Microbiological characteristics of clinical isolates of Cryptococcus spp. in Bahia, Brazil: molecular types and antifungal susceptibilities.

To determine the profiles of susceptibility to antifungal and the genotypes of clinical isolates of Cryptococcus in Bahia, Brazil, 62 isolates were collected from cases of meningitis in the period from 2006 to 2010. Their susceptibilities to fluconazole, itraconazole, amphotericin B and 5-flucytosine were determined by the broth microdilution technique described by the Clinical and Laboratory Standards Institute and genotyping of the URA5 gene was accomplished by restriction fragment length polymorphism. C. neoformans accounted for 79% of the identified yeast and C. gattii represented the remaining 21%. Evaluation of the genotypes determined that 100% of the C. gattii isolates belong to the VGII genotype, and 98% of the C. neoformans isolates belong to the VNI genotype. Determination of susceptibility revealed isolates resistant to fluconazole (4.8%), 5-flucytosine (1.6%) and amphotericin B (3.2%); the stratification of sensitivity results for each species showed significant differences in susceptibility to azoles. This study is the first to describe the susceptibility profiles of molecular and clinical isolates of Cryptococcus in Bahia, Brazil. The high percentage of C. gattii isolates belonging to the VGII genotype and its lower susceptibility to antifungal agents highlight the importance of knowing which species are involved in cryptococcal infections in northeastern Brazil.

Introduction

Cryptococcal meningitis, an important opportunistic infection in HIV-positive patients in developing countries, is caused by Cryptococcus spp., which are encapsulated yeasts that are spread worldwide. The genus has two species most commonly associated with infection in humans, Cryptococcus neoformans (serotypes A, D and the AD hybrid; VNI – IV genotypes) and Cryptococcus gattii (serotypes B and C; VGI – IV genotypes), which differ genotypically, phenotypically and epidemiologically [1-3].

Immunocompromised patients are more frequently infected by C. neoformans, while C. gattii has emerged as an important cause of infection in immunocompetent individuals, as illustrated by the recent outbreak on Vancouver Island (Canada) by the VGII genotype, with molecular evidence of spread to the northwestern United States [4-6]. In Brazil, there are few studies that illustrate the distribution of this genotype [7, 8]. The yeast is acquired from the environment, and its tropism for the central nervous system leads to meningitis [9, 10]. The American Society of Infectious Diseases suggests that the treatment of cryptococcal meningitis be initiated with an induction therapy of amphotericin B alone or combined with 5-flucytosine, followed by consolidation and maintenance therapy with an azole [11].

There are few reports of antifungal resistance in isolates of Cryptococcus [12-14]. Although some authors suggest that it remains stable [15-17], the ARTEMIS global antifungal surveillance study showed increased fungal resistance to fluconazole. In the period from 1997–2000, the rate of resistance to fluconazole was 7.3%; in 2005–2007, this rate increased to 11.7%, with the highest rates found in Latin America (13.6%) and Africa (12.4%) [18]. The authors propose that the increased resistance may be limited to areas where antiretroviral therapy is still not effective, such as Spain [13], Cambodia [12] and Africa [19, 20].

Little is known about the frequency and susceptibility profile of Cryptococcus species in northeastern Brazil. Thus, this study aims to determine the antifungal susceptibility profiles and the frequencies of species and genotypes of clinical isolates of Cryptococcus spp. in the state of Bahia in northeastern Brazil.

Materials and methods

Source and identification of clinical isolates

We collected 62 clinical isolates from 62 cases of cryptococcal meningitis diagnosed at the Hospital Couto Maia, Bahia, from 2006 to 2010. After 48 hours of growth on Sabouraud dextrose agar (Acumed, New York, NY), the genus was identified based on the demonstration of typical encapsulated cells in India ink preparations and the development of brown pigmentation on the surface of Niger seed agar; classical biochemical tests were also conducted for the identification of yeasts [21]. The identification of C. neoformans and C. gattii was accomplished using canavanine-glycine-bromothymol blue agar [22]. This study was approved by the ethics committee of the Hospital Couto Maia.

Reference strains

Cryptococcus complex reference strains were kindly provided by the Mycology Laboratory of the Institute of Clinical Research Evandro Chagas - Oswaldo Cruz Foundation and included WM 148 (VNI), WM 626 (VNII), WM 628 (VNII), WM 629 (VNIV), WM 179 (VGI), WM 178 (VGII), WM 161 (VGIII) and WM 779 (VGIV) [23]. Reference strains from the American Type Culture Collection (ATCC) included Candida parapsilosis 22019 and C. krusei 6258, which were also used for quality and reproducibility controls during susceptibility testing.

Antifungal susceptibility tests

The susceptibility tests against fluconazole, itraconazole, amphotericin B and 5-flucytosine (Sigma Aldrich Quimica SA, St. Louis, MO) were performed as described in document M27-A3 of the Clinical and Laboratory Standards Institute (CLSI) [24]. The minimum inhibitory concentrations (MICs) were determined by the lowest antifungal agent concentrations that inhibited 50% fungal growth compared to the control growth (without antifungal) for fluconazole, itraconazole and 5-flucytosine; the MICs for amphotericin B were determined by the lowest concentrations that inhibited 100% growth.

DNA extraction

For the extraction of genomic DNA, the mechanical lysis method was used after digestion of the capsule with urea buffer [25].

Restriction fragment length polymorphism (RFLP)—URA5 gene

RFLP analysis using the URA5 gene was performed as described by Meyer and colleagues (2003) using the primers URA5 (5′ATGTCCTCCCAAGCCCTCGACTCCG3′) and SJ01 (5′ TTAAGACCTCTGAACACCGTACTC3′) [23]. RFLP patterns were assigned visually by comparing them to the patterns obtained from the standard-type strains (VNI-VNIV and VGI-VGIV).

Statistical analysis

Statistical analysis was performed by the least squares method using SPSS 17.0 software (SPSS Inc, Chicago, Illinois). Statistical significance was defined at p < 0.05.

Results

All yeasts were characterized as belonging to the genus Cryptococcus spp., and there were no biochemical differences among isolates; 79% (49/62) of the isolates were C. neoformans and 21% (13/62) were C. gattii. The distributions of the MIC values of the 62 Cryptococcus spp., separated by species, to the antifungals tested are shown in Fig. 1. The antifungal susceptibility profiles are summarized in Table 1, showing isolates of C. neoformans that were resistant to fluconazole, amphotericin B and 5-flucytosine and C. gattii isolates that were significantly resistant to fluconazole.

Fig. 1

The distributions of MIC (mg/L) values for fluconazole (a), itraconazole (b), amphotericin B (c) and 5-flucytosine (d) for 62 isolates of Cryptococcus spp. C. neoformans (empty bars) and C. gattii (black bars)

Table 1

Susceptibility profiles and in vitro activities of antifungal agents against clinical isolates of Cryptococcus spp

Antifungal agent	Species	Susceptibility profile %			p value*
		S	S-DD	R
Fluconazole	C. neoformans	83.7 (41/49)	14.3 (7/49)	2 (1/49)	0.013
	C. gattii	38.5 (5/13)	46 (6/13)	15.5 (2/13)
Itraconazole	C. neoformans	89.8 (44/49)	10.2 (5/49)	0	0.001
	C. gattii	46.1 (6/13)	53.9 (7/13)	0
Amphotericin B	C. neoformans	96 (48/49)	0	4 (2/49)	0.622
	C. gattii	100 (13/13)	0	0
5-Flucytosine	C. neoformans	24.5 (12/49)	7.5 (36/49)	2 (1/49)	0.163
	C. gattii	53.9 (7/13)	46.1 (6/13)	0

S susceptible, S-DD susceptible, dose-dependent, R resistant

The URA5 gene RFLP analysis revealed that 98% (48/49) of isolates of C. neoformans presented the profile of the molecular genotype VNI and 2% (1/49) of genotype VNII, while 100% (13/13) of isolates of C. gattii showed VGII genotype profile (Fig. 2).

Fig. 2

Representative RFLP profiles of the URA5 genes from Cryptococcus spp. obtained by double-digestion with HhaI and Sau96I. Lanes 1 and 14 are the molecular markers; lanes 2, 3, 4, and 5 are genotypes VGI, VGII, VGIII and VGIV, respectively; lanes 6, 7, 8, and 9 are genotypes VNI, VNII, VNII and VNIV, respectively; lane 10 is the clinical isolate of C. gattii; lanes 11 and 12 are clinical isolates of C. neoformans; lane 13 is the URA5 gene amplified product

Discussion

The epidemiology of Cryptococcus has been widely studied around the world, but in Brazil, information regarding the distribution of species is still fragmentary and incomplete, reflecting only differences in frequency. Cryptococcosis caused by C. neoformans is an important cause of morbidity and mortality in immunocompromised individuals worldwide, but the number of published reports of infection by C. gattii in patients without immunosuppression is increasing [3–5, 26].

C. gattii was initially considered to be restricted to tropical and subtropical Australia, Southeast Asia and some African regions. Currently, however, it is expanding to the northwest Pacific and North America, which suggests it is adapting to a new climatic niche or that global warming may provide favorable conditions for its growth [5, 27, 28]. In Brazil, the analysis of Cryptococcus spp. isolates from various regions has shown that C. gattii is significantly more prevalent in the north-northeast of the country and that C. neoformans is more prevalent in the south-southeast of the country; however, there are few studies in the north-northeast that report the distributions of these pathogens [29].

Santos and colleagues (2008) analyzed the distribution of species in 56 isolates from 43 patients with meningitis in the state of Pará and identified a frequency of 35.7% (20/56) for C. gattii infection [30]. In the southeast, all isolates from different areas of São Paulo belonged to C. neoformans [31]. However, Almeida and colleagues (2007) analyzed 83 samples from 38 patients and found only four isolates of C. gattii from only one patient [32]. In Minas Gerais, 11.4% of 35 isolates belonged to the species C. gattii [33]. In the state of Mato Grosso, in the Midwest, 16.6% of 26 samples of 26 cases of cryptococcosis were C. gattii isolates [34]. These results reveal a high number (13) of isolates of C. gattii (21%) compared with other studies [30-34] and suggest that Bahia is an important source of infection by C. gattii.

Genotypic determination revealed a predominance of the VNI genotype among C. neoformans isolates, which have been frequently reported in various regions of the world [35-40]. All C. gattii isolates belonged to the VGII genotype. This genotype was reported as the causal agent of the cryptococcosis outbreak that occurred on Vancouver Island (BC, Canada) [4, 41]. This genotype has also been reported in Brazil [7, 8] demonstrating its potential to cause severe disease in immunocompetent hosts, as it was recognized as the main agent of a primary cryptococcosis endemic in the north [29, 30].

The determination of susceptibility profiles revealed the presence of antifungal-resistant C. neoformans isolates. Antifungal resistance is among the factors that can contribute to treatment failure because previous exposure to a risk factor often leads to resistance [42, 43]. Our findings are consistent with authors who suggest that there are differences between C. neoformans and C. gattii sensitivities [44-46], although others disagree [14, 47, 48]. It is known that an infection caused by C. gattii has a less favorable response to antifungal therapy and a relatively worse prognosis compared to infection by C. neoformans [49, 50]; however, the mechanisms underlying this difference are not clear.

Conclusion

This study is the first to describe the susceptibility profiles of molecular and clinical isolates of Cryptococcus in Bahia, Brazil. The high percentage of isolates from the C. gattii genotype VGII, in addition to its lower susceptibility to antifungal agents, highlights the importance of knowing the species involved in cryptococcal infection in this state. Surveillance studies are needed to elucidate the environmental niche of this yeast and to monitor trends in antifungal susceptibility and the distribution of species in Bahia, Brazil.