Living in a biofilm is a natural and universal form of existence for a broad variety of microorganisms capable of infecting the human host. Bacterial and fungal pathogens that form multicellular communities might gain protection against unfavorable environmental conditions, applied antimicrobial therapies, and the host’s defense mechanisms (Nobile and Johnson, 2015). Microbial consortia also incessantly affect the host immune system and modulate its function and response, often becoming the cause of recurrent and persistent infections difficult to treat (Chandra et al., 2007; Pereira et al., 2021). One of the most widespread fungal pathogens capable of forming biofilm and causing a wide spectrum of different types of infections in humans are the opportunistic pathogenic yeasts from the Candida genus, including the most frequently identified species C. albicans (Netea et al., 2015; Tsui et al., 2016). However, other Candida species, including C. tropicalis, C. glabrata, and C. parapsilosis, are becoming an increasing epidemiological threat, predominantly due to the distinct collection of virulence mechanisms, as well as emerging resistance to antifungal drugs typically used in the treatment of candidiasis (Silva et al., 2012; Arendrup, 2013; Giacobbe et al., 2020; Zhang et al., 2020). Non-albicans Candida (NAC) species also possess the ability to produce biofilm; however, a considerable species- and strain-specific variety of biofilm production capacity for C. tropicalis, C. glabrata, and C. parapsilosis was observed, reflected in the differences in biofilm architecture, complexity, and composition of the biofilm matrix (Silva et al., 2011; Cavalheiro and Teixeira, 2018). Nevertheless, despite some interspecies diversity, generally, the ability to create a biofilm gives pathogenic Candida fungi an advantage in maintaining infection and enhancing virulence potential compared with planktonic cells (Cavalheiro and Teixeira, 2018). The adherence of microbial cells to biotic or abiotic surfaces, their coexistence in the community in high density, and embedding in a composite biofilm-supporting matrix not only might make the biofilm-forming cells different from free-floating ones but also contribute to acquiring novel virulence properties by the complex biofilm structure (Xu et al., 2022). These are mainly related to the increased difficulty in the removal of biofilm-forming pathogens from the inhabited surfaces and to the disturbances in the effective action of antimicrobial drugs (Ponde et al., 2021). In studies recently presented by Sasani et al., in 2021, C. tropicalis strains isolated from patients with candidemia and characterized by increased metabolic activity and high biofilm mass were related to higher mortality rates, and biofilms formed by these strains showed increased resistance to azole drugs (Sasani et al., 2021). Moreover, research presented previously by Tumbarello et al. (2007) demonstrated the correlation between the biofilm-forming ability of C. albicans and C. parapsilosis and the mortality of patients with candidemia caused by these species. Such reports indicate the urgent necessity to study in detail the characteristics of this sedentary form of existence of fungal pathogens, which constitute a real threat to human health and life.

In addition to various well-known mechanisms contributing to fungal virulence, the microbial cells that create biofilm produce also extracellular vesicles (EVs)—nanometric spherical structures surrounded by a lipid bilayer, carrying a diversified biologically active cargo—that are involved in intercellular communication, biofilm matrix production, and interaction with the host (Bleackley et al., 2019; Garcia-Ceron et al., 2021; Piffer et al., 2021). EVs produced by C. albicans biofilm differ from those produced by planktonic cells not only in the size and heterogeneity of the population but also in the protein composition. In studies presented by Zarnowski et al. (2018), 34% of the identified EV-related proteins were unique to the biofilm state, and several other shared proteins were also enriched in biofilm-derived EVs compared with vesicles from planktonic cells. Proteins carried by vesicles released by C. albicans biofilm might be involved in the production of the biofilm matrix through their enzymatic activity, and EVs might directly provide other matrix-building components like lipids and polysaccharides—glucan and mannan—thus contributing to the resistance of C. albicans biofilm to fluconazole by entrapping the antifungal drug within the matrix structure (Taff et al., 2012; Zarnowski et al., 2018; Zhao et al., 2021). Moreover, C. albicans biofilm-derived EVs play a role not only in biofilm matrix production and in reducing biofilm susceptibility to antifungals, but they are also involved in other stages of biofilm formation, like fungal cell adhesion and biofilm dispersion (Zarnowski et al., 2021). Hence, they contribute strongly and in multiple ways to the creation of this complex microbial community.

