Deidentified strains from a collection of 20 C. albicans strains ( Table 1 ) isolated from patients with systemic candidiasis (UNMC strains), 20 strains from a previously published work (Hirakawa et al., 2015), and the type strain SC5314 were used in this analysis. All strains were grown on yeast extract-peptone-dextrose (YPD), 10% FBS media with glucose (solid only), 10% FBS in YPD (liquid only), Lee’s media (Lee et al., 1975), RPMI media with 2.1 mM L-glutamine and buffered with 165 mM MOPs, and spider medium (10 g D-mannitol, 10 g nutrient broth, 2 g K₂HPO₄, in 1 L of H₂O) in both liquid or on solid agar plates as previously described (Azadmanesh et al., 2017). For yeast-like cell growth, strains were grown in YPD at 30°C and for filamentous growth strains were grown at 37° C in 10% FBS, Lee’s, RPMI-MOPS, and spider media.

Cells from a single colony of each of the UNMC strains were used to inoculate overnight YPD cultures. Genomic DNA was extracted from these cultures using the Smash and Grab method (Hoffman and Winston, 1987). DNA was measured by Nanodrop and all genomic DNA was diluted to a concentration of 250 ng/µL to proceed with PCR for Multilocus sequence typing (MLST) analysis.

Primers and housekeeping genes for MLST analysis in C. albicans were previously optimized by others in the field ( Table S1 ) (Bougnoux et al., 2003). 250 ng of genomic DNA was used as a template and DNA for each gene was amplified using the Q5 polymerase (New England Biolabs). The amplification began with an initial denaturation at 98˚C for 3 minutes, followed by 30 cycles of denaturation at 98˚C for 10 seconds, annealing at 54˚C for 30 seconds, and extension at 72˚C for 1 minute, and completed with a final extension at 72˚C for 6 minutes.

Amplicons were confirmed by DNA gel eletrophoresis. The successful PCR products were then purified using a GeneJet PCR purification kit (Fisher Scientific). 1-2 ng of purified PCR product were submitted to the UNMC sequencing core for Sanger sequencing using the primers listed in Table S1 .

Consensus sequences were generated from the Sanger sequence results using Geneious Prime software (version 2022.2.2). The ends of the forward and reverse sequences were trimmed to the core sequence identified in previous work (Bougnoux et al., 2003; Odds and Jacobsen, 2008), and the sequences were analyzed for heterozygosity. Heterozygous bases were determined if two peaks were at least 50% of the height of the initial base call and there were only two peaks (Odds and Jacobsen, 2008). De Novo assembly with high sensitivity and slow assembly was used to generate the final consensus sequence.

The consensus sequence for each gene in each strain was then entered into the MLST database (Jolley et al., 2018) to determine if the sequence was a known allele within the database. For strains with heterozygous alleles, the heterozygous genes were entered into the database using the degenerate nucleotide code to designate the heterozygous alleles. Novel allele sequences were trimmed to the MLST length and submitted for addition to the database. Allelic profiles for each strain were entered into the MLST database to compare against known strains.

In addition to the use of the MLST database for allele analysis, sequences for each gene were concatenated together for each clinical strain for phylogenetic analysis. Sequences were concatenated in the following order: AAT1, AAC1, ADP1, MPIb, SYA1, VPS13, and ZWF1. The concatenated sequences for the novel strains were compared to concatenated sequences from known reference strains ( Table S2 ), and a dendrogram was generated using the unweighted pair group method with arithmetic mean in Geneious Prime.

Clinical strains were assigned a clade designation based on the reference strains they were clustering with in accordance with methods established by others (Tavanti et al., 2003; Bougnoux et al., 2004; Odds et al., 2007; Odds, 2010).

Overnight cultures from a single colony were grown in 3 mL of YPD at 30°C, shaking at 200 rpm in a Benchmark Incu-Shaker 10L. The cells were washed twice in 1 mL of phosphate buffered saline (PBS) pH 7.4 and resuspended in 1 mL of PBS.

For liquid assays, 10µL of cell suspension was added to 2mL of pre-warmed media (10% FBS, Lee’s, RPMI, and Spider) in 35mm petri dishes with 14mm microwell glass bottom (MatTek Corporation) and incubated with shaking (100 rpm) at 37°C for 2 hours. 10µL of washed cells were added to a glass slide for control phenotypes. Cells were imaged on a Zeiss Axiovision fluorescence microscope at 20X magnification. Five images were collected from each assay. For the UNMC strains, each strain was tested in triplicate with 5 images collected for each replicate.

