SEARCH

SEARCH BY CITATION

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The fungal pathogen Candida albicans forms therapeutically challenging biofilms on biomedical implants. Using a transcript profiling approach genes whose expression is favoured upon biofilm growth compared with planktonic growth have been previously identified. Knock-out mutants for 38 of these genes were constructed, six of which showed a specific defect in biofilm formation. Among these genes, TYE7 that encodes a transcriptional activator of glycolytic genes in planktonic and biofilm growth conditions was identified as being required for the cohesiveness of biofilms. Biofilms formed by the tye7Δ knock-out mutant showed a hyperfilamentous morphology, and growth of this mutant on solid medium under hypoxia was also associated with the production of hyphae. Similar to TYE7 inactivation, inhibition of glycolysis or ATP synthesis using oxalate or an uncoupler, respectively, triggered morphogenesis when a wild-type strain was grown under hypoxia. These treatments also induced the formation of weakly cohesive, hyper-filamentous biofilms by a wild-type strain. Our data indicate that a hypoxic environment is generated within C. albicans biofilms and that continued biofilm development requires a Tye7p-dependent upregulation of glycolytic genes necessary to adapt to hypoxia and prevent uncontrolled hyphal formation. Thus, adaptation to hypoxia is an integral component of biofilm formation in C. albicans.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Biofilms are three-dimensional communities of microorganisms that form on artificial or biological surfaces (Donlan and Costerton, 2002). These mono- or poly-species communities are characterized by an extracellular matrix (ECM), in which microorganisms are embedded. Moreover, they display specific properties, including an intrinsic tolerance to a variety of antimicrobials and to host immune defences. Biofilms play an important role in microbial pathogenesis, especially in the hospital setting where the increasing usage of medical devices favours the development of biofilms and the consecutive failure of antimicrobial treatments (Schierholz and Beuth, 2001). In this respect, the formation of biofilms by the fungal pathogen Candida albicans and the resistance of these biofilms to several key antifungal agents such as azoles, amphotericin B and 5-flucytosine are well-recognized phenomena (Ramage et al., 2005). The presence of a central venous catheter is one of the risk factors in the establishment and persistence of nosocomial C. albicans infections (Rex et al., 2000; Filler and Kuhlberg, 2002), and catheter withdrawal is considered a prerequisite to a successful antifungal treatment (Mermel et al., 2001).

The sequence of events that yields mature C. albicans biofilms involves the adherence of yeast cells to an artificial surface, the formation of a microcolony of yeast cells from which a network of hyphae emerges and the production of an ECM (Douglas, 2003). Soluble β-glucans are components of the ECM and have been shown to play a role in antifungal tolerance of the biofilms through their ability to trap azoles and polyenes en route to the biofilm cells (Nett et al., 2007; 2010a,b; Vediyappan et al., 2010). Investigation of the molecular mechanisms that control the different events of biofilm formation has been in particular through the characterization of C. albicans knock-out mutants (Nobile and Mitchell, 2006; ten Cate et al., 2009; Ramage et al., 2009). Early studies of strains defective for the Efg1 and Cph1 regulators of the yeast-to-hypha transition have confirmed the anticipated role of this process in the establishment of C. albicans mature biofilms (Baillie and Douglas, 1999; Ramage et al., 2002). Since this observation, numerous mutants with defects in morphogenesis have been shown to be impaired for biofilm formation (Nobile and Mitchell, 2005; Richard et al., 2005; Firon et al., 2007; Goyard et al., 2008). Yet, investigation by A. Mitchell and collaborators of a collection of C. albicans knock-out mutants defective for transcription factors has revealed that the requirement for biofilm formation of hyphal morphogenesis is not solely related to the shape of hyphal cells that favours the formation of a three-dimensional structure. The Bcr1 transcription factor was shown to be necessary for biofilm formation and the expression of a subset of cell wall-associated GPI-modified proteins, such as members of the agglutinin-like Als protein family and Hwp1 (Nobile and Mitchell, 2005). Subsequent investigation of the role of Als proteins, in particular Als3 and Als1, and Hwp1 has revealed that these adhesins contribute to biofilm formation through heterotypic interactions (Nobile et al., 2006; 2008). Other adhesins such as Eap1 appear to participate at earlier stages of biofilm formation (Li and Palecek, 2008), while Ywp1 might have an opposite role or contribute to dispersal of biofilm cells at late stages of biofilm maturation (Granger et al., 2005). Additionally, a role in biofilm formation for the mitogen-activated protein kinase Mkc1p, a component of the cell integrity pathway activated by cell wall stress, has been shown (Kumamoto, 2005). This role may pertain to the interaction of yeast cells with surfaces where they initiate biofilm formation or interaction between hyphal cells within the biofilm. Recently, the C. albicans zinc-response transcription factor Zap1 was shown to be a negative regulator of both soluble β-glucan and ECM production (Nobile et al., 2009). Zap1 transcriptional targets with a role in ECM production have been characterized, including two glucoamylases, Gca1 and Gca2, and three alcohol dehydrogenases, Adh5, Csh1 and Ifd6 (Nobile et al., 2009). The contribution of alcohol dehydrogenases has been proposed to reflect quorum sensing processes that might in turn control different events of biofilm maturation (Nobile et al., 2009).

Transcript profiling is another approach that was used to shed light into the events occurring upon C. albicans biofilm formation. Different stages of biofilm formation have been investigated, from early phases to late phases along with an early-phase detachment event (Garcia-Sanchez et al., 2004; Murillo et al., 2005; Yeater et al., 2007; Sellam et al., 2009). Importantly, biofilm populations formed in different environments display similar and specific transcript profiles that are largely different from those of planktonic populations (Garcia-Sanchez et al., 2004). In general, biofilm formation was associated with increased expression levels of genes involved in glycolysis and amino acid and lipid metabolism (Garcia-Sanchez et al., 2004; Murillo et al., 2005; Yeater et al., 2007; Sellam et al., 2009). In this respect, the transcriptome of Candida parapsilosis biofilm cells was reminiscent of that of planktonic cells grown under hypoxia that also show an upregulation of glycolytic genes (Rossignol et al., 2009), suggesting that Candida biofilm formation might be associated to the generation of a hypoxic environment as shown for bacterial biofilms (Xu et al., 1998; An and Parsek, 2007). In C. albicans, the TYE7 gene has been shown to encode a positive regulator of glycolytic genes and to be required for hypoxic adaptation (Askew et al., 2009). Yet, the role of TYE7 in biofilm formation is unknown. In contrast, a role for the Efg1 transcriptional regulator in hypoxic adaptation upon C. albicans biofilm growth has been documented (Stichternoth and Ernst, 2009).

Transcript profiling has identified genes that show increased or decreased expression in biofilms relative to planktonic cells and it has been hypothesized that these genes might play some role in biofilm formation. For instance, C. albicans SUN41 was among the genes that showed high upregulation upon biofilm growth (Garcia-Sanchez et al., 2004) and was subsequently shown to be required for biofilm formation, possibly through its contribution to morphogenesis and/or matrix production (Firon et al., 2007; Hiller et al., 2007; Norice et al., 2007). Yet, other examples showed no strict correlation between differential regulation in biofilm versus planktonic growth and contribution to biofilm formation (Moreno-Ruiz et al., 2009; Sellam et al., 2009). Thus, a systematic investigation of genes upregulated upon biofilm growth might be required to assess their contribution to this process.

