Phenotypical properties associated with virulence from clinical isolates belonging to the Candida parapsilosis complex


  • Érika A. Abi-chacra,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Lucieri O.P. Souza,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Lucas P. Cruz,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Lys A. Braga-Silva,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    2. Programa de Pós-Graduação em Bioquímica, Instituto de Química, UFRJ, Rio de Janeiro, Brazil
    Search for more papers by this author
  • Diego S. Gonçalves,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    2. Programa de Pós-Graduação em Bioquímica, Instituto de Química, UFRJ, Rio de Janeiro, Brazil
    Search for more papers by this author
  • Cátia L. Sodré,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    2. Departamento de Biologia Celular e Molecular, Instituto de Biologia, Universidade Federal Fluminense (UFF), Niterói, Brazil
    Search for more papers by this author
  • Marcos D. Ribeiro,

    1. Laboratório de Bacteriologia/Micologia, Instituto de Biologia do Exército (IBEx), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Sergio H. Seabra,

    1. Laboratório de Tecnologia em Cultura de Células, Centro Universitário Estadual da Zona Oeste (UEZO), Duque de Caxias, Brazil
    Search for more papers by this author
  • Maria H.G. Figueiredo-Carvalho,

    1. Laboratório de Micologia, Instituto de Pesquisa Clínica Evandro Chagas, Fundação Oswaldo Cruz (FIOCRUZ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Leonardo S. Barbedo,

    1. Laboratório de Micologia, Instituto de Pesquisa Clínica Evandro Chagas, Fundação Oswaldo Cruz (FIOCRUZ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Rosely M. Zancopé-Oliveira,

    1. Laboratório de Micologia, Instituto de Pesquisa Clínica Evandro Chagas, Fundação Oswaldo Cruz (FIOCRUZ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • Mariangela Ziccardi,

    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    2. Laboratório Interdisciplinar de Pesquisas Médicas, Instituto Oswaldo Cruz, Fundação Oswaldo Cruz (FIOCRUZ), Rio de Janeiro, Brazil
    Search for more papers by this author
  • André L.S. Santos

    Corresponding author
    1. Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, Brazil
    2. Programa de Pós-Graduação em Bioquímica, Instituto de Química, UFRJ, Rio de Janeiro, Brazil
    • Correspondence: André L.S. Santos, Laboratório de Investigação de Peptidases (LIP), Departamento de Microbiologia Geral, Instituto de Microbiologia Paulo de Góes (IMPG), Bloco E-subsolo, sala 05, Centro de Ciências da Saúde (CCS), Universidade Federal do Rio de Janeiro (UFRJ), Rio de Janeiro, RJ 21941-902, Brazil. Tel.: +55 21 25626740; fax: +55 21 25608344; e-mail:

    Search for more papers by this author


The production of virulence attributes in three reference strains and 11 clinical isolates primarily identified as Candida parapsilosis was evaluated. Morphological and phenotypical tests were not able to discriminate among the three species of the C. parapsilosis complex; consequently, molecular methods were applied to solve this task. After employing polymerase chain reaction-based methods, nine clinical strains were identified as C. parapsilosis sensu stricto and two as C. orthopsilosis. Protease, catalase, and hemolysin were produced by all 14 strains, while 92.9% and 78.6% of strains secreted, respectively, esterase and phytase. No phospholipase producers were detected. Mannose/glucose, N-acetylglucosamine, and sialic acid residues were detected at the surface of all strains, respectively, in high, medium, and low levels. All strains presented elevated surface hydrophobicity and similar ability to form biofilm. However, the adhesion to inert substrates and mammalian cells was extremely diverse, showing typical intrastrain variations. Overall, the strains showed (1) predilection to adhere to plastic over glass and the number of pseudohyphae was more prominent than yeasts and (2) the interaction process was slightly enhanced in macrophages than fibroblasts, with the majority of fungal cells detected inside them. Positive/negative correlations were demonstrated among the production of these virulence traits in C. parapsilosis complex.


Candida parapsilosis can be isolated from domestic animals, insects, soil and marine environments (Trofa et al., 2008; Silva et al., 2012). Moreover, C. parapsilosis is a commensal yeast from human that frequently colonizes the skin; however, it can become pathogenic if changes occur in the host defense mechanisms or if any interference breaks up the anatomical barriers of the skin like burn or usage of invasive devices (Almirante et al., 2006; Trofa et al., 2008; Nosek et al., 2009; van Asbeck et al., 2009; Silva et al., 2012). In the clinical arena, C. parapsilosis is an important pathogen of fungemia, particularly among infants and immunocompromised patients hospitalized in intensive care units, causing different clinical manifestations as endocarditis, meningitis, peritonitis, arthritis, eye infections, otomycosis, onychomycosis, vulvovaginitis, and urinary tract infections (Trofa et al., 2008). In some regions/countries, C. parapsilosis is the second most frequent Candida species, reaching 20.5% of cases in Latin America (Nucci et al., 2010) and 21.2% in the United Kingdom (Das et al., 2011).

Based on genetic analysis, it was found that C. parapsilosis forms a complex composed of three genetically distinct species, namely C. parapsilosis sensu stricto, C. orthopsilosis and C. metapsilosis, which are physiologically and morphologically indistinguishable (Tavanti et al., 2005). It is noteworthy that C. metapsilosis and C. orthopsilosis also have clinical importance as nosocomial pathogens of fungemia, because they are also responsible for several infections around the world (Lockhart et al., 2008). Although the incidence of fungi belonging to the complex C. parapsilosis has increased in recent decades, little is known about the virulence factors involved in the colonization, adhesion, invasion, dissemination and escape from host defenses. Similarly, there is a large gap in the current knowledge of the pathogenesis of C. parapsilosis infections. Among the virulence factors proposed in the literature for Candida spp., adherence to host cells and/or tissues as well as to inert supports, morphological changes and secretion of a large array of hydrolytic enzymes are included (Ghannoum, 2000; Haynes, 2001; Hube & Naglik, 2001; Braga-Silva & Santos, 2011; Singh & Mukhopadhyay, 2012).

The aim of the present study was to evaluate the production of virulence attributes in three type strains and 11 clinical isolates belonging to the C. parapsilosis complex.

Materials and methods

Microorganisms and growth conditions

Candida metapsilosis (ATCC 96143), Candida orthopsilosis (ATCC 96141), and Candida parapsilosis sensu stricto (ATCC 22019) were obtained from American Type Culture Collection (ATCC, Rockville; Table 1). The 11 clinical isolates were kindly provided by Dr Marcos Dornelas Ribeiro (Table 1; Instituto de Biologia do Exército, IBEx, Rio de Janeiro, Brazil). Fungal cells were cultured into 1.2% yeast carbon base (YCB) medium (HiMedia Laboratories Ltd, India) supplemented with 0.1% bovine serum albumin (BSA; Sigma-Aldrich, New York, NY), pH 4.0, at 37 °C for 48 h (Hrusková-Heidingsfeldová et al., 2009). Growth was estimated by counting the fungal cells in a Neubauer chamber.

Table 1. Biochemical and morphological characteristics of the Candida parapsilosis complex included in this study
CodeSampleColony color on CHROMagarBiochemical identification (VITEK® 2 system)Morphological and growth pattern
IdentificationProbability (confidence)Contraindicating typical biopatternNumber of cells × 106Size (FSC)Granularity (SSC)
  1. a

    ATCC, American Type Culture Collection. ATCC 22019, C. parapsilosis; ATCC 96141, C. orthopsilosis; ATCC 96143, C. metapsilosis.

  2. IBEx, Instituto de Biologia do Exército; –, not informed; EI, excellent identification; VGI, very good identification; dXYLa, d-xylose assimilation; dGATa, d-galacturonate assimilation; GRTas, glucuronate assimilation; IARAa, L-arabinose assimilation; IMLTa, L-malate assimilation; NAGa, N-acetyl glucosamine assimilation; URE, urease; FSC, forward light scatter; SSC, side scatter.

