Agreement between RAPD, API20C AUX, CHROMagar Candida and microculture on oral Candida identification

Aim: To measure the agreement of methods for identification of Candida species in oral cavity samples, comparing the CHROMagar Candida , microculture, API 20C AUX and RAPD techniques. Methods : Ninety-one colonies of Candida were isolated and presumptively identified in CHROMagar Candida, submitted to microculture, API 20C AUX and RAPD techniques. After this, agreement among methods using Kappa test was performed. Results: Agreement rates between RAPD and CHROMagar Candida , showed significant accuracy for C. albicans , C. tropicalis , C. dubliniensis and C. krusei (Kappa: 0.760, 0.640, 0.416 and 0.360, respectively, p<0.05). Comparing RAPD results with microculture, the highest agreement was for C. albicans (Kappa: 0.851 - p<0.05) but no significant agreement for C. lusitaniae , C. krusei and C. guilliermondii was obtained (p>0.05). The agreement was significant for all identified species when RAPD (OPE-18) and API 20C AUX (p<0.05) were used. Critical levels of agreement between RAPD and microculture were observed when C. lusitaniae , C. krusei and C. guilliermondii were identified. Conclusions : API 20C AUX presented the best agreement with molecular random identification and CHROMagar showed good agreement for C. albicans , C. tropicalis , C. dubliniensis and C. krusei identification.


Introduction
Candida species are commensal microorganisms of the oral cavity.They have several virulence factors, which in the presence of local and systemic host failures may result in their transition from commensal to pathogenic organisms 1 , causing oral and systemic infections that pose significant public health problems.Their isolation is used in investigations related to salivary disfunction, oral candidiasis, orofacial pathologies, and immune suppressant status [2][3][4] .
There is a variety of methods for identifying Candida species from clinical samples in the oral cavity 5 .The CHROMagar Candida differential medium is commonly used to isolate and identify presumptive C. albicans, C. dubliniensis, C. tropicalis and C. krusei.Their sensitivity and specificity are considered satisfactory for these species 2,[6][7] .
The microculture analysis has considerable accuracy and presents low cost 2 , 150 150 150 150 150 but it requires visual experience, sometimes limited by the resolution of optical microscopy and confused by similarities among species' expressions.The biochemical characterization could be performed using the API® 20C AUX (BioMerieux, France) which relies on variations in the assimilation of carbohydrates [7][8][9] .However, it presents limitations related to cost and to distinguish between some species 2 .A study of 159 clinical isolates of Candida species identified by the very similar kit API® Candida AUX (BioMerieux, Marcy l'Etoile, France) reported that 12 isolates (7.5%) were incorrectly identified [10][11] .
In recent decades, traditional methods of microorganism phenotyping have been replaced or added by the procedures associated to recombinant DNA [12][13][14] .Methods based on molecular markers are useful not only for phenotyping, but also for differentiation of Candida species [15][16] .The RAPD (Random Amplification Polymorphic DNA) allows the amplification of DNA sequences and is a simple and quick technique that does not require prior knowledge on the genomes to characterize organisms, using one randomly determined (usually a decamer) primer 17 .It is used for genetic characterization of a range of organisms, plants, animals or microorganisms, including Candida species, for different purposes [18][19][20][21] .The sensitivity, specificity and resolution of the OPE-18 primer for identification of Candida species has been reported and could be used for epidemiological Candida identification 11,22 .
Due to scarce information about presumptive, biochemical and molecular agreement on Candida identification, this study aimed to measure the assertive correlation between the presumptive identification of Candida species from oral cavity using CHROMagar Candida, microculture, API® 20C Aux and OPE-18 genotyping.

Ethical procedures
Ethical considerations in accordance with Helsinki Declaration have been observed.This research was conducted according to the ethical principles of research involving human participants, as stipulated by Resolution 196/96 of the National Health Council of the Ministry of Health of Brazil.The collection and analysis of data in this study were certified by the research ethics committee of the State University of Montes Claros, MG, Brazil, protocol CEP nº.1111/08.

Origin of samples
The Candida isolates resulted from salivary collections of oral cavity of patients irradiated on head and neck due to malignant neoplasms (n=29) and elderly volunteers (n=63).The collection comprised 91 isolates of Candida species.

Isolation and presumptive identification of Candida species
The isolation of yeasts was made in salivary samples collected from the buccal mucosa and tongue with a swab and sterile saline solution (NaCl, 0.85%) as diluent.The isolation and presumptive identification was made by drawing aliquots (100 µL) from each sample and placing them on plates containing CHROMagar Candida and incubated at 37 °C for 24 to 48 h, in duplicate.Yeast identification was made by considering the morphology and color of the colonies 2,23 .Each colony of Candida was cataloged and then stored at -20 °C in Sabouraud Dextrose Broth (DSB, Oxoid Ltd., London, England) amended with glycerol (40% v/v).ATCC 10231 of C. albicans was used as quality control (QC).

Identification by API 20C AUX
The inoculum used to this procedure was obtained from cultured yeast on Sabouraud Agar.The procedures for inoculation and interpretation were performed according to the instructions provided by the manufacturer (BioMerieux, France).Identification list on these indexes was considered as excellent (%ID>99.9,T>0.75), very good (>99.0%ID and T>0.5) or acceptable (%ID>90.0 and T >0.5) 7 .

Identification of isolates by RAPD (Random Amplification Polymorphic DNA)
The extraction and purification of DNA from isolates of Candida spp was made with the Purelink Genomic DNA® kit (Invitrogen K1820-02, Brazil).The used DNA was obtained from cells grown in YPD broth (1% Malt Extract Powder, 2% bacteriological peptone and 2% dextrose -D-glucose) at 37 °C and shaking (150 rpm for 24 h) 11 .A total of 50 µL of concentrated suspension of each isolate was obtained by centrifugation (3,500 rpm for 30 min).The purification of DNA was made by adding 200 µL of digestion buffer, 20 µL proteinase K and 20 µL RNase.We added to 200 µL of binding buffer and then the tubes were heated for 10 min at 80 °C in a water bath.To neutralize the detergent and to allow the connection with the silica column, 200 µL of absolute ethanol was added (Merck, Darmstadt, Germany).The tubes were centrifuged at 13,000 rpm for 1 min and the pellet was discarded.Subsequently were added 500 µL of the first washing buffer and centrifuged again at 13,000 rpm for 1 min and the precipitate discarded.The column with the silica was passed to the second tube and added 500 µL of the second washing buffer and centrifuged to 13,000 rpm for 1 min and for 3 additional minutes to enhance drying.
For the first extraction, 200 µL of sterile water were added to Milli-Q heated to 60 °C in the column in a second High Table 1.Table 1.Table 1.Table 1.Table 1.Comparation of methods used in this study.tube.Then it was centrifuged at 13,000 rpm for 1 min.For the second extraction, 200 µL of elution buffer of the same column were placed in a third tube and centrifuged it at 13,000 rpm for 1 min.
PCR products were separated by agarose gel eletroforesis (1.4% / v -5µL ethidium bromide -10 mg/mL), 80 V for 5 h.We used ATCC 10231 as QC and two molecular weights were incorporated (100 bp and 250 bp -Invitrogen, São Paulo, SP, Brazil).The DNA bands were observed and photographed in transillumination and the images were analyzed considering the literature reports 11,22 .
Table 1 shows comparison of the methods used in this study.
In the RAPD technique only one isolate (1.1%) presented an undefined pattern, followed by five (5.5%) in API 20C AUX® and 8 (8.8%) in microculture.Considering the comparative analysis between the RAPD characterization and microculture evaluation, we can observe that the identifications of C. albicans, C. dubliniensis, C. tropicalis, C. glabrata, C. kefyr and C. parapsilosis were significantly concordant in decreasing levels among the methods, in that order.When RAPD and API 20C AUX® were compared, the species C. tropicalis, C. albicans, C. glabrata, C. kefyr, C. dubliniensis, C. lusitaniae, C. krusei and C. guilliermondii showed significant decreasing agreement, in that order.Table 3 shows the Kappa coefficient among RAPD, API 20C AUX and microculture.

Discussion
The presumptive identification of yeasts may be crucial in the diagnosis and treatment of fungal infections.It is a fundamental recognition and validation of methods that should be fast, accurate and inexpensive.Bernal et al. 25 (1996) using the CHROMagar Candida for presumed identification of 593 colonies, revealed 341 (57.5%) C. albicans, 339 (57.2%) of them featuring green characteristic color.All 35 (5.9%) C. krusei and 73 (12.3%) of C. tropicalis presented specific characteristics identified on CHROMagar Candida.In the present study, among the 91 isolates, 35 (38.46%) were pale green, 18 (30.7%)pale pink with white halo and 5 (5.5%) were blue on CHROMagar Candida.Using RAPD, 29 (31.8%) were identified as C. albicans, 8(8.8%) as C. krusei and 9 (9.9%) as C. tropicalis (Kappa coefficient 0.760, 0.360 and 0.640 respectively -p<0.05), showing a good accuracy of CHROMagar Candida identification of these species.
* Kappa coefficient probability.and