For traditional solid assays, 1µL of washed cell suspension was plated in triplicate on agar plates (YPD, 10% FBS, Lee’s, RPMI, and Spider). The plates were incubated for five days at 37°C, and colony edges were imaged at 4X on an Evos FL inverted microscope. Each strain was tested with 3-5 replicates, with one image per replicate. For the short filamentation assays, 10 µl of the washed cell suspension and 40 µl of autoclaved dH₂0 were spread on the surface of pre-warmed agar plates (YPD, 10% FBS, Lee’s, RPMI, and Spider) using 4 mM glass beads. The plates were incubated at 37˚C for 3 hours and imaged at 10X on an Evos FL inverted microscope. Each strain was tested in triplicate, with 5 images from each replicate.

Strains from the solid assays were scored as previously described (Azadmanesh et al., 2017). Briefly, each strain was scored from 0-4 on the ability to form filaments compared to the control strain SC5314. A score of 0 correlated to abnormal morphology and a score of 4 indicated a morphology similar to the control. To obtain the average score for each strain and condition, all images were scored independently by 3-5 scorers. Computational methods were utilized to score liquid filamentation, as described in the supplemental methods. An average score of 2.5 or below indicated a severe phenotype defect.

MLST analysis was used to characterize the genetic relatedness of clinical C. albicans strains collected from patients with systemic infections at Nebraska Medicine. Fragments of the genes AAT1, AAC1, ADP1, MPIb, SYA1, VPS13, and ZWF1 were amplified in order to compare allelic sequences to the C. albicans MLST database (Jolley et al., 2018). Exact matches for the allele sequences were found in the database for most gene fragments, except for 4 genes in 3 strains, which were added to the MLST database ( Table 1 , highlighted numbers). Strain B1559-15 had a novel allele sequence for VPS13 and was assigned allele number 313 in the MLST database. B687-15 had novel allele sequences for both SYA1 and VPS13 and these were assigned allele numbers 240 and 310, respectively. Lastly, strain B1486-15 had a novel allele sequence for MPIb and was assigned 176 in the MLST database. For known allele sequences, the allele number in the MLST database was used to generate allelic profiles for each strain ( Table 1 ).

Two approaches were used to identify clade designations for the UNMC clinical strains. The allelic profiles for each strain were entered into the MLST database to search for known Strain Types (ST). Six strains in the set had STs with clade designations ( Table 1 , unhighlighted numbers). The allelic profiles of the remaining 14 strains were submitted to the MLST database. These strains were assigned unique strain ID number as well as STs upon addition to the database, which are indicated in Table 1 (highlighted numbers). In addition, the UPGMA algorithm was used to generate an unrooted dendrogram using 20 clinical isolates (UNMC strains) and 224 representative reference strains to identify the clade designations of the unknown clinical isolates ( Figure 1 ). The reference strains spanned a broad genetic background and included strains from all 17 clades and Singletons identified by Odds et al. (Odds and Jacobsen, 2008). Clades 1, 3, and 2 are represented by 30%, 15%, and 10% of the UNMC clinical isolates, respectively ( Figure 1 ; Table 1 ). The clade 1 data is proportional to the 35% of infections from clade 1 seen nationally within the United States (Soll and Pujol, 2003).

The ability of C. albicans to filament has been linked to pathogenesis for systemic infections. However, there are a variety of assays used to measure filamentation, and genetic requirements for filamentation vary between conditions (Azadmanesh et al., 2017; Wang et al., 2021; Sharma et al., 2023). We therefore wanted to look at the filamentation profile of our clinical strains across a number of in vitro filamentation conditions. As a comparison, we also profiled filamentation in a clinical strain set that was previously sequenced, representing a variety of infection sites (Hirakawa et al., 2015). We hypothesized that strains collected from systemic infections would exhibit robust filamentation across the filamentation assays, mirroring the phenotype of the type strain SC5314, but that we might see variability in the filamentation phenotype of strains collected from other sites.