222ToeWhite-cream C. parapsilosis 97% (EI)dGATa, GRTas8.3 ± 3.9313.3 ± 46.0360.4 ± 26.6
225ToeWhite-cream C. parapsilosis 96% (EI)dXYLa, GRTas10.7 ± 1.9300.3 ± 19.8396.3 ± 63.1
229FingerWhite-cream C. parapsilosis 95% (VGI)URE9.5 ± 4.5298.9 ± 19.5385.7 ± 42.9
234White-cream C. parapsilosis 98% (EI)NAGa22.5 ± 1.1306.6 ± 16.4370.1 ± 13.4
235FingerWhite-cream C. parapsilosis 97% (EI)dXYLa, GRTas8.3 ± 2.4321.2 ± 52.8413.2 ± 70.4
241EarWhite-cream C. parapsilosis 94% (VGI)URE, NAGa6.1 ± 1.8275.1 ± 1.3421.1 ± 40.0
248FingerWhite-cream C. parapsilosis 94% (VGI)URE, NAGa9.9 ± 3.4307.3 ± 25.4428.7 ± 93.0
251GluteusWhite-cream C. parapsilosis 96% (EI)dXYLa, GRTas15.5 ± 4.1322.2 ± 12.7370.5 ± 12.1
275FingerWhite-cream C. parapsilosis 96% (EI)dXYLa, GRTas15.2 ± 1.3324.5 ± 31.0390.6 ± 21.1
276FingerWhite-cream C. parapsilosis 94% (VGI)URE, NAGa13.3 ± 1.2295.8 ± 8.0350.0 ± 50.4
454White-cream C. parapsilosis 94% (VGI)URE, NAGa18.5 ± 1.6321.2 ± 58.0332.0 ± 23.1
22019aFecesWhite-cream C. parapsilosis 94% (VGI)IMLTa, dXYLa, NAGa, dGATa11.7 ± 6.8308.1 ± 25.2409.9 ± 20.0
96141aBloodWhite-cream C. parapsilosis 96% (EI)dXYLa, IARAa13.0 ± 1.3310.0 ± 31.4349.6 ± 12.6
96143aWhite-cream C. parapsilosis 98% (EI)IARAa10.3 ± 2.0339.9 ± 45.0467.6 ± 31.7

Morphological parameters

Flow cytometry

Fungal cells were processed for flow cytometry to measure two morphological parameters: cellular size and granularity. Briefly, cells (106) were harvested by centrifugation at 4000 g for 10 min at 4 °C, washed in cold phosphate-buffered saline (PBS; 150 mM NaCl, 20 mM phosphate buffer, pH 7.2), and fixed in 4% paraformaldehyde in PBS at 4 °C for 30 min. Then, fungal cells were washed three times in PBS and analyzed in a flow cytometer (FACSCalibur; BD Bioscience) equipped with a 15-mW argon laser emitting at 488 nm. Each experimental population was mapped (10 000 events) using a two-parameter histogram of forward-angle light scatter (FSC) vs. side scatter (SSC), to evaluate size and granularity, respectively (Braga-Silva & Santos, 2011).

Calcofluor white staining

PBS-washed fungal cells were stained with calcofluor white to detect chitin at the fungal cell wall. With this purpose, fungi (106 mL−1) were mixed with 100 μL of calcofluor white (Sigma-Aldrich) solution (1 mg mL−1) and incubated for 5 min at room temperature (Laffey & Butler, 2005). Then, cells were washed three times with PBS and resuspended in 100 μL of the same buffer. Five microliters of the stained cell suspension were spotted onto a glass slide and visualized under bright field and UV/DAPI using an Observer Z1 (Zeiss, Germany) fluorescence microscope. Images were acquired with a Color View AxioCam MRm digital camera.

Scanning electron microscopy (SEM)

PBS-washed fungal cells (106) were fixed overnight in 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.2) at 4 °C, rinsed three times with PBS, and post-fixed with 2% osmium tetroxide for 2 h. Samples were then dehydrated in graded concentrations of acetone (25–100%). Cells were dried by the critical point method, mounted on stubs, coated with gold (20–30 nm), and observed in a JEOL JSM 6490LV scanning electron microscope (Sangetha et al., 2009).

Biochemical identification

Initially, all fungal isolates were phenotypically identified by plating onto CHROMagar Candida® medium (CHROMagar Candida Company, Paris, France). The plates were incubated for 48 h at 37 °C under aerobic conditions, and the interpretation of results was based according to the manufacturer's guidelines. Furthermore, the metabolic properties, such as sugar assimilation and enzymatic reactions, were analyzed by VITEK® 2 system (bioMérieux, France) using YST card according to the manufacturer's guidelines.

Molecular identification

As phenotypical methods are not sufficient to distinguish between the members of the C. parapsilosis complex, four distinct methodologies were performed in order to detect the presence of the two cryptic species (C. metapsilosis and C. orthopsilosis).

DNA extraction

Fungal cells were recovered from Sabouraud dextrose agar (Difco, Becton, Dickinson and Company) and used for genomic DNA extraction with the Gentra® Puregene® Yeast and G+ Bacteria Kit (Qiagen®, Maryland), according to the manufacturer's protocol. DNA concentration was estimated with a spectrophotometer (NanoVue Plus™; GE Healthcare) at A260, and its integrity was checked under UV light after through a 1% agarose gel electrophoresis stained with ethidium bromide (Sigma-Aldrich). DNA extracts were stored at −20 °C until used.

PCR with species-specific primers

Each isolate was tested with three different pairs of species-specific primers: CORF (5′-TTTGGTGGCCCACGGCCT-3′) and CORR (5′-TGAGGTCGAATTTGGAAGAATT-3′), CMEF (5′-TTTGGTGGGCCCACGGCT-3′) and CMER (5′-GAGGTCGAATTTGGAAGAATGT-3′), CPAF (5′-TTTGCTTTGGTAGGCCTTCTA-3′) and CPAR (5′-GAGGTCGAATTTGGAAGAAGT-3′) targeting the ITS1-5.8S-ITS2 region of the rRNA gene of C. orthopsilosis, C. metapsilosis and C. parapsilosis sensu stricto, respectively, as previously described by Asadzadeh et al. (2009). PCR products (concentrated to 20 μL) were mixed with 0.2 volumes of loading buffer and separated through 1% agarose gel electrophoresis in 1× TBE buffer (89 mM Tris borate, 2 mM EDTA, pH 8.4) at 90 V for 80 min. Next, the gel was immersed in an ethidium bromide solution (Invitrogen™, Carlsbad, CA) at a final concentration of 0.5 μg mL−1 for 30 min and visualized under a UV light.

BanI digestion of secondary alcohol dehydrogenase (SADH) gene fragment

Primers S1F (5′-GTTGATGCTGTTGGATTGT-3′) and S1R (5′-CAATGCCAAATCTCCCAA-3′), which amplify a fragment of the SADH gene, were used for the PCR as published by Tavanti et al. (2005). The expected c. 720 bp amplicon was subsequently digested with the restriction endonuclease BanI according to the manufacturer's instructions (Biolabs). The resulting DNA fragments were separated and visualized on a 2% agarose gel stained with ethidium bromide. Candida orthopsilosis, C. metapsilosis, and C. parapsilosis sensu stricto SADH fragments showed one band (720 bp), four bands (390, 190, 100, and 40 bp) and two bands (200 and 520 bp), respectively.

EcoRI digestion of β(1,3)-glucan synthase subunit 1 (FKS1) gene

Primers REA-F (5′-GATGACCAATTYTCAAGAGT-3′) and REA-R (5′-GTCAACATAAATGTAGCATTCTAGAAATC-3′), which amplify a fragment of the FKS1 gene, were used for the PCR as published by Garcia-Effron et al. (2011). The expected 1032 bp amplicon was subsequently digested with the restriction endonuclease EcoRI according to the manufacturer's instructions (Sigma-Aldrich). The resulting DNA fragments were separated and visualized on a 2% agarose gel stained with ethidium bromide. Candida orthopsilosis, C. metapsilosis, and C. parapsilosis sensu stricto FKS1 fragments showed three bands (474, 306, and 252 bp), two bands (474 and 558 bp), and one band (1032 pb), respectively.

DNA sequencing of the D1/D2 region of the 28S rRNA gene

The results of species-specific identification of clinical isolates were confirmed, when necessary, by DNA sequencing of the D1/D2 region of the 28S rRNA gene and both strands were sequenced. The amplified product was obtained with the NL-1 (5′-GCATATCAATAAGCGGAGGAAAAG-3′) and NL-4 (5′-GGTCCGTGTTTCAAGACGG-3′) primers as described previously by Asadzadeh et al. (2009). The amplicons obtained were purified with the QIAquick® PCR Purification Kit (Qiagen®) according to the manufacturer's protocol. Automated sequencing was carried out using the DNA sequencing platform (ABI-3730; Applied Biosystems) (PDTIS/FIOCRUZ – Rio de Janeiro, Brazil). Sequences from both DNA strands were generated and edited with the Sequencher version 4.9 software package (Genes Codes Corporation), followed by alignment by means of the Mega version 4.0.2 software. The sequences of our strains were compared by blast with sequences available from NCBI GenBank.

Production of hydrolytic enzymes

The enzymatic activities were measured by spectrophotometer assay or plate method. In the latter, aliquots (10 μL) of 48-h-old cultured fungal cells (107) were spotted on the surface of each agar medium (see below) and then incubated for 7 days at 37 °C. The colony diameter (a) and the diameter of colony plus precipitation zone (b) were measured by a digital paquimeter and the enzymatic activities were expressed as Pz value (Pz a/b) as described by Price et al. (1982). According to this definition, low Pz values mean high enzyme production and, inversely, high Pz values indicate low enzymatic production. The Pz value was scored into four categories: Pz of 1.0 was evaluated as negative; high Pz between 0.999 and 0.700; moderate Pz between 0.699 and 0.400; low Pz between 0.399 and 0.100.

Protease activity

Determination of protease production was assayed using the albumin agar plate (1.17% yeast carbon base (YCB) medium supplemented with 0.2% BSA, pH 4.0) as previously described by Rüchel et al. (1982). The protease activity results in a clear zone around the colony, which correspond to the hydrolysis of the BSA present in the medium.

