Pneumocystis stimulates MCP-1 production by alveolar epithelial cells through a JNK-dependent mechanism

Jing Wang, Francis Gigliotti, Samir P. Bhagwat, Sanjay B. Maggirwar, Terry W. Wright

Abstract

Pneumocystis carinii is an opportunistic fungal pathogen that causes pneumonia (PCP) in immunocompromised individuals. Recent studies have demonstrated that the host's immune response is clearly responsible for the majority of the pathophysiological changes associated with PCP. P. carinii interacts closely with alveolar epithelial cells (AECs); however, the nature and pathological consequences of the epithelial response remain poorly defined. Monocyte chemotactic protein-1 (MCP-1) is involved in lung inflammation, immunity, and epithelial repair and is upregulated during PCP. To determine whether AECs are an important source of MCP-1 in the P. carinii-infected lung, in vivo and in vitro studies were performed. In situ hybridization showed that MCP-1 mRNA was localized to cells with morphological characteristics of AECs in the lungs of infected mice. In vitro studies demonstrated that P. carinii stimulated a time- and dose-dependent MCP-1 response in primary murine type II cells that was preceded by JNK activation. Pharmacological inhibition of JNK nearly abolished P. carinii-stimulated MCP-1 production, while ERK, p38 MAPK, and TNF receptor signaling were not required. Furthermore, delivery of a JNK inhibitory peptide specifically to pulmonary epithelial cells using a recombinant adenovirus vector blocked the early lung MCP-1 response following intratracheal instillation of infectious P. carinii. JNK inhibition did not affect P. carinii-stimulated production of macrophage inflammatory protein-2 in vitro or in vivo, indicating that multiple signaling pathways are activated in P. carinii-stimulated AECs. These data demonstrate that AECs respond to P. carinii in a proinflammatory manner that may contribute to the generation of immune-mediated lung injury.

  • inflammation
  • epithelial
  • AIDS

pneumonia induced by Pneumocystis carinii (PCP) continues to be the most common AIDS-defining illness as well as an important cause of morbidity and mortality in patients with a wide array of immunosuppressive conditions. Most recent studies indicate that, although the presence of P. carinii is obviously a factor in the development of lung injury, pulmonary inflammation is a major determinant of the severity of PCP (5, 27, 53, 54). In AIDS patients with profound reductions in CD4+ T cell numbers, bronchoalveolar lavage (BAL) fluid IL-8 and neutrophil concentrations, but not organism numbers, correlate with severity of PCP (5, 23, 28). In addition, clinical studies of immune-reconstituted PCP patients and controlled animal studies have both demonstrated that inflammatory mediators are released, and immune and inflammatory cells are recruited to the lung in response to P. carinii (2, 25, 38, 39). More defined studies in mice have identified specific T cell subsets as having prominent roles in the lung injury associated with PCP (54, 57). For example, CD8+ T cells are responsible for much of the PCP-associated lung injury that occurs in CD4-deficient hosts (17). Therefore, recent studies have focused on the mechanisms by which T cells accumulate in the lung during PCP.

While the close interaction of P. carinii with the alveolar epithelium was one of the first observations offering insight into the pathogenesis of PCP (26, 32, 58, 59), very little is known about the epithelial response to P. carinii. In vivo studies have most often noted the attachment of P. carinii to the type I pneumocyte. However, this observation does not preclude an important role for type II cells in the response to P. carinii. Type II cells are closely positioned near the type I cells and have been reported to interact with P. carinii in vivo (29). These findings are particularly relevant given that the type II cell is becoming increasingly recognized as an immune effector cell in the alveolus (16, 46). Our prior studies have found that immune and inflammatory cells are recruited specifically to alveolar sites of P. carinii infection in mice, suggesting that the interaction of P. carinii with the alveolar epithelium targets the immune response (55, 56). Other groups have found that disruption of alveolar epithelial cell (AEC) signaling in vivo modifies pulmonary immune and inflammatory responses (21, 46). Evidence has also accumulated that transformed human lung epithelial-like cell lines can produce inflammatory mediators in response to P. carinii stimulation (4, 41, 60). In addition, a murine AEC line and primary rat and mouse type II cells undergo NF-κB-dependent macrophage inflammatory protein-2 (MIP-2) production when stimulated with P. carinii or purified P. carinii glucan (15, 52). However, recent evidence indicates that neutrophils are not critical to the development of lung injury during PCP (47). Thus, focus has shifted to CC chemokines, including monocyte chemotactic protein-1 (MCP-1), which have a role in the tissue recruitment of T cells. Pulmonary epithelial cells are capable of secreting MCP-1 in response to infectious stimulation, and epithelial MCP-1 mediates the pulmonary recruitment of CD8+ T cells (42, 62). In addition, a role for MCP-1/CC chemokine receptor 2 (MCP-1/CCR2) signaling in the repair of damaged pulmonary epithelium has been suggested (10). These findings, combined with the fact that lung MCP-1 levels are dramatically elevated in mice with PCP, make this chemokine of interest for its contribution to the pathological T cell response that is critical to the progression of PCP.

Identification of epithelial-specific responses to P. carinii, and the signaling cascades leading to these responses, will aid in understanding the role of AECs as immune effector cells in the generation of pulmonary immune and inflammatory responses. The hypothesis of the current study is that the interaction of P. carinii with type II AECs induces MCP-1 production through NF-κB and mitogen-activated protein kinase (MAPK)-dependent mechanisms. To test this hypothesis, primary murine type II cell cultures, in vivo mouse models, and specific pharmacological and protein inhibitors of NF-κB, p38, JNK, and ERK signaling were utilized. This study will help determine the mechanism by which P. carinii-epithelial interaction promotes and targets the pathological immune/inflammatory response and also determine whether this interaction might be exploited as a therapeutic intervention.

