Growth of Candida albicans in human saliva is supported by low

FEMS Yeast Research, 15, 2015, fov088

doi: 10.1093/femsyr/fov088Advance Access Publication Date: 21 September 2015Research Article

RESEARCH ARTICLE

Growth of Candida albicans in human saliva issupported by low-molecular-mass compoundsMarianne Valentijn-Benz, Kamran Nazmi, Henk S. Brand, Wim van ’t Hofand Enno C. I. Veerman∗

Section of Oral Biochemistry, Academic Centre for Dentistry Amsterdam (ACTA), Vrije Universiteit andUniversiteit van Amsterdam, Gustav Mahlerlaan 3004, 1081 LA Amsterdam, the Netherlands∗Corresponding author: Academic Centre for Dentistry Amsterdam (ACTA), Department of Oral Biochemistry, Gustav Mahlerlaan 3004, 1081 LAAmsterdam, the Netherlands. Tel: +31-20-5980883; E-mail: [email protected] sentence summary: The yeast Candida albicans can thrive on salivary constituents as nutrients, to maintain itself in the oral cavity.Editor: Carol Munro

ABSTRACT

Saliva plays a key role in the maintenance of a stable oral microflora. It contains antimicrobial compounds but alsofunctions as a substrate for growth of bacteria under conditions of low external nutrient supply. Besides bacteria, yeasts, inparticular Candida albicans, commonly inhabit the oral cavity. Under immunocompromised conditions, instantaneousoutgrowth of this yeast occurs in oral carriers of C. albicans, suggesting that this yeast is able to survive in the oral cavitywith saliva as sole source of growth substrate. The aim of the present study was to identify the salivary constituents thatare used by C. albicans for growth and survival in saliva. In addition, we have explored the effect of growth in saliva on thesusceptibility of C. albicans to histatin 5, a salivary antifungal peptide. It was found that C. albicans was able to grow inhuman saliva without addition of glucose, and in the stationary phase could survive for more than 400 h. Candida albicansgrown in saliva was more than 10 times less susceptible for salivary histatin 5 than C. albicans cultured in Sabouraudmedium.

Keywords: saliva; C albicans; glucose; culturing; histatin; peptides

INTRODUCTION

The numbers of microorganisms living within the body of theaverage healthy adult human outnumber human cells 10 to 1.One of the most heavily colonized parts of the human body isthe oral cavity, which harbors more than 108 microorganisms.Although saliva contains several antimicrobial systems (NieuwAmerongen andVeerman 2004; Van ‘t Hof et al. 2014), it obviouslydoes not render the mouth aseptic, but rather seems to supportthemaintenance of amultispecies benign oralmicroflora. In thisprocess saliva can play a key role, because on one hand it con-tains a wide variety of defensive systems that protect oral tis-sues against excessive microbial colonization, whereas on the

other hand it is a source of (glyco)proteins and peptides po-tentially serving as growth substrate for microorganisms underconditions of low external nutrient supply. It is conceivable thatmicroorganisms with the ability to survive and proliferate onthese substrates under nutrient-poor conditions have an eco-logical advantage. Indeed, various oral bacteria are able to growin vitro to a certain extent in saliva or agar media prepared fromheated saliva (Van der Hoeven and Camps 1991; De Jong et al.1997; Shelburne et al. 2005). A number of studies have demon-strated that oral bacteria are capable of utilizing the complexcarbohydrates of human salivary mucins as a nutrient source(Wickstrom and Svensater 2008; Wickstrom, Hamilton and

Received: 13 April 2015; Accepted: 18 September 2015C© FEMS 2015. All rights reserved. For permissions, please e-mail: [email protected]

1

Downloaded from https://academic.oup.com/femsyr/article-abstract/15/8/fov088/2467730by gueston 06 April 2018

http://www.oxfordjournals.org

mailto:[email protected]

mailto:[email protected]

2 FEMS Yeast Research, 2015, Vol. 15, No. 8

Svensater 2009; Wickstrom et al. 2009). This process requires awide variety of glycosidases.

Besides bacteria, yeasts, in particular Candida albicans, com-monly inhabit the oral cavity. Candida albicans is present in theoral cavity of 40–60% of the general population (Samaranayake,Leung and Jin 2009), yet uncontrolled outgrowth of C. albicans oc-curs much less frequently and is related to several, often host-related, factors. For instance, suppression of the immune sys-tem, e.g. by inhalation of corticosteroids, frequently results in in-stantaneous outgrowth of the yeast in oral carriers of C. albicans.This suggests that under healthy conditions C. albicans is able tosurvive in a dormant state in the oral cavity. Previous studiessuggest that C. albicans is not able to thrive on human saliva inthe absence of added glucose, suggesting that, unlike bacteria,the yeast cannot use salivary constituents as growth substrate.Recently, however, it was described that C. albicans can survivein tap water containing trace amounts of saliva (Gerhardts et al.2012).

