Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Exploring ecological modelling to investigate factors governing the colonization success in nosocomial environment of Candida albicans and other pathogenic yeasts


Two hundred seventy seven strains from eleven opportunistic species of the genus Candida, isolated from two Italian hospitals, were identified and analyzed for their ability to form biofilm in laboratory conditions. The majority of Candida albicans strains formed biofilm while among the NCAC species there were different level of biofilm forming ability, in accordance with the current literature. The relation between the variables considered, i.e. the departments and the hospitals or the species and their ability to form biofilm, was tested with the assessment of the probability associated to each combination. Species and biofilm forming ability appeared to be distributed almost randomly, although some combinations suggest a potential preference of some species or of biofilm forming strains for specific wards. On the contrary, the relation between biofilm formation and species isolation frequency was highly significant (R2 around 0.98). Interestingly, the regression analyses carried out on the data of the two hospitals separately were rather different and the analysis on the data merged together gave a much lower correlation. These findings suggest that, harsh environments shape the composition of microbial species significantly and that each environment should be considered per se to avoid less significant statistical treatments.


Candida bloodstream infection (BI) is an important cause of morbility and mortality in health care settings and represents the fourth cause of nosocomial sepsis in the USA and in most developed countries1,2. Candidemia is responsible for unacceptable percentages of attributable and overall mortality rate ranging from 30–81% and from 5–71%3, respectively. The incidence of candidemia rose in the last decades of the 20th century due to several risk factors4. Candida albicans remains the most common species causing BI, followed by several “non Candida albicans Candida species” (NCAC) among which Candida glabrata, Candida parapsilosis and Candida tropicalis are usually the most common and show increasing incidence5,6,7. The various species are a further factor to interpret the origin of infection, in fact C. albicans, C. glabrata and C. tropicalis are considered predominantly commensal and therefore more present in cases of endogenous infections. On the contrary, species found in the natural and anthropic environment, as C. parapsilosis, Meyerozyma guilliermondii (telomorph of C. guilliermondii) and Wickerhamomyces anomalus, are more present in exogenous infections8. The recent finding that medical and food isolates of M. guilliermondii cluster differently, according to the ITS and LSU markers, poses the question on whether the exogenous infection is caused by strains of the nosocomial environment or of other niches9. The increasing frequency of NCAC species has been extensively reported in the last years4,6,10,11,12,13 with significant epidemiological and ecological differences among various geographic areas7,14. This situation represents a serious threat, complicated by a significantly lesser knowledge of biofilm and resistance mechanisms in NCAC then in C. albicans.

The ability of some fungal species to form biofilm is considered an important factor for their persistence on medical devices15,16 and in general in the nosocomial environment17, however, the actual extent of the fungal persistence in the hospital is still unclear. Furthermore, the fungal biofilm displays higher resistance to drugs18,19,20, with a complex and yet not totally understood mechanism7, involving the Hsp90, a protein also responsible for cell dispersion21, making the biofilm a system to persist in harsh environments and resist to the associated stresses.

From what briefly reported above, the biofilm forming ability represents a severe risk factor1,22 and adds more problems to the need of a timely and appropriate therapy to reduce the mortality23.

Altogether, Candida infections represent a serious problem and their ability to form biofilm seems to represent not only a medical, but also an ecological problem. In fact, the infecting cells can be present in different niches spanning from the devices to the surfaces, the air, some foods and the patients themselves. The cell circulation in the environment is an essential point to understand the complex ecology represented by the interaction of fungal cells with patients, different substrates and drugs. Furthermore, only a good ecological insight can lead to the actual possibility of “catching” these pathogens in their actual niches before the infection. In fact, once the infection occurred there are relatively few therapeutics to treat these diseases successfully, whereas environmental treatments with harsh biocidal compounds can be effective and decrease significantly the incidence and the mortality caused by these fungi.

With the above rationale, the present study has been designed around two hypotheses: i. the hospital and the various departments, i.e. specific environments, are key factors for the frequency of candidemias; ii. the ability to form biofilm has a measurable effect on the incidence of these diseases.

For this purpose, 277 strains of eleven Candida species have been isolated from the various departments of two Italian hospitals (Pisa and Udine) 450 km apart, identified at the species level and tested for biofilm formation.