Clinical isolates and the control strain SC5314 were tested for filamentation across 8 inducing conditions in solid and liquid media as well as non-inducing solid and liquid conditions. Strains in solid media were scored from 0-4, with 0 representing no filamentation and a score of 4 representing filamentation as robust as the control strain. Strains in liquid media were scored computationally, normalized to the SC5314 filamentation score (set at 4) for each condition. Strains with filamentation scores of 2.5 or below in either liquid or solid media were considered defective in filamentation. Contrary to our hypothesis, both the systemic isolates from UNMC and the Hirakawa et al. strain set displayed variable filamentation profiles, which did not look different from non-systemic strains ( Figure 2A ; Table S3 ).

Filamentation phenotypes varied between strains, but there were a few broad observations. Almost all of the clinical isolates of C. albicans exhibited severe filamentation defects on solid media, with only P37039 and B527-15 filamenting to some degree across all solid conditions ( Figures 2 , 3A , S2 ). It was notable that these strains are hyperfilamentous on non-inducing solid media and also filamented across all liquid conditions ( Figures 2 , 3A, B ; Table S3 ). One other strain, P37037, also filamented on 3 of the 4 solid conditions ( Figures 2 , 3A, B ). All three of those strains are, like type-strain SC5314, members of Clade 1, although we did not find any correlation between clades in terms of filamentation phenotypes ( Figure S8 ). There were limited representatives of each clade, however, and it is possible that clade-specific patterns might become apparent with a larger sample size. Filamentation was much more robust in liquid media, and indeed, a number of strains exhibited filamentation scores above 4 in Lee’s and Spider media (it should be noted that our solid scoring did not allow for scores above 4) ( Figure S5 ). Only two strains, P37037 and P94015 were defective in filamentation in all liquid inducing conditions tested ( Figures 2 , 3A ). P94015 was unable to filament in any condition, while P37037, as mentioned above, was hyperfilamentous in non-inducing solid media and across 3 of 4 filamentation conditions.

When we focus on filamentation conditions rather than the strains, a few clear trends emerge. There is little variability between solid conditions in the number of strains that are considered filamentous in each condition, ranging from 7 to 10 strains ( Figure 3C ). The number of filamentatous strains in liquid filamentation conditions were much higher. Filamentation was more robust across strains in Lee’s and Spider media (38 strains with scores above 2.5) compared to FBS (33 strains) and RPMI (32 strains) ( Figure 3C ). The significance of this variation in score is unclear, but it is worthwhile to note.

One thing that differentiates the standard liquid and solid filamentation assays in C. albicans, besides the solidity of the media, is the time between induction initiation and analysis of filamentation. Liquid filamentation is measured just a few hours after induction, while the standard solid filamentation assay is measured days after induction. Therefore, standard liquid assays determine initial response to induction in liquid media while standard solid assays measure sustained filamentation on solid surfaces. We found in previous work (Creger and Blankenship, 2018), that solid filamentation can be measured at the same timepoints used in liquid assays. We also found that two of the clinical strains tested above (P87 and P76067), which do not filament in standard solid filamentation assays, do filament at these earlier time points (Creger and Blankenship, 2018). By examining filamentation in solid media at the same timepoint that we examined liquid filamentation, we eliminated the confounding time factor in our comparisons. We hypothesized that clinical strains from systemic infections would filament in solid media at these earlier timepoints as robustly as they do in liquid filamentation conditions.

For these shortened filamentation assays, we focused on filamentation in the UNMC set of clinical strains. These strains were all isolated from patients with systemic infections. The expectation was that all of these strains should be filamentatous in vivo and thus filamentous in vitro in assays that model in vivo induction. Seven of the clinical strains (B212-12, B444-12, B564-15, B733-15, B1257-15, B1486-15, and B1559-15) filamented across all solid conditions in this shortened solid filamentation assay and seven of the clinical strains (B404-15, B421-15, B618-15, B808-15, B1091-15, B1168-15, and B2527-12) had filamentation defects across all solid inducing media ( Figures 4A–C , S3 ). Strain 527-15 was notable because it did filament in the standard solid assay but was defective in most shortened solid filamentation assays ( Figure 4D ). It is possible that this strain responds more slowly to filamentation signals or that it was growing, and thus filamenting, more slowly. Clearly, it can filament in solid media as evidenced in the longer assay ( Figures 2 , 4D ). The rest of the strains showed varying patterns of filamentation in the shortened assays. Overall, it appears that a number of the clinical strains can filament on solid media, but few are able to maintain that filamentation over a period of days.