Here, we report the functional characterization of 38 C. albicans genes that showed upregulation in mature biofilms relative to planktonic cells (Garcia-Sanchez et al., 2004). C. albicans knock-out mutants for these 38 genes have been generated and investigated for their ability to form biofilms and undergo the yeast-to-hypha transition. This analysis allowed the identification of six novel genes whose inactivation impairs biofilm formation independently of a role in morphogenesis. In particular, we showed that inactivation of the TYE7 gene results in the formation of fragile biofilms. We found that TYE7 is necessary for efficient glycolysis during biofilm formation and that inactivation of TYE7 results in enhanced filamentation during biofilm formation and under hypoxia. Strikingly, phenotypes similar to those associated to TYE7 inactivation were observed when glycolysis or ATP biosynthesis were chemically inhibited. These data suggest that maintenance of ATP production, mediated by Tye7p-dependent upregulation of glycolytic genes, is necessary to allow C. albicans cells to adapt to the hypoxic environment that is generated during biofilm maturation and to circumvent the unrestricted formation of hyphae that loosens the biofilm structure.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Identification of genes contributing to biofilm formation

A transcript profiling study by Garcia-Sanchez et al. (2004) has previously identified 317 C. albicans genes that showed elevated expression upon biofilm growth relative to planktonic growth. Forty-three genes were selected for further investigation with the help of knock-out mutants. Genes listed in Table 1 were selected as they (i) showed the highest difference in expression in biofilm and planktonic samples (Garcia-Sanchez et al., 2004); (ii) were unlikely to be essential for C. albicans growth based on the knowledge gained from the study of orthologous genes; and (iii) did not belong to a gene family. In particular, genes that were likely to be involved in protein biogenesis, a functional category over-represented among the 317 genes identified by Garcia-Sanchez et al. (2004), were excluded from this study. Knock-out mutants were constructed by two successive gene replacements in C. albicans strain BWP17U (Moreno-Ruiz et al., 2009) using PCR-generated cassettes with the HIS1 and ARG4 auxotrophic markers (Walther and Wendland, 2008). Gene replacements were confirmed by PCR and Southern hybridization (see Experimental procedures). Homozygous knock-out mutants were obtained for all selected genes except PGA56, ERG6, SSZ1, NHP6A and orf19.4149.

Table 1.  Phenotypic characterization of 38 C. albicans knock-out mutant strains.
StrainGene expression level in the WT: log(B/P)aBiofilm formationHyphal formationGrowth curvec
ClassNameDeleted geneNumber of MFBiomass relative to WT (%)P-valuebLeeRPMIEmbeddedAt 30°CAt 37°C
KineticsSaturationKineticsSaturation
  • a.

    Gene expression level was previously measured by Garcia-Sanchez et al. 2004, in Table S2.

  • b.

    P-value was obtained by Mann–Whitney test.

  • c.

    Cells were grown in flasks at 30°C and 37°C for 24 h in SD + 0.4% glucose at pH 5.4. Growth was measured manually by spectrophotometer at OD600. Kinetics was defined in OD per hour between 2 and 8 h of exponential phase of growth. Saturation corresponds to the stationary phase of growth, and measured at 24 h.

  • B, biofilm; MF, microfermentor; P, planktonic culture; WT, wild-type strain.

  • +, phenotype of the wild-type strain; −, defect in phenotype compared with the wild-type strain.

IICEC298orf19.74591.7520 
IICEC314orf19.60905.26621.70.002+++++
IICEC300orf19.22861.851434.77.152e-06+++++
IICEC246RHR26.02459.60.029+++++++
IICEC203orf19.25271.81462.70.029+++++++
IICEC294NPL32.35865.20.024+++++++
IICEC269orf19.70692.03867.00.007+++++++
IICEC217TYE71.79667.80.013+++++++
IICEC302FAA41.83869.20.003+++++++
ICEC199QDR12.62270.00.333+++++++
ICEC200orf19.19641.99274.50.667+++++++
ICEC265HMX11.81274.70.333+++++
ICEC201MP652.33476.20.486+++++++
ICEC243HGT79.52178.8 +++++++
ICEC247RBT52.94479.30.343+++++++
ICEC267CLG12.09282.20.414+++++++
ICEC299HGT63.16282.40.333+++++++
ICEC270orf19.71071.75283.00.667+++++++
ICEC268CUP92.13486.80.886+++++++
ICEC297orf19.65072.06487.00.343++++
ICEC306orf19.12872.06290.50.667++++
ICEC220FLC11.71491.30.309+++++++
ICEC289MUP12.18493.10.771+++++++
ICEC290PTR23.49394.10.658+++++++
ICEC218MET32.01298.31.000+++++++
ICEC305CAN13.491100 ++++
ICEC214GNP12.3521000.333+++++++
ICEC195FRE104.6121001.000+++++++
ICEC242orf19.16912.821001.000+++++++
ICEC244ALP11.7921000.667+++++++
ICEC216orf19.34831.8121000.667+++++++
ICEC315YWP12.2221000.333+++++
ICEC223HXT52.6421001.000+++++++
ICEC238PTC82.221000.667+++++++
ICEC221orf19.47921.7721000.667+++++++
ICEC312orf19.55252.7721000.333+++++++
ICEC292orf19.73281.721000.333+++++++
ICEC291orf19.74451.8821001.000+++++++

The resulting 38 homozygous knock-out mutants were tested in a microfermentor model of biofilm formation (Garcia-Sanchez et al., 2004), in assays to induce hyphal formation, and in planktonic growth kinetics. Results are summarized in Table 1 and Fig. 1. Two classes of genes were identified (Table 1). Class I included 29 genes whose inactivation had no significant impact on biofilm formation irrespective of a defect in morphogenesis or growth (Fig. 1A, C and D and Table 1). The second class included nine genes whose inactivation resulted in a statistically significant reduction of the biofilm biomass (Fig. 1A and Table 1). Inactivation of orf19.7459, an orthologue of the S. cerevisiae RMD9 gene that encodes a mitochondrial protein required for respiratory growth (Nouet et al., 2007) had a strong impact on both biofilm and planktonic growth, suggesting that the biofilm defect was a consequence of the overall requirement of orf19.7459 for growth (Fig. 1B and D). Inactivation of orf19.6090, an orthologue of S. cerevisiae NSR1 encoding a nucleolar protein required for pre-rRNA processing and ribosome biogenesis (Lee et al., 1992), and orf19.2286, an orthologue of S. cerevisiae LIA1 encoding a deoxyhypusine hydroxylase (Park et al., 2006), did not impair planktonic growth but resulted in a hyphal formation defect in different conditions (Fig. 1B), suggesting that the biofilm deficiency of the two corresponding mutants resulted from an inability to undergo the yeast-to-hypha transition. Inactivation of the six remaining genes – namely RHR2 that encodes a putative glycerol 3-phosphatase with a role in osmo-tolerance (Fan et al., 2005), NPL3 that encodes a RNA-binding protein whose orthologue in S. cerevisiae is required for mRNA maturation (Kress et al., 2008), TYE7 that encodes, like its orthologue in S. cerevisiae, a transcriptional regulator of glycolytic genes (Sato et al., 1999; Benanti et al., 2007; Askew et al., 2009), FAA4 that encodes a putative acyl-CoA-synthase (Lorenz et al., 2004), orf19.2527 and orf19.7069 that encode proteins of unknown function – resulted in a modest but significant reduction of the biofilm biomass (30–40% reduction; Table 1 and Fig. 1A). Yet, inactivation of these genes did not result in hyphal formation or planktonic growth defects (Fig. 1B), suggesting that biofilm deficiency is independent of a role of these genes in yeast-to-hypha transition or planktonic growth.

image

Figure 1. Phenotypic characterization of 38 C. albicans knock-out mutant strains in a microfermentor model of biofilm formation, in assays to induce hyphal formation and in planktonic growth kinetics. A. Dry weights of biofilms formed by each of the 38 knock-out mutant strains compared with the wild-type strain (SC5314) in a microfermentor model. Dark gray colour indicates a statistically significant reduction of the biofilm biomass between mutant and wild-type strains, defining class II genes (P-value < 0.05 by Mann–Whitney test). B. Representative images of Class II mutant strains on RPMI and embedded solid media and in Lee's liquid medium, compared with the wild-type strain (SC5314). C. Representative images of three Class I mutant strains displaying a defect in hyphal formation, on RPMI and embedded solid media and in Lee's liquid medium, compared with the wild-type strain (SC5314). D. Representative growth curves of mutant strains in planktonic culture, at 30°C, compared with the wild-type strain (SC5314). Strains which are not shown present a growth curve similar to that of the wild-type strain (mean doubling times < 2.5 h compared with 2 h for the wild-type). *The orf19.6507, CAN1 and YWP1 knock-out mutants have similar growth kinetics (mean doubling time of 3 h) and are represented by only one growth curve.

