Monitoring ALS1 and ALS3gene expression during in vitroCandida albicans biofilm formation under continuous flow conditions

Heleen Nailis, RoosmarijnVandenbroucke, Kelly Tilleman, Dieter Deforce, Hans NelisTom Coenye

Heleen Nailis, Hans Nelis, Tom Coenye (corresponding author)

Laboratory for Pharmaceutical Microbiology, GhentUniversity, Harelbekestraat 72, B-9000, Ghent, Belgium. Tel.: +32 9 264 81 41; Fax: +32 9 264 81 95; e-mail: .

Roosmarijn Vandenbroucke

Laboratory for General Biochemistry and Physical Pharmacy, GhentUniversity, Gent, Belgium

Kelly Tilleman, Dieter Deforce

Laboratory for Pharmaceutical Biotechnology, GhentUniversity, Ghent, Belgium

Keywords

Candida albicans,biofilms,ALS1, ALS3, gene expression, filamentation

Revised manuscript MYC0892

Abstract

ALS1 and ALS3 encodecell-surface associated glycoproteinsthat are considered to be important forCandida albicans biofilm formation. The main goal of the present study was to monitor ALS1 and ALS3 gene expression during C. albicansbiofilm formation(on silicone) under continuous flow conditions, using the Centers for Disease Control biofilm bioreactor. For ALS1 we found little changes in gene expression until at later stages of biofilm formation (72 and 96 h) when this gene appeared to be down-regulatedrelative to the gene expression level in the start culture. Weobservedan induction of ALS3 gene expression in the initial stages of biofilm formation (0.5, 1 and 6 h), whereas at later stages this genewas also downregulatedrelative to the gene expression level in the start culture. We also found that biofilms of an als3/als3 deletion mutant contained less filaments at several time points (1, 6, 24 and 48 h), although filamentation as such was not affected in this strain. Together, our data indicate an important role for ALS3 in the early phases of biofilm formation in the CDC reactor (probably related to adhesion of filaments), while the role of ALS1 is less clear.

Introduction

Candida albicans can colonize medical devices and multiple human tissues, resulting inbiofilm formation and biofilm-related infections [1,2]. In most cases antifungal therapy is not effective since cells in a biofilm (sessile cells) are highly resistant to antimycotics [1,3]. It has already been shown that in vitroC. albicans biofilm formation is a complex process involvingmultiple regulatory mechanisms [4]. Initially, yeast cells adhere to a surface and form microcolonies. Subsequently, cells in these microcolonies form filaments and produce an extracellular matrix; both are considered to be essential for the structural stability of biofilms [5,6]. Finally, filamentation and matrix production increase further until a mature biofilm with a distinct three-dimensional structure isformed[7-9].It has recently been shown that ALS1 and ALS3are important forC. albicans biofilm formation and both proteins are likely to have multiple functions in this complex developmental process [10,11].Als1p and Als3p are large, cell-surface-associated, glycoproteins, encoded by genes of the ALS(Agglutinin-Like Sequence) family [12,13]. Results from previous studies already showed that ALS1 and ALS3 gene expressionis altered in C. albicanssessile cells compared to planktonic cells[8, 14-16]. More recently, micro-array analysisperformed at different time points during biofilm formationrevealed a relative downregulation of ALS1 at 6 h in biofilms but no differential expression between planktonic and sessile cells was observed at the other time points[17]. In the same study an increase in ALS1 expression was noted in sessile cells between 6 and 12 h, after which the expression declined [17]. These observations are in agreement with data from earlier studies in which it was shown that ALS1 gene expression is induced when cells are transferred to fresh media but then gradually declines[18].Additional data regarding the importance of ALS1 and ALS3 come from studies with mutant strains. It has been shown that anals3/als3deletion mutant was impaired in biofilm formation (i.e. less biofilm biomass was observed in this mutant) due to the absence of an intertwining hyphal network and that als3/als3 mutant biofilms were comprised mainly of yeast cells, although the als3/als3 deletion mutant was not defective in filamentation [10,11]. In addition, an als1/als1deletion mutant was found to have a partial defect in biofilm formation since this mutant produced biofilms thatcan more easily be dispersed from the surface [10]. It was also observed that the expression of various ALS genes is influenced by the growth medium (e.g. transfer of cells to fresh medium was associated with increased ALS1expression) and other environmental conditions [17,18]. As such it can be anticipated that the biofilm model system can have a considerable impact on expression levels of particular genes (including the ALS genes), as some systems do not incorporate fluid flow (e.g. biofilm formation in a microtiterplate [15,19]), others do incorporate this aspect but in these systems the medium is not replaced (e.g. biofilm formation in microtiter plates on a rocking table [17,20]), while still others allow the continuous replacement of growth medium while subjecting the biofilms to a continuous flow (for example microfermentors [14], the Modified Robbins Device [21] and the Centers for Disease Control Bioreactor [22,23]).The Centers for Disease Control biofilm reactor (CDC reactor) has been used to study biofilm formation of various bacterial (e.g. Streptococcus pneumoniae) and fungal (e.g. C. albicans) pathogens [16,22,23].Futhermore, this system provides a means for collecting and quantifying cells from biofilms formed under flow conditions, over time.The aim of the present study was to determine the relative expression of ALS1 and ALS3 during biofilm formation in the CDC reactor, using Reverse Transcriptase-quantitative PCR (RT-qPCR).

