Candida albicans ssp. dubliniensis stat.et new combination for Candida dubliniensis based on genetic criteria

One accredited species, Candida albicans subspecies dubliniensis , has been proposed to replace the existing designations of Candida dubliniensis . The study of the genetic diversity among the clinical isolates of C. albicans and C. dubliniensis was performed based on the amplified transposable intron region in the 25S rRNA gene. This study attempts to verify the unequivocal understanding of the genetic relationship between C. albicans and C. dubliniensis . Twenty (20) isolates of C. albicans and C. dubliniensis were studied using the method of typing by rDNA, restriction fragment length polymorphism (RFLP), random amplification polymorphism DNA-polymerase chain reaction (RAPD-PCR) and intron sequencing in the 25S gene. The results reveals that the specific primer pair CABF59F and CADBR125R was the successfully amplified target for all the C. albicans isolates and three isolate of C. dubliniensis . The Candida isolates revealed a genetic pattern based on the analysis of the RAPD-DNA fingerprinting pattern. The RFLPs generated by Hhal and Hae 111 enzymes elucidated similar recognition sites for both the C. albicans and C. dubliniensis isolates. Analysis of the intron sequence in the 25S gene region of the genotype C. albicans and C. dubliniensis showed identical with only a few differences in the base substitution. The sequence variations appear among the same isolates in each species. In all the cases, the clinical isolates of both species showed a percentage sequence similarity of >99.5%. This result emphasizes a high indication of similarity between C. dubliniensis and C. albicans . It was concluded that the taxonomic position of C. dubliniensis was puzzled due to insufficient genetic and phenotypic characters to warrant species status. Variations were occasionally observed to occur among the same isolates, within the same species; however, this indication is applied to other taxonomic criteria between them, with no credibility for the great differences observed between C. dubliniensis and C. albicans . This is the final taxonomic decision for C. dubliniensis to merit an amendment in order to be included as C. albicans subspecies dubliniensis stat. et comb. nov. C. dubliniensis with a revised synonymy for C. albicans .


