INTRODUCTION
Microsporidia are obligate intracellular parasitic fungi known to cause a wide range of human infections, involving various organ systems including the eye. Modified trichrome staining of fixed ocular specimens followed by examination under light microscopy remains the most widely used method for diagnosis of microsporidia ocular infections. Sensitivity is comparable to that of the gold standard electron microscopy, but does not allow for definitive species identification. Electron microscopy is the traditional gold standard for identification of microsporidia to species level but requires expertise and the use of expensive and sophisticated equipment (Weber, Reference Weber, Schwartz, Deplazes, Wittner and Weiss1999). Recently developed molecular techniques provide a simple means of identification of microsporidia to species level (Muller et al. Reference Muller, Stellermann, Hartmann, Schrappe, Fätkenheuer, Salzberger, Diehl and Franzen1999; Joseph et al. Reference Joseph, Sharma, Murthy, Krishna, Garg, Nutheti, Kenneth and Balasubramanian2006). To our knowledge, there is currently no literature describing DNA extraction and identification of microsporidia species from modified trichome-stained clinical slides.
We modified previously described methods of DNA extraction and PCR sequencing for microsporidial identification from non-clinical, museum Giemsa-stained and unstained slides (Hyliš et al. Reference Hyliš, Weiser, Obornik and Vavra2005), fresh corneal scrapings (Joseph et al. Reference Joseph, Sharma, Murthy, Krishna, Garg, Nutheti, Kenneth and Balasubramanian2006), stool specimens (Muller et al. Reference Muller, Stellermann, Hartmann, Schrappe, Fätkenheuer, Salzberger, Diehl and Franzen1999) and environmental samples (Dowd et al. Reference Dowd, Gerba and Pepper1998), to develop a protocol for species identification of microsporidia from trichrome-stained slides of clinical specimens originally used for microscopical diagnosis.
MATERIALS AND METHODS
Modified trichome slides of corneal scrapings, which had been used to diagnose microsporidial infection, were first cleaned with xylene to remove the adherent immersion oil film. Following the protocol previously described by Hyliš et al. (Reference Hyliš, Weiser, Obornik and Vavra2005), approximately 80 μl of TE buffer (pH 8·0) were added to the fixed specimen on the slides and spread evenly over the surface (about 1·5 cm in diameter), and then allowed to stand for at least 5 min. Using a clean Eppendorf pipette tip, all visible pieces of fixed tissue were scraped off the slide and, along with all the applied TE buffer, transferred into a clean 1·5 ml Eppendorf tube containing an equal amount of 0·5 mm glass beads (Sigma-Aldrich Chemie GmbH, Steinheim, Germany) and 180 μl of ATL buffer from the DNeasy Tissue kit (Qiagen GmbH, Hilden, Germany). The tube was then vortexed for 30 sec and 20 μl of proteinase K added before being vortexed again and incubated at 55°C for at least 4 h. Thereafter, following the protocol described by Dowd et al. (Reference Dowd, Gerba and Pepper1998), 200 μl of AL buffer were added, the sample vortexed thoroughly and incubated at 70°C for 10 min and then at 98°C for another 10 min. then 200 μl of lysate were transferred to a new 1·5 ml Eppendorf tube with 200 ml of absolute ethanol. The remaining steps followed the protocol provided by the Qiagen DNeasy kit for tissue DNA extraction up to the elution step. For elution, 100 μl of AE buffer were added to the column which was incubated for 10 min at 70°C, followed by centrifugation at 20 000 g for 1 min. The eluate was replaced in the same column for incubation for a further 10 min at 70°C before final centrifugation at 20 000 g for 1 min.
The polymerase chain reaction (PCR) was carried out in a reaction volume of 50 μl containing 25 pmol each of pan-microsporidian primers (Muller et al. Reference Muller, Stellermann, Hartmann, Schrappe, Fätkenheuer, Salzberger, Diehl and Franzen1999; Joseph et al. Reference Joseph, Sharma, Murthy, Krishna, Garg, Nutheti, Kenneth and Balasubramanian2006), V1 (5′-CAC CAG GTT GAT TCT GCC TGA C-3′) and PMP2 (5′-CCT CTC CGG AAC CAA ACC CTG-3′), dNTP 200 μm, 2·5U Taq polymerase, 10 mm Tris-HCl (pH 9·0), 50 mm KCl, 1·5 mm MgCl2, and 40 μl of DNA extract to boost the DNA load. The PCR conditions were as follows: 94°C for 3 min initial denaturation, 40 cycles of 94°C for 1 min, 60°C for 2 min, and 72°C for 3 min, and 72°C for 10 min termination cycle. DNA extract from Encephalitozoon intestinalis (ATCC 50651) was used as a positive control. For the negative controls, the DNA extraction procedure described above was employed, but starting out with the 80 μl of TE buffer (pH 8·0) without the slides, to obtain blank extracts.
For samples showing no PCR product with the first primer pair V1::PMP2, a further PCR using ‘universal’ microsporidian rDNA1 primer pair (Weiss and Vossbrinck, Reference Weiss, Vossbrinck, Wittner and Weiss1999), ss18f (V1)::ss1492r (5′-GGT TAC CTT GTT ACG ACT T-3′) was run to detect microsporidia species which were not amplified by the V1::PMP2 primer pair.
