Background
Pseudomonas aeruginosa, a non-fermenting Gram-negative rod, is one of the most important bacterial pathogens responsible for a multitude of opportunistic infections in humans, especially in hospitalized patients [
1]. The strictly aerobic pathogen is the primary cause of ventilator-associated pneumonia or superinfection of burn wounds associated with a mortality of more than 30% [
2]. Its worldwide increasing resistance to manifold antibiotic classes particularly due to extended-spectrum β-lactamases, carbapenemases and 16S rRNA methylases makes
P. aeruginosa to a continuously growing threat for immune-compromised individuals, patients suffering from cystic fibrosis and other pulmonary diseases or patients admitted to intensive care units [
2‐
4]. A single flagellum and multiple cell surface pili allow motility and adherence of
P. aeruginosa with surfaces [
5], whereas biofilm formation, alginate secretion, quorum-sensing and an elaborated secretion system contribute to the virulence of
P. aeruginosa [
1,
5]. Earlier studies revealed that ingested
P. aeruginosa are detectable in human fecal samples of healthy volunteers up to 6 days following oral administration without occurrence of clinical symptoms. Fecal
P. aeruginosa loads were lower than those that had been ingested and decreased over time post challenge [
6]. However, admission to a surgical ward may increase the risk of peroral
P. aeruginosa acquisition, since the percentage of patients’ stool samples carrying
P. aeruginosa increased during hospital stay [
7]. In addition to objects and materials present on wards, the human gastrointestinal tract may be an important source of
P. aeruginosa infection [
7]. Recent clinical surveys revealed that
P. aeruginosa detection rates were significantly higher in fecal and mucosal samples derived from patients suffering from irritable bowel syndrome [
8] or in the colonic mucosa of a pediatric patient with ulcerative colitis [
9] as compared to a healthy individual. Whereas the pathogenic potential of
P. aeruginosa is well-known, its potential contribution to initiation and perpetuation of intestinal immunopathological conditions are not yet understood. Hence, there is currently a large gap in knowledge regarding the interplay of
P. aeruginosa, the host microbiota and immune system, particularly under conditions of intestinal inflammation.
Within 1 week following peroral high dose infection with the intracellular parasite
Toxoplasma gondii, susceptible mice develop acute inflammation of the terminal ileum with massive necrosis (pan-ileitis) as well as extra-intestinal and systemic sequelae and succumb to infection within 7–10 days [
10‐
12]. This fatal hyper-inflammatory scenario is due to a typical T helper cell 1 (Th1)-type immunopathology and characterized by an overproduction of pro-inflammatory mediators including interferon (IFN)-γ, tumor necrosis factor (TNF)-α, nitric oxide, interleukin (IL)-6, IL-12 and monocyte chemoattractant protein (MCP)-1 evolving upon parasitic interacting with antigen presenting cells with subsequent activation of CD4+ T cells, whereas the
T. gondii induced anti-inflammatory measures include IL-10 expression (reviewed by [
13]). We showed earlier that the gut microbiota is essential for the initiation and progression of
T. gondii induced ileitis and that small intestinal inflammation is accompanied by distinct changes in the commensal microbiota composition with Gram-negative species such as enterobacteria and
Bacteroides/Prevotella spp. overgrowing the inflamed ileal lumen [
11]. Toll-like receptor (TLR)-4 dependent signaling of lipopolysaccharide derived from Gram-negative intestinal commensals further perpetuates the hyper-inflammatory process [
14,
15]. Hence, the Th1-type immunopathology underlying
T. gondii induced ileitis and associated intestinal microbiota shifts resemble key features of acute episodes in patients suffering from Crohn’s disease [
10,
13]. In the present study we investigated whether acute small intestinal inflammation predisposes the vertebrate host with a human microbiota to infection with multidrug resistant (MDR)
P. aeruginosa. To address this, mice harboring a complex human microbiota were challenged with high dose
T. gondii infection for ileitis induction and subsequently infected with a MDR
P. aeruginosa strain. The presented results shed further light onto the interplay between MDR
P. aeruginosa, the host innate and adaptive immunity and human intestinal microbiota during acute small intestinal inflammation.
Methods
Ethical statement
All animal experiments were conducted according to the European Guidelines for animal welfare (2010/63/EU) with approval of the commission for animal experiments headed by the “Landesamt für Gesundheit und Soziales” (LaGeSo, Berlin; registration numbers G0097/12 and G0039/15). Animal welfare was monitored twice daily by assessment of clinical conditions and weight loss of mice. Mice suffering from weight loss >20% were euthanized by isoflurane treatment (Abbott, Germany) in accordance with the guidelines of the local authorities.
Generation of gnotobiotic (secondary abiotic) mice
Female C57BL/6j mice were bred under specific pathogen-free conditions in the Forschungsinstitute für Experimentelle Medizin (Charité-University Medicine, Berlin, Germany). Gnotobiotic (i.e. secondary abiotic) mice with a virtually depleted microbiota were generated as described previously [
11]. In brief, 8 weeks old mice were transferred to sterile cages and subjected to a broad-spectrum antibiotic treatment for 8–10 weeks by adding ampicillin plus sulbactam (1 g/L; Ratiopharm, Germany), vancomycin (500 mg/L; Cell Pharm, Germany), ciprofloxacin (200 mg/L; Bayer Vital, Germany), imipenem (250 mg/L; MSD, Germany) and metronidazole (1 g/L; Fresenius, Germany) to the drinking water (ad libitum).
Generation of human microbiota-associated mice
Fresh fecal samples free of enteropathogenic bacteria, viruses and parasites were collected from five individual healthy volunteers, dissolved in sterile phosphate buffered saline (PBS; Gibco, life technologies, UK), aliquoted and stored at −80 °C as described earlier [
16]. Immediately before reconstitution experiments, individual fecal aliquots were thawed, pooled, and the main bacterial communities within the donor suspension quantitatively assessed by cultural and molecular methods [
16]. To generate human intestinal microbiota-associated (hma) mice, gnotobiotic animals were subjected to peroral fecal transplantations with 0.3 mL of the donor suspension by gavage on 3 consecutive days. The total load of bacterial groups between independent experiments counted around 10
10 colony forming units (CFU) and varied less than 0.5 logarithmic orders of magnitude. To assure proper establishment of the human microbiota in the murine host, mice were kept for at least 3 weeks until ileitis induction. Immediately before peroral
T. gondii infection individual fecal samples were collected for quantitative cultural and molecular analyses of main intestinal bacterial communities.
Induction of acute ileitis
In order to induce acute ileitis mice harboring a human intestinal microbiota were infected perorally with 50 cysts of
T. gondii (ME49 strain) by gavage as described previously [
11,
14,
17].
Pseudomonas aeruginosa infection
Three days following ileitis induction mice were perorally infected with 109 CFU of a MDR P. aeruginosa strain by gavage in a total volume of 0.3 mL PBS. The P. aeruginosa isolate was cultured from respiratory material of a patient suffering from nosocomial pneumonia and kindly provided by Prof. Dr. Bastian Opitz (Charité-University Medicine, Berlin, Germany). Of note, the bacterial strain displayed antimicrobial sensitivity to fosfomycin and colistin only.
Cultural analysis of the intestinal loads of P. aeruginosa
On days 2, 3 and 4 post P. aeruginosa infection individual fecal samples were homogenized in sterile PBS, and serial dilutions streaked onto Columbia agar supplemented with 5% sheep blood (Oxoid, Germany) and Cetrimid agar (Oxoid) and incubated in an aerobic atmosphere at 37 °C for at least 48 h to assess intestinal P. aeruginosa loads.
Clinical conditions
Body weights as well as macroscopic and/or microscopic abundance of fecal blood were assessed in individual mice on a daily basis by the Guajac method using Haemoccult (Beckman Coulter/PCD, Germany).
Sampling procedures
Mice were sacrificed 7 days after ileitis induction by isoflurane treatment (Abbott, Germany). Cardiac blood and tissue samples from spleen, liver, lung, kidney, mesenteric lymph nodes (MLN), ileum and colon were removed under sterile conditions. Ileal and colonic samples from each mouse were collected in parallel for microbiological, immunological, immunohistochemical and histopathological analyses. Experiments were repeated at least twice.
Small intestinal lengths and histopathological scores
Small intestinal lengths were determined by measuring the distance from the duodenum leaving the stomach to the ileal-caecal transition. Ex vivo biopsies derived from the terminal ileum were immediately fixed in 5% formalin and embedded in paraffin. Sections (5 µm) were stained with hematoxylin and eosin (H&E) and subjected to a standardized histopathological scoring system ranging from 0 to 6 as described earlier [
11,
14].
Immunohistochemistry
5 µm thin paraffin sections of ileal ex vivo biopsies were used for in situ immunohistochemical analysis as described previously [
18‐
20]. Primary antibodies against CD3 (Polycl.rabbit anti human, DAKO, Denmark; 1:10), FOXP3 (FJK-165, eBioscience, Germany; 1:100), B220 (eBioscience; 1:200) and F4/80 (biot. Clone BM 8 rat anti mouse, Life Technologies, USA; 1:100) were used to assess T lymphocytes, regulatory T cells (Treg), B lymphocytes and macrophages/monocytes, respectively. The average number of positively stained cells within at least six high power fields (HPF, 0.287 mm
2; 400× magnification) were determined by an independent blinded investigator.
Cultural survey of intestinal microbiota and bacterial translocation
For comprehensive quantitative survey of intestinal microbiota composition and translocation of viable bacteria to extra-intestinal compartments, colonic and ileal luminal contents as well as ex vivo biopsies derived from MLN, spleen, liver and lung, respectively, were homogenized in sterile PBS and analyzed in serial dilutions on different solid culture media as described earlier [
11,
14,
21]. Cardiac blood was incubated in thioglycolate enrichment broths (BD Bioscience, Germany) for 1 week at 37 °C and streaked onto solid media thereafter. Bacteria were grown at 37 °C for at least 2–3 days under aerobic, microaerobic and anaerobic conditions.
Molecular analysis of the ileal microbiota
DNA was extracted from fecal samples as described previously [
11,
22]. In brief, DNA was quantified by using Quant-iT PicoGreen reagent (Invitrogen, UK) and adjusted to 1 ng per µL. Then, main bacterial groups abundant in the murine and human intestinal microbiota were assessed by quantitative real-time polymerase chain reaction (qRT-PCR) with species-, genera- or group-specific 16S rRNA gene primers (Tib MolBiol, Germany) as described previously [
16,
18,
23] and numbers of 16S rRNA gene copies per ng DNA of each sample determined.
Cytokine detection in colon, ileum, mesenteric lymph nodes, liver, kidney, spleen and serum
Ex vivo biopsies of approximately 1 cm
2 (ileum cut longitudinally) were washed in PBS and placed in 24-flat-bottom well culture plates (Falcon, Germany) containing 500 mL serum-free RPMI 1640 medium (Gibco, life technologies) supplemented with penicillin (100 U/mL, Biochrom, Germany) and streptomycin (100 µg/mL; Biochrom). After 18 h at 37 °C, culture supernatants and serum samples were tested for IFN-γ, TNF-α, MCP, IL-12p10, IL-6 and IL-10 by the Mouse Inflammation Cytometric Bead Assay (CBA; BD Bioscience) on a BD FACSCanto II flow cytometer (BD Bioscience). Nitric oxide was determined by the Griess reaction as described previously [
11].
Statistical analysis
Mean values, medians, standard deviations (SD) and levels of significance were determined using appropriate tests as indicated (two-tailed Student’s t test, Mann–Whitney U test, ordinary one-way ANOVA and Kruskal–Wallis test). Two-sided probability (p) values ≤0.05 were considered significant.
Discussion
In the present study we were able to show that an acute intestinal inflammatory condition facilitates MDR P. aeruginosa infection of the vertebrate host. Whereas (already fatal) ileal inflammatory changes were similar, extra-intestinal sequelae of high dose T. gondii infection were further aggravated by subsequent P. aeruginosa infection.
Applying a well-known infection model of acute
T. gondii induced ileitis in (with respect to their microbiota) “humanized” mice, we first verified that prior to ileitis induction, human microbiota had stably established within the murine host and that microbiota composition was comparable in experimental groups that were subsequently infected with MDR
P. aeruginosa or remained uninfected. These results are in accordance with our previous studies, where gnotobiotic (i.e. secondary abiotic) mice were replenished with human microbiota in order to overcome colonization resistance against
Campylobacter jejuni. Quantitative molecular analysis further revealed that following peroral fecal transplantation human microbiota could stably establish within the intestinal tract for more than 6 weeks [
16]. Moreover, we showed previously that following peroral infection with high doses (i.e. >50 cysts) of
T. gondii mice developed a distinct Th1-type immunopathology and exhibited an overgrowth of the inflamed ileal lumen with commensal Gram-negative bacterial species such as
Escherichia coli (
E. coli) and
Bacteroides/Prevotella spp. [
11]. Due to progressive virulence of the applied
T. gondii ME49 strain by repeated passages in NMRI “bank” mice for several generations, we also subjected hma mice to 50 cysts in the present study instead of 100 cysts as in our previous reports [
10‐
12,
14,
22,
24‐
29].
Remarkably, acute ileitis induced by 50 cysts of
T. gondii rendered hma mice susceptible to MDR
P. aeruginosa infection. This is in accordance with our previous studies demonstrating that hma mice suffering from small intestinal inflammation could be stably colonized by
C. jejuni [
27]. It is hence tempting to speculate that inflammation induced changes of the intraluminal milieu within the intestinal tract predisposes the host for infection with obligate enteropathogenic species such as
C. jejuni, but also opportunistic pathogens including
P. aeruginosa. Our hypothesis is further supported by a study showing that a nonpathogenic intestinal
E. coli isolate as well as the pathogen
Salmonella typhimurium were able to outcompete the endogenous microbiota of IL-10
−/− mice, a murine model of chronic colitis, suggesting that growth of Gram-negative enterobacteria is enhanced by host mediated intestinal inflammation and following modification of the composition of intestinal microbiota [
30,
31]. Furthermore, in a murine co-infection model intestinal colonization capacity of
S. typhimurium was enhanced by inflammatory changes in the intestinal mucosa and dysbiosis elicited by
Plasmodium yoelii infection [
32].
Given that the intestinal intraluminal milieu is determined by a plethora of factors including the bacteria residing in the intestinal compartments, we performed a comprehensive quantitative survey of the intestinal microbiota composition. Cultural and molecular analyses of the predominant intestinal bacterial groups revealed that differences in the overall microbiota composition of
P. aeruginosa infected and uninfected hma mice with induced ileitis were rather subtle at the first glance. Ileal loads of bifidobacteria, however, were more than two orders of magnitude lower following
P. aeruginosa infection. As commensal residents with beneficial effects for intestinal homeostasis, bifidobacteria contribute to mucosal barrier functions directed against pathogenic colonization of the vertebrate host [
33,
34]. Two bifidobacterial strains that were isolated from resident infant human gastrointestinal microbiota exerted antibacterial activity against several pathogens such as
S. typhimurium, E. coli, Klebsiella pneumoniae, Yersinia pseudotuberculosis, Staphylococcus aureus and
P. aeruginosa in vitro [
34]. In addition, in vitro as well as in vivo studies revealed that
Bifidobacterium animalis AHC7, for instance, exerts an anti-inflammatory effect through the attenuation of NF-κB activation in response to murine
S. typhimurium infection and that stimulation of dendritic cells with
B. animalis AHC7 significantly increased CD25+Foxp3+ T cell (Treg) numbers [
35]. We could further show previously that conventionally colonized mice deficient in the innate immune-receptor nucleotide-oligomerization-domaine-2 (NOD2) were virtually lacking the bifidobacterial population in their intestines and (at least in part) exhibited compromised host resistance, reduced local anti-inflammatory and increased systemic pro-inflammatory immune response upon peroral high dose
T. gondii infection [
29].
Given that acute
T. gondii induced ileitis is highly dependent on Toll-like Receptor (TLR)-4 mediated signaling of lipopolysaccharide derived from Gram-negative intestinal commensals [
11,
15], one could speculate that additional infection with a Gram-negative bacterium such as
P. aeruginosa might further exacerbate the induced inflammatory process. Our study, however, revealed that clinical and ileal changes, the profound influx of innate and adaptive immune cells into the distal small intestines and increased secretion of pro-inflammatory cytokines in ileum and MLN following
T. gondii infection were not further aggravated by MDR
P. aeruginosa infection. This is surprising considering the plethora of virulence factors of
P. aeruginosa. For instance,
Pseudomonas lipid A, a core component of bacterial lipopeptide, has been shown to activate NFκB signaling through TLR-4 and subsequent pro-inflammatory cytokine secretion [
36]. In turn, neutrophils are recruited to the infection site and contribute to the inflammatory host response to
P. aeruginosa [
1]. One needs to take into consideration, however, that peroral high dose
T. gondii infection (irrespective whether performed with 50 or 100 cysts) results in a profound Th1-driven hyper-inflammatory scenario (“cytokine tsunami”) [
13] that cannot further be deteriorated by additional MDR
P. aeruginosa infection. Strikingly, we were able to observe elevated IFN-γ secretion not only in ileal, but also colonic ex vivo biopsies 7 days post ileitis induction. To our best knowledge, the terminal ileum has been reported as exclusive predilection site following peroral high dose
T. gondii infection of conventionally colonized mice so far [
13].
Remarkably, extra-intestinal sequelae of acute ileitis induction were further amplified by P. aeruginosa challenge as indicated by even more pronounced increases in hepatic TNF-α concentrations following P. aeruginosa infection of T. gondii pre-infected mice. Potential inflammatory effects of MDR P. aeruginosa infection in the small intestines (and beyond) should therefore be more distinctly deciphered either in a less acute (i.e. less severe) or more chronic infection model following peroral low dose infection with less than 10 cysts of T. gondii, for instance. To the best of our knowledge, however, a chronic T. gondii ileitis model has not been established so far. Alternatively, experimental models of large intestinal inflammation could be applied to investigate host susceptibility and the pro-inflammatory potential of peroral MDR P. aeruginosa infection during intestinal inflammation in a better discriminatory way than with the model used here. To address this, the murine dextran sulfate sodium induced colitis model would be a promising candidate, for instance.
In the present study we were further able to demonstrate that anti-inflammatory IL-10 levels were multifold increased in the MLN following ileitis induction, but even more distinctly upon additional P. aeruginosa infection, whereas conversely, serum IL-10 concentrations were lower in P. aeruginosa infected as compared to non-infected hma mice suffering from acute ileitis. Hence, the more pronounced intestinal anti-inflammatory response upon P. aeruginosa application was not sufficient to counteract the pro-inflammatory sequelae caused by T. gondii and P. aeruginosa co-infection. One might have also expected comparable increases in systemic IL-10 levels, given that translocation of viable bacteria originating from the intestinal microbiota to systemic compartments such as spleen and blood had occurred more frequently in P. aeruginosa co-infected versus non-infected hma mice with acute ileitis.
We were finally able to show that viable intestinal bacteria were more frequently detected in systemic compartments such as spleen and blood of
P. aeruginosa infected as compared to uninfected hma mice with induced ileitis indicating that (even though not otherwise determined in our study) epithelial barrier leakage was supposably even more pronounced by
P. aeruginosa co-challenge than by acute ileitis induction alone, further facilitating bacterial translocation to extra-intestinal including systemic sites. Notably, several studies illustrate a link between the etiology of inflammatory bowel diseases (IBD) including Crohn’s disease and the abundance of
Pseudomonas species such as
P. fluorescens within the intestinal mucosa [
37‐
39]. Solid data regarding the role of
P. aeruginosa in IBD pathogenesis, however, are scarce. One study applying molecular analysis of
Pseudomonas specific 16S RNA in ileal tissue samples derived from children suffering from Crohn’s disease revealed a higher prevalence of several
Pseudomonas species including
P. proteolytica and
P. brenneri in pediatric Crohn’s disease patients as compared to control individuals without IBD, whereas interestingly
P. aeruginosa could be detected in non-IBD patients only [
40]. In patients suffering from irritable bowel disease, however, detection rates of
P. aeruginosa specific 16S RNA were increased in duodenal mucosa-associated biopsies and fecal samples [
8]. Furthermore, a case report of a child suffering from ulcerative colitis revealed that
P. aeruginosa 16S rRNA could be identified in colonic biopsies [
9]. Already an older study from 1966 suggested a temporal correlation between identical
P. aeruginosa strains isolated from patients’ lesions and feces pointing towards a spread of viable
P. aeruginosa via the blood stream [
7]. Meanwhile, there is evidence that the risk of developing a clinically manifest
P. aeruginosa infection is significantly higher upon rectal colonization of patients in intensive care units [
41].
Authors’ contributions
Conceived and designed the experiments: EVK, MMH. Performed the experiments: EVK, IE, MMH. Analyzed the data: EVK, IE, MMH. Wrote the paper: EVK, SB, MMH. All authors read and approved the final manuscript.