Materials and Methods

Strains

Strains C. albicansSC5314 (American Type Culture Collection MYA-2876, wild type), CAYC2YF1U (an als1/als1 deletion mutant) [24]and CAYF178U(an als3/als3 deletion mutant) [10]were used. Cells were stored at -80°Cin Microbank tubes (Prolab Diagnostics, Richmond Hill, ON) and were routinely grownon Sabouraud Dextrose Agar (BD, Erembodegem, Belgium) plates at 37°C.

Biofilm formation

Start cultureswere prepared by incubating cellsin Sabouraud Dextrose Broth (SDB) (BD)for 16 h at 37°C in a shaking water bath. After 16 h of incubation, cells were washed three times with and finally resuspended in 1 mlof 0.9% (w/v) NaCl andfurther used to grow biofilms. Biofilms were grownin the CDC reactor (Biosurface Technologies, Bozeman, MT),as described previously [16].The CDC reactor consists of a 1-1 jacketed vessel with an effluent spout connected to a waste bottle. The jacket has two ports, which are connected to a circulating water bath set at 37°C. The lid of the vessel contains eight independent rods, each holding three silicone disks and three stainless steel ports allowing growth medium to be continuously pumped through the reactor vessel. Cells were added to 500 ml growth medium (1x YNB,50 mM glucose), transferred to the reactor vessel and stirred at 80 rpm. After 24 h of biofilm formation, diluted growth medium (0.2x YNB, 10 mM glucose) was continuously pumped through the reactor at a flow rate of 400 ml/h.

Confocal Laser Scanning Microscopy

After several time points of biofilm formation by the wild type and both deletion mutants, silicone disks were taken out of the reactor vessel and transferred to 24-well microtiter plates. One ml of a 250 µg per mlstock solution of Concanavalin A, bound to Alexa Fluor 647 (Invitrogen, Carlsbad, CA) (ex/em : 647 nm/668 nm), was added to each well and the plates were incubated for 90 min at 30°C. The biofilm structure was visualized using a Nikon C1si confocal laser scanning module attached to a motorized Nikon TE2000-E inverted microscope. After excitation of the samples using the 647 nm laser line from a DPSS laser, images were captured with a 60 x water immersion (numerical aperture 1.4) or a 40 x (numerical aperture 0.95) Nikon Plan Apochromat objective lens and the emitted fluorescence was captured using a Nikon 655LP detector. A non-confocal transmission image was collected simultaneously with the confocal images.