INTRODUCTION
Over the past 19 years, several studies have been done to evaluate the relationship between Candida dubliniensis and Candida albicans, a typical Candida strain difficult to identify up to species level, because of the heterogeneous morphological, biochemical and genetic characteristics they exhibit (Sullivan et al., 1995;Pujol et al., 1997;Sullivan and Coleman, 1997).
C. dubliniensis had been described as a separate species in 1995 by Sullivan et al. (1995). Retrospective studies revealed that it had been earlier commonly identified as C. albicans, to which C. dubliniensis is closely related, and with which it shares several characteristics including those of growth conditions, germ tube formation, chlamydospore formation and color interaction on CHROMagar (Tamura et al., 2001). Systematic studies of Candida spp. based on phenotypic criteria alone have been revealed to be unreliable markers, although they enabled us to elucidate some taxonomic complications between C. albicans and C. dubliniensis, such as similarities in the phenotypic characters, especially in their color on chromogenic agar, as well as following other conventional criteria, although not the perfect solution for the differences between them (Ahmed et al., 2002;Marot-Leblond et al., 2006;Imran and Al Asadi, 2014). In fact, misidentifications between some Candida spp., particularly between C. dubliniensis and C. albicans have frequently been observed (Tamura et al., 2001;Abaci et al., 2008). Coronado-Castellote and Jiménez-Soriano (2013) referred to C. dubliniensis as exhibiting similarity with C. albicans in their germ tubes and chlamydoconidia, as well as the high probability for mating between them and the similarity in some of their sequences at different loci (Pujol et al., 2004). Most of these reasons were critical in the identification of the taxonomic position of C. dubliniensis. Ribosomal DNA, considered an essential marker in Candida and other fungi, is ideally suited for the development of molecular studies. The high discriminatory power of the m olecular tools like polymerase chain reaction (PCR) PCR, PCR-RFLP, RAPD-PCR, as well as sequencing, have provided change barter fast, relatively simple to perform, precise and reliable methods for the diagnosis of the Candida spp. (Mirhendi et al., 2005;Santos et al., 2010;Mijiti et al., 2010;Shokohi et al., 2010). McCullough et al. (1999) utilized CABF59F and CADBR125R primers designed to span the region that includes the site of the transposable intron of the 25S rRNA gene (rDNA). The molecular target, the transposable intron and the design CA-INT primer were the highly reproducible markers for typing the C. albicans subgenotypes and differentiated the C. dubliniensis from the closely related isolates of the C. albicans compartment with a selection of the ITS or other regions (Tamura et al., 2001). The simple PCR attached to the inserted intron in 25S rDNA classified the strains of C. albicans into three or four subgroups as the given genotype A, genotype B, genotype C and sometimes even genotype D (McCullough et al., 1999). A special genotype was also assigned to C. dubliniensis (Tamura et al., 2001). However, Tamura et al. (2001) showed that no group I intron was observed in the other Candida species tested, including those of Candida glabrata, Candida krusei, Candida parapsilosis and Candida tropicalis. Imran and Al. Asadi (2014) revealed the presence of introns in most non albicans species. In contrast, if reproducible markers like the restriction enzyme are utilized, they could facilitate solving the difficult cases. By analyzing the restriction fragment length polymorphisms (RFLPs), the unique polymorphism in the monomorphic PCR bands can be identified (Mirhendi et al., 2005). From this perspective, the intron inserted in the region of the 25S rRNA gene sequences offers several advantages to the Candida spp. genotypes (McCullough et al., 1999). The intron in the 25S rRNA gene has been shown to have a high heterogeneity within the Candida species (Hanafy and Morsy, 2012). The contribution of the intron inserted in the region of the 25S rRNA in clinical diagnosis remains to be determined, due to the lack of a complete molecular database that could enable the systematic comparison of the inter-and intra-species variations in the different isolates among the Candida spp.
Most of the regions of the large ribosomal subunit genes (LRSg) of yeasts are reproducible markers, which provided very useful information concerning the phylogenetic relationships among the various marine yeasts (Fell and Kurtzman, 1990).
It is not surprising that some irregularities are seen in the taxonomy of Candida. Several recent studies have described the Candida isolates, whose properties do not correspond precisely with the descriptions of the classical species, leading to further confusion (Mahrous et al., 1990;McCullough et al., 1994McCullough et al., , 1995Boerlin et al., 1995). It is, therefore, the right time to assess the potential contribution that other techniques could make towards the identification of the relationships between C. albicans and C. dubliniensis. Sullivan and Coleman (1997) indicated the necessity of further confirmation, which can be obtained by conducting any of the several DNA fingerprint techniques available, as well as by RFLP and RAPD analysis. These are also effective, as well as quicker and easier to perform in order to discriminate between C. albicans and C. dubliniensis.
Indeed, the comparative nucleotide sequence analysis of the rDNA has been used extensively to study the evolutionary relationships among a wide variety of fungi. Most of these studies have been performed on small ribosomal subunit gene sequences (Hendriks et al., 1991;Fleischmann et al., 2004). A search conducted in the Gene Bank nucleotide sequence databases over the past few decades revealed that the sequence data on the rDNA genes have been reported only for a large number of Candida spp. However, these studies indicate that the 25S gene sequences can be used to confirm the natural relationships within the genus, such as the close evolutionary relationship between C. albicans and C. dubliniensis, based on biochemical and phenotypic criteria (Kumar et al., 2006;Nawrot et al., 2010).
The aim of this study was to achieve a detailed and unequivocal understanding of the evolutionary relationships between C. albicans and C. dubliniensis, by performing rapid genotyping based on simple PCR, RFLP-PCR, RAPD-PCR and emphasizing the identification of genotype patterns for both C. dubliniensis and C. albicans by using the sequencing tools.

Yeast collection and cultural characters
A total of 60 clinical vaginal swabs were collected from the clinics in the province of Babylon, Iraq, during the study conducted in 2013-2014. Clinical samples using a sterile cotton swab were taken from the vagina of patients exhibiting clinical signs of the vaginal candidiasis based on Imran and Al. shukry (2014). They were transferred to the biotechnical laboratory where they were directly streaked on Sabouraud agar medium (SDA) supplemented with chloramphenicol and streptomycin (50:50 µg/ml). After inoculation, the Petri plates were incubated for 24-48 h at 37°C (Sivakumar et al., 2009;Baveja, 2010). All the sixty isolates in this study were subjected to preliminary identification was done based on CHROMagar Candida (Ghelardi et al., 2008;Marsh and Martin, 2009;Nadeem et al., 2010).