Download figure to PowerPoint

The Tye7p transcriptional regulator of glycolytic genes negatively regulates hyphal formation under hypoxia

It has been proposed that Candida cells upregulate glycolytic genes during biofilm formation because of the probable generation of a hypoxic environment (Rossignol et al., 2009). Interestingly, we noted that inactivation of the C. albicans TYE7 gene encoding a basic-helix-loop-helix (bHLH) transcription factor shown to positively regulate glycolytic genes, in particular when cells are grown planktonically under hypoxia (Askew et al., 2009), resulted in a defect in biofilm formation (Fig. 1A and Table 1). Inactivation of TYE7 also appeared to decrease biofilm adherence as biofilms were particularly easy to detach from the plastic surface (data not shown). These defects could reflect the impaired ability of this mutant to cope with the biofilm hypoxic environment because of impaired activation of glycolysis. Therefore, the basis for the biofilm formation defect associated to the inactivation of TYE7 was further investigated.

Three C. albicans strains were used to investigate the role of TYE7 in biofilm formation: a wild-type strain (TYE7/TYE7, BWP17AHU; Table 2), a mutant strain where both TYE7 alleles have been inactivated by homologous replacement (Δtye7, CEC1207; Table 2), and a complemented strain (with re-introduction of a wild-type copy of TYE7 at the RPS10 locus; Δtye7 + TYE7, CEC1222, Table 2). Consistent with the observations of Askew et al. (2009), the Δtye7 mutant showed growth kinetics similar to that of the Δtye7 + TYE7 and TYE7/TYE7 strains when grown under strong aeration (SD 0.4% glucose at 37°C in Erlenmeyer, Fig. S1A), but had increased doubling times relative to the TYE7/TYE7 and Δtye7 + TYE7 strains when grown (i) under low aeration (SD 0.4% glucose at 37°C in microtitre plates, Fig. S1B), (ii) under strong aeration in the presence of antimycin A that impairs the respiratory chain (Fig. S1A) and (iii) in an anaerobic atmosphere on YPD solid medium at 37°C (Fig. 2A). Noticeably, a fringe of hyphal cells was clearly visible around colonies of the Δtye7 strain grown under hypoxia while this crown of hyphae was less dense for the Δtye7 + TYE7 strain, and absent for the TYE7/TYE7 strain (Fig. 2B). Hyphal formation at the edge of colonies was also observed at 30°C under hypoxia, although to a lesser extent, but was not observed when these strains were grown under normoxia at 37°C or 30°C (data not shown). Taken together, these data confirmed the role of TYE7 in sustaining C. albicans growth under hypoxia (Askew et al., 2009), but additionally showed that Tye7p negatively regulates the yeast-to-hypha transition induced by hypoxic conditions (Brown et al., 1999; Setiadi et al., 2006; Ernst and Tielker, 2009). That our wild-type strain (BWP17AHU, a derivative of strain BWP17 where the URA3, HIS1 and ARG4 genes have been reintegrated in single copy) did not undergo hypoxia-induced filamentation as observed by others (Land et al., 1975; Mulhern et al., 2006) might reflect differences in experimental conditions and genotypes.

Table 2.  Strains used in this study.
StrainDeleted orfDeleted geneGenotypePhenotypeSource
  1. +, prototroph; −, auxotroph; A, arginine; H, histidine; NR, nourseothricine resistant; U, uracile.

SC5314Wild-type strainA+H+U+Gillum et al. (1984)
BWP17SC5314 ura3Δ::λimm434/ura3Δ::λimm434 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA−H−U−Wilson et al. (1999)
BWP17Uura3Δ::λimm434/ura3Δ::λimm434 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisG RPS10/rps10::CIp10-URA3A−H−U+Moreno-Ruiz et al. (2009)
BWP17AHUura3Δ::λimm434/ura3Δ::λimm434 arg4Δ::hisG/ARG4 his1Δ::hisG/HIS1 RPS10/rps10::CIp10-URA3A+H+U+Firon et al. (2007)
CEC29819.7459orf19.7459orf19.7459Δ::HIS1/orf19.7459Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24519.3642SUN41sun41Δ::HIS1/sun41Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC31419.6090orf19.6090orf19.6090Δ::HIS1/orf19.6090Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC30019.2286orf19.2286orf19.2286Δ::HIS1/orf19.2286Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24619.5437RHR2rhr2Δ::HIS1/rhr2Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC20319.2527orf19.2527orf19.2527Δ::HIS1/orf19.2527Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29419.7238NPL3npl3Δ::HIS1/npl3Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC26919.7069orf19.7069orf19.7069Δ::HIS1/orf19.7069Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC21719.4941TYE7tye7Δ::HIS1/tye7Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC30219.7592FAA4faa4Δ::HIS1/faa4Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC19919.508QDR1qdr1Δ::HIS1/qdr1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC20019.1964orf19.1964orf19.1964Δ::HIS1/orf19.1964Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC26519.6073HMX1hmx1Δ::HIS1/hmx1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC20119.1779MP65mp65Δ::HIS1/mp65Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24319.2023HGT7hgt7Δ::HIS1/hgt7Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24719.5636RBT5rbt5Δ::HIS1/rbt5Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC26719.6146CLG1clg1Δ::HIS1/clg1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29919.2020HGT6hgt6Δ::HIS1/hgt6Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC27019.7107orf19.7107orf19.7107Δ::HIS1/orf19.7107Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC26819.6514CUP9cup9Δ::HIS1/cup9Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29719.6507orf19.6507orf19.6507Δ::HIS1/orf19.6507Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC30619.1287orf19.1287orf19.1287Δ::ARG4/orf19.1287Δ::HIS1 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC22019.2501FLC1flc1Δ::HIS1/flc1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC28919.5280MUP1mup1Δ::HIS1/mup1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29019.6937PTR2ptr2Δ::HIS1/ptr2Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC21819.5025MET3met3Δ::HIS1/met3Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC30519.97CAN1can1Δ::HIS1/can1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC21419.1193GNP1gnp1Δ::HIS1/gnp1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC19519.1415FRE10fre10Δ::HIS1/fre10Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24219.1691orf19.1691orf19.1691Δ::HIS1/orf19.1691Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC24419.2337ALP1alp1Δ::HIS1/alp1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC21619.3483orf19.3483orf19.3483Δ::HIS1/orf19.3483Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC31519.3618YWP1ywp1Δ::HIS1/ywp1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC22319.4384HXT5hxt5Δ::HIS1/hxt5Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC23819.4698PTC8ptc8Δ::HIS1/ptc8Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC22119.4792orf19.4792orf19.4792Δ::HIS1/orf19.4792Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC21919.5032SIM1sim1Δ::HIS1/sim1Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC31219.5525orf19.5525orf19.5525Δ::HIS1/orf19.5525Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29219.7328orf19.7328orf19.7328Δ::HIS1/orf19.7328Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC29119.7445orf19.7445orf19.7445Δ::HIS1/orf19.7445Δ::ARG4 ura3Δ::λimm434/URA3 arg4Δ::hisG/arg4Δ::hisG his1Δ::hisG/his1Δ::hisGA+H+U+This study
CEC120719.4941TYE7BWP17 tye7Δ::HIS1/tye7Δ::ARG4 RPS10/rps10::CIp10A+H+U+This study
CEC1222  BWP17 tye7Δ::HIS1/tye7Δ::ARG4 RPS10/rps10::CIp10-TYE7A+H+U+This study
CEC536ura3Δ::λimm434/ura3Δ::λimm434 arg4Δ::hisG/ARG4 his1Δ::hisG/HIS1 RPS10/rps10::CIp10-SAT1-pTEF1-YFPA+H+U+ NRThis study
CEC141219.4941TYE7BWP17 tye7Δ::HIS1/tye7Δ::ARG4 RPS10/rps10::CIp10-SAT1-pTEF1-YFPA+H+U+ NRThis study
CEC1416  BWP17 tye7Δ::HIS1/tye7Δ::ARG4 rps10::CIp10-TYE7/rps10::CIp10-SAT1-pTEF1-YFPA+H+U+ NRThis study
Δgal419.5338GAL4BWP17 gal4::Tn7-UAU1/gal4::Tn7-URA3A+H−U+Davis et al. (2002)
image

Figure 2. Tye7p is required for efficient growth and negative regulation of the yeast-to-hypha transition under hypoxia. A. Representative images of the wild-type (TYE7/TYE7), knock-out (Δtye7) and complemented (Δtye7 + TYE7) strains after 4 days of growth in an anaerobic atmosphere on YPD solid medium at 37°C by spotting of serial dilutions (from right to left). B. Representative images showing the colony edge of the wild-type (TYE7/TYE7, left), knock-out (Δtye7, centre) and complemented (Δtye7 + TYE7, right) strains, in culture conditions cited in (A), using an inverted microscope.

