Substance P Delays Apoptosis, Enhancing Keratitis after Pseudomonas aeruginosa Infection

Abstract

purpose. Apoptosis was examined after Pseudomonas aeruginosa corneal infection in C57BL/6 (B6, susceptible) and BALB/c (resistant) mice.

methods. TUNEL staining, real-time RT-PCR, polymorphonuclear neutrophils (PMNs) and macrophage (Mφ) depletion, and immunostaining were used.

results. Intense TUNEL staining was seen in BALB/c versus B6 cornea at 1 versus 3 days after infection (PI) and correlated with mRNA levels for caspase-3. TUNEL staining (with or without PMN depletion) and PMN immunostaining revealed the PMN as the major apoptotic cell for both groups. Next, B6 mice with high corneal levels of the antiapoptosis neuropeptide, substance P (SP), were treated with the SP antagonist, Spantide I (with/without Mφ depletion), resulting in earlier apoptosis and diminished disease only when Mφs were present. SP interactions with Mφs were explored further by eliciting cells from both groups and stimulating them with lipopolysaccharide (LPS), with or without SP. LPS with SP treatment decreased the number of apoptotic Mφs in B6 but not BALB/c mice and correlated with reduced mRNA expression of NK-1R (major SP receptor) on BALB/c cells. In addition, mRNA expression for IL-12 was upregulated in LPS-stimulated B6 Mφs, although cells from BALB/c mice expressed more IL-10.

conclusions. These studies provide evidence that PMN apoptosis is delayed in the cornea of B6 versus BALB/c mice after bacterial infection; that in B6 mice, blocking SP interaction with the NK-1R promotes earlier apoptosis and improves disease outcome; that Mφs regulate PMN apoptosis; and that Mφs from B6 versus BALB/c mice differ in expression of the NK-1R and cytokines produced after LPS challenge.

The bacterium Pseudomonas aeruginosa is an opportunistic pathogen associated with microbial keratitis, especially in extended-wear contact lens usage. ¹ The outcome of this sight-threatening disease has been largely determined by the host inflammatory response. In this regard, previous work has shown that C57BL/6 (B6) versus BALB/c mice respond to many antigens disparately, with B6 mice classified as dominant T helper type 1 (Th1) and BALB/c mice, dominant Th2, responders. ^{2 3} After a challenge with P. aeruginosa, the cornea of B6 mice perforates (susceptible), although similarly challenged BALB/c (resistant) mice are able to restore corneal clarity. ⁴ The latter result has provided insight that in this model, as in many others, an intense inflammatory response may be necessary to eradicate the pathogen, but if uncontrolled, is detrimental. Thus, tight regulation of infiltrating immune cells, such as macrophages (Mφs) and polymorphonuclear neutrophils (PMNs), and local balance between pro- and anti-inflammatory factors released by these and other cells is paramount for disease resolution and restoration of tissue homeostasis. ⁵ Whether cells undergo necrosis or apoptosis also may contribute to disease outcome. This contribution has been shown in studies examining Leishmania major–induced subcutaneous infection in mice, in which earlier apoptosis of infiltrating PMNs leads to rapid resolution of inflammation. ⁶ P. aeruginosa also was found to induce apoptosis in mouse lung epithelial cells by the endogenous Fas/Fas ligand system and knocking out Fas (or Fas ligand) resulted in rapid development of sepsis in C3H mice. ⁷ These findings suggest that apoptosis is protective against infection and tissue damage. Nonetheless, no studies have tested whether apoptosis of infiltrating cells in P. aeruginosa–induced keratitis differs in susceptible versus resistant mice and whether neuropeptides, such as substance P (SP) are involved. 

SP is an 11-amino acid polypeptide, a member of the tachykinin family. ^{8 9} It mainly functions through its interaction with the NK-1R, its major physiological receptor. ¹⁰ Evidence has shown that SP is a potent antiapoptosis regulator, and it has been shown that SP-enhanced proliferation of the human laryngeal carcinoma cell line HEp-2 and application of an NK-1R antagonist induce tumor cell apoptosis. ¹¹ Moreover, SP promoted mucosal healing from colitis by stimulating antiapoptosis pathways in human colonic epithelial cells. ¹² SP levels are higher in B6 versus BALB/c mice, ¹³ suggesting that this could contribute to the pattern of apoptosis of infiltrated cells, including PMNs and Mφs, and disease outcome after infection. 

Mφs release inflammatory mediators, including cytokines (e.g., IL-12), prostaglandins, and leukotrienes, ^{14 15} and prolonged and poorly regulated activation of these cells may exacerbate inflammation. In this regard, it has been reported ¹⁶ that alveolar Mφ apoptosis decreases PMN recruitment and improves disease outcome in pneumococcal pneumonia. In addition, other evidence has shown that the inflammatory regulating function of Mφs is regulated by SP. ¹⁵ However, whether the apoptotic and cytokine secretory patterns of activated Mφs differs in B6 versus BALB/c mice after P. aeruginosa infection and the subsequent impact on the inflammatory process also has remained unexplored until this study. 

Our studies provide evidence that apoptosis of PMNs is delayed in the cornea of B6 versus BALB/c mice after P. aeruginosa infection; that in B6 mice, blocking SP interaction with the NK-1R promotes earlier apoptosis and improves disease outcome; that PMN apoptosis in vivo is regulated by SP regulation of Mφs; and that B6 versus BALB/c mouse Mφs differ in expression of the NK-1R, response to SP stimulation and cytokine production after LPS challenge. 