Candida albicans EVs have been characterized the most thoroughly so far regarding Candida fungi. For other species, the available data are incomplete. A detailed comparison has been recently performed by Zamith-Miranda et al. (2021) between EVs released by C. albicans and C. auris grown as planktonic cells, and the vesicle content was identified as significantly different for these two Candida species, including the varied abundance of 393 vesicular proteins when the two species were compared. Such interspecific variability of vesicles resulted also in diverse effects on the host cells (Zamith-Miranda et al., 2021). In our recent paper, we characterized the EVs produced by free-floating cells of three medically important NAC species—C. tropicalis, C. parapsilosis, and C. glabrata. However, given the possible dissimilarity of the EVs produced by the biofilm form, it is also particularly important to characterize the vesicles produced by the biofilm fungal community, especially that the abovementioned NAC species showing the ability to form complex sessile structures are currently the subject of growing interest and importance. Therefore, in the current research, we aimed to characterize the biofilm-derived EVs of C. tropicalis, C. parapsilosis, and C. glabrata along with the analysis of their immunoregulatory properties.

The fungal culture initiated by the defined number of Candida cells was further maintained in rotating culture bottles in RPMI 1640 medium to generate a complex 48-h biofilm, and then the vesicles were isolated from the concentrated post-culture supernatants. The presence of EVs in the obtained preparations was confirmed by TEM imaging, and the achieved pictures for C. glabrata ( Figure 1A ), C. parapsilosis ( Figure 1C ), and C. tropicalis ( Figure 1E ) indicated the presence of spherical structures of various sizes.

NTA that was performed to determine the number of particles in EV-containing samples showed that biofilms started by 1 × 10⁹ C. glabrata cells produced 7.73 ± 0.07 × 10⁹ particles (about 8 per cell) and those by 1 × 10⁹ C. parapsilosis cells produced 2.64 ± 0.37 × 10¹⁰ particles (about 26 per cell), while biofilms initiated by 1 × 10⁸ C. tropicalis cells produced 4.36 ± 0.1 × 10⁹ particles (about 44 per cell). The average particle sizes were also measured using NTA and the diameters of EVs produced by C. glabrata ( Figure 1B ) and C. parapsilosis ( Figure 1D ) ranged from around 80 to 180 nm for 90% of the total particle population, while for C. tropicalis ( Figure 1F ), these values were almost up to 245 nm.

The measured average concentrations of proteins, phospholipids, and carbohydrates per 1 × 10¹⁰ EVs for each Candida species are listed in Table 1 . The highest protein, lipid, and carbohydrate contents were observed in EVs produced by C. tropicalis biofilms. The lowest protein and lipid concentrations were identified in EVs from C. parapsilosis biofilms, and the lowest average carbohydrate content was indicated for C. glabrata biofilm-originated EVs.

Further characterization of fungal EVs included vesicular protein identification using liquid chromatography-coupled tandem mass spectrometry analysis preceded by EV lysis and protein hydrolysis with trypsin. There were 288 proteins identified from C. glabrata, 86 from C. parapsilosis, and 87 from C. tropicalis biofilms ( Supplementary Table 1 ). Among the proteins identified for C. glabrata EVs, an abundant group of proteins involved in cell wall organization might be distinguished, including among others the cell wall protein Scw11, the cell wall mannoprotein Cwp1, the 1,3-beta-glucan-linked cell wall protein Pir1, the GPI-anchored protein Ecm33, and glycoside hydrolase of the Gas/Phr family Gas1. For EVs from C. parapsilosis, this group of proteins was represented by the cell surface mannoprotein Mp65, the GPI-anchored cell wall protein Ecm33, the secreted yeast cell wall protein Ywp1, and 1,3-beta-glucosyltransferase Bgl2. Proteins involved in cell wall organization identified in C. tropicalis EVs were the cell surface mannoprotein Mp65, 1,3-beta-glucosidase Bgl2, GPI-anchored cell wall transglycosylase Crh11, and beta-1,3-glucanosyltransferase Pga4. There were some proteins identified in EVs which were shared between species ( Figure 2 ). Five proteins were shared among all the tested species: protein similar to alpha agglutinin anchor subunit Tos1, cell wall protein Scw11, protein from the Hsp70 chaperone family, 60S ribosomal protein L12, and glyceraldehyde-3-phosphate dehydrogenase Tdh3. Five proteins—ABC transporter, the GPI-anchored cell wall protein Ecm33, mitochondrial F-ATPase beta subunit Atp2, and two proteins from the SUR7 family—were common to C. glabrata and C. parapsilosis. Moreover, five proteins—cell surface mannoproteins Mp65, Ywp1, and Bgl2; autophagy-related protein 9 Atg9; and the adhesin-like protein Sim1—were common to C. parapsilosis and C. tropicalis. Four proteins shared by C. glabrata and C. tropicalis were enolase, alcohol dehydrogenase, elongation factor 2, and ATP synthase alpha subunit Atp1.