Download figure to PowerPoint

The TYE7 gene is required for efficient biofilm formation by C. albicans

The Δtye7, Δtye7 + TYE7 and TYE7/TYE7 strains were investigated for their ability to form biofilms. Data presented in Fig. 3A show biomass measurements obtained from five independent series of cultures in microfermentors, with a total of at least 10 replicates for each of the three strains. These data indicated that inactivation of TYE7 was indeed responsible for the reduction in biofilm biomass observed previously as the Δtye7 strain showed a 54% reduction in the biofilm biomass when compared with the TYE7/TYE7 strain (the previous mutant from the screen showed a 1/3 reduction compared with wild-type) and reintroduction of a wild-type copy of TYE7 restored biofilm formation, although not to wild-type levels, possibly reflecting an haploinsufficiency phenotype. Representative images of biofilms formed in the microfermentor model at 40 h of development are shown in Fig. 3B. From 32 h, biofilms of the Δtye7 strain appeared thinner than those formed by the TYE7/TYE7 and Δtye7 + TYE7 strains and this difference increased at 40 h (data not shown).

image

Figure 3. Impact of TYE7 inactivation on biofilm production. A. Dry weights of biofilms formed by the wild-type (TYE7/TYE7), deleted (Δtye7) and complemented (Δtye7 + TYE7) strains, after 40 h of development in a microfermentor model. Weight differences are statistically significant (P-value < 0.05 by Mann–Whitney test, comparing strain to strain). B. Representative images of biofilms formed by the Δtye7 and the wild-type (TYE7/TYE7) strains after 40 h of development in a microfermentor model. C. Representative images of biofilms formed by the deleted (Δtye7) and wild-type (TYE7/TYE7) strains at 60 h of growth in a silicone square model. D. CSLM top views of biofilms formed by the wild-type (TYE7/TYE7, left), complemented (Δtye7 + TYE7, centre) and mutant (Δtye7, right) strains after 60 h of development in a silicone square model, and staining with concanavalin A conjugate. The scale bar represents 20 µm. The analysed thickness of each biofilm in the observed region was 120 µm.

Download figure to PowerPoint

In order to further document the biofilm deficiency of the Δtye7 mutant, a second biofilm model was used that allows the production of biofilms on serum-coated silicone squares incubated in SD 0.4% glucose medium under mild agitation (Richard et al., 2005). Results presented in Fig. 3C indicated that under these conditions, the Δtye7 strain formed fragile biofilms that detached from the silicone surface in contrast to the biofilms formed by the TYE7/TYE7 strain. Biofilms formed on silicone by derivatives of these three strains expressing YFP constitutively were observed by confocal scanning laser microscopy (CSLM). CSLM did not allow an accurate measurement of these biofilms thickness because of important variations throughout the surface of the biofilms. However, representative CSLM images shown in Fig. 3D indicated that the biofilm formed by the Δtye7 strain contained more hyphae than those formed by the two other strains. Measurements of the ratio of hyphal cells in the biofilms formed on silicone by the Δtye7 and TYE7/TYE7 strains confirmed these observations. If one sets aside the packets of uncountable hyphae that were more numerous in the deleted strain, ratios were 0.71 ± 0.17 and 0.11 ± 0.03 for the Δtye7 and TYE7/TYE7 strains, respectively, indicating a higher proportion of hyphal cells in the mutant biofilm (also see Fig. 5A). Inactivation of TYE7 did not result in enhanced filamentation in liquid Lee's medium, RPMI or in embedded conditions as shown in Fig. 1B. Thus, the TYE7 gene appeared to be required at a relatively late stage of biofilm formation, preventing uncontrolled filamentation and contributing to the eventual robustness of these structures and their persistence on abiotic surfaces. That hyperfilamentation of the Δtye7 had also been observed under hypoxia (see above) suggested that the hyperfilamentous phenotype of the Δtye7 biofilms is a consequence of the generation of a hypoxic environment at late stages of biofilm formation.

image

Figure 5. Impairment of glycolysis and mitochondrial ATP synthesis triggers C. albicans morphogenesis. A. Top, representative images of biofilms formed by the wild-type (TYE7/TYE7) and knock-out (Δtye7) strains after 60 h of development in a silicone square model. Bottom, representative images of biofilm cells obtained after vortexing. B. As in (A) but with the addition of 10 mM oxalate. C. As in (A) but with the addition of 0.05 mM DNP. D. Representative images of the wild-type (TYE7/TYE7), homozygous mutant (Δtye7) and complemented (Δtye7 + TYE7) strains after 4 days of growth in an anaerobic atmosphere on YPD solid medium complemented with oxalate (10 mM) at 37°C by spotting of serial dilutions (from right to left). E. Representative images showing the colony edge of the wild-type (TYE7/TYE7, left), homozygous mutant (Δtye7, centre) and complemented (Δtye7 + TYE7, right) strains, in culture conditions cited in (D). F. As is (D) but with 0.05 mM DNP. G. As is (E) but with 0.05 mM DNP.

Download figure to PowerPoint

Tye7p is required for the efficient expression of glycolytic genes in biofilm and planktonic cultures

The TYE7 gene encodes a conserved bHLH transcription factor that regulates the expression of glycolytic genes in S. cerevisiae and C. albicans (Nishi et al., 1995; Sato et al., 1999; Benanti et al., 2007; Askew et al., 2009). As the identification of transcriptional targets of C. albicans Tye7p has been conducted under planktonic growth conditions (Askew et al., 2009), we extended these studies to biofilm growth conditions. For this purpose, we compared the transcriptome of the Δtye7 and Δtye7 + TYE7 strains growing in SD 0.4% glucose medium as biofilms at 24, 32 and 40 h of development in the microfermentor model, and under planktonic conditions in the exponential phase of growth. Data presented in Fig. S2 and Tables S1 and S2 showed that Tye7p functionality was associated with higher expression (≥ 1.5-fold; P < 0.05) of 67 genes at 24 h, 139 genes at 32 h, 219 genes at 40 h of biofilm growth, and 134 genes in planktonic cultures. There were 22 genes whose upregulation was common to the three time points of biofilm growth. Sixty-one genes were upregulated in at least one of the biofilm samples and in planktonic cultures. The grouping of these 61 genes according to gene ontology categories (Arnaud et al., 2007) revealed that a number of them, particularly those upregulated in the three biofilm and the planktonic datasets, were engaged in all stages of glycolysis (Fig. 4A and Table S2). Glycolytic genes had increased expression in the complemented strain at 24 and 32 h of biofilm growth, and the difference between the complemented and mutant strain decreased at 40 h when the biofilm was mature (Fig. 4A, Table S1). These results confirmed the data of Askew et al. (2009) and extended the role of Tye7p in activating the expression of glycolytic genes in C. albicans to biofilm growth conditions.

image

Figure 4. TYE7 inactivation results in changes in the expression of glycolytic genes at all stages of biofilm development and of a large number of genes in mature biofilms. A. Regulation of genes upregulated by Tye7p involved in the metabolic pathway of glycolysis. Each graph represents expression ratio of a gene encoding a glycolytic enzyme in the revertant (Δtye7 + TYE7) versus knock-out (Δtye7) strains at the three time points of biofilm growth (24, 32 and 40 h) and in planktonic culture in exponential phase of growth (P-values not taken into account). B. Comparison of gene expression profiles at the three time points of biofilm growth (24, 32 and 40 h) in the revertant (Δtye7 + TYE7) versus knock-out (Δtye7) strain. One-way anova with P-value ≤ 0.05 using Genespring GX software (Agilent Technologies) was performed and allowed identification of 669 regulated genes. Hierarchical clustering on this list of genes was used to create a heat map of expression profiles.