Phytase activity

Determination of phytase (myo-inositol hexakisphosphate or myo-inositol 1,2,3,4,5,6-hexakis dihydrogen phosphate) production was assayed using the calcium phytate agar plate (glucose, 10 g; (NH4)2SO4, 0.5 g; KCl, 0.2 g; MgSO4.7H2O, 0.1 g; calcium phytate, 2 g; yeast extract, 0.5 g; MnSO4, 0.005 g; FeSO4, 0.005 g, pH 7.0), as previously described by Tsang (2011). The phytase activity was directly observed around the colony able to solubilize phosphate.

Lipase activity

Production of two distinct lipases, phospholipase and esterase, was also investigated. The determination of phospholipase activity was performed using egg yolk agar plate (1 M NaCl, 5 mM CaCl2 and 8% sterile egg yolk emulsion, pH 7.0) according to Price et al. (1982). The esterase production was assayed using the Tween agar plate (1 g of peptone, 0.5 g of NaCl, 0.01 g of CaCl2, pH 7.0, 1.5 g of agar and 100 mL of distilled water, which was autoclaved, then cooled to about 50 °C, and added of 0.5 mL of autoclaved Tween) according to Aktas et al. (2002). In these two methods, the hydrolysis of lipid substrates present in egg yolk or Tween results in the formation of a calcium complex with fatty acids released by the action of the secreted enzymes, resulting in a precipitation zone around the colony.

Hemolytic activity

Determination of hemolysis was assayed using the blood agar plate, in which the medium was prepared by adding 7 mL of fresh sheep blood to 100 mL of Sabouraud dextrose agar supplemented with 3% glucose, as previously described by Luo et al. (2001). The presence of a distinct translucent halo around the inoculum site, viewed with transmitted light, indicated positive hemolytic activity.

Catalase activity

Fungal cells (106) were added to lysis buffer (100 mM Tris HCl, pH 7.5, 1 mM ethylenediaminetetraacetic acid (EDTA), 1 mM phenylmethylsulfonyl fluoride (PMSF), 1 mg mL−1 pepstatin A, 1 mg mL−1 aprotinin) and lysed in a cell homogenizer (FastPrep) by five cycles of 30 s, alternating with the ice bath for 2 min. Then, the mixtures were centrifuged and the supernatants obtained were added of 1.8 mL H2O2 (Merck, Brazil) at 17 mM in a sterile tube. The mixture was kept at room temperature for 30 min. The activity of catalase was determined by monitoring the removal of H2O2 in a spectrophotometer at 240 nm. Catalase activity was determined as proposed by Aebi (1984), and its specific activity was expressed as ΔA240/min per milligram of protein. Protein concentration was determined by the method described by Lowry et al. (1951), using BSA as standard.

Detection of cell surface glycoconjugates

Paraformaldehyde-fixed fungal cells (106) were incubated for 1 h with three distinct fluorescein isothiocyanate (FITC)-labeled lectins (Sigma-Aldrich) at 5 μg mL−1: concanavalin A (Con A), wheat germ agglutinin (WGA), and Limax flavus agglutinin (LFA). After these incubations, the fungal-associated fluorescence was excited at 488 nm and quantified on a flow cytometer. Fungal cells treated only with PBS were run in parallel as controls. The mapped population (10 000 events) was analyzed for log green fluorescence using a single-parameter histogram. The level of surface glycoconjugates was shown as percentage of fluoresceinated cells and as the mean of the fluorescence intensity (MFI; Braga-Silva et al., 2010).

Cell surface hydrophobicity

Cell surface hydrophobicity was measured by the water-octane two-phase assay (Hazen et al., 1986). Briefly, 2.5 mL of PBS-washed fungi (108 cells mL−1) was mixed vigorously with 0.5 mL of octane and the two liquids were allowed to partition during 15 min at room temperature. ABS570 values of species in PBS without octane overlay were used as negative controls. The percentage of exclusion of the cells from the aqueous phase (% change in ABS570) corresponding to relative cell hydrophobicity was calculated as: [(ABS570 of the control – ABS570 after octane overlay)/ABS570 of the control] × 100. As previously reported, high, moderate, and low cell surface hydrophobicity corresponds to respective changes 80–100%, 20–80%, and 0–20% in ABS570.

Adhesion to abiotic substrates

Fungal cell suspensions (100 μL containing 106 cells) were placed on glass slides, which were previously washed with Extran for 2 h, incubated in 70% ethanol for 30 min, and then sterilized at 180 °C for 2 h, as well as in 24-well plastic plates (polystyrene) for 2 h at 37 °C. Afterward, the systems were washed three times in PBS to remove nonadherent cells. Glass slides were examined in an optical microscope (Zeiss) and the 24-well plates were analyzed in an inverted microscope (Zeiss). The experiments were performed in triplicate, and five different microscopic fields were counted in each system to express the number of total fungi adhered to these substrates (Reinhart et al., 1985). In parallel, the number of yeasts and pseudohyphae were individually analyzed to check the possible induction of morphological transition due to the adhesion process to these inert substrates.

Biofilm production and viability

Fungal cell suspensions (100 μL containing 106 cells) were transferred into each well of a 96-well polystyrene microtiter plates for 48 h at 37 °C. Subsequently, the wells were washed three times in PBS to remove nonadherent cells. The biomass of the resultant biofilms was then assessed using the crystal violet assay (Peeters et al., 2008). Briefly, 100 μL of 99% methanol was added to each well for 15 min to fix the biofilm and then the supernatants were discarded. Microplates were air-dried and then 100 μL of crystal violet solution (1 : 50 from stock solution, Sigma-Aldrich) were added to wells and incubated at room temperature for 20 min. The extra dye was washed away with tap water and then 150 μL of acetic acid 33% was added to the wells. The absorbance was measured at 590 nm using a Thermomax Molecular Device microplate reader. In parallel, the viable cells in biofilm were assessed by the colorimetric assay that investigates the metabolic reduction of 2,3-bis (2-methoxy-4-nitro-5-sulfophenyl)-5-[(phenylamino) carbonyl]-2H-tetrazolium hydroxide (XTT; Sigma-Aldrich) to a water-soluble brown formazan product. In this sense, 100 μL of the XTT/menadione solution [4 mg XTT in 10 mL prewarmed PBS was dissolved and supplemented by 100 μL menadione stock solution (Sigma-Aldrich), which contained 55 mg menadione in 100 mL acetone] was added to all wells and incubated in the dark at 37 °C for 3 h. The contents of the wells were transferred to microcentrifuge tubes and centrifuged at 4000 g for 5 min. A total of 100 μL of supernatant from each well was transferred to a new microplate and the colorimetric changes were measured at 492 nm using a microplate reader (Peeters et al., 2008). In parallel, the biofilm was developed on glass slides for 24 and 48 h and then processed to be analyzed by means of SEM.

Interaction to cell lineages

The cell lineages L929 (code number CRL-2148) from mouse C3H/An connective tissue (mouse fibroblasts) and RAW264.7 (code number TIB-71) from mouse BALB/c (macrophages) were obtained from ATCC and kindly provided by Dr Maria Teresa Villela Romanos (Departamento de Virologia, IMPG – UFRJ). The animal cells were maintained and grown to confluence in 25-cm2 culture flasks containing Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS; Gibco, Life Technologies) at 37 °C with 5% CO2. The pH was maintained at 7.2 by the addition of 3 g L−1 HEPES and 0.2 g L−1 NaHCO3 to the medium. The initial inoculum was 5 × 104 cells mL−1; this was subcultured every 2 days, and the cells were maintained in log-phase growth. For adhesion assays, animal cells were plated onto 24-well multidishes at a density of 105 cells per well and incubated at 37 °C for 24 h in DMEM supplemented with 10% FBS. Before interaction with animal cells, the wall of fungal cells (106) was stained with 0.5 mg mL−1 FITC in PBS at room temperature for 10 min (Chaka et al., 1995). The cells were then washed twice with PBS and finally rinsed in DMEM. Fungal suspensions were prepared in DMEM to generate a ratio of 10 yeasts per animal cell. Interactions between fungal and animal cells occurred at 37 °C with 5% CO2 for 2 h. Cells were washed three times in PBS to remove nonadherent yeasts. Fungus–animal cell complexes were treated for 10 min at 25 °C with trypan blue (200 μg mL−1) to discriminate between surface-associated and intracellular fungal cells. After removal from the plastic surface with a cell scrapper, the interaction systems were analyzed by flow cytometry as described previously (Chaka et al., 1995). Control preparations were developed using uninfected animal cells and nonstained yeasts.


All experiments were performed in triplicate, in three independent experimental sets. The data were analyzed statistically by means of Student's t-test using epi-info 6.04 (Database and Statistics Program for Public Health) computer software. The correlation tests were determined by Pearson's correlation coefficient (r). In all analyses, P values of 0.05 or less were considered statistically significant.