Distribution of the studied characters

Species in the hospitals and wards

Four species were isolated in both hospitals: C. albicans, C. glabrata, C. parapsilosis and C. tropicalis (Fig. 1, panel a,b). In the Pisa hospital were isolated specifically Pichia kudriavzevii (telomorph of Candida krusii), C. rugosa and some strains yet to be attributed to a possibly new yeast species (hereinafter referred to as Candida. spp) (panel a). The species isolated uniquely in Udine were Clavispora lusitaniae (telomorph of Candida lusitaniae), C. sake, Cyberlindnera jadinii (telomorph of Candida utilis) and Meyerozyma guilliermondii (telomorph of Candida guilliermondii) (panel b).The four most represented species present in both hospitals accounted for 86.36% and 96.81% in Pisa and Udine hospitals, respectively. As expected, C. albicans was the most frequently isolated species (panel c), absent only from the Rehabilitation department of Pisa (panel d). The second most frequent species was C. parapsilosis, followed by C. glabrata and C. tropicalis. These data are in good agreement with the current literature24,25. The departments with the highest incidences were the specialized medicine, Surgery and ICU in Pisa, while general medicine in Udine was by far the ward with more isolations, followed by specialized medicine, Surgery and ICU (panel d,e). Merging those data produced a picture very similar to that described for Udine from which came 188 out of the 276 strains analyzed in this study (panel f).

Figure 1

Species presence in the two hospitals studied and relatively frequency of isolation of the different species in the hospitals wards.

Biofilm forming strains in hospital departments

The various departments in the two hospital under study showed different frequencies of biofilm forming strains (Fig. 2). Namely, in Pisa the Rehabilitation ward had only non-biofilm forming isolates, whereas the other four departments showed frequencies ranging from 60 to more than 80% (panel a). The situation in Udine was more variable with frequencies of biofilm forming strains ranging from 20% (Rehabilitation) to 80% (Surgery) (panel b). Merging the data from the two hospitals gave a synthetic view with biofilm forming strain frequency ranging from less than 20% (Rehabilitation) to the 75% of the Surgery departments (panel c).

Figure 2

Relative occurrence of biofilm forming capability in different hospital departments and distribution of biofilm forming strains among species.

Biofilm forming strains and species

The four prominent species (C. albicans, C. glabrata, C. parapsilosis andC. tropicalis) showed different levels of biofilm forming ability, whereas the lesser present species (C. rugosa, P. kudriavzevii, C. jadinii, C. sake, C. lusitaniae and M. guilliermondii) did not produce biofilm (Fig. 2, panel f). C. albicans biofilm forming strains (hereinafter referred to as BF, in contrast with non BF referred to as NBF) were 97.78% and 87.71% in Pisa and Udine, respectively (panels d,e). Since these figures are particularly high, the biofilm formation test was repeated with two different methods using XTT and Tetrazonium Blu as indicators, obtaining no significant differences (data not shown). C. glabrata showed quite different frequencies of BF strains in Pisa (25%) and in Udine (4.3%), similarly C. tropicalis had all BF strains in Pisa and some 55% in Udine. Finally, C. parapsilosis showed 35% and 57% in Udine and Pisa, respectively. In general, Pisa had 75% BF strains and Udine 63.63%.

Contingency analysis studied characters

Species vs. hospital departments

Different opportunistic yeast species can be distributed randomly in the various hospital departments or show specific preferences for the environment represented by the various wards. In order to test the null hypothesis, i.e. that species are distributed randomly within the hospital, a two-way contingency analysis was carried out, following the rationale described in Legendre & Legendre26, as outlined in the Experimental Procedures section. This analysis is the gold standard in ecological studies when non continuous qualitative data are employed, as in our case and allows to calculate the probability associated to the χ2 (p-value) to accept or reject that the association between species and hospital department are independent. When two descriptors are non-independent, then their combination is indicative of some sort of specific occurrence. This test was carried out considering the two hospitals separately and then by merging all data for a joint analysis (Table 1).

Table 1 Probabilities associated with the χ2 statistics calculated on contingency tables.

The general χ2 test for the two-way contingency table, obtained with the data of the Pisa hospital, showed no statistical significance (p = 0.32), indicating that the distribution of the yeast species was largely due to random effect. However, the frequency of C. rugosa in the ICU and of Candida. spp in the Rehabilitation ward, were significantly non independent with 0.0489 and 0.0049 p-values, respectively. These data suggested that these species have a certain level of specificity for these environments. Almost significant (p = 0.11741) was the frequency of isolation of C. tropicalis in the Surgery department (panel a).

The general χ2 test carried out with all the data from the Udine hospital gave 0.1189 p-value indicating independence between species and wards, assuming p = 0.10 as the minimum for statistical significance. However, the frequency of isolation of M. guilliermondii in Oncohematology and Rehabilitation showed high χ2 values corresponding to p = 0.081 and p = 0.0011, respectively.

C. tropicalis in Onchohaematology showed a significantly non-independence with p = 0.0002. Interestingly, three species (C. albicans, C. parapsilosis, C. tropicalis) in the Surgery ward displayed elevated χ2 values with p values ranging from 0.17 to 0.22, indicating that their frequency of isolation cannot be entirely ascribed to a randomness (panel b).

