Metabolic profile of Candida albicans and Candida parapsilosis interactions within dual-species biofilms

Abstract

Within the oral cavity, the ability of Candida species to adhere and form biofilms is well-recognized, especially when Candida albicans is considered. Lately, a knowledge gap has been identified regarding dual-species communication of Candida isolates, as a way to increase virulence, with evidences being collected to support the existence of interactions between C. albicans and Candida parapsilosis. The present work evaluated the synergistic effect of the two Candida species, and explored chemical interactions between cells, evaluating secreted extracellular alcohols and their relation with yeasts' growth and matrix composition. A total of four clinical strains of C. albicans and C. parapsilosis species, isolated from single infections of different patients or from co-infections of a same patient, were tested. It was found that dual-species biofilms negatively impacted the growth of C. parapsilosis and their biofilm matrix, in comparison with mono-species biofilms, and had minor effects on the biofilm biomass. Alcohol secretion revealed to be species- and strain-dependent. However, some dual-species cultures produced much higher amounts of some alcohols (E-nerolidol and E, E-Farnesol) than the respective single cultures, which proves the existence of a synergy between species. These results show evidence that interactions between Candida species affect the biofilm matrix, which is a key element of oral biofilms.

biofilm matrix proteins/carbohydrates, biofilm monoculture/co-culture, Candida spp, extracellular alcohols, oral cavity, single-/dual-species biofilm

Introduction

The pioneer oral mycobiome investigations (Ghannoum et al. 2010, Mukherjee et al. 2014) revealed that when microbial and interpatient variability is considered, Candida is the number one prevalent genus (present commensally in numbers up to 70% of healthy individuals). Candida albicans was identified in healthy and immunocompromised individuals, even though a change in the fungal taxonomic profile of both populations was registered, attributed to interspecies interactions. Candida albicans infections have some predisposing factors associated, such as immunosuppressive therapy, antibiotic therapy, human immunodeficiency virus infection, diabetes, and old age (Redding et al. 1999, Fidel 2006, Perlroth et al. 2007). Among the pathogenic non-Candida albicans species, Candida parapsilosis was commonly isolated and co-isolated with C. albicans from both healthy and immunocompromised individuals (Bonassoli et al. 2005, Ghannoum et al. 2010, Ahuja et al. 2019, da Silva et al. 2020, Wu et al. 2021). In our previous study (Gerós-Mesquita et al. 2020), we analyzed 181 healthy individuals, including 29 couples, regarding oral yeast colonization using a culture-based approach, and results showed that 39% of the individuals were yeast carriers, 89% of them being colonized with C. albicans, and 3% with C. parapsilosis. Indeed, these results highlight the need to further investigate the interactions between different members of the mycobiome to gain further insights into their pathogenicity, and to understand the appearance of an increasing number of Candida strains resistant to conventional treatments.

One of the challenges in fundamental research is to identify the components that mediate interactions between Candida species and to understand their effect in the regulation of virulence factors, together with the search for molecules that act to inhibit biofilm formation (Ramage et al. 2002, Redding et al. 2009, Vera-González and Shukla 2020, Costa et al. 2021). Within the oral cavity, the ability of Candida species to adhere and form biofilms on tooth-associated and soft-tissue sites is well-recognized, especially considering C. albicans (Williams et al. 2011). In these communities, the cells are enmeshed in a matrix of exopolymeric substances, mainly proteins and carbohydrates, but also lipids and nucleic acids. As suggested for C. albicans, the biofilm matrix has metabolic and antifungal protective roles (Zarnowski et al. 2014). Accordingly, this lifeform may facilitate microbial survival and offer an advantage in growth in the presence of an adverse oral environmental of the host (Williams et al. 2011) and enable the physical and chemical interaction with other species. In the last years, some evidences have been collected to support the existence of interactions between C. albicans and C. parapsilosis:(i) the differential adhesion and extracellular enzymes activity of co-cultures vs. monocultures, although in a strain dependent fashion (Seabra et al. 2013); (ii) adherence capability of C. parapsilosis and C. albicans within the biofilms and their synergistic growth (El-Azizi et al. 2004, Kovács et al. 2019); (iii) the C. parapsilosis pheromone-induction ofC. albicans biofilm formation (Alby and Bennett 2011). Furthermore, chemical signaling may also occur through diffusible molecules, as demonstrated for Candida–bacteria interactions (Mallick and Bennett 2013). In this case, dimorphism occurring in C. albicans is controlled by quorum sensing, a mechanism of microbial communication (Mallick and Bennett 2013, Padder et al. 2018, Costa et al. 2021). Fungal quorum sensing molecules are released in culture media, and by achieving a certain threshold, they regulate the genes involved in virulence and in biofilm formation (Hornby et al. 2001, Ramage et al. 2002, Martins et al. 2007, 2010a, 2012). Our previous work (Costa et al. 2020) presented an in-depth mapping of the cellular metabolites of C. albicans and C. non-albicans (C. glabrata and C. tropicalis), identifying 126 putatively metabolites, which are produced by this species during growth. However, there is a knowledge gap on how the production of these molecules by Candida species adapt in the presence of other Candida species.

