Practical Method for Detection and Identification of Candida, Aspergillus, and Scedosporium spp. by Use of Rolling-Circle Amplification

Abstract

A sensitive rolling-circle amplification (RCA)-based method utilizing species-specific padlock probes targeted to the internal transcribed spacer 2 region of the fungal ribosomal DNA gene complex was developed. The assay was rapid (2 hours) and specific. Of 28 fungal isolates (16 of Candida, six of Aspergillus, and six of Scedosporium spp.), all were all identified correctly.

Fungal pathogens cause life-threatening infections in critically ill and immunosuppressed patients. Contemporary epidemiological trends reveal a shift toward species of Candida and Aspergillus other than Candida albicans and Aspergillus fumigatus and a range of emerging fungi including Scedosporium spp. and the zygomycetes (6, 19). Given the reduced susceptibility of many of these pathogens to antifungal agents, timely identification to species level is essential for clinical management. However, standard culture-based identification methods are insensitive and slow (15).

To overcome both problems, PCR-based tools have been developed. In particular, the internal transcribed spacer (ITS) regions, ITS1 and ITS2, of the fungal ribosomal DNA gene complex have shown promise as targets for species identification in a variety of formats including multiplex and/or real-time PCR assays (9, 16), DNA sequence analysis (1, 2, 12), and probe-based techniques (5, 7). The latter range from Southern blotting (5, 7) and reverse line blot (RLB) hybridization methods (23) to sophisticated microarray formats (10, 11, 17).

Recently, the utility of circularizable oligonucleotide (padlock) probes has been demonstrated for the detection of target nucleic acid sequences, including nucleotide polymorphisms that differ by only a few base pairs, with high sensitivity (4, 13, 20). Such probes comprise two sequences complementary to the 5′ and 3′ termini of the target sequence joined by a linker region (Fig. 1A). Upon hybridization to the target, the probe ends are joined by DNA ligase to form a closed molecule. The intensity of the probe-specific signal is then increased exponentially by rolling-circle amplification (RCA) (13) (Fig. 1B). There are few data on the application of padlock probes in the detection of polymorphisms in fungi. We report on a sensitive, RCA-based method using real-time PCR for species identification of clinically important Candida, Aspergillus, and Scedosporium spp.

FIG. 1.

(A) Typical design of a circularizable padlock probe as exemplified by the Candida albicans-specific (CAL) probe. The probe comprises (i) a 5′-phosphorylated end, (ii) a “backbone” containing binding sites for the RCA primers (RCA primers 1 and 2, respectively; designated by bold uppercase letters) as well as the nonspecific linker regions (designated by bold lowercase letters), and (iii) a 3′ end. The 5′ and 3′ ends of the probe are complementary to the 5′ and 3′ termini of the target sequence in reverse, in this example to the C. albicans sequence (GenBank accession no. AF455531). Abbreviations: 5′-P, 5′-phosphorylated binding arm; 3′, 3′ binding arm. (B) Pictorial representation of the RCA method. Step 1, hybridization. Hybridization of padlock probe, containing target-complementary segments, to a target DNA sequence. Step 2, ligation. The probe is circularized by DNA ligase. Step 3, RCA and primer extension I. Ligated probe and binding of RCA primer 1 for RCA. Tandem repeat sequences complementary to the circular probe are generated by RCA. The reverse primer (RCA primer 2) binds to each tandem repeat generated by the rolling circle. Step 4, RCA and primer extension II. As the original RCA strand elongates, further priming events are initiated by primer 2, generating displaced DNA strands. As a result, new priming sites for the first primer (primer 1) are generated. The two primers thus function to generate a self-propagating pattern of DNA fragment release events (20). Step 5, detection of amplified product. RCA may be monitored using real-time PCR or agarose gel electrophoresis. ssDNA, single-stranded DNA.

Twenty-eight clinical isolates were studied: two of C. albicans, two of Candida glabrata, three of Candida krusei, three of Candida tropicalis, three of Candida dubliniensis, three of Candida guilliermondii, four of A. fumigatus, two of Aspergillus flavus, and three strains each of Scedosporium apiospermum and Scedosporium prolificans. Species identity was confirmed by standard laboratory methods (3, 21) and ITS sequence analysis (23). Isolates were stored in sterile water at 25°C until required. DNA extraction and amplification of the ITS (ITS1, 5.8S rRNA, and ITS2) region, using the primers ITS1 and ITS4 (22), in preparation for hybridization with padlock probes (see below) were performed as previously described (23).

DNA sequences spanning the ITS2 region of all major Candida, Aspergillus, and Scedosporium species were accessed from the GenBank database and compared using the Clustal W program (BioManager, ANGIS; http://biomanager.angis.org.au/) to identify informative nucleotide polymorphisms. Ten padlock probes targeting the ITS2 region (six specific for each Candida sp., one each for A. fumigatus and A. flavus, and one each for S. apiospermum and S. prolificans) were then designed. To optimize binding to target DNA, the melting temperature of the 5′-end probe binding arm was higher than the temperature used for probe ligation (62°C). To increase specificity, the 3′-end binding arm was designed with a melting temperature of 50 to 57°C (i.e., below the ligation temperature). Furthermore, the linker region of each probe was designed to minimize similarity to other closely related pathogens and to allow primer binding during RCA (Fig. 1A). Table 1 gives the sequences of the two primers used for RCA (RCA primer 1 and RCA primer 2).

TABLE 1.

Oligonucleotide padlock probes and primers used for RCA

Probe or primer	Target species	GenBank accession no.	Sequences and locations of the two binding arms in comparison with relevant reference GenBank sequences (5′-3′) or sequence of primera
Probes
CAL	C. albicans	AF455531	pa-457AATGTTTTTGGTTAGACCTAAGCCATT431
			gatcabTGCTTCTTCGGTGCCCATtacgaggtgcggatagctaccCGCGCAGACACGATAgtcta
			467CGCCGCAAGC458
CGL	C. glabrata	AF167993	pa-719GTAAAACCTAATACAGTATTAACCCCCGC691
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			736AGATCAACACCGAGTTG720
CKR	C. krusei	AF246989	pa-406GCCAGCTTCGCTCCCTTT389
			gatcaTGCTTCTTCGGTGCCCATtacgag gtgcggatagctacCGCGCAGACACGATAgtcta
			423AAAGTCTAGTTCGCTCG407
CTR	C. tropicalis	AF268095	pa-365GCGTATTGCTCAACACCAAACC344
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			383TTCTTTCAAACAAACCTA366
CDU	C. dubliniensis	AJ249485	pa-450CGCCTTAGCAATGTTTTTGGTTA428
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			463CGACGCCAGAGAC451
CGU	C. guilliermondii	AF455495	pa-513CGGGCCAACAATACCAGAAATAT491
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			528TGGTTGTTGTAAGGC514
AFUM	A. fumigatus	AY660923	pa-144CGGCTGGCGCCGGCCGGG127
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			162AAAAATAAAGTTGGGTGT145
AFLA	A. flavus	EF197070	pa-176TTTGCGTTCGGCAAGCGC159
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			193ACCTGGAAAAAGATTGA177
SAPI	S. apiospermum	AY228118	pa-582AACCGCGGCGGGACCGCC565
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			598TTACTACGCAGAAGGC583
SPRO	S. prolificans	AY228117	pa-443GTAAATCTACTACGCAGAAGGCGC420
			gatcaTGCTTCTTCGGTGCCCATtacgaggtgcggatagctacCGCGCAGACACGATAgtcta
			457ACCCAATGCGAGTT444
Primers
RCA primer 1b			ATGGGCACCGAAGAAGCAb
RCA primer 2c			CGCGCAGACACGATAc

^aRefers to the 5′ end of the probe; “p-” indicates 5′ phosphorylation. The sequences of the 5′ and 3′ binding arms of the probes (derived from reference GenBank sequences) are underlined; these are joined by the backbone of the probe including the nonspecific linker region where the sequence is shown in lowercase.

^bBinding site of RCA primer 1 to the padlock probe, generating a long single-stranded DNA. The sequence of RCA primer 1 is the complement of the segments, in reverse, represented by bold italic nucleotides.

^cBinding site of RCA primer 2. RCA primer 2 binds to nascent single-stranded DNAs as their binding sites become available. Its sequence is the same as that of the segments shown in nonitalic bold.

Purified amplified PCR product (10¹¹-copy-number DNA template [DNA calculator; http://www.uri.edu/research/gsc/resources/cndna.html]) was mixed with 2 U of Pfu DNA ligase (Stratagene, La Jolla, CA) and 0.1 μM padlock probe as previously described (18) and subjected to multiple cycle ligation followed by exonucleolysis to remove unligated probe and template PCR product (18). RCA reactions of the circularized probes were then performed according to previously described experimental conditions (18, 20). Probe signals were amplified by incubation at 65°C for 30 min, and accumulation of double-stranded DNA products was monitored using a Corbett RotorGene 6000 machine (Corbett Research, Mortlake, Australia); probe signals were also visualized on a 1.5% agarose gel to verify the specificity of probe-template binding.