Extraction of genomic DNA
Twenty isolates out of 60 isolates of Candida spp. were subjected to DNA extraction and PCR assays .A loop full of Candida colony was suspended in the lysis buffer (200 mM Tris-HCl, 20 mM EDTA, 150 mM NaCl and 0.5% SDS) and heated in water bath at 95°C for 2 min. The suspension was centrifuged at 5000 rpm for 2 min and the supernatants were decanted into new sterile tubes, and precipitated with an absolute alcohol and then, washed DNA pellet by 70% ethyl alcohol, dried pellet of DNA dissolved in elution buffer and preserved at -20°C until use (Fredricks et al., 2005).
1 μL of DNA (20 µg/ml) from each of 20 Candida isolates were mixed with PCR mixture ( final reaction volume 25 μL) consisted of 12 μL of 2x Master Mix (Promega), 2 μL of primers (10 pmole) and rest molecular-grade water. The PCR conditions for CA-INT-L and CA-INT-R primers were 95°C for 3 min followed by 30 cycles 94°C for 1 min, annealing temperature 65°C for 1 min. Extensions temperature 72°C for 2.5 min followed by final extension temperature 72°C for 7 min. The PCR conditions for primer pairs CABF59F and CADBR125R was similar to previous cycle except annealing temperature which was 55°C in place of 65°C. The PCR mixture was amplified by thermal cycler PCR System (Labnet, USA).
The PCR products for each target region were run on 1.2% agarose gel (Bio Basic Canada Inc.). Electrophoresis was performed at 100 V. in TBE buffer. The gel was pre-stained with 0.05% ethidium bromide. The DNA bands were detected by Desktop Gel imager scope 21 ultraviolet transilluminator (Korea Com.).

PCR-RFLP assay
The PCR-RFLP assay was performed as described earlier by Mirhendi et al. (2006). In brief, the incubation of a 10 μL aliquot of t h e PCR products consisted of 12 μL of 2x Master Mix (Promega), 2 μL of primers (10 pmole) and rest molecular-grade water of amplified intron region of 25S rRNA gene with 10 μL of Hae111 and Hhal cocktail restriction enzymes (Promega, USA) was performed in single reaction, at 37°C for 3 h, using both enzymes. Next, 8 μL of the RFLP-PCR products were run on 1.5% agarose gel at 70 V for 60 min. The gel was pre-stained using 0.05% ethidium bromide and visualized under UV light and photographed by the Desk Gel imager scope 21 ultraviolet transilluminator (Korea Com).

RAPD-PCR assay
RAPD-PCR was accomplished by utilizing a total volume of 34 μL consisting of 0.7 μL (20 µg/ml) genomic DNA, 18 μL of 2x master mix (Promega USA) 12 μL molecular-grade water and 1.5 μL (50 pmole) of random primer GGTGTAGTGT. The mixture was amplified under the following conditions: 95°C for 4 min; 38 cycles at 94°C for 1 min; 36°C for 1.5 min; 72°C for 1.30 min and 72°C for 8 min (Labnet PCR System). Further, 8 μL of the PCR products were run on 1.5% agarose gel at 70 V for 60 min. The gel was prestained with 0.05% ethidium bromide.

Sequencing assay
To study the relationship and similarity at morphological and molecular level that are sometimes exhibited between the C. albicans and C. dubliniensis isolates particularly because C. albicans and C. dubliniensis also possess the transposable intron in the 25S rDNA, the genomic DNA of the representative isolates for both C. albicans and C. dubliniensis were amplified with the CA-INT primers. After PCR amplification, the purified products for 8 isolates were sequenced. The PCR primers CA-INT-L was used for DNA sequencing of transposable intron in the 25S rDNA of Candida isolates. Sequence analysis was performed using the Macro Gene Company, USA. The sequence alignment of C. albicans and C. dubliniensis was compared with the BLAST database and were aligned with sequences from the BLAST database derived from the following reference strains: (C. dubliniensis sequence ID: embІ FM992695.1 United Kingdom isolate; C. albicans sequence ID: gbІ DQ465844.1 New Zealand isolate).