The main objective of the present work was to evaluate the synergistic effect of C. albicans and C. parapsilosis biofilm physiology, in terms of growth and matrix composition, but also to explore the chemical interactions between cells through the evaluation of secreted extracellular alcohols.

Materials and methods

Yeast strains and growth conditions

A total of four clinical isolates were used in this study, as described in Table 1, obtained previously from patients using an oral prosthesis (Seabra et al. 2013). A total of two species of Candida were considered—C. albicans and C. parapsilosis—, and two different strains of each species were tested, being one recovered from single infections of different patients, and another from co-infections of a same patient (Martins et al. 2010b). Yeasts were subcultured for 24 h in Sabouraud dextrose agar (Liofilchem, Italy) at 37°C. Cells were then cultured in Sabouraud dextrose broth (Liofilchem) for 18 h (37°C, 120 r/m), washed with phosphate saline buffer (PBS; 0.1M, pH 7), and quantified using hemocytometer counts. Cellular pellets were suspended in artificial saliva [per liter: 2 g yeast extract (Liofilchem), 5 g peptone (Liofilchem), 2 glucose (Applichem, Germany), 1 g mucin (Sigma-Aldrich), 0.35 g NaCl (Applichem), 0.2 g CaCl₂ (Riedel-de-Häen, Germany), and 0.2 g KCl (Pronalab, Portugal)].

Table 1.

Yeast species and strains used in this study.

Yeast species	Strain code	Type of infection	Mono-culture experiments	Dual-culture experiments
C. albicans (Ca)	AC	Single	CaAC	CaAC | CpAD
	AM	Co-infection	CaAM	CaAC | CpAM2
C. parapsilosis (Cp)	AD	Single	CpAD	CaAM | CpAD
	AM2	Co-infection	CpAM2	CaAM | CpAM2

Biofilm formation and evaluation

Biofilms were formed onto wells precoated with artificial saliva for 4 h (37°C, 120 r/m). The wells were then washed and stored in PBS at 4°C until needed, for a maximum of a week. For single species biofilms, cell concentration was 4 × 10⁷ cells/well (total volume of 2 ml), whereas for dual species biofilms, 2 × 10⁷ cells/well of each species were used (final 4 × 10⁷ Candida cells/well; total volume of 2 ml). Candida strain pairs included: CaAC|CpAD, CaAM|CpAM2, CaAC|CpAM2, and CaAM|CpAD, as summarized in Table 1. Biofilms were formed during 24 h, at 37°C and 120 r/m. After, the supernatants were pooled, filtered (0.22 μm pore size) and stored at −80°C until further use. The biofilm's adherent cells were scrapped, slightly dispersed by sonication (45 s at 30 W, Ultrasonic Processor, Cole-Parmer), vortexed (30 s), and plated in CHROMagar Candida medium (CHROMagarTM Candida, France) to determine the number of colony forming units (CFU). The total biofilm biomass production was assessed through crystal violet staining; the biofilm matrix was extracted by sonication and the corresponding protein and carbohydrate concentration assessed by the bicinchoninic acid and Dubois methods, respectively, as previously described by our group (Silva et al. 2009). The amount of biofilm matrix's proteins or carbohydrates was normalized per the total number of CFU in the biofilm [(μg/CFU) × 10⁶].

