Species-Specific Identification of Microsporidia in Stool and Intestinal Biopsy Specimens by the Polymerase Chain Reaction

Article Vol. 16, No. 5 369 Eur. J. Clin. Microbiol. Infect. Dis., 1997, 16:369-376

N.R Kock 1., H. Petersen 2, T. Fenner 2, I. Sobottka 3, C. Schmetz 4, R Deplazes 5, N.J. Pieniazek 6, H. Albrecht 7, J. Schottelius I

In view of the increasing number of cases of human microsporidiosis, simple and rapid methods for clear identification of microsporidian parasites to the species level are required. In the present study, the polymerase chain reaction (PCR) was used for species- specific detection of Encephalitozoon cunicufi, Encephalitozoon hellem, Encephafitozoon (Septata) intestinalis, and Enterocytozoon bieneusi in both tissue and stool. Using stool specimens and intestinal biopsies of patients infected with Enterocytozoon bieneusi (n = 9), Encephalitozoon spp. (n = 2), and Encephafitozoon intestinalis (n = 1) as well as stool spiked with spores of Encephafitozoon cunicufi and Encephalitozoon hellem and tissue cultures of Encephalitozoon cuniculi and Encephalitozoon hellem, three procedures were developed to produce PCR-ready DNA directly from the samples. Specific detection of microsporidian pathogens was achieved in the first PCR. The subsequent nested PCR permitted species determination and verified the first PCR products. With- out exception, the PCR assay confirmed electron microscopic detection of Enterocyto- zoon bieneusi and Encephalitozoon intestinalis in stool specimens and their corresponding biopsies and in spiked stool samples and tissue cultures infected with Encephalito- zoon cuniculi and Encephafitozoon hellem. Moreover, identification of Encephafito- zoon spp. could be specified as Encephalitozoon intestinalis. Whereas standard methods such as light and transmission electron microscopy may lack sensitivity or require more time and special equipment, the PCR procedure described facilitates species- specific identification of microsporidian parasites in stool, biopsies, and, probably, other samples in about five hours.

Microsporidia are ancient eukaryotic, obligate in- tential human pathogens since 1959 (3). As part of tracellular, spore-forming parasites infecting a the evolving pandemic of human immunodeficien- broad range of invertebrates and vertebrates, in- cy virus (HIV) infection, microsporidia have cluding primates (1). First identified in silkworms been identified with increasing frequency as (2), these protozoans have been recognized as po- causative agents of diarrhea and disseminated infections, mainly in AIDS patients (4). Over 400 1 Section of Parasitology, and 4Electron Microscopy Labora- cases of HIV-associated microsporidiosis have tory,Bernard Nocht Institute for TropicalMedicine, Bernard- been documented; however, only a few cases Nocht-Strage 74, D-20359 Hamburg, Germany. have been described among persons not infected 2 Institute for Clinical Pathology and Microbiology, Dres. with HIV (1, 5). Fenner and Partners, Hamburg, Germany. 3 Institute of Microbiologyand Immunology,and 7Department The six genera Encephalitozoon, Enterocyto- of Internal Medicine, University Hospital Eppendorf, Ham- zoon, Nosema (6), PIeistophora (7), Trachipleisto- burg, Germany. phora (8), and Vittaforma (9), as well as unclassi- 5 Institute of Parasitology,University of Zt~rich,Zfirich, Swit- fied microsporidian organisms (referred to by the zerland. 6 Division of Parasitic Diseases, National Center for Infectious collective term Microsporidiurn), have been asso- Diseases, Centers for Disease Control and Prevention,Atlan- ciated with human microsporidiosis. Intestinobil- ta, Georgia, USA. iary infections with Enterocytozoon bieneusi (10) 370 Eur. J. Clin. Microbiol. Infect. Dis.