Methods

Mice

Female, 8-week-old B6 and BALB/c mice, purchased from the Jackson Laboratory (Bar Harbor, ME), were housed according to National Institutes of Health guidelines. All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Bacterial Infection and Ocular Response

P. aeruginosa strain 19660 was purchased from the American Type Culture Collection (ATCC, Manassas, VA). Cultures were prepared as described before for a final concentration of 1.0 × 10⁶ CFU/μL. ¹⁷ For infection, the mice were anesthetized with ethyl ether and the left central cornea scarified with a 25 \({5}/{8}\) -gauge needle. A 5-μL aliquot containing a 1.0 × 10⁶ CFU/μL bacterial suspension was then applied. Disease was graded at 1, 3, and 5 days after infection (PI), according to an established scale. ^{17 18}  

PMN Depletion

B6 and BALB/c mice (n = 3/group/time) were injected intraperitoneally (IP) with 0.5 mg/mouse of anti-mouse PMN antibody (clone RB6-8C5, IgG2b) 5 hours before corneal infection. Control mice were similarly injected with 0.5 mg/mouse of anti-human HLA DR5 antibody (SFR3-DR5, IgG2b). ¹⁹ The eyes were enucleated for TUNEL staining 18 to 24 hours after infection. 

Mφ Depletion

For depletion of Mφs, prepared liposomes composed of phospholipid bilayers and containing dichloromethylene diphosphonate (clodronate) were prepared ²⁰ and used as described before. ²¹ The clodronate was a gift from Roche Diagnostics, GmbH (Mannheim, Germany). A total of 8 μL of the clodronate-containing liposome suspension was injected subconjunctivally in B6 (n = 32) mice with a 50-μL syringe (Hamilton, Reno, NV) with a 30-gauge needle, as described before. ²¹ On day −4 (day 0 is the day of infection), clodronate-containing liposomes (5 μL/mouse/group) were injected; on day −2, 3 μL more were injected similarly. Half of the mice were treated with Spantide I and control mice were similarly treated with PBS. Corneas were infected (day 0) as described earlier. Eyes (n = 3/group/time) were enucleated on days 1 and 3 PI for TUNEL staining, and corneas (n = 5/group/time) were removed at the same time points for real-time RT-PCR analysis. 

Spantide I Treatment

B6 mice were injected IP with 36 μg/mouse of Spantide I (Sigma-Aldrich, St. Louis, MO) on days −1 and 0 (day of infection) and daily through 5 days PI, ¹³ with or without Mφ depletion. Control mice were similarly injected with PBS. The eyes were enucleated for TUNEL staining, and normal and infected corneas were also collected for real-time RT-PCR. 

Mφ Isolation and Stimulation Assay

Peritoneal Mφs were elicited and isolated from B6 and BALB/c mice, as described before. ^{5 22} Briefly, to induce Mφs into the peritoneal cavity, 1.0 mL of 3% Brewer’s thioglycollate medium (Difco, Detroit, MI) was injected IP in both mouse groups 5 to 7 days before death. The cells were collected by peritoneal lavage and stained with trypan blue, and viable cells (>95%) were counted with a hemocytometer. Mφs were seeded in 12-well plates at a density of 1 × 10⁶ cells/well for RT-PCR analysis, or onto eight-well culture slides (BD BioCoat, Bedford, MA) at a density of 5 × 10⁵ cells/well for TUNEL staining. Nonadherent cells were removed 4 hours later. Isolated Mφs were stimulated with LPS alone (1 μg/mL) or together with SP (10⁻⁸, 10⁻¹⁰, and 10⁻¹² M) for 18 hours. For TUNEL staining, 600 cells per treatment were counted (three separate slides) and the number of TUNEL-positive (apoptotic) cells expressed as a percentage. Cells from 12-well plates were collected and mRNA extracted and assayed by real-time RT-PCR or nested PCR for selected cytokines, caspases, and NK-1R. 

TUNEL Assay

Normal uninfected and infected B6 and BALB/c mouse eyes (n = 3/group/time) were enucleated at 1 and 3 days PI for TUNEL staining (per the manufacturer’s recommendation, TACS TdT kit; R&D System, Minneapolis, MN), to detect corneal apoptotic cells. In another experiment, normal uninfected and infected eyes from B6 mice (n = 3/group/time) treated with Spantide I versus PBS were enucleated at 18 hours and 1, 3, and 5 days PI (with Mφ) and at 1 and 3 days PI (Mφ depleted) for TUNEL analysis. Briefly, eyes were fixed in a 3.7% formaldehyde solution and embedded in paraffin. Sections (10 μm-thick) were cut, deparaffinized, rehydrated, and rinsed with DNase-free water. The sections were permeabilized with proteinase K solution (15 minutes), followed by quenching of endogenous peroxidase with 3% H₂O₂ in methanol (5 minutes). They were incubated with 1× terminal deoxynucleotidyl transferase (TdT enzyme) solution (1 hour) at 37°C in a humidified chamber, to label DNA nick ends. The reaction was stopped with TdT stop buffer (5 minutes). Color was developed with streptavidin-HRP detection solution (10 minutes) followed by incubation with TdT blue labeling solution (5 minutes). The sections were counterstained with nuclear Fast Red (5 minutes). All procedures were performed at room temperature, unless otherwise indicated. Control samples were treated similarly but without TdT enzyme. Samples were photographed by microscope (Axiophot with Axiocam digital imagery; Carl Zeiss Meditec GmbH, Jena, Germany). 