Identification of fungal strains

All the 14 fungal isolates (three ATCC reference strains and 11 clinical isolates) were initially rechecked to certify their authentication by mycology methodologies. Analysis of growth in a chromogenic medium culture facilitated the determination of colony purity and the identification of Candida spp. In this context, the fungal strains yielded only white-cream colonies after 48 h of cultivation on CHROMagar Candida medium (Fig. 1a), which suggests a presumptive identification of C. parapsilosis (Table 1). The carbohydrate assimilation and metabolic enzymatic profiles evaluated by VITEK 2 system corroborated the preliminary result, showing identity probability ranging from 94% (very good identification) to 98% (excellent identification) in which the most contradictory tests were d-xylose assimilation (6/14 strains, 42.9%), N-acetylglucosamine assimilation (6/14, 42.9%), glucuronate assimilation (5/14, 35.7%), urease (5/14, 35.7%), d-galacturonate assimilation (2/14, 14.3%), L-arabinose assimilation (2/14, 14.3%), and L-malate assimilation (1/14, 7.1%; Table 1).

Figure 1.

Macroscopic and microscopic aspects of species belonging to the Candida parapsilosis complex. Macroscopic colony cultured for 48 h at 37 °C on CHROMagar Candida medium (a) and YCB-BSA (b). Note in (b) a clearance zone around the colony, due to albumin hydrolysis by fungal secreted aspartic-type proteases. In the present study, all the C. parapsilosis complex strains exhibited smooth phenotypes cultured on both culture media. For this reason, only representative images of C. parapsilosis sensu stricto are shown. Microscopy structures observed when fungal strains were cultured in YCB-BSA medium (c–g). Light (c) and scanning electron (d) microscopies were performed to visualize the different morphotypes of C. parapsilosis complex, including oval/rounded yeasts (open white arrows), elongated cells (white arrowheads), and pseudohyphae (closed white arrows). All these macroscopic colonial aspects as well as the microscopic structures were identical in the three species belonging to the C. parapsilosis complex (C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis); consequently, it was showed only representative images concerning these morphological characteristics. To corroborate these findings, it was presented the epifluorescence photocomposition of the different morphological growth forms of C. parapsilosis complex stained with calcofluor white to visualize chitin along the surface of all morphotypes: oval/rounded yeasts (e), elongated cells (f), and pseudohyphae (g). The panels (e–g) were photographed at the same magnification.

Some morphological and growth characteristics were also examined. All the 14 strains of C. parapsilosis complex formed smooth colonies with no apparent filamentous extensions from the edge of the colonies on YCB-BSA agar medium (Fig. 1b). However, yeasts, displaying oval, round or cylindrical shapes, and pseudohyphae were detected during the in vitro growth in YCB-BSA broth in all the fungal isolates studied (Figs. 1c and d). Staining with calcofluor white showed that chitin was distributed along the length of all morphotypes as well as at the bud scars (Fig. 1e–g). Regarding the size and granularity, all the strains presented similar measurements for these two morphological parameters as judged by flow cytometry analyses, showing mean size of 310.3 ± 15.6 and mean granularity of 389.0 ± 36.9 (Table 1). In addition, all the strains attained the exponential growth phase at 48 h of cultivation in YCB-BSA medium; however, they yielded different amount of cells ranging from 6.1 × 106 to 22.5 × 106 (Table 1).

As previously proposed (Gomez-Lopez et al., 2008; Lockhart et al., 2008; Asadzadeh et al., 2009), the three species belonging to the C. parapsilosis complex are both morphologically and phenotypically indistinguishable (Fig. 1e–g); consequently, molecular methods are needed for their precise detection (Fig. 2). Firstly, the fungal isolates were screened by PCR using species-specific primers derived from unique sequences within the internally transcribed spacer 1 (ITS1)–5.8 rRNAITS2 region developed by Asadzadeh et al. (2009). As expected, PCR amplifications of the reference ATCC strains with the CPAF+CPAR, CORF+CORR, and CMEF+CMER primers yielded amplicons of 379, 367, and 374 bp only with DNA from C. parapsilosis sensu stricto, C. orthopsilosis, and C. metapsilosis, respectively (Fig. 2a). Of the 11 clinical isolates, nine were amplified only by a single pair of primers, CPAF+CPAR, identifying them as C. parapsilosis sensu stricto; however, the remaining strains (248 and 276) were imprecisely identified, because two distinct pair of primers (CPAF+CPAR and CORF+CORR) were able to generate amplifications (Fig. 2a). To solve this identification inaccuracy, we performed the PCR-restriction fragment length polymorphism (RFLP) of two distinct genes: SADH and FKS1. PCR amplification of a fragment of the SADH gene generated an amplicon of c. 720 bp in all fungal strains (Fig. 2b, SADH), which was subsequently digested by the restriction enzyme BanI. As expected, C. parapsilosis sensu stricto (ATCC 22019) presented one BanI restriction site at position 196, C. metapsilosis (ATCC 96143) had three restrictions at positions 96, 469 and 529, while C. orthopsilosis (ATCC 96141) had none cleavage site (Fig. 2b, SADH-BanI). Among the clinical isolates, all samples (n = 11) were identified as C. parapsilosis sensu stricto (Fig. 2b, SADH-BanI). The FKS1 gene was successfully amplified in all isolates tested, generating an amplicon of 1032 bp (Fig. 2c, FKS1) that was enzymatically digested with endonuclease EcoRI. As predicted, the restriction pattern of C. metapsilosis (ATCC 96143) showed only one site of cleavage at position 474, C. orthopsilosis (ATCC 96141) two sites at positions 474 and 778, while C. parapsilosis sensu stricto (ATCC 22019) showed no EcoRI digestion sites (Fig. 2c, FKS1-EcoRI). Based on the FKS1 restriction profile, nine clinical isolates were identified as C. parapsilosis sensu stricto and two (248 and 276) as C. orthopsilosis (Fig. 2c, FKS1-EcoRI). DNA sequencing of the D1/D2 region of the 28S rRNA gene was subjected to a nucleotide blast search and sequence analysis, which resulted in identification of the strains 248 and 276 as C. orthopsilosis due to 100% sequence similarity in comparison with the D1/D2 sequence of two reference strains (FM172989.1 and ATCC 96141) of C. orthopsilosis deposited at the GenBank database (Fig. 3).

Figure 2.

Molecular discrimination of species belonging to the Candida parapsilosis complex. (a) Agarose gel showing the PCR-amplified products obtained from 11 clinical isolates and three reference strains amplified with species-specific primers from C. parapsilosis (CPAF+CPAR), C. orthopsilosis (CORF+CORR), and C. metapsilosis (CMEF+CMER). Note that the first lane of each gel represents the 100-bp ladder; (−), negative control (PCR mix with no fungal DNA); (+), positive control (reference strain specific for each pair of primes); and clinical strains (1–11). Lanes: (1) 222, (2) 225, (3) 229, (4) 234, (5) 235, (6) 248, (7) 241, (8) 251, (9) 276, (10) 275, (11) 454. (b) Amplification of a 720-bp SADH gene fragment from the genomic DNA of fungal strains and BanI digestion of the SADH-PCR product obtained from clinical isolates (1–11) and reference strains (12–14). (c) Amplification of a 1032-bp FKS1 gene fragment from the genomic DNA of fungal strains and EcoRI digestion of the FKS1-PCR product obtained from clinical isolates (1–11) and reference strains (12–14). Note that in (b) and (c), the lanes represent the following strains: (1) 222, (2) 225, (3) 229, (4) 234, (5) 235, (6) 241, (7) 248, (8) 251, (9) 275, (10) 276, (11) 454, (12) C. orthopsilosis ATCC 96139, (13) C. metapsilosis ATCC 96143, (14) C. parapsilosis ATCC 22019.

Figure 3.

Phylogenetic neighbor-joining dendrogram generated from a genetic similarity matrix based on comparison of D1/D2 region of the 28S rRNA gene sequences from two cutaneous clinical strains (designated as 248 and 276) and six type strains belonging to the Candida parapsilosis complex, which sequences were obtained from GenBank.

After employing all these methodologies, among 11 presumed C. parapsilosis clinical isolates examined, nine were identified as C. parapsilosis sensu stricto and the remaining two isolates were found to belong to the cryptic species C. orthopsilosis. No C. metapsilosis was found in the clinical isolates analyzed in this study.

Production of hydrolytic enzymes

All the 14 fungal isolates were able to secrete protease after 7 days of growing in YCB-BSA medium. Except for the reference strain C. orthopsilosis (ATCC 96141), all the remaining isolates presented low Pz values ranging from 0.352 to 0.228, characterizing them as excellent protease producers (Table 2). Weak hemolytic activity was detected in all fungal strains (mean Pz of 0.862 ± 0.042; Table 2). Regarding the phytase activity, 11 strains (78.6%) were able to solubilize phosphate when incubated for 7 days in calcium phytate agar, while the following strains ATCC 96141 (C. orthopsilosis), 225, and 229 (C. parapsilosis) showed no activity for phytase (Table 2). Almost all phytase-producing strains (9/11) had moderate Pz values varying from 0.530 to 0.420 (Table 2). Esterase activity was detected in 92.9% (13/14) of fungal strains. Curiously, the three ATCC reference strains produced considerable (< 0.05) lower amount of esterase activity than the other esterase-producing clinical isolates (Table 2). The activity of catalase was measured in all the studied strains ranging from 0.031 to 0.593 (Table 2). In contrast, any fungal strain with ability to produce phospholipase was not observed.