Analyses of extracellular alcohols

Extracellular alcohols were quantified in the biofilm supernatants and in planktonic cell cultures, considering the presence of isoamyl alcohol ( = 3-methyl-1-butanol), 2-phenylethanol, 1-dodecanol, E-nerolidol, and E, E-farnesol. The five alcohols were chosen according to our previous results (Costa et al. 2020), one of the most detailed studies on the volatile composition of Candida species, and also due to previous reports about their presence and importance in C. albicans. Alcohols’ content was determined by headspace solid-phase microextraction (HS-SPME) combined with comprehensive 2D gas chromatography–time-of-flight mass spectrometry, with time-of-flight analyzer (GC × GC–ToFMS). The SPME device included a fused silica fiber, coating partially cross-linked with 50/30 μm divinylbenzene/carboxen/poly (dimethylsiloxane) StableFlex™ fiber, with 1 cm of length (DVB/CAR/PDMS; Supelco, Aldrich, Bellefonte, PA). Each sample (8.3 ml of artificial saliva or supernatant) was inserted into a 25-ml glass vial with 1.7 g NaCl and a stirring bar (ratio of the volume of the liquid phase to the headspace volume (1/β) of 0.5). The capped vial was placed in a thermostatted bath, adjusted to 50.0 ± 0.1°C for 30 min, to promote the transference of the compounds to the SPME fiber. After that, the SPME coating fiber was manually introduced into the GC × GC-ToFMS (LECO Pegasus 4D; LECO, St. Joseph, MI) injection port at 250°C and kept for 30 s for desorption. The equipment setup and operation conditions were followed as previously described (Salvador et al. 2013). The mass spectrometer was operated in the EI mode at 70 eV using a range of m/z 33–350 and the detector voltage was −1741 V. The identification of the alcohols was performed by comparison of their mass spectrum with that of pure standards and by their retention index. The Deconvoluted Total Ion Current GC × GC area data were used as an approach to estimate the relative content of each alcohol. The peak areas were corrected against those of artificial saliva and were normalized per the total number of CFU in the corresponding biofilm (peak area/ CFU).

Statistical analyses and multivariate projection

Biofilm growth and matrix analyses were performed in triplicate and repeated in three independent assays. Extracellular alcohols analyses were performed in duplicate and repeated in two independent assays. A conservative approach was taken and only the alcohols consistently detected were statistically examined.

Data were statistically analyzed using GraphPad Prism, version 6.0 (GraphPad Software, San Diego, CA). The outliers were determined using the ROUT method and the normality was evaluated by the D'Agostino-Pearson omnibus test. A Kruskal–Wallis test was applied followed by Dunn's test for multiple comparisons between each Candida species in mono and co-culture. P-values lower than .05 were considered as statistically significant.

Data visualization was performed with Orange data mining suite [v. 3.24.1; (Demšar et al. 2004)], using scatterplot and FreeViz (Demšar et al. 2007) tools.

Results

Planktonic vs. biofilm Candida cultures

In this study, four different Candida strains were used, isolated from patients’ oral cavities (Table 1), with the aim to analyze C. albicans and C. parapsilosis interactions within biofilms. First, in order to evaluate if alcohols produced by mono- and dual-species cultures was dependent on how cells have been grown, we assessed the production of these compounds in planktonic and biofilm Candida cultures. Interesting, noticeable differences were found when comparing both types of cell culture (Table 2). In general, higher alcohol production was detected in planktonic cells, in comparison with biofilms. Statistically significant differences were obtained only between planktonic forms of C. parapsilosis for the production of E, E-farnesol, isoamyl alcohol ( = 3-methyl-1-butanol), and 2-phenylethanol. The opposite was detected for C. albicans for which the differences between strains (for these three alcohols) were only detected in biofilm cultures. Candida albicans cells when growing in biofilms showed a much lower production of all the detected alcohols, even though the production of almost all of them was significantly higher than the one of C. parapsilosis. When comparing single- vs. dual-species cultures, one can conclude that planktonic yeasts showed lower production of isoamyl alcohol and 2-phenylethanol in co-cultures (statistically significant at P < .05). Results of biofilm yeasts suggest that the co-culture production of these alcohols was lower in comparison with the C. albicans single species growth, but was significantly higher than the isolated production by C. parapsilosis species.

Table 2.

Alcohols secretion by Candida species planktonic and biofilm supernatant fraction [(a.u./CFUs) × 10⁵] in mono-species and dual-species culture.

		CaAC	CaAM	CaAC|CpAD	CaAM|CpAM2	CaAM|CpAD	CaAC|CpAM2	CpAD	CpAM2
Planktonic	E,E-farnesol	34.95 Δ	39.32	20.75	17.84	30.46	13.39	ND*Δ	21.92*Δ
	isoamyl alcohol	46.24Δ	45.27	25.54 ◼◇	22.98 ◼◇	40.61 ◼◇	28.12 ◼ ◇	51.07*Δ	32.51*Δ
	2-phenylethanol	45.90 Δ	45.11	24.74 ◼◇	22.60 ◼ ◇	40.63 ◼ ◇	27.77 ◼◇	49.64 *Δ	30.52 *Δ
	E-nerolidol	ND	ND	9.19	ND	ND	ND	ND	ND
	1-dodecanol	ND	18.81	10.53	ND	16.74	ND	ND	21.37
Biofilm	E,E-farnesol	4.91 *Δ	ND *	0.92 ◇	1.13 ◼	1.23	5.20 ◼◇	NDΔ	NDΔ
	isoamyl alcohol	6.30* Δ	5.35*	1.33◼◇	1.71 ◼◇	3.55◼◇	7.98 ◼◇	1.07 Δ	1.00 Δ
	2-phenylethanol	6.76*Δ	5.65*	1.38 ◼◇	1.75 ◼ ◇	3.77 ◼◇	8.31 ◼ ◇	1.10 Δ	1.05 Δ
	E-nerolidol	2.12	3.73	0.57 ◇	0.63◼	ND	5.98 ◼◇	ND	ND
	1-dodecanol	ND*Δ	5.15*	1.12 ◼◇	1.60 ◇	ND◼◇	6.03 ◇	1.03 Δ	0.40

*Strains of the same species that are significantly different.

ΔStatistically differences between C. albicans AC and C. parapsilosis strains.

Statistically difference between C. albicans AM and C. parapsilosis strains.

◼

Statistical differences between co-cultures of different C. albicans strains.

◇

Statistical differences between co-cultures of different C. parapsilosis strains.

Statistical differences between co-cultures with both C. albicans strains and C. parapsilosis strains different (P< .05).

To verify which compounds are responsible for the larger differences observed between planktonic and biofilm cultures, a multivariate projection using the FreeViz algorithm was performed (Fig. 1). Each single- and dual-species inoculation was projected in a 2D surface in respect to the concentration of the five alcohols—isoamyl alcohol, 2-phenylethanol, 1-dodecanol, E-nerolidol, and E, E-farnesol. Results of the algorithm revealed a complete separation between biofilm and planktonic cells, corroborating the statistical differences analyzed before. Cultures stratification is explained by the contribution of all the tested alcohols, but mainly by the concentrations of E, E-farnesol, isoamyl alcohol, and 2-phenylethanol, as discussed. Higher variability (represented by higher dispersion in the FreeViz surface) was observed for planktonic cells, in relation to biofilm quantifications. In the later, only a slight separation of CaAC|CpAM2, CaAM, and CaAC from the remaining group was observed, which was mainly caused by a combination of a higher production of E-nerolidol with a lower production of 1-dodecanol.

Figure 1.

FreeViz plots of multivariate projections using concentration of alcohols secreted by Candida species, identified in planktonic (dark color) and biofilm (light color) supernatant fraction, both considering monoculture (CaAC, CaAM, CpAD, and CpAM2) and co-cultures (CaAC|CpAD, CaAC|CpAM2, CaAM|CpAD, andCaAM|CpAM2). The direction and size of each vector indicates the relative prevalence of that feature to explain group stratification. Background color is an intensity gradient related to cluster positioning and size.

Biofilm growth and matrix composition

With the aim of testing the hypothesis of a synergy established during dual-species colonization, including an increase in the virulence, C. albicans and C. parapsilosis biofilm growth and matrix composition were compared, in monocultures and co-cultures. Results are summarized in Fig. 2. In general, the co-culture of C. albicans and C. parapsilosis did not have a major effect on the biofilm cell growth of C. albicans, with the exception of CaAC|CpAM2 co-culture (statistically significant at P< .05; Fig. 2A, left panel). However, there was a reduction in the number of C. parapsilosis culturable cells (statistically significant at P < .05), in comparison with the respective monocultures, in all combinations tested (Fig. 2A, right panel). Nevertheless, this did not result in major changes of the overall biofilm biomass of dual-species biofilms, with the exception of the biofilm formed by the pair C. albicans AC and C. parapsilosis AD (Fig. 2B), where a synergistic effect on biofilm biomass was observed (P< .05). Interestingly, dual-species biofilms had a suppressive effect on the production of the matrix proteins for three out of the four strains studied (CaAC, CaAM, and CpAD; Fig. 2C). On the contrary, these dual-species biofilms had a null or moderate suppressive impact on the production of the matrix carbohydrates, that was Candida pair dependent (Fig. 2D), when compared with single-species biofilms.

Figure 2.

Candida albicans (Ca) and C. parapsilosis (Cp) biofilm growth and matrix composition in monoculture (CaAC, CaAM, CpAD, andCpAM2) and co-culture (CaAC + CpAD, CaAC + CpAM2, CaAM + CpAD, andCaAM + CpAM2). The number of culturable cells of each species within the biofilm (A), total biofilm biomass (B), the crude biofilm matrix protein (C), and carbohydrate (D) content were determined. The data represents mean ± standard error of the mean. The full line denotes statistically significant differences between C. albicans in monoculture vs. in co-culture; the dashed line denotes statistically significant differences between C. parapsilosis in mono vs. in co-cultures; the asterisk highlights a mean value below the method quantification limit.

However, with the dual-species pair CaAM|CpAM2, an increase in the quantified carbohydrates was observed, in relation to the single-species production and in opposition to the remaining yeast pairs (statistically insignificant).

Metabolic profile: analytical determination of extracellular alcohols

Bioanalytical analysis (GC × GC–ToFMS after HS-SPME extraction) was accomplished with biofilm supernatants, to identify extracellular alcohols, in particular isoamyl alcohol, 2-phenylethanol, 1-dodecanol, E-nerolidol, and E, E-farnesol (Table 3). Generally, single species derived biofilms led to inconsistencies, such as the detection of the 1-dodecanol, E-nerolidol and E, E-farnesol only in some replicates (highlighted with *), or even the absence of detection of these compounds in all replicates (emphasized with ^‡). In contrast, in the samples recovered from the co-culture of both Candida species, alcohols were consistently identified. Notably, E, E-farnesol was found in the supernatants of all the Candida pairs evaluated.

Table 3.

Candida albicans (Ca) and C. parapsilosis (Cp) biofilm monoculture (CaAC, CaAM, CpAD, and CpAM2) and co-culture (CaAC|CpAD, CaAC|CpAM2, CaAM|CpAD, and CaAM|CpAM2) derived extracellular alcohols identified in culture supernatants by GC × GC–ToFMS and expressed as mean peak area (10^–4)/CFU. Standard deviations are detailed after the mean peak area values.

Alcohols	CaAC	CaAM	CaAC|CpAD	CaAM|CpAM2	CaAM|CpAD	CaAC|CpAM2	CpAD	CpAM2
Isoamyl alcohola (3-methyl-1-butanol)	318.4 (33.2)	457.0 (30.0)	305.8 (73.5)	250.5 (52.0)	230.7 (29.0)	350.6 (40.0)	259.2 (24.8)	220.0 (31.0)
2-Phenylethanolb	1381.9 (119.9)	1393.6 (222.3)	604.7 (91.4)	383.8 (103.5)	722.2 (193.4)	1150.4 (192.1)	482.9 (117.4)	495.0 (58.5)
1-Dodecanolc	0‡	387.0 (223.6)*	59.2 (49.3)	139.6 (67.6)	0‡	0.01 (0.01)*	346.6 (203.3)*	3.4 (2.0)*
E-Nerolidold	0.6 (0.4)*	1.0 (0.8)*	0.5 (0.4)	1.0 (0.6)*	0.1 (0.1)*	4.3 (2.6)*	0.3 (0.2)*	0‡
E,E-Farnesole	4.7 (2.7)	0*	2.1 (1.0)	0.7 (0.2)	0.5 (0.3)	13.7 (7.7)	0.4 (0.2)*	0‡

a

Retention time for first (¹t_R)/second (²t_R) dimensions = 140/1.280, Retention Index reported in the literature (RI_lit)/Retention Index obtained through the modulated chromatogram (RI_calc) = 720 [52] /734;

b1

t_R/²t_R = 465/3.680, RI_lit/RI_calc = 1136 [52] /1110;

c1

t_R/²t_R = 750/1.180, RI_lit/RI_calc = 1476 [53] /1466;

d1

t_R/²t_R = 825/1.150, RI_lit/RI_calc = 1570 [52] /1564;

e1

t_R/²t_R = 960/1.600, RI_lit/RI_calc = 1733 [52] /1722;

‡

the compound was not detected;

*

the compound was detected in one assay.

In general, the dual-species biofilm cultures did not affect the secretion of isoamyl alcohol and 2-phenylethanol, in comparison to single species cultures (Table 3). Nevertheless, a strain dependent impact in the secretion of specific alcohols was observed: (i)C. albicans AM co-culture with C. parapsilosis AD or AM2 resulted in a decrease in the overall secretion of isoamyl alcohol (in comparison with C. albicans AM, P = .03) or 2-phenylethanol (in comparison to C. albicans AM, P = .04); (ii) the production of 1-dodecanol was largely higher in the co-culture of C. albicans CA with C. parapsilosis AD, in comparison with single culture C. albicans AC, in which no production was detected, or with the co-culture of this strain with C. parapsilosis AM2, in which a residual detection was observed; (iii)E-nerolidol and E, E-Farnesol production increased almost four times in the co-culture of CaAC|CpAM2, in comparison with the amounts quantified in the respective single cultures.

To validate if the observed differences are significant enough to clearly discriminate between dual- vs. single-species, a multivariate projection using the FreeViz algorithm was performed (Fig. 3), for the five quantified alcohols. With the exception of the single species inoculation CpAM2, the remaining three single inoculations were completely separated from the dual-species ones, mainly due to an increase in the concentrations of 1-Dodecanol, isoamyl alcohol or 2-phenylethanol, having these compounds an almost neutral effect in the positioning of the reaming yeast combinations. The majority of dual-species inoculations, showing lower concentrations of the mentioned compounds, showed neutral or increased concentrations of E, E-Farnesol and E-Nerolidol. This last set of results should be analyzed with caution due to the inconsistencies detected in single species biofilm quantification, as referred before.

Figure 3.

FreeViz plots of multivariate projections using extracellular alcohols concentration, identified in culture supernatants, considering C. albicans (Ca) and C. parapsilosis (Cp) biofilm monoculture (light color; CaAC, CaAM, CpAD, and CpAM2) and co-culture (dark color; CaAC|CpAD, CaAC|CpAM2, CaAM|CpAD, andCaAM|CpAM2). The direction and size of each vector indicates the relative prevalence of that feature to explain group stratification. Background color is an intensity gradient related to cluster positioning and size.

Discussion

The main goal of the present work was to shed light on the conditions that may favor the association of two different Candida species, namely C. albicans and C. parapsilosis, in infecting the oral cavity. In our previous study, it was demonstrated that single infection is the most common situation (Martins et al. 2010b), and so we hypothesized that a dual infection only occurs when a synergy is established, determining an increase in virulence. Virulence traits of four Candida clinical oral isolates were evaluated, considering two C. albicans strains (AM and AC) and two C. parapsilosis strains (AM2 and AD), with the particularity that C. albicans AM was co-isolated with C. parapsilosis AM2 (Table 1).

Differences were reported before between biofilm and planktonic cell cultures, mainly due to the fact that cells dispersed from a biofilm are usually the ones originating from its top layer. In detail, biofilm dispersed yeast cells have been shown to present enhanced virulence-associated features, extending also to transcriptomic alterations (Uppuluri et al. 2018). In the present work, mono- and dual-species cultures were compared, testing both biofilm derived cells and planktonic cultures. Great heterogeneity was observed in our work regarding the higher production of E, E-farnesol, isoamyl alcohol ( = 3-methyl-1-butanol) and 2-phenylethanol by C. parapsilosis, but only in planktonic cells, while for C. albicans the production of these alcohols varied only in biofilm cultures. The effect of growth form/mode of Candida spp., C. albicans, and C. parapsilosis (biofilm vs. planktonic) was already reported before (Pires et al. 2011, Zhou et al. 2012, Souza et al. 2016, Weerasekera et al. 2016, Uppuluri et al. 2018, Modiri et al. 2019, Lemos et al. 2020), which is line with our findings. Farnesol has been described as a quorum sensing molecule in C. albicans, which regulates cell morphogenesis, inhibits filamentation and, most importantly, hinders biofilm formation (Hornby et al. 2001, Ramage et al. 2002, Costa et al. 2021). This metabolite was already reported for C. tropicalis (Martins et al. 2010a). Other secreted alcohols have also been shown to act as signaling molecules in these yeasts, such as 1-dodecanol or E-nerolidol (Martins et al. 2007, 2010a, Rodrigues and Černáková 2020).

All Candida strains were able to form single biofilms on the abiotic surface, although in different extents, depending on the strain, species and culture counterparts (Fig. 2), as previously shown (Silva et al. 2009). Even though very similar in terms of total biofilm biomass (Fig. 2B), the number of viable cells of C. albicans was lower in comparison to C. parapsilosis (Fig. 2A and B). This result indicates that C. albicans strains produced higher amounts of extracellular matrix, which is in correlation with their significantly higher concentration of biofilm matrix proteins (Fig. 2C). The association of C. albicans AC with C. parapsilosis AD led to a significantly higher biofilm biomass, even though the matrix proteins and carbohydrates were much lower (significant at P < .05). In addition, the association of the two co-isolated species (C. albicans AM + C. parapsilosis AM2) displayed a lower number of viable cells (Fig. 2A), but an increase in total biomass (Fig. 2B), which is an indication of enhanced virulence. Additionally, it should be noted that C. albicans strains (AC and AM) seem to have an impact on the regulation of growth of C. parapsilosis (Fig. 2A). Even though, the secretion of alcohols by yeast in mono-species biofilms has been studied, there is a lack in research involving alcohol production in dual-species biofilms. In the current work, results show that the alcohol secretion profile in mono- or dual-species cultures is markedly different (Table 3 and Fig. 3). E, E-farnesol was detected on the supernatant of the C. albicans AM + C. parapsilosis AM2 biofilm, but it was not detected when each strain was cultured alone (Table 3; in CaAM this compound was only detected in one assay, as already referred). Some studies report that this alcohol may have an inhibitory effect on C. parapsilosis (Navarathna et al. 2007, Rossignol et al. 2007, Cordeiro et al. 2013), which can explain the reduced number of viable cells of C. parapsilosis when cultured with C. albicans strains, in comparison with the number of viable cells of these species cultured alone (Fig. 2). Differences on the secretion of 1-dodecanol were observed on dual-species cultures, particularly for the dual-species biofilm with C. albicans AC, which only produced this alcohol when cultured with C. parapsilosis strains.

The synergistic effect of C. albicans in C. parapsilosis growth and alcohol production is in accordance with our previously obtained results, in which the enzymatic activity of these strains in single and dual-species inoculations was shown (Seabra et al. 2013). In the cited work, for dual-species cultures, enzymatic activity of C. parapsilosis, in particular for hemolysin, phospholipase, and proteinase, was higher in the presence of C. albicans AC, than in the presence of C. albicans AM (isolated from the same patient). Results suggest that in the presence of C. albicans AM, the enzymatic activity of C. parapsilosis AM2 seems to be inhibited, corroborating the hypothesis that metabolites produced by C. albicans AM influence the expression of virulence factors in C. parapsilosis AM2. These results are correlated with the ones obtained in the present work, since the co-culture of CaAM with CpAM2 led to lower alcohol production than the one obtained with single-species cultures. The exceptions are the production of E-Nerolidol and E, E-Farnesol that were similar or slightly higher, as discussed. Interestingly, in comparison to mono-species biofilms, Seabra et al. (2013) also observed a higher enzymatic activity in dual-species biofilms containing C. albicans AC, producing a higher amount of E, E-Farnesol and a lower amount of the other alcohols. These results suggest a synergism between both species when in biofilms, since an increase of the concentration of one of the alcohols seem to be balanced by the decrease of other alcohols’ concentration.

From this work, it is possible to collect vast evidence about a synergy between Candida species that leads to an increased virulence, even though being strain-dependent. Due to the diversity of the compounds involved, the dynamic regulation and the interaction between species, it is difficult to determine the exact nature of this synergy and the mechanism that leads to its increased virulence. Although an important achievement was obtained to reckon and understand this dynamism. The results obtained in this work are a step forward to unravel how the virulence factors are stripped and enhanced in dual-species structures.

ACKNOWLEDGEMENTS

A especial thanks to my dearest Professor Rosário Oliveira, who was one of the mentors of this work, but unfortunately passed away.

Funding

This work was supported by the Portuguese Foundation for Science and Technology (FCT) under the scope of the strategic funding of UIDB/04469/2020 unit.

Conflicts of interest statement

None declared.

Author notes

Present address: LAQV/REQUIMTE, Departamento de Ciências Químicas, Faculdade de Farmácia, Universidade do Porto, Rua de Jorge Viterbo Ferreira, 228, 4050-313, Porto, Portugal.