Because of the presumably suboptimal growth conditions inthe oral cavity, C. albicans present in this natural ecological habi-tat will likely differ from those cultured in vitro. This may haveprofound effects on their virulence as well as on their resis-tance against innate immune proteins. For instance, it has beendemonstrated that, when the energy metabolism is suppressed,the yeast’s resistance against antimicrobial peptides, includinghistatins and defensins, is enhanced (Veerman et al. 2007). Thepresent study was undertaken to explore whether C. albicans isable to remain viable in saliva as sole nutrient source, and towhich extent this affects its susceptibility to histatin 5, a sali-vary antifungal peptide. We found that C. albicans was able togrow in human saliva without addition of glucose, and in thestationary phase could survive for more than 400 h. In addition,C. albicans cultured in saliva was more than 10 times less sus-ceptible for salivary histatin 5 than those cultured in Sabouraudmedium.

MATERIALS AND METHODSCollection of saliva from healthy human volunteers

Saliva was collected on ice, without conscious stimulation, fromhealthy volunteers (3 males, 2 females, 23–61 y) according to themethod ofNavazesh et al. (1992). In thismethod, saliva is allowedto accumulate in the floor of the mouth and the subject spit outin ice-chilled test tubes every 60 s. Collection of saliva was doneunder a VUmc Ethical Committee approved protocol (OB 96 01).After collection, the samplewas homogenized on a Vortexmixerfor 1 min to decrease the viscosity, and subsequently clarifiedby centrifugation at 10 000 g for 10 min to remove bacteria andcellular debris. The supernatant (CHWS) was directly used orstored at −20oC until use. In some experiments, pooled clari-fied human whole saliva (CHWS) was sterilized by incubationin boiling water for 15 min. Sterilized pooled CHWS was frozenin 1 ml aliquots at −20◦C and thawed immediately before use.Secretions from the parotid glands (PAR) were collected usingLashley cups as described previously (Veerman et al. 1996). Sub-mandibular/sublingual (SM/SL) and palatal (PAL) secretionswerecollected using custom-made devices, as described previously(Veerman et al. 1996).

Growth of C. albicans in human saliva

Candida albicans (ATTC 10231) was cultured aerobically at 30oCon Sabouraud dextrose agar plates (SDA, Oxoid, Hampshire, UK).

Yeast was inoculated in 25ml of Sabouraud dextrose broth (SDB,Oxoid) in a 100ml Erlenmeyer flask. After 20 h of growth at 30oC,1 ml of this suspension was subcultured for 1–2 h in 20 ml ofSDB, to obtain a mid-log phase culture. After harvesting, cellswere washed three times in 1mM PBS and resuspended in salivain Falcon culture tubes with caps loosely tightened, under gen-tle shaking (30oC). Aliquots were taken at indicated time points,homogenized on a Vortex mixer to disrupt aggregates and OD600

was measured. For determination of CFU, samples were diluted100-fold, homogenized on a Vortex mixer and plated on SDA.The plates were incubated at 30◦C and after 2 or 3 days colonieswere counted.

Biochemical analysis

Salivary glucose was determined with an enzymatic method(Boehringer Mannheim, Mannheim, Germany) according to themanufacturer’s instructions.

Sialyl-Lewisa and sulfo-Lewisa antigens on salivary mucinswere determined by ELISA as described previously (Veermanet al. 1997). In short, saliva samples obtained before and aftergrowth were 2-fold serially diluted in 0.1 M NaHCO3, (pH = 8.0),starting from 1:200. After incubation overnight at 4oC, plateswere washed, and subsequently incubated with MAb E9 (anti-sialyl-Lewisa) or MAb F2 (anti-sulfo-Lewisa). Bound antibodieswere detected using anti-mouse immunoglobulin conjugated toHRP in combination with OPD/H2O2.

SDS-PAGE analysis

SalivaSaliva samples, before and after growth of C. albicans, were sup-plemented with four times concentrated sample-solubilizingbuffer (50 mM Tris-HCl, pH 6.8, containing 1% SDS and 10mM DTT), boiled for 5 min and examined by SDS-PAGE onprecast 4–20% polyacrylamide gradient gels (Novex, Life Tech-nologies, Bleiswijk, the Netherlands). After electrophoresis, gelswere protein-stained with Coomassie Brilliant Blue R250 in10% (v/v) acetic acid and carbohydrate-stained using the Pe-riodic Acid Staining (PAS) reagent. Gels were destained in10% acetic acid.