Phylogenetic analysis
The phylogenetic tree dendrograms of RFLP and RAPD-PCR products for isolates of C. albicans and C. dubliniensis was created by clustering methods applied on distance matrix unweighted pair group method with arithmetic mean (UPGM) which offers automatic lane/band detection, band matching and molecular weight computation. The phylogeny tree computation was analyzed based on UVI band software GD/45230 for gel image analysis. The software is able to analyze gel image patterns of bands for different isolates or species and generates phylogenetic tree based on the information available in gel image. It also evaluated the similarity coefficient factor according to Mackenstedt et al. (1994) and Ute et al. (1994). The phylogenetic tree based on sequencing and sequence table were constructed employing the Mega 6 software.

Phenotypic and molecular diagnosis for Candida spp.
All the sixty isolates in this study showed up in green color on the CHROMagar Candida. The results of molecular assay showed that 20 isolates of Candida were identified as C. albicans. The amplification of the targeted region produced an amplicon of size 665 bp (Figure 1). The target regions of three isolates for Candida showed faint bands as seen in Figure 1, lanes B, J and Q.

Genotyping of Candida spp. by CA-INT primer pair
The specific primer CA-INT was designed to flank the transposable intron region of the 25S rRNA gene. PCR was successfully amplified the target region of the genomic DNA of the 20 isolates. The amplification result designated five isolates as C. dubliniensis, which had a high PCR product (1080 bp). Thus, 16 isolates of C. albicans, with low PCR products could be classified and three genotypes could be designated viz., (i) genotype C, (ii) (450 and ~840 bp), (iii) A genotype of (~840 bp) of C. albicans (450 bp) (Figure 2).

RFLP-PCR assay
Both restriction enzymes (Hhal and Hae 111 enzyme) have an equal chance of making a cut anywhere in the PCR product. However, the restriction banding patterns by using the Hhal enzyme showed large fragments (500 bp) of C. dubliniensis, as in Figure 3a (lanes B, H, J, K and Q).
This enzyme also revealed a similar basal band with the PCR fragments of average length approximately 380 bp in all the isolates for C. albicans and C. dubliniensis. However, the use of the Hhal enzyme cut PCR products into many short PCR fragments (<100 bp), as seen in Figure 3a. The restriction banding patterns by using the Hae 111 enzyme showed characteristic cleavage profile (350, 300, 180 and 60 bp fragments) for C. dubliniensis.
However, the PCR products of C. albicans also showed variation in their RFLP patterns. The first pattern revealed two fragments such as C, E-G, I and M, whereas the second pattern was composed of four fragments such as D, L, N, O, PT and U (Figure 3a). The isolates of C. dubliniensis B, H, J, K and Q showed variation in their RFLP-PCR patterns, in which the H, J and K appeared closely related, while the B and Q showed differences ( Figure 3b). The Hae 111 enzyme resulted in a large fragment of the PCR, of C. dubliniensis, of about 350 bp as in the B, H, J, K and Q lanes. The PCR product of the albicans isolates showed fragments of about 300 bp, as in C, E-G, I, M, R and R-S, as well as a fragment of 270 bp in D, L, N-P, T-U; This enzyme also revealed a similar basal band of average length of the PCR fragments, approximately 100 bp, in all the isolates for C. albicans and C. dubliniensis ( Figure 4 a).

RAPD-PCR assay
The results show that the oligo primer GGTGTAGTGT successfully genotyped the 20 isolates of C. albicans and C. dubliniensis into 7 genotypes: C. dubliniensis revealed three genotypes, while C. albicans showed four genotypes ( Figure 5). RAPD-PCR produced multiple bands, the main band of which was consistently present in all the

Sequence analysis
The results of sequence analysis for eight isolates of Candida spp. showed a similarity with the entry in the percentage sequence of >99.5% with the intron region of the 25S rRNA gene. Genotypes A and C of C. albicans isolates (D, T and E) showed high similarity of about 99.98% in their sequence at the same time, the genotype of C. dubliniensis isolates (B and Q) showed high similarity of about 99.97% with the A and C genotypes of C. albicans isolates (D, T and E), genotype B of C. albicans isolate (R) showed similarity of 99.96 when compare with the genotypes C and A of C. albicans isolates (C and L) ( Figure 6).  Figure 7 showed sequence analysis for eight isolates of Candida spp. C. albicans isolates D, T and E showed high similarity in their sequence with C. dubliniensis isolates B and Q except difference in two adenine nucleotides at the same time, the C. albicans isolates C, L and R showed high difference in their sequence with C. albicans isolates D, T and E and C. dubliniensis isolates B and Q.

