantially more massive, and mycobacterial phagocytosis was significantly higher in I/St mice (Fig. 4a). These differences were nonspecific and based solely on genetic differences between mouse strains, as analogous results were obtained with zymosan particles (Fig. 4b). About half of I/St neutrophils contained intracellular mycobacteria (Fig. 4c), whereas A/Sn neutrophils contained only a few bacilli (Fig. 4d). To examine whether the capacity of neutrophils to ingest mycobacteria and/or whether interstrain differences in phagocytosis efficacy depend upon opsonization, we assessed phagocytosis in cultures supplemented with HI FCS or normal mouse serum, as well as autologous or heterologous (reciprocal) fresh mouse serum. I/St and A/Sn neutrophils ingested mycobacteria much more efficiently in the presence of fresh serum than those with HI serum, with inactivated FCS providing almost no opsonization (Fig. 4e and f). The efficacy of phagocytosis in vitro was similar in the presence of I/St and A/Sn sera and in cultures supplemented with autologous and heterologous sera (data not shown), indicating that interstrain differences in opsonin-dependent mycobacterial phagocytosis are not due to the lack of opsonins in the A/Sn serum. Capacity of neutrophils to inhibit mycobacterial growth is low. The ability of neutrophils to kill virulent M. tuberculosis remains controversial (7, 9, 17). The presence of numerous neutrophils containing mycobacteria in the lungs of I/St mice may reflect a compensatory line of defense to replace other mechanisms or a convenient niche for mycobacterial survival in the face of an inflammatory response. Thus, it was important to find out whether neutrophils are capable of inhibiting mycobacterial growth and whether I/St and A/Sn mice differ in this regard. To this end, we used a surrogate [3H]uracil assay which provided an accurate evaluation of mycobacterial growth in phagocytic cells (26). Infection of neutrophils with mycobacteria at various MOIs showed that neutrophils from two strains of mice display equally low levels of bacteriostatic activity (Fig. 5a). In A/Sn mice versus in I/St mice, the capacity of neutrophils to inhibit mycobacterial growth was significantly lower than that of infected lung macrophages (Fig. 5b) (see also reference 26), and, by contrast to the latter, did not increase after adding exogenous gamma interferon (IFN-␥) (Fig. 5a and b). No detectable nitric oxide production by mycobacterium-infected neutrophils was observed. Chemokine and cytokine production by lung cells. Neutrophil migration from the bloodstream depends upon a complex network of chemokines and cytokines with a high degree of functional redundancy, so quantitative differences in the production of some of these mediators by lung cells may or may not underlie interstrain differences in neutrophil inflammation.
1748
ERUSLANOV ET AL.
INFECT. IMMUN.
FIG. 3. I/St neutrophils are more resistant to apoptosis and have longer lifespan than A/Sn neutrophils. (A) I/St blood and lung neutrophils remained CD95 negative at week 3 following TB challenge, whereas a substantial proportion of A/Sn neutrophils acquired the CD95⫹ phenotype. (B) The proportion of live Annexin V⫺ propidium iodide⫺ cells was significantly (P ⬍ 0.05) higher, whereas the proportion of late-apoptotic-stage Annexin V⫹ propidium iodide⫹ cells was lower (P ⬍ 0.01) in I/St neutrophil cultures than in A/Sn neutrophil cultures after 8-h incubation. FITC, fluorescein isothiocyanate. (C) Microscopic appearance of live I/St and dead A/Sn neutrophils in 20-h cultures is shown. The results of one out of three similar experiments are displayed.
Nevertheless, we estimated in vitro the level of a few key mediators involved in the regulation of neutrophil influx. The production of a key murine neutrophil attractant, MIP-2, by unseparated lung cells from I/St and A/Sn mice before and after TB challenge was similarly high (results not shown).
However, irrespective of TB challenge, I/St lung cells produced significantly more IL-6 (Fig. 5c), a proinflammatory cytokine capable of affecting the deformability of neutrophils and promoting their sequestration in the lungs (45). In addition, I/St lung cells extracted before or shortly after infection produced
VOL. 73, 2005
NEUTROPHIL RESPONSE TO MYCOBACTERIA AND TB SEVERITY
1749
FIG. 4. Phagocytic capacity of I/St neutrophils is significantly higher that that of A/Sn neutrophils. Following injection of peptone with mycobacteria (a, c, and d) or peptone with zymosan (b), neutrophil yield, percent phagocytosis, and phagocytic numbers were significantly higher (P ⬍ 0.01, mean ⫾ standard error of the mean from four independent experiments, three mice each, Mann-Whitney U test) in peritoneal cavities of I/St mice than in those in A/Sn mice. Mycobacterial phagocytosis is opsonin dependent, and the serum from A/Sn mice has unimpaired opsonizing properties (e and f). hi, heat inactivated; NMS, normal mouse serum.
