Targeting Candida albicans in dual-species biofilms with antifungal treatment reduces Staphylococcus aureus and MRSA in vitro

Polymicrobial biofilms consisting of fungi and bacteria are frequently formed on endotracheal tubes and may contribute to development of ventilator associated pneumonia (VAP) in critically ill patients. This study aimed to determine the role of early Candida albicans biofilms in supporting dual-species (dual-kingdom) biofilm formation with respiratory pathogens in vitro, and investigated the effect of targeted antifungal treatment on bacterial cells within the biofilms. Dual-species biofilm formation between C. albicans and three respiratory pathogens commonly associated with VAP (Pseudomonas aeruginosa, Escherichia coli and Staphylococcus aureus) was studied using quantitative PCR. It was shown that early C. albicans biofilms enhanced the numbers of E. coli and S. aureus (including methicillin resistant S. aureus; MRSA) but not P. aeruginosa within dual-species biofilms. Transwell assays demonstrated that contact with C. albicans was required for the increased bacterial cell numbers observed. Total Internal Reflection Fluorescence microscopy showed that both wild type and hyphal-deficient C. albicans provided a scaffold for initial bacterial adhesion in dual species biofilms. qPCR results suggested that further maturation of the dual-species biofilm significantly increased bacterial cell numbers, except in the case of E.coli with hyphal-deficient C. albicans (Ca_gcn5Δ/Δ). A targeted preventative approach with liposomal amphotericin (AmBisome®) resulted in significantly decreased numbers of S. aureus in dual-species biofilms, as determined by propidium monoazide-modified qPCR. Similar results were observed when dual-species biofilms consisting of clinical isolates of C. albicans and MRSA were treated with liposomal amphotericin. However, reductions in E. coli numbers were not observed following liposomal amphotericin treatment. We conclude that early C. albicans biofilms have a key supporting role in dual-species biofilms by enhancing bacterial cell numbers during biofilm maturation. In the setting of increasing antibiotic resistance, an important and unexpected consequence of antifungal treatment of dual-species biofilms, is the additional benefit of decreased growth of multi-drug resistant bacteria such as MRSA, which could represent a novel future preventive strategy.

Introduction Biofilm formation is a phenotype of many microorganisms, linked to the ability to survive in a hostile host environment [1]. Biofilm-related infections are considered a significant and increasingly prevalent source of morbidity and mortality within the healthcare system and thus much research effort is required towards aiding their prevention and management [2]. Indeed the polymicrobial nature of biofilms associated with infections such as cystic fibrosis [3,4] and chronic wounds [5] is increasingly recognised as an important contributor to disease pathogenesis. Despite our knowledge that biofilms in vivo generally contain multiple species encased within an extracellular polymeric matrix, much remains to be learned about the development of mixed species biofilms and the beneficial, parasitic or antagonistic interactions that exist between the microorganisms within them [6][7][8][9].
In-dwelling medical devices, many of which are used in critically ill patients, have also been well documented to support polymicrobial biofilm growth [10,11], leading to infections ranging from catheter-related urinary tract infections [10], to ventilator associated pneumonia (VAP) [11,12]. Biofilm colonisation of the endotracheal tube in VAP patients has been reported to occur rapidly after intubation [13], with biofilms acting as reservoirs for pathogenic bacteria [14]. Indeed it has been shown that there is microbiological continuity between airway colonization, biofilm formation and VAP development [15]. Emerging technologies to detect pathogens associated with VAP have identified a broad range of Gram positive and Gram negative bacteria including, Staphylococcus aureus, Escherichia coli and Pseudomonas aeruginosa, amongst the common bacterial pathogens associated with VAP [16]. Of particular interest is the finding that Candida albicans [17], is often present in endotracheal tubes [11], and is associated with an increased risk of VAP (although not as the pathogen causing VAP), as well as prolonged intensive care unit and hospital stays [18]. Given that oral carriage of Candida is reported in up to 40-60% of the population [19,20], it is plausible that C. albicans could adhere to endotracheal tubes on their insertion through the mouth. Thus, establishment of an early C. albicans biofilm could facilitate respiratory pathogen integration, leading to subsequent polymicrobial biofilm formation and ultimately drive the development of VAP.
An emerging interest in interactions between fungi and bacteria has highlighted the complexity of fungal-bacterial interactions in polymicrobial and dual-kingdom biofilms, particularly those involving dimorphic fungi. Polymicrobial biofilms are notoriously difficult to eradicate and tend to be recalcitrant to both antimicrobials and host defences [21]. It has been suggested that the morphological plasticity of C. albicans between yeast and hyphal forms has a major influence on its virulence [22] and that the hyphal form of C. albicans could provide architecture to the developing biofilm [8,23].
In this study, we investigated dual-species biofilms between C. albicans and three bacterial species with a view to determining if C. albicans could enhance respiratory bacterial pathogen numbers in dual-species biofilms, using qPCR for bacterial enumeration. We used a biofilm model in which C. albicans was inoculated first, to develop an early biofilm before the addition of the bacterial species. We also investigated whether contact was needed between the early C. albicans biofilm and bacterial cells to enhance bacterial cell numbers. Using Total Internal Reflection Fluorescence (TIRF) microscopy, we imaged bacterial attachment to wild type and hyphal-deficient C. albicans and by qPCR determined bacterial cell numbers during dual-species biofilm maturation. Furthermore, we examined whether a targeted preventative approach with an antifungal could reduce bacterial cell numbers in dual-species biofilms. The current work has potential clinical implications as it could open up new preventive and therapeutic targets for the management of polymicrobial biofilms infections such as VAP, with the possibility of decreasing bacterial numbers of S. aureus and MRSA without antibiotics, thereby lowering the risk of developing antibiotic resistance.

Micro-organism strains and growth conditions
C. albicans (NCTC 3179) was sub-cultured aerobically on Sabouraud agar plates and propagated in yeast peptone dextrose (YPD) broth (US Biological). P. aeruginosa (ATCC 27853), E. coli (ATCC 29522), S. aureus (NCTC 6571) and methicillin resistant S. aureus (MRSA; MRSA 4D) were grown on blood agar plates and propagated in brain heart infusion (BHI) broth. The laboratory strain (wild type) of C. albicans (NCTC 3179) was used in all experiments, except where otherwise stated, such as dual-species biofilm treatment with liposomal amphotericin, in which a clinical isolate of C. albicans (CA 239SB) was also studied. In experiments on dualspecies biofilm maturation, a C. albicans hyphal-deficient mutant (Ca_gcn5Δ/Δ) [24] (gift from Prof Karl Kuchner, Medical University of Vienna) was studied along with the laboratory strain of C. albicans. The clinical isolate and hyphal-deficient mutant of C. albicans were subcultured as described for the laboratory strain.

Preparation of single species bacterial biofilms
Overnight cultures (18 hours) of P. aeruginosa, E. coli, S. aureus and MRSA were resuspended in BHI to yield inocula of 5.0 x 10 6 cells/ml. A total volume of 100 μl of each inoculum in BHI was added to microtitre plate wells (Thermo Fisher Scientific, Roskilde, Denmark), and the plates were incubated at 37˚C for 4 hours to allow initial biofilm formation under static growth conditions. Wells were washed carefully, three times with PBS, to remove planktonic cells and the biofilms incubated with 100 μl fresh broth for a further 24 hours to allow biofilm maturation. Biofilms were then washed to remove planktonic cells and quantified by qPCR as outlined below. Single species bacterial biofilms were grown in BHI throughout (4 hours for initial biofilm formation, plus an additional 24 hours for early biofilm maturation).

Preparation of dual-species biofilms
A schematic outline of dual-species biofilm formation, treatment and quantification is shown in S1 Fig. Overnight cultures (18 hours) of C. albicans were washed with phosphate-buffered saline (PBS) and resuspended in a modified Roswell Park Memorial (RPMI) medium (RPMI-1640; Sigma-Aldrich, St Louis, USA), referred to subsequently as RPMI, to yield an inoculum of 1.0 x 10 6 cells/ml [25]. The C. albicans inoculum (100 μl) was added to microtitre plate wells (Thermo Fisher Scientific, Roskilde, Denmark) and incubated under static growth conditions at 37˚C for 4 hours to allow initial biofilm formation. The biofilm was washed three times with PBS to remove planktonic C. albicans cells and spent RPMI, prior to inoculation with 100 μl of P. aeruginosa, E. coli, S. aureus, or MRSA (5 x 10 6 cells/ml) in BHI. Bacteria were allowed to adhere to the initial C. albicans biofilms for 4 hours to facilitate dual-species biofilm formation. Following washing to remove planktonic cells and spent media as outlined above, biofilms were incubated in BHI for a further 24 hours to allow dual-species biofilm maturation. Wells were then washed with PBS as outlined above and the biofilms were quantified by qPCR (S1A Fig). Single species bacterial biofilms served as controls.
In selected experiments, to further investigate the role of C. albicans in enhancing bacterial cell numbers during biofilm maturation, we prepared dual-species biofilms as outlined above with wild type C. albicans or C. albicans gcn5Δ/Δ hyphal-deficient mutant. We then quantified bacterial numbers after initial adhesion (4 hours), or following a combined initial adhesion period (4 hours) and further maturation period (24 hours). Quantification of bacterial cell numbers was undertaken by qPCR as described below.

Bacterial cell quantification in biofilms by qPCR
To quantify cell numbers in single species or dual-species biofilms, the biofilms were detached from microtitre wells into 100 μl of broth by sonicating for 5 min in an ultrasonic bath (Dawe, Middlesex, UK). Any remaining cells were then collected into a further 100 μl broth. DNA was extracted using the microLYSIS 1 -Plus kit (Microzone, Haywards Heath, UK) as per the manufacturer's instructions and individual monoplex qPCR assays were performed using an Mx3005P qPCR System (Agilent Technologies, California USA) as detailed in S1-S5 Tables (Supporting information). The candidate genes selected were expressed by the specific microorganisms used in this study and the primers employed for their quantification had been designed and published previously: Primers for P. aeruginosa were against the oprL gene

Generation of standard curves for qPCR
To allow for quantification of bacterial cell numbers, DNA standards were prepared by extraction of DNA from planktonic organisms using the microLYSIS1-Plus kit and purified using the DNeasy kit (Qiagen, Manchester, UK). DNA standards, as previously described by us [29], were used in all qPCR assays to generate standard curves from which the numbers of organisms within the biofilms could be determined. Cell numbers within the biofilm were thus generated from and were equivalent to cell numbers from planktonic cultures used in the standard curves.

Transwell assays for determining the influence of contact with C. albicans biofilms on E. coli, S. aureus and MRSA cell numbers
To determine whether contact with early C. albicans biofilms was required to enhance respiratory pathogen cell numbers, Transwell assays were employed in which C. albicans biofilms grown in the upper chamber, were physically separated from bacteria (E. coli, S. aureus, or MRSA) in the lower chamber. A 1 ml inoculum of C. albicans (1.0 x 10 6 cells/ml in RPMI) was added to the upper chamber (Costar Transwell 12 mm, 0.4 μm Polyester Membrane, Corning NY) and incubated for 4 hours to allow initial C. albicans biofilm formation. The C. albicans biofilm was washed three times with PBS and the lower chamber was inoculated with 2 ml of bacteria (5 x 10 6 cells/ml in BHI) before replacing the C. albicans-coated upper chamber containing 1 ml BHI. Transwells were incubated for a further 4 hours before washing both chambers three times with PBS and adding fresh BHI (1 ml upper chamber and 2 ml lower chamber). Biofilms were incubated for a further 24 hours to allow maturation before bacterial biofilm quantification by qPCR as outlined above. For control experiments, the upper chamber was not inoculated with C. albicans, but all other steps were undertaken as outlined above.
No Transwell experiments were undertaken with P. aeruginosa, as cell numbers of P. aeruginosa did not increase in the dual-species C. albicans-P. aeruginosa biofilms that we studied.

Dual-species biofilm preparation for TIRF microscopy
Overnight cultures of C. albicans and mCherry-labelled E. coli, GFP-labelled S. aureus, or GFP-labelled MRSA were prepared as outlined above and biofilms were formed in μ-Slide 8 well ibiTreat chamber slides (ibidi, Germany) by inoculating wells with 100 μl C. albicans laboratory strain or C. albicans gcn5Δ/Δ hyphal-deficient mutant (1.0 x 10 6 cells/ml in RPMI). Wells were incubated for 4 hours to allow initial C. albicans biofilm formation. The biofilm was washed with PBS to remove planktonic cells and RPMI, prior to inoculation with 100 μl mCherry-or GFP-labelled bacteria (5.0 x 10 6 cells/ml in BHI). Bacteria were allowed to adhere to the C. albicans biofilms for 4 hours to facilitate initial dual-species biofilm formation. Images were acquired by TIRF microscopy (Leica UK) in the epifluorescence mode using a Leica EL6000 external light source for fluorescent images and LED lamp for bright field images. Fluorescent and bright field images were overlaid using LAS-X software (Leica Application suite).

Dual-species biofilm treatment with liposomal amphotericin and subsequent quantification by PMA-qPCR
To determine the effect of antifungal treatment on bacterial numbers in dual species biofilms, liposomal amphotericin (1 μg/ml; reflecting the clinical breakpoint recommended by EUCAST) was added to C. albicans (NCTC 3179) or clinical isolate of C. albicans (CA 239SB) inoculum preparations, before addition to the wells of the microtitre plates (S1B Fig). The remaining steps for dual-species biofilm were as outlined above, the only exception being that after the last washing step at each stage of the protocol, 1 μg/ml liposomal amphotericin was added to the BHI media (S1B Fig). Following biofilm treatment with liposomal amphotericin, it was important to quantify only living cells within the biofilm and thus a PMA (Biotium Inc., California, USA) qPCR protocol was employed as previously described by us [29]. Briefly, following biofilm detachment from microtitre plates, as outlined above, the DNA-intercalating agent PMA (200 μM) was added to each tube and incubated at 37˚C for 5 min, prior to photoactivation with a broad-spectrum LED flood light [32]. PMA binds to DNA from dead cells and prevent its amplification by qPCR, thereby allowing quantification of DNA from living cells only. DNA was then extracted and quantified by qPCR using an Mx3005P qPCR System (Agilent Technologies, California USA) with the reaction conditions outlined in S1-S5 Tables (Supporting information).

Statistical analysis
Datasets were analysed by nonparametric Mann-Whitney statistical tests or Kruskal-Wallis tests for multiple comparisons, as detailed in figure legends. Data from a total of three independent experiments were analysed for each dataset. Statistical analysis and graphing of data was performed using GraphPad Prism version 8 for Windows. A p value of <0.05 was considered statistically significant.

C. albicans early biofilms enhance cell numbers of selected respiratory pathogens
We initially studied whether the presence of early biofilms of C. albicans could enhance cell numbers of P. aeruginosa, E. coli, S. aureus or MRSA (S2 Fig). In C. albicans-P. aeruginosa dual-species biofilms, P. aeruginosa cell numbers (4.3x10 6 ) were not significantly different to those in axenic P. aeruginosa biofilms (3.4x10 6 ) (S2A Fig) and it was not studied further. However, numbers of E. coli (8.9x10 6 ) were significantly increased in the presence of early C. albicans biofilms compared with axenic E. coli biofilms (3.0x10 6 ) (S2B Fig). Likewise, S. aureus numbers were significantly increased in dual-species (2.6x10 7 ) compared with axenic biofilms (1.2x10 7 ) (S2C Fig). To demonstrate that results obtained with laboratory strains were also applicable to clinical isolates we showed that cell numbers of a clinical isolate of MRSA were significantly increased when grown in the presence of a clinical isolate of C. albicans (2.

Contact with C. albicans early biofilms is required to enhance cell numbers of respiratory pathogens
To determine whether contact between C. albicans and bacteria was required for the increased bacterial cell numbers observed, C. albicans early biofilms were separated from respiratory pathogenic bacteria in Transwell assay experiments. In the absence of contact between C. albicans and bacteria, no increase in E.coli, S. aureus or MRSA cell numbers was observed (Fig 1) and indeed MRSA numbers were shown to decrease.

Dual species biofilm formation with C. albicans wild type and hyphaldeficient mutant
To confirm the presence of C. albicans biofilms after 4 hours initial biofilm formation, we demonstrated the presence of extracellular polymeric matrix, using SYPRO™ Ruby Biofilm matrix stain (S3 Fig). Having established the presence of any early biofilm, we wanted to determine if hyphae had a role in enhancing bacterial numbers following inoculation of respiratory pathogens on the pre-formed C. albicans biofilm. TIRF microscopy of wild type and hyphaldeficient C. albicans showed attachment of E. coli, S. aureus and MRSA following 4 hours initial adhesion (Fig 2A-2F). To further elucidate the role of C. albicans hyphae, we quantified cell numbers of E. coli, S. aureus and MRSA after 4 hours initial adhesion to wild type or hyphal-deficient mutant C. albicans biofilms and then after an additional 24 hours (corresponding to further maturation of the dual-species biofilms). Following the initial bacterial adhesion period to the pre-formed C. albicans biofilms, no significant differences in bacterial numbers attached to wild type or hyphal-deficient mutant C. albicans were observed. Following dual species biofilm maturation for a further 24 hours E. coli, S. aureus and MRSA numbers increase significantly in biofilms with wild type C. albicans. However, during the maturation phase with the hyphal-deficient C. albicans only S. aureus and MRSA increased significantly, with E. coli numbers failing to do so (Fig 2G-2I).

Targeting C. albicans with liposomal amphotericin in dual-species biofilms decreased S. aureus and MRSA cell numbers
In view of our results showing that C. albicans early biofilms enhanced E.coli, S. aureus and MRSA numbers, we were prompted to test a novel approach aimed at decreasing pathogenic bacterial cell numbers by targeting the architecturally supporting micro-organism, C. albicans, with the antifungal drug liposomal amphotericin. Following treatment of dual-species with liposomal amphotericin, no significant reduction in E.coli cell numbers was observed (Fig 3A). However, liposomal amphotericin treatment of dual species C. albicans-S. aureus biofilms resulted in significant reduction of S. aureus (Fig 3B) cell numbers. Moreover, in dual-species biofilms containing clinical isolates of C. albicans and MRSA, liposomal amphotericin treatment significantly reduced MRSA numbers (Fig 3C). Cell numbers in axenic S. aureus biofilms were not significantly altered by liposomal amphotericin treatment (Fig 3D), suggesting it did not have a direct effect on S. aureus within the biofilms.

Discussion
C. albicans is often overlooked as a bystander in polymicrobial biofilm-associated infections. However, using qPCR to determine bacterial cell numbers within dual-species biofilms [29] we showed that C. albicans early biofilms enhanced cell numbers of S. aureus, MRSA and E. coli, but not P. aeruginosa. Our results, which focus on quantifying bacterial cell numbers, agree with previous studies in which C. albicans was shown to enhance S. aureus dual-species biofilm formation in the presence of serum [33,34] and that an antagonistic relationship exists between P. aeruginosa and C. albicans [35,36].
Using a Transwell assay, we demonstrated that contact between early C. albicans biofilms and E. coli, S. aureus or MRSA was required to enhance bacterial cell numbers. Previously, using a Transwell assay Harriott & Noverr [33] showed contact requirement when C. albicans and S. aureus were inoculated simultaneously in Transwell upper and lower chambers respectively. In our Transwell assay (as in our biofilm model) C. albicans was inoculated first (upper chamber) to allow early biofilm development before bacterial inoculation (lower chamber). Despite the presence of an early C. albicans biofilm, soluble factors from the early biofilm did not support increased numbers of bacteria. Our results therefore concur that contact is required between C. albicans and S. aureus [33] and also between C. albicans and E. coli or MRSA, even when an early C. albicans biofilm is present. The decreased cell numbers of MRSA observed in the Transwell assay results, highlight that additional factors, such as the production of farnesol by C. albicans [37], which is recognised to have inhibitory effects against MRSA [38], may be more evident in the absence of direct fungal-bacterial contact.
In the presence of direct contact, the architecture of C. albicans has been reported to have an important role in polymicrobial biofilms [8,23,34,39,40]. C. albicans adhesion proteins such as the agglutinin-like sequence 3 protein (Als3p) [41,42], as well as Staphylococcal adhesins such as fibronectin binding protein B (FnpB), S. aureus surface protein F (SasF) or a putative N-acelymuramoul-1-alanine amidase (Atl) [43,44] may contribute to physical interactions in dual-species biofilm formation. Although our results do not appear to concur with previous work on the importance of the hyphal adhesion protein Als3p in S. aureus adhesion, multiple additional C. albicans adhesins are likely to have roles in biofilm formation [45]. Moreover, yeast wall protein 1 (Ywp1), a protein downregulated during filamentation [46] and previously reported to regulate dispersion in C. albicans, has recently been shown to function in maintaining adhesion following initial attachment [45]. Thus, biofilm formation is likely to involve multiple adhesins, with complex temporal and spatial roles in the initial attachment and biofilm maturation stages. Furthermore, although mediators such as prostaglandin E2 (PGE2) have been reported to enhance C. albicans-S. aureus biofilm formation [47], the concentration of PGE2 produced by a 24 hour culture of C. albicans has been shown to be insufficient to do so [47]. Thus, while soluble factors may enhance bacterial numbers in maturing dual-species biofilms, our results suggest that during the initial stages of dual-species biofilm formation, contact is essential.
