Candida albicans biofilms and polymicrobial interactions

Abstract Candida albicans is a common fungus of the human microbiota. While generally a harmless commensal in healthy individuals, several factors can lead to its overgrowth and cause a range of complications within the host, from localized superficial infections to systemic life-threatening disseminated candidiasis. A major virulence factor of C. albicans is its ability to form biofilms, a closely packed community of cells that can grow on both abiotic and biotic substrates, including implanted medical devices and mucosal surfaces. These biofilms are extremely hard to eradicate, are resistant to conventional antifungal treatment and are associated with high morbidity and mortality rates, making biofilm-associated infections a major clinical challenge. Here, we review the current knowledge of the processes involved in C. albicans biofilm formation and development, including the central processes of adhesion, extracellular matrix production and the transcriptional network that regulates biofilm development. We also consider the advantages of the biofilm lifestyle and explore polymicrobial interactions within multispecies biofilms that are formed by C. albicans and selected microbial species.


Introduction
A biofilm is a consortium of microbes attached to a biotic or abiotic substrate embedded within a matrix of extracellular polymeric substances. For many microbes, the ability to form biofilms is an important virulence factor. Biofilms are found on natural substrates including plant and mammalian tissues and form readily on synthetic polymers such as plastics found in catheters, prostheses and implanted heart valves . Rapid medical intervention, high concentrations of antimicrobials and replacement of the infected device are required to manage a biofilm infection. However, such procedures are not without significant patient risk, as toxicity issues may arise from antimicrobial therapy and replacement of infected devices may sometimes require major surgery (for instance, when replacing coronary stents or heart valves). Thus, biofilm-related infections contribute significantly to the economic burden of global healthcare (Pittet 1994;Crump and Collignon 2000;Schierholz and Beuth 2001;Orsi et al. 2002;Yousif et al. 2015) and are associated with poor clinical outcome.
Candida species, which include the opportunistic pathogen Candida albicans, are one of the main causes of nosocomial infections worldwide, particularly in immunocompromised individuals (Hajjeh et al. 2004). Candida albicans can form complex biofilms on medical devices and, often, device removal is recommended (Kojic and Darouiche 2004;Pappas et al. 2016;Cavalheiro and Teixeira 2018;Nett and Andes 2020). Biofilms are also associated with recurrent candidaemia, which affects more than 400 000 people per year and has an alarming mortality rate of approximately 50% (Tumbarello et al. 2012;Bongomin et al. 2017;Li et al. 2018).
Candida albicans forms biofilms comprised of yeast cells, pseudohyphae and hyphae. The transition from planktonic growth to biofilm is accompanied by a complex remodelling of phenotypic behaviour underpinned by myriad changes in gene expression. The development of C. albicans biofilm occurs progressively, and is typically classified into four stages; (i) adsorption and adhesion of C. albicans yeast cells to a substrate, (ii) formation of microcolonies and production of the extracellular matrix, (iii) maturation and (iv) dispersal of cells from the mature biofilm. Once established, C. albicans biofilms are highly tolerant to antifungal therapy and can serve as a reservoir for recurrent infection. Indeed, the medical impact of C. albicans biofilm is underscored by a positive correlation between biofilm formation, increased virulence and higher patient mortality (Tumbarello et al. 2012;Rajendran, May, et al. 2016).
In this review, we summarise the current understanding of the processes involved in C. albicans biofilm formation and the regulatory circuits that are pivotal to C. albicans biofilm development. We explore multispecies biofilms, the advantages conferred to microbes during biofilm growth, persister cells and resistance to host immune clearance.