MATERIALS AND METHODS

Animals.

CB.17 wild-type and severe combined immunodeficient (SCID) mice were purchased from Taconic. C57BL/6 mice were purchased from the Jackson Laboratory. B6.129-Tnfrsf1atm1Mak/J and B6.129S2-Tnfrsf1btm1Mwm/J mice on a C57BL/6 background were also purchased from the Jackson Laboratory. These mice were then crossed to produce mice deficient in both TNFR1 and TNFR2 (“TNFR-deficient”) on a C57BL/6 background.

All animal protocols were preapproved by University Committee for Animal Research (UCAR) at the University of Rochester Medical Center.

Isolation and culture of primary murine type II cells.

Primary type II pneumocytes were isolated from mouse lungs using a modification of the method of Corti et al. (11). Briefly, the lung was perfused with saline. Two milliliters of dispase solution (BD Biosciences) was instilled by tracheal catheter, followed immediately by slow insertion of 0.45 ml of low-melting-point agarose (GIBCO-BRL) at 45°C. The lungs were cooled briefly on ice and then incubated at room temperature in dispase for 45 min. The lung tissue was microdissected, incubated briefly in DMEM with 0.01% DNase at room temperature, filtered through nylon monofilament screens (100, 40, and 25 μm; BD Falcon), and centrifuged at 150 g at 4°C. Type II cells were purified from inflammatory cells by incubation with biotin-conjugated antibodies against CD32 and CD45 followed by recovery with streptavidin-conjugated magnetic beads in a magnetic separator. With this procedure, type II cell yield was ∼1 × 106 cells/mouse. Typically the type II cells were >95% viable and >92% pure as assessed by papanicolaou staining. In addition, >95% of the isolated cells were typically positive for surfactant protein C (SP-C) expression as assessed by intracellular staining and fluorescence-activated cell sorting (FACS) analysis.

The isolated type II cells were cultured under conditions previously demonstrated to maintain a type II phenotype as described by Rice et al. (43). The cells were cultured on Matrigel/rat tail collagen-coated plates (ratio 70/30, vol/vol) (BD Biosciences). Cells were maintained at 37°C with 6% CO2 in bronchial epithelial cell growth medium (BEGM) without hydrocortisone (Cambrex), supplemented with 5% charcoal-stripped FBS (Hyclone, Logan, UT) and 10 ng/ml keratinocyte growth factor (KGF) (Calbiochem) to promote maintenance of a type II cell phenotype.

Intracellular SP-C staining.

Intracellular SP-C expression was measured to determine the purity of type II cells. Briefly, primary AECs were gently dislodged from wells and washed with PBS + 1% FBS staining buffer, spun at 250 g for 10 min, and resuspended in staining buffer. The cells were incubated with Cytofix/Cytoperm (BD Biosciences) solution on ice for 20 min. The cells were stained with rabbit-anti-human SP-C (Chemicon International), which cross-reacts with mouse SP-C, followed by goat anti-rabbit conjugated with allophycocyanin or primary antibody alone. Unstained cells were used as an additional negative control. SP-C expression was measured by a FACS caliber cytofluorometer and analyzed by CellQuest software (Becton-Dickinson, San Jose, CA).

Isolation and enumeration of mouse P. carinii.

P. carinii was isolated from the lungs of heavily infected SCID mice and enumerated by the Gomori methenamine silver staining as described before (52).

Antibody-mediated depletion of P. carinii.

Since there is no reliable in vitro culture system for P. carinii, the organisms used in these experiments were purified from the lungs of infected SCID mice. Therefore, to ensure that MCP-1 secretion by AECs was a response to P. carinii and not to potentially copurified contaminants or mouse lung proteins, an antibody/magnetic bead-based technique for the specific removal of P. carinii from the partially purified P. carinii preparation was employed. The purified preparation was depleted of P. carinii using magnetic beads coated with a pool of anti-P. carinii antibodies. Enumeration of cysts before and after P. carinii depletion, as well as real-time PCR analysis, demonstrated that this method removed >96% of the P. carinii organisms as previously described (52).

In vitro AEC treatments.

Primary murine type II cells were grown to >90% confluence and then incubated in BEGM without KGF for 6 h prior to experimental treatments. In some experiments cells were pretreated with sulfasalazine (SSA, 2 mM; Calbiochem), SB203580 (0.6 μM), SP600125 (10 μM), or PD20359 (50 μM; Calbiochem), to inhibit IKK, p38 MAPK, JNK, and ERK, respectively. Cells were pretreated with inhibitors in serum-free media for 2 h prior to P. carinii stimulation. In addition, JNK inhibitor 1, l-stereoisomer (l-JNKI1) (Alexis Biochemicals), was used as a more specific inhibitor of JNK than SP600125. A protein named JNK interacting protein 1/IB1 (JIP-1/IB1) has been described that competitively blocks the interaction between JNK and c-Jun, thereby inhibiting the signaling events downstream of JNK (1, 7). To convert JIP-1/IB1 into a cell-permeable inhibitor of JNK (l-JNKI peptides), the minimal 20-amino-acid inhibitory sequence of JIP-1/IB1 was covalently linked to the 10-amino-acid HIV Tat transporter sequence recognized by TAT transporter (8, 50). l-JNKI1 is a potent and specific inhibitor of JNK. The control peptide (l-TAT), which lacks JNK inhibitory activity, was also used in the culture experiments. l-JNKI1 and l-TAT (Alexis Biochemicals) were used at a concentration of 20 μM. Monolayers of type II cells were pretreated with inhibitors or control in serum-free DMEM media without KGF for 1 h and then stimulated with freshly isolated mouse P. carinii organisms suspended in serum-free DMEM. In some experiments, cells were treated with β-glucan (Sigma) as a control for the whole freshly isolated P. carinii. Experiments were terminated at 6, 12, or 24 h. Cell supernatants were recovered for MCP-1 ELISA, and cells were used to either isolate total RNA for ribonuclease protection assay (RPA) or produce cell lysates for Western analysis.