Merging the data of the two hospitals, the general χ2 test was significant with p = 0.00819, indicating that the hypotheses of random distribution of the species in the departments of the two hospitals can be rejected. In fact, several species-department combinations showed high and significant χ2 values, such as C. tropicalis in Surgery and Oncohaematology and M. guilliermondii in Rehabilitation and Oncohaematology (panel c). The difference between the χ2 tests of the two hospitals considered separately and then jointly can be ascribed to the fact that merging all data together produces larger numbers and a consequently more solid statistics. Interestingly, expectable random effects explain the frequency of isolation of the various species (null hypothesis) only in the specialized medicine departments. C. tropicalis was overrepresented in Oncohaematology and underrepresented in the Surgery department (p = 0.0504 and 0.0016 respectively). M. guilliermondii was isolated with a frequency higher than expected in Rehabilitation and Oncohaematology (p = 0.0002 and 0.022 respectively), although the absolute frequencies are low, that is typical for this rather ubiquitous species found on fruit9.

Most of the cases in which the frequency of isolation were different than expected (null hypothesis rejected) interested mostly the less representative species which account for ca. 13% in Pisa and 3% in Udine Hospitals. Based on this observation, the cases of statistically supported over- and under-isolation represent only few cases, maybe of little epidemiological importance. Yet these preferences are worth better insight, because these few cases could underline yet to be described phenomena of preferential environmental (nosocomial) colonization.

Biofilm forming strains vs. hospital departments

The question on whether the biofilm forming strains can dwell preferentially in some hospital departments was addressed with the same contingency approach as above (Table 1). The general χ2 test for the two-way contingency table, obtained with the data of the Pisa hospital, showed no statistical significance (p = 0.44), once again indicating that the combination of biofilm forming strains and hospital departments were largely due the randomness. In fact, no combination showed p values below 0.10 (panel d). On the contrary, in the Udine hospital, the χ2 test was associated with p = 0.0607, indicating that the frequencies of biofilm forming strains and the various departments of isolation cannot be considered totally independent. Although no combination showed significant p values below 0.10, some weak signal of non-independence existed for those departments with less biofilm to isolates as in Rehabilitation and Oncohaematology of the Udine hospital (p = 0.11 and 0.14, respectively) (panel e).

Merging the two datasets led to a two-way contingency table for which the hypothesis of independence cannot be rejected (p = 0.14). In general, these tests indicated the frequency of isolation of biofilm forming strains is based on the randomness, with some significant non independence for the combinations of the Udine hospital presented above (panel f).

Biofilm forming strains vs. species

The χ2 analysis of two-way contingency tables between the frequency of biofilm forming strains and species allowed to reject the null hypothesis with p = 3 × 10−8, 8.7 × 10−6 and 7.2 × 10−22 for Pisa, Udine and merged data, respectively. This extremely high statistical significance indicates that the frequency of isolation of the single species is influenced by their ability to form biofilm. More specifically, in the Pisa hospital, the random effect could be excluded for the BF strains of Candida spp. and for the NBF isolates of C. glabrata, C. tropicalis and Candida. spp. Particularly interesting the fact that C. glabrata had three times more NBF than expected on the basis of a random effect. In Udine, BF and NBF strain frequencies were significantly associated with C. albicans, C. parapsilosis and C. glabrata. Namely, C. albicans had 28% more BF than expected, while C. glabrata and C. parapsilosis had respectively 14 times more and 50% less BF strains than expected. The situation of both hospitals considered together was substantially similar to that of Udine, given the preponderance of the strains deriving from that hospital.

Modelling the biofilm vs. isolation frequencies

The biofilm formation frequency was compared with the frequency of isolation of the single species because the combination of these two characters were shown to be non-random with extremely high statistical significance. These two descriptors had 98.51% and 97.37% Pearson correlation (R) in Pisa and Udine hospitals, respectively. This was confirmed by the linear correlation analysis with R2 of 97.04% and 94.8% for Pisa and Udine, but with two totally different correlation equations. In fact, in Udine, the correlation curve could be described by the equation

in which IF and BF indicate, respectively, the isolation frequency of single species and the biofilm formation frequency. Formula 3 specifies that the frequency of isolation of each species is 94.2% of the frequency of BF strains of the same species.

In Udine, the correlation analysis produced the same R2 as in Pisa, but a different equation:

This equation indicates that in the Udine hospital the frequency of isolation of each yeast species is 108% of its frequency of BF strains. This observation indicates that indeed the ability to form biofilm plays a key role in the presence of the species in the hospital environment, but with different dynamics in Pisa and Udine. It was therefore not surprising that the correlation analysis of the Pisa and Udine data merged together produced an intermediate situation, described by Formula 5 with a lower R2 (60.94%) than those obtained with the two hospitals analyzed separately.