TUNEL and PMN Immunostaining

Infected B6 and BALB/c mouse eyes (n = 3/group/time) were enucleated at 1 and 3 days PI, fixed, embedded, and sectioned. Alternate serial sections were stained for TUNEL, and PMNs were stained by specific immunostaining. For PMN staining, the sections were rehydrated and blocked with rabbit IgG (1:100, Sigma-Aldrich, St. Louis, MO) diluted in 2.5% bovine serum albumin (BSA) for 30 minutes. The sections were then incubated with a primary rat anti-PMN antibody NIMP-R14 (1:80; Abcam, Cambridge, MA), followed by incubation with a biotinylated secondary rabbit anti-rat IgG antibody (1:1000; Vector Laboratories, Burlingame, CA) for 1 hour each and then incubated (30 minutes) with a peroxidase-conjugated staining kit (Extravidin; Sigma-Aldrich) for color development. Control sections were similarly processed, omitting the primary antibody. In TUNEL and PMN staining, the sections were counterstained with Fast Green (1:100; Fisher, Pittsburgh, PA). Sections were photographed (Axiophot; Carl Zeiss Meditec), as described for the TUNEL assay. 

Immunofluorescent Staining and Confocal Microscopy

Normal uninfected and infected B6 and BALB/c mouse eyes (n = 3/group/time) were enucleated at 1 and 3 days PI. The eyes were immersed in optimal cutting temperature compound (OCT, Tissue Tek; Miles, Elkhart, IN) and frozen in liquid nitrogen. Cryostat sections (10-μm thick) were cut and collected onto freshly coated polylysine glass slides. Nonspecific staining was blocked with goat IgG (1:100; Invitrogen-Molecular Probes, Eugene, OR) diluted in 2.5% BSA for 30 minutes. Slides then were incubated with primary rabbit anti-mouse cleaved caspase-3 (Asp 175) antibody (1:200; Cell Signaling Technology, Inc., Danvers, MA) for 1 hour, and incubated for 1 hour with secondary Alexa Fluor 633 goat anti-rabbit IgG (1:1500; Invitrogen-Molecular Probes). The slides were incubated for 2 minutes with nuclear acid stain (Sytox Green, 1:200,000; Fisher Scientific, Pittsburgh, PA). Negative controls were similarly treated, but with omission of the primary antibody, and all procedures were performed at room temperature, unless otherwise indicated. The sections were visualized and digital images captured with a confocal laser scanning microscope (TSC SP2; Leica Microsystems, Bannockburn, IL). 

Real-Time RT-PCR

Normal, uninfected, and infected corneas from B6 and BALB/c mice (n = 5/group/time) were removed for real-time RT-PCR analysis. The corneas were collected at 1 and 3 days PI for detection of caspase-3 and -9, and at 7 and 18 hours and 1 and 3 days PI for Bcl-2 analysis. In another series of experiments, normal uninfected and infected corneas of Spantide I- and PBS-treated B6 mice (n = 5/group/time) were collected at 1, 3, and 5 days PI for detection of mRNA expression for caspase-3 and -9. Spantide- versus PBS-treated mice also were depleted of Mφs (n = 5/time/group) and the corneas collected at 1 and 3 days PI for detection of caspase-3 and -9. For these experiments, the corneas were removed and stored in RNA stabilization solution (RNAlater; Ambion Inc., Austin, TX) at −20°C before processing. Total corneal RNA was extracted (RNA STAT-60; Tel-Test, Friendsville, TX), per the manufacturer’s instructions, precipitated with isopropanol (0.5 mL/1 mL RNA STAT-60; Sigma-Aldrich), and washed with 75% ethanol. One microgram of each RNA sample was reverse transcribed with M-MLV reverse transcriptase (Invitrogen, Carlsbad, CA) in a 20-μL volume. cDNA products were diluted 1:25 with diethylpyrocarbonate (DEPC)-treated H₂O and 2 μL of each cDNA dilution was used for real-time RT-PCR (20-μL reaction volume). Real-time RT-PCR was performed with a single-color real-time RT-PCR detection system (MyiQ; BioRad, Hercules, CA). SYBR Green mastermix (iQ Supermix; Bio-Rad) was used for each PCR reaction with a primer concentration of 5 μM. The PCR parameters used were preheated at 95°C for 3 minutes followed by 40 cycles of amplification (15 seconds at 95°C and 40 seconds at 60°C). Primer pair sequences, designed with Primer 3 software (Whitehead Institute for Biomedical Research; MIT, Cambridge, MA), are shown in Table 1 . Changes in mRNA level were calculated after normalization with β-actin. 

Nested PCR

To detect NK-1R in Mφ samples, we performed nested PCR analysis, as described before. ¹³ In brief, 2 μL of sample cDNA was added to a total volume of 50 μL PCR buffer containing appropriate concentrations of Taq DNA polymerase (Invitrogen), dNTPs, and external NK-1R primers (450 bp). Thirty-five cycles were run, with 95°C denaturation, 58°C annealing, and 72°C extension temperatures. Then, 1 μL of a 1:20 dilution of the previous PCR reaction mixture was combined with internal NK-1R primers (432 bp) and other reaction components at appropriate concentrations. PCR cycles were repeated, and 15 μL of each amplified sample was electrophoresed on an ethidium bromide-stained agarose gel and visualized under UV illumination. Primer sequences are shown in Table 1 . 

