We present an experimental procedure of Candida albicans biofilm development in a mouse subcutaneous model. Fungal biofilms were quantified by determining the number of colony forming units and by a non-invasive bioluminescence imaging, where the amount of light that is produced corresponds with the number of viable cells.
Candida albicans biofilm development on biotic and/or abiotic surfaces represents a specific threat for hospitalized patients. So far, C. albicans biofilms have been studied predominantly in vitro but there is a crucial need for better understanding of this dynamic process under in vivo conditions. We developed an in vivo subcutaneous rat model to study C. albicans biofilm formation. In our model, multiple (up to 9) Candida-infected devices are implanted to the back part of the animal. This gives us a major advantage over the central venous catheter model system as it allows us to study several independent biofilms in one animal. Recently, we adapted this model to study C. albicans biofilm development in BALB/c mice. In this model, mature C. albicans biofilms develop within 48 hr and demonstrate the typical three-dimensional biofilm architecture. The quantification of fungal biofilm is traditionally analyzed post mortem and requires host sacrifice. Because this requires the use of many animals to perform kinetic studies, we applied non-invasive bioluminescence imaging (BLI) to longitudinally follow up in vivo mature C. albicans biofilms developing in our subcutaneous model. C. albicans cells were engineered to express the Gaussia princeps luciferase gene (gLuc) attached to the cell wall. The bioluminescence signal is produced by the luciferase that converts the added substrate coelenterazine into light that can be measured. The BLI signal resembled cell counts obtained from explanted catheters. Non-invasive imaging for quantifying in vivo biofilm formation provides immediate applications for the screening and validation of antifungal drugs under in vivo conditions, as well as for studies based on host-pathogen interactions, hereby contributing to a better understanding of the pathogenesis of catheter-associated infections.
Candida albicans is a commensal organism, which can be found at different sites of healthy individuals, for example on the skin or as a part of the gastrointestinal and vaginal flora. However, in hospitalized, and especially immunocompromised patients, it may cause a wide range of infections 1. In such individuals, the weakened immune system allows Candida cells to disseminate into the bloodstream and to invade deeper tissues causing life-threatening infections. In addition, the presence of abiotic substrates such as central venous and urinary catheters, artificial heart valves and joints may provide a niche for Candida attachment 2. Adhesion to such substrates is a prerequisite for further biofilm development, which represents a layer of yeast and hyphal cells embedded in extracellular polymeric material, mainly consisting of polysaccharides 2. C. albicans catheter –associated infections are associated with high mortality rate. A general characteristic of biofilms is their decreased susceptibility to known antifungals, such as azoles 3,4. Only newer classes of antifungal drugs, such as echinocandins and liposomal formulation of amphotericin B proved to be active against catheter-associated infections 5-7. Because of biofilm resilience to antifungals, therapeutic approaches are very limited, often leading to catheter removal and its subsequent replacement as a sole solution.
Most of our current understanding of C. albicans biofilm development originates from in vitro studies on abiotic substrates such as polystyrene, or plastics used for the manufacture of above-mentioned devices, i.e., silicone, polyurethane 2. These models are quite advanced and try to mimic the situation in vivo as closely as possible. However, these systems do not involve the continuous blood flow and the immune system of the host. This resulted in the development of in vivo model systems, such as the central venous catheter (CVC) model 8-10, the denture stomatitis model of oral candidiasis 11 and a murine model for catheter-associated candiduria 12. Additionally, C. albicans biofilm development was studied in vivo on the mucosal surfaces, such as those from the vagina 13 and oral cavity 14. Our laboratory contributed with the establishment of a subcutaneous C. albicans biofilm model, which is based on the implant of infected catheter pieces on the back of Sprague Dawley rats 15. This model was successfully used in our laboratory to test biofilm susceptibility to fluconazole and echinocandin drugs 5,16, to study the effect of combinatorial therapy of diclofenac and caspofungin 17. More recently, we adapted this system for use in BALB/c mice 18,19. In comparison with other in vivo models, the main advantage of this subcutaneous model is the possibility to study multiple biofilms per animal developed inside the lumen of implanted catheter pieces.