The differences between the two hospitals can be pin-pointed by comparing the IF/BF ratio of the single species. C. albicans showed 103% and 114% IF/BF ratios in Pisa and Udine respectively. This means that, given a frequency of BF C. albicans strains, the probability of isolating this species in Pisa is slightly lower than in Udine. C. parapsilosis had 175% and 283% IF/BF ratios in Pisa and Udine, showing that the biofilm forming ability influences the frequency of isolation more in Udine than in Pisa. A similar situation was observed for C. tropicalis with 100% and 183% IF/BF ratios in Pisa and Udine. Finally, C. glabrata showed 400% and 2300% IF/BF ratios in the two hospitals, making it the species for which the biofilm formation ability has the maximum effect on the frequency of isolation.


The distribution of opportunistic species of the genus Candida presented in this work is largely in accordance with the current literature. In fact, the number of isolates of C. albicans ranged from 51–61% in Pisa and Udine, showing figures similar to those previously reported5,25. C. tropicalis and C. parapsilosis showed 12.5% average frequency and C glabrata 10.6% with a significant difference from the 20% of the formers and the 5% of the latter recently reported in Brazil7. The seven species isolated in only one of the two hospitals showed low frequencies around 1%, again in accordance with most of the current literature27. These differences can be justified by the different geographic places where the studies were carried out and by a general increase of NCAC species vs C. albicans14, indicating that epidemiological data differ significantly over the time and the geography. The question on how the geography and the hospital ward influence these species frequencies is difficult to address, due to a literature concentrating either on aggregated data or on specific wards, with very few papers reporting the distribution within the hospital. This suggests that more detailed reports will be necessary in future to track these aspects.

Candida infections have an endogenous origin based on the growth of the cells already dwelling on or in the body of the patient as a commensal, while the exogenous origin derives from the surrounding environment28. Early studies on these aspects suggested that different areas of the hospital can be more interested to one of the two types of Candida infection29. Exogenous nosocomial infections can be triggered by the ability of these fungi to persist in the environment, by forming biofilm. Kramer and colleagues30 reported ca 4 months persistance for C. albicans and C. glabrata, whereas these figures dropped to a few days according to an older study31. The persistence of C. parapsilosis was estimated on a couple of weeks by the above two studies. There is currently little if any information on whether the cells move as pieces of biofilm (sessile cells) or as planktonic cells liberated during the maturation of the biofilm and in their way to colonize another surface, initiating a new biofilm plaque32. In both cases, dispersed cells have showed enhanced adherence and produce a more robust biofilm than planktonic cells not deriving from a biofilm33. The biofilm architecture and resistance varies among species34,35,36,37 and sometimes with the genetic setting38, making any generalization quite difficult. The whole situations seems to be further complicated by the diversity of resistance mechanisms in young and mature biofilms, whereas the former is relatively well elucidated and based on multidrug pumps, the latter seems based on a dormancy mechanism leading to the production of “persister” cells5,24,25,39. The fungal biofilm displays higher resistance to drugs18,19,20, with a complex and yet not totally understood mechanism7, involving the Hsp90, a protein also responsible for cell dispersion21. The double function of this protein suggests that the complex regulation of the biofilm formation40 is responsible for dissemination and resistance: two key factors for the success of these cells in the hospital environment.

The Biofilm forming ability has been suggested as one of the major risk factors for mortality due to C. albicans1 and in general to the other yeast species analyzed in this article7. However, it seems that, in some settings, the biofilm forming ability does not increase the mortality nor the probability of getting catheter related candidemia16.

The study of the factors facilitating the formation of biofilm is an important and little investigated aspect22 to understand whether the biofilm formation is triggered by the environmental conditions. In this paper, we have considered different hospital departments to test if they have an influence in the biofilm formation. The contingency analysis has shown that there is no differential effect of the various departments in the presence of biofilm forming strains. Some instances, such as the Oncohematology of Udine, showed less biofilm than expectable. The statistical significance of these variations is relatively good when considering the aggregated data of the two hospitals (p = 0.031), whereas there is not significance when considering the disaggregated data of Udine alone. This suggests that we have only hints of an effect exerted by some departments, but higher number of strains for each hospital will be necessary for a careful determination of the statistical significance.

In our analytical conditions, C. albicans, C. tropicalis, C. parapsilosis and C. glabrata formed biofilm in 90%, 75%, 42% and 10% of the isolates with 67% among all the studied strains. These figures are in agreement with a recent study carried out in Scotland, in which the biofilm formation was reported as quartiles of the spectrophotometric quantification after Crystal Violet staining1. The high and intermediate producers accounted for 67%, 59 and 100% of the C. albicans, C. tropicalis and C. parapsilosis strains, whereas in both cases C. glabrata produced biofilm poorly. Interestingly, the high and intermediate producers reported for these three species in this study match exactly with the 67 of biofilm formers of our investigation. The differences in the portion of biofilm formers within the various species can be due to different environmental conditions.

