Production of White Colonies on CHROMagar Candida Medium by Members of the Candida glabrata Clade and Other Species with Overlapping Phenotypic Traits

Abstract

We hypothesized that species of the Candida glabrata clade and species with phenotypic traits that overlap those of C. glabrata would produce white colonies on CHROMagar Candida medium. Of 154 isolates (seven species) tested, C. bracarensis, C. nivariensis, C. norvegensis, C. glabrata, and C. inconspicua produced white colonies; the Pichia fermentans group and C. krusei did not. Many of these species are difficult to identify phenotypically; white colonies may signal the need for the use of molecular approaches.

Chromogenic media have been formulated to discern common Candida species on the basis of the color of the colonies (7). Beginning in Europe in the 1990s, investigators published reports of the development of several chromogenic media; many of the media are now commercially available products in the United States. These publications include several on the performance of CHROMagar Candida medium (CAC; Becton Dickinson, Sparks, MD, and Hardy Diagnostics, Santa Maria, CA). These data support the conclusions that there are relatively specific colony colors for Candida albicans, Candida tropicalis, and Candida krusei (2, 8, 11, 12). In addition, Candida glabrata and Candida dubliniensis can often be distinguished by their specific colony colors (5, 9). A major advantage of these media is that they provide identifications (presumptive or definitive) in times shorter than those required for identification by many standard mycologic phenotypic assays (11). In addition, these media permit the detection of the growth of multiple yeast species from a single clinical specimen due to the colony color differences (8, 11). However, the lack of a wide range of colors limits the expansion to more yeast genera or species, and the increased cost (higher than that for standard mycologic media) has prevented some laboratories from implementing these media.

Since its original descriptions, there have been reports of yeast species that produce white colonies on CAC (8). However, white colony-producing yeasts are not frequently encountered among yeasts recovered from clinical specimens (7, 8, 12). We recently observed white colonies while investigating Candida bracarensis and Candida nivariensis (3), two species within the C. glabrata clade recently defined molecularly (1, 4). These species are difficult to separate by the use of phenotyping due to their significantly overlapping traits and also due to the relatively few positive reactions for these species (e.g., by a carbohydrate assimilation assay). From these observations, we hypothesized that white colonies would be produced on CAC by other strains or members of the C. glabrata clade and also by Candida spp. that are not in the clade but that have few positive reactions and that share many phenotypic traits.

A total of 154 isolates was tested in this study, including 13 C. krusei and 127 C. glabrata recent clinical isolates identified by phenotypic characteristics at the Clinical Mycology Laboratory of The Johns Hopkins Hospital. Phenotypic identification of C. glabrata and C. krusei was made by the germ tube assay; the urease test; determination of microscopic morphology on cornmeal agar with Tween 80; the phenoloxidase enzyme assay; and fermentation patterns on glucose, maltose, sucrose, lactose, galactose, trehalose, and cellobiose. Fourteen reference type strains and clinical laboratory reference strains of seven other Candida species were also tested (Table 1). Each isolate was coded to blind the reader and streaked onto the surfaces of three CAC plates; one set of plates was incubated at 25°C, one at 30°C, and one at 37°C. The plates were read independently by at least two readers after 24 and 48 h of incubation. An isolate was designated positive if it produced white colonies on CAC at two or three of the temperatures. All isolates that produced white colonies were reidentified by molecular assays, D1 and D2 sequencing (6), or peptide nucleic acid fluorescence in situ hybridization (3) and phenotypically by using the carbohydrate patterns obtained with the API 20C system (BioMerieux, Hazelwood, MO).

TABLE 1.

Type strains and reference strains of Candida glabrata and related species

Yeast species	Class culturea	Strain designation
C. glabrata	JHH lab control	CGL-C
C. krusei	IVST control	ATCC 6258
C. bracarensis	Type	NRRL Y-48270
C. bracarensis	Reference	NRRL Y-27794
C. nivariensis	Type	NRRL Y-48269
C. (Pichia) norvegensis	Reference	NRRL YB-3899
C. (Pichia) norvegensis	Reference	NRRL Y-27195
C. (Pichia) norvegensis	Reference	NRRL Y-27202
C. (Pichia) norvegensis	Reference	NRRL Y-12804
C. (Pichia) norvegensis	Type	NRRL YB-2029
C. (Pichia) norvegensis	Reference	NRRL YB-3924
C. monosa (Pichia fermentans)	Type	NRRL Y-7180
C. lambica (Pichia fermentans)	Type	NRRL Y-7181
Pichia fermentans	Type	NRRL Y-1619

^aJHH lab, The Johns Hopkins Hospital Clinical Mycology Laboratory; IVST, in vitro susceptibility testing.

Numerous white colonies were noted after 24 h of incubation, but many of these colonies developed into shades of pink and occasionally purple after an additional 24 h of incubation. Therefore, all subsequent results are from data collected after 48 h of incubation. The color distinction was the most pronounced at 37°C and was the least pronounced at 25°C. Similarly, although white colonies were noted at all three temperatures, more white colonies were observed at 37°C than at 25°C and 30°C. In addition, all “positives” (i.e., isolates that grew as white colonies at two of the three temperatures) were white at both 30°C and 37°C.

White colonies were produced by a total of 18 isolates representing five of the seven species tested (Table 2). The five species were C. glabrata, C. bracarensis, C. nivariensis, Candida norvegensis, and Candida inconspicua. The first three species are members of the C. glabrata clade, whereas C. inconspicua and C. norvegensis share phenotypic traits with C. glabrata but, along with C. krusei and C. inconspicua, are members of the Pichia membranifaciens/P. fermentans clade (6). The two species studied that did not produce any white colonies were C. krusei and the P. fermentans group, two species that also share phenotypic traits with members of the C. glabrata clade. These data support our hypothesis that members of the C. glabrata clade as well as distantly related species with overlapping phenotypic traits would produce white colonies on CAC.

TABLE 2.

Production of white colonies on CAC

Type of isolate and yeast speciesa	No. of isolates positive/total no. tested	Species reidentifiedb	CHO assimilatedc
Recent clinical C. glabrata isolates	13/127	C. glabrata (8/8)	− − − +
		C. norvegensis (1/0)	−/+ − − −
		C. bracarensis (2/0)	+ − − +
		C. bracarensis (1/0)	− − − +
		C. bracarensis (1/0)	+ − − −
Reference or type strains
C. bracarensis	2/2	C. bracarensis (1/0)	+ − − −
		C. bracarensis (1/0)	+ − − +
C. nivariensis	1/1	C. nivariensis (1/0)	+ − − +
C. inconspicua	2/2	C. inconspicua (2/0)	−/+ − − −

^aThe species that did not produce white colonies were C. glabrata (1 control isolate and 119 recent clinical isolates), C. krusei (1 control isolate and 13 recent clinical isolates), P. fermentans (3 reference strains), and C. norvegensis (4 reference strains).

^bThe data in parentheses indicate the number of isolates identified by molecular methods/number identified by phenotypic methods.

^cThe assimilation results are for the five carbohydrates (CHO) glycerol, xylose, N-acetylglucosamine, and trehalose, respectively.

Confirmations of the identifications of the 18 isolates by an additional phenotypic assay (carbohydrate assimilation patterns with the API 20C system) and by sequencing of the large subunits D1 and D2 yielded interesting results (Table 2). The sequencing data identified all 18 isolates to the species level, and the identifications of 13 isolates were reconfirmed. Five isolates whose identifications were not confirmed by sequencing were among the eight recent clinical isolates of C. glabrata. Molecularly, four were C. bracarensis and one was C. norvegensis. The use of standard mycologic phenotyping, including the assimilation patterns, correctly identified only 4 of the 18 white colony-producing isolates; these 4 isolates were the same 4 C. glabrata clinical isolates whose identifications were also reconfirmed by sequencing. Unfortunately, the assimilation patterns were not helpful for the identification of the other isolates since C. bracarensis and C. nivariensis are not in the API database and since this group of species assimilates only a few carbohydrates (0 to 2 carbohydrates, not including glucose), yielding multiple species with the same biocode and low probabilities of correct identifications. This low level of reactivity not only limits the discriminatory patterns for identification but also may contribute to the lack of color production by these species.

The 13 isolates that produced white colonies were recovered from specimens representing five anatomic sites (Table 3). Ten isolates were recovered from three nondeep sites (urine, stool, and throat) and included all three species. For seven of these isolates, a review of the patients' charts did not reveal evidence of infection; therefore, the positive cultures likely represented colonizations. For two of the C. glabrata isolates recovered from urine, there was insufficient clinical information to determine whether the positive cultures represented infection or colonization. The remaining four isolates (two C. glabrata isolates, one C. bracarensis isolate, and one C. norvegenesis isolate) were recovered from patients with clinical evidence of infections caused by these organisms. The C. bracarensis isolate was recovered operatively from an intra-abdominal abscess in an immunocompetent 67-year-old woman with a history of perforated diverticulitis (3). C. norvegenesis was recovered via straight catheterization from the urine of a 43-year-old woman with a history of end-stage renal disease and kidney transplantation who had recurrent urinary tract infections caused by a yeast. One isolate of C. glabrata was recovered in a sterile fashion from a peripancreatic fluid collection recovered from a 67-year-old man with end-stage renal disease and chronic pancreatitis with infected pseudocysts. Finally, another isolate of C. glabrata was recovered from the blood of a 49-year-old man undergoing chemotherapy for relapsed chronic myelogenous leukemia.

TABLE 3.

Clinical significance of 11 of the 13 isolates of Candida species that produced white colonies on CAC

Yeast species (no. of isolates)	No. of isolates
	Colonizations			Infections
	Urine	Stool	Throat	Urine	Blood	Abscess
C. glabrata (6)a	2	1	1	0	1	1
C. bracarensis (4)	0	2	1	0	0	1
C. norvegenesis (1)	0	0	0	1	0	0
Total (11)	2	3	2	1	1	2

^aFor two additional isolates of C. glabrata, insufficient clinical information was available to determine whether a positive urine culture represented infection or colonization.

In conclusion, we have generated data that supported our hypothesis. Five of seven selected species produced white colonies on CAC. These included C. glabrata; two other members of the C. glabrata clade, C. bracarensis and C. nivariensis; and two of four of the species with overlapping phenotypic traits, C. norvegensis and C. inconspicua. Four of the five species that produced white colonies were not identifiable by the commonly used phenotypic assays. These data may contribute to the use of chromogenic media in the clinical mycology laboratory. Isolates with white colonies might indicate that molecular assays might be required for identification.