Table 2. Hydrolytic enzyme production by clinical strains belonging to the Candida parapsilosis complex
  1. a

    The protease, phytase, esterase, and hemolysis activities were measured by the formation of a clear halo around the colony and expressed as Pz value as previously proposed by Price et al. (1982). The Pz value was scored into four categories: Pz of 1.0 was evaluated as negative; high Pz between 0.999 and 0.700; moderate Pz between 0.699 and 0.400; low Pz between 0.399 and 0.100.

  2. b

    The catalase activity was determined by monitoring the removal of H2O2 at 240 nm and expressed as ΔA240/min mg−1 of protein.

C. parapsilosis
ATCC 220190.310 ± 0.0200.530 ± 0.0330.670 ± 0.0220.849 ± 0.0300.248 ± 0.056
2220.262 ± 0.0500.363 ± 0.0150.438 ± 0.0050.877 ± 0.0340.457 ± 0.025
2250.285 ± 0.03110.414 ± 0.0200.851 ± 0.0520.183 ± 0.010
2290.302 ± 0.02510.456 ± 0.0170.806 ± 0.0360.232 ± 0.003
2340.295 ± 0.0310.420 ± 0.0200.434 ± 0.0180.926 ± 0.0800.593 ± 0.013
2350.292 ± 0.0380.480 ± 0.00310.911 ± 0.0280.048 ± 0.015
2410.228 ± 0.0250.392 ± 0.0130.448 ± 0.0130.858 ± 0.0340.367 ± 0.006
2510.330 ± 0.0410.450 ± 0.0020.446 ± 0.0110.857 ± 0.0250.183 ± 0.021
2750.335 ± 0.0300.467 ± 0.0130.476 ± 0.0090.875 ± 0.0280.047 ± 0.018
4540.333 ± 0.0240.433 ± 0.0150.456 ± 0.0170.866 ± 0.0390.513 ± 0.024
C. orthopsilosis
ATCC 961410.480 ± 0.02610.870 ± 0.0220.927 ± 0.0190.370 ± 0.029
2480.324 ± 0.0180.480 ± 0.0020.336 ± 0.0230.776 ± 0.0760.144 ± 0.002
2760.352 ± 0.0330.473 ± 0.0120.364 ± 0.0110.846 ± 0.0680.384 ± 0.012
C. metapsilosis
ATCC 961430.350 ± 0.0190.530 ± 0.0080.720 ± 0.0210.844 ± 0.0640.031 ± 0.012

Expression of surface glycoconjugates and cell surface hydrophobicity

Glycoconjugates containing units of mannose/glucose were found in the surface of all fungal strains in similar proportions (regarding the percentage of fluorescent cells of each fungal population, mean of 89.8 ± 4.6) and comparable amounts (regarding the mean of fluorescence intensity of each fungal population, mean of 109.9 ± 12.8; Table 3). Also, N-acetylglucosamine was evidenced in all fungal strains, presenting a proportion of fluorescent cells around 44–58% (mean of 52.6 ± 3.9), exception with the reference strains ATCC 22019 (C. parapsilosis) and ATCC 96141 (C. orthopsilosis) that presented higher proportion of this monosaccharide (Table 3). Sialic acids were expressed in low percentage at the surface of fungal cells varying from 0.3% to 21% (Table 3). The partition of cells in water-octane solution revealed a high cell surface hydrophobicity in all the fungal strains (mean of 95.3 ± 4.3; Table 3).

Table 3. Surface properties of clinical strains belonging to the Candida parapsilosis complex
CodeMannose/GlucoseaN-acetylglucosamineaSialic AcidaHydrophobicityb
  1. a

    The results were expressed as percentage of fluorescence cells (FC) as well as mean of fluorescence intensity (MFI).

  2. b

    The results were expressed as percentage of hydrophobicity.

C. parapsilosis
ATCC 2201983.7 ± 9.8139.4 ± 12.384.1 ± 0.596.8 ± 2.62.5 ± 0.4134.1 ± 1.595.9 ± 4.2
22296.0 ± 0.3108.2 ± 0.855.1 ± 0.1102.5 ± 4.20.4 ± 0.1122.4 ± 6.399.7 ± 3.3
22592.6 ± 1.4107.0 ± 0.353.9 ± 0.2123.7 ± 0.91.6 ± 0.5126.9 ± 4.499.9 ± 2.0
22991.1 ± 0.1104.3 ± 0.257.1 ± 0.2116.4 ± 2.08.4 ± 1.0125.7 ± 1.190.9 ± 5.7
23484.4 ± 2.0107.5 ± 0.254.3 ± 4.4121.5 ± 1.12.6 ± 1.2156.8 ± 6.797.2 ± 2.0
23590.4 ± 4.0105.3 ± 0.151.0 ± 1.0127.2 ± 2.020.9 ± 2.3124.0 ± 3.499.1 ± 2.6
24195.8 ± 0.198.4 ± 1.243.8 ± 6.5126.8 ± 2.51.8 ± 0.1144.0 ± 1.694.7 ± 4.8
25193.0 ± 0.9102.2 ± 0.254.0 ± 1.6111.8 ± 2.20.9 ± 0.4113.5 ± 5.492.1 ± 7.8
27589.4 ± 6.1106.9 ± 1.647.0 ± 5.3121.9 ± 4.51.3 ± 0.8130.3 ± 2.394.4 ± 9.0
45488.6 ± 1.6104.5 ± 0.352.5 ± 13.3115.2 ± 3.52.4 ± 0.957.6 ± 4.084.9 ± 12.8
C. orthopsilosis
ATCC 9614186.0 ± 0.4136.2 ± 4.279.0 ± 0.2127.2 ± 19.08.1 ± 1.2131.6 ± 7.392.0 ± 5.6
24896.5 ± 0.198.7 ± 1.352.8 ± 0.1122.3 ± 2.80.3 ± 0.1117.4 ± 11.195.9 ± 1.7
27683.6 ± 5.7100.9 ± 1.557.5 ± 2.2123.6 ± 0.811.7 ± 1.4160.4 ± 0.899.2 ± 3.3
C. metapsilosis
ATCC 9614385.5 ± 5.4118.4 ± 8.952.4 ± 0.7116.6 ± 4.912.5 ± 1.4121.0 ± 0.398.9 ± 1.5

Adhesion to abiotic substrates

The ability to adhere to abiotic substrates was performed using glass and polystyrene. All the fungal strains were able to bind to both inert surfaces with a marginally predilection to the plastic substrate (Table 4). Interestingly, the number of total fungi adhered to both substrates was immensely varied, showing a typical strain-specific adhesion property (Table 4). For instance, the number of fungi adhered to glass and plastic fluctuated from 4 to 127 and 8 to 300 cells per microscopic field, respectively (Table 4). In the adhesion experiments, yeasts and pseudohyphae were regularly observed in all studied fungal strains, which led us to quantify the number of these two distinct morphotypes. The number of pseudohyphae exceeded 50% in 8 (57.1%) and 13 (92.9%) fungal strains, respectively, during the interaction with glass and polystyrene for 2 h (Table 5). In general, the percentage of pseudohyphae was higher during the interaction with polystyrene than glass in all fungal strains, with the exception of C. metapsilosis (ATCC 96143; Table 5).

Table 4. Adhesion to abiotic substrates by clinical strains belonging to the Candida parapsilosis complex
  1. a

    The results were expressed as number of fungal cells per microscopic field.

  2. b

    The biomass and viability of biofilm were measured by crystal violet incorporation at 540 nm and XTT reduction at 492 nm, respectively.

C. parapsilosis
ATCC 2201923.8 ± 1.3104.0 ± 11.30.792 ± 0.0060.122 ± 0.004
22216.0 ± 1.640.9 ± 3.90.732 ± 0.0170.124 ± 0.023
22552.8 ± 4.7114.6 ± 10.20.788 ± 0.0060.125 ± 0.004
22917.4 ± 2.462.0 ± 3.60.686 ± 0.0300.120 ± 0.007
234142.8 ± 5.7105.1 ± 12.90.822 ± 0.0040.122 ± 0.010
23530.6 ± 3.921.9 ± 2.00.801 ± 0.0180.116 ± 0.024
24127.7 ± 2.427.9 ± 1.90.789 ± 0.0030.100 ± 0.001
2514.3 ± 0.912.4 ± 1.40.766 ± 0.0150.102 ± 0.005
27511.9 ± 1.08.6 ± 0.30.758 ± 0.0230.118 ± 0.011
45413.4 ± 0.852.3 ± 4.00.781 ± 0.0050.118 ± 0.011
C. orthopsilosis
ATCC 9614163.7 ± 3.4300.4 ± 37.30.560 ± 0.0030.134 ± 0.030
24825.5 ± 1.321.6 ± 0.60.804 ± 0.0120.125 ± 0.010
27676.7 ± 3.717.9 ± 1.30.741 ± 0.0150.136 ± 0.019
C. metapsilosis
ATCC 96143126.6 ± 7.8244.9 ± 29.90.524 ± 0.0040.126 ± 0.017
Table 5. Cell differentiation of clinical strains belonging to the Candida parapsilosis complex
Yeasts (%)Pseudohyphae (%)Yeasts (%)Pseudohyphae (%)
  1. a

    The results were expressed as number of fungal cells (yeasts or pseudohyphae) per microscopic field.