These high levels of BF vs NBF strains can be due to at least three causes: the environment or the niche, methods and clonality. The study of the environmental effect is of great importance in these infections, very often caused by exogenous environmental contaminations. The other two factors, briefly outlined below, can seriously interfere with the effective determination of environmental effects on biofilm formation. The Crystal Violet, the XTT and the SYTO9 methods showed correlations ranging from 0.8 to 0.4, meaning poor R2 regression values ranging from 0.64 to 0.161. This fact indicates that the three methods yield rather independent measures of biofilm formation and the results obtained with different methods can diverge significantly. The clonality can account for the same strain being isolated repeatedly as different isolates. This problem has not an easy solution, because very accurate analyses are necessary to rule out the hypothesis of two strains being identical. Even the current barcode marker ITS41 is scarcely effective differentiating isolates and combinations with more marker genes can be necessary to address the clonality problem effectively9,42. Until more light will not be shaded on this issue, we can only state that the isolates of these studies are very frequently able to form biofilm, regardless of their genetic relationships.

In the two hospital studied, the ability to form biofilm was directly correlated with the different frequency of the various isolated species according to solid linear correlations. The regression between species frequencies and biofilm formation studied separately in the two hospitals had better R2 (0.97 and 0.95 for Pisa and Udine, respectively) than when all data were merged together (R2 = 0.6).

This evidence indicates that the dynamics governing yeasts biofilm presence in the two hospitals considered are quite different. Although this situation cannot be generalized on the basis of a single study, care should be taken in epidemiological studies to analyze the data pertinent to the specific place from which they derive by condensing data because merging data from different environments (hospitals, cities etc.) might be induce serious bias.

Biofilm formation is considered of paramount importance in medicine and in a number of environmental applications. Altogether, this study has demonstrated that, in the two hospitals analyzed, the biofilm is the major factor triggering the persistence of the yeast species in these environments. A number of questions are still open, as the problem of clonality and the definition of a comprehensive working model to explain the role played by biofilm in persistence, resistance and spreading of the cells in the four most common opportunistic pathogens of the genus Candida.

The finding that biofilm formation can be an important factor to favor the presence of microbial cells in harsh environments further improves its general and applied biological importance.


Strains and growth condition

277 strains belonging to opportunistic species of the Candida genus, isolated in a clinical (medical) environment, were employed in this study (Tables S1 and S2, Supplementary materials). All strains were isolated from patient bloodcultures, with the exception of C tropicalis 6184a and 6184b isolated from peritoneal fluid, C. parapsilosis 6551 from pharyngeal swab and C. albicans 8158 and 8158/C from vascular prosthesis. Isolates are kept frozen at −80 °C in 17% glycerol. Short term storage was carried out on YEPDA (YEPD added with 1.7% agarose) at 4 °C. Strains were grown in YEPD (Yeast Extract 1%, Peptone 1%, dextrose 1% - all products from Biolife- at 37 °C with 150 rpm shaking.

Molecular analysis ITS and bioinformatics tools

Genomic DNA was extracted as indicated by Cardinali et al.43. ITS1, 5.8S, ITS2 rDNA genes were amplified with FIREPol® Taq DNA Polymerase (Solis BioDyne, Estonia), using ITS1 (5′-TCCGTAGGTGAACCTGCGG) - ITS4 (TCCTCCGCTTATTGATATGC) primers.

The amplification protocol was carried out as follows. Initial denaturation at 95 °C for 4 min, 35 amplification cycles (94 °C for 1 min, 53 °C for 1 min and 72 °C for 1 min) and final extension at 72 °C for 10 min. Amplicons were purified with the GFX PCR DNA purification kit (GE Healthcare) and subject to electrophoresis on 1.5% agarose gel (Gellyphor, EuroClone, Italy). Amplicons were sequenced in both directions with ABI PRISM technology by MACROGEN ( with the same primers used for the generation of the amplicons. Consensus sequences for each strain and trimming of the ends with low sequencing quality were carried out with Geneious R6 (v. 6.17, Biomatters, Auckland, New Zealand, ITS-based species identification was carried out with BLAST search44 in GenBank ( and refined with specialized databases, RefSeq45 and ISHAM-ITS reference database (ref. 46).

Biofilm protocol

Biofilm activity was assessed with an XTT method47 and using Resazurin with slight modification on the XTT protocol. Briefly, each strain was grown over night in bottles containing YEPD (Yeast Extract 1%, Peptone 1%, Dextrose 2% - Difco Laboratories, USA) medium, at 30 °C in an orbital shaker at 150–180 rpm and then harvested and centrifuged at 3,000 × g for 5 minutes at 4 °C. The supernatant was removed and the pellet was washed twice with PBS. Washed cells were then resuspended in RPMI-1640 medium (Sigma Aldrich), previously warmed at 37 °C, in order to obtain a final density of 1.0 × 106 cells/mL. 100 μL of this standardized cell suspension were seeded in each selected well of 96-well microtiter plate; the wells on column 12 remained unseeded, in order to act as negative background control for the subsequent steps.