Statistical Analysis

An unpaired, two-tailed Student’s t-test was used to analyze real-time RT-PCR and in vitro Mφ assay data for significance (P < 0.05). Experiments were performed twice, to ensure reproducibility, and data from a representative experiment are shown. 

Results

TUNEL Staining

At 1 day PI, BALB/c mouse cornea showed more intense TUNEL-positive staining than did the B6 mice (Fig. 1) . However, at 3 days PI, the cornea of the B6 mice had more intense TUNEL staining than did those of BALB/c mice. No TUNEL staining was detected in normal, uninfected mouse cornea, nor in control sections in the absence of the TdT enzyme in either group (data not shown). 

TUNEL and PMN Staining

TUNEL and PMN staining (Figs. 2 3)showed that infiltrating PMNs were the predominant cells undergoing apoptosis in both groups of mice and confirmed that the kinetics of apoptosis of these cells was disparate, reflecting the TUNEL staining described earlier. Sequential sections showed that few PMNs were TUNEL positive at 1 day PI in B6 mice (Figs. 2A 2B) , whereas many PMNs were TUNEL positive in BALB/c mice (Figs. 2C 2D) . At 3 days PI, more TUNEL-positive PMNs were seen in B6 cornea (Figs. 3A 3B)than in the cornea of BALB/c mice (Figs. 3C 2D) . No positive staining was detected when omitting the TdT enzyme for TUNEL or the primary antibody for PMN staining (data not shown). 

PMN Depletion and TUNEL

To confirm further that the PMNs are the major cell type undergoing apoptosis, we depleted the cells in B6 and BALB/c mice before corneal infection. Figures 4A 4B 4C 4Dshow that after PMN depletion, no TUNEL-positive cells were detected in the cornea of either group at 1 day PI. In contrast, mice treated with an irrelevant antibody showed a pattern of TUNEL-positive staining in the cornea as shown in Figure 1 . No TUNEL-positive staining was observed after the TdT enzyme was omitted (data not shown). 

Real-Time RT-PCR

The TUNEL studies led to testing mRNA expression of apoptosis-related genes in normal and infected B6 and BALB/c mouse cornea by real-time RT-PCR (Figs. 5A 5B 5C) . Caspase-3 (effector caspase) mRNA expression (Fig. 5A)was significantly increased at 1 day PI in BALB/c over B6 cornea (P = 0.007), whereas at 3 days PI, the corneas of B6 versus BALB/c mice expressed significantly more mRNA for caspase-3 (P = 0.004). Expression of caspase-9 (Fig. 5B ; initiator caspase) was significantly elevated at 1 day PI in BALB/c over B6 mouse corneas (P = 0.01). However, no significant change in expression was detected between the two groups at 3 days PI. Bcl-2 (antiapoptosis gene; Fig. 5C ) was elevated in the cornea of B6 versus BALB/c mice at 7 hours (P = 0.32) and was significantly upregulated at 18 hours PI (P = 0.03). No difference in expression levels of Bcl-2 was detected at later time points (1 and 3 days PI). mRNA corneal expression levels for these three genes were not significantly different in the normal uninfected cornea of either mouse group (Figs. 5A 5B 5C) . 

Immunofluorescent Staining and Confocal Microscopy

The mRNA data for caspase-3 were confirmed by performing immunostaining and confocal laser scanning microscopy for detection of activated caspase-3 (Fig. 6A 6B) . Immunostaining was more intense at 1 day PI in BALB/c over B6 mouse cornea (Fig. 3A) . In contrast, at 3 days PI (Fig. 3B) , more intense immunostaining was detected in B6 cornea. Negative control sections, processed in the absence of primary antibody, showed no detectable specific staining and appeared similar to Sytox Green–stained sections. 

Spantide I Treatment and TUNEL Staining

Spantide I treatment of B6 mice blocks SP interaction with the NK1-R and has been shown to decrease bacteria-induced disease, ¹³ but whether apoptosis is involved has not been examined. Thus, to test whether SP regulates apoptosis and whether its effects are mediated through the Mφs in the infected B6 mouse cornea, TUNEL staining was performed in Spantide I- versus PBS-treated (Fig. 7A)and in Spantide I- versus PBS-treated (with Mφ depletion, Figs. 8A 8B 8C 8D ) mice. In the former, at 18 hours and 1 day PI, more intense TUNEL staining was detected in the corneal epithelium and anterior corneal stroma in Spantide I- versus PBS-treated mice. At 3 days PI, the cornea of PBS- versus Spantide I-treated mice showed intense TUNEL staining, as anticipated. At 5 days PI, few apoptotic cells were identified in tissue sections from either group. In the latter (Spantide I- versus PBS-treated), few to no TUNEL-positive cells were detected in the cornea at 1 or 3 days PI with Mφ depletion. No TUNEL-positive cells were detected in control tissue sections from any treatment group in the absence of exposure to the TdT enzyme (data not shown). 

Real-Time RT-PCR

To confirm TUNEL staining, we tested corneal mRNA expression of caspase-3 and 9 in Spantide I- and PBS-treated B6 mice at 1, 3, and 5 (caspase-3 only) days PI (Figs. 7B 7C)and in Spantide I- and PBS-treated mice (with Mφ depletion, Figs. 8E 8F ). As shown in Figure 7B , caspase-3 mRNA levels were significantly increased in cornea at 1 day PI in Spantide I- over PBS-treated mice (P = 0.05). At 3 days PI, the cornea of PBS-treated mice expressed significantly more caspase-3 compared with Spantide I-treated mice (P < 0.001). No significant differences were detected between the two groups at 5 days PI (P = 0.61). No differences were detected in mRNA levels of caspase-9 at 1 and 3 days PI in Spantide I- versus PBS-treated mice. In contrast, in Spantide I- versus PBS-treated mice (with Mφ depletion), no significant differences in mRNA levels of either caspase-3 or -9 were detected at 1 or 3 days PI (Figs. 8E 8F) . 

TUNEL Staining of Stimulated Mφ

Because Mφs are also important cells in the host response to bacterial keratitis, we next tested whether SP differentially regulated apoptosis of Mφs from B6 versus BALB/c mice. For this, peritoneally elicited cells from the two groups were exposed to LPS (1 μg/mL) and/or SP (10⁻⁸, 10⁻¹⁰, and 10⁻¹² M) for 18 hours. TUNEL staining detected apoptosis (600 cells per treatment were counted) and the percentage of apoptotic cells were calculated per treatment. LPS+SP (at 10⁻¹⁰ and 10⁻¹² M) significantly decreased the number of apoptotic cells from B6 mice compared with LPS stimulation alone (P = 0.003 for both, Fig. 9A ). In contrast, SP treatment, at all tested concentrations, did not significantly change the number of TUNEL-positive cells in LPS-stimulated BALB/c cells compared with LPS treatment alone (Fig. 9B) . 

Nested PCR Assay

The disparate expression of the vasoactive intestinal peptide receptor (VIPR1) on Mφ between the two mouse groups has been reported before, and it is important in disease resolution. ⁵ We next tested Mφs for NK-1R expression (Fig. 9C) . mRNA expression of the NK-1R was strongly expressed in agarose gels of PBS- versus LPS (1 μg/mL)-treated B6 mouse Mφs. In contrast, the BALB/c mice exhibited less intense mRNA expression of the NK-1R in PBS treated Mφ, and expression was not detectable after LPS stimulation. 

Real-Time RT-PCR for Mφ-Derived Cytokines

Because of the disparate expression patterns of the NK1-R on Mφs from the two groups, we next tested mRNA expression levels of IL-10 and -12, cytokines that also are disparately upregulated in the cornea after infection. ^{1 13 21} After LPS stimulation (Fig. 10A 10B) , mRNA expression for IL-10 was significantly upregulated in BALB/c compared with B6 mouse Mφs at all concentrations tested except 100 ng/mL (P = 0.002, P < 0.001, and P = 0.002, respectively). In contrast, mRNA expression of IL-12 was significantly increased in B6 over BALB/c mouse Mφs after similar LPS stimulation and at all concentrations (P = 0.001, P < 0.001, P = 0.001, and P = 0.01). No difference in expression levels were detected for either gene between the two groups after PBS treatment. 

Discussion

The infected cornea of resistant BALB/c mice showed earlier apoptosis of PMNs compared with susceptible B6 mice (1 vs. 3 days PI), and the disparate pattern of apoptosis was confirmed by several approaches, including TUNEL staining and real-time RT-PCR analysis. These data are the first to show that a delay in apoptosis in bacterial keratitis contributes to corneal perforation and are complimentary to those in other studies that have provided evidence that the ExoS toxin of P. aeruginosa can induce apoptosis in host cells ²³ and that the bacterium triggers apoptosis in lung epithelial cells by activating Fas/Fas ligand. ⁷ In the latter study, the induced epithelial apoptosis was found to be protective in vivo, as deficiency of Fas or Fas ligand resulted in fatal sepsis in infected mice. In addition, we found that Bcl-2, an antiapoptosis gene, was significantly upregulated in B6 mouse cornea at 18 hours PI, suggesting that the delayed onset of apoptosis in B6 mouse cornea may be, in part, due to upregulation and signaling of this gene. These data also are consistent with previous studies showing that overexpression of Bcl-2 reduces lymphocyte apoptosis in P. aeruginosa–induced pneumonia. ²⁴  

In the process of apoptosis, execution of cells largely depends on proteolytic cleavage and activation of caspase-3. ²⁵ In this regard, we showed that the BALB/c versus B6 mouse cornea expressed more intense staining for activated caspase-3 at 1 day PI compared with the delayed peak intensity in B6 mice. The two mouse groups also were disparate in the expression of caspase-9, with significantly more mRNA expression in BALB/c over B6 mice. Although the theory suggested by these data is only hypothetical, it may be that different pathways of apoptosis are operative in the infected cornea of the two groups of mice. Others ²⁶ have reported that in an LPS-induced acute lung injury mouse model, caspase-1 deficiency delayed PMN apoptosis and prolonged the inflammatory response. Our studies, with TUNEL and PMN immunostaining, as well as PMN depletion, provided evidence that the corneal apoptotic cells identified in both groups of mice were predominantly PMNs and confirmed that apoptosis of these cells is delayed in B6 mice. We hypothesize that earlier apoptosis of PMNs in resistant BALB/c mice is consistent with effective elimination of invading bacteria, while inducing minimal tissue damage due to inflammation. ⁶  

The balance between apoptosis and cell survival, as well as the tissue milieu and timing of apoptosis, is critical in immune defense. In this regard, the neuropeptide SP is a potent antiapoptosis regulator and can exacerbate inflammation. ²⁷ SP, a member of the tachykinin neuropeptide family, ^{8 9} is widely distributed throughout the body. In this regard, the murine cornea contains SP nerve fibers, but their distribution is disparate, with B6 over BALB/c mice. SP levels are increased after infection in both groups, but for the most part with higher levels in B6 mouse cornea. ¹³ An exception to this was seen at 5 days PI. At this time, SP levels peaked in the BALB/c mouse cornea, which has been shown previously to correlate with peak levels of NK cell IFN-γ production and disease resolution. ²⁸  

Tachykinin neuropeptide family members, namely SP, neurokinin A (NKA), neurokinin B (NKB), and others, function by their interaction with tachykinin receptors, including NK-1, -2, and -3, but SP is the major ligand for the NK-1R type. ¹⁰ SP has been shown to stimulate phosphorylation of the antiapoptosis molecule Akt in colonic mucosa both in vivo and in vitro, preventing apoptosis in humans with colitis. ¹² In another in vitro study, ²⁹ SP induced p53, Bcl-2 and nitric oxide expression in peritoneal Mφ, blocking apoptotic signals. These data are consistent with our findings that Bcl-2 was significantly upregulated in B6 versus BALB/c mouse cornea at 18 hours PI, but we did not test for nitric oxide or p53 levels. Bockmann et al. ²⁷ also reported that SP delays spontaneous apoptosis of PMNs in a dose-dependent fashion by its interaction with the NK-1R in vitro, and that this effect is inhibited by application of the NK-1R antagonist GR82334. 

In this regard, B6 mice treated with another NK-1R antagonist, Spantide I, showed significantly improved disease outcome, ¹³ and the present study provides evidence that an earlier onset of apoptosis, similar to the pattern observed in naturally resistant BALB/c mice, may in part, account for the decreased pathogenesis. Consistent with earlier onset of apoptosis, mRNA expression of caspase-3 also was significantly upregulated earlier in the cornea of Spantide I, versus control treated animals. In contrast, caspase-9 levels were not different at 1 or 3 days PI, suggesting that earlier time periods would need to be tested to confirm or negate these data. The data for caspase-3, however, suggest that the protective mechanism of Spantide I treatment in B6 mice involves induction of earlier PMN apoptosis in the infected cornea, with less bystander tissue damage. Whether in B6 mice the effects of Spantide I are directly mediated via the PMNs or indirectly through Mφ regulation of PMNs was also tested. Clodronate depletion of Mφ with or without Spantide I treatment revealed that in the absence of Mφs, apoptosis was reduced or absent in the cornea. In this regard, apoptosis is an effective regulatory mechanism that can eliminate cells other than PMNs, including activated Mφs. These cells are potent immune-regulating cells that function by the release of an array of mediators, including proinflammatory cytokines such as TNF-α, IL-12, IL-1, and IL-6 and anti-inflammatory cytokines such as IL-10 and TGF-β. ¹⁴ However, unregulated and imbalanced production of Mφ-derived pro- and anti-inflammatory cytokines may exacerbate inflammation and result in enhanced disease. ³⁰ Therefore, Mφs from both groups of mice were incubated in the presence of SP together with LPS or with LPS alone. Significantly fewer apoptotic cells were detected in cells from B6 mice in the presence of SP and LPS versus LPS alone, whereas the same combined treatment (SP and LPS) did not decrease the number of LPS-induced apoptotic Mφs in BALB/c mice. To determine the mechanism for the disparate response to SP treatment between Mφs from the two groups, we conducted comparative testing of the expression levels of the NK-1R. Although cells from both groups expressed the receptor, with a slightly weaker signal in BALB/c cells, after LPS stimulation, mRNA expression for the NK-1R was detected only in cells from B6 mice. These data suggest that the possible mechanism for the absence of the antiapoptotic effects observed in BALB/c Mφs after SP treatment may involve a low level of NK-1R expression on the cells and possible rapid depletion of the receptor on LPS stimulation. In this regard, VIPR1, the major receptor for the neuropeptide VIP, was also reported to be expressed disparately in Mφ from B6 (less) versus BALB/c (more) mice. ⁵ In addition, others have reported that after LPS treatment, IL-1RII, a negative regulator of the TLR/IL-1R superfamily, is downregulated in monocytes. ^{31 32} Our own work also showed that expression of Ig IL-1R-related molecule (SIGIRR) is significantly decreased at 12 hours PI (mRNA level) and at 1 day PI (protein level) after bacterial infection. ³³ Collectively, these data, suggest that certain signaling molecules are downregulated after LPS or bacteria induced inflammation, due either to consumption of the molecule or other regulatory processes. 