Cytokine enzyme-linked immunosorbent assay (ELISA).

Culture supernatants were collected, centrifuged at 12,000 g for 5 min to remove debris, and then stored at −80°C. MCP-1 and MIP-2 concentrations were measured using a commercially available ELISA kit according to the manufacturer's instructions (R&D, Minneapolis, MN).

In situ hybridization.

In situ hybridization was performed as previously described (54, 56). Murine clones for MCP-1 were subcloned into the plasmid vector, pBluescript II SK+ (Stratagene, La Jolla, CA), for the in vitro transcription of RNA. Sense and antisense orientations were confirmed by DNA sequencing. MCP-1 antisense RNAs were transcribed from 1 μg of linearized plasmid template according to the procedure. After ethanol precipitation, the riboprobes were dissolved in diethylpyrocarbonate-treated water. Full-length transcripts were ∼0.7 kb for MCP-1. Prior to hybridization, limited alkaline hydrolysis was performed to create riboprobes ranging in length from 0.1 to 0.3 kb. Hydrolyzed transcripts were sized by denaturing agarose gel electrophoresis.

The lungs from P. carinii-infected SCID mice were inflation-fixed with 15 cmH2O gravity flow pressure of 10% buffered formalin (Sigma-Aldrich, St. Louis, MO). Tissue sections of 4 μm were cut, then treated by the method of Angerer et al. (1a) with modification as previously described (54, 56).

RNA isolation and RPAs.

Primary AECs were grown to confluence in 24-well plates, then stimulated for the indicated times. Total RNA was isolated from the cells using TRIzol reagent according to the manufacturer's instructions (Life Technologies, Grand Island, NY). A custom RPA template was purchased (BD Biosciences) and used to transcribe radiolabeled, antisense riboprobes for murine MCP-1 and the murine ribosomal protein L32, as previously described (56).

Western blot analysis.

Primary type II cells were cultured and treated in 12-well cell tissue culture plates. At the indicated time points, cells were washed with 1× cold PBS and lysed in ELB buffer (50 mM HEPES, pH 7, 250 mM NaCl, 0.1% NP-40, 5 mM EDTA, 10 mM NaF, 0.1 mM Na3VO4, 50 μM ZnCl2, supplemented with 0.1 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, and a mixture of protease and phosphatase inhibitors), and cellular debris was removed by high-speed centrifugation. Cell lysates containing equal amounts of total protein were fractionated by 10% SDS-PAGE and electrophoretically transferred to methanol-pretreated polyvinylidene difluoride membrane. Immunoblot assays were performed as described previously (30). Phospho-JNK primary antibodies were purchased from Cell Signaling Technology and diluted 1:1,000 in PBS with 5% Tween and nonfat dry milk. Secondary anti-mouse and anti-rabbit antibodies were purchased from Amersham and used at a dilution of 1:2,000.

Administration of recombinant adenovirus.

Replication-deficient recombinant adenovirus (rAd) type 5 was used to construct a rAd vector overexpressing the Jun binding domain (JBD) of JNK interacting protein 1 (Ad-JBD) (22). This peptide interrupts the activation of the downstream JNK signal. A rAd-vector expressing firefly luciferase (Ad-LUC) was used as a control. Mice were intratracheally instilled with rAd vectors or an equal volume of sterile PBS under anesthesia. Each mouse was given a single dose of 109 plaque-forming units of Ad-JBD or Ad-LUC diluted in sterile PBS. Recent data has demonstrated that intratracheal instillation of this dose of rAd delivers genes selectively to the lung epithelium (46). Forty-eight hours after virus infection, mice were intratracheally inoculated with 106 purified P. carinii or an equal volume of sterile PBS as a control. Three hours after P. carinii inoculation, mice were euthanized and lungs were lavaged with four 1-ml aliquots of sterile PBS. MCP-1 levels in the lavage fluid were determined by ELISA.

Statistical analysis.

A one-way analysis of variance was performed with the SigmaStat 2.0 software (Jandel, San Rafael, CA) to determine the confidence intervals of observed variations in chemokine protein and mRNA levels in the experimental animals. The Student-Newman-Keuls method was used for all pair-wise multiple comparisons of experimental groups.