Determination of the number of culturable sessile cells

The number of culturable sessile cells was determined at different time points (after 0.5 h and 1 h and then after each additional 2 h of biofilm formation for up to 96 h) by transferring disks to 10 ml SDB. New sterile disks were placed into the rodsto make sure that the flow through the reactor vessel was not disturbed. Sessile cells were removed from the silicone disks by three cycles of 30 sec sonication (Branson 3510, 42 kHz, 100 W, Branson Ultrasonics Corporation, Danburry, CT) and 30 sec vortex mixing.Using this procedure all cells are removed from the disks and clumps of cells are broken apart (data not shown). Serial 1:10 dilutions of the resulting cell suspensions were plated on Sabouraud Dextrose Agar. The plates were incubated for 24 h at 37°C, after which the colonies were counted.The number of culturable sessile cells was determined on at least 6 disks for each time point.The mean number of colony forming units (CFU) per disk and corresponding standard deviationwas calculated for each time point tested. Statistical analysis (one way ANOVA) was carried out using SPSS 15.0 software (p<0.05).

Measuring gene expression using RT-qPCR

Start cultures and biofilms were grown as described above. After different time points of biofilm formation (0.5 h, 1 h , 6 h , 12 h, 24 h, 36 h, 48 h, 72 h and 96 h) disks were transferred to 0.9% (w/v). Following sonication and vortex mixing (see above), cells were washed three times with and finally resuspended in 1 mlof 0.9% (w/v) NaCl. RNA extraction, DNase treatment and quantification of the RNA samples, cDNA synthesis and qPCR analysis were performed as described previously (Nailis et al., 2006).No fluorescent signal was detected in the RNA aliquots which confirms that all genomic DNA was fully degraded during the DNase reaction. Samples were obtained from at least three independently-grown biofilms. For accurate and reliable normalization of qPCR data, the use of the geometric mean of multiplehousekeeping genesis recommended[25]. In the present study we used ACT1, RIP, RPPB2B, PMA1 and LSC2 for this purpose [16]. Followingnormalization, normalized relative quantities were used to calculate ALS1 and ALS3 gene expression levelsat different stages during biofilm formation relative to the expression in the start culture (cells harvested after 16 h of growth in SDB). Mann-Whitney U tests were carried out using SPSS 15.0 software to determine whether differences in ALS1 and ALS3 gene expression were statistically significant (p0.05).

Determination of the fraction of filaments using solid phase cytometry

To quantify the fraction of filaments(germ tubes, true hyphae or pseudohyphae) at different time points of biofilm formation and in the start culture, a rapid method based on solid phase cytometry (SPC)was developed.Cells were grown and harvested as described above.Serial 1:10 dilutions of cell suspensions were made in sterile and particle free 0.9% (w/v) NaCl. One ml of the diluted suspensions was filtered over a Cycloblack-coated polyester membrane (diameter 25 mm, 2.0μm pore size; Chemunex, Ker Lann, France), followed by 1 h incubation at 37°C of the filter on a cellulose pad (diameter 25 mm) (Chemunex), saturated with 600 µL ofthe viability substrate ChemChrome V6(1:100 in ChemSol B2) (Chemunex). Following incubation, the entire surface of the Cycloblack-coated polyester filter was scanned using the ChemScanRDI device (Chemunex). The fluorescent signals were processed by a computer, using software discriminants that allow differentiation between valid signals and fluorescent background. Scan results were displayed on the computer in a primary window map representing an image of the membrane filterwith accurate position information for each detected spot. The software discriminates between fluorescent spots which do and do not complywith the pre-setparameters. The retained spotswere displayed in the secondary window mapand each fluorescent spot was visually inspected using a computer-driven moving stagecoupled to an Olympus BX40 epifluorescence microscope. Highlighting a spot in the secondary scan map directs the motorized stage to the corresponding position on the membrane filter and allowed us to rapidly determine whether the fluorescent spot is a yeast cell ora filament. For each strain, experiments were carried out at least in six-fold for each time point. Results were analysed with independent student t-tests or one-way ANOVA using SPSS 15.0 software; differences were considered significant if p<0.05.

Results and discussion

C. albicans biofilm growth in the CDC reactor

The use of the CDC reactor allowed the formation of high-density C. albicans biofilms on silicone disks under continuous flow conditions with constant medium replacement. Young biofilms (up to 1 h) contained approximately 104 CFU per silicone disk and the number of culturable sessile cells increased exponentially during the initial stage of biofilm formation in the CDC reactor (1 – 12 h). Later (up to 48 h) the number of culturable sessile cells increased further but more slowly than in the initial stages, until a mature biofilm containing approximately 108 CFU per disk was obtained after 48 h. Additional incubation (up to 96 h) did not result in a further increase in the number of culturable sessile cells despite the incomplete colonization of the silicone surface, as demonstrated by CLSM (Fig. 1). It is likely that at later stages of C. albicans biofilm development a steady-state condition is reached, in which multiplication and detachment of sessile cells result in a mature biofilm in which no net increase in the number of culturable sessile cells occurs [26,27]. We found that deletion of ALS1 or ALS3 did not affect the number of culturable sessile cells in CDC-grown C. albicans biofilms. After 1 h of wild type biofilm growth, concanavalin A staining combined with CLSM revealed the presence of mainly filaments (Fig. 1A). Further biofilm development (48 h and 96 h) resulted in a complex three-dimensional network of yeasts and filaments, embedded in a dense matrix (Fig. 1B and 1C). No major differences were observed between the als1/als1 mutant and the wild type strain, but the als3/als3 deletion mutant biofilm had an altered structure with less filaments both after 1 h (Fig. 1D) and 48 h of biofilm growth (Fig. 1E). However, after 96 h no differences in the structure between biofilms formed by the als3/als3 deletion mutant (Fig. 1F) and the wild type were observed. It should be noted that it has previously been shown that the als3/als3 mutant is not defective in hyphal formation [10] and this was confirmed in the present study using planktonic cultures and hyphae-inducing conditions (data not shown).Taken together, C. albicans biofilms formed on silicone disks in the CDC reactor under continuous flow conditions with constant medium replacement containeda distinct three-dimensional structure of both yeast cells and filaments embedded in an extracellular matrix. Similar biofilm structures were previously observed for C. albicans in other in vitro model systems [7, 8, 10, 11].

ALS1 and ALS3 gene expressionin C. albicansbiofilms

A relatively high basal ALS1 and ALS3 gene expression level was detected in the biofilmsfor C. albicans SC5314. The Ctvalues ranged between 17.2 and 25.8 for the ALS1 gene and between 20.6 and 27.5 for the ALS3 gene. ALS1 and ALS3 gene expression levels (relative to the expression in the start culture) monitored at different time points during C. albicans SC5314 biofilm formation in the CDC reactor are shown in Fig.2. Nosignificantchanges in ALS1 gene expression were detected up to 48 hof biofilm formation, whereasthis gene was downregulated in older biofilms (72 h and 96 h) (p<0.05).In contrast, a significant increase of ALS3 gene expression was observedin the initial stages of biofilm formation as a 12-fold and 17-fold induction was detected after 0.5 h and 1 h, respectively (p<0.05). After 6 h, ALS3 gene expression decreasedbut we still observed a 3-fold induction at this stage of biofilm development compared to the start culture (p<0.05). At later stages (48to 96 h of growth) asignificant underexpression of ALS3 was found (p<0.05). Most studies regarding gene expression in C. albicans biofilms have focussed on comparisons between planktonic and sessile cells or between sessile cells formed in various model systems (see for example [8,14-16]) and in most of these studies a marked overexpression of ALS1 was observed in sessile cells. However, less information is available concerning ALS3 expression and/or kinetics of ALS1 and ALS3 expression during biofilm formation. Murillo et al. [28] showed that ALS1 was underexpressed in young biofilms (up to 2.5 h) and was not different from the start culture in older (4.5 – 6.5 h) biofilms. Yeater et al. [17] noted an increase in ALS1 expression in sessile cells at 12 h compared to earlier (6 h) and later (48 h) stages. Combined, these data suggest that increased expression of ALS1 is not required for the initial attachment of planktonic cells to the surface. In the present study we did not observe significant differences in ALS1 expression between the start culture and biofilm samples taken after up-to 48 h, confirming that an increase in ALS1gene expression is not required during the initial stages of C. albicans biofilm formation. However, we did observe a marked decrease of ALS1 expression in later stages (72 h and 96 h) of biofilm formation. To our knowledge there are no other gene expression data available for late-stage C. albicans biofilms (e.g. 72 h and 96 h old), and it remains to be determined whether the reductions at these later stages observed are a general phenomenon in C. albicans biofilm formation or whether this is specific for biofilm formation in the CDC reactor. In addition, it has already been shown that ALS1 gene expression level changes are influenced by the growth stage of a culture [17,18]. It is possible that the decrease of ALS1 gene expression at later stages is associated with a reduced growth rate of sessile cells in mature biofilms (72 h and older). Until now, no information is available about the growth rate of sessile cells in biofilms grown in the CDC reactor so this hypothesis requires further investigation. In the microarray study of Murillo et al.,ALS3 was reported to be considerably overexpressed in young biofilms compared to the start culture, however as this overexpression was also noted in planktonic cells, its reason was unclear [28]. In the present study we noted a marked increase in early stages of biofilm formation (0.5 – 6 h), followed by a period in which expression returned to the levels observed in the start culture (12-36 h), after which the expression decreased further (48 – 96 h). Collectively, these data seem to indicate that ALS3 mainly plays a role during early stages of biofilm formation in the CDC reactor. However, it has already been shown that ALS3 is a hyphae-specific gene and that ALS3 mRNA levels are correlated with hyphal development [29,30]. Therefore, it is likely that the overexpression of ALS3 observed at initial stages of biofilm formation in the CDC reactor is associated with the morphology of sessile cells in biofilms since we also found an increase in the fraction of filaments at these stages.

Filamentation in biofilms formed by als1/als1 and als3/als3 deletion mutants

To further determine the role of both genes inC. albicans biofilm formation, filamentation during biofilms formation of the wild type and the als1/als1 and als3/als3 deletion mutants was investigated. Overall C. albicans SC5314 and both mutant strains showed a similar filamentation pattern (Fig. 3), although the fraction of filaments in the als3/als3 mutant biofilm was significantly lower for most time points (1, 6, 24 and 48 h) (p<0.05). Start cultures for all strains contained appr. 5% of filaments and a considerable increase in filamentation was observed in the early stages (up to 6 h) of biofilm formation, with 40-60% of filaments after 1 h and 6 h of biofilm formation (Fig. 3). During later stages, the fraction of filaments decreased to lower levels; however these levels (10-30%) were still considerably higher than the levels in the start culture. As mentioned above, we foundno differences in the number of culturable sessile cellsbetween biofilms formed by the wild type and by the als1/als1 or als3/als3 deletion mutant. Using biofilm dry mass determinations, a similar observation was previously made for the als1/als1 mutant but an approximate 4-fold reduction in biofilm dry mass had been observed for the als3/als3 mutant after 60 h [10]. Since theals3/als3 mutant contained significantly less filaments than the wild type after 1, 6, 24 and 48 of biofilm formation in the CDC reactor, this could explain the discrepancies observed between the two quantification approaches (determination of number of culturable cells vs. dry mass determinations), as a filament and a yeast cell will both be counted as a single colony forming unit although the former contributes more to the biofilm mass due to its larger size.As discussed above, the als3/als3 mutant is not defective in hyphal formation, suggesting that Als3p has a major role in biofilm formation [10,31].Our results confirm this and suggest that Als3p may play a role in the specific adhesion of filaments. However, various processes including transport and degradation will influence the fate of Als1p and Als3p on the C. albicans cell surface and studying the levels of both proteins using specific antibodies [32] will be required to further confirm their role in various stages of biofilm formation in different model systems.