The microtiter plate was closed, sealed and incubated for 24 h at 37 °C. After biofilm formation, the medium in each well was removed carefully with a multi-channel pipette, taking care of not disrupting the biofilm; each well was subsequently washed three times with PBS. After each washing step, the plate was drained in an inverted position by blotting with paper towels.

Using a multichannel pipette, 100 μl of fresh 0.001 mg/mL resazurin solution was added to each well of the drained plated, included the negative control wells, to assess the biofilm formation. After 1 h incubation at 37 °C, the plate was visually inspected to highlight the presence of pink color gradient, resulting from the reduction of the blue dye resazurin to the pink dye resorufin by living biofilm-forming cells. Both systems produced quite similar results. Strains were considered able to form biofilm when a visible color change was detectable.

Statistical data analysis

Contingency analysis and chi squared test

Contingency analysis and chi squared test is the gold standard approach in ecology for the treatment of qualitative non-continuous data, as those treated in this paper. Contingency analysis was carried out according to Legendre & Legendre26, as detailed in Supplementary materials and depicted in the four parts of Table S3.

Additional Information

How to cite this article: Corte, L. et al. Exploring ecological modelling to investigate factors governing the colonization success in nosocomial environment of Candida albicans and other pathogenic yeasts. Sci. Rep. 6, 26860; doi: 10.1038/srep26860 (2016).


  1. Rajendran, R. et al. Biofilm formation is a risk factor for mortality in patients with Candida albicans bloodstream infection—Scotland, 2012–2013. Clinical Microbiology and Infection 22.1, 87–93 (2015).

    Google Scholar 

  2. Wisplinghoff, H. et al. Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study. Clinical infectious diseases 39, 309–317 (2004).

    Article  Google Scholar 

  3. Horn, D. L. et al. Epidemiology and outcomes of candidemia in 2019 patients: data from the prospective antifungal therapy alliance registry. Clinical infectious diseases 48, 1695–1703 (2009).

    CAS  Article  Google Scholar 

  4. Sobel, J. D. The emergence of non-albicans Candida species as causes of invasive candidiasis and candidemia. Current infectious disease reports 8, 427–433 (2006).

    Article  Google Scholar 

  5. Dimopoulos, G., Ntziora, F., Rachiotis, G., Armaganidis, A. & Falagas, M. E. Candida albicans versus non-albicans intensive care unit-acquired bloodstream infections: differences in risk factors and outcome. Anesthesia & Analgesia 106, 523–529 (2008).

    Article  Google Scholar 

  6. Miceli, M. H., Díaz, J. A. & Lee, S. A. Emerging opportunistic yeast infections. The Lancet infectious diseases 11, 142–151 (2011).

    Article  Google Scholar 

  7. Sardi, J., Scorzoni, L., Bernardi, T., Fusco-Almeida, A. & Giannini, M. M. Candida species: current epidemiology, pathogenicity, biofilm formation, natural antifungal products and new therapeutic options. Journal of medical microbiology 62, 10–24 (2013).

    CAS  Article  Google Scholar 

  8. Merseguel, K. B. et al. Genetic diversity of medically important and emerging Candida species causing invasive infection. BMC infectious diseases 15, 57, doi: 10.1186/s12879-015-0793-3 (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  9. Chen, S. C. A. et al. Candidaemia with uncommon Candida species: predisposing factors, outcome, antifungal susceptibility and implications for management. Clinical Microbiology and Infection 15, 662–669, doi: 10.1111/j.1469-0691.2009.02821.x (2009).

    CAS  Article  PubMed  Google Scholar 

  10. Trofa, D., Gácser, A. & Nosanchuk, J. D. Candida parapsilosis, an emerging fungal pathogen. Clinical microbiology reviews 21, 606–625 (2008).

    CAS  Article  Google Scholar 

  11. Gácser, A., Schäfer, W., Nosanchuk, J. S., Salomon, S. & Nosanchuk, J. D. Virulence of Candida parapsilosis, Candida orthopsilosis and Candida metapsilosis in reconstituted human tissue models. Fungal Genetics and Biology 44, 1336–1341, doi: 10.1016/j.fgb.2007.02.002 (2007).

    CAS  Article  PubMed  Google Scholar 

  12. Pfaller, M. et al. Candida guilliermondii, an opportunistic fungal pathogen with decreased susceptibility to fluconazole: geographic and temporal trends from the ARTEMIS DISK antifungal surveillance program. Journal of clinical microbiology 44, 3551–3556 (2006).

    CAS  Article  Google Scholar 

  13. Papon, N., Courdavault, V., Clastre, M. & Bennett, R. J. Emerging and emerged pathogenic Candida species: beyond the Candida albicans paradigm. PLoS Pathog 9, e1003550 (2013).

