Rapid Identification of Candida glabrata Based on Trehalose and Sucrose Assimilation Using Rosco Diagnostic Tablets

Historically, Candida glabrata has been considered a saprophyte of the normal flora of healthy individuals, rarely causing serious infection in humans. Within the past two decades however, non-Candida albicans Candidaspecies have emerged as important opportunistic pathogens in immunocompromised patients. Although C. albicansremains the most frequent cause of severe candidiasis, C. glabrata is now recognized as an important nosocomial agent (2, 10). In our laboratory, this species accounts for 16% of all Candida species isolated, and as such, it is the second most frequent yeast cultured from clinical samples. Due to the commonly occurring innate or acquired resistance to fluconazole, rapid identification of C. glabrata is essential to guide antifungal therapy (2, 3).

Although morphological features of the colonies in cultures, such as glossy, smooth, and dome-shaped colonies, and the presence of small spherical yeasts upon microscopic examination may be an aid for the presumptive identification of C. glabrata, definitive identification requires additional tests. However, commercially available methods are time-consuming and expensive. Different methods for rapid screening and identification ofC. glabrata have thus been developed. They are based on the ability of C. glabrata to rapidly use trehalose. By these techniques, false-positive results have been reported, especially with Candida tropicalis (1, 7, 8). To overcome these difficulties, we developed a new and simple method, using Rosco diagnostic tablets (purchased from EUROBIO, Les Ulis, France), based on trehalose and sucrose assimilation. Indeed, among yeast species commonly isolated in a clinical mycology laboratory, C. glabrata is the only one which utilizes trehalose but not sucrose (6).

Altogether, 440 clinical isolates were tested to characterize the present method (see Table 1). These included C. glabrata (n = 180) , C. albicans(n = 84) , C. tropicalis (n = 62) , Candida kefyr (n = 38) ,Candida krusei (n = 22) , Candida parapsilosis (n = 19) , Candida sake(n = 9) , Candida famata (n = 7) , Candida lusitaniae (n = 4) ,Candida guilliermondii (n = 2) ,Candida sphaerica (n = 1) , Saccharomyces cerevisiae (n = 11) , and Trichosporon cutaneum (n = 1) . These isolates originated from a variety of clinical samples: stool (n = 149) , throat swab (n = 82) , urine (n = 87) , tracheal secretion and sputum (n = 57) , skin (n = 23) , blood and catheter (n = 12) , vaginal (n = 10) , nose (n = 8) , and other (n = 12) samples. The reference strains C. albicans ATCC 26278, C. parapsilosis ATCC 22019,C. glabrata ATCC 66032, C. krusei ATCC 6258, Candida zeylanoides CBS 619 and 947, Candida conglobata CBS 2018 and 2019, Pichia farinosa CBS 185 and 2001, and Prototheca wickerhamii IP 1202-79 were also included in the study (see Table 2). In addition, a series of clinical isolates obtained from patients receiving antifungal therapy at the time of sampling was tested (see Table 3). C. albicanswas identified by germ tube formation at 37°C for 4 h in human serum. Other Candida species were identified by their carbohydrate assimilation profiles with the API ID 32C kit (bioMerieux sa, Marcy l'Étoile, France). In order to avoid any error possibly resulting from the presence of multiple yeast species on primary cultures of the clinical samples, yeasts were recovered from the API ID 32C or germ tube test samples and subcultured on Sabouraud dextrose agar for the trehalose-sucrose assimilation tests.

The Rosco diagnostic tablets used in this investigation contain 2.5 mg of a specific carbohydrate substrate, together with a weak buffer and a pH indicator (phenol red) that turns from red to yellow upon utilization of the carbohydrate. For each isolate tested, a colony was suspended in 600 μl of sterile 0.85% NaCl, and if necessary the suspension was adjusted to a no. 2 McFarland standard by dilution with 0.85% NaCl or the addition of yeasts from a second colony. A 300-μl aliquot of the resulting suspension was incubated at 37°C with either trehalose or sucrose tablets. Results were read after 4 h and 24 h of incubation. A test was positive if a yellow color was observed in the reaction tube. Tests were considered negative if a pink or red color was observed.

The results of the tests performed with clinical isolates are shown in Table 1. All C. glabrataisolates gave a positive reaction with trehalose tablets after 4 or 24 h of incubation. None of the C. albicans, C. kefyr, C. krusei, C. parapsilosis,C. famata, C. lusitaniae, C. guilliermondii, C. sphaerica, or T. cutaneum isolates were able to utilize trehalose. OneC. sake isolate and one S. cerevisiaeisolate gave positive reactions with trehalose after 24 h of incubation. Among the 62 C. tropicalis strains tested, 8 gave positive reactions after 4 h of incubation and 21 gave positive reactions after 24 h of incubation. All C. tropicalis, C. sake, and S. cerevisiaeisolates that gave positive reactions with trehalose at either 4 or 24 h were also positive for sucrose utilization at 4 h. The trehalose-sucrose assimilation assay was also applied to reference strains of C. conglobata, C. zeylanoides, P. farinosa, and P. wickerhamii, organisms that are occasionally isolated in clinical laboratories and that have the ability to utilize trehalose (6, 9). None of these four species gave a positive result by the trehalose assimilation test (Table 2) in four separate experiments that included C. glabrata and S. cerevisiae as controls for trehalose or sucrose assimilation. Finally, analysis of clinical isolates from patients receiving antifungal treatment at the time of sampling showed that allC. glabrata isolates obtained after antifungal treatment had the trehalose-positive, sucrose-negative profile at 4 h (Table 3). The combination of the trehalose and sucrose assimilation tests thus allowed identification of C. glabrata in 4 h with 100% sensitivity and specificity.

Several groups have described trehalose assimilation-based assays for the identification of C. glabrata (Table4). Most of these methods, however, present one or several disadvantages such as the need for homemade reagents, the need for large yeast inocula, false-positive results withC. tropicalis, or the need for the use of a microtiter plate format, which implies the simultaneous testing of a large number of isolates for cost-effectiveness. The Rosco system for determination of the carbohydrate metabolism of microorganisms was previously developed in clinical bacteriology laboratories to identify fastidious bacteria (4, 5), and our goal was to adapt this system to identification of a yeast organism. In preliminary experiments, the proportions of C. glabrata isolates at 1.5, 1, or 0.5 McFarland standard suspensions that showed trehalose assimilation at 4 h were 100, 60, and 0%, respectively. Suspensions of C. glabrata at a no. 3 McFarland standard suspension always produced the trehalose-positive, sucrose-negative profile at 4 h. However, higher densities resulted in cloudy suspensions that made interpretation of colors difficult. Any inoculum from a no. 1.5 to a no. 3 McFarland standard suspension thus appeared adapted to the 4-h detection of trehalose assimilation by C. glabrata. We therefore decided to characterize the method with a no. 2 McFarland inoculum, which provided a reasonable margin of error, since a no. 1.5 McFarland inoculum might, due to inaccurate calibration, result in a suspension at no. 1 McFarland standard, which could cause a false-negative result. The approach described herein presents several advantages over previously described techniques: (i) results are obtained in 4 h with a 100% sensitivity and specificity, (ii) the test format is adapted to the work flow of a clinical laboratory, (iii) when the test is performed with a 48-h-old culture, a unique middle-size colony is sufficient, (iv) all reagents are commercially available, and (v) the method is cost-effective. Further evaluation will be necessary to determine the usefulness of the proposed method in a clinical mycology laboratory, with special emphasis on yeast species that were not tested during the current investigation.

• Copyright © 2001 American Society for Microbiology