Proteins identified from EVs produced by C. glabrata, C. parapsilosis, and C. tropicalis biofilms were grouped, as previously described (Karkowska-Kuleta et al., 2020), into seven functional classes on the basis of protein descriptions and Gene Ontology (GO) annotations included in the Candida Genome Database (CGD, http://www.candidagenome.org) (Arnaud et al., 2005), UniProtKB Database (https://www.uniprot.org) (UniProt Consortium, 2019), and KEGG (Kyoto Encyclopedia of Genes and Genomes; https://www.genome.jp/kegg/) (Kanehisa and Goto, 2000): i) fungal-type cell wall organization, ii) membrane transport and organization, iii) stress response, iv) pathogenesis, v) cell metabolism, vi) transcription and translation, and vii) proteins of unknown function. Bar charts representing the percentage contents of proteins for each functional group are shown in Figure 3 . The largest group of proteins found in EVs was transcription and translation (21%) for C. glabrata and cell metabolism (22%) for C. parapsilosis and C. tropicalis (29%). The least represented functional group was pathogenesis (1%) for C. glabrata, transcription and translation (8%) for C. parapsilosis, and stress response for C. tropicalis (3%). Among the identified proteins, there were no proteins whose main function will indicate that they are involved in stress response and pathogenesis for C. parapsilosis.

Previous reports showed that vesicles produced by C. albicans free-floating cells exhibited immunoregulatory properties, inducing membrane expression of co-stimulatory molecules by the host’s immune cells and influencing the production of cytokines (Vargas et al., 2015; Vargas et al., 2020; Martínez-López et al., 2022). There is still a lack of studies on the impact of candidal biofilm-originated vesicles on the host response, and the influence of EVs released by C. glabrata, C. parapsilosis, and C. tropicalis cells existing in the form of sedentary fungal communities has not yet been investigated. In the current study, we examined the effect of these particles on cytokine production by cells from the human monocytic cell line THP-1 differentiated into macrophage-like cells. Analysis of the secretion of proinflammatory and anti-inflammatory cytokines by THP-1 cells stimulated with EVs of each Candida species showed a quite comparable tendency ( Figure 4 ). There has been a significant increase in the level of TNF-α production after contact of human cells with EVs released by the tested Candida fungi. The highest increase in the production of this cytokine, compared with the control, was observed after stimulation with EVs produced by C. glabrata biofilm, and less amount of cytokine was produced by the other two species. Similarly, the level of IL-8 secretion after incubation of human cells with candidal EVs was higher than the level for cells not stimulated with EVs, and the highest amount of cytokine was again produced under the influence of C. glabrata EVs. On the contrary, the decrease in the amount of cytokine produced was observed for IL-10 after incubation of THP-1 cells with EVs from each tested Candia species, when compared with non-stimulated cells.

In the case of C. albicans biofilm, it has been shown that vesicles produced by cells in the biofilm contribute to the formation of the biofilm matrix and the occurrence of fluconazole resistance (Zarnowski et al., 2018; Zarnowski et al., 2021). Therefore, in this study we examined whether EVs might influence formation of biofilm by C. tropicalis, which of the three tested Candida species produced the most stable biofilm and was susceptible to the applied antifungal drugs ‒ fluconazole (FLU) and caspofungin (CASP) ( Figure 5 ).

Antifungals were used at concentrations of 0.5 µg/ml for fluconazole and 0.01 µg/ml for caspofungin (Bizerra et al., 2008; Ferreira et al., 2009). In the first step, 10⁵ cells were added to the wells of a microplate and incubated for 90 min to allow them to adhere to the surface. After this time, the non-adherent cells were washed away and the adherent cells were further incubated with or without vesicles produced by the 48-h C. tropicalis biofilm. Additionally, after preincubation of the adherent cells with EVs for 90 min, the antifungal drugs fluconazole or caspofungin were added to particular wells, and biofilms were further formed for 24 h, all the time in the presence of EVs where they were originally introduced. Reference samples were biofilms formed without EVs added. To investigate the impact of EVs on the thickness of biofilm formed by C. tropicalis cells in the presence of antifungals, optical density measurements in the area scan mode (OD₆₀₀) were performed ( Figure 5A ). Biofilm metabolic activity and cell viability were analyzed with the XTT test ( Figure 5B ) and CFU counting ( Figure 5C ), respectively. It was indicated that in the presence of both antifungal drugs, the addition of EVs increased the thickness of the biofilm and the metabolic activity of the cells, as well as cell viability (only for caspofungin), when compared with samples where EVs were not added; however, in the absence of drugs, the addition of EVs did not significantly affect these parameters under the conditions tested.

Fungi from the Candida genus, including C. glabrata, C. parapsilosis, and C. tropicalis, are important commensal pathogens able to cause various types of infections in humans. Although currently these species are not as common cause of fungal diseases as C. albicans, they pose a growing threat to the immunocompromised patient population due to the everincreasing frequency of infections and emerging resistance to antifungals. During the colonization of different niches of the human organism, Candida yeasts exploit many various mechanisms of virulence, including the ability to form biofilm and the production of EVs (Yu et al., 1994; Zuza-Alves et al., 2017). EVs are nanometric structures surrounded by a lipid bilayer and filled with a plenteous and assorted cargo (Joffe et al., 2016). These structures have become more widely described in the last few decades primarily for human cells (Faure et al., 2006; Bobis-Wozowicz et al., 2017; You et al., 2020), but also for numerous bacteria (Knox et al., 1967; Lee et al., 2009; Rivera et al., 2010; Tashiro et al., 2010; Lee et al., 2013; Brown et al., 2014; Olaya-Abril et al., 2014), as well as for some fungi (Albuquerque et al., 2008; Vallejo et al., 2012; Wolf et al., 2014; Vargas et al., 2015; Joffe et al., 2016; de Toledo Martins et al., 2018; Zamith-Miranda et al., 2018). Nonetheless, fungi are the least studied so far. There are many possible functions that fungal EVs might perform, which are suggested on the basis of published but still limited research. One of the proposed functions of candidal vesicles is acting as carriers of virulence factors and signaling molecules for intercellular communication (Oliveira et al., 2010). EVs and their content might participate in the interaction between fungal pathogens and human cells during infection development, inducing also immune system response (Gil-Bona et al., 2015; de Toledo Martins et al., 2018). Moreover, Candida EVs might also play an important role in the communication between cells of the same species forming a complex community, influencing fungal cell adhesion to different surfaces, formation of biofilm structures, and further development and propagation of infection (Zarnowski et al., 2018; Zarnowski et al., 2021). In addition, the communication through EV release between Candida yeasts and other microorganisms living in the same niche might be crucial in the development of multispecies infections.

EVs might be produced by yeast cells grown in different morphological forms. Previously, we described the population of EVs released by planktonic cells and free-floating aggregates of non-albicans Candida species (Karkowska-Kuleta et al., 2020), consisting of several subpopulations of vesicles differing in size. In the current study, the heterogeneity of biofilm vesicle population has been demonstrated. Candida glabrata EVs produced by biofilm structures were rather smaller in diameter than the ones produced by planktonic cells. For C. tropicalis, the sizes were similar when comparing both forms, and in the case of C. parapsilosis, the mean sizes of biofilm EVs were larger than those determined by the NTA for the EVs released by free-floating cells (Karkowska-Kuleta et al., 2020). The dimensions of C. parapsilosis, C. glabrata, and C. tropicalis biofilm-derived vesicles were in a similar size range compared with previously described EVs released by C. albicans biofilm, constituting a population of structures of 30–200 nm in diameter (Zarnowski et al., 2018), although such smallest particles have not been measured for NAC species. The concentration of proteins and lipids was higher in EVs released by C. glabrata biofilm than in those produced by planktonic cells. In the case of C. parapsilosis EVs, the opposite was true and higher concentrations of proteins and lipids were proven for vesicles released by planktonic cells, whereas C. tropicalis biofilm vesicles contained a higher lipid content, and the protein concentration was comparable for both forms. The concentration of vesicular carbohydrates was not analyzed in a previous study; however, the presence of surface-exposed mannoproteins for EVs produced by planktonic fungal cells was demonstrated with flow cytometry (Karkowska-Kuleta et al., 2020). The current results indicate that in the case of biofilm-originated EVs, the highest concentration of polysaccharides was observed for C. tropicalis, while in the other two tested Candida species, the carbohydrate concentrations were rather within a similar range.

If the abundance of protein categories in planktonic vesicles (Karkowska-Kuleta et al., 2020) and biofilm vesicles is compared, for C. parapsilosis, a depletion of proteins related to pathogenesis is noticeable. However, for all species, a large proportion of the proteins are not well recognized and have no assigned functions in the database; hence, the actual categorization might be slightly different. In the case of C. glabrata, the presence of hydrolytic enzymes related to the adhesive properties and virulence of this species, i.e., yapsin 1 aspartyl proteinase (Yps1) and phospholipase 1 (Plb1), is noteworthy (Kaur et al., 2007; Gómez-Molero et al., 2015). Interestingly, in the case of C. glabrata and C. parapsilosis, two proteins from the SUR7 protein family have been identified, while for C. albicans, representatives of this family have been indicated as markers for candidal EVs (Dawson et al., 2020). Despite the fact that the determined concentration of vesicular proteins was the highest for C. tropicalis, the greatest protein diversity was observed for EVs of C. glabrata, and also with regard to their functionality, where each of the distinguished categories was represented, and the most numerous was the group of proteins related to cellular metabolism and transcription and translation. Moreover, the group of proteins responsible for cell wall organization, along with proteins involved in membrane transport, was represented in each species, which certainly highlights the role of fungal vesicles in the organization of this outer compartment of the cell and, in the case of the currently studied structures, most likely also in the formation of the biofilm matrix. The role of vesicles in building the biofilm matrix was confirmed previously for C. albicans (Zarnowski et al., 2018); additionally, the involvement of particular proteins constituting the cargo of biofilm EVs in biofilm regulation has been demonstrated, including the cell wall protein Ecm33 and the cell surface mannoprotein Mp65 (Zarnowski et al., 2021). The ortholog of the Ecm33 protein in our study was identified in both C. glabrata and C. parapsilosis, while Mp65 was identified for C. parapsilosis and C. tropicalis. The latter protein in C. albicans is a well-known antigen located at the cell wall, mainly of the hyphae form, stimulating T-cell proliferation of human peripheral blood mononuclear cells and inducing dendritic cell maturation and their production of proinflammatory cytokines—TNF-α and IL-6 (Gomez et al., 1996; Nisini et al., 2001; Pietrella et al., 2006).