1750
ERUSLANOV ET AL.
INFECT. IMMUN.
FIG. 5. Inhibition of mycobacterial growth by I/St and A/Sn neutrophils is not elevated in the presence of exogenous IFN-␥ (a) in contrast to the inhibitory activity of lung macrophages (b). Lung cells from infected I/St mice produce more IL-6 (c) and less active TGF-1 (d) in response to mycobacterial sonicate in vitro than their A/Sn counterparts. The results represent means ⫾ standard deviations from two independent experiments, three mice per point each, total n ⫽ 6, assessed in enzyme-linked immunosorbent assay format individually. *, P ⬍ 0.05 by the Mann-Whitney U test.
⬃2 times less active TGF-1 (Fig. 5d), a mediator capable of inhibiting endothelial adhesiveness and transmigration of blood neutrophils (16, 43). Thus, interstrain differences in the level of neutrophil inflammation may be due, at least in part, to genetic variability in the level of immune mediators in the lungs of mice. Neutrophil inflammation and TB severity are controlled by the same QTLs. Earlier, we reported that, in the I/St-A/Sn strain combination, several epistatically interacting QTLs located on chromosomes 3, 9, 17, and X control the postinfection body weight loss and time to death (22, 37). It was important to find out whether or not the degree of the lung neutrophil inflammation is also controlled by this genetic network. We genotyped a set of segregating (I/St ⫻ A/Sn)F2 mice for the alleles of I/St (i) and A/Sn (a) origins at the microsatellite markers D3Mit215, D9Mit89, and D17Mit175 (linkage peaks with the polar TB severity phenotypes), infected these animals with TB, and individually estimated the percentage of Ly-6Gpositive cells in their lungs at week 3 postinfection. The proportion of lung neutrophils varied substantially between F2 mice, suggesting polygenic control of this trait (Fig. 6A). Following genotypic stratification, it appeared that the recessive i allele at the D3Mit215 locus and the dominant i allele at the D17Mit175 locus determine a high level of neutrophil inflammation; the proportion of neutrophils was mark-
edly higher in D3Mit215i/i (Fig. 6B) and in D17Mit175i/⬃ (Fig. 6C) mice. Accumulation of neutrophils in the lungs following TB challenge was significantly higher in mice bearing the combined D3Mit215i/iD17Mit175i/⬃ genotype than in animals bearing all other allelic combinations (Fig. 6D). A formal genetic confirmation that i alleles at the loci which control TB severity also determine a high level of neutrophil influx strongly suggests that neutrophil inflammation indeed underlies an unfavorable anti-TB response. DISCUSSION The importance of interactions between neutrophils and mycobacteria is highlighted by the fact that, very early after infection via the respiratory route, neutrophils are recruited to the lungs. Nevertheless, animal models have failed to demonstrate a clear mechanism whereby neutrophils contribute to defense against mycobacteria, due primarily to conflicting data (2, 14, 32). The use of TB-resistant A/Sn and TB-susceptible I/St mouse strains, which represent polar extremes of the susceptibility spectrum, has facilitated the study of functional and genetic aspects of neutrophil responses to mycobacteria. In the present study, we show for the first time that accumulation of neutrophils in the lungs shortly after infection is characteristic of susceptible but not resistant mice (Fig. 1).
VOL. 73, 2005
NEUTROPHIL RESPONSE TO MYCOBACTERIA AND TB SEVERITY
1751
FIG. 6. Neutrophil influx to the lungs following TB challenge (3 weeks postinfection) is controlled by the QTLs involved in TB severity control. Variation in the neutrophil influx between F2 mice indicates polygenic control (A). i alleles were of I/St origin; a alleles were of A/Sn origin. The recessive i allele of the D3Mit215 QTL (B) and the dominant i allele of the D17Mit175 QTL (C) determine a high level of neutrophil inflammation. In mice bearing their combination, accumulation of neutrophils in the lungs was significantly (P ⬍ 0.05, Mann-Whitney U test) higher than that in mice with other genotypes (D).