Biofilm formation
Biofilm formation begins with the adhesion of C. albicans yeast cells to a suitable surface. These early contact events comprise several non-specific factors including attractive and repulsive forces (such as van der Waals forces, hydrophobic interactions and Brownian movement forces), and electrostatic interactions that facilitate initial adherence (Douglas 1987;Hazen 1989;Hazen et al. 1990;Hobden et al. 1995;Jones et al. 1995;Williams et al. 2013). While these factors may not be the predominant mechanism for yeast cells to adhere, they are nevertheless an important prerequisite for biofilm formation (Park et al. 2008;Silva-Dias et al. 2015).
Following initial contact, cell wall-associated adhesion molecules are expressed to further strengthen fungal adhesion. There are three key families of C. albicans adhesins that play a major role in mediating adherence during biofilm formation: the agglutinin-like sequence (Als) family, the hyphal wall protein (Hwp) family, and the individual protein file family F/hyphally regulated (Iff/Hyr) family (de Groot et al. 2013). The most widely studied are the Als adhesins (Hoyer and Cota 2016) comprising eight members (Als1p-Als7p and Als9p), which are glycosylphosphatidylinositol (GPI)-linked to b-1,6 glucans in the fungal cell wall. The Als family of adhesins is key in both yeast and hyphal adhesion during biofilm formation and development.
During the initial adhesion of yeast cells to surfaces, Als1p and Als5p facilitate adhesion via their amyloidforming region that is critical for yeast cell-cell aggregation and cell-substrate adhesion (Rauceo et al. 2004;Ramsook et al. 2010;Garcia et al. 2011). These yeast cell aggregates form quickly and are key to the establishment of mixed-species biofilms with other fungal species and bacteria (Klotz et al. 2007;Alsteens et al. 2010;Ramsook et al. 2010;Lipke et al. 2012Lipke et al. , 2014. In addition to the Als family, other adhesins are important in mediating the initial attachment of C. albicans yeast cells during biofilm formation (Figure 1(A)). The enhanced adhesion to polystyrene protein (Eap1p) is significant in the clinical setting due to its ability to mediate attachment of C. albicans to plastics found in medical devices (Chong et al. 2018;McCall et al. 2019). Deletion of EAP1 renders C. albicans unable to attach to plastic surfaces and epithelial cells in vitro. In addition, Eap1p is required for biofilm formation in a rat central venous catheter model (Li et al. 2007).
Upon contact with a suitable surface, yeast cells germinate to form hyphae and pseudohyphae. The adhesion of these distinct morphological forms is crucial during the later stages of biofilm development. A vast array of adhesins orchestrate the adherence of yeast, pseudohyphae and hyphae, thereby contributing to biofilm maintenance. Als1p and Als3p have critical roles in facilitating hyphal adhesion during biofilm formation (Nobile et al. 2008). A C. albicans mutant unable to express ALS3 forms weaker and more disorganised biofilms compared to wild-type fungi in vitro (Zhao et al. 2006). Furthermore, disruption of both ALS3 and ALS1 (als3DD/als1DD) reduces the biofilm biomass when compared with controls (McCall et al. 2019).
Other Als family members also make significant contributions to biofilm development. Increased expression of ALS6, ALS7, and ALS9 rescues the biofilm defect of an als3DD/als1DD mutant (Nobile et al. 2008), while mutations in the amyloid-forming region of Als5p prevent the formation of adhesion nanodomains, cellular aggregates and biofilm development on polystyrene surfaces in vitro (Garcia et al. 2011). Conversely, decreased expression of ALS2 but not ALS4 results in impaired biofilm formation in a catheter biofilm model (Zhao et al. 2005). However, overexpression of ALS2 or ALS4 fails to rescue biofilm formation in an als3DD/als1DD null mutant in vivo (Nobile et al. 2008). This suggests that only certain Als proteins may function equivalently to Als1p and Als3p in an in vivo model of biofilm formation (Nobile et al. 2008).
Hyphal wall protein 1 (Hwp1p) is the most thoroughly characterised member of the Hwp family. Hwp1p is a molecular mimic of the substrate recognised by mammalian transglutaminase enzymes which cross-link Hwp1p to the surface of the host cell (Staab 1999), facilitating adhesion. A C. albicans hwp1DD mutant is unable to produce hyphae or form biofilms , underscoring the critical role of Hwp1p in these processes. Interestingly, overexpression of HWP1 is unable to promote biofilm formation in an als3DD/als1DD mutant in vitro or in vivo (Nobile et al. 2008), though a mixture of biofilm defective hwp1DD and als3DD/als1DD strains can form a hybrid biofilm. This suggests that while Hwp1p has a distinct function to the Als proteins, the three adhesins may have complementary roles in mediating adherence of C. albicans cells during biofilm formation (Nobile et al. 2008). Other Hwp family members including Hwp2p and Rbt1p contribute to initial cell attachment and the maintenance of adhesion (Li et al. 2007;Ene and Bennett 2009;Younes et al. 2011;Desai and Mitchell 2015;McCall et al. 2019). Mutants defective in HWP2 are unable to adhere efficiently to epithelial cells and present a decrease in biofilm mass , while RBT1 overexpressing strains exhibit an increase in cell surface hydrophobicity, which is essential for C. albicans adherence to polystyrene (Monniot et al. 2013).
The Iff/Hyr family of adhesins includes the hyphae specific Hyr1p (Bailey et al. 1996), which appears to be the only member of this family associated with biofilm formation and development. Mutants defective in HYR1 exhibit reduced adhesion in a flow system during the initial stages of biofilm development, as well as a reduction in biomass due to defective attachment (McCall et al. 2019). Furthermore, a significant reduction in both initial attachment and subsequent adhesion was observed following the loss of HYR1, while the loss of ALS3/ALS1 only affected the maintenance of adhesion, suggesting that Hyr1p is key in mediating adhesion during both early and late stages of biofilm development (McCall et al. 2019).

Biofilm development
Following adhesion, biofilm development is accompanied by changes in morphology, cell number, the secretion of extracellular polymeric substances and production of an extracellular matrix (ECM). Once adhered, yeast cells proliferate to form microcolonies, forming the basal layer of the biofilm. As the biofilm matures, the biomass increases with yeast cells, hyphae, pseudohyphae and the ECM accounting for a large proportion of the biofilm (Figure 1(B)). Hyphae are fundamental components that support the structural integrity of the biofilm and provide a scaffold for the attachment of additional yeast cells, pseudohyphae, other hyphae, as well as bacteria in multispecies biofilms (Chandra et al. 2001;Harriott and Noverr 2009). Mutants defective in the production of hyphae or hypha-specific adhesins are also defective in the production of substantial biofilms (Baillie and Douglas 1999;Ramage, VandeWalle, et al. 2002;Richard et al. 2005). Accordingly, the ability to form hyphae is critical for the normal development and maintenance of C. albicans biofilms. In a typical biofilm, an adherent layer of yeast cells is covered by an entangled mesh of hyphal filaments and pseudohyphae and encased in an ECM (Baillie and Douglas 1999;Chandra et al. 2001) (Figure 1(C)). Indeed, the ECM is essential for biofilm development and maturation. The ECM is composed of extracellular polymeric substances secreted by cells within the biofilm, including a complex mixture of macromolecules comprising proteins (55%), carbohydrates (25%), lipids (15%) and nucleic acids (5%) (Zarnowski et al. 2014). The most common polysaccharide found in the ECM of C. albicans biofilms are mannans, which are often associated with b-1,6 glucans (Pierce et al. 2017). Proteomic analysis of C. albicans extracellular polymeric substances also reveals the presence of numerous host molecules, including those found in epithelial cells and neutrophils, suggesting a potential role for host components in biofilm development (Andes et al. 2004;Dongari-Bagtzoglou et al. 2009;Zarnowski et al. 2014;Nett et al. 2015).
The presence of nucleic acids in C. albicans biofilms contributes to the structural and protective properties of the ECM. Extracellular deoxyribonucleic acid (eDNA) released by C. albicans within the ECM may contribute to the maintenance and stability of mature biofilms, but is not required to establish a C. albicans biofilm (Al-Fattani and Douglas 2006;Martins et al. 2010;Rajendran et al. 2014), suggesting an important role for eDNA in ECM assembly. The release of eDNA confers heterogeneity during biofilm formation, with higher levels of eDNA observed in isolates with high biofilmforming capacity compared to isolates with lower capacity ). Moreover, eDNA may also play a role in modulating the host immune response by activating bone marrow-derived myeloid dendritic cells (Kasai et al. 2006;Miyazato et al. 2009;Martins et al. 2010). Treatment of C. albicans biofilm with deoxyribonuclease I (DNase) improved the efficacy of amphotericin B in vitro (Martins et al. 2012), particularly in the presence of the chitinase inhibitor acetazolamide , suggesting that eDNA also contributes to recalcitrance. Accordingly, targeting eDNA within the ECM may present a novel therapy to improve the activity of antifungal drugs and decrease biofilm formation (Martins et al. 2012;Sapaar et al. 2014).
The biofilm matrix has pleiotropic functionality; it contributes to biofilm structure, provides physical support Pierce et al. 2017) and is critical in enhancing tolerance to antifungal drugs (described below).