To reduce the number of laboratory animals, we have adapted this model to study the development of C. albicans biofilms non-invasively by using bioluminescence imaging (BLI) 18,19. This method proved to be a powerful technique, which can be used to quantify biofilms by measuring the specific BLI signal at the region of interest (in our case the area of implanted catheters), avoiding animal sacrifice. In comparison to bacteria, which can express both the gene and the substrate required for the bioluminescence reaction due to the introduction of a specific lux operon 20, most of the eukaryotic organisms, including C. albicans, are dependent on the heterologous expression of a luciferase gene coupled with the external administration of a specific substrate, such as D-luciferin or coelenterazine 21. Probably due to the presence of the fungal cell wall and C. albicans morphogenesis, the intracellular delivery of the substrate for the luciferase enzyme was a main challenge 21. In order to solve this problem, Enjalbert et al. 22 engineered a strain where a synthetic C. albicans codon-optimized version of the gene for the naturally secreted Gaussia princeps luciferase (gLuc) was fused to to the C. albicans PGA59 gene, a GPI- anchored cell wall protein. Because of the presence of luciferase at the cell wall, problems concerning the intracellular availability of the substrate could be avoided. This particular system was used to study superficial infections caused by C. albicans 22. Very recently, BLI was also used to follow the progression of oropharyngeal candidiasis and its possible treatment 23. Such findings support the use of BLI as a promising technique to study infections caused by free-living cells but also device-associated infections.
In this study, we describe the C. albicans biofilm development on polyurethane catheter pieces in BALB/c mice and its quantification using BLI. We provide a detailed protocol of in vitro colonization of polyurethane catheters during the period of adhesion followed by implantation in mice and subsequent biofilm development in live animals. Apart from measuring the BLI signal emitted by the C. albicans cells, we also determine the colony forming units for comparison with the standard technique for biofilm fungal load quantification.
NOTE: All animal experiments were approved by the ethical committee of KU Leuven (project number 090/2013). Maintain animals in accordance with the KU Leuven animal care guidelines.
1. C. albicans Growth
2. Catheter Pieces Preparation
3. Animals and Suppression of the Immune System
4. Ex vivo C. albicans Adhesion on FBS-coated Polyurethane Substrates
5. Anesthesia
6. Animal Surgery
7. Bioluminescence Imaging: Preparation of Coelenterazine (CTZ), the Substrate for G. princeps luciferase
8. Bioluminescence Imaging
9. Catheter Explant
10. Quantification of Biofilm-associated Cells by Colony Forming Units Count (CFUs) and Statistical Analyses of Results
In this study, we show the surgical procedure of catheter implant and explant during in vivo C. albicans biofilm development in a mouse. Moreover, we display the quantification of mature biofilms not only by classical CFUs enumeration but also by BLI.
As shown in Figure 1A, non-phosphorescent polyurethane catheter pieces were cut into 1 cm devices and subsequently coated with serum. This step is very important because it allows Candida cells to attach to the substrate more rapidly in comparison with non-serum coated implants 15. It is important to mention that prior to any C. albicans experiment we first documented the background luminescence of our devices 18. This step is crucial before performing any BLI because high phosphorescence of the device will interfere with the evaluation of the specific bioluminescent signal intensity and signal kinetics from gLuc-expressing cells. Next, C. albicans cells are incubated with the serum-coated catheter pieces. Catheter fragments are then implanted on the back part of the animal after the subcutaneous incision (Figure 2A). In this in vivo set up, two incisions are performed (one on the right and another one on the left side of the back of a mouse) followed by formation of subcutaneous tunnels inside each incision (Figure 2A (3)). Subsequently, three devices, previously infected with Candida cells, are implanted inside each operation site (Figure 2A (3 and 4)). This set up allows us to study up to six biofilms per animal. It is noteworthy that two operation sites provide a possibility to study biofilm formation by two different strains of interest, for example wild type and mutant in the same animal. Figure 2B displays the wound containing catheters prior to the device explant and subsequent washing steps. Biofilms developed inside the back part of the animal were assessed for the bioluminescence signal measurements. One of the representative animals displaying the bioluminescence signal together with rectangles characterizing the regions of interest (ROIs) is shown in Figure 3. After the last BLI time point, catheters are explanted, sonicated and subsequently vortexed and further assessed for the biofilm-forming cells quantification by CFUs. Data analyses demonstrating the Log10 CFUs obtained from each catheter piece after biofilm development and explantation are shown in Figure 4A. Next to that, CFUs were compared with the BLI signal intensity and these data are shown in Figure 4B. Ninety min after implantation a clear BLI signal is produced by the ACTgLuc-expressing biofilms whereas there in the wild type strain, hardly any light is produced; The intensity of the light detected from the ACTgLuc-expressing biofilms is significantly increasing as the biofilm is formed, here following the same mouse (Figure 4C). This increase in light follows the same trend as the increase in CFUs per biofilm (Figure 4A). In our experiments we also observed that a normal wild type strain (not engineered to express luciferase) also results in the production of light upon addition of the substrate. However, the photon flux was significantly lower than that obtained for the engineered strain.
Taken together, our data illustrate that BLI is a powerful technique to monitor and quantify in vivo mature C. albicans biofilm formation in a subcutaneous mouse model.
Figure 1: Preparation of polyurethane devices. Polyurethane part of the catheter cut into 1 cm pieces. Plastic pocket containing catheter is open under the sterile conditions and all parts, except catheter are removed from the package. Secondly, place ruler under the plastic pocket and cut exactly 1 cm polyurethane pieces. Such devices are subsequently distributed to microcentrifuge tubes (max 15 pieces/tube) and submerged in 100% fetal bovine serum followed by overnight incubation at 37 °C.
Figure 2: Major steps during the animal surgery. (A) Procedure of catheter implant. (1) Place anesthetized animal on a warm pad containing paper tissue and apply the ophthalmic ointment on eyes. (2) Shave lower part of the back and disinfect. (3) Create two small (approx. 0.5 cm) incisions through the skin on the left and on the right side of the back. Create subcutaneous tunnel inside each incision and place 3 catheters inside. (4) Close the wound with sutures and place animal on a warm pad to recover. (B) Procedure of catheter removal. (1) Place sacrificed animal on a pad and disinfect the operated side containing catheters. Cut the wound right above the catheters. (2) Take out each catheter gently and wash twice with 1 ml of PBS. Place to the separate microcentrifuge tube.
Figure 3: Bioluminescence signal measurement. In vivo BLI image from one representative mouse imaged after 6 days of biofilm development. The quantification of signal intensity is performed by placing a region of interest (ROI) (rectangular) around the catheters followed by measurement of the photon flux per second through every ROI.
Figure 4: Quantification of mature Candida albicans biofilms by colony forming units (CFUs) and by BLI. (A) In vivo mature C. albicans SKCA23-ACTgLuc and SC5314 (WT) biofilm quantification by Log10 CFUs after 90 min, 2 days and 6 days of biofilm development. (B) Bioluminescence signal intensity quantification from in vivo biofilms formed by SKCA23-ACTgLuc and by C. albicans WT. Signal was determined after 90 min (period of adhesion), 2 days and 6 days of biofilm development. As a control, we used mice that were implanted with non-colonized catheters and that were subcutaneously injected with CTZ. The photon flux obtained in these mice results in our background signal (BG). Data were expressed as mean ± standard deviation (SD). Statistical significance is indicated as follows: * p <0.05, ** p <0.005, *** p <0.0005. (C) In vivo bioluminescence images from one representative mouse that was implanted with catheter fragments containing wild type C. albicans cells on the left side and ACTgLuc-expressing cells on the right side. The mouse was imaged at three different time periods, i.e. 90 min, 2 days and 6 days after implantation of the catheter fragments.