Furthermore, tissue neutrophils from susceptible mice have a significantly higher migration capacity in vivo and in vitro (Fig. 1), survive longer (Fig. 3), and contain significantly more intracellular mycobacteria (Fig. 1 and 4). Importantly, neutrophils display a weak antimycobacterial activity and, unlike lung macrophages, are resistant to activation by IFN-␥ (Fig. 5). Given that, in our TB model, lung cells from I/St mice produce significant amounts of IFN-␥ (⬎40 ng/ml [11]) up to the week 3 post-TB challenge (i.e., around the neutrophil peak [Fig. 1a and b]), we suggest that, in the infected I/St mice, an abundant and relatively long-living lung neutrophil population may promote mycobacterial survival by temporarily “hiding” bacilli from ingestion by activated macrophages. Thus, neutrophils contribute to the disease’s progression rather than to host protection. The majority of studies addressing neutrophil functions have been performed using human peripheral blood PMNs, traditionally referred to as resting PMNs (15). This experimental system has at least two disadvantages which probably account for the conflicting results obtained by different research groups. First, neutrophil surface phenotype and functional activity strongly depend upon the genetic background of the donor. As shown above, mice of different inbred strains differ substantially with regard to the neutrophils’ migration capacity, phagocytic activity, expression of apoptotic markers, and lifespan. Thus, caution should be exercised when generalizing data obtained in genetically heterogeneous populations and in estimating the size of the sample necessary to obtain sufficient statistical power. Second, as the result of cytokine exposure, shifts in surface molecule expression during migration, and the change in microenvironment, tissue neutrophils differ profoundly from blood neutrophils (15). For example, compared to circulating PMNs, neutrophils recruited to peripheral tissues have a longer survival time because they are rescued from the constitutive apoptotic pathway (10). Prolonged survival of neutrophils in tissues may result in protection against some infectious agents but may also cause tissue damage. We found
that, in A/Sn-resistant (but not in I/St-susceptible) mice, both anatomic location and exposure to TB infection influence the expression of the CD95 apoptotic marker (Fig. 3). Thus, not only do neutrophils recovered from the lungs and blood differ physiologically but their site-specific features depend upon the genetics of the host. The functional role of neutrophils is often assessed by their short-term depletion in vivo using anti-Ly-6G RB68C5 MAbs. It was shown that neutrophil depletion in the very early stages of infections caused by Toxoplasma gondii and Candida albicans exacerbates the disease (4, 35); similar results were obtained in a murine TB model (32, 42). However, neutrophil depletion started later than day 2 of infection showed no effect, suggesting that neutrophils exert their protective function exclusively during the initiation of response. In our system, neutrophil influx into the lungs of susceptible mice occurred between weeks 1 and 3 of infection (Fig. 1) and at this stage was deleterious rather than protective. Considering antineutrophil immunotherapy with antibodies, we concluded that multiple injections would be needed to achieve a significant neutrophil exhaustion throughout a 2-week interval. It was shown, however, that sera from mice that received multiple injections of antiLy-6G RB6-8C5 antibodies contain high titers of blocking anti-RB6-8C5 immunoglobulin (14). Moreover, it was shown that anti-RB6-8C5 antibodies readily arise after as few as two anti-Ly-6G antibody injections (42). Thus, the immunogenic properties of antineutrophil antibodies preclude their usage when neutrophil inflammation is prolonged. In addition, the anti-Ly-6G antibody was shown to remove a subset of CD8⫹ T cells (28), thus hampering interpretation of the results. An important aspect of neutrophil inflammation is the balance between attraction of neutrophils to, and their elimination from, the inflamed tissue. Neutrophil apoptosis, followed by their phagocytosis by macrophages, provides a mechanism for the safe removal of neutrophils from the site of inflamma-
1752
ERUSLANOV ET AL.
tion, thereby minimizing the risk of tissue damage (38). In this study, we showed that I/St neutrophils have an increased mobility which allows a more rapid migration into the lung tissue, have a longer survival time than their A/Sn counterparts, and do not up-regulate CD95 expression when leaving circulation (Fig. 3). As infection progresses, a higher proportion of blood and lung A/Sn, but not I/St, neutrophils expresses the CD95 apoptotic receptor, the ligation of which results in Fas-induced cell death. Thus, the turnover of A/Sn neutrophils in the lung appears to be strictly controlled, whereas in I/St mice, overwhelming neutrophil inflammation may lead to lung pathology and an increase in the severity of TB disease. The involvement of some QTLs which control the disease severity (37) in the regulation of the neutrophil influx into infected lung (Fig. 6) clearly points at a pathogenic role for this aspect of the inflammatory response.