Biofilm dispersal
Dispersal is the final stage of the biofilm lifecycle and describes a process where cells are released from a mature biofilm and are thus free to disseminate and establish secondary sites of infection ( Figure  1(D)). The majority of dispersal occurs from the uppermost layers of biofilm and is associated with yeast cells that have an elongated morphology (Uppuluri, Chaturvedi, et al. 2010). Several factors contribute to the process of dispersal. Candida albicans Nrg1p, Pes1p and Ume6p are known regulators of dispersal with opposing functions. Nrg1p is a negative regulator of the yeast-to-hypha morphological switch (Braun et al. 2001;Murad et al. 2001). Consistent with the release of yeast cells from biofilm during dispersal, overexpression of NRG1 promotes dispersal from biofilms in vitro (Uppuluri, Pierce, et al. 2010), while overexpression of the Pescadillo homologue PES1 also increases dispersal (Uppuluri, Chaturvedi, et al. 2010). In contrast, overexpression of the transcriptional regulator UME6 reduces dispersal (Uppuluri, Chaturvedi, et al. 2010), while transcriptional repression of the chaperone protein Hsp90p also reduces dispersal of C. albicans cells from biofilms in vitro (Robbins et al. 2011).
Candida albicans yeasts express Ywp1p, a cell-surface glycoprotein that is covalently linked to b-glucans in the fungal cell wall. Intriguingly, Ywp1p is reported to exhibit anti-adhesive properties consistent with a role in dispersal from biofilms. A C. albicans mutant unable to express YWP1 displays enhanced adhesion and biofilm formation (Granger et al. 2005). Members of the C. albicans histone deacetylase complex have also been implicated in biofilm dispersal, with C. albicans mutants unable to express HOS2, SET3, SIF2 or SNT1 exhibiting increased cohesiveness and impaired dispersal over a 60 h period .
Notably, dispersed cells have a number of altered characteristics when compared to planktonic cells including an increased ability to filament and form biofilms, increased adherence to biotic and abiotic substrates, reduced susceptibility to antifungals and an increased ability to damage endothelial cells in vitro (Uppuluri, Chaturvedi, et al. 2010). Dispersed cells also upregulate the expression of genes whose products are involved in the acquisition of micronutrients, drug resistance, and hydrolysis of host substrates (Uppuluri et al. 2018). Moreover, dispersed cells display enhanced pathogenicity in vivo when compared with matched planktonic controls (Uppuluri, Chaturvedi, et al. 2010). Collectively, these observations suggest that dispersed cells are "preconditioned" for maximum virulence. More recent work suggests that the degree of fungal dispersal is influenced by the extracellular environment. Investigations with mixed-species biofilms reveal that the Gram-positive bacterium Rothia dentocariosa can inhibit the release of C. albicans cells in vitro (Uppuluri et al. 2017), thus the amount of dispersal that occurs in vivo is likely to be influenced by biofilm composition.