DISCUSSION
Our results concurred with several recent studies and demonstrated a wide degree of genetic homogeneity between C. dubliniensis and C. albicans (Jackson et al., 2009). The results of Boucher et al. (1996) found that both the intron and ribosomal RNA nucleotide sequences were almost perfectly identical between the different C. albicans strains, as well as between the C. albicans and C. dubliniensis. Although it is difficult to distinguish between the C. albicans and C. dubliniensis colonies formed on CHROMagar which are green in color, the CHROMagar medium can be unstable following subculture or storage (Schoofs et al., 1997;Sullivan and Coleman, 1998). This result indicated that the specific primer pair (CABF59F and CADBR125R) for C. albicans amplified its target in C. dubliniensis as well as in 16 isolates of C. albicans. Despite the concurrence in the results for C. albicans with those of Costa et al. (2010), the result with C. dubliniensis indicated the similarity of the sequence target region for C. dubliniensis and C. albicans. This result concurred with the report of Pujol   Both the restriction enzymes, Hhal and Hae 111, revealed a high degree of polymorphisms with respect to RFLP-fingerprinting in both C. albicans and C. dubliniensis. In total, 16 isolates of C. albicans showed polymorphic RFLP-patterns. However, the use of the Hhal enzyme revealed a similar basal band with an average length of the PCR fragments at approximately 380 bp in all the isolates for C. albicans and C. dubliniensis, the phylogenetic tree highlighting the degree of homogeneity in the sequence of recognition site of enzyme between C. albicans and C. dubliniensis. The Hae111 enzyme elucidated polymorphic RFLPpatterns in both C. albicans and C. dubliniensis and yielded different fragment sizes with C. dubliniensis having other short PCR fragments. On the other hand, the PCR products of C. albicans also showed variation in their RFLP-patterns. This enzyme also revealed a similar basal band of average length of the PCR fragments, approximately 100 bp in all the isolates for C. albicans and C. dubliniensis. This implied a similarity in the recognition site and the same sequence in this site for both species.
The phylogenetic tree supported the natural variation such as mating, mutations and recombination which may occurred in many isolates related to the same species. Our result coincidence with many studies (Mirhendi et al., 2005;Mirhendi et al., 2006;Shokohi et al., 2010).
The sequence introns of 25S from eight isolates of C. albicans and C. dubliniensis were retrieved through BLAST analysis, the sequences in C. albicans and C. dubliniensis were found to differ by approximately 0.02%. This difference is not similar to the differences in the sequences found between C. tropicalis and C. maltosa (2.8%) and between C. parapsilosis and the ascosporic species, Lodderomyces elongisporus (3.2%). Based on this, results were contrary to those of Peterson and Kurtzman (1991), in which they earlier suggested that strong evidence for a separate species exists when this region contains a substitution of greater than 1% of the nucleotide between t h e two organisms. Most of the phenotypic characters of C. dubliniensis d i d n o t serve to confirm the taxonomic status of a distinct species.

Amendment of C. dubliniensis: subspecies novus
In the light of these results, despite the C. dubliniensis had been described as a separate species over the past decade (Sullivan et al., 1995;Tamura et al., 2001), the understanding that prevailed was that a few variations occasionally occurred in many strains due to natural selection, which were, however, insufficient to justify the emergence of a new species. besides strong confirmation from many studies including the work of Coleman et al. (1997) who revealed the very close similarity among the isolates of C. dubliniensis to those of C. albicans. Particularly, because C. dubliniensis and C. albicans are phenotypically very similar, it is highly likely that the isolates of C. dubliniensis had been misidentified as C. albicans or C. stellatoidea in the past (Anthony et al., 1995;Boerlin et al., 1995;Coleman et al., 1997). On the other hand, it is shown that it is impossible to consider any variation in the phenotypic and genetic properties, which were not contingent upon their definitive identification as C. albicans, based on the views of Coleman et al. (1997) to be the emergence of a new species.
In spite of the C. albicans and C. dubliniensis, isolates produced chlamydospores on the TOC agar medium based on Tamura et al. (2001). Our results showed the density of the chlamydospores produced by both species were not a good taxonomic character from which to draw any conclusion.
Our results demonstrate that C. dubliniensis shares a very close relationship with C. albicans based on the results of both the RFLP, RAPD-PCR patterns and the sequence marker shown in Figures 6 and 7, this result coincides with prior studies that suggested that C. dubliniensis isolates were not merely mutant derivatives of C. albicans; in the same trend, we do not observe sufficient differences to separate C. dubliniensis from C. albicans to warrant a species status. Therefore, we provided further support for its designation and confirmed that C. dubliniensis should be considered as a subspecies of C. albicans. This judgment, based on the molecular RFLP patterns, such as the Hhal and Hae 111 enzymes revealed a similar basal band, this indicated the presence of the same sequence and recognized the region in both species. Sequence also confirmed part of this truth based on sequence analysis. These results are in agreement with those of Jackson et al. (2009) and  where they refer to the requirement for further confirmation, which can be obtained by performing any of the several DNA fingerprint techniques available. The phylogenetic tree, based on the sequencing of the introns on the 25S gene, showed close similarity (99.5%) between C. dubliniensis and C. albicans as shown in Figure 7, with only subtle differences in sequence between the two species.
The amendment of C. dubliniensis taxonomic state agrees with early and recent studies that are closely related C. dubliniensis to C. albicans which was routinely misidentified as C. albicans (Sullivan et al., 1995;Moran et al., 2004;Jones et al., 2004). Based on the results of Tamura et al. (2001), the genotype 1080 bp was elucidated only as C. dubliniensis on typing a transposable intron region in the 25S rRNA gene from other four genotype strains viz.: genotypes of C. albicans (genotype C = 450 and 840 bp, genotype A = ~ 840bp, genotype B = 450 bp and genotype E= 1400 bp), these genotypes continued to remain as different strains of C. albicans. Tamura et al. (2001) referred to the genotype E strain which showed a high degree of similarity to C. dubliniensis when compared with the degree of similarity of the strains of the other C. albicans genotypes, in which the similarity was determined based on the group I intron sequence; however, from his results, he neglected to include this genotype within C. dubliniensis, as was expected. From our view, with his erroneous taxonomic judgment along with his temperament and individuality, based on the trend of Tamura et al. (2001), each one of the all the genotypes of Candida (450, ~840, 450+850 and 1400 bp) was merited to be included as a new species at the same time. We think it is insufficient to justify the emergence and support genotype 1080 bp of a new species by Tamura et al. (2001). Therefore, the differentiation of the two taxa was based on the color of the colony on CHROMagar. Candida, thus, may not be as reliable as was considered earlier, to utilize CHROMagar to differentiate between C. albicans and C. dubliniensis. Tamura et al. (2001) revealed a dark blue color of the colonies on CHROMagar, which could not confirm the differentiation of four of the C. dubliniensis isolates out of five. They also used the growth at 45°C as a criterion for the differentiation between the two species. This was confuted Imran 1213 by Tamura et al. (2001) when he referred to all the C. albicans genotypes, including the five C. dubliniensis strains, which grew well at 45°C on a culture media such as PDA and Sabouraud dextrose agar. We concluded that, it is impossible to consider any few variations in the phenotypic and genetic properties of the Candida strains, and showed that a few variations occasionally occurred in many strains due to natural selection, which were however, insufficient to justify the emergence of a new species. Besides strong confirmation from previous studies which revealed the very close similarity among the isolates of C. dubliniensis to those of C. albicans, C. dubliniensis is not an emerging new species. We provided further support for its designation and confirmed that C. dubliniensis do not merit being included as a new species and should be considered as C. albicans subspecies dubliniensis stat. et.comb. nov.

Ethical approval
Author hereby declare that all actions have been examined and approved by the appropriate ethics committee and have therefore been performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki.