Subspecies Discrimination

Transcription

1 JOURNAL OF CLINICAL MICROBIOLOGY, OCt. 1992, p /92/ $02.00/0 Vol. 30, No. 10 Colony Morphotype on Sabouraud-Triphenyltetrazolium Agar: a Simple and Inexpensive Method for Candida Subspecies Discrimination GUILLERMO QUINDOS,* MANUEL FERNANDEZ-RODRIGUEZ, ANGEL BURGOS, MATILDE TELLAETXE, RAMON CISTERNA, AND JOSE PONTON Departamento de Microbiologia e Inmunologia, Facultad de Medicina y Odontologia, Universidad del Pais Vasco, Bilbao, Spain Received 27 January 1992/Accepted 27 June 1992 A new method of Candida subspecies discrimination on Sabouraud-triphenyltetrazolium agar is reported. Eive hundred sixty-two strains of Candida and Torulopsis glabrata, previously identified by conventional mycological methods, were studied. Each strain received a three-letter code and a number based on its colonial morphology. Sixteen morphotypes were found for Candida albicans, 6 were found for Candida parapsilosis, 4 were found for both Candida guilliermondii and Candida krusei, and 12 were found for Candida tropicalis. None of the 56 T. glabrata strains studied grew on this agar. A reproducibility of 95% was found for C. albicans. The simplicity and low cost could make this method useful for typing Candida spp. Candida albicans and other Candida spp. are opportunistic pathogens involved in an increasing number of infections of patients with immunodeficiency. Different typing methods, including biotyping (11), enzyme typing (15), morphotyping (13), serotyping (1), resistotyping (8), killer typing (14), protein typing (2, 7), and karyotyping (10), have been described previously, and attempts have been made to detect a possible correlation of some C. albicans types with the virulence of the isolates (5, 9). However, it is still not possible to type C. albicans with a degree of certainty, and the lack of a suitable typing scheme has undoubtedly hindered studies of the epidemiology and virulence of this pathogenic yeast (3). These typing methods have usually not been applied to other Candida spp. (9), and most of them involve a relatively complicated methodology. In this article, we report the development of a new and inexpensive typing method and its application to evaluation of the morphotypes present in 562 strains of Candida spp. from different clinical specimens. Data were obtained from 562 unrelated strains of C. albicans, Torpulopsis glabrata, and other medically important Candida spp., previously identified by conventional mycological methods, from a variety of clinical samples. Most of them were from skin and mucosae, with 384 isolates from oral, vaginal, and respiratory samples and 178 from blood, peritoneal fluid, or organ biopsies. C. albicans and 3153 and their mutants IR2h (proteinase deficient) and CA2 (germ tube deficient), respectively, sent by A. Cassone (Istituto Superiore di Sanita, Rome, Italy), were also evaluated. Fungal identification was based on morphology on Sabouraud dextrose agar (Difco, Detroit, Mich.) and corn meal extract-tween 80 agar (Oxoid, Basingstoke, United Kingdom), germ tube formation, as well as biochemical characterization with the use of the API 20C Aux (API System S.A., Montalieu Vercieux, France) and the 16-disc carbon auxanogram (Diagnostics Pasteur, Paris, France) systems. The Sabouraud-triphenyltetrazolium (STTZ) agar, con- * Corresponding author taining 2,3,5-triphenyltetrazolium chloride (Sigma Chemical Co., St. Louis, Mo.) (0.1 g), peptone (10 g), glucose (20 g), agar (20 g), and distilled water (1 liter), was distributed in 40- to 50-ml quantities into 15-cm petri dishes. Plates were either inoculated immediately or stored at 4 C for no longer than 8 weeks. Each isolate was maintained by monthly transfers in Sabouraud dextrose agar slants and storage at 4 C. Before preparation of inocula, each isolate was grown on a fresh Sabouraud dextrose agar plate at 37 C for 24 h. Three or four colonies were resuspended in 1 ml of distilled water and adjusted to a MacFarland standard of 5 to prepare each inoculum. Before inoculation, all plates were incubated inverted at 37 C for 1 h to dry agar surfaces. Plates were inoculated in triplicate with 10,ul from the cell suspension by using a