TIRF microscopy of wild type or hyphal-deficient mutant C. albicans showed attachment of E. coli, S. aureus and MRSA in early dual-species biofilms. Quantification of bacterial numbers in early and maturing dual-species biofilms showed that hyphal-deficient C. albicans was capable of significantly enhancing S. aureus and MRSA, but not E. coli numbers, suggesting that subtle differences may exist in the ability of the C. albicans strains to increase bacterial numbers in the maturing biofilms.
It is well recognised that liposomal amphotericin has enhanced efficacy against C. albicans biofilms [48,49] and as expected, C. albicans numbers were decreased in all dual species biofilms studied, following liposomal amphotericin treatment (results not shown). Interestingly, we found that liposomal amphotericin treatment of dual-species C. albicans-S. aureus or C. albicans-MRSA biofilms also resulted in a significant reduction in S. aureus and MRSA cell numbers, despite their being no direct effect of the antifungal on S. aureus axenic biofilms. Notwithstanding some previous indications of potential synergistic relationships between C. albicans and S. aureus [35,50,51], antifungal treatments have not been actively pursued as a potential mechanism to reduce bacterial burden. Several elegant studies on C. albicans dual-species biofilms have previously reported that fungal cells may modulate the action of antibiotics [33,39,40,52]. However, no studies to date have successfully targeted C. albicans as a means of reducing bacterial cell numbers in dual-species biofilms. It had previously been suggested that in the absence of a known antibacterial effect of miconazole against S. aureus, the clinical success of topical treatment with the antifungal, miconazole, could potentially be attributed to direct activity against the C. albicans biofilm meshwork, which could destabilise S. aureus colonization [42]. However, more recently, the imidazole antifungal miconazole has indeed been shown to have direct efficacy against S. aureus [53] and therefore the clinical efficacy of topical miconazole could involve direction antibacterial action. What we show in the current study, which has not been reported previously, is that targeting C. albicans with liposomal amphotericin in dual-species biofilms reduces S. aureus and MRSA numbers, without direct antibacterial/off-target effects on S. aureus.
Interesting, liposomal amphotericin treatment of dual-species C. albicans-E.coli did not reduce E. coli numbers. Our data on dual-species biofilm development in either wild type or hyphal-deficient C. albicans, showed that E.coli numbers failed to increase significantly during biofilm maturation with the hyphal-deficient mutant, suggesting that hyphae may have a more prominent role in dual-species C. albicans-E.coli biofilm formation. With this in mind, it is possible that, following liposomal-treatment, dead hyphae may still be able to provide architectural support for E. coli, a phenomenon that has been described in other fungal species [54]. Such an effect could potentially contribute to the lack of reduction in E. coli numbers that we observed following liposomal-treatment of C. albicans-E.coli biofilms.

Conclusions
In conclusion, we demonstrate that early C. albicans biofilms facilitated increased numbers of E. coli, S. aureus and MRSA in dual-species biofilms via direct contact. Furthermore, treatment of dual-species biofilms with liposomal amphotericin significantly reduced S. aureus and MRSA cell numbers but not E. coli numbers. Given the importance of fungal-bacterial biofilms in a wide range of human diseases [55], the identification of C. albicans as a target microorganism in polymicrobial biofilm infections may have important clinical consequences. Targeted treatment of early C. albicans biofilms could be developed to reduce not only fungal but also bacterial burdens in polymicrobial biofilm infections. The controlled release [56,57] of surface-bound antifungal drugs may prove useful in this respect. To demonstrate the presence of an early biofilm containing extracellular matrix, chamber slides were stained with SYPRO™ Ruby Biofilm Matrix Stain following 4 hours incubation with C. albicans (see S1 Methods). Image acquired using a Leica TCS SP8 confocal laser scanning microscope (Leica, UK). To preserve the image data (without modification) images were processed in 3D, using LAS-X software (Leica Application suite), for addition of a 3D scale. (DOCX) S1