are the most common microsporidial diseases, at 37~ using M-199 medium with Hanks' salts and L-gluta- but disseminated infections with Encephalito- mine (Gibco BRL, Germany) supplemented with 2 to 5% heat-inactivated fetal bovine serum (Gibco BRL), nonessen- zoon cuniculi (11), Encephalitozoon hellem (12), tial amino acid solution (Sigma, Germany), and penicillin- and Encephalitozoon (Septata) intestinalis (13, streptomycin (Gibco BRL). Spores were harvested by centri- 14) are increasingly being recognized (15-17). fugation (1,500 x g for 20 min) of the cell culture supernatant, Enterocytozoon bieneusi and Encephalitozoon washed three times in 20 mM (pH 7.2) of phosphate-buffered intestinalis have been found in 10 to 40% of HIV- saline (PBS), diluted in PBS to a concentration of 106 infected patients with chronic diarrhea, mal- spores/mI, and stored at 4~ Tissue specimens were harvested with a cell scraper after removing the medium, suspended absorption, and severe weight loss (18, 19). in PBS, centrifuged (600 x g for 10 min), and stored at 4~ Identification of these parasites to the species lev- Uninfected E6 cells were used as a negative control. el is of clinical importance, since epidemiological Stool Specimens. Stool specimens were obtained from HIV- data are limited and few studies have document- positive male patients with acute diarrhea caused by Entero- cytozoon bieneusi (n = 8; 1 patient was coinfected with Cam- ed the response of pathogenic microsporidia to pylobacter jejuni), Encephalitozoon spp. (n = 2), and Enceph- chemotherapeutic agents (17, 20). Despite a varie- alitozoon intestinalis (n = 1) and from an immunocompetent ty of methods for the detection of microsporidian HIV-negative female child infected with both Enterocyto- infections, definitive diagnosis depends upon ultra- zoon bieneusi and Cryptosporidium parvum (5). Microspori- structural examination using the transmission dia were detected by a fluorescent stain with Calcofluor electron microscope (TEM) (4). However, it is not White M2R (32, 33) and a modified chromotrope-based Weber stain (34). Identification to the genus level was achieved always possible to identify the species by TEM. by TEM as described previously (16). Encephalitozoon intes- Furthermore, in specimens such as stool, urine, and tinalis was determined by TEM examination of an intestinal sputum, microsporidian pathogens can only be biopsy. Furthermore, the diagnosis was confirmed by align- identified to the genus level (1). In addition, En- ment of the cloned and partially sequenced SSU rRNA gene cephalitozoon cuniculi and Encephalitozoon hel- to published sequences of this species (35). Soft stool used for spiking with spores (200 spores/g stool) of Encephalitozoon lem are morphologically indistinguishable, even in cuniculi isolate Budejovice, Encephalitozoon cuniculi isolate tissue samples (21). Since light and electron micro- CH-K-4379, and Encephalitozoon hellem isolate CH-H-3, re- scopic techniques may lack sensitivity and the use spectively, was obtained from an asymptomatic person with- of immunodiagnostic tests may be limited by the out microscopically detectable parasites. In addition, this absence of Enterocytozoon bieneusi-specific anti- stool sample was used as a negative control. For determination of the influence of fixatives on DNA isolation and PCR bodies, DNA amplification methods such as the amplification, stool specimens of two patients infected with polymerase chain reaction (PCR) have been de- Enterocytozoon bieneusi and one with Encephalitozoon intes- veloped for the diagnosis of microsporidian infec- tinalis were fixed in 100% ethanol, 4% formalin, 2% glutaral- tions (22-30). Using primers targeted to the small dehyde (Piano, Germany), and 4% paraformaldehyde (Serva, subunit (SSU) rRNA gene of microsporidia, Germany), respectively. All stool samples were stored up to identification of Enterocytozoon bieneusi and one week at 4~ and then either processed immediately or frozen at -20~ until examination. Encephalitozoon intestinalis in intestinal biopsies (23, 24, 28, 30) and of Enterocytozoon bieneusi, En- Intestinal Biopsy Specimens. Biopsy specimens were obtained from the distal duodenum of two patients infected with Entero- cephalitozoon cuniculi, and Encephalitozoon intes- cytozoon bieneusi and one with Encephalitozoon intestinalis tinalis in stool specimens (22, 25, 29, 30) has been by flexible fiberglass endoscopy. One part of the biopsied ma- successfully demonstrated. The present report terial was kept in physiological NaC1 solution; the other parts describes a simple and rapid nested PCR assay for were fixed with 4% formalin and 2% glutaraldehyde, respec- species-specific detection of Enterocytozoon bie- tively. Microsporidia were detected by TEM. The diagnosis of neusi, Encephalitozoon cuniculi, Encephalito- Encephalitozoon intestinalis infection was confirmed as al- ready described. All biopsy specimens were stored up to one zoon hellem, and Encephalitozoon intestinalis in week at 4~ and then either processed immediately or frozen both tissue and stool samples. at -20~ until examination. DNA Isolation. DNA was isolated from biopsy specimens and tissue cultures using the QIAamp Tissue Kit (Qiagen, Germa- Materials and Methods ny) as described previously (23). Stool specimens (250 mg) were thoroughly diluted in 1.5 ml of TE buffer (10 mM of Tris-HC1, 1 mM of EDTA, pH 7.2), centrifuged in an Eppen- Parasite Cultures. The Vero E6 green monkey kidney cell line doff centrifuge 5415C (7,500 x g for 5 min), and resuspended as well as Encephalitozoon cuniculi isolate Budejovice were in 0.5 ml of TE buffer. Washed stool was stored up to one kindly provided by Jiri Lom and Oleg Ditrich of the Institute week at 4~ and then either processed immediately or frozen of Parasitology, Czechoslovak Academy of Sciences, Budejo- at -20~ until DNA isolation. vice, Czech Republic. Encephalitozoon cuniculi isolate CH-K- 4379 of rabbit origin and Encephalitozoon hellem isolate Two methods generally suited to DNA extraction were mod- CH-H-3 were isolated and identified as described previously ified to prepare DNA from washed stool specimens: the GIT- (31). All microsporidian parasites were cultivated in E6 cells Celite method (36) and the QIAamp Blood Kit (Qiagen). Us- Vo1.16,1997 371

ing the GIT-Celite method, 200 Ixl of the stool suspension was (L17072; 266 to 285), Encephalitozoon hellem (L19070; 277 to added to 40 p~l of Celite (Merck, Germany) solution (10 g of 296), and Encephalitozoon intestinalis (U09929; 268 to 287). Celite in 50 ml of distilled water and 0.5 ml of 32% HC1) and Eb379, Ec378, Eh410, and Ei395 were directed against posi- 1 ml of GIT (Roth, Germany) lysis buffer (5 M of guanidine tions 359 to 379, 358 to 378, 390 to 410, and 376 to 395 of SSU thiocyanate, 50 mM of Tris-HC1, 20 mM of EDTA, 1% Triton rRNA sequences of Enterocytozoon bieneusi (L16868), X100). The sample was thoroughly vortexed, incubated in a Encephalitozoon cuniculi (L17072), Encephalitozoon hellem thermomixer at 60~ for about 30 rain, vortexed again, and (L19070), and Encephalitozoon intestinalis (U09929), respec- centrifuged in an Eppendorf centrifuge 5415C (full speed, 1 tively. Sizes of nested PCR fragments varied from 113 bp min). The supernatant was discarded, and the pellet was ( Encephalitozoon cuniculi) to 128 bp (Encephalitozoon intes- washed once with 1 ml of GIT washing buffer (5 M of guani- tinalis) and from 132 bp (Enterocytozoon bieneusi) to 134 bp dine thiocyanate, 50 mM of Tris-HC1) and twice with 1 ml of (Encephalitozoon hellem ). 70% ethanol (Eppendorf centrifuge 5415C; full speed, 1 min). For the first PCR, 20 ~1 of DNA isolate was added to 25 p~l of Afterwards, the pellet was resuspeuded in 1 ml of acetone, 2 x PCR mixture [300 ~1 of dNTP mix (2 mM each; Pharma- centrifuged (full speed, 2 rain), and dried at 60~ for about 10 cia-Biotech, Germany), 300 Ixl of 10 x GeneAmp PCR buffer min. The dried pellet was thoroughly resuspended in 120 ~1 of (Perkin-Elmer, Germany), 12 ~tl of AmpliTaq Gold DNA preheated (60~ 10 mM of Tris buffer (pH 8), incubated in a polymerase (5 U/pA; Perkin-Elmer), 300 rzl of glycerol, and thermomixer at 60~ for 10 min, and centrifuged (full speed, 588 ~1 of distilled water per 1.5 ml], 5 ~1 of 10 x primer solu- 3 rain). Finally, the supernatant was carefully removed and tion (5 ~M of both Mic3U and Mic421U in 10 mM of Tris buf- transferred to a precooled (4~ Eppendorf tube. Using the fer; pH 8), and one drop of mineral oil (Sigma). The reactions QIAamp Blood Kit, 200 ixl of the stool suspension was added were run in a Perkin-Elmer 9600 thermocycler. After initial to 200 p~l of buffer AL and 25 ~1 of proteinase K solution (25 denaturation of the DNA and activation of the AmpliTaq mg of proteinase K in 1.3 ml of distilled water). The sample Gold DNA polymerase (hot start) at 95~ for 10 min, 40 cy- was thoroughly vortexed, incubated in a thermomixer at 70~ cles were performed with denaturation at 94~ annealing at for about 30 min, and centrifuged in an Eppendorf centrifuge 65~ and extension at 72~ (each step, 20 sec). The first PCR 5415C (4,000 x g for 2 min). The supernatant was mixed with was terminated with a 5 min extension at 72~ 210 txl of isopropanol and used for DNA isolation by QIAamp spin columns. All DNA isolates were stored up to For the nested PCR, four reaction mixtures that differed one week at 4~ and then either processed immediately or only with regard to their reverse primer were prepared as fol- frozen at -20~ until PCR amplification. lows: 2 ~1 of the first PCR product was added to 20 ~1 of 1 x PCR mixture [150 ixl of dUNTP mix (2 mM of dATE dCTE Polymerase Chain Reaction Amplification. Primer construction and dGTR 4 mM of dUTP; Pharmacia-Biotech), 150 ~1 of 10 was based on a multiple alignment (MacMolly Tetra software; x GeneAmp PCR buffer (Perkin-Elmer), 6 p~l of AmpliTaq Soft gene, Germany) of published (EMBL Nucleotide Se- DNA polymerase (5 U/pA; Perkin-Elmer), 15 ~1 of HK-UNG quence Database, Germany) SSU rRNA gene sequences of Thermolabile Uracil N-Glycosylase (UNG; 1 U/~I; Biozym (GenBank/EMBL accession numbers Enterocytozoon bieneusi Diagnostik, Germany), 150 ~1 of glycerol, and 1029 pA of dis- L07123 and L16868), Encephalitozoon cuniculi (L07255, tilled water per 1.5 ml] and 1 p~l of 20 x primer solution [10 L17072, Z19563), (L19070), Encephalitozoon hellem Encepha- I~M of each primer (Mic266 and Eb379, Ec378, Eh410, and (L19567, U09929), spp. litozoon intestinalis Encephalitozoon Ei395, respectively) in 10 mM of Tris buffer; pH 8]. The reac- [L16866, L16867; this species was later confirmed as Encepha- tions were run in a Perkin-Elmer 9600 thermocycler without (37)], and litozoon intestinalis Nosema corneum [Vittaforma oil overlay. After digestion of Mic3U and Mic421U (as well as (Ul1046)]. A first PCR with the forward primer corneae of nonspecific amplifications and cross-contaminations by Mic3U (5'-CCA GGU TGA TUC TGC CUG ACG-3') and the other nested PCR fragments) by preincubation with UNG at reserve primer Mic421U (5'-TUA CCG CGG CUG CUG 50~ for 10 min and an initial denaturation of the DNA (com- GCA C-3') was used to amplify a fragment of the SSU rRNA bined with a complete and irreversible inactivation of the gene of all the above-mentioned microsporidia. Mic3U was de- UNG) at 95~ for 2 rain, 35 cycles were performed with designed to be complementary to the conserved 5'-end (positions naturation at 94~ annealing at 62~ and extension at 72~ 3 to 23) of the SSU rRNA gene, which has been used previous- (each step, 20 sec). The nested PCR was terminated with a 5 ly for primer construction (22, 23, 28, 29, 38). Mic421U was tar- min extension at 72~ All PCR products were stored for up geted to positions 394 to 412,403 to 421,417 to 435, and 405 to to one week at 4~ or frozen at -20~ Polymerase chain re- 423 of SSU rRNA sequences of Enterocytozoon bieneusi action products were visualized by agarose gel electrophore- (L16868), (L17072), Encephalitozoon cuniculi Encephalito- sis (FMC Reliant NuSieve 3:1; Biozym Diagnostik) and ethi- (L19070), and zoon hellem Encephalitozoon intestinalis dium bromide (0.5 ~g/ml gel solution) staining. (U09929), respectively. Depending on the species, the size of the PCR products varied from 410 bp (Enterocytozoon bieneu- To avoid cross-contamination during PCR testing, DNA iso- si) to 419 bp (Encephalitozoon cuniculi) and from 421 bp (En- lation, the first and the nested PCR, and gel electrophoresis cephalitozoon intestinalis) to 433 bp (Encephalitozoon hellem). were performed in separate rooms. In addition, positive as well as negative controls were used for all manipulations, and A nested PCR with the forward primer Mic266 (5'-AAG PCR was performed three times to verify all results. In addi- GAG CCT GAG AGA TGG CT-3') and reverse primers tion, dUNTP and UNG were used for all nested PCRs. Eb379 (5'-CAA TTG CTT CAC CCT AAG GTC-3'), Ec378 (5'-GAC CCC TTT GCA CTC GCA CAC-3'), Eh410 (5'-TGC CCT CCA GTA AAT CAC AAC-3'), and Ei395 (5'-CCT CCA ATC AAT CTC GAC TC-3') was used to amplify species-specific fragments of the first PCR products. Results Mic266 was designed to be complementary to a conserved re- gion of the SSU rRNA gene of Enterocytozoon bieneusi Using the described methods (GIT-Celite meth- (L16868; positions 248 to 267), Encephalitozoon cuniculi od, QIAamp Blood Kit, QIAamp Tissue Kit), 372 Eur. J. Clin. Microbiol. Infect. Dis.