Transcriptional regulation of biofilm formation and maintenance
The formation and development of C. albicans biofilms are underpinned by complex changes in transcriptional circuitry. While 58 transcription factors have been identified to contribute to C. albicans biofilm formation and development (Table 1), an interdependent core network of transcriptional regulators has been identified (Bcr1p, Tec1p, Efg1p, Ndt80p, Rob1p and Brg1p) that is critical for biofilm formation Glazier et al. 2017; Figure 1). Candida albicans mutant strains unable to express BCR1, TEC1, EFG1, NDT8 or ROB1 fail to form biofilms in vivo, while ablation of BRG1 results in biofilms with abnormal structure (Ramage, Vande Walle, et al. 2002;Nobile and Mitchell 2005;Nobile et al. 2012;Yano et al. 2016;Panariello et al. 2017). Collectively, this core network of transcriptional regulators influences the expression of approximately 1000 genes involved in hypha formation, adhesion, metabolism, production of extracellular polymeric substances and drug resistance Nobile and Johnson 2015).
More recent work has identified additional master regulators of biofilm formation including Flo8p, Gal4p and Rfx2p (Fox et al. 2015). Candida albicans mutant strains unable to express GAL4 or RFX2 form thicker biofilms (Fox et al. 2015), suggesting that these factors function as negative regulators. In contrast, strains unable to express FLO8 were markedly defective in biofilm formation in a rat central venous catheter model Fox et al. 2015). Bcr1p is critical for biofilm formation and influences the expression of several adhesins including Als3p, Hwp1p and Hyr1p Dwivedi et al. 2011). A C. albicans bcr1DD mutant produces a defective biofilm in vitro and in vivo, but overexpression of ALS3 rescues the defective biofilm phenotype, highlighting the importance of Als3p Liu and Filler 2011). Interestingly, Bcr1p also targets ECE1, which encodes the peptide toxin candidalysin Moyes et al. 2016). RNA transcriptome analysis reveals that several genes (including ECE1) are upregulated in high biofilm-forming isolates of C. albicans (Rajendran, May, et al. 2016).
The production of biofilm ECM is predominantly controlled by two transcription factors: Zap1p and Rlm1p. Zap1p is a zinc regulator that negatively regulates the formation of b-1,3 glucan through its target genes GCA1 and GCA2 (Nobile et al. 2009;Mayer et al. 2013;Nobile and Johnson 2015). In contrast, Rlm1p is a positive regulator of ECM production Nobile and Johnson 2015). Another transcription factor, Stp2p, is a key activator of amino acid permease genes that influences adherence and morphogenesis, adaptation to nutrients and biofilm longevity (B€ ottcher et al. 2020).
Candida albicans biofilms in vitro and in vivo: one morphology does not fit all Most biofilm studies have been conducted in vitro. For the most part, data obtained from in vitro biofilm studies correlate well with data obtained from in vivo and ex vivo models. The structure of biofilms from rat denture stomatitis and indwelling urinary catheter models is similar to the architecture of biofilms produced in vitro (Andes et al. 2004;Johnson et al. 2012). The formation and structure of C. albicans biofilm is influenced by numerous factors including the nature of the contact surface, environmental factors, and C. albicans morphology. Maturation of C. albicans biofilms on polyvinyl chloride catheter discs occurs within 24-48 h (Hawser and Douglas 1994), whereas maturation on polymethylmethacrylate strips requires longer timeframes (38-72 h) (Chandra et al. 2001). Interestingly, Andes et al. reported differences in temporal developments between in vivo and in vitro studies. Utilising a rat central venous catheter model, the duration of the early phase and maturation of biofilm formation in vivo was shorter than that observed in vitro (Andes et al. 2004). While C. albicans forms biofilms readily on abiotic substrates and on the mucosal surfaces of the oral cavity, conflicting observations have been made regarding the formation of biofilms on the vaginal mucosa (Harriott et al. 2010;Paiva et al. 2012;Sherry et al. 2017;Noverr and Fidel 2019;Swidsinski et al. 2019;Xu et al. 2019), suggesting that tissue/anatomical location and surrounding microbiota are likely to influence biofilm development. These observations have led to some contention regarding the definition of clinical diseases that specifically involve biofilms as opposed to diseases that are associated with a diverse or altered microbiota (Harriott et al. 2010;Paiva et al. 2012;Sobel 2015;Sherry et al. 2017;Noverr and Fidel 2019;Swidsinski et al. 2019;Xu et al. 2019). Rather than being a simple binary event, biofilms comprise a spectrum of morphologies, the composition of which varies between diseases. Accordingly, the provision of an unambiguous definition of C. albicans biofilms in the clinical setting is not without significant challenge.