micropipette and allowing the inoculum drop to dry on the agar surface. Cultures were incubated at 37 C for 6 days. After incubation, a three-letter code and a number were given to each strain, according to its colonial morphology (Fig. 1): texture, S = smooth and R = rough; color, P = pink (P1 = pale pink = Pantone 176C, P2 = dark pink = Pantone 1785C), V = violet (Vi = pale violet = Pantone 251C, V2 = dark violet = Pantone 262C) (Pantone Inc.), 0 = orange, W = White, etc.; and presence of mycelial halo (hyphae or pseudohyphae), Y = yes and N = no. The code NG was used if no growth was observed. Pantone is a registered trademark for color specification that is oriented to facilitation and color standardization of publishing and graphic design work. We were not able to make a direct comparison between the morphotype method with STTZ agar and reference methods, which are not well established for Candida typing; therefore, we attempted to subdivide the strains within C. albicans species on the basis of their likelihood of diversity. We compared strains from various geographical and epidemiological situations with those from infected and uninfected persons. With this purpose, 277 additional C. albicans strains from Barcelona (J. M. Torres Rodriguez, Instituto Municipal de Investigaciones Medicas-Hospital del Mar, Barcelona, Spain), London (G. C. White, Candida Unit, Division of Hospital Infection, Central Public Health Laboratory, Colindale, London, United Kingdom), and Madrid

2 VOL. 30, 1992 NOTES 2749 FIG. 1. Candida morphotypes on S1TZ agar. (A) Inoculated 15-cm petri plate with different morphotypes; (B) RP2N (rough, dark pink, without mycelial halo) morphotype; (C) SPlN (smooth, pale pink, without mycelial halo) morphotype; (D) SPlY (smooth, pale pink, with mycelial halo) morphotype; (E) SV2N (smooth, dark violet, without mycelial halo) morphotype. (M. C. Cabronero, Departamento de Microbiologia, Facultad de Medicina, Universidad Complutense, Madrid, Spain) were studied. Twenty-three isolates from oral, rectal, and skin samples from eight children at 15 days and 3 and 6 months of age (three patients with oral thrush and five healthy children with colonization) were also studied (I. Contreras, Unidad Pediatrica, Centro de Salud, Algorta- Getxo, Spain). The intralaboratory reproducibility of the test was studied with 78 strains. Each strain was tested in triplicate twice, first at the beginning of the experiment and then after the strain was subcultured on Sabouraud dextrose agar for 60 days. This second culture was read blindly with respect to the first test results. The test reproducibility is the percentage of times that the code of a strain remained the same on repeated testing. This was calculated from the binary standard deviation. The variance of a binary distribution is given by the equation s2 = p(l - p)in, where p is the proportion positive and n is the total number (3). The discriminatory index (D) used here is derived from that of Hunter and Gaston (6). The discriminatory index is the probability that two randomly chosen strains, sampled consecutively, would be distinguished by the test, and it is based on Simpson's diversity index. Sixteen morphotypes were found among the 394 C. albicans strains, the predominant (88.33%) morphotypes being SPiN (162 strains), SPlY (68 strains), SP2Y (60 strains), and SP2N (58 strains) (Table 1). Strain and its mutant IR2h (proteinase deficient) showed morphotypes SPiN and RP1N, respectively. Strain 3153 and its mutant CA2 (germ tube deficient) had the same morphotype (SP1N). None of the 56 T. glabrata strains studied grew on STTZ agar. The absence of growth was a characteristic feature for this species. However, this method has no value for identification of Candida species because the same morphotypes are observed with different species of the genus. The 13 C. guilliermondii strains studied were also divided into different morphotypes. As shown in Table 1, the distribution was more homogeneous than in other species, the strains belonging to four morphotypes: SV1N (38.46%), SPiN (30.78%), SV2N (15.38%), and SPlY (15.38%). Candida krusei showed the least homogeneous distribution, since 14 of 20 strains belonged to biotype SPlY (70%). Although rare, morphotypes SOY and SON were unique for this species. Eight morphotypes were observed for 34 Candida parapsilosis strains, 85.29% of the strains being grouped into morphotypes SV2N, SV1N, RV1N, and SP1N. Twelve morphotypes were found among 45 C. tropicalis strains, with morphotype SV2Y found for 45% of the strains. The distribution of the morphotypes according to the strains' geo-

3 2750 NOTES TABLE 1. Morphotypes of 562 strains of Candida on STTZ agar Species Morphotype' No. of (no. of isolates) Mrhteastrains (%) C. albicans (394) SPlN 162 (41.12) SPlY 68 (17.26) SP2Y 60 (15.23) SP2N 58 (14.72) NG 11 (2.79) RP1N 6 (1.52) RP2Y 6 (1.52) RP1Y 6 (1.52) SV1Y 4 (1.03) RP2N 3 (0.76) SV2Y 3 (0.76) RV2Y 2 (0.51) RV2N 2 (0.51) RV1Y 1 (0.25) SV1N 1 (0.25) SV2N 1 (0.25) T. glabrata (56) NG 56 (100) C. guilliernondii (13) SV1N 5 (38.46) SPlN 4 (30.78) SV2N 2 (15.38) SPlY 2 (15.38) C. krusei (20) SPlY 14 (70) SPlN 4 (20) SOY 1 (5) SON 1 (5) C. parapsilosis (34) SV2N 11 (32.35) SV1N 8 (23.53) RV1N 5 (14.71) SPlN 5 (14.71) RP1N 2 (5.88) RV2N 1 (2.94) SPlY 1 (2.94) NG 1 (2.94) C. tropicalis (45) SV2Y 18 (45) SV1Y 5 (12.5) RV2Y 5 (12.5) SV2N 4 (10) NG 4 (10) SP2Y 3 (7.5) SPlN 1 (2.5) SPlY 1 (2.5) RP1N 1 (2.5) RP2Y 1 (2.5) RV1Y 1 (2.5) RV1N 1 (2.5) graphical and clinical origins is shown in Tables 2 and 3. Although differences in morphotypes were observed, SPiN was predominant among strains from the four geographical locations and from most of the clinical sites. Morphotype SV2N, which was very infrequent among our C. albicans strains, was the second most frequent among strains from Madrid. Morphotype SPiN was predominant in the pediatric isolates (data not shown). However, in all patients, the same morphotypes were present among isolates from different locations and at different times of sampling if the cultures were positive. Patients with oral thrush had strains with the SPiN morphotype isolated from the mouth, anus, and skin TABLE 2. Morphotypea J. CLIN. MICROBIOL. Morphotypes of C. albicans isolates from different geographical origins No. of strains (%) from source or geographical origin Our study Barcelona London Madrid SPlN 162 (41.12) 29 (52.73) 38 (33.63) 61 (55.96) SPlY 68 (17.26) 7 (12.73) 12 (10.62) 1 (0.92) SP2Y 60 (15.23) 2 (3.64) 30 (26.55) 0 (0) SP2N 58 (14.72) 4 (7.27) 24 (21.24) 0 (0) NG 11 (2.79) 5 (9.09) 2 (1.77) 0 (0) RP1N 6 (1.52) 1 (1.82) 2 (1.77) 12 (11.01) RP2Y 6 (1.52) 1 (1.82) 2 (1.77) 0 (0) RP1Y 6 (1.52) 1 (1.82) 0 (0) 0 (0) SV1Y 4 (1.03) 1 (1.82) 0 (0) 3 (2.75) RP2N 3 (0.76) 0 (0) 1 (0.88) 0 (0) SV2Y 3 (0.76) 0 (0) 0 (0) 0 (0) RV2Y 2 (0.51) 3 (5.44) 2 (1.77) 0 (0) RV2N 2 (0.51) 0 (0) 0 (0) 0 (0) RV1Y 1 (0.25) 0 (0) 0 (0) 0 (0) SV1N 1 (0.25) 1 (1.82) 0 (0) 28 (25.69) SV2N 1 (0.25) 0 (0) 0 (0) 0 (0) RV1N 0 (0) 0 (0) 0 (0) 4 (3.67) of the napkin area. In some oral and rectal samples, two different morphotypes were present, but this was not observed with skin samples. Ten SPlN-morphotype C. albicans isolates from eight patients with denture stomatitis (two strains from one patient, another two strains from another patient, and six from six different persons) were evaluated by Susan A. Howell TABLE 3. Relationship of morphotypes of C. albicans isolates to their clinical and geographical origins Origin of clinical Morphotypea (no.) of isolates from: sample Madrid Barcelona London Blood SPlN (4) SPlN (2) SPlN (12) SV1N (2) SP2N (1) SP2Y (10) RP1N (1) SP2N (9) SV1Y (1) RV2Y (2) SPlY (2) Other (4) Mouth and SPlN (7) SPlN (12) SPlN (18) digestive tract SV1N (6) SPlY (4) SP2Y (14) RP1N (4) NG (4) SPlY (9) Other (4) SP2N (7) Other (3) Respiratory tract SPlN (7) SV2Y (3) SPlN (4) RV1N (3) SPlY (2) SP2N (3) SV1N (2) Other (3) SP2Y (3) SV1Y (2) SPlY (1) RP1N (1) Nail and skin SPlN (13) SPlN (4) SP2N (3) SV1N (7) RP1N (1) SPlN (2) RP1N (2) Other (2) SP2Y (1) Urogenital tract SPiN (30) SPlN (11) SPlN (2) SV1N (10) Other (4) SP2N (2) RP1N (4) SP2Y (2) SPlY (1)