The products of PCR amplification were identified through electrophoresis on a 1·5% agarose gel and stained with ethidium bromide for visualization under UV trans-illumination. For all DNA extracts from specimens showing no PCR products, a repeat PCR reaction was performed under the same conditions but spiked with a positive control to rule out the presence of inhibitors. The PCR products were purified by using the QIAquick PCR purification kit (Qiagen) and both the forward and reverse PCR primers were then used for dye termination PCR sequencing. The obtained SSU-rDNA sequences were then edited using FinchTV (http://www.geospiza.com/finchtv/index.htm) and then compared with sequences in GenBank (http://www.ncbi.nlm.nih.gov) using BLAST 2.0.
RESULTS
We selected modified trichrome-stained slides of corneal scrapings from 26 different patients diagnosed clinically with keratoconjunctivitis, from September 2006 to May 2007. All these patients had no history of Human Immunodeficiency Virus (HIV) infection. In total, 13 of these slides were deemed positive and 13 negative for microsporidia, based on examination by light microscopy. Of the 13 positive slides, 11 showed PCR products with the V1::PMP2 primer pair and subsequent sequencing identified all 11 as being Vittaforma corneae (GenBank Accession numbers: U11046 and L39113), with sequence identities ranging from 96% to 99% on BLAST analysis (GenBank Accession numbers for the 11 sequences of our PCR products: EU436724 to EU436734). The remaining 2 positive slides which showed no PCR products with primer pair V1::PMP2 also showed no PCR products on subsequent PCR with the primer pair ss18f (V1)::ss1492r. All the 13 negative slides went through the same process as the positive slides and all showed no PCR products for both primer pairs and, along with the 2 positive slides with no PCR products, were all subsequently tested negative for the presence of inhibitors.
We validated our protocol on the ATCC strain of E. intestinalis (ATCC 50651) with both primer pairs and achieved the expected results of Encephalitozoon(Septata) intestinalis (GenBank Accession numbers: U09929 and L39113), with a blast sequence identity of 99% (GenBank Accession number for the sequence of our positive control PCR product: EU436735).
DISCUSSION
Identification of microsporidia in ocular infections to species level may have an impact on the management and prognosis of these infections. For instance, Encephalitozoon cuniculi, Encephalitozoon hellum and Encephalitozoon intestinalis are associated with conjunctivitis or superficial keratopathy in immunocompetent contact lens wearers as well as in immunocompromised patients with disseminated infection (Weiss, Reference Weiss, Mandell, Bennett and Dolin2005). On the other hand, Vittaforma corneae, Microsporidium ceylonesis, Microsporidium africanum, Nosema ocularum are associated with deeper stromal keratitis and occasionally uveitis, and direct traumatic inoculation has been implicated (Weiss, Reference Weiss, Mandell, Bennett and Dolin2005). Identification of ocular microsporidia infection to species level also sets the stage for further epidemiological studies.
To our knowledge, there was no prior data describing the make up of microsporidial species causing ocular infections in our population. From our findings, the sensitivity of this proposed protocol (84·6%) appears to be comparable with previously published data using fresh clinical samples (Joseph et al. Reference Joseph, Sharma, Murthy, Krishna, Garg, Nutheti, Kenneth and Balasubramanian2006). The specificity in our observation (100%) appears to be excellent although, in future, a larger series of patients may better ascertain this result, and thus this method may have a role as a confirmatory tool following microscopy.
One limitation of our study is that only a single species of microsporidia, Vittaforma corneae, was identified from our clinical slides. However, the ATCC strain of E. intestinalis (ATCC 50651), which is another microsporidial species, was also correctly identified using this method. The analytical sensitivity and specificity of the pan-microsporidial primer pair (V1::PMP2) which targets a portion of small subunit rRNA genes of E. cuniculi, E. hellum, E. intestinalis, V. corneae and Enterocytozoon bieneusi (Weiss and Vossbrinck, Reference Weiss, Vossbrinck, Wittner and Weiss1999) to yield a 250–279 bp product (depending on the species) had previously been evaluated by Joseph et al. (Reference Joseph, Sharma, Murthy, Krishna, Garg, Nutheti, Kenneth and Balasubramanian2006). The inclusion of the ‘universal’ primer pair ss18f(V1)::1492r, which also targets the small subunit rRNA gene of many microsporidia (Weiss and Vossbrinck, Reference Weiss, Vossbrinck, Wittner and Weiss1999) but producing much larger amplicons of 1200–1500 bp (depending on the species), did not pick up any additional microsporidia species from our samples. The positive slides which showed no PCR products may be attributed to the initial scant amount of scraping materials present on the slides, thus resulting in insufficient DNA load in these sample extracts. Furthermore, fragmentation of the microsporidial genome from slide fixation, as part of the staining process, may have also resulted in the inability to amplify a PCR product.
Nonetheless, PCR sequencing of DNA extracted from modified trichrome-stained slides for species identification of microsporidia provides an alternative to electron microscopy and does not require additional sampling from patients. The results obtained from PCR and sequencing are also objective and reproducible. As microscopical examination of clinical specimens stained by a modified trichrome method remains the current mainstay method for diagnosis of ocular microsporidia infection, these slides are readily available in many laboratories. Therefore, this technique could be useful for retrospective epidemiological studies with the added advantage that the material is unlikely to be contaminated with HIV.
We would like to thank Dr A. L. Tan and Miss B. S. Low (both from the Department of Pathology, Singapore General Hospital) for their support and assistance.