Advantages of the biofilm lifestyle
The environment within a biofilm differs markedly from that encountered during planktonic growth and offers numerous advantages to the microbes that reside within it. The morphological variation within biofilms (yeasts, pseudohyphae and hyphae) combined with the presence of other fungi and bacteria collectively confer an increased competitive fitness that is absent in planktonic cells (Lohse et al. 2018). Microbes encased within the ECM are more likely to withstand minor mechanical stress and are better protected from the external environment and hostile host immune responses (Math e and Van Dijck 2013; Cavalheiro and Teixeira 2018;Kernien et al. 2018). Mixed-species biofilms can produce and utilise a greater variety of nutrients through the formation of metabolic consortiums in which the waste products of one species may be metabolised by another (Elias and Banin 2012). Such communities exhibit tremendous genomic diversity and, consequently, biofilms provide a suitable environment in which new genetic material can be acquired through the process of horizontal/lateral gene transfer, particularly between species of bacteria (Roberts and Kreth 2014).
Numerous differences have been reported between C. albicans biofilms and planktonic cells including alterations in transcriptional network activity (Yeater et al. 2007;Nobile et al. 2012;Fox et al. 2015), altered cell wall architecture (Nett et al. 2007), differences in ergosterol content of biofilm cell plasma membranes (Mukherjee et al. 2003) and, most notably, differences in sensitivities to antifungals. Candida albicans biofilms are substantially more difficult to eradicate with antifungals as compared to planktonic cells. Numerous mechanisms enable planktonic cells to withstand the toxicity of antifungals including constitutive expression and activation of drug efflux pumps, mutation of genes encoding drug targets, the acquisition of drug tolerance (Berman and Krysan 2020), alterations in the copy number of genes via aneuploidy (Yang et al. 2019) and adaptive mistranslation (Weil et al. 2017). Mature C. albicans biofilms are highly tolerant of fluconazole, amphotericin B, and caspofungin at concentrations that are lethal to planktonic cells (Vediyappan et al. 2010). Although numerous drug efflux pumps are expressed both transiently and permanently during the formation and maturation of C. albicans biofilm in vitro and in vivo (Mukherjee et al. 2003;Mateus et al. 2004;Yeater et al. 2007;Nett et al. 2009;Fox et al. 2015), the overall contribution of drug efflux to biofilm longevity appears to be relevant only during the early stages of biofilm formation (Mukherjee et al. 2003). Importantly, while C. albicans strains deleted for specific ATP-binding cassette transporters and major facilitator superfamily transporters (cdr1DD, cdr2DD, cdr1/cdr2DD, mdr1DD and mdr1/ cdr1DD) were highly susceptible to fluconazole during planktonic growth, the same mutants exhibited resistance to fluconazole when grown as a biofilm (Ramage, Bachmann, et al. 2002), suggesting that factors distinct from drug efflux pumps contribute to drug tolerance within a biofilm.
Specific components of the biofilm ECM are instrumental in the orchestration of tolerance to antifungals. Candida albicans mutant strains unable to deliver b-1,3 glucan to the biofilm ECM are susceptible to fluconazole (Nett, Sanchez, et al. 2010;Nett et al. 2011;Taff et al. 2012) while reducing FKS1 (glucan synthase) expression renders cells susceptible to amphotericin B, anidulafungin, and flucytosine (Nett, Crawford, et al. 2010). Indeed, b-glucan is capable of binding fluconazole and amphotericin B (Nett et al. 2007;Vediyappan et al. 2010), while treatment of C. albicans biofilms with the enzyme b-1,3 glucanase increases sensitivity to fluconazole (Nett et al. 2007). Sequestration of fluconazole by b-glucan in the ECM combined with a reduction in antifungal penetrance into the biofilm often results in fungi receiving a sub-lethal dose of the drug, which can lead to the emergence of resistance.
Collectively, these observations are consistent with the notion that the ECM functions in part, as a "molecular trap" that impedes the diffusion of antifungals into the biofilm, causing a reduction in the local concentration of antifungal that can engage with target cells. Exposure of biofilms to diminished concentrations of an antifungal is likely to impose a positive selection pressure towards the acquisition of mutations that reduce susceptibility to subsequent therapy, an outcome that may have serious clinical implications should mutated cells be dispersed from a recalcitrant biofilm.
More recently, studies have highlighted an essential role for extracellular vesicles (EVs) in conferring the drug-resistant properties of C. albicans biofilms. Analyses of EVs released from C. albicans biofilms reveal a proteome that is distinct from that present in EVs released from planktonic cells, and which contain high levels of mannan and glucan (Zarnowski et al. 2018). Candida albicans phr1DD and sun41DD null mutants are devoid of specific EV cargo proteins and form biofilms that are sensitive to fluconazole. Strikingly, EVs that were harvested from wild type C. albicans biofilms and applied to phr1DD and sun41DD biofilms restored the ability to grow in the presence of the drug (Zarnowski et al. 2018), demonstrating a functional role for EVs in conferring drug resistance.
The effectiveness of the host immune response to biofilms is also diminished when compared to planktonic cells. The biofilm ECM protects underlying fungal cells from the innate immune system, particularly from killing by neutrophils and monocytes. Impaired recognition of biofilms by neutrophils and monocytes is evidenced by altered patterns of cytokine secretion when compared to those produced when encountering planktonic cells ). Monocytes can phagocytose planktonic C. albicans but not fungal cells associated with biofilms, while neutrophils fail to eradicate mature biofilms Katragkou et al. 2010;Xie et al. 2012;Pierce et al. 2017). Furthermore, neutrophils release extracellular traps in response to planktonic C. albicans yeast and short hyphae but fail to release extracellular traps in response to biofilm ECM (Nett 2016;Johnson et al. 2016Johnson et al. , 2017Pierce et al. 2017;Kernien et al. 2018).

Persister cells: a disputed presence in C. albicans biofilms
Persister cells are a contentious topic in the context of C. albicans biofilms. Defined as a subpopulation of metabolically dormant biofilm cells, these cells are only detected in the presence of antifungals. Persister cells express high levels of alkyl hydroperoxide reductase 1 (Ahp1p), which is positively correlated with prolonged survival following exposure to amphotericin B (Truong et al. 2016). When antifungal therapy is discontinued, persister cells proliferate to restore the biofilm. Patients with long-term oral carriage of C. albicans have a higher incidence of persister cells (LaFleur et al. 2010), suggesting that persister cells may be relevant in the context of continual colonisation or recurrent infections in vivo (Wuyts et al. 2018). However, despite observations which support the existence of persister cells they remain a controversial topic. Indeed, some studies suggest that persister cells are not a general characteristic of Candida biofilms (Al-Dhaheri and Douglas 2008; Denega et al. 2019).

Heteroresistance of C. albicans biofilms
As previously mentioned, C. albicans biofilms are difficult to eradicate. An issue that clinicians are facing is biofilm heterogeneity. Indeed, clinical isolates of C. albicans biofilms from the bloodstream or the vaginal mucosa seem to differ in biomass and therefore isolates can be defined as either high-or low-biofilm formers (Sherry et al. , 2017. This inherent biological heterogeneity has led to patients responding differently to antifungals (Kean et al. 2018) and is thus called heteroresistance. Heteroresistance is defined as the presence of a subpopulation with lower susceptibility to specific antifungals within a larger microbial population. A recent study of C. glabrata clinical strains suggests that this heteroresistance might be linked to persistent infections (Ben-Ami et al. 2016).

Multispecies biofilms: C. albicans and polymicrobial interactions
Interactions between C. albicans and the microbes found on biotic and abiotic substrates are varied and complex, comprising synergistic, antagonistic, and neutral relationships (Delaney et al. 2018). These interactions involve direct cell to cell contact, chemical interactions via the secretion of small molecules, enhancement of colonisation, changes in the host environment, the use of metabolic by-products, changes in the host response or a combination of these. Unsurprisingly, these interactions are particularly relevant in the context of biofilms and are discussed below.

Adhesion in mixed species biofilms
The interactions of C. albicans with other microorganisms can occur via co-aggregation and co-adhesion. Candida albicans adhesins (e.g. Als1p, Als2p, Als3p, Hwp1p) facilitate interaction with bacterial species such as Streptococcus gordonii and Staphylococcus aureus (Peters et al. 2012;Hoyer et al. 2014;Schlecht et al. 2015;Allison et al. 2019; Figure 2(A)). The adherence of C. albicans to pre-attached Streptococcus spp. in the oral cavity is mediated by polysaccharide receptors on the streptococcal cell surface (SspA and SspB) and C. albicans hypha-associated adhesins (Als1p and Als3p) (Holmes et al. 1996;Bamford et al. 2009;Silverman et al. 2010). Moreover, S. gordonii SspA and SspB mutants exhibit reduced expression of streptococcal adhesins for C. albicans, with less co-aggregation of C. albicans with these S. gordonii mutants (Klotz et al. 2007).

Microbial synergism and antagonism within mixed species biofilm
Interkingdom interactions between C. albicans and other microbes can be synergistic, mutually beneficial or antagonistic, and these interactions can dictate the composition of the microbiota. One of the most thoroughly characterised synergistic relationships is that of C. albicans and Streptococcus mutans. Within plaque biofilms, S. mutans glucosyltransferases bind directly to mannans on the surface of C. albicans yeast and hyphal cell walls, promoting the development of an extensive ECM and the formation of a mixed species biofilm (Hwang et al. 2015;Ellepola et al. 2017). Consistent with this, S. mutans mutants lacking glucosyltransferase genes demonstrate reduced ability to form mixed species biofilms with C. albicans (Falsetta et al. 2014;Hwang et al. 2017; Figure 2(A)). The presence of both S. mutans and C. albicans in the oral microbiota significantly increases bacterial colonisation and supports the development of dental caries, suggesting that this interaction may be a pivotal factor in increasing the risk of disease (de Carvalho et al. 2006;Khoury et al. 2020). In addition to physical associations, streptococci excrete Streptococcus gordonii can utilise C. albicans adhesins to directly bind to hyphae. In contrast, glycosyltransferases (Gtfs) secreted by Streptococcus mutans within the oral cavity can bind to C. albicans mannans, increasing the production of glucans and ECM. Consequently, the glucan increases the ability of the bacterium to bind to C. albicans and forms a C. albicans-S. mutans biofilm on the tooth surface (dental plaque). (B) Signalling molecules produced by C. albicans and bacterial species enable inter-kingdom communication within multispecies biofilms. For example, S. mutans and Pseudomonas aeruginosa can secrete quorum sensing molecules that influence the behaviour of C. albicans within the biofilm. Likewise, the C. albicans quorum sensing molecule farnesol, can influence the behaviour of interacting bacteria. lactate that acts as a carbon source for C. albicans yeast growth (Ene et al. 2012;Metwalli et al. 2013). The C. albicans quorum sensing (QS) molecule farnesol has known antibacterial effects, however, within a dual-species biofilm, farnesol enhances S. mutans cell growth and microcolony development . Conversely, the S. mutans QS molecules mutanobactin A, trans-2-decenoic acid and competence-stimulating peptide can inhibit C. albicans germ tube formation and the yeast-to-hypha transition (Jarosz et al. 2009;Joyner et al. 2010;V ılchez et al. 2010; Figure 2(B)). Overall, this demonstrates a mutually beneficial relationship between C. albicans and S. mutans with a synergistic partnership that can alter the oral microbiota and influence disease.
Within a dual-species biofilm, C. albicans can protect anaerobic bacteria from high concentrations of oxygen by providing a hypoxic microenvironment that supports the growth of Bacteroides fragilis, B. vulgatus and Clostridium perfringens Valentine et al. 2019). Accordingly, C. albicans affected the biofilm composition, suggesting a role for the fungus in maintaining homeostasis within the microbiota. Additionally, C. albicans b-1-3-glucan facilitates bacterial tolerance to ofloxacin  and vancomycin (Kong et al. 2016) in a dual-species biofilm with Escherichia coli and S. aureus respectively, demonstrating antimicrobial tolerance and microbial persistence within multispecies biofilms. However, interactions between C. albicans and bacteria are not always synergistic. Antagonistic effects have been reported between Lactobacillus spp. and C. albicans, with both hypha formation and biofilm development affected by the bacteria (Orsi et al. 2014). In the normal vaginal microbiota, Lactobacillus spp. compete with C. albicans for receptors present on the surface of genitourinary epithelium for adhesion; they also secrete lactic acid and hydrogen peroxide, lowering the pH, which in turn inhibits fungal attachment to the vaginal epithelium (Boris et al. 1998;Boris and Barb es 2000;Aroutcheva et al. 2001).
The antagonistic relationship between C. albicans and Pseudomonas aeruginosa is particularly well defined. Both P. aeruginosa and C. albicans are frequently co-isolated from contaminated catheters and chronic lung infections (El-Azizi et al. 2004;Williams and C amara 2009). Numerous interactions occur between C. albicans and P. aeruginosa, the outcomes of which are greatly influenced by the environment. Pseudomonas aeruginosa can form dual-species biofilms with C. albicans in liquid medium through a physical association with C. albicans hyphae and germ tubes but not yeast cells. Once formed, C. albicans hyphae are killed by P. aeruginosa secreted virulence factors (phospholipase C and redox-active oxygen species) and quorum sensing molecules (3-oxo-C12-homoserine lactone (C12AHL)) (Hogan and Kolter 2002;Hogan et al. 2004;Gupta et al. 2005;Gibson et al. 2009; Figure 2(B)). The destruction of hyphae but not C. albicans yeast cells suggests a potential survival mechanism developed by the fungus that allows yeast to evade destruction by P. aeruginosa.
Metabolic interactions play an important role in microbial adaptation within polymicrobial environments. Metabolic dependencies between microbes can be direct or indirect and involve the production of metabolites by one species and their subsequent consumption by another, or the localised depletion of a toxic metabolite by one species, enabling a second species to persist (Krom et al. 2014). Competition for nutrients is endemic within polymicrobial communities. In a mixed-species biofilm, P. aeruginosa sequesters iron from C. albicans, reducing the metabolic activity of the fungus (Purschke et al. 2012). Lactobacillus rhamnosus can protect oral epithelial cells from fungal damage by inducing metabolic reprogramming of C. albicans. This protective effect comprises a reduction in C. albicans adhesion, impaired hypha formation and glucose depletion (Mail€ ander-S anchez et al. 2017). As the nature of fungal metabolism is adaptable, these metabolic interactions may be significant in maintaining homeostasis.

Intra-genus interactions within Candida biofilms
While the establishment of mixed-species biofilms involving C. albicans and bacteria has been extensively described, intra-genus interactions between Candida species are not as well characterised. Although C. albicans has been studied more extensively than any other Candida species, other species such as C. dubliniensis, C. glabrata, C. krusei, C. tropicalis and C. auris are becoming an increasing concern in the clinical setting. Though poorly studied, C. albicans, and other non-C. albicans species have been shown to form biofilms with each other (Pereira-Cenci et al. 2008;Pathak et al. 2012). Mixed species biofilms of C. albicans and non-C. albicans species have been identified on dental acrylic resin strips in which C. albicans was the predominant species (Pathak et al. 2012). Moreover, aggregates of C. albicans and C. dubliniensis have been isolated from the oral cavity of immunocompromised patients (Kirkpatrick et al. 2000), while C. albicans and C. glabrata were frequently co-isolated from patients with dentureassociated stomatitis (Coco et al. 2008). In another study, the presence of C. albicans in a mixed biofilm facilitated the invasion of C. glabrata in an in vitro model of denture stomatitis and promoted significant tissue damage to the epithelium during vulvovaginal candidiasis (Silva et al. 2011;Alves et al. 2014). The increasing number of mixed Candida species biofilms poses a major problem in the clinical setting due to the resistance of Candida species to a variety of antifungal drugs (Al-Fattani and Douglas 2006;Bizerra et al. 2008). These interactions highlight the importance of further research into the mechanisms and consequences of C. albicans and non-C. albicans species biofilm formation.

Communication within mixed-species biofilms
Numerous chemical interactions occur between the microbes found within a mixed-species biofilm via the secretion of small molecules and the exchange of nutrients. Both bacterial and fungal populations within a biofilm can modulate their behaviour via quorum sensing; a mechanism of chemical interaction in which microbes communicate through the secretion of signalling molecules (Wongsuk et al. 2016; Figure 2(B)). Quorum sensing induces changes in transcriptional activity, which in turn influences the behaviour, survival, or virulence of either microbe. Candida albicans secretes farnesol, a quorum-sensing molecule that inhibits the Ras1-cyclic AMP (cAMP)-protein kinase A (PKA) (Ras1-cAMP-PKA) signalling pathway to repress hyphal development (Hornby et al. 2001;Davis-Hanna et al. 2008), inhibit biofilm formation (Ramage, Saville, et al. 2002) and upregulate the expression of FCR1 and PDR16, which are involved in resistance to azole therapy (Cao et al. 2005). Notably, the addition of farnesol to cultures of P. aeruginosa decreases the production of Pseudomonas quinolone signalling (PQS) molecule, a known suppressor of C. albicans biofilm formation, and the PQS-controlled virulence factor pyocyanin and inhibits swarming motility (Cugini et al. 2007;McAlester et al. 2008). As these P. aeruginosa factors are toxic to C. albicans, the release of farnesol within this interaction is likely to promote C. albicans survival.
In the oral cavity, S. gordonii uses quorum sensing to suppress farnesol-mediated repression of C. albicans filamentation, enhancing biofilm formation (Bamford et al. 2009). Such behaviour is far from altruistic. Indeed, S. gordonii promotes fungal growth by providing a carbon source to C. albicans (Bamford et al. 2009) and, in return, C. albicans increases the likelihood that S. gordonii can survive and persist in the oral cavity by lowering oxygen tension and releasing bacterial growth factors derived from nutrient metabolism (Shirtliff et al. 2009).
Molecules secreted by microbes can have an inhibitory effect on C. albicans. In the gastrointestinal tract, Bacteroides thetaiotaomicron stimulates the host to express HIF-1a and LL-37, which inhibit gastrointestinal colonisation by C. albicans (Fan et al. 2015). Similarly, C. albicans is susceptible to type VI secretion system (T6SS)-mediated antifungal activity perpetrated by the Gram-negative opportunistic bacterium Serratia marcescens, which uses its T6SS to penetrate the fungal cell wall to deliver the toxic effector proteins Tfe1 and Tfe2 that disrupt nutrient uptake, amino acid metabolism and plasma membrane potential (Trunk et al. 2018). However, whether this occurs in the context of a multispecies biofilm has yet to be investigated.
Overall, polymicrobial interactions within multispecies biofilms are of great importance and further studies may aid better understanding of C. albicans co-infection and possible therapeutic strategies to combat inter and intra-kingdom biofilm infections. Moreover, our understanding from next-generation sequencing studies, accompanied by greater bioinformatics power, is likely to aid our discovery of key interactions between Candida spp. and bacteria from an array of clinical diseases.

