Abstract
The restoration of blood flow, i.e., reperfusion, is the treatment of choice to save viable tissue following acute ischemia of a vascular territory. Nevertheless, reperfusion can be accompanied by significant inflammatory events that limit the beneficial effects of blood flow restoration. To evaluate the potential role of the intestinal microbiota in facilitating the development of tissue injury and systemic inflammation, germ-free and conventional mice were compared in their ability to respond to ischemia and reperfusion injury. In conventional mice, there was marked local (intestine) and remote (lung) edema formation, neutrophil influx, hemorrhage, and production of TNF-α, KC, MIP-2, and MCP-1. Moreover, there was an increase in the concentration of serum TNF-α and 100% lethality. In germ-free mice, there was no local, remote, or systemic inflammatory response or lethality after intestinal ischemia and reperfusion and, in contrast to conventional mice, germ-free animals produced greater amounts of IL-10. Similar results were obtained after administration of LPS, i.e., little production of TNF-α or lethality and production of IL-10 after LPS in germ-free mice. Blockade of IL-10 with Abs induced marked inflammation and lethality in germ-free mice after ischemia and reperfusion or LPS administration, demonstrating that the ability of these mice to produce IL-10 was largely responsible for their “no inflammation” phenotype. This was consistent with the prevention of reperfusion-associated injury by the exogenous administration of IL-10 to conventional mice. Thus, the lack of intestinal microbiota is accompanied by a state of active IL-10-mediated inflammatory hyporesponsiveness.
The restoration of blood flow, i.e., reperfusion, is the treatment of choice to save viable tissue following acute ischemia of a vascular territory. Nevertheless, reperfusion of ischemic tissues can be accompanied by significant local, remote, and systemic inflammatory events that may limit the beneficial effects of blood flow restoration (1, 2, 3, 4). Thus, it is believed that strategies limiting the acute inflammatory response that follows reperfusion may be useful therapeutic adjuncts in the treatment of acute ischemia (4). Factors influencing the severity of the inflammatory response include the vascular territory, the duration of ischemia, the duration of reperfusion, and the presence of collateral blood flow (1, 2, 3, 4). Several mediators of the inflammatory process have been shown to participate in the cascade of events leading to ischemia and reperfusion (I/R)3 injury, including CXC chemokines, platelet-activating factor and cytokines, especially TNF-α and IL-10 (5, 6, 7, 8, 9, 10, 11, 12). Treatment of animals with anti-TNF-α Abs or experiments in TNF-α receptor 1-deficient animals (TNFR1−/−) clearly show a central role of TNF-α in mediating tissue injury, systemic inflammation, and lethality following intestinal reperfusion injury (6, 7, 13). As TNF-α release and action is central in the pathogenesis of the local and systemic inflammatory response that occurs after intestinal reperfusion, a greater understanding of the mediators that induce or modulate TNF-α production in this process is clearly of interest.
A few studies in rats have suggested that bacterial translocation following disruption of intestinal epithelial lining would account for the increase in TNF-α production and tissue injury that occurs after intestinal I/R (14, 15). Of note, most of these studies used shorter ischemia and prolonged reperfusion times that could allow greater loss of the ability of epithelial cells to maintain their barrier function. In contrast, other studies using shorter ischemia times have failed to demonstrate an important role for LPS or the LPS receptor TLR4 during I/R injury (16, 17). Thus, an aim of the present study was to investigate whether bacterial and/or LPS translocation occurred and underlined the local and systemic inflammatory injury following prolonged intestinal I/R in mice. To this end the concentrations of LPS in blood and blood and liver bacterial load after intestinal I/R injury were quantified. Moreover, experiments evaluating the inflammatory injury after I/R were also conducted in germ-free mice, which have no bacteria (and indeed, no other known pathogen) in their gut. In the latter animals, we wished to evaluate the relevance of the intestinal microbiota to acute inflammation, as measured in a model of I/R injury. As germ-free mice were shown not to inflame or die after I/R injury, additional experiments were conducted to evaluate a role for IL-10 in mediating the lack of inflammatory responsiveness of these animals. For comparison, responses of germ-free and conventional mice to LPS and the role of IL-10 were also investigated.
Materials and Methods
Animals
Germ-free Swiss/NIH mice were derived from a germ-free nucleus (Taconic Farms, Germantown, NY) and maintained in flexible plastic isolators (Standard Safety Equipment, Palatine, IL) using classical gnotobiology techniques (18). Conventional Swiss/NIH mice are derived from germ-free matrices, and considered conventional only after two generations in the conventional facility. All animals were 8- to 10-wk-old males and females. All experimental procedures in germ-free mice were conducted under aseptic conditions to avoid infection of animals and had prior approval from the local animal ethics committee.
Ischemia and reperfusion
Mice were anesthetized with urethane (140 mg/kg, i.p.) and laparotomy was performed. The superior mesenteric artery (SMA) was isolated and ischemia was induced by totally occluding the SMA for 60 min. For measuring percentage of surviving mice, reperfusion was re-established, and mice were monitored for indicated time periods. For the other parameters, reperfusion was allowed to occur for 40 min (I60R40) when mice were sacrificed. This time of reperfusion (40 min) was chosen based on the presence of significant tissue injury without unduly high mortality rates. Sham-operated animals were used as controls. In some experiments, recombinant murine IL-10 (PeproTech, Rocky Hill, NJ) or vehicle (PBS, 100 μl/mouse) were administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. In other experiments, anti-IL-10 polyclonal (rabbit anti-rat/murine-IL-10, 1 μl/g) or monoclonal (SCX-2 clone, 100 μg/mouse) Abs were administered s.c. 45 min before reperfusion. As controls for the latter experiments, mice received preimmune rabbit serum (1 μl/g) or purified rat IgM (100 μg/mouse, clone R4-22; BD Pharmingen, San Diego, CA).