INFECT. IMMUN.
19. 20. 21.
22. 23. 24.
25.
ACKNOWLEDGMENTS We thank David McMurray for critically reading the manuscript. This work was supported by National Institutes of Health grant HL68532 and by HHMI grant 75301-564101 (to A.S.A. as an HHMI International Research Scholar). REFERENCES 1. Appelberg, R. 1992. T cell regulation of the chronic peritoneal neutrophilia during mycobacterial infections. Clin. Exp. Immunol. 89:120–125. 2. Appelberg, R., A. G. Castro, S. Gomes, J. Pedrosa, and M. T. Silva. 1995. Susceptibility of beige mice to Mycobacterium avium: role of neutrophils. Infect. Immun. 63:3381–3387. 3. Benedek-Spat, E., R. Di Felice, E. Andersen, and G. Cimasoni. 1991. In vitro release of elastase from human blood and gingival crevicular neutrophils. Arch. Oral Biol. 36:507–510. 4. Bliss, S. K., L. C. Gavrilescu, A. Alcazar, and E. Y. Denkers. 2001. Neutrophil depletion during Toxoplasma gondii infection leads to impaired immunity and lethal systemic pathology. Infect. Immun. 69:4898–4905. 5. Bliss, S. K., Y. Zhang, and E. Y. Denkers. 1999. Murine neutrophil stimulation by Toxoplasma gondii antigen drives high level production of IFNgamma-independent IL-12. J. Immunol. 163:2081–2088. 6. Borregaard, N., L. Kjeldsen, H. Sengelov, M. S. Diamond, T. A. Springer, H. C. Anderson, T. K. Kishimoto, and D. F. Bainton. 1994. Changes in subcellular localization and surface expression of L-selectin, alkaline phosphatase, and Mac-1 in human neutrophils during stimulation with inflammatory mediators. J. Leukoc. Biol. 56:80–87. 7. Brown, A. E., T. J. Holzer, and B. R. Andersen. 1987. Capacity of human neutrophils to kill Mycobacterium tuberculosis. J. Infect. Dis. 156:985–989. 8. Cassatella, M. A. 1995. The production of cytokines by polymorphonuclear neutrophils. Immunol. Today 16:21–26. 9. Denis, M. 1991. Human neutrophils, activated with cytokines or not, do not kill virulent Mycobacterium tuberculosis. J. Infect. Dis. 163:919–920. 10. Edwards, S. W., D. A. Moulding, M. Derouet, and R. J. Moots. 2003. Regulation of neutrophil apoptosis, p. 204–224. In M. A. Cassatella (ed.), The neutrophil: an emerging regulator of inflammatory and immune response, vol. 83. Karger, Basel, Switzerland. 11. Eruslanov, E. B., K. B. Majorov, M. O. Orlova, V. V. Mischenko, T. K. Kondratieva, A. S. Apt, and I. V. Lyadova. 2004. Lung cell responses to M. tuberculosis in genetically susceptible and resistant mice following intratracheal challenge. Clin. Exp. Immunol. 135:19–28. 12. Flynn, J. L., and J. Chan. 2001. Immunology of tuberculosis. Annu. Rev. Immunol. 19:93–129. 13. Fujie, K., Y. Shinguh, N. Inamura, R. Yasumitsu, M. Okamoto, and M. Okuhara. 1999. Release of neutrophil elastase and its role in tissue injury in acute inflammation: effect of the elastase inhibitor, FR 134043. Eur. J. Pharmacol. 374:117–125. 14. Fulton, S. A., S. M. Reba, T. D. Martin, and W. H. Boom. 2002. Neutrophilmediated mycobacteriocidal immunity in the lung during Mycobacterium bovis BCG infection in C57BL/6 mice. Infect. Immun. 70:5322–5327. 15. Galligan, C., and T. Yoshimura. 2003. Phenotypic and functional changes of cytokine-activated neutrophils. Chem. Immunol. Allergy 83:24–44. 16. Gamble, J. R., and M. A. Vadas. 1988. Endothelial adhesiveness for blood neutrophils is inhibited by transforming growth factor-. Science 242:97–99. 17. Jones, G. S., H. J. Amirault, and B. R. Andersen. 1990. Killing of Mycobacterium tuberculosis by neutrophils: a nonoxidative process. J. Infect. Dis. 162:700–704. 18. Kasahara, K., I. Sato, K. Ogura, H. Takeuchi, K. Kobayashi, and M. Adachi.
