Antifungal susceptibility and phenotypic virulence markers of Candida species isolated from Nepal

Candida species are part of the commensal microflora in many anatomical sites of the human body; however, breach in the integrity of the body part and impaired immunity of the host can lead to invasive candidiasis. A number of virulence determinants could contribute towards its pathogenicity. Thus we attempted to evaluate the in vitro expression of different virulence factors among clinical isolates of Candida species and assayed their susceptibility patterns against a range of antifungal agents. Of the total of 71 isolates we obtained, 48 (67.6%) were Candida albicans, 11 (15.49%) Candida tropicalis, 09 (12.67%) Candida glabrata and 03 (4.22%) were Candida krusei. Proteinase, phospholipase and esterase production could be revealed amongst 43 (60.56%), 44 (61.97%) and 49 (69.01%) isolates respectively. None of the isolates showed DNAase activity. Fifty-five (77.39%) isolates were biofilm producers, and 53 (74.6%) exhibited high cell surface hydrophobicity.


Introduction
Candida species are part of the commensal microflora in many anatomical sites of the human body [1]. If host immunity is compromised, or there is disruption in the skin or mucosal site where Candida remains as a commensal, there is always a chance for Candida to invade and cause a wide range of infections with significant morbidity and mortality [2,3]. Though Candida albicans has been associated with most human infections, there has been increasing reports of infections due to non-albicans Candida species in the recent past [4,5]. A number of virulence attributes such as biofilm formation, proteinase, esterase, phospholipase activities and drug resistance contributing towards the pathogenicity of Candida have been proposed. Thus our study was conducted to determine and compare in vitro production of virulence factors by Candida species and their antifungal the biofilm activity. Qualities of all assays were checked using known positive and negative controls.

Biofilm formation
The method standardised by Malek et al. [6] was followed to develop biofilms in 96 well microtiter plates. Measurement of biofilm mass by quantitative method was performed using crystal violet for staining the biomass and metabolic activity of the biofilm cells was assessed colorimetrically based on reduction of sodium 39-[1-(phenylamino-carbonyl)-3,4-tetrazolium]-bis(4methoxy-6-nitro)benzene sulfonic acid hydrate (XTT) as described elsewhere [7]. Biomass was also demonstrated by fluorescent microscopy with calcofluor white staining (Fig. 1). Known biofilm producer and non-biofilm producer Candida strains served as positive and negative controls respectively.

Proteinase and phospholipase activities
Proteinase activity was assessed by bovine serum albumin (BSA) agar based assay as described previously [8]. The presence of halo surrounding the growth representing proteinase activity was observed by staining with amido black. The egg yolk agar method as described earlier was employed for determining the phospholipase activity [9]. Pz (precipitation zone) values for both the tests were calculated according the parameters noted earlier [8,9]. Known proteinase and phospholipase positive and negative Candida strains served as controls.

Esterase, deoxyribonuclease and cell surface hydrophobicity
Esterase activity was noted in Tween-80 agar as described previously [10] and test for cell surface hydrophobicity (CSH) was performed in accordance with the earlier devised technique [11]. DNase production was measured according to the standard protocol [11], using ATCC 25923 standard strain of Staphylococcus aureus as positive control. Known Esterase and CSH positive and negative Candida strains served as controls.

Antifungal susceptibility testing
All 71 isolates were subjected to antifungal susceptibility testing against amphotericin B, voriconazole, fluconazole and caspofungin by microbroth dilution method based on the Clinical and Laboratory Standards Institute M27-A3 standard [12]. Tests were interpreted by visual method with the help of reading mirror after 24 h of incubation at 37 °C. Candida parapsilosis ATCC 95142 and C. albicans ATCC 90028 were used as controls. Antifungal compounds were obtained as pure powders from the manufacturer, Sigma-Aldrich Laborchemikalien GmbH, Germany.

Statistical analysis
Descriptive statistics were used for analyzing the data entered in Microsoft Excel 2010 by Statistical Analysis System (SAS) and Origin Pro 2016. MIC values of different antifungal agents against C. albicans and non albicans Candida were expressed in terms of range, median and geometric mean. Variations among MIC values of antifungal agents against biofilm producing and non-biofilm producing Candida albicans and non-albicans Candida species were assessed by using minimum, maximum, median, and 90th percentile and box plot.

Results
Out of a total of 71 Candida isolates, 48 (67.6%) were C. albicans; 11 (15.5%) were C. tropicalis; 9 (12.7%) C. glabrata and 3 (4.2%) were C. krusei. As depicted in Table 1, proteinase, phospholipase and esterase activity could be detected amongst 43 (60.6%), 44 (62%), and 49 (69%) of the isolates respectively. None of the isolates produced DNase. CSH, was observed among 54 (76%) of the 71 isolates. As many as 55 (77.4%) out of the total 71 isolates were found to be biofilm producers as evidenced by metabolic activity and biomass production ( Fig. 1). Majority, i.e. 40 (74.07%) of the 54 having high cell surface hydrophobicity produced biofilms. 36 (75%) out of 48 C. albicans strains produced proteinase in contrast to only 7 (30%) of the 23 non-albicans Candida species. Similarly, higher numbers of C. albicans strains were found to be phospholipase and esterase producers as compared to non albicans Candida (Table 1). Isolation rates of C. albicans from blood and indwelling devices were found to be much higher as compared to non albicans Candida species. Similarly biofilm production was seen among 84-100% of the blood and device isolates (Additional file 1: Table S1). Tables 2 and 3 denote the antifungal susceptibility patterns of the isolates. Isolates were classified as sensitive, intermediately sensitive and resistant to each antifungal agent in accordance with the break point criteria laid down by CLSI [10]. During data analysis, both sensitive and intermediately sensitive isolates were categorized as one group, i.e. sensitive. As high as 97.9% (47/48), 85.4% (41/48) and 77% (37/48) C. albicans isolates were sensitive to amphotericin B, caspofungin and voriconazole respectively. Overall, 95.7% (22/23) of non-albicans strains were found to be susceptible to amphotericin B and caspofungin. Amongst C. tropicalis all 11, i.e. 100% were sensitive to amphotericin B and caspofungin (Table 3). Fluconazole sensitivity of C. albicans, C. tropicalis and C. krusei ranged between 33.3 and 52%. A total of 82.6% (19/23) of the non-albicans Candida were sensitive to voriconazole, only 56.5% (13/23) were sensitive to fluconazole.
Median MICs and geometric mean MIC (GMM) values of fluconazole were found much higher as compared to  (Table 3). While the median MICs and GMM values were found to be the lowest with respect to caspofungin in both C. albicans (0.00625 and 0.068 μg/ml) and non-albicans Candida (0.0625 and 0.0628 μg/ml) species, those for voriconazole were 0.125 and 0.1 μg/ml respectively against C. albicans and 0.0625 and 0.077 μg/ml respectively against non-albicans Candida. Based upon the median MIC data, we determined the number of strains showing high MIC values (higher than the median MIC) and those exhibiting low MIC values (lower than the median MIC), in order to see if there was any correlation between biofilm production and drug resistance. A significant difference could be noted amongst the non-albicans Candida, against amphotericin B, fluconazole and caspofungin. As depicted in Fig. 2, a large number of the non-biofilm producing C. albicans strains showed high MIC values against amphotericin B, fluconazole and voriconazole. Similarly, non-albicans Candida that were non-biofilm producers exhibited moderately higher MICs against amphotericin B.

Discussion
Many invasive Candida infections are attributable to some of the potential virulence factors of the organism such as proteinase, phospholipase and esterase. Majority (72-77%) of the C. albicans strains in this study were capable of producing proteinase, phospholipase and esterase. A high rate of phospholipase (94.7%) and a moderately high rate of proteinase (73.7%) production amongst C. albicans clinical isolates were reported earlier [13][14][15]. Proteinase as a major virulence determinant of both C. albicans and non albicans Candida in invasive infections was documented earlier [16]. Gokee et al. [17] detected proteinase in 89.7% of C. albicans isolates, and only in 25.8% of the non-albicans isolates. Inci et al. [18] reported that 95% C. albicans and 24% non-albicans Candida were proteinase producers. We noted proteinase production among 75% of our C. albicans isolates and only 30% of the non-albicans isolates.
The role of esterase in the pathogenesis of invasive candidiasis is debatable [14][15][16]. However, earlier studies [18] demonstrated that both C. albicans and non albicans Candida species showed esterase activity. In our study, esterase was detected amongst 77% of C. albicans isolates as compared to 52% non-albicans isolates, difference being marginally higher among the C. albicans isolates. Tellapragada et al. [19] found no significant difference in the esterase activities among invasive and non-invasive Candida. They did not, however, compare this observation between C. albicans and non albicans isolates.
In the present study, resistance rates for the azoles were substantially higher as compared to amphotericin B or caspofungin, especially in C. tropicalis (Table 2). Additionally MIC ranges, median MICs and geometric mean titres were higher for fluconazole both for C. albicans and non-albicans Candida (Table 3). These findings were similar to those reported by others [20], who proposed that azole resistance in Candida was of concern because azoles like fluconazole happened to be the most common antifungal agent used for the treatment and prophylaxis of candidiasis in organ transplant recipients. In yet another study [5], nosocomial isolates of C. albicans were shown to have far lower sensitivity rates towards fluconazole. Seneviratne et al. [21] very recently reported that  (23) 14 (29)  31.7% of the Candida isolated from blood were resistant to fluconazole. Apart from antifungal drug resistance in Candida, another major virulence attribute of this organism is production of biofilm that could lead to treatment failure and recurrence of infection. Tallaprageda et al. [19] noted high rate of biofilm production among the Candida isolates from blood stream and other invasive infections. Hassan et al. [22] found that significantly larger number of C. albicans isolates were biofilm producers as compared to the non-albicans Candida. While the exact reason for the higher rate of biofilm production among C. albicans was ill-understood, scanning electron microscopy studies of complex biofilm architectures attributed the integrity and strength of these biofilms to the higher number of hyphal elements produced by C. albicans than C. tropicalis and C. parapsilosis. The latter two species formed biofilms of lesser strength and the biofilm formed by these two species primarily consisted of micro colony aggregates of yeast cells [19]. In another recent investigation, Sariguzel et al. [11] detected biofilm among 33% of non albicans Candida as compared to 25% of C. albicans. We also observed a comparatively higher rate of biofilm production among non albicans Candida as opposed to C. albicans (83% vs. 75%; Table 1). Notwithstanding the aforementioned variability in the rate of biofilm production among different Candida species, it is noteworthy that such high degree of biofilm forming ability among clinical Candida isolates reflects the potential of these organisms to cause invasive disease [4,22]. Thus, biofilm production could be a classic prototypical phenotypic marker of pathogenicity of a distinct population of Candida, differentiating these from mere commensals [1,2,23].
We observed that non-biofilm producing C. albicans and non albicans Candida showed high MICs towards fluconazole, amphotericin B and voriconazole. This correlation, however, could not be detected among C. albicans isolates (Fig. 2). Unlike in bacterial pathogens [24], studies involving correlation between biofilm production and multidrug resistance among Candida are scanty [20, [25][26][27]. Our study, however, highlighted that majority of the non albicans Candida strains that were biofilm producers had shown high MICs towards fluconazole (Fig. 2).

Conclusion
Non-albicans Candida species are emerging as potential threats to cause invasive disease and posing a therapeutic challenge. Detection of high rate of biofilm activities among non-albicans Candida species along with high level of fluconazole resistance warrant wider surveillance of Candida isolates in order to clearly define the exact role of biofilms and drug resistance in invasive candidiasis.

Limitations
Non-albicans Candida isolates in our study were few. Thus in order to hypothesise that more of non-albicans Candida were capable of forming biofilms as compared to C. albicans, further studies, including higher numbers, would be required. Testing of hypothesis of inferential statistics was not applicable for this study because of the inadequate sample size.