Introduction
Ventilator-associated pneumonia (VAP) occurs in a considerable proportion of patients undergoing mechanical ventilation and is associated with substantial morbidity, a two-fold increase in mortality rate, and excess cost [
1]. Tracheobronchial colonization (TBC) and duration of mechanical ventilation are the two most important risk factors for VAP [
2,
3].
Pseudomonas aeruginosa is one of the most frequent causative microorganisms of VAP [
2‐
4]. Several studies have reported the presence of
Candida species in the airway specimens of immunocompetent ventilated patients [
5,
6].
Candida TBC occurs in 17% to 28% of ICU patients receiving mechanical ventilation for more than 48 hours [
7‐
9]. Although the relationship between tracheal biofilm and VAP is based on one small observational study,
P. aeruginosa is the most common pathogen retrieved from endotracheal tube biofilm in patients with VAP [
10].
P. aeruginosa and
C. albicans coexist predominantly as biofilms rather than as free-floating (planktonic) cells on abiotic medical devices (catheters and prostheses) [
11,
12].
The question of their interplay has been addressed by several experimental and clinical studies. So far,
in vitro studies suggest that the interaction between
C. albicans and
P. aeruginosa is likely to be antagonistic. When mixing
in vitro cultures,
P. aeruginosa is involved in killing
C. albicans filaments associated with biofilm formation [
13]. Additionally, quorum-sensing signaling molecules of
P. aeruginosa impair
C. albicans yeast-to-hyphae transition [
14]. The relative
C. albicans hyphal-binding affinity within biofilm is reported to be lower for
P. aeruginosa than for
Staphylococcus aureus [
15]. In contrast, a synergistic relationship is described
in vivo with a recent study showing that
C. albicans TBC facilitates
P. aeruginosa pneumonia occurrence in a rat model [
16]. A recent clinical study suggested an interaction between
C. albicans and
P. aeruginosa [
8]. The authors identified
Candida spp. tracheobronchial colonization as an independent risk factor for
P. aeruginosa pneumonia. No cause-and-effect relationship was demonstrated in that study. In addition,
Candida spp. tracheobronchial colonization and
P. aeruginosa pneumonia could both be a consequence of prior antibiotic treatment. Further, the median duration of mechanical ventilation in that study was 13 days. Therefore, the results could not be generalized to patients with shorter duration of mechanical ventilation. Another recent preliminary case-control study suggested that antifungal treatment might be associated with reduced risk for VAP or TBC related to
P. aeruginosa [
9], although no definite conclusion can be drawn from this observational retrospective single-center study including a small number of patients.
The study of P. aeruginosa and C. albicans interactions in the respiratory tract aims at more effectively understanding the balance between microbial ecology and bacteria-related pathogenesis. This issue has major environmental and medical consequences. The present study proposes to investigate P. aeruginosa-related lung injury in mice previously colonized with C. albicans and to evaluate the impact of caspofungin antifungal treatment.
Materials and methods
Animals
BALB/c mice (20 to 25 g) purchased from Charles River Laboratories (Domaine des Oncins, L'Arbresle, France) were housed in a pathogen-free unit of the Lille University Animal Care Facility and allowed food and water ad lib. All experiments were performed with the approval of the Lille Institutional Animal Care and Use Committee.
Growth conditions for bacterial and yeast strains
The wild type strain
Pseudomonas aeruginosa PAO1 was grown in Luria-Bertani medium at 37°C for 16 h and was centrifuged at 3,000 × g for 10 minutes. The bacterial pellets were washed twice and diluted in an isotonic saline solution to obtain an optical density of 0.63 to 0.65 nm determined by spectrophotometry [
17].
The reference strain
C. albicans SC5314 was maintained at 4°C on Sabouraud dextrose agar (SDA) [
18]. For the study, cell of broth test isolates were grown in SDA at 37°C in a shaking incubator for 18 h.