26.
27. 28.
29. 30. 31. 32.
33. 34. 35. 36. 37.
38.
39. 40. 41. 42.
43.
1998. Expression of chemokines and induction of rapid cell death in human blood neutrophils by Mycobacterium tuberculosis. J. Infect. Dis. 178:127–137. Kaufmann, S. H. E. 2001. How can immunology contribute to the control of tuberculosis? Nat. Rev. Immunol. 1:20–30. Kisich, K. O., M. Higgins, G. Diamond, and L. Heifets. 2002. Tumor necrosis factor alpha stimulates killing of Mycobacterium tuberculosis by human neutrophils. Infect. Immun. 70:4591–4599. Lasco, T. M., O. C. Turner, L. Cassone, I. Sugawara, H. Yamada, D. N. McMurray, and I. M. Orme. 2004. Rapid accumulation of eosinophils in lung lesions in guinea pigs infected with Mycobacterium tuberculosis. Infect. Immun. 72:1147–1149. Lavebratt, C., A. S. Apt, B. V. Nikonenko, M. Schalling, and E. Schurr. 1999. Severity of tuberculosis in mice is linked to distal chromosome 3 and proximal chromosome 9. J. Infect. Dis. 180:150–155. Lee, W. L., and G. P. Downey. 2001. Neutrophil activation and acute lung injury. Curr. Opin. Crit. Care 7:1–7. Liles, W. C., P. A. Kiener, J. A. Ledbetter, A. Aruffo, and S. J. Klebanoff. 1996. Differential expression of Fas (CD95) and Fas ligand on normal human phagocytes: implications for the regulation of apoptosis in neutrophils. J. Exp. Med. 184:429–440. Lyadova, I. V., E. B. Eruslanov, S. V. Khaidukov, V. V. Yeremeev, K. B. Majorov, A. V. Pichugin, B. V. Nikonenko, T. K. Kondratieva, and A. S. Apt. 2000. Comparative analysis of T lymphocytes recovered from the lungs of mice genetically susceptible, resistant, and hyperresistant to Mycobacterium tuberculosis-triggered disease. J. Immunol. 165:5921–5931. Majorov, K. B., I. V. Lyadova, T. K. Kondratieva, E. B. Eruslanov, E. I. Rubakova, M. O. Orlova, V. V. Mischenko, and A. S. Apt. 2003. Different innate ability of I/St and A/Sn mice to combat virulent Mycobacterium tuberculosis: phenotypes expressed in lung and extrapulmonary macrophages. Infect. Immun. 71:697–707. May, M. E., and P. J. Spagnuolo. 1987. Evidence for activation of a respiratory burst in the interaction of human neutrophils with Mycobacterium tuberculosis. Infect. Immun. 55:2304–2307. Montes de Oca, R., A. J. Buendia, L. Del Rio, J. Sa ´nchez, J. Salinas, and J. A. Navarro. 2000. Polymorphonuclear neutrophils are necessary for the recruitment of CD8⫹ T cells in the liver in a pregnant mouse model of Chlamydophila abortus (Chlamydia psittaci serotype 1) infection. Infect. Immun. 68:1746–1751. Neufert, C., R. K. Pai, E. H. Noss, M. Berger, W. H. Boom, and C. V. Hardling. 2001. Mycobacterium tuberculosis 19-kDa lipoprotein promotes neutrophil activation. J. Immunol. 167:1542–1549. Nikonenko, B. V., M. M. Averbakh, Jr., C. Lavebratt, E. Schurr, and A. S. Apt. 2000. Comparative analysis of mycobacterial infections in susceptible I/St and resistant A/Sn inbred mice. Tuber. Lung Dis. 80:15–25. Orme, I., and A. M. Cooper. 1999. Cytokine/chemokine cascades in immunity to tuberculosis. Immunol. Today 20:307–312. Pedrosa, J., B. M. Saunders, R. Appelberg, I. M. Orme, M. T. Silva, and A. M. Cooper. 2000. Neutrophils play a protective nonphagocytic role in systemic Mycobacterium tuberculosis infection of mice. Infect. Immun. 68: 577–583. Petrofsky, M., and L. E. Bermudez. 