Download figure to PowerPoint

As shown in Fig. 4B, while there were minor differences in the transcriptomes of 24 and 32 h biofilms formed by the Δtye7 and Δtye7 + TYE7 strains, there were major differences in the transcriptomes of 40 h biofilms formed by these strains. Indeed, Tye7p functionality was associated with higher or lower expression (≥ 1.5-fold or ≤ 0.67-fold; P < 0,05) of 83 genes at 24 h, 206 genes at 32 h and 570 genes at 40 h of biofilm growth. This was consistent with our prior observation that differences in phenotypes between the Δtye7 and Δtye7 + TYE7 strains were predominantly observed after 40 h of biofilm growth. Even though TYE7 inactivation resulted in upregulation of ALS1 in all biofilm samples and in upregulation of the ALS3, HWP1 ad RBT1 genes in the 40 h-biofilm (Table S1), no significant enrichment for specific functional categories, especially those related to hyphal development, was observed despite enhanced hyphal formation in mutant biofilms (Table S3). Askew et al. (2009) have not observed Tye7p binding at the promoter region of hypha-specific genes suggesting that it is not a direct regulator of these genes. The absence of enrichment for hypha-specific genes among the Tye7p-downregulated genes was consistent with this observation and, taken together, suggested that hypha-specific gene expression triggered by the hypoxic environment generated within the biofilm is an indirect consequence of Tye7 inactivation.

Hypoxia-induced hyperfilamentation is a consequence of the glycolysis defect associated to TYE7 inactivation

Data presented above showed that inactivation of TYE7 resulted in uncontrolled hyphal formation in conditions of strict hypoxia or in a presumably low oxygen environment such as biofilms. In order to test whether this phenotype was a direct consequence of TYE7 inactivation or, more generally, of the reduction of the glycolytic flux, we used oxalate, a well-known competitive inhibitor of eukaryotic pyruvate kinase, the enzyme that catalyses the ultimate step in glycolysis, yielding ATP and pyruvate (Nowak, 1978; Hynne et al., 2001; Dombrauckas et al., 2005). As the inhibitory activity of oxalate on yeast pyruvate kinases was never reported, we first confirmed that oxalate inhibited C. albicans pyruvate kinase activity in whole-cell extracts of the TYE7/TYE7 strain (data not shown). Second, we tested the effect of 10 mM oxalate on the TYE7/TYE7 and Δtye7 strains when grown in planktonic cultures, in the silicone square model of biofilm formation, and in an anaerobic atmosphere on YPD solid medium. Results are summarized in Fig. S1C and in Fig. 5B, D and E. When grown in SD 0.4% glucose at 37°C in Erlenmeyer flasks, 10 mM oxalate addition did not change the growth kinetics of the two strains (Fig. S1C). In contrast, addition of 10 mM oxalate during biofilm growth on silicone squares resulted in the formation of a thinner and less adherent biofilm by the TYE7/TYE7 strain, similar to that formed by the Δtye7 mutant in the absence of oxalate, and in a strong reduction in biofilm formation by the Δtye7 strain (Fig. 5B). When biofilm cells were detached from the silicone square by vortexing and observed by light microscopy, the ratio of hyphal cells in the biofilm of the TYE7/TYE7 strain was increased in the presence of oxalate (0.94 ± 0.35 versus 0.11 ± 0.03 in the absence of oxalate; Fig. 5A and B) and was comparable to that obtained for biofilms of the Δtye7 strain grown without oxalate (0.71 ± 0.17; Fig. 5A). The same experiment, performed using the microfermentor model, showed a 60% decrease in the biomass of biofilms formed by the wild-type strain when 10 mM oxalate was added (P-value < 0.05 by Mann–Whitney test). Finally, when the mutant strains were grown in an anaerobic atmosphere on YPD solid medium supplemented with oxalate, the Δtye7 strain still showed a reduction in growth compared with the TYE7/TYE7 and Δtye7 + TYE7 strains (Fig. 5D), while growth of these three strains was similar on the same medium in normoxia (data not shown). In this condition, a fringe of hyphal cells was visible around colonies of the three strains (Fig. 5E), whereas no hyphal cells could be seen for these strains grown in normoxia in the same medium containing oxalate (data not shown). Similar results were obtained at 30°C, with a less dense filamentation for the three strains (data not shown). Therefore, growth of the wild-type strain in the presence of oxalate, an inhibitor of glycolysis, recapitulated all of the phenotypes observed for the Δtye7 strain grown in the absence of oxalate, i.e. the formation of loose and hyper-filamentous biofilms and hypoxia-induced hyper-filamentation. This suggested that these phenotypes are not a consequence of a direct role of Tye7p in regulating hyphal morphogenesis but rather a consequence of a reduced glycolytic flux due to TYE7 inactivation or oxalate treatment. Importantly, filamentation was only observed in hypoxic environments and upon glycolysis inhibition, suggesting that it is a consequence of a drastic reduction in ATP synthesis.

Hypoxia-induced hyperfilamentation of the Δtye7 mutant is a consequence of a drastic reduction in ATP synthesis

In order to evaluate whether reduction in cellular energy production is responsible for the observed phenotypes, we used the uncoupling agent, 2,4-dinitrophenol (DNP), which inhibits ATP synthesis in the mitochondrial respiratory chain without disturbing electron transport (Korshunov et al., 1997). The TYE7/TYE7 and Δtye7 strains were tested in the presence or absence of 0.05 mM DNP in planktonic growth kinetics, in the silicone square model of biofilm formation, and in an anaerobic atmosphere on YPD solid medium. Results are summarized in Fig. S1D and in Fig. 5C, F and G. DNP addition at concentrations below 0.05 mM did not alter the growth rate of the two strains (Fig. S1D). In contrast, addition of 0.05 mM DNP resulted in the formation of thinner and less adherent biofilms by the TYE7/TYE7 strain (Fig. 5A and C). Yet, this phenotype was less pronounced than when oxalate was used (Fig. 5B). In particular, the ratio of hyphal cells in DNP-treated biofilms of the wild-type strain was 0.34 ± 0.14 (Fig. 5A and C, bottom). Finally, when cells were grown in an anaerobic atmosphere and in the presence of 0.05 mM DNP, results similar to those obtained in the presence of oxalate were obtained, i.e. the presence of a fringe of hyphal cells was seen around colonies of the Δtye7, Δtye7 + TYE7 and TYE7/TYE7 strains (Fig. 5G), whereas no hyphal cells could be seen for these strains grown in normoxia with DNP (data not shown). Thus, similar to oxalate but to a lesser extent, DNP addition recapitulated all of the phenotypes due to inactivation of TYE7. Consequently, formation of hyper-filamentous biofilms and hypoxia-induced hyperfilamentation is a likely consequence of genetically or biochemically reducing ATP production at the level of glycolysis and/or the respiratory chain.