4 VOL. 30, 1992 from the Department of Microbial Diseases, Lambeth Hospital (London, United Kingdom) by restriction fragment length polymorphism analysis. DNA was digested with restriction endonucleases EcoRI, Hinfl and BglII (12). Eight different karyotypes were found, and only isolates from the same patient had the same karyotype (data not shown). Morphotype SPlN appeared to be that for a subspecies composed by genetically different strains. The typeabilities for the different species tested, if absence of growth was not considered a morphotype, were 91.11% for C. tropicalis, 97.21% for C. albicans and C. parapsilosis, and 100% for C. guilliennondii and C. krusei. The reproducibilities of the test after the strains were subcultured on STTZ agar for 2 months were 85% for 6 C. guilliennondii strains, 92% for 12 C. tropicalis strains, 95% for 49 C. albicans strains, and 100% for 4 C. krusei strains and 7 C. parapsilosis strains. According to Simpson's index of diversity, biotyping on STTZ agar had a D = for C. albicans. The method showed discriminatory indices for Candida spp. different from that of C. albicans; for C. guilliermnondii, for C. krusei, for C. parapsilosis, and for C. tropicalis. Sixty strains of six species of Candida were cultured in different conditions on S1TZ agar for the selection of the appropriate conditions of test performance. A 10-,ul inoculum from a cellular suspension adjusted to a MacFarland standard of 5 was chosen since lesser concentration suspensions gave unsatisfactory results. Incubation of cultures for 6 days at 37 C was selected versus incubation at 37 C for 1, 2, 4, 5, 8, and 10 days or the same periods of incubations at 37 C plus 1 additional day at 25 C. The reason for this choice was the consolidation of the colonial features used for the code (texture, color, or presence of mycelial halo) on day 6 of incubation in 98% of the strains. No improvement of these results was observed with other incubation conditions. When assessing the value of any typing method, the three main characteristics that need to be considered are typeability, reproducibility, and discriminatory power. The cost, ease of use, and rapidity should also be considered. The typeability of a method is the proportion of a population of isolates that can be typed by that method. All existing typing methods for C. albicans have 100% typeability (2). Our method gave 97.21% typeability, since 11 of 394 C. albicans strains typed did not grow in the medium. These strains may be considered to be a different morphotype (NG). The method described here was not able to discriminate different morphotypes of T. glabrata, since none of the strains tested grew in STTZ agar. Reproducibility of the method for expression and stability of strain codes ranged from 85 to 100%, depending on the species tested, and was 95% for C. albicans. Overheating of the medium during its preparation and marked changes in the incubation temperature resulted in differences among morphotypes. As the evaluation of a morphotype is subjective and different appreciations of color intensity are unavoidable, we recommend the use of Pantone color standards. The same results are thus obtained with different readers. At the same time, variation in colony morphology of some strains during subculturing could be explained by high-frequency switching (16). Estimates of reproducibility of different physiochemical typing methods for C. albicans have shown variations better and worse than those described here for the typing on STTZ agar. McCreight et al. (8) found a 96% in vitro reproducibility for resistotyping tests. However, Hunter and Fraser (4), using a modification of that system, found reproducibilities ranging from 77 to 100%, depending NOTES 2751 on the number of test differences required to distinguish among strains. In contrast, a reproducibility lower than 95% has been found for morphotyping on malt extract agar and different biotyping systems (3, 13). Other typing methods, based on assimilation of carbon sources, detection of extracellular enzymes, or biochemical reactions, have also shown wide variations in their reproducibilities (11, 15, 17, 18). The discrimination of a method is an estimate of its ability to differentiate between two unrelated strains. According to Simpson's index of diversity, the typing on SITZ had ad = for C. albicans. This means that if two strains of C. albicans are taken randomly from the population, then 75.7% of the time they would fall into different types. The method showed higher discriminatory indices for C. guilliermondii, C. parapsilosis, and C. tropicalis. The lowest discriminatory index was found for C. krusei, since 70% of the isolates fell into the same biotype. The discriminatory ability of C. albicans typing on STTZ agar seems to be in an intermediate position among previously published methods. It was lower than resistotyping (D = 0.899) and DNA typing (D = 0.868) but higher than the killer system (D = 0.724), immunoblotting (D = 0.679), enzyme biovars (D = 0.549), monoclonal antibodies (D = 0.492), and serotyping (D = 0.438) (3, 6). Compared with other systems for strain differentiation, the S1TZ agar morphotyping system has the advantages of simplicity, low cost (morphotype determination for nine strains costs approximately the same as a Sabouraud dextrose agar plate), and the potential for typing C. albicans and other medically important Candida species. The main disadvantage is the impossibility of typing T. glabrata strains. We thank D. W. R. Mackenzie and Tyrone Pitt for critical revisions of the manuscript; A. Cassone for strains 10261, 3153, CA2, and IR2h; Susan Howell for work on karyotyping; J. M. Torres Rodriquez, G. C. White, and M. C. Cabronero for strains; and E. Gonzalez-Miranda for estimable help with the graphic work. This work was supported by grant UPV /89 from the Universidad del Pais Vasco-Euskal Herriko Unibertsitatea and grant SGV SC 04/90 from the Departamento de Educaci6n, Universidades e Investigaci6n del Gobierno Vasco. REFERENCES 1. Hasenclever, H. F., and W. 0. Mitchell Antigenic studies of Candida. IV. The relationship of the antigenic groups of Candida albicans to their isolation from various clinical specimens. Sabouraudia 2: Howell, S. A., and W. C. Noble Typing tools for the investigation of epidemic fungal infection. Epidemiol. Infect. 105: Hunter, P. R A critical review of typing methods for Candida albicans and their applications. Crit. Rev. Microbiol. 17: Hunter, P. R., and C. A. M. Fraser Use of modified resistogram to type Candida albicans isolated from cases of vaginitis and from faeces in the same geographical area. J. Clin. Pathol. 40: Hunter, P. R., C. A. M. Fraser, and D. W. R. Mackenzie Morphotype markers of virulence in human candidal infections. J. Med. Microbiol. 28: Hunter, P. R., and M. A. Gaston Numerical index of the discriminatory ability of typing systems: an application of Simpson's index of diversity. J. Clin. Microbiol. 26: Lehmann, P. F., B. J. Kemker, C. B. Hsiao, and S. Dev Isoenzyme biotypes of Candida species. J. Clin. Microbiol. 27: McCreight, M. C., D. W. Warnock, and M. V. Martin Resistogram typing of Candida albicans isolates from oral and cutaneous sites in irradiated patients. Sabouraudia 23: Merz, W. G Candida albicans strain delineation. Clin.

5 2752 NOTES Microbiol. Rev. 3: Merz, W. G., C. Connelly, and P. Hieter Variation of electrophoretic karyotypes among clinical isolates of Candida albicans. J. Clin. Microbiol. 26: Odds, F. C., and A. B. Abbott A simple system for the presumptive identification of Candida albicans and differentiation of the strains within the species. Sabouraudia 18: Pearce, M., and S. A. Howell Restriction fragment length polymorphism analysis of azole-resistant and azole-susceptible Candida albicans strains. J. Clin. Microbiol. 29: Phongpaichit, S., D. W. R. Mackenzie, and C. Fraser Strain differentiation of Candida albicans by morphotyping. Epidemiol. Infect. 99: Polonelli, L., C. Archibusacci, M. Sestito, and G. Morace J. CLIN. MICROBIOL. Killer system: a simple method for differentiating Candida albicans strains. J. Clin. Microbiol. 17: Roman, M. C., and M. J. L. Sicilia Preliminary investigation of Candida albicans biovars. J. Clin. Microbiol. 18: Slutsky, B., J. Buffo, and D. R. Soll High-frequency switching of colony morphology in Candida albicans. Science 230: Williamson, M. I., L. P. Samaranayake, and T. W. MacFarlane Biotypes of oral Candida albicans and Candida tropicalis isolates. J. Med. Vet. Mycol. 24: Williamson, M. I., L. P. Samaranayake, and T. W. MacFarlane A new simple method for biotyping Candida albicans. Microbios 51: