Saccharomyces cerevisiae biofilm tolerance towards systemic antifungals depends on growth phase

Rasmus Bojsen 0 1 Birgitte Regenberg 2 Anders Folkesson 0 0 National Veterinary Institute, Technical University of Denmark , Frederiksberg , Denmark 1 Department of Systems Biology, Technical University of Denmark , Kgs. Lyngby , Denmark 2 Department of Biology, University of Copenhagen , Copenhagen , Denmark Background: Biofilm-forming Candida species cause infections that can be difficult to eradicate, possibly because of antifungal drug tolerance mechanisms specific to biofilms. In spite of decades of research, the connection between biofilm and drug tolerance is not fully understood. Results: We used Saccharomyces cerevisiae as a model for drug susceptibility of yeast biofilms. Confocal laser scanning microscopy showed that S. cerevisiae and C. glabrata form similarly structured biofilms and that the viable cell numbers were significantly reduced by treatment of mature biofilms with amphotericin B but not voriconazole, flucytosine, or caspofungin. We showed that metabolic activity in yeast biofilm cells decreased with time, as visualized by FUN-1 staining, and mature, 48-hour biofilms contained cells with slow metabolism and limited growth. Time-kill studies showed that in exponentially growing planktonic cells, voriconazole had limited antifungal activity, flucytosine was fungistatic, caspofungin and amphotericin B were fungicidal. In growth-arrested cells, only amphotericin B had antifungal activity. Confocal microscopy and colony count viability assays revealed that the response of growing biofilms to antifungal drugs was similar to the response of exponentially growing planktonic cells. The response in mature biofilm was similar to that of non-growing planktonic cells. These results confirmed the importance of growth phase on drug efficacy. Conclusions: We showed that in vitro susceptibility to antifungal drugs was independent of biofilm or planktonic growth mode. Instead, drug tolerance was a consequence of growth arrest achievable by both planktonic and biofilm populations. Our results suggest that efficient strategies for treatment of yeast biofilm might be developed by targeting of non-dividing cells. - Background Nosocomial fungal infections are a major problem for immune compromised patients with a severe underlying disease [1]. Fungi can cause infections by colonizing mucosal surfaces in the oral cavity, airways, wounds and the gastrointestinal tract [2]. Fungi can also adhere to invasive medical devices and cause severe septicemia upon detachment [3]. The hallmarks of biofilms are surface attachment and production of an extracellular matrix (ECM) [4]. Failure to eradicate microbial infections is often attributed to the unique lifestyle of cells in biofilms and it is widely accepted that cells in a biofilm possess antimicrobial tolerance mechanisms that are distinct from their planktonic counterparts [2]. Drugs currently being used to treat systemic mycoses belong to four major classes. The azoles target cytochrome P450 and inhibit cell membrane ergosterol biosynthesis, resulting in accumulation of toxic ergosterol intermediates [5]. Azoles have poor efficacy against Candida species other than C. albicans, such as C. glabrata [6]. The number of nosocomial blood isolates of these non-susceptible Candida species has increased in the past decades, possibly because of the selection that frequent azole use impose [7]. The echinocandins inhibit 1,3--glucan synthases, resulting in a reduction in cell wall 1,3--glucan [8], and the polyenes target ergosterol and cause pore formation in the fungal cell membrane [9]. The fourth class is the antimetabolite flucytosine. Flucytosine is deaminated upon uptake in susceptible cells and converted to 5-fluorouridine triphosphate, which is incorporated into RNA, inhibiting protein synthesis [10]. Flucytosine can also be converted to 5-fluorodeoxyuridine monophosphate which acts on thymidylate synthase to inhibit DNA synthesis [10]. Despite the pronounced diversity in antifungal mechanism of action and chemical structure, most antifungal agents are inactive against fungal biofilms [11]. Several mechanisms have been suggested to be responsible for drug tolerance of yeast biofilms. One of them is the ECM layer that contains -1,3 glucans and extracellular DNA [12,13]. Treatment of biofilm cells with glucanases or DNase result in increased efficacy of antifungal agents, which indicate a role of ECM on antifungal drug tolerance [13,14]. However, it has been shown that antifungal susceptibility is independent of amount of matrix produced and antifungal drugs can diffuse through the matrix layer in inhibitory concentrations [15,16]. The ECM, in combination with the nutrient-limited environment that results from a large number of microbial cells, might induce expression of genes that help cells cope with stressful conditions. Altered gene expression could involve differential regulation of general stress-response genes that affect drug tolerance. For example, efflux pumps are reported to be upregulated in young and intermediate [17,18] biofilms in Candida species. However, efflux pump knockout mutants remain drug resistant [18,19] and up-regulation is lost in mature biofilms [17,18]. Furthermore, since polyenes and echinocandins are not a substrate of any known efflux pumps [20], efflux pumps are not responsible for biofilm-mediated tolerance to these drug classes. None of the suggested tolerance mechanisms are solely responsible for the multidrug tolerance associated with biofilm, and it might be a combination of several individual mechanisms that cause multidrug tolerance in yeast biofilms. Candida is the most frequent cause of fungal infections and extensive research has been performed with this organism to investigate regulation of biofilm formation and antifungal drug recalcitrance [3]. However, due to a limited repertoire of genetic and molecular techniques available for some Candida species, the knowledge about yeast biofilm regulation and drug tolerance is incomplete. The genetic tractability of another fungus, Saccharomyces cerevisiae, has made it a model organism for the study of fundamental issues in fungal biology [21]. Transition from yeast to filamentous morphology is correlated to virulence in Candida albicans and key signaling pathways controlling this process is conserved in S. cerevisiae [22]. Candida glabrata is phylogenetically more closely related to S. cerevisiae than to other Candida species [23] and they have homologous cell-surface adhesins [24]. C. glabrata and S. cerevisiae furthermore form biofilms as haploids with similar biofilm architecture: thin layer of biofilm cells with yeast morphotype surrounded by a low density of ECM [25,26]. S. cerevisiae is therefore relevant for the study of C. albicans virulence and C. glabrata biofilm. S. cerevisiae has previously been used as a model organism to study yeast biofilm development and regulation by taking advantage of the molecular tools available fo (...truncated)