Macrophage cultures
Resident macrophages were obtained from the peritoneal cavity of naive germ-free or conventional mice and were cultured in DMEM supplemented with 10% FBS. The cells obtained from peritoneal washes were seeded in 96-well plates at 2 × 106 cells/ml. After 3 h of incubation at 37°C and 5% CO2, nonadherent cells were washed off and the adherent macrophages were incubated with or without LPS (100 ng/ml, from Escherichia coli serotype 0111:B4; Sigma-Aldrich, St. Louis, MO). After 48 h, the supernatant was removed and stored at −20°C until further analysis. The concentration of TNF-α and IL-10 was measured in supernatants using commercially available Abs and according to the procedures supplied by the manufacturer (R&D Systems, Minneapolis, MN).
Conventionalization of germ-free mice
The process of colonizing germ-free mice with microbiota from conventional mice is a process referred to as conventionalization and this was performed as previously described (18). Briefly, fecal samples removed from the large intestine of conventional mice were homogenized in saline (10%) and administered by oral gavage to germ-free mice. Seven, 14, or 21 days later, intestinal I/R was conducted in these animals, as described above. To assess whether there was adequate conventionalization of germ-free mice, fecal samples were cultured using a thioglycolate test (18).
Antibiotic treatment
A combination of large spectrum antibiotics was used to sterilize the intestinal microbiota of conventional mice. To this end a combination of metronidazole, ciprofloxacin, vancomycin, and imipenem (all at 50 mg/kg body weight/day) was administered in the drinking water. Animal beds were changed daily and all experimental procedures were conducted under aseptic conditions to avoid reinfection of animals. To assess whether the antibiotic treatment was adequate to deplete the intestinal microbiota, fecal samples were cultured using a thioglycolate test after 7, 14, and 21 days of antibiotic treatment (18). On day 21, the animals were submitted to intestinal I/R injury.
Measurement of bacteria and LPS after I/R
Blood was collected from the brachial plexus and the liver was harvested and grinded in a vented hood. The homogenates and blood samples were placed on ice, and serial 1/10 dilutions were made. One hundred microliters of each dilution were plated on McConkey agar plates (Difco, Detroit, MI) and incubated for 24 h at 37°C and then the number of CFU was counted. The detection limit of the assay was 100 bacteria ml−1 or 100 bacteria per 100 mg of tissue. LPS in plasma samples was determined using the Limulus amebocyte assay performed by Dr. S. Poole (National Institute of Biological Standard and Control, Potter Bar, U.K.).
LPS-induced lethality
LPS (1 and 10 mg/kg, from E. coli serotype 0111:B4; Sigma-Aldrich) was administered i.p. to conventional or germ-free mice. In these animals, lethality was observed at various times after injection or serum was obtained for TNF-α and IL-10 measurements. In some experiments, control Ab (rat IgM) or anti-IL-10 monoclonal (SCX-2 clone, 100 μg/mouse) Abs were administered s.c. 60 min before the administration of LPS (10 mg/kg).
Evaluation of changes in vascular permeability
The extravasation of Evans blue dye into the tissue was used as an index of increased vascular permeability, as previously described (19, 20). Briefly, Evans blue (20 mg/kg) was administered i.v. (1 ml/kg) via a tail vein 2 min before reperfusion of the ischemic artery. Forty minutes after reperfusion, a segment of the duodenum (∼3 cm) or the flushed left lung were cut in small pieces and Evans blue extracted using 1 ml of formamide. The amount of Evans blue in the tissue (micrograms of Evans Blue per 100 mg of tissue) was obtained by comparing the extracted absorbance with that of a standard Evans blue curve read at 620 nm in an ELISA plate reader.
Myeloperoxidase concentrations
The extent of neutrophil accumulation in the intestine and right lung tissue was measured by assaying myeloperoxidase activity, as previously described (7, 21). Briefly, a portion of duodenum and the flushed right lungs of animals that had undergone I/R injury were removed and snap frozen in liquid nitrogen. Upon thawing and processing, the tissue was assayed for myeloperoxidase activity by measuring the change in OD at 450 nm using tetramethylbenzidine. Results were expressed as total number of neutrophils by comparing the OD of tissue supernatant with the OD of casein-elicited murine peritoneal neutrophils processed in the same way.
Measurement of hemoglobin concentrations
The determination of hemoglobin concentrations in tissue was used as an index of tissue hemorrhage. After washing and perfusing the intestines to remove excess blood in the intravascular space, a sample of ∼100 mg of duodenum was removed and homogenized in Drabkin’s color reagent according to instructions of the manufacturer (Analisa, Belo Horizonte, Brazil). The suspension was centrifuged for 15 min at 3000 × g and filtered using 0.2-μm filters. The resulting solution was read using an ELISA plate reader at 520 nm and compared against a standard curve of hemoglobin.
Measurement of cytokine/chemokine concentrations in serum, intestine, and lungs
Serum was obtained from coagulated blood (15 min at 37°C, then 30 min a 4°C) and stored at −20°C until further analysis. Serum samples were analyzed at a 1/3 dilution in PBS. One hundred milligrams of duodenum or lung of sham-operated and reperfused animals were homogenized in 1 ml of PBS (0.4 M NaCl and 10 mM NaPO4) containing anti-proteases (0.1 mM PMSF, 0.1 mM benzethonium chloride, 10 mM EDTA, and 20 Kallikrein inhibitor units of aprotinin A) and 0.05% Tween 20. The samples were then centrifuged of 10 min at 3000 × g and the supernatant was immediately used for ELISA at a 1/3 dilution in PBS. The concentration of TNF-α, KC, MIP-2, MCP-1, and IL-10 was measured in serum and tissue of animals using commercially available Abs and according to the procedures supplied by the manufacturer (R&D Systems).
Statistical analysis
Results are shown as means ± SEM. Percent inhibition was calculated by subtracting the background values obtained in sham-operated animals. Differences were compared by using ANOVA followed by Student-Newman-Keuls posthoc analysis. Results with a p < 0.05 were considered significantly different.
Results
No evidence of systemic bacterial translocation after I/R
To evaluate the possible role of the translocation of bacteria and/or bacterial products as a means of triggering or potentiating reperfusion-associated systemic inflammatory response, blood and liver were cultured and evaluated for the presence of LPS. There were no detectable CFUs or measurable amounts of LPS in blood and liver tissue of mice following reperfusion of the ischemic SMA (data not shown).
Lack of tissue injury and lethality following I/R in germ-free mice
The local (intestine), remote (lung), and systemic inflammatory response after I/R injury in germ-free was compared with that in their conventional counterparts (i.e., animals normally exposed to the indigenous microbiota). As seen in Fig. 1, there was a massive increase in vascular permeability and neutrophil influx in the lungs and intestine of conventional, but not germ-free, animals submitted to I/R. Similarly, intestinal hemorrhage did not occur after reperfusion injury in germ-free animals, but was marked in conventional animals (Fig. 1,E). The concentration of TNF-α and chemokines (KC, MIP-2, and MCP-1) markedly increased after reperfusion in the intestine and lungs of conventional, but not germ-free, mice (Fig. 2 and Table I). In addition to the changes in tissue, there were marked systemic alterations in conventional mice submitted to I/R. Thus, serum TNF-α concentrations were strikingly elevated after reperfusion in conventional mice (Fig. 3,A) and this was associated with marked and early reperfusion-associated lethality (Fig. 3,B). In contrast, germ-free mice had no detectable amounts of TNF-α in serum and did not die (Fig. 3), even after 4 h of reperfusion (data not shown). Whereas tissue inflammation and proinflammatory cytokine/chemokine production were clearly suppressed in germ-free mice, an increase in reperfusion-associated IL-10 was detected in the intestine and lungs of germ-free, but not conventional, mice (Fig. 2).
Reperfusion-associated increase in neutrophil influx, vascular permeability, and hemorrhage in the intestine and lungs of conventional (CV) and germ-free (GF) mice. Neutrophil accumulation (A and B) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (C and D) by the extravasation of Evans blue dye, and hemorrhage (E) by the intestinal levels of hemoglobin. In CV mice, recombinant murine IL-10 (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as number of neutrophils, micrograms of Evans blue dye, or micrograms of hemoglobin per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Reperfusion-associated increase in neutrophil influx, vascular permeability, and hemorrhage in the intestine and lungs of conventional (CV) and germ-free (GF) mice. Neutrophil accumulation (A and B) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (C and D) by the extravasation of Evans blue dye, and hemorrhage (E) by the intestinal levels of hemoglobin. In CV mice, recombinant murine IL-10 (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as number of neutrophils, micrograms of Evans blue dye, or micrograms of hemoglobin per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Reperfusion-associated increase in the concentrations of TNF-α and IL-10 in the intestine and lungs of conventional (CV) and germ-free (GF) mice. TNF-α (A and B) and IL-10 (C and D) were measured using specific ELISA. In CV mice, recombinant murine (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as picograms of the cytokine per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Reperfusion-associated increase in the concentrations of TNF-α and IL-10 in the intestine and lungs of conventional (CV) and germ-free (GF) mice. TNF-α (A and B) and IL-10 (C and D) were measured using specific ELISA. In CV mice, recombinant murine (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as picograms of the cytokine per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Concentration of chemokines in the intestine and lungs of mice after reperfusion of the ischemic SMAa
. | MCP-1 . | . | KC . | . | MIP-2 . | . | |||
---|---|---|---|---|---|---|---|---|---|
. | Intestine . | Lung . | Intestine . | Lung . | Intestine . | Lung . | |||
CV-sham | 372.2 ± 29.6 | 1002.3 ± 98.5 | ND | 18.3 ± 1.6 | 149.9 ± 15.2 | 312.6 ± 29.6 | |||
CV-I/R | 1956.2 ± 187.6b | 3756.2 ± 298.5b | ND | 365.3 ± 32.6b | 155.2 ± 12.4 | 629.6 ± 59.8b | |||
CV-I/R + IL-10 | 824.3 ± 75.4c | 2223.6 ± 196.8c | ND | 112.4 ± 13.5c | 98.7 ± 9.4 | 457.0 ± 43.5c | |||
GF-sham | 196.4 ± 15.4 | 1009.0 ± 113.0 | ND | ND | 141.9 ± 7.6 | 250.6 ± 13.5 | |||
GF-I/R | 245.3 ± 18.3c | 1021.0 ± 97.8c | ND | 12.2 ± 1.0c | 200.6 ± 8.8 | 187.7 ± 6.63c |
. | MCP-1 . | . | KC . | . | MIP-2 . | . | |||
---|---|---|---|---|---|---|---|---|---|
. | Intestine . | Lung . | Intestine . | Lung . | Intestine . | Lung . | |||
CV-sham | 372.2 ± 29.6 | 1002.3 ± 98.5 | ND | 18.3 ± 1.6 | 149.9 ± 15.2 | 312.6 ± 29.6 | |||
CV-I/R | 1956.2 ± 187.6b | 3756.2 ± 298.5b | ND | 365.3 ± 32.6b | 155.2 ± 12.4 | 629.6 ± 59.8b | |||
CV-I/R + IL-10 | 824.3 ± 75.4c | 2223.6 ± 196.8c | ND | 112.4 ± 13.5c | 98.7 ± 9.4 | 457.0 ± 43.5c | |||
GF-sham | 196.4 ± 15.4 | 1009.0 ± 113.0 | ND | ND | 141.9 ± 7.6 | 250.6 ± 13.5 | |||
GF-I/R | 245.3 ± 18.3c | 1021.0 ± 97.8c | ND | 12.2 ± 1.0c | 200.6 ± 8.8 | 187.7 ± 6.63c |
Cytokine concentrations were assessed by using specific ELISA. IL-10 (0.5 μg/animal) was given s.c. to conventional mice (CV) 45 min prior to reperfusion of the ischemic SMA (I/R). Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Parallel experiments were conducted in germ-free mice (GF). Results are shown as picograms of cytokine per 100 mg of tissue and are the mean ± SE mean of five to six animals in each group.
p < 0.01 when compared to sham-operated animals.
p < 0.05 when compared to severe I/R animals.
Reperfusion-associated increase in the serum concentrations of TNF-α and lethality in conventional (CV) and germ-free (GF) mice. TNF-α (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. In conventional mice, recombinant murine IL-10 (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. There were at least 10 animals in each group for the survival experiments. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Reperfusion-associated increase in the serum concentrations of TNF-α and lethality in conventional (CV) and germ-free (GF) mice. TNF-α (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. In conventional mice, recombinant murine IL-10 (shown as IL-10) was administered s.c. at the dose of 0.5 μg/animal 45 min before reperfusion. Control animals (shown as IR) received IL-10 vehicle (PBS, 100 μl). Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. There were at least 10 animals in each group for the survival experiments. ∗, p < 0.01 when compared with sham-operated animals; #, p < 0.01 when compared with vehicle-treated animals submitted to I/R.
Prolonged contact with the intestinal microbiota is necessary to induce an inflammatory response in germ-free mice
Intestinal bacteria were already detectable on day 7 after conventionalization of germ-free (data not shown). However, there was no tissue injury (Fig. 4) or lethality (Fig. 5) after intestinal I/R in germ-free 7 days after the administration of bacteria. In contrast, there was significant intestinal inflammation 14 and 21 days later, as assessed by vascular permeability, neutrophil influx, intestinal hemorrhage, and local production of TNF-α (Fig. 4). There was also marked pulmonary inflammation in the latter animals (data not shown). Overall, the inflammatory changes were more marked after 21 than 14 days after conventionalization (Fig. 4). In addition to inducing tissue inflammation, there was also a systemic inflammatory response 14 and 21 days after reinstatement of the microbiota, as assessed by the reperfusion-induced increase in serum TNF-α concentrations and lethality (Fig. 5). As noted above, germ-free mice produced large amounts of IL-10 after intestinal I/R (Fig. 2). Conventionalization was accompanied by a time-dependent inhibition of IL-10 production that was already maximal 14 days after reposition (Fig. 4). Overall, there was a good correlation between the loss of the ability to produce IL-10 and the gain of the ability to inflame in response to the reperfusion injury.
Reversal by microbiota reposition of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in neutrophil influx, vascular permeability hemorrhage, TNF-α and IL-10 concentration in the intestine. Neutrophil accumulation (A) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (B) by the extravasation of Evans blue dye, hemorrhage (C) by the intestinal levels of hemoglobin, TNF-α (D) and IL-10 (E) were measured using specific ELISA. To colonize germ-free mice, feces from conventional mice were administered orally 7, 14, or 21 days prior to when the animals were submitted to I/R. Results are shown as number of neutrophils, micrograms of Evans blue dye, micrograms of hemoglobin per 100 mg of tissue, picograms of the cytokine per 100 mg of intestine and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with germ-free animals submitted to I/R.
Reversal by microbiota reposition of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in neutrophil influx, vascular permeability hemorrhage, TNF-α and IL-10 concentration in the intestine. Neutrophil accumulation (A) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (B) by the extravasation of Evans blue dye, hemorrhage (C) by the intestinal levels of hemoglobin, TNF-α (D) and IL-10 (E) were measured using specific ELISA. To colonize germ-free mice, feces from conventional mice were administered orally 7, 14, or 21 days prior to when the animals were submitted to I/R. Results are shown as number of neutrophils, micrograms of Evans blue dye, micrograms of hemoglobin per 100 mg of tissue, picograms of the cytokine per 100 mg of intestine and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with germ-free animals submitted to I/R.
Reversal by microbiota reposition of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in the concentrations of TNF-α and lethality. TNF-α in serum (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. To colonize germ-free mice, feces from conventional mice were administered orally 7, 14, or 21 days prior to when the animals were submitted to I/R. Results are shown in picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. There were at least 10 animals in each group for the survival experiments. ∗, p < 0.01 when compared with germ-free animals submitted to I/R.
Reversal by microbiota reposition of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in the concentrations of TNF-α and lethality. TNF-α in serum (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. To colonize germ-free mice, feces from conventional mice were administered orally 7, 14, or 21 days prior to when the animals were submitted to I/R. Results are shown in picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. There were at least 10 animals in each group for the survival experiments. ∗, p < 0.01 when compared with germ-free animals submitted to I/R.
Antibiotic treatment is not sufficient to prevent injury in conventional mice
Treatment of conventional mice with large spectrum antibiotics resulted in no detectable bacteria on days 7, 14, and 21 after the start of treatment (data not shown). Although these animals had no detectable bacteria, reperfusion-associated lethality was similar in untreated and antibiotic-treated conventional mice on days 7, 14, and 21 (data not shown).
Exogenous administration of IL-10 protects against I/R injury
The next series of experiments were performed to examine whether the exogenous administration of recombinant murine IL-10 would be capable of modulating the tissue and systemic injuries and lethality that occurred after reperfusion of the ischemic SMA. The systemic administration of IL-10 before reperfusion was associated with an increase in the concentrations of IL-10 immunoreactivity in the intestine and lungs of reperfused animals (Fig. 2). The tissue levels of IL-10 were 70–80% of those found in germ-free mice, suggesting that the chosen dose of IL-10 was sufficient to reach the affected tissues. In IL-10-treated conventional mice, there was a marked inhibition of reperfusion-induced increases in vascular permeability and neutrophil influx in the lungs and intestine and intestinal hemorrhage (Fig. 1). Exogenous IL-10 also prevented the increases in serum, intestinal and pulmonary concentrations of TNF-α (Figs. 2 and 3,A). The reperfusion-induced increase in the concentrations of the chemokines KC, MIP-2, and MCP-1 were significantly attenuated by IL-10 treatment (Table I). More importantly, lethality rates in IL-10-treated and reperfused conventional mice were significantly delayed and ∼75% of the animals were still alive 120 min after reperfusion (Fig. 3,B). Altogether the experiments above demonstrate that IL-10 is effective in preventing tissue injury and lethality due to intestinal I/R injury in mice (Fig. 6).
Reversal by anti-IL-10 of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in neutrophil influx, vascular permeability, and hemorrhage in the intestine and lungs. Neutrophil accumulation (A and B) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (C and D) were assessed by the extravasation of Evans blue dye, and hemorrhage (E) was assessed by the intestinal levels of hemoglobin. Anti-IL-10 mAbs (aIL-10) or rat IgM (shown as IR) were administered s.c. 45 min before reperfusion. Results are shown as number of neutrophils, micrograms of Evans blue dye or micrograms of hemoglobin per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. #, p < 0.01 when compared with germ-free animals submitted to I/R.
Reversal by anti-IL-10 of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in neutrophil influx, vascular permeability, and hemorrhage in the intestine and lungs. Neutrophil accumulation (A and B) was assessed by measuring the tissue contents of myeloperoxidase, changes in vascular permeability (C and D) were assessed by the extravasation of Evans blue dye, and hemorrhage (E) was assessed by the intestinal levels of hemoglobin. Anti-IL-10 mAbs (aIL-10) or rat IgM (shown as IR) were administered s.c. 45 min before reperfusion. Results are shown as number of neutrophils, micrograms of Evans blue dye or micrograms of hemoglobin per 100 mg of tissue and are the mean ± SEM of at least six animals in each group. #, p < 0.01 when compared with germ-free animals submitted to I/R.
Anti-IL-10 reverses the hyporesponsiveness of germ-free mice to reperfusion-associated injury and lethality
As IL-10 prevented tissue injury and lethality and germ-free mice produced significantly greater levels of IL-10 following I/R of SMA, the next series of experiments were conducted to evaluate whether endogenous IL-10 played a role in the lack of tissue injury and lethality observed in germ-free mice. To this end, germ-free mice were treated with anti-IL-10 Abs before reperfusion and tissue injury and lethality examined. Postischemic treatment of germ-free mice with anti-IL-10 was accompanied by a significant increase in reperfusion-induced tissue injury, as assessed by a marked increase in intestinal and pulmonary vascular permeability, neutrophil influx, and hemorrhage (Fig. 7). Indeed, tissue injury in germ-free mice treated with anti-IL-10 reached ∼70% of the injury observed in control conventional mice (compare Figs. 1 and 7). The concentrations of TNF-α in tissue and serum of anti-IL-10-treated germ-free mice is shown in Fig. 7. Whereas untreated germ-free mice have undetectable concentrations of TNF-α in tissues and serum after I/R injury, there was a marked and significant increase in TNF-α in animals treated with anti-IL-10 (Fig. 7,A). Similarly, whereas germ-free mice did not die after I/R injury, treatment with anti-IL-10 was accompanied by significant reperfusion-induced lethality, which was similar to that seen in conventional mice (Fig. 7 B). Virtually identical results were obtained when germ-free mice were administered an anti-IL-10 polyclonal Ab (data not shown). Overall, our results argue that the lack of reperfusion-induced tissue inflammation observed in germ-free mice is largely due to their innate ability to produce IL-10 and consequent IL-10-mediated inhibition of the systemic inflammatory response.
Reversal by anti-IL-10 of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in the concentrations of TNF-α and lethality. TNF-α in serum (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. Anti-IL-10 mAbs (a-10) or rat IgM (shown as IR) were administered s.c. 45 min before reperfusion. Results are shown as picograms of the cytokine per 100 mg of tissue or 100 μl of serum and are the mean ± SEM of at least six animals in each group. In B, the survival curves of conventional mice (CV) are also shown for comparison and there were at least 10 animals in each group. #, p < 0.01 when compared with germ-free animals submitted to I/R.
Reversal by anti-IL-10 of the hyporesponsiveness of germ-free (GF) mice to reperfusion-associated increase in the concentrations of TNF-α and lethality. TNF-α in serum (A) was measured using specific ELISA. Survival (B) was monitored as indicated and survivors were sacrificed after 120 min. Anti-IL-10 mAbs (a-10) or rat IgM (shown as IR) were administered s.c. 45 min before reperfusion. Results are shown as picograms of the cytokine per 100 mg of tissue or 100 μl of serum and are the mean ± SEM of at least six animals in each group. In B, the survival curves of conventional mice (CV) are also shown for comparison and there were at least 10 animals in each group. #, p < 0.01 when compared with germ-free animals submitted to I/R.
The hyporesponsiveness of germ-free mice to LPS is also mediated by IL-10
The IL-10-dependent inability of germ-free mice to respond to such a potent inflammatory stimulus, as that triggered by the reperfusion of an ischemic vascular, was a striking finding. It was, thus, of interest to examine whether the lack of inflammatory responsiveness of germ-free mice would be also applicable to other stimuli capable of eliciting a systemic inflammatory response, such as those derived from microbes. To this end, germ-free or conventional mice were injected i.p. with LPS (10 mg/kg) and lethality rates and cytokine (TNF-α and IL-10) concentrations examined at different time points. Concentrations of TNF-α rose rapidly (within 1.5 h) in the serum of conventional mice and were still elevated by 6 h, whereas only small amounts of TNF-α (<10% of the concentrations found in conventional mice) were detected in the serum of GF mice after systemic administration of LPS (Fig. 8,A). In contrast, concentrations of IL-10 were much greater and more persistent in serum of germ-free than conventional mice (Fig. 8,B). Indeed, measurable amounts of IL-10 were detected even 24 h after injection of LPS in germ-free (112 ± 8 pg/100 mg of tissue) but not conventional (below detection limit of the assay) mice. All conventional mice injected with LPS were dead by 12 h after injection, whereas none of the LPS-injected germ-free or PBS-injected mice were dead by 96 h after LPS injection (Fig. 8 C). In fact, there was no lethality even 10 days after LPS injection (data not shown).
LPS-induced increase in serum concentrations of TNF-α and IL-10, and lethality in conventional (CV) and germ-free (GF) mice. A and B, LPS was administered i.p. at the dose of 10 mg/kg, and TNF-α (A) and IL-10 (B) measured at the indicated times using specific ELISA. Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with PBS-injected animals; #, p < 0.01 when compared with LPS-treated conventional animals. C, The concentration of TNF-α in serum was evaluated 6 h after challenge in germ-free mice injected with LPS or LPS and anti-IL-10. Anti-IL-10 mAbs (aIL-10) or rat IgM (shown as LPS) were administered s.c. 45 min before the administration of LPS. Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. #, p < 0.01 when compared with LPS-treated GF mice given control Ab. D, Survival was monitored as indicated and survivors were sacrificed after 96 h. There were at least 10 animals in each group. After LPS treatment, survival of GF mice given control Ab was significantly (p < 0.01) different from that of CV mice and anti-IL-10-treated germ-free mice.
LPS-induced increase in serum concentrations of TNF-α and IL-10, and lethality in conventional (CV) and germ-free (GF) mice. A and B, LPS was administered i.p. at the dose of 10 mg/kg, and TNF-α (A) and IL-10 (B) measured at the indicated times using specific ELISA. Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. ∗, p < 0.01 when compared with PBS-injected animals; #, p < 0.01 when compared with LPS-treated conventional animals. C, The concentration of TNF-α in serum was evaluated 6 h after challenge in germ-free mice injected with LPS or LPS and anti-IL-10. Anti-IL-10 mAbs (aIL-10) or rat IgM (shown as LPS) were administered s.c. 45 min before the administration of LPS. Results are shown as picograms of the cytokine per 100 μl of serum and are the mean ± SEM of at least six animals in each group. #, p < 0.01 when compared with LPS-treated GF mice given control Ab. D, Survival was monitored as indicated and survivors were sacrificed after 96 h. There were at least 10 animals in each group. After LPS treatment, survival of GF mice given control Ab was significantly (p < 0.01) different from that of CV mice and anti-IL-10-treated germ-free mice.
To examine whether the endogenous production of IL-10 in response to LPS treatment in germ-free mice would also account for the lack of responsiveness of these animals to LPS, experiments were performed in the presence of an anti-IL-10 Ab. Similarly to our findings in the I/R injury model, germ-free treated with anti-IL-10 became very responsive to LPS administration and produced large amounts of TNF-α and died in a similar manner to conventional mice (Fig. 8, C and D).
Macrophages from germ-free mice preferentially produce IL-10
The stimulation of resident peritoneal macrophages obtained from germ-free or conventional mice with LPS induced a distinct profile of cytokine production (Table II). There was no difference in basal production of cytokines. However, whereas LPS-stimulated macrophages derived from conventional mice produced large amounts of TNF-α and little IL-10, macrophages from germ-free mice produced large amounts of IL-10 and little TNF-α (Table II).
Concentration of TNF-α and IL-10 in supernatant of macrophagesa
. | Conventional Mice (pg/ml) . | . | Germ-Free Mice (pg/ml) . | . | ||
---|---|---|---|---|---|---|
. | Medium . | LPS . | Medium . | LPS . | ||
TNF-α | 331.1 ± 33.7 | 1852.8 ± 118.2b | 284.5 ± 24.0 | 390.8 ± 2.5c | ||
IL-10 | 26.4 ± 2.3 | 170.4 ± 8.0b | 27.7 ± 9.5 | 494.0 ± 38.1bc |
. | Conventional Mice (pg/ml) . | . | Germ-Free Mice (pg/ml) . | . | ||
---|---|---|---|---|---|---|
. | Medium . | LPS . | Medium . | LPS . | ||
TNF-α | 331.1 ± 33.7 | 1852.8 ± 118.2b | 284.5 ± 24.0 | 390.8 ± 2.5c | ||
IL-10 | 26.4 ± 2.3 | 170.4 ± 8.0b | 27.7 ± 9.5 | 494.0 ± 38.1bc |
Cytokine concentration was assessed by using specific ELISA. Resident macrophages were obtained from the peritoneal cavity of naive germ-free or conventional mice and were cultured in DMEM supplemented with 10% FBS. After 48 h, the supernatant was removed for analyses of the concentration of TNF-α and IL-10. Results are shown as mean ± SD of six replicates and are given as picograms of cytokine per milliliter of supernatant. The experiment was repeated three times.
p < 0.01 when compared to medium.
p < 0.01 when compared to LPS-stimulated macrophages from conventional mice.
Discussion
The restoration of blood flow of an ischemic vascular bed is the treatment of choice to prevent the death of tissue cells during acute ischemic events. Nevertheless, reperfusion itself can induce a local inflammatory response that limits the benefits of blood flow restoration. The present study was conducted to investigate the relevance of bacterial translocation and of the intestinal microbiota for the inflammatory response and tissue injury that occurs after intestinal I/R.
Inflammatory or ischemic injury to the intestine may be of sufficient intensity to cause disruption of the intestinal epithelial lining and consequent loss of the barrier function of epithelial cells. This may occur, for example, after hypovolemic shock and facilitate the translocation of bacteria and/or bacterial products, such as LPS, with ensuing activation of macrophages and infection of the organism (22, 23). Whereas some investigators have found elevated concentrations of LPS in systemic or portal system blood following reperfusion of an ischemic vascular bed (14, 17, 24, 25), others, including the present study, have not been able to detect either LPS or significant levels of circulating bacteria (26). The conflict between these studies is largely due to the different durations of I/R, in such way that LPS/bacteria are more likely to be detected after prolonged reperfusion times. However, it is still possible that sufficient amounts of bacteria and/or LPS reach and activate inflammatory and endothelial cells within the intestinal mucosa (16). In this respect, studies in LPS-hyporesponsive C3H/HeJ mice have demonstrated that tissue inflammation and E-selectin expression are similar to those of a matched LPS-responsive background (Ref.16 , our unpublished results). C3H/HeJ mice do not possess a functional TLR4 and are, thus, not capable of responding to LPS or other TLR4 ligands, but still respond normally to ligands at other TLRs (27, 28, 29). Germ-free mice provide an alternative approach to investigate the possible contribution of intestinal bacteria that contain molecular patterns capable of activating TLRs distinct from TLR4. As discussed in more detail below, germ-free mice did not inflame or die after intestinal I/R. Nevertheless, when these animals were administered an anti-IL-10 Ab just before reperfusion, tissue inflammation and lethality were similar to those found in conventional animals. Of note, anti-IL-10 was given just before reperfusion and was, hence, unable to modify the microbiological status of the animals. Thus, it is clear from the present study that translocation of bacteria and/or bacterial-derived products is not necessary to activate the cascade of inflammatory events leading to acute tissue injury, systemic inflammation, and lethality after reperfusion of the ischemic intestine.
In mice, intestinal reperfusion is followed by severe local (intestine) and remote (lungs) tissue pathology, characterized by marked neutrophil influx, edema formation, hemorrhage, and tissue destruction (shown here and in Ref.7). Not only is there tissue damage, but also marked systemic inflammation, as assessed by the elevation in the serum concentration of proinflammatory cytokines and chemokines (7). In this model, TNF-α appears to play a major pathophysiological role, as demonstrated by the protective effects of the administration of soluble TNF-α receptors or experiments in p55−/− mice (7). Similarly, other studies have clearly established a role for TNF-α in mediating both local and systemic inflammation after I/R injury in several vascular beds (e.g., Refs.13 , 30 , 31). In contrast to these findings and to the lack of a role of the translocation of bacteria/bacterial products, germ-free mice presented little evidence of local or systemic injury after intestinal I/R. Indeed, intestinal or pulmonary injury and concentrations of proinflammatory cytokines and chemokines in germ-free mice after reperfusion were virtually identical to those found in sham-operated mice (which were not submitted to I/R). Furthermore, there was no lethality even after 240 min of reperfusion, whereas all conventional mice were dead by 90 min. Despite the absence of local or systemic inflammation, germ-free mice were actively sensing the inflammatory stimulation, as demonstrated by the marked increases in reperfusion-induced IL-10 production.
Several studies have now reported on the ability of endogenous or exogenous IL-10 in limiting injury secondary to I/R to various vascular beds (e.g., Refs.12 , 32 , 33). In these models, IL-10 acts by limiting TNF-α production and neutrophil influx (34). In our model, administration of IL-10, at a dose that led to tissue concentrations of IL-10 similar to those found in germ-free mice, was effective in diminishing reperfusion-associated tissue damage and delaying lethality. More importantly, the administration of anti-IL-10 Abs before the reperfusion of the ischemic SMA in germ-free mice was followed by marked reperfusion-associated inflammation–roughly 70% of that observed in conventional mice–and lethality. In our experiments, the ability of endogenous or exogenous IL-10 to prevent lethality and tissue injury was clearly correlated with the concentrations of local or systemic TNF-α (see Figs. 3 and 5). The situation in LPS-injected germ-free and conventional mice was virtually identical. Contrary to conventional mice, germ-free animals produced little TNF-α, did not die, and produced large amounts of IL-10 following LPS. Moreover, administration of anti-IL-10 to germ-free mice enhanced the LPS-induced TNF-α production and was followed by markedly increased lethality. Thus, exogenously administered IL-10 could mimic, and blockade of endogenous IL-10 could prevent, the expression of a “no-inflammation” phenotype in two models of systemic inflammation in germ-free mice. The likelihood that other anti-inflammatory molecules and/or pathways are involved is given by the inability of anti-IL-10 to reverse completely and IL-10 to mimic fully the phenotype of germ-free mice.
The present study did not investigate in any detail the cell types responsible for the production of TNF-α and/or IL-10. However, several studies have now shown that macrophages are an important source of the latter cytokines in the setting of I/R injury (35, 36, 37). To evaluate whether macrophages derived from germ-free and conventional mice differed in their ability to respond to inflammatory stimulation, resident peritoneal macrophages were obtained from both strains and stimulated with LPS. As noted above, germ-free do not die after LPS injection in vivo and respond by producing large quantities of IL-10. In contrast, LPS-stimulated macrophages derived from germ-free mice responded by producing IL-10, whereas those from conventional mice produced TNF-α. Overall, the latter results show that the function of macrophages may be fundamentally between conventional and germ-free mice and suggest that studies of macrophage function may provide clues as to why germ-free preferentially produce IL-10.
It is interesting to note that the innate ability of germ-free to produce IL-10 and, possibly, other molecules, was an active process that prevented the inflammatory phenotype, i.e., animals not exposed to an intestinal microbiota have an innate ability to produce molecules with anti-inflammatory properties that suppress the development of an inflammatory response. In contrast, mice that had a normal intestinal microbiota or in which the microbiota had been reinstated were capable of responding to inflammatory stimulation, suggesting that the presence of microorganisms in the gut induce a “state of alert” that is characterized by the loss of the innate ability to produce IL-10 and, possibly, other molecules (e.g., TGF-β). Indeed, colonization of the gastrointestinal tract of germ-free with gut bacteria of conventional mice was capable of preventing the preferential production of IL-10 and restoring inflammatory responsiveness. It was interesting to notice that there is a need for prolonged colonization of the gastrointestinal tract as an effect of conventionalization was only seen after 14 days and was maximal at 21 days. In contrast, even prolonged treatment with large spectrum antibiotics was not capable of modifying the inflammatory responsiveness of conventional mice, even if bacteria could not be detected in feces of treated mice. The test used for bacterial detection is sensitive, suggesting that there was a decrease of the number of bacteria below the number necessary to establish an efficient ecological relation between host and microbiota (usually around 107 bacteria) (18). It is not clear why the antibiotic treatment was ineffective in reversing the inflammatory responsiveness. It is possible that more prolonged treatment is required to reverse the inflammatory reactivity of germ-free mice, but this is experimentally difficult to achieve. Alternatively, once the organism is programmed to inflame in response to stimulation–and not produce IL-10–it may not be possible to reverse to a “no inflammation” phenotype as that observed in germ-free mice. Further studies are clearly necessary to resolve this issue.
In conclusion, our studies demonstrate that the inability of germ-free mice to inflame in response to systemic LPS or reperfusion-induced injury is largely due to the innate capacity of these mice to produce IL-10 and, possibly, other anti-inflammatory molecules. The mechanisms underlying the innate ability of germ-free mice to produce IL-10 have not been investigated here any further but our recent studies suggest that lipoxin production underlies the greater production of IL-10 in germ-free mice (D. G. Souza and M. M. Teixeira, unpublished data). The IL-10 produced then switches off proinflammatory cytokine production, inflammatory cell influx, and consequent tissue injury and lethality. The detailed understanding of the molecular interactions underlying innate IL-10 production is a fundamental question that may unravel novel targets for treatment of acute and chronic inflammatory disorders.
Acknowledgements
We are grateful to Dr. Ricardo Gazzinelli for his helpful comments and to Dr. Jacqueline Alvarez-Leite for the supply of conventional mice.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by Fundação do Amparo a Pesquisa do Estado de Minas Gerais (FAPEMIG), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), and Pró-Reitoria de Pesquisa da UFMG (PRPq/UFMG).
Abbreviations used in this paper: I/R, ischemia and reperfusion; SMA, superior mesenteric artery.