Mice infection
Mice were infected by direct intratracheal inoculation under short anaesthesia with inhaled sevoflurane (Servorane™, Abbott, Queenborough, UK) as previously described [
17]. For each mouse, 50 μl of fungal or bacterial suspension containing 2 × 10
6 or 2 × 10
7 or 2 × 10
8 colony-forming units (CFU)/ml of yeasts or 2 × 10
8 CFU/ml of bacteria respectively, was instilled. Control mice received 50 μl of sterile saline solution.
Treatment with caspofungin
Caspofungin (Merck & Co. Inc., Whitehouse Station, NJ, USA) was injected intraperitoneally once daily either from T0 or from 24 h post-C. albicans challenge. The full recommended dose of 1 mg/kg was administered the first day of treatment and then 0.8 mg/kg was administrated on Days 2, 3, and 4.
Quantitative blood culture and pulmonary bacterial and fungal loads
For bacterial blood culture, 100 μl of blood was plated on bromocresol purple (BCP) agar plates for 24 h at 37°C to allow for P. aeruginosa growth. In co-infected groups, BCP agars were treated with 50 μg per plate of caspofungin. For fungal blood culture, the same amount was plated on yeast peptone dextrose (YPD) agar plates containing 1% yeast extract, 1% peptone, 2% D-glucose and 500 mg/l amikacin sulphate and incubated for 48 h at 37°C to allow for C. albicans growth.
For quantification of lung bacterial loads, lungs were removed after exsanguination via intracardiac puncture and homogenized in 0.9 ml of sterile isotonic saline solution. Viable bacteria were counted after serial dilutions of 100 μL of lung homogenate on BCP agar plates for 24 h at 37°C to allow for P. aeruginosa growth. Similarly, another 100 μL of lung homogenate was plated on YPD plates for 48 h to allow for C. albicans growth. In co-infected groups, agar was treated with caspofungin or amikacin.
In vivo quantification of acute lung injury: alveolar-capillary barrier permeability
125I-albumin was injected as a vascular protein tracer and its leakage across the endothelial barrier and accumulation in the extravascular spaces of the lungs was measured using a previously described permeability index [
19]. More details are provided in the Additional file
1.
Determination of histological score
At Days 2 and 4, the lungs were removed and fixed overnight in 4% paraformaldehyde-acid and embedded in paraffin for histologic analysis. Cross-sections (3 μm thick) were stained with hematoxylin and eosin stain (Sigma-Aldrich Europe, Saint-Quentin Fallavier, France) and periodic acid Schiff. Two independent blinded investigators graded the inflammation score. The degree of peribronchial and perivascular inflammation was evaluated on a subjective scale of 0 to 3, as described elsewhere [
20].
Fluorescence staining of C. albicans in situ
Paraffin-embedded lung sections were stained with either the monoclonal antibody (mAb) 5B2 or the galenthus nivalis lectin [
21,
22] and examined by immunofluorescence microscopy (Leica Microsystems AG, Heerbrugg, Switzerland).
Experimental groups
Animals were randomly assigned to the following groups: Ca: mice infected with 1 × 10
5 CFU of
C. albicans at T0 and sacrificed at Day 2 or 4; Pa: mice infected with 1 × 10
7 CFU of
P. aeruginosa at Day 2 and sacrificed at Day 4; CaPa: mice infected with 1 × 10
5 CFU of
C. albicans at T0, infected with 1 × 10
7 CFU of
P. aeruginosa at Day 2 after infection by
C. albicans, and sacrificed at Day 4; CaPaCasp0 and CaPaCasp1: mice infected with 1 × 10
5 CFU of
C. albicans at T0, treated with caspofungin from T0 or from Day 1 to Day 4, infected with 1 × 10
7 CFU of
P. aeruginosa at Day 2 after infection by
C. albicans, and sacrificed at Day 4. The experimental design is further detailed in Table
1. The sample size was four (microbial count assay), five (mortality assay), and eight animals (permeability index assay) per group. Each experiment was performed in duplicate.
Table 1
Experimental design of the study
Ctr
| / | Saline solution | none | Day 2, Day 4 |
Ca
| T0 | 1 × 105 C. albicans | none | Day 2, Day 4 |
Pa
| Day 2 | 1 × 107 P. aeruginosa | none | Day 4 |
CaPa
| T0, Day 2 | T0: 1 × 105 C. albicans d2: 1 × 107 P. aeruginosa | none | Day 4 |
CaPaCasp0
| T0, Day 2 | T0: 1 × 105 C. albicans d2: 1 × 107 P. aeruginosa | T0 to Day 4 | Day 4 |
CaPaCasp1
| T0, Day 2 | T0: 1 × 105 C. albicans d2: 1 × 107 P. aeruginosa | Day 1 to 4 | Day 4 |
Statistical analysis
Mortality rates were compared between groups by using the log rank test with Kaplan-Meier analysis. Data were analyzed by Kruskal-Wallis one-way analysis of variance test using Dunn's method to compare differences between groups (GraphPad Prism, v5.0, La Jolla, California, USA). Data are expressed as means ± standard error of the mean (SEM). P-values below 0.05 were considered significant.
Discussion
The present study was designed to determine the contribution of C. albicans airway colonization to P. aeruginosa pathogenicity in immunocompetent mice. Our results indicate that prior short-term C. albicans airway colonization reduced P. aeruginosa-induced lung injury and the amount of live P. aeruginosa in lungs. This effect is reversed by fungicidal drug caspofungin when initiated concomitantly to C. albicans infection.
The prerequisite to the study was the set-up of tracheobronchial colonization by
C. albicans according to the definition of colonization, which is the presence of a pathogen that does not cause damages on the lung parenchyma. The dose of 1 × 10
5 CFU of
C. albicans per mouse matched with this criterion as no invasive disease occurred. After
P. aeruginosa infection, a trend toward a higher survival rate in the
C. albicans-colonized mice was observed. This result is consistent with data comparing groups of mice instilled simultaneously with
P. aeruginosa and
C. albicans or with
P. aeruginosa alone showing a significant difference of survival in favor of the
C. albicans-colonized group at Day 7 [
23]. Then, it was found that previous
C. albicans airway colonization was associated with an increase in lung
P. aeruginosa clearance compared to the non-colonized-group. These data differ from the previous study, which did not detect a significant decrease in quantitative bacterial burden in the group receiving simultaneous administration of
C. albicans along with
P. aeruginosa [
23]. However, a major difference is that bacterial loads were recorded early after the co-infection between 3 and 20 h. Another study, which addressed the issue of prevalence of
P. aeruginosa pneumonia in rats colonized by
C. albicans, evaluated the quantitative bacterial cultures of
P. aeruginosa in rat lungs at 48 h post-infection [
16]. Subsequent to
C. albicans colonization obtained by intratracheal instillation (2 × 10
6 CFU per rat three days in a row), a low dose of
P. aeruginosa (1 × 10
4 CFU per rat) was delivered at Day 2 post-colonization. The bacterial burden was significantly higher at 48 h in rats instilled with
C. albicans before
P. aeruginosa compared to rats instilled with saline solution or ethanol-killed
C. albicans before
P. aeruginosa. Contrary, in our experimental model,
P. aeruginosa dissemination in the bloodstream showed a trend toward a decrease in the case of prior
C. albicans colonization. Although bacterial dissemination is multifactorial depending on the magnitude of alveolar-capillary barrier injury [
24] as well as the strain virulence and the size of the inoculum [
25], the decrease was most likely due to the decrease of alveolar-capillary barrier injury since the
P. aeruginosa strain used and the size of the inoculum administrated were identical in both groups. Regarding histolopathologic results, the inflammation score decreased in the case of previous
C. albicans airway colonization in comparison to
P. aeruginosa infection alone suggesting that primary immune activation could reduce
P. aeruginosa pathogenicity. This observation was consistent with a decrease in
P. aeruginosa lung loads and a decrease in the lung permeability index in the CaPa group in comparison to the Pa group. These results differ from a study previously mentioned which concluded that previous
C. albicans colonization lowered the threshold of
P. aeruginosa load necessary to induce parenchymal injury since in rats given
C. albicans, histologic aspect of
P. aeruginosa pneumonia was significantly more frequent than in controls or ethanol-killed
C. albicans rats [
16]. Overall, the results may differ between mice and rats, and between different strains of
P. aeruginosa owing to differential susceptibility to pneumonia.
In the second part of this study, the influence of fungicidal caspofungin was tested. The use of caspofungin aimed at detecting a difference between colonization with live or killed C. albicans. For that purpose, two target times for treatment initiation were chosen, from T0 or from Day 1. Regarding the alveolar-capillary barrier injury, the use of caspofungin resulted in distinct effects: reversal of the decrease in the protein tracer leakage when initiated at T0 or maintenance of the decrease in the protein tracer leakage when initiated at Day 1. The difference observed between the T0- and the Day 1-treated group suggests that the viability and/or the growth of C. albicans makes a difference in reducing the magnitude of alveolar-capillary barrier injury.
Overall, these results raise three hypotheses: first, a competitive effect regarding the adhesion of the pathogens to lung epithelial surface. Indeed, they both use ligands to recognize the glycoconjugates at the surface of epithelial cells [
26,
27]. Recently, it has been demonstrated that
P. aeruginosa lectins LecA and LecB, which are involved in adhesion to epithelial cells, contribute to
P. aeruginosa-induced lung injury [
17]. The neutralization of these lectins by the administration of specific lectin inhibitors was remarkably effective in improving lung injury.
C. albicans adherence to host tissue is controlled by the ALS (agglutinin-like sequence) gene family which encodes a group of glycosyl-phosphatidyl-inositol (GPI)-linked cell surface proteins that function as adhesins that bind to the cell surface [
28]. The second hypothesis is a bactericidal effect mediated by higher-inducible lung mucosal innate response by live
C. albicans. This hypothesis is supported by the decrease in inflammation score in case of previous
C. albicans airway colonization in comparison to
P. aeruginosa infection alone, and by the fact that T0 caspofungin treatment resulted in a higher rate of bacterial growth. Finally,
C. albicans produces farnesol, a cell-to-cell signaling molecule that could act as a quorum-sensing antagonist of
P. aeruginosa [
14,
29]. The addition of farnesol to cultures of
P. aeruginosa leads to decreased production of the
Pseudomonas quinolone signal (PQS) and the PQS-controlled downstream virulence factor, pyocyanin [
30]. Furthermore, it has been shown that the
C. albicans farnesol has also the ability to inhibit swarming motility in
P. aeruginosa cystic fibrosis clinical isolates [
31]. All together, the reduction in PQS-pyocyanin production and swarming mobility may also have implications for the interaction between
P. aeruginosa and the host.
The present study has several limitations that prevent extrapolating the results to the chronically colonized and/or critically ill patients at risk for VAP. First, the short term
C. albicans colonization in the model does not correctly reflect the situation of these patients. Indeed, the amount of
C. albicans in lungs cannot be substantially sustained over time in immunocompetent BALB/c mice, as already described elsewhere [
32]. Consequently,
P. aeruginosa pneumonia had to be generated only 48 hours after the prior fungal colonization. The addition of a control experimental group testing the impact of killed
C. albicans would have been of interest to assess the need of live
C. albicans to produce the effects described. Concern can also be raised regarding some
in vitro data indicating a decrease of
P. aeruginosa growth following exposure to halogenated anesthetics [
33], although it occurred after several hours of exposure and has not been investigated
in vivo. The short duration of mutual contact and interaction of
C. albicans and
P. aeruginosa in the airways (48 h) represents another potential bias of the present study. Furthermore, a dose/effect study testing various doses of
P. aeruginosa to generate pneumonia could better document the
in vivo dynamics of bacterial-fungal interactions. Performing microbial CFU counts in spleen and liver could better assess microbial dissemination. Finally, this relationship is studied in normal lungs and in the absence of any airway prosthetic device, which largely promotes microbial community networking [
12].
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
FA participated in the design of the study, carried out the in vivo experiments, performed the statistical analysis and drafted the manuscript. SJ carried out the histological and immunofluorescence assays and helped to draft the manuscript. SN, BS KF, FV, CC, DP and BG participated in the design and coordination of the study and helped to draft the manuscript. All authors read and approved the final manuscript.