    CAS  Article  Google Scholar 

  14. Kojic, E. M. & Darouiche, R. O. Candida infections of medical devices. Clinical microbiology reviews 17, 255–267 (2004).

    Article  Google Scholar 

  15. Guembe, M. et al. Is biofilm production a predictor of catheter-related candidemia? Medical Mycology 52, 407–410 (2014).

    Article  Google Scholar 

  16. Weinstein, R. A. & Hota, B. Contamination, disinfection and cross-colonization: are hospital surfaces reservoirs for nosocomial infection? Clinical infectious diseases 39, 1182–1189 (2004).

    Article  Google Scholar 

  17. Ramage, G., Rajendran, R., Sherry, L. & Williams, C. Fungal biofilm resistance. International journal of microbiology 2012, doi: 10.1155/2012/528521 (2012).

  18. d’Enfert, C. Biofilms and their role in the resistance of pathogenic Candida to antifungal agents. Current drug targets 7, 465–670 (2006).

    Article  Google Scholar 

  19. Mukherjee, P. K. & Chandra, J. Candida biofilm resistance. Drug Resistance Updates 7, 301–309, doi: 10.1016/j.drup.2004.09.002 (2004).

    CAS  Article  PubMed  Google Scholar 

  20. Robbins, N. et al. Hsp90 governs dispersion and drug resistance of fungal biofilms. PLoS pathogens 7, e1002257 (2011).

    CAS  Article  Google Scholar 

  21. Tumbarello, M. et al. Risk factors and outcomes of candidemia caused by biofilm-forming isolates in a tertiary care hospital. PloS one 7, e33705 (2012).

    ADS  CAS  Article  Google Scholar 

  22. Bassetti, M., Molinari, M., Mussap, M., Viscoli, C. & Righi, E. Candidaemia in internal medicine departments: the burden of a rising problem. Clinical Microbiology and Infection 19, E281–E284 (2013).

    CAS  Article  Google Scholar 

  23. Leroy, O. et al. Epidemiology, management and risk factors for death of invasive Candida infections in critical care: a multicenter, prospective, observational study in France (2005–2006). Critical care medicine 37, 1612–1618 (2009).

    Article  Google Scholar 

  24. Weinberger, M. et al. Characteristics of candidaemia with Candida albicans compared with non-albicans Candida species and predictors of mortality. Journal of Hospital Infection 61, 146–154 (2005).

    CAS  Article  Google Scholar 

  25. Legendre, P. & Legendre, L. F. Numerical ecology. Second English Edition Ch. 6, 207–245 (Elsevier, 1998).

  26. Corte, L. et al. Phenotypic and molecular diversity of Meyerozyma guilliermondii strains isolated from food and other environmental niches, hints for an incipient speciation. Food Microbiology 48, 206–215 (2015).

    CAS  Article  Google Scholar 

  27. Lortholary, O. et al. Worrisome trends in incidence and mortality of candidemia in intensive care units (Paris area, 2002–2010). Intensive care medicine 40, 1303–1312 (2014).

    Article  Google Scholar 

  28. Gallè, F., Catania, M. & Liguori, G. Nosocomial Candida infections: epidemiology of candidaemia. Journal of preventive medicine and hygiene 47, 119–126 (2015).

    Google Scholar 

  29. Fridkin, S. K. & Jarvis, W. R. Epidemiology of nosocomial fungal infections. Clinical microbiology reviews 9, 499–511 (1996).

    CAS  Article  Google Scholar 

  30. Kramer, A., Schwebke, I. & Kampf, G. How long do nosocomial pathogens persist on inanimate surfaces? A systematic review. BMC infectious diseases 6, 130, doi: 10.1186/1471-2334-6-130 (2006).

    Article  PubMed  PubMed Central  Google Scholar 

  31. VandenBergh, M. F. Q., Verweij, P. E. & Voss, A. Epidemiology of nosocomial fungal infections: invasive aspergillosis and the environment. Diagnostic Microbiology and Infectious Disease 34, 221–227, doi: 10.1016/S0732-8893(99)00026-7 (1999).

    CAS  Article  PubMed  Google Scholar 

  32. Nikolaev, Y. A. & Plakunov, V. Biofilm—“City of microbes” or an analogue of multicellular organisms? Microbiology 76, 125–138 (2007).

    CAS  Article  Google Scholar 

  33. Uppuluri, P. et al. Dispersion as an important step in the Candida albicans biofilm developmental cycle. PLoS Pathog 6, e1000828 (2010).

    Article  Google Scholar 

  34. Silva, S. et al. Adherence and biofilm formation of non-Candida albicans Candida species. Trends in microbiology 19, 241–247 (2011).

    ADS  CAS  Article  Google Scholar 

  35. Silva, S. et al. Biofilms of non-Candida albicans Candida species: quantification, structure and matrix composition. Medical Mycology 47, 681–689 (2009).