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

The aim of this study was to evaluate the contribution to C. albicans biofilm formation of genes previously shown to have elevated expression in biofilms relative to planktonic cultures (Garcia-Sanchez et al., 2004). Thirty-eight genes have been studied through the construction of corresponding knock-out mutants and their evaluation in a model of biofilm formation under continuous flow. These mutants were also evaluated for their growth rate in planktonic cultures and ability to undergo the yeast-to-hyphal switch that is necessary for efficient biofilm formation (Baillie and Douglas, 1999; Ramage et al., 2002). If one excludes a knock-out mutant that showed a reduced planktonic growth rate and hence was unable to form biofilms, eight knock-out mutants, i.e. 22% of the tested mutants, were significantly impaired in their ability to form biofilms. Of these eight mutants, two had defects in morphogenesis under different hypha-inducing conditions while six did not show any impairment in morphogenesis. Previous studies aimed at identifying genes involved in biofilm formation have focused on a random set of genes (Richard et al., 2005) and on genes encoding transcription factors (Nobile and Mitchell, 2005). Richard et al. (2005) identified four biofilm-defective mutants among 197 and all four mutants had defects in morphogenesis. Nobile and Mitchell (2005) identified two biofilm-defective mutants among 83 and only one did not show any impairment in hyphal morphogenesis. Therefore, our approach to select genes on the basis of an elevated expression in C. albicans biofilms showed an increased likelihood to identify genes that have an actual role in biofilm formation. Yet, this result might also be a consequence of the biofilm models and analytical methods that were used in the three studies. While we used a continuous flow biofilm model and quantitative evaluation of biofilm formation through dry mass measurement, the studies of Richard et al. (2005) and Nobile and Mitchell (2005) used a static biofilm model and visual examination of the biofilms. This latter approach might prevent identification of mutants that have a moderate biofilm defect although work by Nobile et al. (2009) has now shown that qualitative inspection of biofilms developed on silicone could identify biofilms with relatively subtle alterations. While our study shows that there is interest in using the quantitative microfermentor model, it should be noted that it is hardly applicable to large collections of mutants as those evaluated by Richard et al. (2005) and Nobile and Mitchell (2005).

As mentioned above, two of the genes that we have studied were required for efficient hyphal morphogenesis. orf19.6090, a orthologue of S. cerevisiae NSR1 encoding a nucleolar protein required for pre-rRNA processing and ribosome biogenesis (Lee et al., 1992), was required for morphogenesis upon embedded growth and in liquid Lee's medium while orf19.2286, a orthologue of S. cerevisiae LIA1 encoding a deoxyhypusine hydroxylase (Park et al., 2006), was required for morphogenesis upon embedded growth and on RPMI solid medium. Interestingly, three other mutant strains were defective for morphogenesis in liquid Lee's medium only and did not show a significant defect in biofilm formation. Thus, it is likely that the biofilm defect of the orf19.6090 and orf19.2286 knock-out mutants is related to their inability to undergo morphogenesis when growing in contact with a surface and/or upon microaerophilic conditions. In this regard, alteration of contact sensing has been shown to prevent biofilm formation and invasive hyphal growth (Kumamoto, 2005). Yet, it remains unclear how the Orf19.6090 and Orf19.2286 proteins specifically contribute to contact-induced or invasive hyphal growth. Noticeably, other genes involved in RNA metabolism have been shown to play a role in hyphal morphogenesis and biofilm formation (Richard et al., 2005).

Additionally, our study showed a requirement for biofilm formation of the FAA4, NPL3, RHR2, TYE7, orf19.2527 and orf19.7069 genes, the inactivation of which was not associated to a defect in the yeast-to-hypha transition or an impairment of hyphal extension. Importantly, inactivation of these genes was not only associated to a modest (30–40%) and yet significant reduction of the biofilm biomass but in some instances with the formation of weakly cohesive biofilms. In this regard, growing the corresponding mutants in the absence of methionine, an enhancer of biofilm growth and cohesiveness in our microfermentor model, resulted in an exacerbation of their biofilm formation defect (data not shown). Thus, the FAA4, NPL3, RHR2, TYE7, orf19.2527 and orf19.7069 genes appear to contribute to biofilm formation, possibly through a role in biofilm cohesiveness. The C. albicans Faa4, Npl3, Rhr2 and Tye7 proteins or their orthologues in S. cerevisiae have demonstrated roles in distinct cellular processes, respectively, fatty acid metabolism, mRNA metabolism, glycerol metabolism and osmo-tolerance, and glycolysis (Sato et al., 1999; Lorenz et al., 2004; Fan et al., 2005; Benanti et al., 2007; Kress et al., 2008; Askew et al., 2009); and it is therefore unclear how they contribute to biofilm formation and cohesiveness. Yet, data presented in this work and discussed below suggest that Tye7p contributes to biofilm formation through its positive regulation of glycolytic genes that is necessary to adapt to the hypoxic environment generated within biofilms and to limit enhanced filamentation that is triggered by hypoxia.

In a recent report, Askew et al. (2009) have shown that the TYE7 gene is a key regulator of glycolysis in C. albicans. Using a combination of transcript profiling and chromatin immunoprecipitation coupled to microarray analysis, these authors could show that Tye7p bound the promoter of glycolytic genes and activated their expression. Tye7p was in particular shown to be required for the upregulation of glycolytic genes under hypoxic conditions and consequently inactivation of TYE7 was associated with a hypoxia-dependent growth defect that was exacerbated when GAL4, a second positive regulator of glycolytic genes, was inactivated (Askew et al., 2009). Upregulation of glycolytic genes in response to hypoxia allows for an increased glycolytic flux that helps circumvent the defects of the respiratory chain due to oxygen limitation. Our results are in agreement with those of Askew et al. (2009) and show that the contribution of Tye7p to the regulation of glycolytic genes is not limited to planktonic growth but also extends to biofilm growth. Interestingly however, our study of the role of Tye7p during biofilm growth revealed that an additional function of Tye7p under this specific growth condition and under hypoxia was to repress the formation of hyphae as indicated by confocal microscopy of biofilms of the tye7Δ mutant, measurements of the hyphal cell ratios in these biofilms and by its growth in anaerobic atmosphere on solid medium. It is noteworthy that inactivation of GAL4 does not lead to the formation of less cohesive biofilms with reduced biomass in the microfermentor model, confirming that this regulator has a marginal role relative to that of Tye7p (data not shown).

Several studies have indicated that biofilm formation might be associated with the generation of a hypoxic environment (Xu et al., 1998; An and Parsek, 2007; Rossignol et al., 2009). In particular, the upregulation of glycolytic genes that is a landmark of Candida biofilms even when they are grown in normoxia has been regarded as evidence of this local hypoxic environment since it is also a landmark of cells grown under hypoxia (Garcia-Sanchez et al., 2004; Murillo et al., 2005; Yeater et al., 2007; Ernst and Tielker, 2009; Rossignol et al., 2009; Sellam et al., 2009). In C. albicans, hypoxia has also been shown to trigger the yeast-to-hypha transition and expression of hypha-specific genes through a signalling pathway that is independent of the classical Efg1-dependent hypha-inducing pathway operating under normoxia (Brown et al., 1999; Sonneborn et al., 1999; Setiadi et al., 2006; Ernst and Tielker, 2009). Therefore, it is most likely that the biofilm alterations associated to TYE7 inactivation, i.e. a reduced biomass and a predominantly hyphal content, are a direct consequence of a reduced glycolytic flux, that does not allow biofilm cells to sustain the hypoxic environment within biofilms and that, by mimicking an exacerbated hypoxic environment, stimulates hyphal growth. Interestingly, the Efg1/Flo8 regulators have also been proposed to contribute to adaptation to the hypoxic environment generated within biofilms even when these are formed under normoxia (Stichternoth and Ernst, 2009). In hypoxia Efg1 appears to repress the expression of hypha-specific genes and to contribute to the upregulation of different classes of genes, including glycolytic genes, that might favour adaptation to hypoxia (Setiadi et al., 2006; Stichternoth and Ernst, 2009). In addition, the regulator Upc2 has been shown to play an important role in the adaptation response of C. albicans to hypoxia by modulating the ergosterol pathway and hyphal development (Dunkel et al., 2008; Hoot et al., 2008; Synnott et al., 2010). Thus, several regulatory pathways seem to operate in concert to limit the deleterious effects of hypoxic environments, such as that established when biofilms mature, and to prevent unrestricted hyphal growth.

Our transcript profiling study indicated that Tye7p was unlikely to directly repress genes involved in morphogenesis, such as hypha-specific genes or those involved in cell wall biogenesis. Indeed, an alteration of their regulation could have explained the tye7Δ mutant phenotype as inactivation of several of these genes has already been shown to result in a defect in biofilm development (Ramage et al., 2002; Richard et al., 2005; Norice et al., 2007, Bernardo and Lee, 2010; Palanisamy et al., 2010). In contrast, we have shown that blockade of glycolysis with a competitive inhibitor of pyruvate kinase mimicked the phenotypes associated to TYE7 inactivation, confirming that hypoxia-induced hyperfilamentation of the Δtye7 mutant was a consequence of genetic downregulation of glycolysis and exacerbation of hypoxia-mediated cellular alterations. The observed synergy between oxalate treatment and inactivation of TYE7 in hypoxia is consistent with the role of Tye7p in increasing the level of expression of glycolytic genes and the maintenance of an oxalate-sensitive basal glycolytic flux in the Δtye7 mutant. Similarly, blockade of mitochondrial ATP synthesis under hypoxic or biofilm conditions led to a hyperfilamentous phenotype, similar to that observed in the Δtye7 mutant. As expected, inhibiting mitochondrial ATP synthesis had a lower impact on biofilm formation and hypoxia-induced filamentation than inhibiting glycolysis since hypoxia already limits energy production at the level of the mitochondrial respiratory chain and glycolysis has an essential compensatory function upon hypoxia. Therefore, unrestricted filamentation is directly linked to a reduced ATP supply. In other organisms, energy reduction has been shown to accelerate morphogenesis, filamentous growth being regarded as a stress phenotype in response to nutrient limitation and starvation (Blackstone and Buss, 1992; Emri et al., 2004). Dirmeier et al. (2002) have shown that S. cerevisiae cells exposed to anoxia experience transient oxidative stress and exposure of C. albicans to reactive oxygen species or genotoxic stress has been shown to trigger hyphal morphogenesis (Phillips et al., 2003; Shi et al., 2007; Nasution et al., 2008; da Silva Dantas et al., 2010). In yeasts and in particular C. albicans, several studies have shown a relationship between metabolism and morphogenesis. In particular, Chattaway et al. (1973) found that enzyme activities related to carbohydrate metabolism differed between yeast and hyphal forms. It was proven that an intact respiratory pathway was preferentially associated with yeast cells, and that the yeast-to-hyphal switch corresponded to a transition from aerobic metabolism to fermentation, supporting the hypothesis that respiration predominates in yeast cells and fermentation in hyphal cells (Nickerson, 1963; Land et al., 1975; Mulhern et al., 2006; Vellucci et al., 2007). Yet, a direct link between energy depletion and hyphal formation has never been demonstrated. Nevertheless, it is reasonable to assume that energy depletion is sensed as a stress by C. albicans and that filamentation provides an escape route from this unfavourable environment. The cellular mechanisms for this adaptation remain to be elucidated.

Finally, an interesting phenotype of biofilms formed by the C. albicans Δtye7 mutant was their relative lack of cohesiveness and propensity to detach from the substratum. This propensity for flakiness might result from a variety of phenomena. First, unrestricted filamentation in the biofilm might contribute to weaken these structures. Baillie and Douglas (1999) have shown that a mutant strain of C. albicans incapable of yeast growth was nevertheless able to produce a hyphal biofilm, which was much more easily detached from the surface. Elongated hyphae, despite their elevated adhesiveness (Kimura and Pearsall, 1980), appeared to interfere with the initial stages of attachment during the process of biofilm formation. Second, the inability of the Δtye7 mutant cells to withstand the hypoxic environment occurring within biofilms might prompt an early detachment phase. Third, impairment of the glycolytic flux and/or downregulation of ADH5 that encodes an alcohol dehydrogenase positively regulating matrix formation (Nobile et al., 2009) might contribute to limit the production of soluble matrix beta-glucans and consequently reduce the cohesiveness of the biofilm structure. Despite repeated attempts we could not obtain reproducible measurements of soluble β-glucans from the ECM of Δtye7 or wild-type biofilms (data not shown). Therefore, we cannot exclude that this latter hypothesis is partly responsible for the altered structure of the Δtye7 biofilms.

In summary, results presented in this study bring credit to the longstanding hypothesis that the formation of C. albicans biofilms is associated with the establishment of a hypoxic environment, and demonstrate that biofilm formation requires functional defence mechanisms necessary to withstand hypoxia. This includes the upregulation of glycolytic genes mediated in part by Tye7p, which ensures sufficient energy to avoid cellular stress reflected by unrestricted filamentation, but also other mechanisms such as the activation of sulphur metabolism that could contribute to detoxification of reactive oxygen species (Garcia-Sanchez et al., 2004; Stichternoth and Ernst, 2009). Interestingly, Askew et al. (2009) have shown that the Tye7p and Gal4 regulators are simultaneously required for C. albicans virulence and suggested that these fungal-specific transcription factors could serve as antifungal targets. That Tye7p is necessary for biofilm formation makes it an even more attractive target.

Experimental procedures

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Strains, media and growth conditions

All C. albicans strains used in this study are listed in Table 2. Strains were grown at 30°C or 37°C on YPD medium (1% yeast extract, 2% peptone, 2% glucose) or SD minimal medium [0.67% yeast nitrogen base (YNB; Difco) with 0.4 or 2% glucose] supplemented, if necessary, with arginine, histidine and uracil, at 20 mg l−1 and 2% agar for solid media. For transformation using nourseothricin as a resistance marker, transformation mixtures were plated in YPD for 24 h at 30°C and then replica plated on YPD  +  50, 100, and 150 mg l−1 nourseothricin (Werner Bioagent). Transformants were obtained after 48 h incubation at 30°C. To induce hyphal formation, Lee's liquid medium, RPMI 1640 solid medium, and embedded culture conditions were used as previously described (Moreno-Ruiz et al., 2009). For inhibition of the mitochondrial respiratory chain (complex III), media were supplemented with antimycin A (10 mg l−1). Hypoxia was obtained by placing the cultures under anaerobic atmosphere using a Gaspak Pouch system (Becton Dickinson). For inhibition of ATP synthesis in the mitochondrial respiratory chain without affecting the respiratory chain, the uncoupling reagent DNP was added at various concentration in media. For glycolysis inhibition, media were supplemented with oxalate (10 mM).

Construction of C. albicans mutant strains

The lithium acetate-PEG protocol was used to transform C. albicans as described previously (Walther and Wendland, 2003). Knock-out mutants were constructed in BWP17 (Wilson et al., 1999) or BWP17U (Moreno-Ruiz et al., 2009) by sequential replacement of the complete ORF of both alleles using PCR-generated HIS1 or ARG4 disruption cassettes flanked by 100 bp of target homology region as described elsewhere (Gola et al., 2003). Gene replacements were verified by PCR, and Southern analyses were performed using HIS1 and ARG4 fragments as probes on EcoRI- and BamHI-digested genomic DNA. All oligonucleotides used in this study are listed in Table S4. If necessary, the strains were transformed to uracil prototrophy by targeting the URA3 marker into the RPS10 locus, using StuI-linearized CIp10 plasmid (Murad et al., 2000). Mono-allelic complementation of the tye7Δ::HIS1/tye7Δ::ARG4 knock-out was carried out by introducing a derivative of CIp10 harbouring the TYE7 gene to the RPS10 locus. A derivative of CIpSAT1 (Reuss et al., 2004; Moreno-Ruiz et al., 2009) carrying the YFP gene under the control of the C. albicans TEF1 promoter was used to transform the three strains BWP17AHU, CEC1207 and CEC1222 in order to constitutively express YFP in wild-type (CEC536), TYE7-deleted (CEC1412) and TYE7-complemented (CEC1416) strains.

Growth assays

Growth curves were performed twice with strains in triplicates in a Tecan Infinite M200 microplate reader, in 96-well plates containing 200 µl of culture. Measurements were made every 20 min, after shaking, for 30 h, at 30°C and 37°C. Growth assays were also carried out at least twice with strains in duplicates in 250 ml Erlenmeyer flasks containing 50 ml of culture, with shaking for 24 h at 150 r.p.m., at 30°C and 37°C, and measurements made manually with a spectrophotometer. In both experiments, cultures had an initial OD600 of 0.1. Doubling time was calculated between 2 and 10 h by the following formula: log2/{[log(OD10) − log(OD2)]/8}. Growth on YPD solid medium, alone or supplemented with oxalate or DNP, was performed by spotting 2 µl of culture (initial OD600 of 10) with serial 1/10th dilutions. Pictures of the edge of the colonies were taken with an inverted microscope with a 10 × objective.

Biofilm formation assays

Biofilms were produced in microfermentors as described previously (Garcia-Sanchez et al., 2004), with a continuous flow (2.5 ml min−1) of SD medium containing 0.4% glucose, 20 mg l−1 arginine, 20 mg l−1 histidine, 20 mg l−1 uracil and 200 mg l−1 methionine. Biofilms were grown for up to 40 h at 37°C. To quantify the biofilm dry mass, the biofilm in each microfermentor was resuspended in water, filtered, dried 3 days at 65°C, and weighted. Statistical differences were estimated using Mann–Whitney test. Alternatively, biofilms were produced on silicone squares in sterile 12-well plates, as previously described (Richard et al., 2005), with some modifications: strains were diluted to an initial OD600 of 1 in the same medium as for microfermentors, supplemented or not with oxalate or DNP. To measure the ratio of hyphal cells, the biofilm in each well was resuspended in 2 ml of water and vortexed for 1 min. The sample was then placed on a Kova Glasstic slide with grid (Hycor). Hyphal and yeast cells were counted by microscopy using a 40 × objective. At least three sets of biofilms on silicone squares were carried out. For each set, two biofilms per strain were used, and 10 squares per slide per biofilm were analysed.

Pyruvate kinase assay

All procedures were performed at 0–4°C. Cell-free extract was prepared from 20 OD600 units of cells, previously grown 8 h in SD 0.4% glucose. After centrifugation 5 min at 3500 r.p.m. at 4°C, the pellet was washed with 1 ml of lysis buffer (50 mM Tris-HCl, pH 8, 150 mM NaCl, 5 mM NaF and 10 mM β-glycerophosphate), vortexed and centrifuged (3500 r.p.m., 5 min, 4°C). The pellet was washed again with 0.5 ml of lysis buffer. Then, cells were mechanically disrupted with glass beads using a Fastprep (MP Biomedicals). After centrifugation (3500 r.p.m., 10 min, 4°C), the supernatant was immediately used for determination of protein concentration by the method of Bradford (1976) and pyruvate kinase activity (Kitanovic et al., 2009). All chemicals and enzymes were obtained from Sigma-Aldrich. The reaction was performed in a total volume of 150 µl containing 50 mM Tris-HCl, pH 7.1, 5 mM KCl, 2.5 mM MgCl2, 10 mM ADP, 0.7 mM NADH, 1.7 mM phosphoenolpyruvate, and 10 U of lactate dehydrogenase, supplemented or not with increasing concentrations of oxalate (0.1, 1, 10 and 100 mM). The reaction was initiated by the addition of 8 µl cell-free extract, absorbance decrease at 340 nm was monitored at 30°C in a Tecan Infinite M200 microplate reader. Negative controls were carried out without phosphoenolpyruvate or without crude extract. Pyruvate kinase activity was defined as ΔOD340 (0–10 min) per minute per mg of proteins. Two biological replicates were used and wild-type strain extracts were tested in triplicate.

Transcriptome analysis by microarray

The study was conducted by comparing deleted and complemented strains for TYE7 (CEC1207 and CEC1222 respectively), during a kinetic in biofilm conditions (24, 32 and 40 h growth) and in planktonic condition (exponential growth phase in SD 0.4% glucose). Briefly, the Thermanox plastic slide harbouring the biofilm was vortexed in 10 ml SD medium, and the resuspended biofilm was centrifuged 5 min at 3500 r.p.m. at 4°C. For planktonic cultures, after 6 h growth in flask, 50 ml of the culture was centrifuged 5 min at 3500 r.p.m. at 4°C. Then, cells were mechanically disrupted with glass beads using a Fastprep (MP Biomedicals). Total RNA was isolated using RNeasy mini kit (QIagen) according to the manufacturer's instructions. The quality and concentration of the isolated RNA were analysed using an Agilent 2100 Bioanalyzer. Ten micrograms of total RNA was used for cDNA synthesis using Superscript III Indirect cDNA labelling system (Invitrogen), according to the manufacturer's instructions. cDNA samples were labelled with Cy3 and Cy5 dyes (Amersham Biosciences). Purified fluorescent cDNAs were hybridized to oligonucleotides-microarrays containing 5896 C. albicans probes (Eurogentec) according to manufacturer's instructions. Arrays were scanned with an Axon 4000A scanner (Axon Instrument). Data were acquired and analysed using Genepix Pro 5.0 (Axon Instrument). For each comparison, two biological replicates were used and each biological replicate was subjected to technical replicates with dye-swaps. Data normalization (Lowess) and statistical analysis (Student's t-test) were performed using Genespring GX 7.3 (Agilent Technologies). Genes that showed at least a 1.5-fold induction in the complemented versus deleted strain for TYE7 and a P-value < 0.05 were defined as activated by Tye7p. Genes that showed at least a 0.67-fold repression in the complemented versus deleted strain for TYE7 and a P-value < 0.05 were defined as repressed by Tye7p. Microarray data have been deposited at ArrayExpress under accession number E-MEXP-2604. Normalized data (for significantly regulated genes) are available in Table S1. Gene ontology analyses were conducted through the Candida Genome Database (http://www.candidagenome.org/cgi-bin/GO/goTermFinder) and are available in Tables S2 and S3.

Confocal laser scanning microscopy of biofilms on silicone squares

Biofilms of wild-type (CEC536), TYE7-deleted (CEC1412) and TYE7-complemented (CEC1416) strains were observed at 60 h growth by direct observation and by confocal laser scanning microscopy (CLSM). These strains constitutively express the YFP under the TEF promoter. For CLSM imaging, biofilms were stained with 10 mg l−1 concanavalin A Alexa Fluor 594 conjugate (Invitrogen), as described previously (Nobile and Mitchell, 2005). Images were acquired using a Zeiss LSM 700 laser scanning confocal microscope on an upright Axio Imager.Z2 stand using a Zeiss W-nACHROPLAN 40X/0.75 working distance 2.1 mm objective, using Texas Red and YFP filters. All images were assembled using the Zeiss Zen 2009 software. The CSLM experiment was performed twice, using each time 2 biofilms per strain and analysing biofilm surface of at least 600 µm2.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Thanks are due to Malcolm Whiteway and colleagues for pointing the hypoxia-dependent morphogenesis phenotype of the Δtye7 mutant. We are grateful to Caroline Proux and Jean-Yves Coppée at the PF2 Platform of Pasteur Génopole Ile-de-France for help with micro-array experiments. We are also grateful to past and present members of the Fungal Biology and Pathogenicity Unit, Guilhem Janbon and Ismaïl Iraqui for continued interest and suggestions during the course of this study. J.B. was the recipient of a fellowship of the Fondation pour la Recherche Médicale (DEA20080713005). T.R. was the recipient of a postdoctoral fellowship in the framework of the NPARI consortium (LSHE-CT-2006-037692). This work was supported by grants from Institut Pasteur (PTR50), the Ministère de la Recherche et de la Technologie (PRFMMIP, ‘Réseau Infections Fongiques’) and the European Commission (QLK2-2000-00795, MRTN-CT-2003-504148, PITN-GA-2008-214004).

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Results
  5. Discussion
  6. Experimental procedures
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
MMI_7626_sm_FigureS1-2_TableS4.pdf3485KSupporting info item
MMI_7626_sm_TableS1.xls529KSupporting info item
MMI_7626_sm_TableS2.xls122KSupporting info item
MMI_7626_sm_TableS3.xls109KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.