Mφs may also provide distinct activation signals for Th1/Th2 differentiation. In this regard, others ³⁴ have reported that L. major–infected Mφs enhance the proliferation and IL-4 secretion of Th2 T cells, but inhibit the response of Th1 T cells. When testing for this possibility, we detected that Mφs from B6 mice expressed significantly more IL-12, whereas BALB/c Mφ expressed more IL-10 after LPS stimulation. IL-10 appears protective in the BALB/c cornea, as after subconjunctival injection of clodronate-containing liposomes to deplete these cells, higher levels of IFN-γ and lower levels of IL-10 were detected, and resistant mice were converted to the susceptible phenotype. ²¹ It was also reported that recombinant (r)VIP-treated B6 mice showed improved disease outcome and increased IL-10 expression after P. aeruginosa corneal infection. ⁵ Furthermore, the data suggest that SP, which acts in an antiapoptotic manner toward activated B6 mouse Mφs, may also enhance an IL-12 driven, Th1 type immune response and thus further contribute to the susceptibility of this mouse strain to P. aeruginosa infection. 

In summary, we provide evidence that susceptible B6 mice show delayed apoptosis of PMNs in the cornea compared with resistant BALB/c mice after P. aeruginosa infection; that SP delays apoptosis in the infected corneas of B6 mice in vivo, and that this antiapoptotic effect can be reversed with the NK-1R antagonist, Spantide I, leading to improved disease outcome, which is abrogated in the absence of the Mφs. We also show in vitro that SP is antiapoptotic in LPS-stimulated B6 but not BALB/c mouse Mφs and that the different response may reflect a weaker expression level of the NK-1R in BALB/c cells. Our data further show that activated B6 mouse Mφ production of IL-12 may drive a Th1 response only in B6 mice. 

 Supported by National Institutes of Health Grants R01 EY02986 and P30 EY04068.

 Submitted for publication February 20, 2008; revised May 12, 2008; accepted August 21, 2008.

 Disclosure: Z. Zhou, None; R.P. Barrett, None; S.A. McClellan, None; Y. Zhang, None; E.A. Szliter, None; N. van Rooijen, None; L.D. Hazlett, None

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

 

Table 1.

 

Nucleotide Sequence of the Specific Primers Used in PCR Amplification

Gene	Primer Sequence	Sense
β-Actin	5′-GAT TAC TGC TCT GGC TCC TAG C-3′	F
	5′-GAC TCA TCG TAC TCC TGC TTG C-3′	R
Caspase 3	5′-TGG GCC TGA AAT ACC AAG TC-3′	F
	5′-AAA TGA CCC CTT CAT CAC CA-3′	R
Caspase 9	5′-GCC AGA GGT TCT CAG ACC AG-3′	F
	5′-TCC CTG GAA CAC AGA CAT CA-3′	R
NK-1R (450bp)	5′-TGG ACT CTG ATC TCT TCC TGC CCA ACA-3′	F
	5′-GGA CCC AGA TGA CAA AGA TGA CCA ACA-3′	R
NK-1R (432bp)	5′-CCA ACA CCT ACT CCA CCA ACA CTT CTG-3′	F
	5′-GGA CCC AGA TGA CAA AGA TGA CCA CTT-3′	R
IL-10	5′-TGC TAA CCG ACT CCT TAA TGC AGG AC-3′	F
	5′-CCT TGA TTT CTG GGC CAT GCT TCT C-3′	R
IL-12	5′-GGT CAC ACT GGA CCA AAG GGA CTA TG-3′	F
	5′-ATT CTG CTG CCG TGC TTC CAA C-3′	R
Bcl-2	5′-GGA CTT GAA GTG CCA TTG GT-3′	F
	5′-AGC CCC TCT GTG ACA GCT TA-3′	R

Figure 1.

 

Apoptosis identified by TUNEL staining. At 1 day PI, more TUNEL labeling (dark blue) was detected in BALB/c than in B6 mouse cornea. At 3 days PI, B6 corneas showed more TUNEL staining than did those of BALB/c mice. No apoptotic cells were detected in normal, uninfected B6 or BALB/c corneas (data not shown). Magnification: low mag, ×38; high mag, ×75.

Figure 2.

 

Serial sections of TUNEL alternating with PMN immunohistochemical staining at 1 day PI. Numerous infiltrating PMNs (orange) were detected in B6 (A) and BALB/c (C) corneas. More TUNEL labeling (dark blue cells) was detected in BALB/c (D) versus B6 (B) corneas. Insets (higher magnification of boxed areas in full images, with arrows); typical overlapped regions in two sequential serial sections. In B6 mouse corneas, only a few PMNs were TUNEL positive (A, B, arrows), whereas in BALB/c mice, more PMNs were TUNEL positive (C, D, arrows). No positive staining was detected when the TdT enzyme for TUNEL or the primary antibody for PMN staining was omitted (data not shown). Magnification: full images, ×75; insets, ×180.

Figure 3.

 

Serial sections of TUNEL combined with PMN immunohistochemical staining at 3 days PI. Numerous infiltrating PMNs were detected both in B6 (A) and BALB/c (C) corneas, and more TUNEL labeling was detected in B6 (B) compared with BALB/c (D) corneas at 3 days PI. Insets: magnified images of the corresponding regions in two sequential serial sections. At this time, most PMNs in B6 mouse cornea were TUNEL positive compared with few cells in BALB/c cornea (arrows). Negative controls, omitting TdT enzyme or primary anti-PMN antibody, showed no positive staining in either case (data not shown). Magnification: full images, ×75; inset, ×180.

Figure 4.

 

PMN depletion. After PMN depletion, few infiltrating cells were detected at 1 day PI in the cornea of B6 (A) or BALB/c mice (C). No TUNEL-positive cells were detected in either cornea. In contrast, as expected for control treated mice at 1 day PI, few TUNEL-positive cells were detected in B6 (B), whereas numerous cells were seen in BALB/c mouse corneas (D). No positive staining was detected when the TdT enzyme was omitted in the TUNEL assay (data not shown). Magnification, ×75.

Figure 5.

 

Corneal mRNA levels of apoptosis-related genes. (A) mRNA expression of caspase-3 was significantly increased at 1 day PI in BALB/c versus B6 corneas (P = 0.007). At 3 days PI, B6 mice expressed more corneal caspase-3 than did BALB/c mice (P = 0.004). (B) BALB/c mouse cornea showed significantly more caspase-9 mRNA expression at 1 day PI than did those of B6 mice (P = 0.01). There was no difference between the two groups at 3 days PI (C) Bcl-2 gene expression was higher at 7 (P = 0.32) and significantly so at 18 (P = 0.03) hours PI in B6 versus BALB/c mouse cornea. No significant difference in corneal Bcl-2 mRNA expression was detected between the two groups at 1 and 3 days PI. No difference was detected between normal corneas from either group for any of the genes.

Figure 6.

 

Confocal microscopy. At 1 day PI, more intense activated caspase-3 staining (blue) was detected in BALB/c versus B6 mouse cornea (A, top two panels), while at 3 days PI, B6 mice expressed more intense staining (B, top two panels). Sytox green nuclear labeling (A, B, bottom) was similar to staining without addition of the primary antibody. Magnification, ×45.

Figure 7.

 

Spantide I treatment. (A) TUNEL staining detected an earlier onset of apoptosis in Spantide I- versus PBS-treated B6 mouse cornea after infection. At 18 hours and 1 day PI, the cornea of Spantide I-treated B6 mice showed more intense TUNEL-positive staining than did those of PBS-treated B6 mice. At 3 days PI, Spantide I-treated B6 mice showed less TUNEL-positive staining in the cornea than did PBS-treated mice. By 5 days PI, both groups had few apoptotic cells in the cornea, similar to the normal, uninfected cornea of either group (data not shown). Magnification, ×75. (B) Corneal mRNA expression of caspase-3 was significantly increased at 1 day PI in Spantide I- versus PBS-treated B6 mice (P = 0.05). At 3 days PI, however, PBS-treated mice expressed significantly higher mRNA levels of caspase-3 in the cornea compared with Spantide I-treated mice (P < 0.001). No difference was detected at 5 days PI. (C) Corneal mRNA expression of caspase-9 was not significantly different at 1 and 3 days PI in Spantide I- versus PBS-treated B6 mice.

Figure 8.

 

Mφ depletion and Spantide I treatment. (A–D) TUNEL staining detected rare/few apoptotic cells in the cornea at 1 day PI in (A) Spantide I- or (B) PBS-treated B6 mice (both Mφ depleted). At 3 days PI, (C) Spantide I- and (D) PBS-treated B6 mice similarly showed slightly increased TUNEL-positive staining in the cornea. No positive staining was detected when the TdT enzyme was omitted in TUNEL staining (data not shown). Magnification, ×75. (E, F) No difference in corneal mRNA expression of caspase-3 (E) or -9 (F) was detected at 1 and 3 days PI in Spantide I- versus PBS-treated B6 mice (both Mφ depleted).

Figure 9.

 

Apoptosis of peritoneally elicited Mφs after in vitro LPS (1 μg/mL, with or without SP stimulation. (A) TUNEL staining was performed on Mφ elicited from B6 mice and the percentage of TUNEL-positive cells was calculated per treatment. In the presence of SP (only at 10⁻¹⁰ and 10⁻¹²M), LPS stimulation induced significantly fewer TUNEL-positive cells than did LPS stimulation alone (P = 0.003 and P = 0.003, respectively). (B) TUNEL staining was performed on BALB/c mouse Mφs, and the percentage of TUNEL-positive cells was calculated per treatment. SP treatment (10⁻⁸, 10⁻¹⁰, and 10⁻¹² M) did not significantly decrease the number of LPS-induced apoptotic cells compared with LPS stimulation alone. (C) Nested PCR was performed to detect mRNA expression of NK-1R in peritoneally elicited Mφs. NK-1R mRNA expression was detected in B6 and BALB/c mouse Mφs after PBS treatment, with a stronger signal in cells from B6 mice. After LPS stimulation, mRNA expression for NK-1R was lower in both groups and was detectable only in cells from B6 mice.

Figure 10.

 

mRNA expression of IL-10 and -12 in peritoneal Mφ. (A) mRNA expression of IL-10 was significantly increased in BALB/c compared with B6 mouse Mφs after LPS treatment at all tested concentrations except 100 ng/mL (P = 0.002, P < 0.001, and P = 0.002, respectively). No difference was detected between the groups with PBS treatment. (B) mRNA expression of IL-12 was significantly increased in B6 versus BALB/c mouse Mφs after LPS stimulation at all tested concentrations (P = 0.001, P < 0.001, P = 0.001, and P = 0.01, respectively). No difference was detected between the groups with PBS treatment.