Figure 1: Agarose gel electrophoresis of first PCR amplification products from nonfixed stool specimens infected with microsporidia. Lane m, markers; lane 1, stool of a patient with Enterocytozoon bieneusi infection; lane 2, stool spiked with spores of Encephalitozoon cuniculi; lane 3, stool spiked with spores of Encephafitozoon hellem; lane 4, stool of a patient with Encephafitozoon intestinalis infection. Sizes (in base pairs) of markers (pUC18 DNA Msp I Digest; Sigma, Germa- ny) are indicated on the left. Figure 2: Agarose gel electrophoresis of nested PCR amplification products from nonfixed intestinal biopsy specimens and E6 tissue culture samples infected with microsporidia. Lane m, markers; lane 1, biopsy of a patient with Enterocy- DNA was easily isolated from all samples tozoon bieneusi infection; lane 2, tissue culture infected with (washed stool specimens, biopsy specimens, tissue Encephalitozoon cuniculi; lane 3, tissue culture infected cultures). In one case the DNA could only be pre- with Encephalitozoon hellem; lane 4, biopsy of a patient with pared from a stool containing just a few spores of Encephalitozoon intestinalis infection. Primer pairs: b, Mic266/Eb379; c, Mic266/Ec378; h, Mic266/Eh410; i, Enterocytozoon bieneusi by using the GIT-Celite Mic266/Ei395. Sizes (in base pairs) of markers (pUC18 DNA method but not the QIAamp Blood Kit. Thus, the Msp I Digest; Sigma, Germany) are indicated on the left. GIT-Celite method appeared to be more efficient at extracting DNA from stool than the QIAamp Blood Kit. Using the DNA isolates as templates, the first PCR To confirm the nature of nested PCR products, amplification with the primer pair Mic3U/ amplicons were cloned and sequenced and Mic421U resulted in the expected DNA frag- found to be homologous to previously pub- ments of 410 to 433 bp (Figure 1). Although the lished DNA sequences of corresponding micro- products differed in size, species determination sporidian species. Using all published sequences was difficult or impossible. The definite identifica- of the Entrez Document Retrieval System (Na- tion of Enterocytozoon bieneusi, Encephalito- tional Center for Biotechnology Information, zoon cuniculi, Encephalitozoon hellem, and En- USA) and the EMBL Nucleotide Sequence Da- cephalitozoon intestinalis was facilitated by the tabase, specificities of Mic3U for microsporidian nested PCR. This PCR, with the primer pairs parasites, Eb379 for Enterocytozoon bieneusi, Mic266/Eb379, Mic266/Ec378, Mic266/Eh410, Ec378 for Encephalitozoon cuniculi, Eh410 for and Mic266/Ei395, exclusively produced the ex- Encephalitozoon hellem, and Ei395 for Encepha- pected DNA fragments of 113 to 134 bp when the litozoon intestinalis were verified. However, first PCR products of samples infected with En- target sequences of Mic421U and Mic266 were terocytozoon bieneusi, Encephalitozoon cuniculi, found to be widespread in the groups of prokar- Encephalitozoon hellern, and Encephalitozoon yotic and eukaryotic organisms (data not intestinalis, respectively, were used as templates shown). The PCR assay was not influenced by (Figure 2). different origins of DNA isolates (stool speci- Vol. 16, 1997 373

st~l biopsy (fi~edtnot fixed) (fixed/taot fixed) @

1 x washing with TE buffer 15 rain 'I I GTr-Ce~te method QIAamp Tissue KIt 60 min (QIAamp Blood Kit} I,, ] I first PCR with Mtc3U/Mic421U 90 rain I, , nested PCR with: 90 rain

l I I' I Mir M1r Mir Mic266/ tzb379 Er Eh410 Ei395 I l I t I ~ ia~ fPCR lane b lane c lane h lane 60 rain I I 1 I I DNA fragments: r bg 132 bp If3 bp ~ ~ 128 bp I I t I I E, E. E.(S.,~ bter~usi cunicu# her/era in~estiuafis

Figure 4: Scheme for the species-specific identification of pathogenic microsporidia in stool and intestinal biopsy specimens by PCR.

eneusi or Encephalitozoon intestinalis could not be visualized by agarose gel electrophoresis and ethidium bromide staining. Using these products as templates, a second PCR as well as the nested PCR produced the expected DNA fragments on ethidium bromide-stained gels (data not shown). Figure 3: Agarose gel electrophoresis of nested PCR amplification products from fixed stool specimens infected with Encephafitozoon intestinalis. Lane m, markers; lane 1, stool fixed with 4% formalin; lane 2, stool fixed with 2% glutaraldehyde; lane 3, stool fixed with 100% ethanol; lane 4, stool Discussion fixed with 4% paraformaldehyde. Primer pairs: b, Mic266/Eb379; c, Mic266/Ec378; h, Mic266/Eh410; i, Different procedures were found to produce Mic266/Ei395. Sizes (in base pairs) of markers (pUC18 DNA PCR-ready DNA directly from stool and tissue Msp I Digest; Sigma, Germany) are indicated on the left. samples. The GIT-Celite method and the QIAamp Blood Kit as well as the QIAamp Tissue mens, biopsies, and tissue cultures) or by fixation Kit facilitate simple and rapid DNA isolation. of samples (Figure 3). Secondary infections of Whereas the QIAamp Tissue Kit was previously two patients due to Campylobacter jejuni and used for the extraction of microsporidian DNA Cryptosporidium parvum did not reduce the from biopsy specimens (23), this report demon- identification of microsporidia. strates DNA isolation from stool by modifications of the GIT-Celite method and the QIAamp Without exception, first and nested PCR verified Blood Kit for the first time.The first attempt at ex- TEM-confirmed detection of Enterocytozoon bi- traction of microsporidian DNA from stool re- eneusi and Encephalitozoon intestinalis in stool quired incubation times of more than 90 hours us- specimens and biopsies of ten patients as well as ing sodium hypochlorite, two different lyric en- in spiked stool samples and tissue cultures infect- zymes, and mechanical disintegration (22). A ed with two Encephalitozoon cuniculi isolates faster and simpler procedure was developed by and one Encephalitozoon hellem isolate. More- adapting a method for the isolation of Entamoe- over, TEM detection of Encephalitozoon spp. in ba histolytica DNA from stool using a phe- stool specimens of two additional patients could nol:chloroform:isoamyl alcohol extraction (25). be specified as Encephalitozoon intestinalis. However, neither the GIT-Celite method nor the By using the PCR assay described in this study, QlAamp Blood Kit required toxic reagents. In species-specific detection of the above-men- comparison with the GIT-Celite method, the tioned microsporidia in stool specimens and biop- QIAamp Blood Kit appeared to be less efficient sies required about five hours (Figure 4). In some at isolating DNA from stool. In addition, DNA ex- cases the first PCR products from stool specimens traction by the GIT-Celite method is much less ex- infected with a few spores of Enterocytozoon bi- pensive. 374 Eur. J. Clin. Microbiol. Infect. Dis.

The exclusive amplification of a DNA fragment of this study, the PCR assay should also enable spe- microsporidian SSU rRNA gene by the first PCR cies-specific detection of microsporidia in urine and of species-specific DNA fragments by the sediment, nasal secretions, sputum, bronchoalve- nested PCR was demonstrated in this study. The olar lavage fluid, paraffin-embedded tissue, and use of Mic3U/Mic421U facilitates the detection of other samples. parasites belonging to the genera Encephalitozoon It is noteworthy that shedding of microsporidia in and Enterocytozoon and also of species of micro- stool can be intermittent (3, 39). Three instances sporidian genera Arneson, Endoreticulatus, Glugea of double infection with two different types of mi- (1 mismatch of Mic421U), Ichthyosporidium (2 crosporidia in humans have been reported (24, 40, mismatches of Mic421U), Nosema, Pleistophora, 41). Spiking of stool with both Vairirnorpha, Vavraia (1 mismatch of Mic421U), Enterocytozoon bieneusi and Encephalitozoon intestinalis was detect- and Vittaforrna (data not shown). Thus, all micro- ed by the present PCR procedure (data not sporidia known to infect humans as well as a broad shown), whereas a dual infection was not found in range of species not pathogenic in humans could any stool sample or biopsy. be detected by the first PCR. To enhance the sensitivity of the assay, a double first PCR (or a com- Previously reported PCR assays have only facili- bination of the first and the nested PCR) should tated the species-specific detection of Enterocyto- be used for the examination of specimens to rule zoon bieneusi (23, 30) or of Enterocytozoon bie- out false-negative results. Species determination of neusi artd Encephalitozoon intestinalis (24, 29), or Enterocytozoon bieneusi, Encephalitozoon cunicu- they have only been used for stool samples and re- li, Encephalitozoon hellern, and Encephalitozoon quired more than 90 hours (22) or toxic reagents intestinalis can be achieved using the nested PCR (25) for DNA extraction. We have described a sim- as described here. Moreover, nested PCR can ple, rapid, and highly sensitive method for diagno- specifically amplify DNA fragments of other hu- sis of microsporidian infections due to Enterocy- man pathogens as well as fragments of microspo- tozoon bieneusi, Encephalitozoon cuniculi, En- ridia not pathogenic in humans using Mic266 cephalitozoon hellem, or Encephalitozoon (modified in at most 4 bases) and a species- intestinalis in both tissue and stool specimens by specific reverse primer. The PCR assay described nested PCR. The entire procedure requires less here does not require prior culture or concentra- technical equipment than does TEM, needs no tion of the microsporidian spores. Infections toxic reagents, and can be performed in about five were identified by first and nested PCR in soft hours. stool spiked with 200 spores/g. However, the detection limit could vary, depending on the partic- ular stool sample. The nested PCR enhances the sensitivity of the method, as does a second per- Acknowledgement formance of the first PCR. In addition, the authen- ticity of the first PCR products is confirmed We thank Dr. Thomas Weitzel from the Department of Mi- crobiology, General Hospital St. Georg, Hamburg, Germany, without the need for hybridizations or the use of for providing three stool specimens infected with Enterocyto- radionucleotides. zoon bieneusi. Although only Enterocytozoon bieneusi, Enceph- alitozoon cuniculi (23), and Encephalitozoon intestinalis are presently recognized as human enteric References pathogens, intestinal infections with Encephalito- zoon hellern (and other microsporidian parasites) 1. Weber R, Bryan RT, Schwartz DA, Owen RL: Human mi- cannot be excluded. Definite identification of all crosporidial infections. Clinical Microbiology Reviews four species (and of other microsporidia pathogen- 1994, 7: 426-461. ic to humans) in both stool and biopsy specimens 2. N~igeli KW: Uber die neue Krankheit der Seidenraupe und by methods such as the present PCR assay would verwandte Organismen. Botanische Zeitung 1857, 15: provide much-needed epidemiological data. 760-761. Moreover, rapid identification to the species lev- 3. Matsubayashi H, Koike T, Mikata T, Hagiwara S: A case of Encephalitozoon-like body infection in man. Archives el helps physicians determine appropriate treat- of Pathology 1959, 67:181-187. ment, since the microsporidian pathogens appear 4. Schwartz DA, Sobottka I, Leitch G J, Cali A, Visvesvara to differ in their susceptibilities to antimicrobial GS: Pathology of microsporidiosis. Archives of Patholo- agents such as albendazole or fumagillin (1). Al- gy and Laboratory Medicine 1996, 120: 173-188. though only stool and tissue samples were used in 5. Sobottka I, Albrecht H, Schottelius J, Schmetz C, Bent- Vol. 16, 1997 375

reid M, Laufs R, Schwartz DA: Self-limited traveller's di- 18. Asmuth DM, de Girolami PC, Federman M, Ezratty CR, arrhea due to a dual infection with Enterocytozoon bie- Pleskow DK, Desai G, Wanke CA: Clinical features of mi- neusi and Cryptosporidium parvum in an immunocom- crosporidiosis in patients with AIDS. Clinical Infectious petent HIV-negative child. European Journal of Clinical Diseases 1994, 18: 819-825. Microbiology & Infectious Diseases 1995, 14:919-920. 19. Kotler DP: Gastrointestinal manifestations of human im- 6. Margileth AM, Strano A J, Chandra R, Neafie R, Blum M, munodeficiency virus infection. Advances in Internal McCully RM: Disseminated nosematosis in an immuno- Medicine 1995, 40:197-242. logically compromised infant. Archives of Pathology 20. Albrecht H, Stellbrink HJ, Sobottka I: Failure of itracon- 1973, 95: 145-150. azole to prevent Enterocytozoon bieneusi infection. Ge- 7. Ledford DK, Overman MD, Gonzalo A, Call A, Mester W, nitourinary Medicine 1995, 71: 325-326. Lockey RF: Microsporidiosis myositis in a patient with ac- 21. Didier PJ, Didier ES, Orenstein JM, Shadduck JA: Fine quired immunodeficiency syndrome. Annals of Internal structure of a new human microsporidian, Encephalito- Medicine 1985, 102: 628-630. zoon hellem, in culture. Journal of Protozoology 1991,38: 8. Hollister WS, Canning EU, Weidner E, Field AS, Kench 502-507. J, Marriott DJ: Development and ultrastructure of Trachi- 22. Fedorko DP, Nelson NA, Cartwright CP: Identification of pleistophora hominis n.g., n.sp. after in vitro isolation from microsporidia in stool specimens by using PCR and re- an AIDS patient and inoculation into athymic mice. Par- striction endonucleases. Journal of Clinical Microbiolo- asitology 1996, 112:143-154. gy 1995, 33: 1739-1741. 9. Silveira H, Canning EU: Vittaforma corneae n.comb, for 23. Franzen C, MQIler A, Hegener P, Salzberger B, Hartmann the human microsporidium Nosema comeum Shad- P, F&tkenheuer G, Diehl V, Schrappe M: Detection of mi- duck, Meccoli, Davis & Font, 1990, based on its ultra- crosporidia (Enterocytozoon bieneusO in intestinal biop- structure in the liver of experimentally infected athymic sy specimens from human immunodeficiency virus-in- mice. Journal of Eukaryotic Microbiology 1995, 42: fected patients by PCR. Journal of Clinical Microbiolo- 158-165. gy 1995, 33: 2294-2296. 10. Desportes I, Le Charpentier Y, Galian A, Bernard F, Co- 24. Franzen C, Keppers R, Meller A, Salzberger B, chand-Priollet B, Lavergne A, Ravisse P, Modigliani R: F&tkenheuer G, Vetten B, Diehl V, Schrappe M: Genetic Occurrence of a new microsporidian: Enterocytozoon bi- evidence for latent Septata intestinalis infection in human eneusi n.g., n.sp., in the enterocytes of a human patient immunodeficiency virus-infected patients with intestinal with AIDS. Journal of Protozoology 1985, 32: 250-254. microsporidiosis. Journal of Infectious Diseases 1996,. 11. Levaditi C, Nicolau S, Schoen R: L'agent 6tiologique de 173:1038-1040. I'encephalite 6pizootique du lapin (Encephalitozoon cu- 25. Katzwinkel-Wladarsch S, Lieb M, Heise W, L6scher T, nicuh). Comptes Rendus des S~ances de la Soci~te de Rinder H: Direct amplification and species determination Biologie et de ses Filiales 1923, 89: 984-986. of microsporidian DNA from stool specimens. Tropical 12. Didier ES, Didier PJ, Friedberg DN, Stenson SM, Oren- Medicine and International Health 1996, 1: 373-378. stein JM, Yee RW, Tio FO, Davis RM, Vossbrinck CR, Mil- 26. Visvesvara GS, Leitch GL, Da Silva A J, Croppo GP, Mou- lichamp N, Shadduck JA: Isolation and characterization ra H, Wallace S, Slemenda SB, Schwartz DA, Moss D, of a new human microsporidian, Encephalitozoon hellem Bryan RT, Pieniazek NJ: Polyclonal and monoclonal an- (n.sp.), from three AIDS patients with keratoconjunctivi- tibody and PCR-amplified small-subunit rRNA identifi- tis. Journal of Infectious Diseases 1991,163:617-621. cation of a microsporidian, Encephafitozoon hellem, 13. Call A, Kotler DP, Orenstein JM: Septata intestinalis n.g., isolated from an AIDS patient with disseminated infec- n.sp., an intestinal microsporidian associated with tion. Journal of Clinical Microbiology 1994, 32: chronic diarrhea and dissemination in AIDS patients. 2760-2768. Journal of Eukaryotic Microbiology 1993, 40:101-112. 27. Weiss LM, Zhu X, Call A, Tanowitz HB, Wittner M: Utili- 14. Hartskeerl RA, van Gool T, Schuitema ARJ, Didier ES, ty of microsporidian rRNA in diagnosis and phylogeny: Terpstra W J: Genetic and immunological characterization a review. Folia Parasitologica 1994, 41: 81-90. of the microsporidian Septata intestinalis Call, Kotler and 28. Zhu X, Wittner M, Tanowitz HB, Kotler D, Call A, Weiss Orenstein, 1993: reclassification to Encephafitozoon in- LM: Small subunit rRNA sequence of Enterocytozoon bi- testinalis. Parasitology 1995, 110: 277-285. eneusi and its potential diagnostic role with use of the 15. Bryan RT: Microsporidiosis as an AIDS-related opportu- polymerase chain reaction. Journal of Infectious Diseas- nistic infection. Clinical Infectious Diseases 1995, 21, es 1993, 168: 1570-1575. Supplement 1: 62-65. 29. Ombrouck C, Ciceron L, Biligui S, Brown S, Marechal P, 16. Schottelius J, Lo Y, Schmetz C: Septata intestinalis and van Gool T, Datry A, Danis M, Desportes-Livage I: Spe- Encephalitozoon cuniculi: cross-reactivity between two cific PCR assay for direct detection of intestinal micro- microsporidian species. Folia Parasitologica 1995, 42: sporidia Enterocytozoon bieneusi and Encephalitozoon 169-172. intestinalis in fecal specimens from human immunodef- 17. Sobottka I, Albrecht H, Schfifer H, Schottelius J, Visves- iciency virus-infected patients. Journal of Clinical Micro- vara GS, Laufs R, Schwartz DA: Disseminated Encepha- biology 1997, 35: 652-655. litozoon (Septata) intestinalis infection in a patient with 30. Vel&squez JN, Carnevale S, Guarnera EA, Labbe JH, AIDS: novel diagnostic approaches and autopsy-con- Chertcoff A, Cabrera MG, Rodriguez MI: Detection of the firmed parasitologicat cure following treatment with al- microsporidian parasite Enterocytozoon bieneusi in bendazole. Journal of Clinical Microbiology 1995, 33: specimens from patients with AIDS by PCR. Journal of 2948-2952. Clinical Microbiology 1996, 34: 3230-3232. 376 Eur. J. Clin. Microbiol. Infect. Dis.

31. Deplazes P, Mathis A, Baumgartner R, Tanner I, Weber heim-Van Dillen PME, van der Noordaa J: Rapid and sim- R: Immunologic and molecular characteristics of En- ple method for purification of nucleic acids. Journal of cephafitozoon-like microsporidia isolated from humans Clinical Microbiology 1990, 28: 495-503. and rabbits indicate that Encephafitozoon cunicufi is a 37. Vossbrinck CR, Baker MD, Didier ES, Debrunner- zoonotic parasite. Clinical Infectious Diseases 1996, 22: Vossbrinck BA, Shadduck JA: Ribosomal DNA se- 557-559. quences of Encephalitozoon hellem and Encephalitozoon 32. Van Gool T, Canning EU, Dankert J: An improved prac- cuniculi: species identification and phylogenetic con- tical and sensitive technique for the detection of micro- struction. Journal of Eukaryotic Microbiology 1993, 40: sporidian spores in stool samples. Transactions of the 354-362. Royal Society of Tropical Medicine and Hygiene 1994, 38. Baker MD, Vossbrinck CR, Didier ES, Maddox JV, 88: 189-190. Shadduck JA: Small subunit ribosomal DNA phylogeny 33. Vavra J, Dahbiova R, Hollister WS, Canning EU: Stain- of various microsporidia with emphasis on AIDS related ing of microsporidian spores by optical brighteners with forms. Journal of Eukaryotic Microbiology 1995, 42: remarks on the use of brighteners for the diagnosis of 564-570. AIDS-associated human microsporidioses. Folia Parasit- 39. Weitz JC, Bothelo R, Bryan RT: Microsporidiosis en pa- ologica 1993, 40: 267-272. cientes con diarrea cronica y SlDA, HIV asintomaticos y 34. Weber R, Bryan RT, Owen RL, Wilcox CM, Gorelkin L, pacientes con diarrea aguda. Revista Medica de Chile Visvesvara GS: Improved light-microscopical detection 1995, 123: 849-856. of microsporidia spores in stool and duodenal aspirates. 40. Blanshard C: Hollister WS, Peacock CS, Tovey DG, El- New England Journal of Medicine 1992, 326: 161-166. lis DS, Canning EU, Gazzard BG: Simultaneous infection 35. Visvesvara GS, Da Silva A J, Croppo GP, Pieniazek NJ, with two types of intestinal microsporidia in a patient with Leitch GJ, Ferguson D, de Moura H, Wallace S, Slemen- AIDS. Gut 1992, 33: 418-420. da SB, Tyrrell I, Moore DF, Meador J: In vitro culture and 41. Van Gool T, Canning EU, Gills H, van den Bergh- serologic and molecular identification of Septata intes- Weerman MA, Eeftinck-Schattenkerk JK, Dankert J: tinalis isolated from urine of a patient with AIDS. Jour- Septata intestinalis frequently isolated from stool of nal of Clinical Microbiology 1995, 33: 930-936. AIDS patients with a new cultivation method. Parasitol- 36. Boom R, Sol CJA, Salimans MMM, Jansen CL, Wert- ogy 1994, 109: 281-289.