RESULTS

P. carinii stimulates epithelial MCP-1 production in vivo.

It has been demonstrated that MCP-1 levels are elevated in the lungs of mice with active PCP (56, 57). However, the intensity of the inflammatory response accompanying PCP has made it difficult to identify the cell types responsible for MCP-1 production. To examine the role of resident lung cells in the MCP-1 response to P. carinii, normal and CD4-depleted mice were intratracheally inoculated with P. carinii and examined for MCP-1 production at 1, 3, 8, 24, 48, and 72 h postinoculation. While MCP-1 was not detected in the BAL fluid of mice inoculated with saline alone, MCP-1 levels were significantly elevated in the BAL fluid of P. carinii inoculated mice at 1, 3, and 8 h postinoculation and had declined toward baseline levels by 24 h (Fig. 1A). MCP-1 protein levels were not different in normal and CD4-depleted mice, demonstrating that CD4+ T cells were not required for MCP-1 production. Importantly, MCP-1 was produced at time points that precede inflammatory cell infiltration into the lungs following P. carinii inoculation. Because neutrophils are the first cells recruited to the lung following the instillation of P. carinii, the percentage of BAL fluid neutrophils was used as a marker of P. carinii-induced inflammatory response. MCP-1 levels were clearly elevated prior to the appearance of neutrophils in the lung (Fig. 1B). Importantly, inoculation of mice with preparations that were depleted of P. carinii did not elicit an MCP-1 response, showing that the organism was driving the response (Fig. 1C). Together, these data showed that resident lung cells were capable of mounting an MCP-1 response to P. carinii.

Fig. 1.

Pneumocystis carinii induces monocyte chemotactic protein-1 (MCP-1) production by resident lung cells in vivo. Nondepleted and CD4+ T-cell-depleted CB.17 mice were intratracheally inoculated with 1 × 106 purified P. carinii (Pc), and then euthanized 1, 3, 8, 24, 48, and 72 h postinoculation. The bronchoalveolar lavage (BAL) fluid (“BALF”) was collected, and MCP-1 levels were determined by ELISA (A). The percentage of neutrophils (PMNs) recovered in the BAL fluid was also determined at the same time points (B). The label “C” on the abscissa indicates the very small bar that shows the percentage of PMNs in the BAL fluid of uninfected mice. C: the BAL fluid MCP-1 levels of CB.17 mice that were inoculated with 1 × 106 P. carinii or an equal amount of the same P. carinii preparation that had been depleted of viable organisms (52). BAL fluid was collected at 3, 8, and 24 h postinoculation, and MCP-1 levels were determined by ELISA. Values are means ± SE. *P < 0.05 compared with mice treated with P. carinii-depleted preparations; n ≥ 3 for each group at each time point.

To more specifically identify the cells producing MCP-1 in vivo, in situ hybridization was performed on inflation-fixed P. carinii-infected and uninfected SCID mouse lung sections. SCID mice exhibit less cellular infiltration in response to P. carinii infection, thereby facilitating localization of mRNA hybridization to cells with histological characteristics of epithelial cells. A radiolabeled MCP-1-specific antisense riboprobe was constructed and used as previously described (54, 56). While only low background hybridization was observed in the uninfected lung (Fig. 2C), focal hybridization was shown in the alveolar epithelium in P. carinii-infected mice (Fig. 2, A, B, DF). The appearance and positioning of cells expressing MCP-1 RNA was consistent with that of type II AECs. Gomori methenamine silver staining of a serial section demonstrated that P. carinii cysts colocalized to the same region of the lung that exhibited an increased steady-state level of MCP-1 mRNA (Fig. 2, G and H). Black asterisks in Fig. 2, A, B, and G, mark the same structures in each micrograph. Red asterisks in Fig. 2, G and H, mark the same alveoli under lower and higher power magnification. These data demonstrated that AECs in P. carinii-infected regions of the lung produce MCP-1 in vivo.

Fig. 2.

Alveolar epithelial cells (AECs) express MCP-1 in vivo during P. carinii infection. Severe combined immunodeficient (SCID) mice were infected with P. carinii by cohousing with heavily infected SCID mice. When the mice began to show increased respiration, they were euthanized, and the lungs were inflation fixed with 10% formalin under 15 cmH2O of pressure. The fixed tissue was paraffin embedded, and 4-μm sections were cut. In situ hybridization was done using a MCP-1 antisense riboprobe as described. A: low-magnification dark-field image of sections from P. carinii-infected SCID mice. B: corresponding bright-field image of A. C: dark-field image of uninfected lung. D–F: high-magnification dark-field and light-field images of sections from P. carinii-infected SCID mice. Arrows show the positive signal of MCP-1 in the dark-field images and type II AECs in the light-field images. G and H: silver-stained serial lung sections from the same P. carinii-infected SCID mice; arrows show the P. carinii cysts in the alveoli. Black asterisks in A, B, and G mark the same structures in each micrograph. Red asterisks in G and H mark the same alveoli under lower and higher power magnification, respectively.

P. carinii stimulates MCP-1 production by primary type II pneumocytes.

To directly assess the ability of AECs to respond to P. carinii, primary type II cells were purified and cultured under conditions that promote a type II phenotype (43). Confluent monolayers of primary murine type II cells were inoculated with freshly isolated murine P. carinii at cyst-to-AEC ratios of 0, 0.5, 1.0, and 2.0. At 6, 12, and 24 h postinoculation, the culture supernatants were removed and assayed for MCP-1 by ELISA, and total RNA was isolated from the cells for RPA analysis. Stimulation of AECs with P. carinii induced a dose- and time-dependent increase in MCP-1 secretion (Fig. 3A). MCP-1 concentrations were significantly elevated in AECs treated with P. carinii at 6, 12, and 24 h postinoculation compared with unstimulated cells (P < 0.05). Because the mouse P. carinii must be isolated from lung tissue, a P. carinii-depleted preparation was used to control for the presence of non-P. carinii contaminants that could contribute to the AEC response. Antibody-mediated removal of the P. carinii from the preparation completely abolished the inducible MCP-1 response, demonstrating that the AECs are responding to P. carinii (Fig. 3A) and not a copurified contaminant.

Fig. 3.

P. carinii stimulates MCP-1 production by type II AECs. Primary type II AECs were isolated and cultured under conditions reported to maintain a type II cell phenotype as described above. Pneumocytes were cultured until monolayers were >90% confluent. The cells were then treated with P. carinii at cyst-to-AEC ratios of 0, 0.5, 1.0, and 2.0. Culture supernatants were collected 6, 12, and 24 h later for MCP-1 ELISA, and cells were lysed for RNA isolation. An amount of P. carinii preparation equal to the highest dose (2.0) was depleted of P. carinii as described in the material and methods and used as a control. A: MCP-1 protein levels in the culture supernatant as determined by ELISA. The data presented are from one of at least three independent experiments with separate AEC and P. carinii preparations and are expressed as means ± SE; *P < 0.05 compared with both unstimulated cells and cells treated with P. carinii-depleted preparations; n = 3 for each condition at each time point. B: RNA expression at 6 h as determined by ribonuclease protection assay (RPA). Lane “C” indicates untreated AECs. Lanes 1, 2, and 3 are representative of cells treated with P. carinii at cyst-to-AEC ratios of 0.5, 1.0, and 2.0, respectively. C: the quantification of 6 h MCP-1 RPA data (B) by PhosphorImager analysis. The data are representative of one of three independent experiments with separate AEC and P. carinii preparations. Values are means ± SE (n = 3). *P < 0.05 compared with unstimulated cells. L32, murine ribosomal protein L32.

RPA analysis demonstrated that the P. carinii-stimulated increase in MCP-1 protein secretion was accompanied by a concomitant increase in the steady-state level of MCP-1 mRNA (Fig. 3, B and C). MCP-1 mRNA was elevated at 6 h postinoculation in AECs stimulated with cyst-to-AEC ratios of 0, 0.5, 1.0, and 2.0. In contrast, unstimulated cells exhibited no detectable MCP-1 mRNA expression. Steady-state MCP-1 mRNA was not significantly elevated at 12 or 24 h (data not shown). These data demonstrate that the primary AECs respond to P. carinii stimulation with MCP-1 production and suggest that primary AECs have the capacity to serve as important inflammatory modulators in response to P. carinii.

JNK signaling is required for P. carinii-stimulated MCP-1 production by type II AECs.

Primary type II cells were isolated and cultured as described above, then stimulated with freshly isolated mouse P. carinii in the absence or presence of the MAPK inhibitors SB203580, PD98059, and SP600125, or the IKK inhibitor SSA. Unstimulated cells and cells stimulated with P. carinii-depleted preparations were used as controls. Experiments were terminated at the 6 h time point to assess MCP-1 levels in the culture supernatants by ELISA. Inhibition of either JNK/SAPK1 with SP600125 or IKK with SSA nearly completely abolished MCP-1 protein secretion by P. carinii-stimulated AECs (Fig. 4). In contrast, neither inhibition of p38 MAPK with PD98059 nor p44/42 MAPK with SB203580 had any effect on P. carinii-stimulated MCP-1 production (Fig. 4). Since it has been reported that SP600125 may not have absolute specificity for JNK, additional experiments were performed using l-JNKI1 as a specific, cell-permeable inhibitor of JNK. Similar to SP600125, l-JNKI1 also nearly completely blocked the MCP-1 response of P. carinii-stimulated AECs (Fig. 5). Western blot analysis of cell lysates from type II AECs confirmed that P. carinii induced increased cellular levels of activated phospho-JNK (Fig. 6). Together, these results demonstrated that P. carinii stimulates the JNK signaling pathway in AECs, leading to MCP-1 production.

Fig. 4.

Inhibition of JNK signaling blocks MCP-1 production by P. carinii-stimulated type II AECs. Primary type II AECs were isolated and cultured until >90% confluent as described above. The cells were then treated with a 2:1 cyst-to-AEC ratio of P. carinii in the absence or presence of specific inhibitors [2 mM sulfasalazine (SSA), 10 μM SP600125, 0.6 nM SB203580, or 50 μM PD98059]. Culture supernatants were collected 6 h later for MCP-1 ELISA. Untreated AECs and AECs treated with an equal amount of the same P. carinii preparation that had been depleted of organisms were used as controls. Values are means ± SE. *P < 0.05 compared with untreated control, P. carinii-depleted control, and P. carinii-stimulated cells pretreated with SSA or SP600125; n = 3 for each condition. The data are from one of three independent experiments with separate AEC and P. carinii isolations.

Fig. 5.

JNK inhibitor 1, l-stereoisomer (l-JNKI1), blocks MCP-1 production by P. carinii-stimulated type II AECs. Primary type II cells were isolated and cultured until >90% confluent as described above. The cells were then treated with a 2:1 cyst-to-AEC ratio of P. carinii in the absence or presence of SP600125, the highly specific JNK inhibitor, l-JNKI1, or a control peptide, l-TAT. Culture supernatants were collected 6 h or 24 h later for MCP-1 ELISA. Untreated AECs and AECs treated with inhibitors alone were used as controls. AECs were pretreated for 2 h with 10 μM SP600125, or for 1 h with 20 μM l-JNKI1 or 20 μM control peptide l-TAT. Values are means ± SE. *P < 0.05 compared with untreated controls, cells treated with inhibitors alone, and P. carinii-stimulated cells treated with SP600125 or l-JNKI1; n = 3 for each condition. The data are from one of three independent experiments with separate AEC and P. carinii isolations.

Fig. 6.

P. carinii stimulates JNK activation in primary type II AECs. Cells were isolated and cultured as described above. When cells were 80% confluent, they were treated with P. carinii at a cyst-to-AEC ratio of 2:1. Cells were washed and harvested 30 min, 1 h, and 2 h later. Cell lysates were separated on a 10% SDS-PAGE gel and electrotransferred to polyvinylidene difluoride membranes. Immunoblotting was performed with anti-phospho-JNK and anti-β-actin primary antibodies, followed by incubation with horseradish peroxidase-conjugated secondary antibody and detection with electrochemical luminescence. One representative experiment out of three is shown.

TNFR signaling is not required for P. carinii-stimulated MCP-1 production by type II AECs.

It has been reported that P. carinii β-glucan can induce a low level induction of TNF by AECs (15). Therefore, to determine whether the MCP-1 response of P. carinii-stimulated AECs was dependent upon autocrine stimulation by TNF, we assessed the MCP-1 response of primary AECs isolated from wild-type and TNFR-deficient mice. Importantly, the MCP-1 response to P. carinii did not require TNFR signaling. Both TNFR-deficient and wild-type AECs exhibited elevated MCP-1 protein secretion and mRNA expression in response to P. carinii (Fig. 7). In a separate experiment, primary AECs produced MCP-1 under recombinant TNF-α stimulation (data not shown). Overall, this data demonstrates that both P. carinii and TNF are potent inducers of MCP-1 production by AECs. Furthermore, since the majority of MCP-1 production by P. carinii-stimulated AECs is TNFR-independent, it suggests that in vivo direct P. carinii-AEC interactions contribute to MCP-1 production.

Fig. 7.

MCP-1 production by P. carinii-stimulated type II AECs is TNFR independent. Primary type II AECs from wild-type (WT) and TNFR-deficient mice were isolated and cultured as described above. AECs were grown to >90% confluence and then treated with a 2:1 cyst-to-AEC ratio of P. carinii. Culture supernatants were collected 6 h later for MCP-1 ELISA. Untreated wild-type and TNFR-deficient AECs were used as controls. Values are means ± SE (n = 3 for each condition). **P < 0.05 compared with both untreated controls and P. carinii-stimulated TNFR−/− AECs. *P < 0.05 compared with untreated controls. The data are from one of three individual experiments.

β-Glucan does not induce MCP-1 production in murine primary type II cells.

Prior studies have demonstrated that purified β-glucan is a more potent inducer of epithelial MIP-2 secretion than whole Pneumocystis (15). Therefore, the relative potency of β-glucan and whole Pneumocystis for the induction of MCP-1 production in primary murine type II cells was assessed. As expected, whole Pneumocystis induced significant MCP-1 secretion at 24 h. In contrast, AECs treated with 200 μg/ml of β-glucan did not exhibit increased MCP-1 production (Fig. 8). These data suggested that P. carinii β-glucan was likely not responsible for the type II AEC MCP-1 response to whole P. carinii. Thus, it is likely that multiple mechanisms of P. carinii-AEC interaction exist.

Fig. 8.

MCP-1 production by P. carinii-stimulated type II AECs is not mediated by β-glucan. Primary type II AECs were isolated and cultured as described above. AECs were stimulated with a 2:1 cyst-to-AEC ratio of P. carinii or 200 μg/ml β-glucan, and culture supernatants were collected 24 h later for MCP-1 ELISA. Untreated AECs were used as controls. Values are means ± SE (n = 3). *P < 0.05 compared with both untreated controls and β-glucan-treated cells. The data are combined from two independent experiments with separate AEC and P. carinii isolations.

P. carinii-stimulated MIP-2 production by AECs is JNK independent.

Prior studies have demonstrated that P. carinii stimulates MIP-2 production by an immortalized murine AEC line through an NF-κB-dependent signaling mechanism (52). Consistent with these previous findings, SSA also blocked MIP-2 production by primary murine type II cells (Fig. 9). In addition, SSA also significantly inhibited MCP-1 production by primary AECs, suggesting that NF-κB is involved (Fig. 4). However, JNK inhibition had no effect on MIP-2 production by type II AECs, suggesting that different signaling requirements exist for P. carinii-induced MIP-2 and MCP-1 production by AECs (Fig. 9). NF-κB activation regulates MIP-2 gene expression, while both the JNK and NF-κB pathway are involved in MCP-1 production by AECs in response to P. carinii.

Fig. 9.

Macrophage inflammatory protein-2 (MIP-2) production by P. carinii-stimulated type II AECs is JNK independent. Primary type II AECs were isolated from mice and cultured for 4 days as described above. Cells were then treated with P. carinii in the absence or presence of specific inhibitors, and culture supernatants were collected 6 h, 12 h, and 24 h later for MIP-2 ELISA. Untreated AECs were used as controls. AECs were stimulated with a 2:1 cyst-to-AEC ratio of P. carinii or an equal amount of the same P. carinii preparation that had been depleted of organisms. Values are means ± SE (n = 3). *P < 0.05 compared with untreated controls, cells treated with P. carinii-depleted preparations, and P. carinii-stimulated cells pretreated with SSA. The data are from one of at least two independent experiments with separate AEC and P. carinii preparations.

Blockade of epithelial JNK signaling blocks the early MCP-1 response in vivo.

Prior studies have demonstrated that recombinant adenovirus vectors can deliver genes specifically to airway and AECs in vivo (44). Therefore, this technique was utilized to deliver DNA encoding the JBD of JIP-1 to the pulmonary epithelium of mice. This peptide acts as a dominant negative inhibitor of JNK signaling. Either PBS, Ad-JBD, or Ad-LUC was intratracheally instilled into the lungs of mice. Mice instilled with PBS were used as uninfected controls. Forty-eight hours later, the mice were inoculated with 1 × 106 P. carinii. At 3 h postinfection, at the time of peak MCP-1 levels in the lung (Fig. 1), the mice were lavaged, and cell-free BAL fluid was collected. Mice pretreated with either PBS or Ad-LUC had similar levels of MCP-1 in the lavage fluid (Fig. 10A). In contrast, mice pretreated with Ad-JBD exhibited a dramatic reduction in lavage MCP-1 levels (Fig. 10A; P < 0.05). In vitro studies found that JNK inhibition did not affect MIP-2 secretion by type II AECs (Fig. 9). This finding was confirmed in vivo. Mice pretreated with either PBS, Ad-JBD, or Ad-LUC all had similar levels of MIP-2 in the lavage fluid (Fig. 10B), indicating that JNK signaling was not required for epithelial production of MIP-2 in vitro or in vivo.

Fig. 10.

Modulation of JNK activity in lung epithelial cells blocks P. carinii-stimulated MCP-1 production in P. carinii-infected mice. Mice were pretreated with Ad-JBD, Ad-LUC, or PBS for 48 h. MCP-1 and MIP-2 levels were measured in the BAL fluid from mice 3 h after P. carinii infection. Values are means ± SE. *P < 0.05 compared with Ad-LUC- and PBS-treated controls; n ≥ 3 for each group. Ad-JBD, recombinant adenovirus vector expressing the JBD of JIP-1 (this peptide interrupts the activation of the downstream JNK signal); Ad-LUC, recombinant adenovirus vector expressing firefly luciferase (used as a control).

DISCUSSION

Located at the boundary between the environment and internal tissues, lung epithelial cells are an important component of the host defense. AECs function not only as a physical barrier but also as biological sensors for invading microorganisms and their products, by producing cytokines, chemokines and other potent inflammatory mediators. The close interaction of P. carinii with the type I pneumocyte is characteristic of PCP (32, 51, 58). In addition, direct and indirect interactions between P. carinii and type II pneumocytes have been demonstrated. Importantly, an increasing body of evidence supports a role for type II AECs in the initiation and targeting of pulmonary inflammation through secretion of proinflammatory cytokines and chemokines. While AEC responses are critical for protective host responses against many infectious agents, the immunomodulatory potential of these cells is also likely to directly impact the immune-mediated lung injury characteristic of PCP. Many studies have focused on the ability of P. carinii to directly damage the alveolar epithelium, but few have examined the potential of AECs to function as cell mediators of P. carinii-driven inflammatory responses. In addition, the difficulties associated with isolating and culturing primary type II AECs have dictated that most of the studies involving P. carinii-AEC interactions utilize transformed cell lines. In the present study we have successfully isolated highly purified primary murine type II cells, and cultured them under conditions that promote the maintenance of a type II phenotype. Using our in vitro system, we have demonstrated that P. carinii stimulates MCP-1 production from type II AECs through a JNK and NF-κB-dependent signaling pathway. This finding was corroborated by in situ hybridization studies demonstrating that P. carinii-stimulated AECs express MCP-1 mRNA in vivo, and also by in vivo studies demonstrating that delivery of a gene encoding the JBD of JIP-1 specifically to the lung epithelium blocked the early MCP-1 response to P. carinii.

This study demonstrated that P. carinii activates the JNK signaling pathway, and alters gene expression in murine primary type II AECs. JNK, p38, and ERK1/2 belong to a superfamily of stress-induced MAPK kinases used by cells to transduce extra-cellular signals into a cellular response (14). The JNK and p38 MAPK are believed to be important in induction of proinflammatory responses and, in contrast to ERK1/2 MAPK, have not been strongly associated with proliferation, transformation and differentiation (24). Consistent with previous data that JNK is involved in MCP-1 production from LPS-stimulated primary microglia (49), our data showed that the JNK specific inhibitors SP600125 and l-JNKI1 block MCP-1 production in P. carinii-stimulated primary AECs. While other studies have demonstrated that P. carinii can stimulate AEC production of the PMN-specific chemokines IL-8 in humans and MIP-2 in rodents (4, 15, 19, 20), a direct role for neutrophils in damaging the lung has not been shown (27, 47). In contrast, MCP-1 is a CC chemokine with the potential to attract and activate T lymphocytes and macrophages, which are critical to both effective immunity and the generation of immune-mediated lung injury. Our prior studies have found that immune and inflammatory cells are recruited specifically to alveolar sites of P. carinii infection in mice, suggesting that the interaction of P. carinii with the alveolar epithelium targets the immune response (55, 56). In vivo studies aimed at defining the role of epithelial MCP-1 production in the generation of PCP-related lung injury are currently underway. Together, these data suggest that the P. carinii-AEC interaction induces changes in AEC gene expression that contributes to the generation and targeting of the host's immune response. In a normal host these changes may lead to the generation of effective immunity, while in a compromised host they may contribute to the immunopathogenesis of PCP.

Consistent with our previous work demonstrating that P. carinii stimulates NF-κB signaling in AECs, this study found that blockade of NF-κB activation reduced MCP-1 production. Activation of NF-κB regulates the expression of a network of mediators involved in activation of inflammatory cells and their recruitment to extravascular tissues. The finding that P. carinii induces MAPK and NF-κB signaling suggests that in addition to the known genes that are altered by the interaction of P. carinii with AECs many other MAPK and NF-κB responsive genes are also regulated. Several groups have shown that there is a crosstalk between the NF-κB and JNK pathways (34). The relevance of the JNK cascade to TNFR-mediated apoptosis is highlighted by the finding that activation of this cascade is controlled by NF-κB (9). The NF-κB-mediated attenuation of JNK signaling is crucial for numerous physiological processes, such as the response of the liver to injury and the survival of cells during an inflammatory reaction, as well as for chronic inflammatory diseases and cancers (34, 37). However, it is still not clear how NF-κB and JNK signaling pathway cooperate in chemokine production during inflammatory responses. Our studies have found that NF-κB signaling is required for both MCP-1 and MIP-2 production by AECs (Figs. 4 and 9) (52). However, JNK inhibition was critical to MCP-1 production by P. carinii-stimulated AECs, but had no affect of MIP-2 production. Thus, multiple epithelial signaling pathways that differentially regulate gene expression are activated by the P. carinii-AECs interaction.

Although the biology of P. carinii has been extensively investigated, the molecular mechanisms by which P. carinii induces inflammatory responses in AECs are not clear. TLR are a group of transmembrane receptors known to be activated by conserved molecular patterns of microbes such as fungi (18, 35). It has been shown TLR2 and TLR4 are known to share a common proinflammatory signal transduction pathway leading to the nuclear translocation of the NF-κB and transcription of various genes (33, 48). TLRs also transduce signals that activate the JNK pathway in response to LPS (31, 36). Furthermore, TLR2 and TLR4 have both been implicated in alveolar macrophage responses to P. carinii (12, 61). Thus, TLRs may mediate P. carinii induced activation of JNK and NF-κB signaling pathways in AECs. The alternate glucan receptor, lactosylceramide, also has been shown to mediate NF-κB activation in response to P. carinii glucan (15). Therefore, it is possible that the interaction of P. carinii with lactosylceramide could also promote AEC inflammatory responses. However, our data indicate that intact P. carinii, but not purified glucan, stimulates an MCP-1 response in purified murine type II cells (Fig. 8).

The consequences of NF-κB and JNK activation in the AECs of P. carinii-infected animals remain unclear. However, it is plausible that altered AEC gene expression promotes the immune-mediated lung injury associated with PCP (54, 55, 56). We have demonstrated that following the immunological reconstitution of P. carinii-infected SCID mice, immune and inflammatory cells are recruited specifically to alveolar regions of infection, and not to uninfected alveoli (54, 56). This finding suggests that the interaction of P. carinii with AECs in vivo produces signals that target the inflammatory response to infected alveoli. Therefore, if the recruitment of inflammatory cells to the lung is a direct consequence of NF-κB and JNK-mediated signal transduction in AECs, then blockade of NF-κB and JNK may alleviate some of the damage caused by immune-mediated lung injury. Several NF-κB inhibitors are already used in humans, including SSA, which is used therapeutically to alleviate the inflammatory consequences of inflammatory bowel disease and rheumatoid arthritis (13, 40). Furthermore, it has been well-documented that NF-κB is important for lymphocyte activation and proliferation (3), and lymphocytes are directly involved in PCP-related lung injury (54). MAPK inhibition is also a very active area of research with regards to the treatment of inflammatory disorders (6, 24). Our previous studies have shown that MCP-1 levels correlate with lung injury during PCP. Therefore, the inhibition of MAPK and/or NF-κB signaling may interfere with the generation of an injurious host response to P. carinii on several fronts, and could provide a promising therapeutic intervention to lessen immune-mediated respiratory impairment during PCP.

In summary, we have demonstrated that stimulation of primary type II AECs with P. carinii induces the production of MCP-1 through JNK- and NF-κB-dependent mechanisms. This finding provides evidence that AECs are potentially important modulators in the lung during PCP. The interaction of P. carinii with AECs may promote the development of immune-mediated lung injury through the production of chemokines and subsequent recruitment of immune cells to infected alveoli. More in-depth studies of the role of epithelial MCP-1, JNK, and NF-κB are needed to provide insight into how AECs may affect immunity to P. carinii and also the role of AEC in promoting the immune-mediated lung injury observed during PCP.

GRANTS

This work was supported by National Heart, Lung, and Blood Institute Grants R01-HL-083761 (T. W. Wright), R01-HL-64559 (T. W. Wright), and P01-HL-71659 (F. Gigliotti) and a Bradford Fellowship awarded to Dr. Jing Wang by the Strong Children's Research Center.

Acknowledgments

We acknowledge the excellent technical support of Stephanie Campbell and Nabilah Khan.

Footnotes

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

REFERENCES

  1. 1a.
  2. 1.
  3. 2.
  4. 3.
  5. 4.
  6. 5.
  7. 6.
  8. 7.
  9. 8.
  10. 9.
  11. 10.
  12. 11.
  13. 12.
  14. 13.
  15. 14.
  16. 15.
  17. 16.
  18. 17.
  19. 18.
  20. 19.
  21. 20.
  22. 21.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
  27. 26.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 33.
  35. 34.
  36. 35.
  37. 36.
  38. 37.
  39. 38.
  40. 39.
  41. 40.
  42. 41.
  43. 42.
  44. 43.
  45. 44.
  46. 45.
  47. 46.
  48. 47.
  49. 48.
  50. 49.
  51. 50.
  52. 51.
  53. 52.
  54. 53.
  55. 54.
  56. 55.
  57. 56.
  58. 57.
  59. 58.
  60. 59.
  61. 60.
  62. 61.
  63. 62.
View Abstract