Vesicles as extracellularly secreted and biologically active structures might participate in the communication between fungi and human cells inducing the immune system response (Herkert et al., 2019; Munhoz da Rocha et al., 2020). For EVs released by C. albicans planktonic cells, a few studies were presented previously showing the ambiguous character of the triggered response (Vargas et al., 2015; Vargas et al., 2020; Zamith-Miranda et al., 2021). Thus far, the effect of vesicles on changes in the production of several different cytokines has been shown, including TNF-α (one of the major proinflammatory cytokines regulating immune cells), which controls apoptotic pathways and stimulates the further release of other cytokines, or IL-10 (a potent anti-inflammatory interleukin), which suppresses the secretion of proinflammatory cytokines (Zhang and An, 2007). Vargas et al. (2015) showed that both TNF-α and IL-10 were released by murine primary phagocytes and cells from cell lineages in greater amounts after stimulation with fungal EVs than in the control samples, making it challenging to have a clear definition of the type of immune response induced. These studies showed also the increased level of IL-10 produced by murine dendritic cells compared with control without EVs added (Vargas et al., 2015); however, in later studies, this tendency was observed with a lower level of IL-10 compared with the previous result, albeit it was still higher than in the control sample (Vargas et al., 2020). In contrast, in the work of Zamith-Miranda et al. (2021), the production of TNF-α and IL-10 by murine dendritic cells was not detected after their stimulation with EVs produced by C. albicans and C. auris. In the current study, the immunomodulating properties of EVs of each investigated non-albicans Candida species were demonstrated by the evaluation of the level of cytokines released by human THP-1 cells differentiated into macrophage-like cells. The obtained results revealed an increasing tendency in the production of the proinflammatory cytokine TNF-α and a decreasing release of the anti-inflammatory cytokine IL-10 compared with cells untreated with EVs. This might be a signal that C. glabrata, C. tropicalis, and C. parapsilosis biofilm-derived vesicles induce a rather proinflammatory response of the human immune cells, thereby making Candida non-albicans biofilm vesicles promising components in anti-Candida vaccines. Reducing the concentration of IL-10 at the site of infection leads to increased apoptosis and inflammation, and although it accelerates the removal of the microbial factor causing infection, it also triggers significant damage associated with the expansion of the inflammatory state (Vazquez-Torres et al., 1999; Cyktor and Turner, 2011). The observed increase in the IL-8 secretion by THP-1 cells allows us to conclude that the presence of EVs of each tested non-albicans Candida species might also stimulate response involving other immune cells, as IL-8 is a chemoattractant affecting neutrophils and stimulating their chemotaxis, degranulation, and cytotoxicity, and the enhanced production of this interleukin is a signal favoring their migration to the infectious niche. Therefore, all these observations indicate rather the stimulation of a defense response of human cells in response to Candida non-albicans vesicles at the studied contact time.

As previously demonstrated for C. albicans, EVs also contribute to the regulation of fungal biofilm (Zarnowski et al., 2018; Zarnowski et al., 2021). The analysis of the formation of biofilm by C. tropicalis cells in the presence of EVs derived from the 48-h biofilm culture, introduced at an early stage of biofilm production, corroborated the substantial role of vesicles in the process of drug resistance that occurs in this complex structure of pathogen existence. Although the introduction of vesicles into cells growing without the presence of antifungal drugs did not cause a statistically significant effect on the biofilm (it also developed without EVs), in the presence of caspofungin and fluconazole, the protective effect of the vesicles was noticeable. In the case of the two drugs tested, both the thickness of the biofilm and the metabolic activity of the cells in the biofilm were partially restored if EVs from the mature biofilm were present at the stage of biofilm initiation; similarly, in the presence of caspofungin, cell viability was higher in biofilms supported by the added EVs. In the case of C. tropicalis biofilm, this may indicate the involvement of vesicles, among other mechanisms, in resistance to antifungals. Related conclusions were presented previously for the susceptibility of C. albicans biofilm to fluconazole, but these were from studies using mutant strains impaired in vesicle secretion and with incomplete cargo (Zarnowski et al., 2018; Zarnowski et al., 2021). With regard to C. tropicalis, the mechanisms of the observed phenomenon still require further detailed studies.

The characteristic and immunobiological properties of candidal vesicles investigated in the present study reveal to some extent the complexity of functions performed by fungal biofilm-derived EVs, thus also indicating the need for further research in this area which could result in a better understanding of the mechanism of development and pathogenesis of candidal infections.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: http://www.proteomexchange.org/, PXD033327; http://www.proteomexchange.org/, PXD033299; and http://www.proteomexchange.org/,PXD033300.

KK, JK-K, EK, OW, and MR-K planned the methodology and experiments. KK, JK-K, EK, and OW executed the experiments. PK and AO provided the resources. KK and JK-K wrote the manuscript draft. MR-K, AK, EZ-S, AO, and EP contributed to the editing and preparation of the final version of the manuscript. All authors contributed to the article and approved the submitted version.

This work was financially supported by the National Science Centre of Poland (grant no. 2019/33/B/NZ6/02284 awarded to MR-K).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.