C. parapsilosis
ATCC 220196.9 ± 1.3 (36.5)12.0 ± 0.5 (63.5)22.5 ± 11.3 (22.7)76.7 ± 5.2 (77.3)
2223.8 ± 0.6 (23.6)12.2 ± 1.4 (76.4)6.7 ± 3.0 (16.3)34.3 ± 1.5 (83.7)
22524.0 ± 4.0 (45.5)28.8 ± 5.2 (54.5)22.2 ± 4.8 (19.4)92.4 ± 5.4 (80.6)
2297.9 ± 0.8 (45.1)9.6 ± 1.0 (54.9)11.4 ± 0.4 (18.4)50.6 ± 1.7 (81.6)
234106.0 ± 6.8 (74.3)36.8 ± 3.8 (25.7)11.3 ± 3.3 (10.7)95.3 ± 11.1 (89.3)
23524.2 ± 6.5 (79.0)6.4 ± 1.2 (21.0)4.9 ± 0.8 (22.3)17.0 ± 2.0 (77.7)
2419.6 ± 1.5 (34.5)18.1 ± 2.8 (65.5)6.5 ± 0.6 (23.2)21.5 ± 2.3 (76.8)
2511.7 ± 1.0 (40.0)2.6 ± 0.8 (60.0)4.4 ± 0.7 (35.7)8.0 ± 1.8 (64.3)
2757.1 ± 1.8 (60.0)4.8 ± 0.5 (40.0)3.1 ± 0.6 (35.9)5.5 ± 0.9 (64.1)
4548.0 ± 0.6 (59.8)5.4 ± 0.8 (40.2)3.9 ± 1.2 (7.4)48.4 ± 4.8 (92.6)
C. orthopsilosis
ATCC 961416.9 ± 1.3 (36.5)12.0 ± 0.5 (63.5)22.5 ± 11.3 (22.7)76.7 ± 5.2 (77.3)
2485.8 ± 0.8 (22.9)19.7 ± 1.7 (77.1)3.7 ± 0.3 (17.0)17.9 ± 0.4 (83.0)
27642.0 ± 4.2 (54.8)34.7 ± 2.8 (45.2)8.7 ± 0.7 (48.7)9.2 ± 1.2 (51.3)
C. metapsilosis
ATCC 9614385.0 ± 7.8 (65.9)44.0 ± 4.2 (34.1)239.0 ± 37.3 (90.2)25.8 ± 12.5 (9.8)

Biofilm formation

The formation of biofilm by fungal strains of C. parapsilosis complex on polystyrene was quite similar regarding either biomass (mean of 0.772 ± 0.038), except for the reference strains C. metapsilosis (ATCC 96143) and C. orthopsilosis (ATCC 96141), or viability (mean of 0.121 ± 0.010; Table 4). As no significant differences were observed in the biomass and viability, we showed representative images of biofilm formed by strains belonging to the C. parapsilosis complex on polystyrene substrate after 24 h (Fig. 4a) and 48 h (Fig. 4b). On light microscopy, the biofilm architecture was comprised by a homogeneous layer of yeasts as well as by some elongated and pseudohyphal forms (Fig. 4a). Cells were distributed widely on the polystyrene surface; however, some gaps can be observed during the first 24 h (Fig. 4a). The ultrastructure of biofilms developed on glass initially (24 h) consisted of irregular clusters of yeast-like cells and short filamentous forms (Fig. 4c). Mature biofilms (48 h) showed a tridimensional layer containing yeasts and pseudohyphae, exhibiting minimal visible extracellular matrix material (Fig. 4d and e).

Figure 4.

Biofilm architecture of species of Candida parapsilosis complex. Light micrographs showing the in vitro biofilm formation of C. parapsilosis sensu stricto on polystyrene substrate after 24 h (a) and 48 h (b). Note in (a) the presence of single-type cells (oval/rounded yeasts and elongated cells) and filamentous forms (pseudohyphae). Scanning electron microscopy images (c–e) showing the biofilm of C. parapsilosis sensu stricto on glass after 24 h (c) and 48 h (d and e). Note that early-phase biofilm (c) presented clustered cells of yeast-shaped morphology (white arrows) as well as elongated cells and pseudohyphae. A typical mature biofilm was observed after 48 h of adhesion (d and e), consisting of a dense network of yeast-like and pseudohyphal elements as well as a thin and irregular extracellular matrix (open white arrows) that appears in micrographs as a fibrous network between fungal cells. Similar biofilm architectures were observed in C. metapsilosis and C. orthopsilosis (data not shown).

Interaction with cell lineages

In these set of experiments, two distinct cell lineages were used: a nonprofessional (fibroblast) and a professional phagocytic cell (macrophage). The association indexes among fungi and both cell lineages were very dissimilar, varying immensely according to each studied strain. However, as a whole, the interaction process was marginally enhanced to macrophages when compared to fibroblasts (Table 6). Also, the majority of fungal cells were detected inside the fibroblasts and macrophages in almost all fungal strains (Table 6).

Table 6. Interaction of clinical strains of Candida parapsilosis complex with mammalian cells
Species/codeL929aRAW 264.7a
Association indexAdhered fungi (%)Internalized fungi (%)Association indexAdhered fungi (%)Internalized fungi (%)
  1. a

    Association index represents the total number of fungi interacted with animal cells.

C. parapsilosis
ATCC 2201974.7 ± 2.312.987.199.3 ± 2.017.582.5
22241.8 ± 2.558.241.856.8 ± 0.142.457.6
22545.5 ± 2.738.261.867.3 ± 1.233.366.7
22945.3 ± 5.850.149.972.9 ± 0.528.072.0
23463.0 ± 2.017.482.681.6 ± 0.118.881.2
23567.5 ± 2.128.371.775.7 ± 2.525.474.6
24166.8 ± 3.942.058.039.0 ± 1.060.339.7
25163.8 ± 3.927.472.653.8 ± 3.847.152.9
27568.2 ± 0.732.068.056.0 ± 0.644.655.4
45450.9 ± 1.258.341.362.7 ± 1.638.661.4
C. orthopsilosis
ATCC 9614178.8 ± 3.616.283.897.8 ± 0.322.777.3
24864.5 ± 3.546.853.267.0 ± 2.233.966.1
27665.9 ± 0.240.359.776.6 ±
C. metapsilosis
ATCC 9614395.3 ± 1.412.088.099.6 ± 1.717.182.9


Statistically significant correlations were found to exist among the surface N-acetylglucosamine residues, protease production or pseudohyphae formation in species belonging to the C. parapsilosis complex and the adherence to the polystyrene. When the data on adhesion of glass were correlated with data from expression of mannose-/glucose-rich surface glycoconjugates or pseudohyphae formation, significant positive relationships were observed between these parameters. The biofilm formation was positively correlated with the production of protease and negatively related with the fungal cell size. The adhesion to fibroblast cells was positively correlated with the expression of glycoconjugates containing mannose/glucose units as well as production of esterase by species belonging to the C. parapsilosis complex. The expression of surface residues of mannose/glucose, N-acetylglucosamine, and esterase production was positively correlated with the adhesion to macrophages. All correlations are shown in Table 7.

Table 7. Correlations between potential fungal virulence attributes and adhesion to different structures/cells
ParametersSurface molecules/properties, P (r)Hydrolytic enzymes, P (r)Morphological characteristics, P (r)
Man/GluNAGSialic acidCSHProteaseEsteraseCatalasePhytaseHemolysisSizeGranularityFilamentation
  1. a

    < 0.05 denotes significance; r, Pearson's correlation coefficient.