Dialysis of saliva

Heat-treated CHWS was placed into dialysis tubings (Spec-tra/Por, Spectrum Europe, Breda, the Netherlands), with cut-offvalues of 8000–10 000 Da, and dialyzed against a 10-fold excessof saliva buffer (1 mM phosphate, 20 mM KCl, 1 mM CaCl2, pH7.0) with three buffer changes. The dialyzed saliva (‘retentate’)was inoculated with C. albicans to a starting OD600 of ∼0.2. For 72h, growth of the yeast was monitored by measuring the OD600 atregular intervals.

In separate experiments, the dialyzable fraction of saliva(‘dialysate’) was obtained by dialyzing 10 ml heat-treated CHWSagainst 3 × 100 ml of a volatile dialysis buffer consisting of 0.1 M(NH4)2CO3 (pH 7.2). After overnight dialysis, the three portions ofvolatile buffer containing the dialyzedmaterial were pooled andlyophilized. For use in the experiments, this material was recon-stituted into the original volume (10ml) of either saliva buffer orof the non-dialyzable fraction of saliva (‘retentate’). These solu-tions were also tested for their capability to support growth of C.albicans, as described above.


Valentijn-Benz et al. 3

Sensitivity of C. albicans to salivary histatin 5

Sensitivity of C. albicans to the membrane-disruptive activityof histatin 5 was determined by monitoring the fluorescenceenhancement of propidium iodide (PI, Invitrogen, Breda, theNetherlands) in peptide-treated cells, as described previously(Veerman et al. 2007). The membrane impermeant dye PI onlyenters membrane-compromized cells, after which the fluores-cence of this probe is enhanced by 20- to 30-fold due to its bind-ing to nucleic acids. Cells were grown for 24 h in saliva or SBDmedium, harvested by centrifugation and incubated with 5 mMpotassium phosphate, pH 7.0 (PPB), supplemented with 5 mMNaCl or NaN3 at a density of 2 × 107 cells ml−1. After 30 min,PI was added to a final concentration of 10 μM and cells weretransferred to wells containing equal volumes of serially dilutedhistatin 5 (0–100 μM) in the corresponding buffer. Developmentof fluorescence was for each peptide concentration monitoredin a Fluostar Galaxy microplate fluorimeter (BMFG Labtechnolo-gies, Offenburg, Germany) at 5min intervals for 1 h, at excitationand emission wavelengths of 544 and 620 nm, respectively.

Statistical analysis

Data are presented as means ± standard error, and were sta-tistically analyzed with IBM SPSS Statistics version 21.0 (IBMCorp., Armonk NY, USA; Chicago, USA), using one-way or two-way ANOVA. All levels of significance were set at P < 0.05.

RESULTS

To examine the ability of C. albicans to use human saliva asgrowth substrate, CHWS from several donors was inoculatedwith C. albicans cells, grown in amid-log phase on SDA (Fig. 1). Asa control, cellswere incubated in PBS and saliva buffer. A startingcell density of approximately 2 × 106 cells ml−1 was used, corre-sponding to an OD600 of ∼0.08. At different time points, aliquotswere taken for OD600 measurement and determination of CFUsby plating. Because C. albicans tended to clump in whole saliva,it was difficult to correlate OD600 measurements with CFUs. Vig-orous vortexing of the suspensions, before measurement andculturing, improved the correlation and was therefore routinelydone. Four hours after inoculation with C. albicans OD600 andCFUs started to increase (Fig. 1). This continued for approxi-mately 20 h, after which the levels remained constant.

Although interindividual variations were present, CHWS ofall individuals tested supported growth of C. albicans (two-wayANOVA, P < 0.0005). Growth of yeast was confirmed by micro-scopic inspection, showing that at the end of the incubationonly yeast cells, often in large aggregates, were present in saliva.Three species of Candida (C. albicans, C. glabrata and C. tropicalis)exhibited similar growth patterns on saliva (not shown). In thecontrols (saliva buffer either with or without glucose), OD600 andCFUs gradually decreased and after 70 h virtually no viable yeastcells were present as determined by culturing.

