Gamma Interferon Does Not Enhance Clearance of Pseudomonas aeruginosa but Does Amplify a Proinflammatory Response in a Murine Model of Postseptic Immunosuppression

Mice.Six- to eight-week-old male wild-type (C57BL/6) and IFN-γ KO (B6.129S7-Ifng^tm1Ts/J) mice were purchased from Jackson Laboratories (Bar Harbor, Maine) and were housed in the animal care facility at the University of Texas Medical Branch (Galveston, Tex.). The animals were allowed to acclimate for 1 week prior to use. Animal protocols were approved by the Institutional Animal Care and Use Committee and met National Institutes of Health guidelines for the care and use of experimental animals.

Nonlethal CLP.Mice were anesthetized with 2.5% isoflurane in an induction chamber and maintained under anesthesia by delivery of isoflurane through a mask. The mice were positioned in dorsal recumbancy, and the ventral abdominal walls were shaved and prepared with 70% isopropanol and 1% iodine. A ventral midline incision (∼1 cm) was made to allow exteriorization of the cecum, and cecal luminal contents were manipulated from the apex to the base of the cecum. The cecum was ligated 1 cm from the apex with 3-0 silk, and a 25-gauge needle was used to penetrate the cecum in a through-and-through fashion. Sham surgery, in which the cecum was exteriorized and manipulated as described but not ligated or punctured, was performed on additional animals.

Administration of IFN-γ to post-CLP mice.Some groups of mice were given subcutaneous injections of 1 mg (∼8,400 U) of recombinant murine IFN-γ (R&D Systems, Minneapolis, Minn.) at the time of challenge with P. aeruginosa. Preliminary studies were performed to determine a dose of recombinant IFN-γ that would approximate the serum IFN-γ response measured for normal or sham mice after Pseudomonas administration (see Fig. 1A).

• Open in new tab
• Download powerpoint

FIG. 1.

Administration of IFN-γ to post-CLP mice attenuated elevated IL-10 production and increased serum concentrations of IL-12 and TNF-α after P. aeruginosa challenge. Mice were subjected to either CLP or sham surgery. Five days later, the mice were given intravenous injections of saline or 5 × 10⁷ CFU of P. aeruginosa and were sacrificed 6 h later for collection of blood samples. Plasma cytokine concentrations were measured by enzyme-linked immunosorbent assay. Graphs show P. aeruginosa-induced plasma concentrations of (A) IFN-γ, (B) IL-10, C) IL-12, and D) TNF-α concentrations in sham and post-CLP mice. *, significantly different from sham group; #, significantly different from CLP group (sham group, n = 4; CLP group, n = 12; CLP + IFN group, n = 14).

Microbiology. P. aeruginosa (ATCC 19660), a clinical isolate originally obtained from a septic patient, was inoculated into tryptic soy broth and allowed to replicate overnight in a shaking 37°C incubator. The bacteria were pelleted by centrifugation. The supernatant was aspirated, and the pelleted bacteria were washed with 10 ml of sterile 0.9% saline and repelleted. The bacteria were resuspended in saline, and the viable number of CFU in this suspension was determined by culturing serial dilutions overnight on tryptic soy agar. Serial dilutions in sterile 0.9% saline were made to produce the desired bacterial concentration for intravenous injection.

To perform the Pseudomonas challenge, mice were anesthetized with isoflurane and placed in dorsal recumbancy. The penis was extruded, and 0.1 ml of either saline alone (controls) or saline with 5 × 10⁷ CFU (sublethal dose) or 1 × 10⁸ to 4 × 10⁸ CFU (lethal doses) of P. aeruginosa was injected intravenously through the dorsal vein of the penis. The mice were returned to their cages. Some of the mice were observed for mortality during the next 7 days. The remaining mice were sacrificed under isoflurane anesthesia 6 h after Pseudomonas injection for collection of samples.

After euthanasia, lungs were excised aseptically and homogenized in sterile saline at a 1:10 ratio (weight/volume). Serial dilutions of the homogenates were plated on tryptic soy agar and MacConkey's agar and were incubated at 37°C. CFU were counted after 24 h of incubation and recounted after another 24 h. Gram staining and API 20 E biochemical strips (BioMerieux, Hazelwood, Mo.) were used to identify bacterial colonies and to confirm the strain of P. aeruginosa used for injection.

Splenocyte preparation and flow cytometry.Spleens were aseptically harvested and transferred to six-well culture plates containing RPMI 1640 supplemented with 10% fetal calf serum and antibiotics. Spleens were homogenized and passed through a nylon mesh strainer. Red blood cells were lysed (erythrocyte lysis kit; R&D Systems), and the remaining splenocytes were washed in phosphate-buffered saline (PBS) prior to counting on a hemocytometer. Fluorochrome-conjugated antibodies against CD3, NK1.1, F4/80 (Caltag, Burlingame, Calif.) and CD11c (BD Biosciences) were used to identify T cells, NK cells, macrophages, and dendritic cells, respectively. IL-12 receptor cell surface expression was analyzed using antibodies against the β1 subunit of the IL-12 receptor (BD Biosciences). Splenocytes (10⁶ in 0.1 ml of PBS) were incubated with 0.5 μg of fluorochrome-conjugated antibodies for 30 min at 4°C and washed in PBS, and samples were fixed in 1% paraformaldehyde. Cells were analyzed in a FACSort flow cytometer (Becton Dickinson, Franklin Lakes, N.J.), and specific staining was determined by comparison with appropriate antibody isotype controls.

Quantitation of plasma cytokines.Enzyme-linked immunosorbent assay kits were used in the sublethal Pseudomonas challenge experiments to assay plasma IL-12 p70 (R&D Systems) and IL-10, IFN-γ, and tumor necrosis factor alpha (TNF-α) (eBiosciences, San Diego, Calif.). In subsequent experiments, plasma cytokines were quantified by multiplex cytokine technology, using the Bio-Plex System (Bio-Rad, Hercules, Calif.) per the manufacturer's instructions. Briefly, plasma was incubated with spectrally addressed polystyrene beads coated with cytokine-specific monoclonal antibodies. After the beads were washed, a second set of fluorochrome-labeled cytokine-specific antibodies were added. The beads were again washed, and cytokine levels were determined by measuring fluorescent signal following laser excitation.

Statistical analyses.All statistical analyses were performed by using the GraphPad InStat software package. An unpaired, two-tailed Student t test was used to compare CLP and IFN-γ-treated CLP groups. Multiple groups were analyzed by analysis of variance, followed by a Tukey-Kramer multiple comparisons test. Incidences of positive lung culture were compared by chi-square analysis. Statistical analyses of survival curves were performed by using the log rank test. A P value of 0.05 or less was considered significant.

Administration of IFN-γ attenuated the IL-10 response and increased the IL-12 response to an intravenous challenge with a nonlethal dose of P. aeruginosa but did not improve bacterial clearance for post-CLP mice.Serum IFN-γ concentrations are undetectable in mice 5 days after CLP or sham CLP without further bacterial challenge (38). Mice that were subjected to CLP and 5 days later were challenged with Pseudomonas (5 × 10⁷ CFU, given intravenously) had a significantly lower IFN-γ response than sham CLP mice after challenge with Pseudomonas. IFN-γ was then administered to one group of CLP mice at the time of bacterial challenge to simulate the IFN-γ response seen in sham mice challenged with Pseudomonas (Fig. 1A). The mice were sacrificed 6 h later, and serum cytokine concentrations were determined. Serum IFN-γ was significantly lower after Pseudomonas injection in mice that had been previously subjected to CLP than for sham mice (Fig. 1A). Serum IFN-γ concentrations were significantly higher in CLP mice that received an IFN-γ injection than in untreated CLP mice and were comparable to those in sham controls (Fig. 1A). IL-10 concentrations were increased in post-CLP mice compared to those in sham mice, and this increase was attenuated by IFN-γ administration (Fig. 1B). Concentrations of IL-12 and TNF-α in serum did not differ between sham and CLP mice, but treatment with IFN-γ caused a significant increase in both cytokines in post-CLP mice after Pseudomonas challenge (Fig. 1C and D).

Cell surface expression of the IL-12 receptor protein (IL-12R) was examined by using flow cytometry. IL-12 is an important modulator of the innate response to infection, and IL-12R β1 expression is induced upon lymphocyte activation (18). Expression of the IL-12 receptor β1 subunit protein on the surfaces of CD3⁺ T cells and NK cells was assessed (Fig. 2). There were significantly more T and NK cells expressing IL-12R in spleens from IFN-γ-treated CLP mice than in those from untreated CLP mice (Fig. 2B and D). In the CD3⁺ lymphocytes, there was a significant increase in the percentage of cells expressing IL-12R (Fig. 2A) after IFN-γ treatment. In the NK1.1⁺ lymphocytes, there was no increase in the percentage of cells expressing IL-12R (Fig. 2C), but there was a general increase in the number of NK cells, resulting in an overall increase in the number of IL-12R⁺ NK cells in the spleens of CLP mice treated with IFN-γ.