The universal fungal primers ITS1 and ITS4 amplified the ITS region of all study isolates. Further, individual species-specific probes, designed on the basis of species-specific signatures identified by alignment of ITS2 sequence data, successfully identified all strains. To assess assay sensitivity, RCA was performed on serial dilutions of target template (10¹¹ copies). For all isolates, a measurable signal was observed using real-time PCR (schematically shown in Fig. 1B). The sensitivity of detection was 10⁷ copies (data not shown). The duration of the RCA procedure was 2 h, but a positive signal was evident 15 min after commencement of the reaction.

The RCA assay was also highly specific. Amplification of probe signals was seen only with matched template-probe mixtures; DNA from species which contained ITS2 polymorphisms not targeted by the padlock probe in use produced no signal (Fig. 2). Species-specific probes correctly identified all six Candida species studied including C. dubliniensis and C. guilliermondii. The assay also clearly differentiated between A. fumigatus and A. flavus and between S. apiospermum and S. prolificans. Concordance with phenotypic identification and ITS sequencing results (23) was 100%.

FIG. 2.

Gel representation of specificity of RCA probes. Amplification of probe signals was seen only with matched template-probe mixtures (empty lanes denote absence of signals with unmatched template-probe mixtures). The species-specific probes are labeled as shown at the top of the figure (CDU, Candida dubliniensis; CAL, Candida albicans; CGL, Candida glabrata; CKR, Candida krusei; CTR, Candida tropicalis; CGU, Candida guilliermondii; SPRO, Scedosporium prolificans; SAPI, Scedosporium apiospermum; AFUM, Aspergillus fumigatus; AFLA, Aspergillus flavus). Lane 1, 100-bp DNA Ladder MW marker (Invitrogen, Mount Waverly, Australia); lanes 2, 8, 14, 20, 26, and 32, C. dubliniensis WM 03.119; lanes 3, 9, 15, 21, 27, and 33, C. albicans WM 01.218; lanes 4, 10, 16, 22, 28, and 34, C. glabrata WM 02.58; lanes 5, 11, 17, 23, 29, and 35, C. krusei WM 03.204; lanes 6, 12, 18, 24, 30, and 36, C. tropicalis WM 233; lanes 7, 13, 19, 25, 31, and 37, C. guilliermondii WM 02.72; lane 38, S. prolificans WM 06.372; lane 39, S. apiospermum WM 06.466; lane 40, S. apiospermum WM 06.466; lane 41, S. prolificans WM 06.372; lane 42, A. fumigatus WM 06.96; lane 43, A. flavus WM 06.93; lane 44, A. fumigatus WM 06.96; lane 45, A. flavus WM 06.93.

The continued development of reliable diagnostic tools for the early detection and identification of fungi remains a priority for improving patient outcomes. This proof-of principle study confirms the great potential of a simple, rapid (2-h), and highly specific RCA-based assay for the identification of pathogenic fungal species, in this instance Candida, Aspergillus, and Scedosporium.

ITS2-targeted padlock probes were employed as the ITS2 region contains species-specific sequences well suited for identification of yeasts and molds (2, 12). Signature polymorphisms between the major Candida spp. were successfully targeted by individual probes. In particular, C. albicans was clearly distinguished from C. dubliniensis, in contrast to the use of phenotypic methods, when separation of these two species is often unreliable (8). The RCA assay also distinguished unambiguously between A. fumigatus and A. flavus and between S. apiospermum and S. prolificans. The latter are two significant pathogens of seriously ill patients which have different epidemiological and clinical associations (14).

Among an increasing number of molecular methods established for species identification of fungal pathogens (especially Candida and Aspergillus), ITS sequence analysis has emerged as the “gold standard.” However, sequencing is expensive, has a 2-to 3-day turnaround time, and is impractical for analyzing large numbers of isolates. Probe-based methods thus far offer a choice between broad-range detection and identification of a limited number of species, although more recently development of sensitive RLB and microarray formats has allowed the simultaneous detection of multiple pathogens in one test (10, 23). RLB approaches are best suited to “batch testing” of isolates, while microarray systems require expensive specialized equipment and are not readily adaptable in clinical laboratories.

Advantages of the RCA-based assay developed in this study include its specificity and flexibility in that probes can be custom made to meet specific requirements, e.g., for a particular clinical setting. While the setup costs of the assay ($A300 per probe) are relatively high, the probes may be used up to 5,000 times and there is no need for specialized equipment. The technique has recently been used to detect single nucleotide polymorphisms within species of bacteria (18), indicating that RCA is a practicable option whereby selective targeting of single nucleotide polymorphisms can discriminate between closely related species. Further evaluation of this system using a broad range of fungal species is indicated to clarify its role in routine diagnosis.