Next we examined the effect of the inoculum density on thegrowth of C. albicans (Fig. 2). CHWS of one donor was inocu-lated with different quantities of C. albicans to OD600 values of0.027, 0.125 and 0.45, respectively. After different periods of time,aliquots were taken for OD600 measurement and determinationof CFUs. Irrespective of the initial density, each of the suspen-sions had reached approximately the same OD600 after 40 h sug-gesting that growth was supported to a certain limit. To exam-ine if C. albicans could persist in saliva for a prolonged periodof time, the growth of the yeast in saliva was monitored for up

Figure 1.Growth pattern ofC. albicans in salivas fromdifferent donors.Candida al-bicans cells, grown in amid-log phase on SDA, were inoculated in saliva collected

from different donors, in PBS and in saliva buffer (SB). Starting cell density was2 × 106 cells ml−1 (OD600 of ∼0.08). A–E: saliva from five different donors. Salivabuffer: 2 mM phosphate, 0.1 mM MgCl2, 2 mM KSCN, 50 mM KCl (pH 6.8). Atthe indicated time points, aliquots were taken. At different time points, aliquots

were taken for OD600 measurement (1A) and determination of CFUs by plating(1B). Values are means ± standard error of three separate incubations. Growth insaliva differed significantly from the growth in the control (saliva buffer) (two-way ANOVA, all P < 0.0005).

Figure 2. Effect of inoculum density on growth of C. albicans in saliva. Saliva of

one donor (final volume 2 ml) was inoculated with C. albicans to starting OD600

values of 0.027, 0.125 and 0.45. At the indicated time points, aliquots were takenfor determination of OD600.

to 600 h (Fig. 3). Because of the large volume required, in thisexperiment a pooled saliva batch was used, obtained from sixdonors. To lower the risk of bacterial infection, it was sterilizedby incubation in boiling water for 15 min. Control experimentsrevealed that sterilization of saliva had no effect on the growthof C. albicans (not shown). After approximately 20 h, the cultureentered the stationary phase, which was maintained for at least600 h. After 200 h, part of the saliva-grown cellswas harvested bycentrifugation and resuspended in a batch of fresh saliva. Thisrestored growth to an approximately two times higher level, as



Figure 3. Prolonged growth of C. albicans in saliva. Candida albicans cells, grown ina mid-log phase on SDA, were inoculated in 9 ml of pooled heat-treated CHWS,

obtained from six donors. Incubations were carried out in triplicate. Startingdensity: approximately 1 × 106 cells ml−1. At different time points, aliquots weretaken for OD600 measurement and CFU determination. After 7 days (arrows),2 × 3 ml aliquots of the suspension were taken and centrifuged to separate

spent saliva and C. albicans. Upper curve: C. albicans in a new batch of heat-treated CHWS. Middle curve: prolonged cultivation in original (spent) CHWS.Lower curve: spent CHWS inoculated with fresh C. albicans. Values are mean ±standard error. ∗P < 0.0005 (vs t = 0 h); #P < 0.0005 (vs t = 168 h) (ANOVA).

evidenced by an increase in both OD600 and CFUs. Inoculationof the resulting supernatant (spent CHWS) with a fresh batchof C. albicans did not induce a further increase in C. albicans(Fig. 3). Spent CHWS supplemented with 100 μM glucose and amixture of all amino acids at a physiological concentration of20 μM (Brand et al. 1997) did not support growth of C. albicans(not shown). The combined results of Figs 2 and 3 indicate thatCHWS is a source of growth substrates on which C. albicans canthrive. Furthermore, C. albicans in the stationary growth phasesurvived in CHWS for more than 600 h (Fig. 3). In contrast, inmineral buffers (PBS or saliva buffer) CFU as well as OD600 de-creased after 20 h.

Mucins do not support growth of C. albicans

Next we wanted to identify the salivary constituents responsi-ble for the saliva-supported growth of C. albicans. Whole salivaoriginates from various glandular sources, each of which con-tributes a characteristic set of (glyco)proteins and peptides(Veerman et al. 1996). We examined if secretions directly col-lected from the parotid (PAR), palatal (PAL) and submandibu-lar (SM) glands supported growth of C. albicans (Fig. 4). Of theglandular secretions, SM supported growth to the largest extent

Figure 4. Growth pattern of C. albicans on secretions from individual salivaryglands. Candida albicans cells, grown in a mid-log phase on SDA, were inoculated

in different glandular secretions collected from one donor. Starting cell densitywas approximately 1 × 106 cells ml−1 (OD600 of approximately 0.08). At differentpoints, time aliquots were taken for determination of CFUs. Values are mean of

triplicates ± standard error. CHWS: cleared whole saliva; PAL: Palatal saliva; PAR:parotid saliva; SM: submandibular saliva. #P = 0.45; ∗∗P = 0.01; ∗∗∗P = 0.0005.

(P = 0.0005), followed by PAL (P = 0.001). In PAR, no significantgrowth occurred (P= 0.45). Chemical analysis of the glucose con-tent in CHWS after growth revealed that after culturing of C. albi-cans, the glucose concentration had dropped to 2 μM, indicatingthat salivary glucose had been used as growth substrate. Still,there was no clear correlation between the growth supportingproperties of the different salivas and their glucose concentra-tion. For instance, the glucose concentrations in CHWS and PARwere comparable, still, hardly if any growth occurs on PAR. Onthe other hand, the growth on CHWS (10 μMglucose) was clearlyhigher than on PAL (110 μM glucose). Furthermore, after sup-plementing spent CHWS or PAR with 100 μM glucose, still nogrowth of C. albicans was observed (not shown). Because CHWS,PAL and SM, but not PAR, contain MUC5B, we hypothesized thatC. albicans might utilize these heavily (>90%) glycosylated gly-coproteins as carbohydrate source, similar to bacteria (Van derHoeven and Camps 1991; Wickstrom and Svensater 2008). Wethus examined if mucin carbohydrates were degraded duringgrowth of C. albicans (Fig. 5), by monitoring the sialic acid con-tent of mucins. Hydrolysis of sialic acid, the terminal residue ofthe carbohydrate side chains, is the first step in the degradationof complex carbohydrates, making the remainder of the chainaccessible for enzymes. To examine if breakdown of the carbo-hydrate moiety of MUC5B occurred during growth of C. albicanson (heat-sterilized) CHWS, aliquotswere taken before and after 7days of growth and tested in ELISAwithmAb E9, directed againstsialic acid containing epitopes (Veerman et al. 1997). Previously, ithas been found that the E9 epitope is a sensitive marker for bac-terial degradation of carbohydrates in mucins (Nieuw Ameron-gen et al. 1993). As a control mAb F2 was used. This antibodyrecognizes the sulfo-Lewisa epitope, which is more robust than



Figure 5. Effect of C. albicans on the concentration of MUC5B-associated carbo-hydrate epitopes. Heat-treated CHWS was incubated with or without C. albicans

for 7 days. Before and after culturing the concentration of sulfated and sialylatedLewisa was determined by ELISA. (A) Sulfo-Lewisa; diamonds: before culturing;filled circles: heat-treated CHWS after 7 days incubation with C. albicans; opencircles: untreated (infected) CHWS incubated in parallel without C. albicans. (B)Sialo-Lewisa; diamonds: before culturing; filled circles: after 7 days incubationwith C. albicans; open circles: untreated (infected) CHWS after 7 days withoutC. albicans.

sialyl Lewisa (Veerman et al. 1991, 1997). After 7 days of growth ofC. albicans, levels of both E9- and F2-epitopes were not changed(Fig. 5). In a batch of non-sterilized saliva incubated in parallel, abacterial infection developed over this period, as was confirmedby visual and microscopic inspection. In this sample, the sialyl-Lewisa epitope was barely detectable after 7 days, whereas thelevel of sulfo-Lewisa was virtually unchanged (Fig. 5). This illus-trates the sensitivity of the E9 epitope for breakdown by bac-terial enzymes. Analysis of the saliva protein composition be-fore and after growth of C. albicans using SDS PAGE confirmedthat little if any degradation of mucins or of other salivary pro-teins occurred during growth of Candida in heat-treated CHWS(Fig. 6). On the other hand, SDS PAGE analysis of the infectedsaliva batch revealed a decrease in the intensity of bands cor-responding to mucin and other specific proteins. To explore ifchanges occurred in low-molecular-mass salivary constituents,which are not detected by SDS PAGE, in another experimentCHWSwas analyzed by MALDI-MS. This revealed that a series of

Figure 6. SDS PAGE analysis of CHWSbefore and after growth ofC. albicans. CHWSwas incubated with or without C. albicans for 7 days. Left panel: PAS staining; –:

no pretreatment of CHWS; 100◦C, heat-sterilization; C. albicans: inoculated withC. albicans; day 0, day 7: before, respectively, after 7 days incubation at 30◦C.∗ ; untreated, infected saliva.

compounds in the mass range of 1000–2000 Da had disappearedafter growth of Candida (Fig. 7).

Dialysis of CHWS removes the growth-supportingcompounds

To further characterize the putative growth substrates in saliva,we fractionated saliva by dialysis using a membrane with acut-off of approximately 8000 Da and subsequently tested thehigh- and low-molecular-mass fractions individually. After re-constitution to the original volume of either water or retentate,the dialyzed fraction supported growth of C. albicans (Fig. 8). Thenon-dialyzable fraction (retentate) no longer supported growth.This again underlines that the growth of C. albicans on salivadepends on low-molecular-mass constituents. Saliva containsa variety of low-molecular-mass compounds including glucose,amino acids and ammonia, which are potential growth sub-strates for C. albicans. Together these results suggest that be-sides glucose, other low-molecular-mass compounds, presum-ably peptides, are required for growth of C. albicans in saliva.

Effect of saliva culturing on susceptibility to salivaryhistatins

In a first attempt to examine phenotypic consequences ofgrowing on a low-energy substrate such as saliva, we tested thesensitivity of C. albicans towards histatin 5, an innate immunitypeptide of saliva, with potent in vitromembrane-disruptive prop-erties against C. albicans. The sensitivity of C. albicans to histatin5 was determined using the viability probe PI (Fig. 9). When themembrane is compromised, PI can enter the cell, after which itsfluorescence is enhanced 20- to 30-fold due to its binding to nu-cleic acids. These experiments revealed that saliva-grownC. albi-canswas more than 10 times less susceptible to salivary histatin5 than C. albicans cultured in Sabouraud medium (Fig. 8). The



Figure 7. MALDI-MS analysis of the low-molecular-weight fraction of saliva be-

fore (upper panel) and after growth (lower panel) of C. albicans for 2 days.

Figure 8. Effect of dialysis of CHWS on growth of C albicans. Candida albicans cells

were inoculated in CHWS fractionated by dialysis. CHWS: startingmaterial (con-trol); R: dialyzed CHWS (retentate); D: dialyzedmaterial (dialysate), reconstitutedinto the original volume of water; R+D: dialysate reconstituted into the originalvolume of retentate.

decreased sensitivity to histatin 5 may be due to a diminishedenergy content of yeast cells grown on saliva as sole source ofnutrients. Previous studies have revealed that under low-energyconditions, C. albicans is virtually completely insensitive to his-tatin 5 (Veerman et al. 2007).

Figure 9. Candidacidal effects of histatin 5 towards C. albicans cultured in CHWS.Candida albicans cells were grown in CHWS or SDB for 24 h. Cells (2 × 107 ml–1)

were incubated with PPB supplemented with 5 mM NaCl or NaN3. After 30 min,PI was added and cells were transferred to wells containing equal volumes ofserially diluted histatin 5 in the corresponding buffer. Development of fluores-cence was monitored for each concentration of the peptides at 5 min intervals

at λexc 485 nm and λem 620 nm. The figure shows the fluorescence, expressed inarbitrary units (AU), after 1 h of incubation. Values are mean of duplicate mea-surements ± standard error.

DISCUSSION

In previous studies, it was found that supplementation withrelatively high amounts of glucose (20 mM) turns saliva intoa suitable growth medium for C. albicans (Samaranayake et al.1994; Lenanderlumikari and Johansson 1995). The present study,however, demonstrates that saliva by itself provides sufficientnutrients for the long-term survival of C. albicans in the oralcavity. The availability of nutrients in saliva seems to be agrowth-limiting factor, but evenwhen salivawas depleted of nu-trients, C. albicans survived for more than 400 h. In vivo salivais continuously replenished, ensuring a constant availability ofnutrients for growth and survival of C. albicans.

Our data suggest that C. albicans, in contrast to oral bacte-ria (Van der Hoeven and Camps 1991; Wickstrom and Svensater2008; Wickstrom et al. 2009), is by itself not able to utilize thecarbohydrate moiety of salivary mucins as carbon source. (Figs 5and 6). It can be envisaged, however, that in vivo, where C. albi-cans is often found in polymicrobial biofilmswith bacteria (Rafay,Homer and Beighton 1996; Harriott and Noverr 2011), it can takeadvantage of glycosidases released by bacteria.

It has been shown that C. albicans can use host proteinsas source of nitrogen by secreting aspartic proteinases (Zaugget al. 2001), which degrade proteins in peptides that are takenup into the cell by specific transporters (Morschhaeuser 2011;Dunkel et al. 2013, 2014). In the present study, however, wedid not find evidence for substantial degradation of salivaryproteins by C. albicans. Rather it seems that growth substratesfor C. albicans reside in the low-molecular-mass, dialyzablefraction of saliva (Fig. 7). This comprises a number of poten-tial low-molecular carbon and nitrogen sources for C. albicans,including glucose (Gough et al. 1996; Andersson et al. 1998;Luke et al. 1999), lactate (Ene et al. 2013), amino acids Brandet al. (1997), oligopeptides (Payne, Barrettbee and Shallow 1991;Dunkel et al. 2013) and ammonium ions (Morschhaeuser 2011;Navarathna et al. 2012; Dunkel et al. 2014). Although salivary



glucose was decreased after growth of C. albicans, several ob-servations suggest that it is not the only growth limiting fac-tor in saliva: (i) supplementation of spent saliva with physio-logical amounts of glucose did not restore growth; and (ii) therewas not a clear correlation between the glucose level in differ-ent glandular secretions and their growth supporting abilities.This indicates that besides glucose other presumably nitrogen-containing nutrients determine the growth and survival of C.albicans in CHWS. A preferred nitrogen source of C. albicansthat is present in sufficient amounts in saliva is ammoniumions (Oliver, Silver and White 2008). The ammonium concentra-tion of saliva varies between 2 and 5 mM (Huizenga and Gips1982), which is well in the range required for growth of C. albi-cans (Vidotto et al. 1991). However, heat treatment and freeze-drying, treatments by which volatile compounds such as am-monia are removed from saliva, had no discernible effect onthe growth of C. albicans in either saliva or the reconstituteddialysate. The MALDI-MS analysis suggests as possible nutri-ents oligopeptides, which are either secreted directly in salivaor originate from breakdown of proteins by bacterial and hostproteases.

The current finding that C. albicans could survive in salivafor more than 400 h is in line with previous observations thatsaliva promotes survival and proliferation of Candida species intap water (Barbot et al. 2011; Gerhardts et al. 2012). This is re-markable as saliva is generally considered a hostile environ-ment for C. albicans because of the presence of antimicrobialproteins, including histatin 5. In the present study, we found,however, that growth in saliva reduced the susceptibility to his-tatin 5 more than 10-fold. This is in line with a previous studyreporting that C. albicans grown under nitrogen-limiting condi-tions (with BSA as sole nitrogen source) was 4- to 16-fold lesssusceptible to amphotericin B and nystatin than when grown ina preferred nitrogen source (Oliver, Silver and White 2008). Theunderlying mechanism for the increased resistance of Candidathat has been cultured in saliva is not known, but we envisagethat the energy charge of Candida is diminished under these con-ditions. It has been shown that under low energetic conditionsCandida becomes resistant against a variety of membrane-activeantimicrobial peptides, including histatin 5 (see e.g. Koshlukovaet al. 1999; Veerman et al. 2007). Together with the fact thatthe candidacidal activity of whole saliva is partially maskedby calcium ions, this may provide an explanation for the abil-ity of C. albicans to colonize the oral tissues, which may evenbe enhanced under conditions when the oral immunity iscompromised.

Conflict of interest. None declared.

REFERENCES

Andersson AB, Birkhed D, Berntorp K, et al. Glucose concentra-tion in parotid saliva after glucose/food intake in individualswith glucose intolerance and diabetes mellitus. Eur J Oral Sci1998;106:931–7.

Barbot V, Migeot V, Rodier MH, et al. Saliva promotes survivaland even proliferation of Candida species in tap water. FEMSMicrobiol Lett 2011;324:17–20.

Brand HS, Jorning GG, Chamuleau RA, et al. Effect of a protein-richmeal on urinary and salivary free amino acid concentra-tions in human subjects. Clin Chim Acta 1997;264:37–47.

De Jong MH, Van der Hoeven JS. The growth of oral bacteria onsaliva. J Dent Res 1987;66:498–505.

Dunkel N, Biswas K, Hiller E, et al. Control of morphogenesis,protease secretion and gene expression in Candida albicansby the preferred nitrogen source ammonium. Microbiology2014;160:1599–608.

Dunkel N, Hertlein T, Franz R, et al. Roles of different peptidetransporters in nutrient acquisition in Candida albicans. Eu-karyot Cell 2013;12:520–8.

Ene IV, Cheng SC, NeteaMG, et al.Growth ofCandida albicans cellson the physiologically relevant carbon source lactate affectstheir recognition and phagocytosis by immune cells. InfectImmun 2013;81:238–48.

Gerhardts A, Hammer TR, Balluff C, et al. A model of thetransmission of micro-organisms in a public setting andits correlation to pathogen infection risks. J Appl Microbiol2012;112:614–21.

GoughH, LukeGA, Beeley JA, et al.Human salivary glucose analy-sis by high-performance ion-exchange chromatography andpulsed amperometric detection. Arch Oral Biol 1996;41:141–5.

Harriott MH, Noverr MC. Importance of Candida-bacterialpolymicrobial biofilms in disease. Trends Microbiol2011;19:557–63.

Huizenga JR, Gips CH. Determination of ammonia in saliva us-ing indophenol, an ammonium electrode and an enzymaticmethod – a comparative investigation. J Clin Chem Clin Bio1982;20:571–4.

Koshlukova SE, Lloyd TL, Araujo MWB, et al. Salivary histatin 5induces non-lytic release of ATP from Candida albicans lead-ing to cell death. J Biol Chem 1999;274:18872–9.

Lenanderlumikari M, Johansson I. Effect of saliva compositionon growth ofCandida-albicans and Torulopsis glabrata.OralMi-crobiol Immun 1995;10:233–40.

Luke GA, Gough H, Beeley JA, et al. Human salivary sugar clear-ance after sugar rinses and intake of foodstuffs. Caries Res1999;33:123–9.

Morschhaeuser J. Nitrogen regulation of morphogenesis andprotease secretion in Candida albicans. Int J Med Microbiol2011;301:390–4.

Navarathna DHMLP, Lionakis MS, Lizak MJ, et al. Urea ami-dolyase (DUR1,2) contributes to virulence and kidney patho-genesis of Candida albicans. PLos One 2012;7:e48475.

Navazesh M, Mulligan RA, Kipnis V, et al. Comparison of wholesaliva flow rates and mucin concentration in healthy Cau-casian young and aged adults. J Dent Res 1992;71:1275–8.

NieuwAmerongen AV, Strooker H, Henskens YM, et al. Biochem-ical changes in (glyco)proteins of saliva of epileptic patients.Ann NY Acad Sci 1993;694:283–5.

Nieuw Amerongen AV, Veerman ECI. Saliva–the defender of theoral cavity. Oral Dis 2002;8:12–22.

Oliver BG, Silver PM, White TC. Polyene susceptibility is depen-dent on nitrogen source in the opportunistic pathogen Can-dida albicans. J Antimicrob Chemoth 2008;61:1302–8.

Payne JW, Barrettbee KJ, Shallow DA. Peptide-substrates rapidlymodulate expression of dipeptide and oligopeptide per-meases in Candida albicans. FEMS Microbiol Lett 1991;79:15–20.

Rafay AM, Homer KA, Beighton D. Effect of mucin and glucoseon proteolytic and glycosidic activities of Streptococcus oralis.J Med Microbiol 1996;44:409–17.

Samaranayake LP, Leung WK, Jin LJ. Oral mucosal fungal infec-tions. Periodontol 2000 2009;49:39–59.

Samaranayake YH, Macfarlane TW, Samaranayake LP, et al.The in-vitro proteolytic and saccharolytic activity of Can-dida species cultured in human saliva. Oral Microbiol Immun1994;9:229–35.



Shelburne SA, Granville C, Tokuyama M, et al. Growth char-acteristics of and virulence factor production by group AStreptococcus during cultivation in human saliva. Infect Im-mun 2005;73:4723–31.

Van ‘t Hof W, Veerman EC, Nieuw Amerongen AV, et al. Antimi-crobial defense systems in saliva. Monogr Oral Sci 2014;24:40–51.

Van der Hoeven JS, Camp PJM. Synergistic degradation of mucinby Streptococcus-oralis and Streptococcus sanguis in mixedchemostat cultures. J Dent Res 1991;70:1041–4.

Veerman ECI, Bolscher JGM, Appelmelk BJ, et al. A monoclonalantibody directed against high M(r) salivary mucins recog-nizes the SO3–3Gal beta 1–3GlcNAc moiety of sulfo-Lewis(a):a histochemical survey of human and rat tissue. Glycobiology1997;7:37–43.

Veerman ECI, Valentijn-Benz M, Nazmi K, et al. Energy de-pletion protects candida albicans against antimicrobial pep-tides by rigidifying its cell membrane. J Biol Chem 2007;282:18831–41.

Veerman ECI, Valentijn-Benz M, Van den Keijbus PAM, et al.Immunochemical analysis of high-molecular-weight human

salivary mucins (MG1) using monoclonal antibodies. ArchOral Biol 1991;36:923–32.

Veerman ECI, van den Keijbus PAM, Vissink A, et al.Human glan-dular salivas: Their separate collection and analysis. Eur JOral Sci 1996;104:346–52.

Vidotto V, Ochoa LG, Cortes JM, et al. Optimal concentrationof ammonium ion in a minimal synthetic medium for thegrowth of Candida albicans. Mycopathologia 1991;113:139–42.

Wickstrom C, Hamilton IR, Svensater G. Differential metabolicactivity by dental plaque bacteria in association with twopreparations of MUC5B mucins in solution and in biofilms.Microbiol-SGM 2009;155:53–60.

Wickstrom C, Herzberg MC, Beighton D, et al. Proteolyticdegradation of human salivary MUC5B by dental biofilms.Microbiol-SGM 2009;155:2866–72.

Wickstrom C, Svensater G. Salivary gel-forming mucin MUC5B –a nutrient for dental plaque bacteria. Oral Microbiol Immun2008;23:177–82.

Zaugg C, Borg-von Zepelin M, Reichard U, et al. Secreted as-partic proteinase family of Candida tropicalis. Infect Immun2001;69:405–12.


Documents

Growth of Candida albicans in human saliva is supported by low