• Open in new tab
• Download powerpoint

FIG. 2.

Administration of IFN-γ caused increased numbers of CD3⁺ T cells and NK cells expressing IL-12R in post-CLP mice after Pseudomonas challenge. Mice were subjected to either CLP or sham surgery. Five days later, the mice were given intravenous injections of saline or 5 × 10⁷ CFU of P. aeruginosa, and they were sacrificed 6 h later. Spleens were aseptically excised and homogenized. Splenic homogenates were analyzed by flow cytometry after staining for cell surface markers. (A) Percentage of CD3⁺ T cells expressing IL-12R in control and IFN-γ-treated mice; (B) number of CD3⁺ IL-12R⁺ T cells as a percentage of all splenocytes; (C) percentage of NK cells expressing IL-12R; (D) number of NK1.1⁺ IL-12R⁺ T cells as a percentage of all splenocytes. *, significantly different from sham group; #, significantly different from CLP group (n = 6 to 10 per group).

To determine if IFN-γ treatment enhances the immune response to Pseudomonas, the bacterial burden in lungs was determined 6 h after Pseudomonas challenge (Fig. 3). All mice that were subjected to CLP prior to the Pseudomonas challenge had bacterial growth from lung tissue. In contrast, only two of the sham CLP mice challenged with Pseudomonas exhibited bacterial growth from lungs (seven of seven CLP mice compared to two of five sham mice; P < 0.05), and the mean number of CFU/lung in these mice was significantly lower than in mice that had been subjected to CLP (P < 0.05). Lung tissue homogenates from CLP mice had approximately 10-fold-greater numbers of bacterial CFU than samples from sham mice (Fig. 3). Treatment of post-CLP mice with IFN-γ at the time of Pseudomonas challenge did not significantly improve bacterial clearance, as indicated by the number of mice exhibiting bacterial growth (eight of eight animals) or number of CFU measured in lung homogenates (Fig. 3).

• Open in new tab
• Download powerpoint

FIG. 3.

Administration of IFN-γ does not reverse impaired bacterial clearance in mice previously subjected to CLP. P. aeruginosa (5 × 10⁷ CFU) was injected intravenously 5 days after CLP (or sham CLP), and mice were sacrificed 6 h later. Lungs were collected aseptically and homogenized in sterile saline, and serial dilutions were plated on tryptic soy agar. All untreated mice injected with Pseudomonas after CLP had positive lung cultures (seven of seven, versus two of five in the sham-CLP group; P < 0.05), and Pseudomonas growth in mice that had been subjected to CLP was approximately 10-fold higher than that of sham-CLP mice. All mice in the IFN-γ-treated group of CLP mice had positive lung cultures (eight of eight), and IFN-γ administration was not associated with a significant improvement in bacterial clearance. *, bacterial counts significantly different from those of sham group; #, number of animals with positive lung cultures significantly different from that for sham group (n = 5 to 8 per group).

Administration of IFN-γ significantly attenuated the increase in circulating IL-10 concentrations and increased the proinflammatory cytokine response but did not improve bacterial clearance or mortality in post-CLP mice after intravenous challenge with a lethal dose of P. aeruginosa.In our initial experiments, IFN-γ administration was associated with a significant effect on the cytokine response and a trend towards improved bacterial clearance in CLP mice after a sublethal Pseudomonas challenge. To further test the efficacy of exogenous IFN-γ in our model of post-CLP immunosuppression, we studied these parameters in post-CLP mice given a dose of Pseudomonas that causes ∼90% mortality. Mice that had been subjected to CLP or sham procedures 5 days earlier were challenged intravenously with P. aeruginosa (4 × 10⁸ CFU) and sacrificed 6 h later for aseptic collection of lung, spleen, and blood samples. CLP caused significant suppression of Pseudomonas-induced IFN-γ concentrations in serum, which was reversed by exogenous IFN-γ administration (Fig. 4A). Serum IL-10 was significantly higher after Pseudomonas injection for mice that had been previously subjected to CLP than for the sham CLP mice (Fig. 4B). Administration of IFN-γ attenuated IL-10 induction, with lower circulating plasma concentrations of IL-10 detected for IFN-γ-treated CLP mice than for untreated CLP mice. At the same time, concentrations of proinflammatory cytokines in plasma were elevated for mice challenged with Pseudomonas after CLP and were relatively unaffected by administration of IFN-γ (Fig. 4C to F). Pseudomonas-induced plasma concentrations of TNF-α, IL-1α, IL-1β, and IL-6 in the CLP mice were all found to be as high as, or higher than, concentrations of those cytokines in plasma of sham-CLP mice. Of those cytokines, only IL-1α differed significantly between the two CLP groups, with higher serum IL-1α levels detected in the plasma of IFN-γ-treated CLP mice.

• Open in new tab
• Download powerpoint

FIG. 4.

Administration of IFN-γ to CLP mice was associated with decreased serum concentrations of IL-10 but no difference in the proinflammatory cytokine profile induced by administration of a lethal dose of P. aeruginosa. P. aeruginosa (4 × 10⁸ CFU) was injected intravenously 5 days after CLP (or sham CLP), and mice were sacrificed 6 h later. Blood samples were collected for plasma cytokine analysis. (A) Plasma IFN-γ induced by Pseudomonas is significantly lower in mice previously subjected to CLP than with sham CLP animals. Administration of IFN-γ resulted in increased concentrations of circulating IFN-γ in CLP mice. B) Plasma IL-10 induced by Pseudomonas is significantly higher in mice previously subjected to CLP when compared to sham CLP animals. Administration of IFN-γ resulted in a decrease in circulating concentrations of IL-10 in CLP mice. The proinflammatory cytokine response induced by Pseudomonas is as great as, or greater than, that seen in sham CLP animals as determined by plasma concentrations of (C) TNF-α;,(D) IL-1α, (E) IL-1β, and (F) IL-6. Administration of IFN-γ did not affect the proinflammatory cytokine response. *, significantly different from sham group; #, significantly different from CLP group (sham group, n = 4; CLP and CLP + IFN-γ groups, n = 7).

To test the functional significance of IFN-γ treatment, we investigated bacterial clearance in lungs of post-CLP mice treated with IFN-γ following a lethal Pseudomonas challenge. Following challenge with a lethal dose of Pseudomonas, all mice had positive bacterial cultures from the lungs. As at the lower dose of Pseudomonas, post-CLP mice had higher bacterial counts in lungs than sham controls (Fig. 5). Significantly higher levels of total bacterial growth (Fig. 5A) and Pseudomonas growth (Fig. 5B) were observed for post-CLP mice than for sham mice. Exogenous administration of IFN-γ did not improve bacterial burdens in lungs after CLP. Furthermore, there was no observable effect of IFN-γ administration on the mortality rate, with 14% of the IFN-γ-treated CLP mice surviving for 7 days after the Pseudomonas challenge compared to 29% of untreated CLP mice (Fig. 6).

• Open in new tab
• Download powerpoint

FIG. 5.

Administration of IFN-γ does not reverse impaired clearance of a lethal dose of bacteria in mice previously subjected to CLP. P. aeruginosa (4 × 10⁸ CFU) was injected intravenously 5 days after CLP (or sham CLP), and mice were sacrificed 6 h later. Lungs were collected aseptically and homogenized in sterile saline, and serial dilutions were plated. (A) Total numbers of bacterial CFU in lung homogenates were significantly higher for mice challenged with Pseudomonas after CLP than for sham CLP mice. Exogenous IFN-γ did not improve bacterial clearance. (B) Numbers of Pseudomonas bacterial CFU in lung homogenates were significantly higher for mice challenged with P. aeruginosa after CLP than for sham CLP mice. Exogenous IFN-γ did not improve clearance of the Pseudomonas bacterial challenge. *, significantly different from sham group (sham group, n = 7; CLP and CLP + IFN-γ groups, n = 14).

• Open in new tab
• Download powerpoint

FIG. 6.

Survival after Pseudomonas challenge for post-CLP mice is not improved by administration of IFN-γ. Mice were subjected to CLP. Five days later, the mice were given intravenous injections of saline or 10⁸ CFU of P. aeruginosa and observed for mortality over the next 7 days. One group of mice was given IFN-γ subcutaneously at the time of Pseudomonas challenge. Mortality was not improved by IFN-γ treatment (no significant difference; n = 7 per group).