  2. Man, mannose; Glu, glucose; NAG, N-acetylglucosamine; CSH, cell surface hydrophobicity; ND, non-determined.

Polystyrene0.0900 (−0.469)0.0491a (0.534)0.5969 (−0.154)0.9736 (0.009)0.0137a (−0.640)0.0619 (−0.511)0.9511 (0.0180)0.0644 (0.524)0.2687 (0.317)0.0859 (0.475)0.9867 (−0.004)0.0019a (0.752)
Glass0.0267a (−0.589)0.8999 (0.037)0.1945 (−0.368)0.1664 (0.391)0.4793 (−0.206)0.6695 (−0.125)0.4937 (0.199)0.9162 (−0.031)0.2657 (0.319)0.1001 (0.457)0.1892 (0.372)< 0.0001a (0.861)
Biofilm (biomass)0.2881 (0.305)0.3389 (−0.276)0.4779 (0.206)0.9297 (0.025)0.0150a (0.633)0.1384 (0.416)0.4930 (0.200)0.1426 (0.412)0.8181 (−0.067)0.0435a (−0.545)0.8586 (−0.052)0.2608 (−0.322)
Fibroblast0.0364a (0.562)0.3846 (0.252)0.6895 (0.117)0.8068 (−0.071)0.0717 (0.495)0.0495a (−0.533)0.1795 (−0.380)0.7495 (−0.093)0.5995 (0.153)0.0524 (−0.527)0.6332 (−0.140)ND
Macrophage0.0018a (0.755)0.0042a (0.713)0.5061 (0.194)0.4023 (−0.243)0.1130 (0.442)0.0368a (−0.561)0.7596 (−0.092)0.2883 (0.305)0.5505 (0.174)0.5792 (−0.162)0.8480 (0.056)ND


Invasive candidiasis is a leading cause of mycosis-associated morbidity and mortality (Pfaller & Diekema, 2007). Among the Candida species causing human infections, C. parapsilosis accounts for a significant proportion of nosocomial infections, with an increasing prevalence in clinic settings due its propensity to easily colonize hospital environments including medical devices and hands of healthcare workers (Trofa et al., 2008; van Asbeck et al., 2009). Early reports demonstrated that C. parapsilosis strains are more heterogeneous than other Candida spp. In this line of thinking, Tavanti et al. (2005) proposed the creation of three related species named C. parapsilosis sensu stricto, C. orthopsilosis and C. metapsilosis based on significant genetic differences. Since then, an increased interest has flourished in the studies focused on epidemiology, biochemical/metabolic properties, antifungal susceptibilities, virulence factors' expression and pathogenesis of the three species belonging to the C. parapsilosis complex.

Identification of fungal pathogens from clinical specimens at species level, particularly clinical isolates, is important to improve antifungal therapy and patient assistance. The 11 clinical isolates used in the present study were first identified based on morphological and physiological criteria as C. parapsilosis. As expected, the three species belonging to the C. parapsilosis complex were not distinguished by either morphological or physiological levels. Consequently, molecular methodologies must be applied to solve this problem. A PCR-based strategy using species-specific primers failed to correctly identify all the 11 clinical strains, because two of them (designated as 248 and 276) were indiscriminately amplified by two distinct pairs of primers. With this result in hands, we tested two PCR-based restriction endonuclease analysis centered on the amplification of (1) a FKS1 region followed by an EcoRI digestion and (2) a SADH region followed by BanI digestion. Contradictory results were detected regarding the strains 248 and 276, whereas the remaining nine strains were identified as C. parapsilosis sensu stricto. Sequencing of the D1/D2 gene region solved this doubt, showing 100% of sequence similarity among the clinical strains (248 and 276) and two reference strains of C. orthopsilosis. In time, the reference strains (C. metapsilosis – ATCC 96143, C. orthopsilosis – ATCC 96141 and C. parapsilosis sensu stricto – ATCC 22019) were correctly identified by all molecular methodologies applied in our study. Although the identification of SADH-RFLP methodology is widely used, Silva et al. (2009) demonstrated that this method can identify wrongly C. metapsilosis as Corthopsilosis because of a missing BanI restriction site in those clinical strains. In the present work, we also showed that this approach can erroneously identify C. orthopsilosis as C. parapsilosis sensu stricto. Contrarily, the digestion of FKS1 gene by the enzyme EcoR1 showed 100% of concordance with sequencing of ITS1/ITS2 region in the study conducted by Garcia-Effron et al. (2011) as well as herein by the two clinical strains incorrectly identify by both SADH-RFLP (Tavanti et al., 2005) and PCR using species-specific primers (Asadzadeh et al., 2009), reinforcing its promising application in the clinic arena.

Within the complex, C. parapsilosis remained to be the predominant species while C. orthopsilosis and C. metapsilosis were recovered at a much lower incidence. Currently, global epidemiologic data indicate that C. parapsilosis sensu stricto represented 70.7–95.6% of the complex, while C. orthopsilosis and C. metapsilosis corresponded to 4.4–20.4% and 0–9.3%, respectively (Lockhart et al., 2008; Cantón et al., 2011; Ge et al., 2012). Despite the fact that the number of strains used was small, our data agree with the worldwide epidemiologic studies where the frequencies of isolation of these species have been found to be low (e.g. C. orthopsilosis) or even null (e.g. C. metapsilosis). Corroborating this finding, the absence of C. metapsilosis was reported by (1) Romeo et al. (2012) among 97 strains recovered from blood and venous central catheter tips, (2) Asadzadeh et al. (2009) among bloodstream (n = 66) and other clinical specimens (n = 48), (3) Hays et al. (2011) among 116 clinical strains from different anatomical sites, and (4) Tavanti et al. (2005, 2007) from various clinical sources (n > 300). The importance of C. orthopsilosis and C. metapsilosis as human pathogens remains unknown; however, studies point to their significant involvement in human candidiasis (Lockhart et al., 2008; Cantón et al., 2011).

Pathogenic Candida species have developed a wide range of putative virulence factors to assist in their ability to colonize host tissues, cause disease, and overcome host defenses. Among them, extracellular secreted hydrolytic enzymes have gained considerable attention due to their potential roles in pathogenesis and as possible targets for future antimicrobial therapies. Proteases (Hube & Naglik, 2001; Braga-Silva & Santos, 2011), phospholipases (Ghannoum, 2000), esterases (Singh & Mukhopadhyay, 2012), phytases (Tsang, 2011), catalases (González-Párraga et al., 2008), and hemolysins (Nayak et al., 2013) not only facilitate the adherence, tissue/cell penetration, invasion, dissemination, and evasion of host immune responses, but also increases bioavailability of essential nutrients, which are usually in short supply. However, the molecular mechanisms of pathogenicity remain poorly explored for the species of the C. parapsilosis complex.

Regarding the production of hydrolytic enzymes, 100% of the strains studied herein produced extracellular protease, when cultured under aspartic protease inducible conditions (Hrusková-Heidingsfeldová et al., 2009; Braga-Silva & Santos, 2011), most of which (13/14) showed very strong enzymatic activity with Pz values lower than 0.355. Extracellular protease production might be common for C. parapsilosis isolates of various clinical origins, especially those from skin (De Bernardis et al., 1999). Secreted aspartic proteases (Saps) play a role in the virulence of pathogenic Candida spp. C. parapsilosis possesses three genes encoding these enzymes: SAPP1, SAPP2, and SAPP3 (Hrusková-Heidingsfeldová et al., 2009). Recently, Horváth et al. (2012) described that SAPP1 gene was duplicated (SAPP1a and SAPP1b) in the genome of C. parapsilosis and acted as a classic virulence factor, because the deletion of both SAPP1a and SAPP1b genes significantly reduced the capacity of the fungus to grow in human serum as well as to survive inside the human macrophages. There have been contradictory findings in terms of phospholipase and esterase activities in C. parapsilosis. For instance, Ge et al. (2012) showed that 90.5% of C. parapsilosis and 91.7% of C. metapsilosis isolates were phospholipase producers and no difference in phospholipase activity was observed between these two species. On the other hand, both species similarly displayed rare esterase activity, with only one C. parapsilosis and two C. metapsilosis isolates being positive. Shimizu et al. (1996) and Kantarcioglu & Yucel (2002) did not find phospholipase activity in their C. parapsilosis isolates. In our study, we detected differences in the frequency of lipolytic enzymes between the phospholipases and esterases of the species of the C. parapsilosis complex. No phospholipase-positive C. parapsilosis isolates were detected, while 92.9% strains were able to produce esterase. Several reasons have been postulated to explain the wide variation in the lypolytic activity of C. parapsilosis complex, including use of different media for enzymatic test and/or inherent biological variations among isolates (van Asbeck et al., 2009; Ge et al., 2012). Similarly, no consistent results were obtained concerning the hemolytic activity of clinical strains of C. parapsilosis. Luo et al. (2001) found that C. parapsilosis (n = 5 isolates) failed to demonstrate any hemolytic activity. França et al. (2011) reported that the source of C. parapsilosis isolates correlated with the ability to produce hemolytic activity, because isolates recovered from tracheal secretion had higher activity than blood isolates, but did not differ of nail/skin isolates. Overall, the majority of C. parapsilosis isolates (n = 34) produced weak hemolytic activity (França et al., 2011). In accordance, all cutaneous clinical isolates tested herein showed weak ability to lysis the erythrocytes. Phytase is a phosphohydrolase that cleaves phytate in a stepwise manner to release inorganic phosphate and inositol, which are essential nutrients for all living cells (Lei & Porres, 2003). In Candida species, maintaining a supply of inositol and phosphate seems to be especially important for survival, propagation, and virulence (Olstorpe et al., 2009; Tsang, 2011). Phytase-positive phenotype was identified in 11 of 14 strains (78.6%) belonging to the C. parapsilosis complex with a mean Pz of 0.456. Tsang reported that two of four strains (50%) of C. parapsilosis produced phytase with a mean Pz of 0.640. Catalase, which is an enzymatic scavenger of H2O2, is presumed to be an important antioxidant defense in Candida species against neutrophils and macrophages (Lefkowitz et al., 1996; Wysong et al., 1998). For instance, deletion of the catalase gene in C. albicans was associated with increased susceptibility to leukocyte-mediated killing of organisms and to decreased virulence for mice in an experimental model of disseminated candidiasis (Wysong et al., 1998). In different extents, all the C. parapsilosis strains used in the present study had the ability to breakdown H2O2.