The use of animal models, and especially rodent models, for studies dedicated to microbial biofilms is very important as the host immune system is an essential factor in biofilm formation that in vitro models cannot account for. In this study, we describe a relatively straightforward subcutaneous C. albicans biofilm mouse model, which can be easily adopted in a research laboratory and does not require strong technical skills. This model was originally developed to study Staphylococcus epidermidis biofilm formation in a rat 24.
In the presented model, serum-coated polyurethane catheters were challenged with Candida cells ex vivo during the period of adhesion (90 min, 37 °C). This first in vitro step might be considered as a limiting factor because of the lack of the immune system at the very first stages of the biofilm infection process. Following adhesion and before catheter implant, non-device associated cells are removed by washing. Avoiding the washing step after the period of adhesion may create uncontrollable amount of cells associated with the device. Because of this reason, we suggest to wash catheters prior to the implant and therefore, to initiate its development from adhered cells. After this initial adhesion period, catheters contain approximately 2.0-2.5 Log10 CFUs/device spread alongside the catheter lumen 15. During this stage Candida forms germ tubes, which are attached onto the substrate.
In order to resemble the situation in hospitalized patients of whom the immune system is often compromised, we immunosuppressed the rats in our subcutaneous model system 15. Partial impairment of the immune system of a rat, prior to the device implant and throughout the period of biofilm development, resulted in increased reproducibility of biofilm-forming cells retrieved from catheters implanted in the same host and also from devices obtained from additional animals. Because of these findings we suggest to immunosuppress the mice before catheter implant and also during the period of biofilm formation. However, our results show that the variability in immunocompetent mice is much less compared to that observed in rats, which makes that mice model system also suitable for studying the role of the host immune system on biofilms. In our original rat model and also in the same model translated to mice, mature C. albicans biofilms develop within 48 hr demonstrated by the similar amount of CFUs and biofilm architecture between 2 and 6 days 15,18,19.
The subcutaneous biofilm model was successfully used to follow biofilm development by several mutants 15. It has been shown previously by Nobile et al. 25 that C. albicans bcr1∆/bcr1∆ failed to form biofilms in a central venous catheter (CVC) model. This strain displayed rudimentary biofilm features in our subcutaneous model pointing to the fact that the phenotypic changes found in the CVC model could be reproduced in the subcutaneous model 15.
In comparison with other existing models, i.e., CVC model or denture stomatitis model 8,9,11, the subcutaneous model allows to follow biofilm development in multiple catheter pieces obtained from one animal, thereby reducing the number of animals needed for biofilm studies. Despite these advantages, a shortcoming may be the lack of blood flow that may lead to certain deprivation of nutrients in the biofilm. Therefore, in term of nutrient supplies and environmental conditions, the subcutaneous model is more related to device-associated infections developed on joint prostheses, voice prostheses and pacemakers than the ones formed on intravenous catheters. Even more importantly, translating the rat subcutaneous biofilm system to mice made this animal model compatible with BLI, which further reduces costs and most importantly, the number of necessary animals. Furthermore, translating the rat model to mice enables the use of transgenic mouse models to study host factors relevant to catheter-associated infection studies.
The major advantage of using BLI to quantify biofilms in live animals lies not only in the non-invasive character of this method, but also in its ability to provide dynamic information on the development of an infection. In this study, gLuc expressed at the cell wall of C. albicans allowed better accessibility and direct interaction with the substrate coelenterazine, which are crucial for a detection of bioluminescent signal in vivo. In the study of Vande Velde et al. 18, we were able to follow the bioluminescent signal from C. albicans biofilms developed on foreign bodies in vitro. The BLI signal intensity strongly corresponded with the data obtained from additional biofilm quantification techniques, i.e., CFUs determination and XTT reduction assay. Next to these results, we showed that the bioluminescent signal from in vivo biofilms was in agreement with the amount of biofilm-associated cells recovered from explanted biofilms. Additionally, Vande Velde et al. 18 demonstrated that BLI can be used to follow biofilm development by a wild type and also by C. albicans bcr1∆/bcr1∆(biofilm-deficient strain) in the same animal at the same time. Such findings strongly support the use of BLI during the studies dedicated to assessing the time course of biofilm development and infection in the same animal.