IFN-γ KO mice have an increased IL-10 response but little difference in the production of proinflammatory cytokines, bacterial clearance, or survival after a Pseudomonas challenge.To further investigate the role of IFN-γ in the immune response to Pseudomonas, wild-type and IFN-γ KO mice were challenged with a nonlethal dose of P. aeruginosa (5 × 10⁷ CFU, given intravenously) and sacrificed 6 h later for aseptic collection of lung tissue and blood samples. IFN-γ was present in the plasma of wild-type mice challenged with Pseudomonas but not in that of IFN-γ KO mice (Fig. 7A). IL-10 concentrations in plasma were elevated in IFN-γ KO mice receiving a Pseudomonas challenge compared to those in wild-type mice (Fig. 7B). At the same time, concentrations of TNF-α, IL-1α, IL-1β, and IL-6 in plasma were induced by Pseudomonas in all mice (Fig. 7C to F). Among the proinflammatory cytokines analyzed, only IL-6 was significantly lower in IFN-γ KO mice than in wild-type controls.

• Open in new tab
• Download powerpoint

FIG. 7.

IFN-γ KO mice exhibit increased serum concentrations of IL-10 but little difference in the proinflammatory cytokine response after Pseudomonas challenge. IFN-γ knockout mice (IFN KO) or wild-type controls (WT) were subjected to an intravenous injection of P. aeruginosa (5 × 10⁷ CFU) and were sacrificed 6 h later. (A) Plasma IFN-γ concentration after Pseudomonas challenge. (B) Significantly higher circulating concentrations of IL-10 were detected after Pseudomonas injection for IFN-γ^−/− mice than for wild-type controls. Concentrations of IL-6 (7F) in serum were significantly lower after injection of Pseudomonas into IFN-γ KO mice than for wild-type controls. However, (C) TNF-α, (D) IL-1α, and (E) IL-1β concentrations were similar between the two groups of mice. *, significantly different from other groups; #, significantly different from IFN knockout group (sham group, n = 4; WT and IFN-γ KO groups, n = 7).

Additional studies were undertaken to measure clearance of Pseudomonas by wild-type and IFN-γ KO mice. Growth of Pseudomonas from the lungs of mice was not significantly different between wild-type and IFN-γ KO mice (Fig. 8A). However, high numbers of gram-positive cocci were cultured from lung samples of all IFN KO mice in both the Pseudomonas challenge group and the saline control group and were significantly higher than those for wild-type mice (Fig. 8B).

• Open in new tab
• Download powerpoint

FIG. 8.

Loss of IFN-γ activity was not associated with diminished ability to clear a Pseudomonas bacterial challenge but was associated with a high baseline level of endogenous gram-positive bacterial background. IFN-γ knockout mice (IFN KO) or wild-type controls (WT) were subjected to an intravenous injection of P. aeruginosa (5 × 10⁷ CFU) (Pseudo) and were sacrificed 6 h later. Lungs were collected aseptically and homogenized in sterile saline, and serial dilutions were cultured. (A) The number of Pseudomonas CFU in lung homogenates did not differ between wild-type and IFN-γ knockout mice after an intravenous injection of Pseudomonas. (B) Loss of IFN-γ activity was associated with a high background of endogenous gram-positive coccus growth in lung homogenates from animals with, or without, a Pseudomonas bacterial challenge.

To further determine the functional role of IFN-γ in resistance to Pseudomonas challenge, we challenged IFN-γ KO and wild-type mice with Pseudomonas at two different doses and recorded the mortality rates over the subsequent 7-day period. Loss of IFN-γ had no discernible effect on mortality, with 22% of IFN-γ KO mice and 17% of wild-type mice surviving for 7 days after the low-dose challenge and no IFN-γ KO mice surviving the higher-dose challenge, compared to 11% of wild-type mice (Fig. 9).

• Open in new tab
• Download powerpoint

FIG. 9.

Survival after intravenous injection of Pseudomonas was not affected by the loss of IFN-γ activity. P. aeruginosa was injected intravenously into IFN-γ KO (IFN KO) or wild-type (WT) mice with no prior injury, and mortality was observed over the next 7-day period. There was no difference between survival of IFN-γ KO mice and that of wild-type mice after challenges of either (A) 1 × 10⁸ CFU or (B) 2 × 10⁸ CFU (no significant difference; n = 18 per group]).