Other physiological factors believed to be important for fungal colonization including the physicochemical properties of the cell surface. Cell surface carbohydrates play critical roles in many fundamental fungal processes including interaction with cells and extracellular matrix components (Free, 2013). For instance, Candida cell surface mannan was found to participate in the adhesion to the epithelial cells, recognition by innate immune receptors, and development of pathogenicity (Shibata et al., 2012). The amino sugar N-acetylglucosamine is the key component of chitin, a polysaccharide involved in important structural roles of cell wall. Interestingly, N-acetylglucosamine stimulates the human fungal pathogen C. albicans to induce the expression of virulence genes and a shift from growing as unicellular budding yeasts to instead forming multicellular filamentous hyphal cells (Konopka, 2012). In several fungal species, sialic acids are thought to function as antirecognition molecules eluding host immune system mechanisms (Alviano et al., 1999). These multifunctional carbohydrate units, mannose/glucose, N-acetylglucosamine and sialic acid, were detected at the cell surface of all strains belonging to the C. parapsilosis complex in high (≈90%), medium (≈55%), and low (≈7%) levels.

Cell surface hydrophobicity is recognized as a physical force playing a major role in the initial events leading to the adherence of Candida to inert surfaces (Klotz et al., 1985). Studies have also shown that hydrophobic yeasts are more virulent than their hydrophilic counterparts (Hazen et al., 1986). Our strains had elevated cell surface hydrophobicity with a mean of 95% as measured by a hydrocarbon assay method. In contrast, Panagoda & Samaranayake (1998) described that the relative cell surface hydrophobicity of 24 clinical isolates of C. parapsilosis varied considerably, ranging from 17.5% to 76.6% with a mean of 35.7%. Moreover, a strong positive correlation was observed between the cell length and the relative cell surface hydrophobicity of C. parapsilosis as well as the cell length and adhesion to acrylic surfaces. Also, the hydrophobic properties of Candida cells may depend on the changes in the glycosylation of the mannoproteins (Shibata et al., 2012). Corroborating this statement, our studied fungal strains possessed elevated cell surface hydrophobicity and high expression of mannose-rich glycoconjugates.

It is well known that C. parapsilosis is associated with pronounced capacity to adhere to plastic surfaces and several other implanted devices and, consequently, to the development of candidemia related to catheters (Douglas, 2003). Our results revealed that the clinical strains of C. parapsilosis presented an immense variation regarding the ability to adhere to both polystyrene and glass, showing typical intrastrain differences. Overall, the strains showed predilection to adhere to plastic (mean of 14 strains was 81 fungi per microscopic field) compared to glass (mean of 14 strains was 45 fungi per microscopic field) as well as the number of pseudohyphae was higher than yeasts. Similarly, larger fungal cells (pseudohyphae) of C. parapsilosis bound more avidly to acrylic surface (Panagoda & Samaranayake, 1998). The environmental signals that trigger pseudohyphal differentiation and the signaling pathways that transduce these signals in C. parapsilosis are not yet well known. We also demonstrated that filamentation was positively correlated with the adhesion of C. parapsilosis complex strains to either polystyrene or glass substrates. Contrarily, different surface glycoconjugates must be involved in the initial (2 h) adhesive steps, because the expression of mannose-/glucose-rich glycoconjugates correlated strongly with interaction to glass as well as the expression of N-acetylglucosamine correlated with polystyrene adherence. However, these glycoconjugates did not influence the biofilm formation in strains belonging to the C. parapsilosis complex, because all of them were able to produce similar biofilm concerning both biomass and cellular viability parameters after 48 h of interaction with polystyrene. Similar biofilm architectures, consisting of different morphotypes such as oval/rounded yeasts, elongated cells, and pseudohyphae as well as a thin and irregular extracellular matrix, were clearly developed on plastic or glass in all strains of C. parapsilosis complex and detected by both light and scanning electron micrographies. These results concur with the findings of some authors (Kuhn et al., 2002; Laffey & Butler, 2005; Oliveira et al., 2010). For instance, Lattif et al. (2010) reported that clinical isolates of C. parapsilosis, C. metapsilosis, and C. orthopsilosis were able to form biofilm with similar surface topography and architecture on abiotic surface (silicone disks). Biofilm formation is considered a virulence factor due to the ability to confer resistance to antifungal therapy and protect the fungal cells from host immune responses (Tumbarello et al., 2007). Tavanti et al. (2010) showed that biofilm production after 24 h of interaction with plastic substrate and protease secretion was negatively correlated in C. parapsilosis sensu strictu strains isolated from different geographical regions and body sites. Those authors also proposed that protease activity plays a role in detachment and release from a mature biofilm, via degradation of C. parapsilosis adhesins and/or extracellular matrix components. Justifying this hypothesis, our data revealed that mature biofilm formed in plastic substrate by C. parapsilosis strains positively correlated with protease production.

For Candida to persist in the host and induce infectious process, it must be able to adhere and invade to the biotic surfaces as host cells, tissues, and extracellular matrix barriers (Sundstrom, 2002). Candida parapsilosis was able to interact with different mammalian cell lineages, in different extents, and in most cases in a strain-dependent manner (reviewed in Nosek et al., 2009; van Asbeck et al., 2009; Silva et al., 2012). To exemplify, Panagoda et al. (2011) demonstrated a significant (< 0.0001) intraspecies variation in adherence among isolates of C. parapsilosis to human buccal epithelial cells ranging from 23.5 to 154.3 fungi per 50 epithelial cells. Our results corroborated these findings, because the fungal strains had the ability to interact with both fibroblasts (association indexes ranging from 45.3% to 95.3%) and macrophages (association indexes ranging from 39.0% to 99.6%). Moreover, a typical strain-dependent variation regarding the association index was clearly reported after 2 h of contact, in which most of the fungi was located inside the mammalian cells. Adding knowledge about the interaction processes between C. parapsilosis complex and host cells, our data revealed that surface glycoconjugates containing both mannose and glucose were relevant to adhesive events with both fibroblast and macrophage cells. Additionally, N-acetylglucosamine also positively favored the binding of fungal strains with macrophages. Based on the saccharide specificity of the agglutinin UEA I, Lima-Neto et al. (2011) suggested that L-fucose residues on cell surface glycoconjugates of C. parapsilosis strains (n = 12) represented recognition molecules for interactions between yeasts and epithelial cells (r = 0.6985, P = 0.0045). The esterase activity of C. parapsilosis complex was linked to interaction with both fibroblast and macrophage cells. The role of lipases (esterases and/or phospholipases) in fungi capable of causing cutaneous and subcutaneous mycoses is probably to hydrolyze fats in the subcutis and the fatty acid residues used as a nutrient source by the invading fungus. Paraje et al. (2008) showed that a 70-kDa extracellular lipase purified from C. albicans directly induced cytotoxicity and promoted the deposition of lipid droplets in the cytoplasm of macrophages and hepatocytes.

According to the studies conducted in vitro with reconstituted human tissue cultures derived from oral and epidermal epithelium cells, C. orthopsilosis isolates caused similar damage as C. parapsilosis, while C. metapsilosis was less virulent (Gacser et al., 2007). In a similar way, Orsi et al. (2010) showed that C. metapsilosis isolates were more susceptible to microglia-mediated antifungal activity, as compared with those of C. parapsilosis and C. orthopsilosis. Also, Bertini et al. (2013) described that while C. parapsilosis and C. orthopsilosis strains showed similar adhesion capabilities, C. metapsilosis isolates displayed a significantly lower ability to adhere to human buccal epithelial cells. Differences in virulence factors between species could not be determined at a statistically significant level because of the small numbers of the newly described isolates. However, some findings could be extrapolated taking into account that all samples belong to the same fungal C. parapsilosis complex. In this sense, we reinforce the heterogeneity of these fungal complex regarding morphological, biochemical, and genetic features. Collectively, growth pattern, production of hydrolytic enzymes, expression/exposition of surface glycoconjugates, adhesion to abiotic substrates, differentiation process, biofilm formation, and interaction with mammalian cells were clearly noticed as a typical strain-specific fashion.


This study was supported by grants from the Brazilian Agencies: Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), Fundação de Amparo à Pesquisa no Estado do Rio de Janeiro (FAPERJ) and Fundação Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES). Automated sequencing was performed using the genomic platform/DNA sequencing at Fundação Oswaldo Cruz – PDTIS/FIOCRUZ (RPT01A), Brazil.