Herewith, we described an experimental procedure for in vivo subcutaneous implant of Candida-infected devices with subsequent monitoring of in vivo biofilm development by BLI. Taken together, BLI showed to be a reliable technique, which allowed us to follow an infection over time avoiding animal sacrifices at each time point of data analyses.
The authors have nothing to disclose.
This work was supported by the KU Leuven PF ‘IMIR’, the FWO Research community on biology and ecology of bacterial and fungal biofilms (FWO: WO.026.11N) and by the FWO project G.0804.11. SK gratefully acknowledges KU Leuven for the PDMK 11/089 fellowship and FWO for the postdoctoral fellowship. We are grateful to Nico Vangoethem for his assistance with preparation of the figures. We would like to acknowledge Celia Lobo Romero for technical assistance during in vivo experimental procedures.
Name of Reagent | Company | Catalog Number | Comments/Description |
Yeast extract granulated | Merck | MERC1.03753.0500 | |
Bacteriological peptone | Oxoid | LP037B | |
Agar granulated | Difco | 214530 | |
D-(+)-glucose | Fluka | 49159-5KG | |
Phosphate buffered saline | Prepared in the laboratory | for 1L of 10x PBS: 80 g NaCl, 2 g KCl, 14.4 g Na2HPO4, 2.4 g KH2PO4 | |
RPMI1640 with L-glutamine and without sodium carbonate | Sigma | R6504-1L | Prepare according the protocol for Candida albicans drug susceptibility testing |
3-(N-Morpholino)propanesulfonic acid (MOPS) | Sigma | M1254 | MOPS is used to adjust the pH of RPMI medium (pH 7.0) |
fetal bovine serum (FBS) | Sigma | F7524 | |
Polyurethane tripe-lumen intravenous catheter piece (2.4 mm diameter, Certofix Trio S730) | BBraun | CV-15703 | Polyurethane part cut into 1 cm pieces |
Dexamethasone | Fagron SAS, France | 611139 | Immunosuppressant (stock solution 10 mg/ml) |
Ampicillin | Duchefa Biochemie, The Netherlands | A0104 | Antibacterial prophylaxis |
Ketamine 1000 | Pfizer | 804 119 | Anesthetic |
Domitor | Pfizer | 134737-1 | Anesthetic |
Antisedan | Pfizer | 134783-2 | Reversal of anesthesia |
Xylocaine gel (2%) – this is Linisol | AstraZeneca | 352 1206 | Local anesthetic for the skin |
Terramycin/ polymyxin-b ophthalmic ointment | To prevent drying and infection of eyes | ||
Coelenterazine | Prolume (Nanolight) | NF-CTZ-FB | Light sensitive agent (must be kept in the dark) |
Iodine isopropanol (1%) | 3M™ DuraPrep™ | Disinfectant for the skin | |
0.5 % chlorhexidine in 70 % alcohol. | Cedium | Disinfectant for the skin | |
Equipment | |||
Cell counting chamber | |||
Insulin syringes (0.3 ml) | Terumo Myjector 29G | 324826 | For injection of coelenterazine |
Electric razor | For small animals | ||
Sterile surgical tools | Scissors, 2 pairs of tweezers, scalpel | ||
Heating pad | Leica | 14042321474 | |
Skin suture | Johnson&Johnson | K890H | Surgical thread, needle |
Water bath sonicator | Branson 2210 | ||
BLI camera (IVIS Spectrum) | Perkin Elmer, Alameda | IVISSPE | |
Living Image software | Perkin Elmer, Alameda | (version 4.2) |