Clinical impact of biofilms
Biofilms are intrinsically associated with highly virulent, persistent and drug-resistant infections. The dispersal of cells from an established biofilm facilitates wholebody dissemination and reduces the likelihood of a positive clinical outcome and compounds options for effective therapeutic intervention. As such, C. albicans biofilms are widely considered a major virulence factor and a key determinant to the high mortality rate attributed with candidiasis (Tumbarello et al. 2007;Tumbarello et al. 2012;. Infections that arise from mixed-species biofilms are considerably more difficult to treat than their singlespecies counterparts, requiring complex multi-drug treatment strategies (Costa-Orlandi et al. 2017;Montelongo-Jauregui et al. 2019). Antimicrobials that target a specific group of species often enable other microbes within the biofilm to thrive and continue infection (Pammi et al. 2014;Kean et al. 2017;Allison et al. 2019); for instance within a S. aureus-C. albicans biofilm, the efficacy of miconazole is reduced, with a much higher concentration of the drug needed to induce an effect against the biofilm (Kean et al. 2017). This increases the challenge of finding effective treatments to tackle C. albicans multispecies biofilms and may exacerbate issues relating to drug resistance.
While the treatment of single species C. albicans biofilms has been studied in vivo, in vitro and in clinical settings (Schinabeck et al. 2004;Shuford et al. 2006;Tsui et al. 2016), considerably less is known about targeted therapeutics to mixed-species biofilms. Understanding these polymicrobial interactions in mixed-species biofilms is necessary in order to improve current therapeutic strategies. Overall, we note a lack of anti-biofilm strategies, but an increasing need for efficacious therapies that specifically target C. albicans biofilms.

Therapeutic intervention
In recent years, attempts have been made to utilise natural or synthetic compounds with improved antimicrobial activity. The use of quaternary ammonium amphiphiles may be a useful application in preventing and treating C. albicans biofilms. Amphiphile screening has revealed putative antimycotic agents that target C. albicans membranes in biofilms (Gupta et al. 2020), and which prevent biofilm formation on acrylics and promote degradation. Natural compounds have served as a source of novel alternatives to C. albicans biofilm treatment (Manoharan et al. 2017). Shikonin, from the roots of the Lithospermum erythrorhizon plant, prevents biofilm formation and maintenance (Yan et al. 2019), while Thymus vulgaris, and its active compound thymol, has demonstrated efficacy against C. albicans biofilms in vitro when used alone or in combination with fluconazole, revealing disaggregation and deformity in C. albicans biofilms (Jafri and Ahmad 2020). Furthermore, thymol acts synergistically with standard antifungals (Sharifzadeh et al. 2018) and could be utilised in combination therapy. Recent studies have shown that miconazole combined with a chitosan-coated iron oxide nanoparticle could be used in anti-biofilm therapy against C. albicans (Arias, Pessan, et al. 2020), and that this was also efficacious within a range of different oral interkingdom biofilms (Arias, Brown, et al. 2020). These alternative methods of treatment may present a strategy to overcome tolerance in the treatment of C. albicans biofilms, but also in more complex interkingdom communities. The potential benefits of combination therapy include a broader spectrum of action, greater potency than agents used in monotherapy, reduction in the number of resistant organisms and lower tolerance to antifungal agents (Lewis and Kontoyiannis 2001;Mukherjee et al. 2005;. More recently, drug combination therapy has been reported as an alternative to conventional treatment. The codelivery of C12AHL (a quorum-sensing molecule) with fluconazole in a liposomal drug carrier has shown promise as a strategy to eliminate C. albicans biofilms (Bandara et al. 2020). In recent work, key components of the host innate immune response and fungal cell wall components have been a focus for the development of new antifungals. The antimicrobial peptides ToAP2 and NDBP-5.7, increased yeast cell permeability and generated morphological alterations in C. albicans cells in both the early and later stages of biofilm formation (do Nascimento Dias et al. 2020). Furthermore, the combination therapy utilised within this study increased the efficacy of fluconazole and amphotericin B, decreasing their active concentrations and potential toxicity (do Nascimento Dias et al. 2020). Finally, vaccination of mice with the N-terminus of recombinant Als3p (designated NDV-3A) markedly reduced the ability of C. albicans to colonise a medically implanted jugular vein catheter (Alqarihi et al. 2019), suggesting that targeting key adhesins may have a positive outcome in the clinical setting.

Conclusion
The ability of C. albicans to form biofilms is a complex process and complicates treatment of C. albicans biofilm-related infections. Biofilm development, antifungal resistance and polymicrobial interactions have been a continual focus of C. albicans biofilm research and while recent advances have begun to highlight the complex mechanisms that underpin the formation of C. albicans single and mixed-species biofilms, there is still much to be discovered. The prevention of biofilm formation and the eradication of mature biofilms in the clinical setting present formidable challenges that must be overcome if the mortality associated with biofilm infection is to be addressed. The question remains as to how we can utilise this knowledge to develop more efficacious strategies to treat C. albicans biofilms and biofilmrelated infections.

Disclosure statement
No potential conflict of interest was reported by the author(s).