1999. Neutrophils from Mycobacterium avium-infected mice produce TNF-alpha, IL-12, and IL-1-beta and have a putative role in early host response. Clin. Immunol. 91:354–358. Riedel, D. D., and S. H. Kaufmann. 1997. Chemokine secretion by human polymorphonuclear granulocytes after stimulation with Mycobacterium tuberculosis and lipoarabinomannan. Infect. Immun. 65:4620–4623. Romani, L., A. Mencacci, E. Cenci, G. Del Sero, F. Bistoni, and P. Puccetti. 1997. An immunoregulatory role for neutrophils in CD4⫹ T helper subset selection in mice with candidiasis. J. Immunol. 158:2356–2362. Rook, G. A. W., and R. Hernandez-Pando. 1996. The pathogenesis of tuberculosis. Annu. Rev. Microbiol. 50:259–284. Sa ´nchez, F., T. V. Radaeva, B. V. Nikonenko, A.-S. Persson, S. Sengul, M. Schalling, E. Schurr, A. S. Apt, and C. Lavebratt. 2003. Multigenic control of disease severity after virulent Mycobacterium tuberculosis infection in mice. Infect. Immun. 71:126–131. Savill, J. S., A. H. Wyllie, J. E. Henson, M. J. Walport, P. M. Henson, and C. Haslett. 1989. Macrophage phagosytosis of aging neutrophils in inflammation. Programmed cell death in the neutrophil leads to its recognition by macrophages. J. Clin. Investig. 83:865–875. Scapini, P., J. A. Lapinet-Vera, S. Gasperini, F. Calzetti, F. Bazzoni, and M. A. Cassatella. 2000. The neutrophil as a cellular source of chemokines. Immunol. Rev. 177:195–203. Schluger, N. W., and W. N. Rom. 1998. The host immune response to tuberculosis. Am. J. Crit. Care Med. 157:679–691. Schwander, S. K., E. Sada, M. Torres, D. Escobedo, J. G. Sierra, S. Alt, and E. A. Rich. 1996. T lymphocytic and immature macrophage alveolitis in active pulmonary tuberculosis. J. Infect. Dis. 173:1267–1272. Seiler, P., P. Aichele, B. Raupach, B. Odermatt, U. Steihoff, and S. H. E. Kaufmann. 2000. Rapid neutrophil response controls fast-replicating intracellular bacteria but not slow-replicating Mycobacterium tuberculosis. J. Infect. Dis. 181:671–680. Smith, W. B., L. Noack, Y. Khew-Goodall, S. Isenmann, M. A. Vadas, and
VOL. 73, 2005
NEUTROPHIL RESPONSE TO MYCOBACTERIA AND TB SEVERITY
J. R. Gamble. 1996. Transforming growth factor-1 inhibits the production of IL-8 and the transmigration of neutrophils through activated endothelium. J. Immunol. 157:360–368. 44. Stewart, G. H., B. D. Robinson, and D. B. Young. 2003. Tuberculosis: a problem with persistence. Nat. Rev. Microbiol. 1:97–105. 45. Suwa, T., J. C. Hogg, M. E. Klut, J. Hards, and S. F. van Eeden. 2001. Interleukin-6 changes deformability of neutrophils and induces their sequestration in the lung. Am. J. Respir. Crit. Care Med. 163:970–976. 46. Tateda, K., T. A. Moore, J. C. Deng, M. W. Newstead, X. Zeng, A. Matsukawa,
Editor: A. D. O’Brien
1753
M. S. Swanson, K. Yamaguchi, and T. J. Standiford. 2001. Early recruitment of neutrophils determines subsequent T1/T2 host responses in a murine model of Legionella pneumophila pneumonia. J. Immunol. 166:3355–3361. 47. Weiss, S. J. 1989. Tissue destruction by neutrophils. N. Engl. J. Med. 320: 365–376. 48. Witko-Sarsat, V., E. M. Cramer, C. Hieblot, J. Guichard, P. Nusbaum, S. Lopez, P. Lesavre, and L. Halbwachs-Mecarelli. 1999. Presence of proteinase 3 in secretory vesicles: evidence of a novel, highly mobilizable intracellular pool distinct froom azurophil granules. Blood 94:2487–2496.