    ADS  CAS  Article  Google Scholar 

  36. Parahitiyawa, N. et al. Interspecies variation in Candida biofilm formation studied using the Calgary biofilm device. Apmis 114, 298–306 (2006).

    CAS  Article  Google Scholar 

  37. Hawser, S. P. & Douglas, L. J. Biofilm formation by Candida species on the surface of catheter materials in vitro. Infection and immunity 62, 915–921 (1994).

    CAS  PubMed  PubMed Central  Google Scholar 

  38. Park, Y.-N., Daniels, K. J., Pujol, C., Srikantha, T. & Soll, D. R. Candida albicans forms a specialized “sexual” as well as “pathogenic” biofilm. Eukaryotic cell 12, 1120–1131 (2013).

    CAS  Article  Google Scholar 

  39. LaFleur, M. D., Kumamoto, C. A. & Lewis, K. Candida albicans biofilms produce antifungal-tolerant persister cells. Antimicrobial agents and chemotherapy 50, 3839–3846 (2006).

    CAS  Article  Google Scholar 

  40. Nobile, Clarissa J. et al. A Recently Evolved Transcriptional Network Controls Biofilm Development in Candida albicans. Cell 148, 126–138, doi: 10.1016/j.cell.2011.10.048 (2012).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. Schoch, C. L. et al. Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proceedings of the National Academy of Sciences of the United States of America 109, 6241–6246, doi: 10.1073/pnas.1117018109 (2012).

    ADS  Article  PubMed  PubMed Central  Google Scholar 

  42. Stielow, J. et al. One fungus, which genes? Development and assessment of universal primers for potential secondary fungal DNA barcodes. Persoonia-Molecular Phylogeny and Evolution of Fungi 35, 242–263, doi: 10.3767/003158515X689135 (2015).

    CAS  Article  PubMed Central  Google Scholar 

  43. Cardinali, G., Bolano, A. & Martini, A. A DNA extraction and purification method for several yeast genera. Annals of Microbiology 51, 121–130 (2001).

    CAS  Google Scholar 

  44. Altschul, S. F., Gish, W., Miller, W., Myers, E. W. & Lipman, D. J. Basic local alignment search tool. Journal of molecular biology 215, 403–410 (1990).

    CAS  Article  Google Scholar 

  45. Schoch, C. L. et al. Finding needles in haystacks: linking scientific names, reference specimens and molecular data for Fungi. Database- the journal of biological datrabase and curation 2014, doi: 10.1093/database/bau061 (2014).

  46. Irinyi, L. et al. International Society of Human and Animal Mycology (ISHAM)-ITS reference DNA barcoding database—the quality controlled standard tool for routine identification of human and animal pathogenic fungi. Medical Mycology 53, 313–337 (2015).

    CAS  Article  Google Scholar 

  47. Pierce, C. G. et al. A simple and reproducible 96-well plate-based method for the formation of fungal biofilms and its application to antifungal susceptibility testing. Nature protocols 3, 1494–1500 (2008).

    CAS  Article  Google Scholar 

Download references


The work was partly supported by the project of National interest (PRIN) number 2010WZ2NJN_002, funded by the Italian Ministry of Research and University. LR was partly supported by a Grant of the Fondazione Cassa di Risparmio di Perugia (Italy); CC was partly supported by a PhD Grant of the Italian Ministry of Research and University. The authors thank Mrs. Cecilia Lucarelli and Mr. Emanuele Pitari for their technical work.

Author information




L.C. participated in planning of the study and experimental design, carried out the experimental activities, participated in data analysis and interpretation and drafted the manuscript. L.R. participated in the experimental activities and data analysis and interpretation. C.C. participated in the experimental activities. C.T. participated in planning of the study and experimental design, carried out the experimental activities, participated in data analysis and interpretation and drafted the manuscript. A.L. participated in the experimental activities and data analysis and interpretation. E.S. participated in the experimental activities and data analysis and interpretation. F.M. participated in planning of the study and experimental design, carried out the experimental activities, participated in data analysis and interpretation and drafted the manuscript. M.M. commented on the manuscript. C.S. commented on the manuscript. W.M. commented on the manuscript. C.G. participated in planning of the study, experimental design, data analysis and interpretation and commented on the manuscript. M.B. participated in planning of the study, experimental design, data analysis and interpretation and commented on the manuscript. All authors read and approved the final version.

Ethics declarations

Competing interests

In the past 2 years, Dr. Carlo Tascini has been paid for lectures on behalf of Pfizer, Novartis, Merck Astra, Angelini, Gilead and Astellas.

Electronic supplementary material

Rights and permissions

This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Corte, L., Roscini, L., Colabella, C. et al. Exploring ecological modelling to investigate factors governing the colonization success in nosocomial environment of Candida albicans and other pathogenic yeasts. Sci Rep 6, 26860 (2016).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing