Simultaneous Detection of Bovine Theileria and Babesia ... · Pharmacology, University of Granada, Campus Universitario de Cartuja, 18071 Granada, Spain2 Received 11 November 1998/Returned

JOURNAL OF CLINICAL MICROBIOLOGY,0095-1137/99/$04.0010

June 1999, p. 1782–1789 Vol. 37, No. 6

Copyright © 1999, American Society for Microbiology. All Rights Reserved.

Simultaneous Detection of Bovine Theileria and BabesiaSpecies by Reverse Line Blot Hybridization

J. M. GUBBELS,1 A. P. DE VOS,1 M. VAN DER WEIDE,1† J. VISERAS,2

L. M. SCHOULS,3 E. DE VRIES,1 AND F. JONGEJAN1*

Department of Parasitology and Tropical Veterinary Medicine, Faculty of Veterinary Medicine, Utrecht University,3508 TD Utrecht,1 and Research Laboratory for Infectious Diseases, National Institute of Public Health

and Environment, 3720 BA Bilthoven,3 The Netherlands, and Department of Parasitology, Faculty ofPharmacology, University of Granada, Campus Universitario de Cartuja, 18071 Granada, Spain2

Received 11 November 1998/Returned for modification 22 January 1999/Accepted 15 March 1999

A reverse line blot (RLB) assay was developed for the identification of cattle carrying different species ofTheileria and Babesia simultaneously. We included Theileria annulata, T. parva, T. mutans, T. taurotragi, andT. velifera in the assay, as well as parasites belonging to the T. sergenti-T. buffeli-T. orientalis group. The Babesiaspecies included were Babesia bovis, B. bigemina, and B. divergens. The assay employs one set of primers forspecific amplification of the rRNA gene V4 hypervariable regions of all Theileria and Babesia species. PCRproducts obtained from blood samples were hybridized to a membrane onto which nine species-specificoligonucleotides were covalently linked. Cross-reactions were not observed between any of the tested species.No DNA sequences from Bos taurus or other hemoparasites (Trypanosoma species, Cowdria ruminantium,Anaplasma marginale, and Ehrlichia species) were amplified. The sensitivity of the assay was determined at0.000001% parasitemia, enabling detection of the carrier state of most parasites. Mixed DNAs from fivedifferent parasites were correctly identified. Moreover, blood samples from cattle experimentally infected withtwo different parasites reacted only with the corresponding species-specific oligonucleotides. Finally, RLB wasused to screen blood samples collected from carrier cattle in two regions of Spain. T. annulata, T. orientalis, andB. bigemina were identified in these samples. In conclusion, the RLB is a versatile technique for simultaneousdetection of all bovine tick-borne protozoan parasites. We recommend its use for integrated epidemiologicalmonitoring of tick-borne disease, since RLB can also be used for screening ticks and can easily be expandedto include additional hemoparasite species.

Tick-borne protozoan diseases (e.g., theileriosis and babesi-osis) pose important problems for the health and managementof domestic cattle in the tropics and subtropics (24). The mostwidespread and malignant Theileria species is Theileria annu-lata, causing tropical theileriosis, which occurs around theMediterranean basin, in the Middle East, and in SouthernAsia. The other malignant Theileria species is T. parva, whichoccurs in East and Southern Africa and causes East Coastfever. Bovine babesiosis is caused by Babesia bovis and B.bigemina, both of which occur worldwide in tropical and sub-tropical regions. B. divergens occurs in cattle in Europe andextends into North Africa (4). In addition, benign forms oftheileriosis are caused by T. mutans, T. velifera, and T. tauro-tragi, which are mainly located in Africa, whereas parasites ofthe T. sergenti-T. buffeli-T. orientalis group, referred to below asthe T. orientalis group, occur worldwide (45).

Several techniques for detection of these hemoparasiteshave been developed separately for each species (reviewed inreference 17). For the T. orientalis group, several sensitive PCRassays based on the gene encoding the major piroplasm surfaceprotein have been developed (27, 42). A similar approachusing the Tams1 gene was followed for T. annulata (13). PCRdetection of T. parva using repetitive DNA sequences or gene-specific primers has also been developed (2). Sensitive PCR

methods are available for B. bovis (6) and B. bigemina (15) butnot yet for B. divergens. Assays to differentiate Theileria specieson the basis of their rRNA genes have been described byAllsopp et al. (1) and Bishop et al. (3), but these assays did notinclude all Theileria species. Moreover, these assays were de-veloped merely for the differentiation of species rather than fordetection in carrier animals. To date, screening for the pres-ence of benign Theileria species, such as T. mutans, T. velifera,and T. taurotragi, occurring in cattle depends on the immuno-fluorescent antibody test and may give rise to cross-reactions(36). Moreover, since serodiagnosis does not detect the para-site itself, the animal may have already cleared the pathogenbut remained seropositive. The first integrated approach tosensitive detection and differentiation of B. bigemina, B. bovis,and Anaplasma marginale in a 16S–18S rRNA gene multiplexPCR was reported by Figueroa et al. (16).

Despite the usefulness of these assays, it would be verypractical to have a universal test to simultaneously detect anddifferentiate all protozoan parasites that could possibly bepresent in the blood of carrier cattle. A reverse line blot (RLB)technique fulfilling these criteria has recently been developedto detect four different Borrelia species in ticks (39). RLB wasinitially developed as a reverse dot blot assay for the diagnosisof sickle cell anemia (40), but the essence of both techniques isthe hybridization of PCR products to specific probes immobi-lized on a membrane in order to identify differences in theamplified sequences. In the “line” approach, multiple samplescan be analyzed against multiple probes to enable simulta-neous detection. This approach was initially developed for theidentification of Streptococcus serotypes (26), followed by anRLB for Mycobacterium tuberculosis strain differentiation (25).

* Corresponding author. Mailing address: Department of Parasitol-ogy and Tropical Veterinary Medicine, Faculty of Veterinary Medi-cine, Utrecht University, P.O. Box 80.165, 3508 TD Utrecht, TheNetherlands. Phone: (31-30) 2532568. Fax: (31-30) 2540784. E-mail:[email protected].

† Present address: Innogenetics, 9052 Ghent, Belgium.

1782

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

In this study we used RLB to detect and differentiate allknown Theileria and Babesia species of importance in cattle inthe tropics and subtropics on the basis of their differences in18S small-subunit (SSU) rRNA gene sequences (Fig. 1). Theconserved domains of the 18S rRNA genes of Theileria andBabesia species were used to amplify the hypervariable V4region by PCR. Within this region, oligonucleotides were de-duced for species-specific detection. For most species one ormore sequences were available from GenBank, whereas the18S rRNA gene of T. velifera needed to be cloned and se-quenced. Since the 18S rRNA gene is reported to be variablein the T. orientalis group (8) and also in T. mutans (1a), theirspecies-specific oligonucleotides were deduced after cloningand sequencing of the 18S rRNA gene of two isolates of eachspecies. The six Theileria species included in the assay are T.annulata, T. parva, T. mutans, T. taurotragi, T. velifera, and theT. orientalis group. The three Babesia species included are B.bovis, B. bigemina, and B. divergens. A catchall Theileria andBabesia species control oligonucleotide is also included.

MATERIALS AND METHODS

GenBank accession numbers of sequences used. The GenBank accessionnumbers of the 18S sequences used for deducing PCR oligonucleotides andspecific RLB oligonucleotides are as follows: for T. annulata, M64243; for T.parva, L02366 and AF013418; for T. taurotragi, L19082; for T. orientalis, U97047(type A), U97048 (type B), U97049 (type B1), U97051 (type C), U97052 (typeD), U97053 (type E), U97050 (type H), AB000274 (Medon, Indonesia),AB000273 (Ipoh, Malaysia), AB000272 (Warwick, Australia), AB000271 (Ikeda,Japan), and Z15106 (Marula, Kenya); for B. bigemina, X59604 (gene A), X59605(gene B), and X59607 (gene C); for B. bovis L31922, L19077, L19078, andM87566; and for B. divergens, U07885, U16370, and Z48751.

Parasite stocks. Parasite stocks used in this study and previously described arelisted in Table 1, whereas those not previously described are given here. Trypano-soma congolense was isolated in Harare, Zimbabwe, from a dog in 1986. B.divergens was isolated from Bos taurus cattle in 1969 in Putten, The Netherlands,and B. major was also isolated in The Netherlands, on the island of Ameland, in1977 from Haemaphysalis punctata ticks. B. bovis was received from E. Pipano inIsrael as vaccine strain C61411. The B. bovis E strain was isolated from cattle inAustralia and received in 1988.

Experimental infections. Experimental infections with one or more parasitestocks were conducted in female Friesian holstein calves of 8 months of age.Monitoring and blood sample storage were performed essentially as describedpreviously (13). Daily rectal temperatures were recorded, and Giemsa-stainedblood smears and lymph node biopsy smears were examined. Serum samples forserodiagnosis and citrate-blood samples for PCR were stored at 220°C. Sixanimals were initially infected with 1 ml of a ground-up tick supernatant (GUTS)stabilate of T. annulata (India). The animals received a second infection 8 weekslater with 2-ml frozen blood stabilates of either T. mutans (Nigeria), T. orientalis(Japan), T. orientalis (Australia), B. bovis (E strain, Australia), B. bigemina(Nigeria), or B. divergens (The Netherlands). The B. bovis infection was treatedwith Imidocarb at day 7 postinfection (p.i.).

Processing of samples for PCR. The DNAs of T. taurotragi, T. parva, T. velifera,T. mutans, B. bovis, B. bigemina, B. divergens, B. major, and all T. orientalis stocks,except the stocks from China, were extracted from EDTA-blood samples thathad been stored in liquid nitrogen. Samples of the T. orientalis stocks from Chinawere prepared from purified merozoites. Samples of T. annulata from Turkey,Portugal, Mauritania, and India were prepared from citrate-buffered blood sam-ples collected from experimentally infected calves. Samples were processed aspreviously described (13). Briefly, 200-ml samples were washed three times with0.5 ml of lysis mixture (0.22% NaCl, 1 mM EDTA, 0.015% saponin) by centrif-ugation at maximum speed for 5 min, resuspended in 100 ml of PCR mixture (10mM Tris-HCl [pH 8.0], 50 mM KCl, 0.5% Tween 20, 100 mg of proteinase K/ml),and incubated overnight at 56°C. Prior to PCR, samples were heated for 10 minat 100°C and centrifuged for 2 min.

Blood samples from cattle in Spain. Heparin-blood samples were collected inthe province of Toledo, central Spain, from 28 Charolais cattle (male and fe-male) ranging in age from 1 month to 7 years. The samples required a proteinaseK treatment overnight at 72°C in order to inactivate PCR-inhibiting components.

FIG. 1. Locations of species-specific oligonucleotides (shaded) in the 18S rRNA V4 hypervariable region. The first T. orientalis isolate listed is type D (underlinedA at position 668), and the second is the T. orientalis isolate described as T. buffeli Warwick (double-underlined T at position 668). A mix of these two oligonucleotideswas used on the blot. The melting temperature (Tm) of each oligonucleotide is indicated. The ratio of each oligonucleotide to the catchall Theileria and Babesiaoligonucleotide (a ratio of 1 correspond to 200 pmol) is also indicated. The catchall Theileria and Babesia oligonucleotide is identical for all 10 sequences. The R atposition 871 in the T. annulata sequence denotes either an A or a G.

VOL. 37, 1999 RLB ASSAY FOR BOVINE THEILERIA AND BABESIA SPECIES 1783

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

EDTA-blood samples were collected in the province of Cadiz, in the south ofSpain, from 17 female Charolais-Retinta crossbred cattle. Their ages rangedfrom 6 months to 3 years. EDTA-blood samples were processed according tostandard procedures as described above.

Cloning and sequencing of 18S genes. A PCR product of approximately 1,750bp was amplified with primer A and primer B, amplifying eukaryotic 18S genes,as described previously (32). The product was directly cloned into the pCR2.1vector according to the manufacturer’s instructions (Invitrogen, Leek, The Nether-lands). Clones from four independent PCRs were sequenced with a PharmaciaBiotech (Uppsala, Sweden) T7 sequencing kit. The sequences were analyzed byalignment using the Multalin program (10). Only the hypervariable V4 regions ofboth T. mutans isolates and both Chinese T. orientalis isolates were sequenced onboth strands, whereas the full-length sequence of T. velifera was determined.

PCR. One set of primers was used to amplify a 460- to 520-bp fragment of the18S SSU rRNA spanning the V4 region. The forward primer, RLB-F (59-GAGGTAGTGACAAGAAATAACAATA-39), and the reverse primer, RLB-R(biotin-59-TCTTCGATCCCCTAACTTTC-39), hybridized with regions con-served for Theileria and Babesia. RLB-F corresponds to nucleotides 437 to 461,and RLB-R corresponds to nucleotides 920 to 939, of the T. annulata 18S rRNAgene sequence (accession no. M64243). Primers were obtained from Isogen(Maarssen, The Netherlands). Reaction conditions in a 100-ml volume were asfollows; 13 PCR buffer (Promega, Madison, Wis.), 1.5 mM MgCl2 (Promega),200 mM each deoxynucleoside triphosphate (Pharmacia Biotech), 2.5 U of Taqpolymerase (Promega), 100 pmol of each primer, and 20 ml of purified DNAsample. The reactions were performed in an automated DNA thermal cycler(Perkin-Elmer, Foster City, Calif.) for 40 cycles. Each cycle consisted of adenaturing step of 1 min at 94°C, an annealing step of 1 min at 50°C, and anextension step of 1.5 min at 72°C. A final extension step of 10 min at 72°Ccompleted the program.

RLB hybridization. All the specific oligonucleotides shown in Fig. 1 containedan N-terminal N-(trifluoracetamidohexyl-cyanoethyl,N,N-diisopropyl phos-phoramidite [TFA])-C6 amino linker (Isogen). The quality of the TFA linker iscrucial and varies among the different oligonucleotide-producing companies. ABiodyne C blotting membrane (Pall Biosupport, Ann Arbor, Mich.) was acti-vated by a 10-min incubation in 10 ml of 16% (wt/vol) 1-ethyl-3-(3-dimethyl-amino-propyl)carbodiimide (EDAC) (Sigma, St. Louis, Mo.) at room tempera-ture. The membrane was washed for 2 min with distilled water and placed in an

MN45 miniblotter (Immunetics, Cambridge, Mass.). Specific oligonucleotideswere diluted to a 200- to 1,600-pmol/150 ml concentration in 500 mM NaHCO3(pH 8.4) and were subsequently covalently linked to the membrane with theamino linker by filling the miniblotter slots with the oligonucleotide dilutions;they were then incubated for 1 min at room temperature. The oligonucleotidesolutions were aspirated, and the membrane was inactivated by incubation in 100ml of a 100 mM NaOH solution for 10 min at room temperature. The membranewas washed with shaking in 125 ml of 23 SSPE–0.1% sodium dodecyl sulfate(SDS) for 5 min at 60°C (203 SSPE contains 360 mM NaCl, 20 mM NaH2PO4,and 2 mM EDTA [pH 7.4]). Before use, the membrane was washed for 5 min at42°C with 125 ml of 23 SSPE–0.1% SDS and placed in the miniblotter with theslots perpendicular on the previously applied specific oligonucleotides. A volumeof 40 ml of PCR product was diluted to an end volume of 150 ml of 23SSPE–0.1% SDS, heated for 10 min at 100°C, and cooled on ice immediately.Denatured PCR samples were applied into the slots and incubated for 60 min at42°C. PCR products were aspirated, and the blot was washed twice in 125 ml of23 SSPE–0.5% SDS for 10 min at 42°C with shaking. Subsequently the mem-brane was incubated in 10 ml of 1:4,000-diluted peroxidase-labeled streptavidin(Boehringer, Mannheim, Germany) in 23 SSPE–0.5% SDS for 30 min at 42°C.The membrane was washed twice in 125 ml of 23 SSPE–0.5% SDS for 10 minat 42°C with shaking. After two rinses in 125 ml of 23 SSPE for 5 min each timeat room temperature, an incubation for 1 min in 10 ml of ECL detection fluid(Amersham, Little Chalfont, Buckinghamshire, United Kingdom) followed be-fore exposure to an ECL hyperfilm (Amersham) for 10 to 60 min. After use, allPCR products were stripped from the membrane by two washes in 1% SDS for30 min each time at 80°C. The membrane was rinsed in 20 mM EDTA (pH 8.0)and stored in fresh EDTA solution at 4°C for reuse.

Sensitivity of the RLB assay. To determine the detection limit of the RLBassay, citrate-buffered blood obtained from an animal infected with T. annulata(India) with a parasitemia of 3.9% was serially diluted with blood of a nonin-fected animal and processed as described above. In order to establish the exactlevel of parasitemia in the diluted samples, the packed cell volume of bothinfected and uninfected blood was measured.

Possible competition between T. annulata and B. bovis in the RLB assay wasdetermined by addition of serial dilutions of T. annulata DNA (0, 0.1, 1.0, 10,100, 1,000, and 10,000 pg) to a fixed amount of 500 fg of B. bovis DNA. As

TABLE 1. Origin and nature of parasite stocks

Genus and speciesStock origin

Reference MaterialCountry Location or name

Theileria annulata Turkey Ankara 41 BloodSpain Caceres 11 CultureIndia Hissar 18 BloodMauritania Nouakchott 21 BloodSudan Soba 33 CulturePortugal Evora 13 BloodBahrain Bahrain 49 Culture

Theileria parva Kenya Muguga 5 BloodTanzania Pugu 1 46 Blood

Theileria taurotragi Zimbabwe McIlwaine 46 GUTSTheileria velifera Tanzania Lugurni 43 BloodTheileria mutans Nigeria Katsina 37 Blood

Tanzania Pugu 44 BloodTheileria orientalis Australia Brisbane 48 Blood

England Essex 34 BloodIran Teheran 47 BloodJapan Fukushima 20 BloodSouth Korea Jeju 48 BloodUnited States Texas 29 BloodChina Unknown 22 BloodChina North-West 38 Blood

Babesia bigemina Nigeria Runka 31 BloodBabesia bovis Israel C61411 Culture

Australia E-strain BloodBabesia divergens The Netherlands Putten BloodBabesia major The Netherlands Ameland BloodTrypanosoma vivax Nigeria Yakawanda 30 BloodTrypanosoma congolense Zimbabwe Harare BloodTrypanosoma brucei Ivory Coast TH114/78E 14 BloodCowdria ruminantium Senegal Niaye 23 CultureAnaplasma marginale Nigeria Zaria 12 BloodEhrlichia canis United States Florida 35 Blood

1784 GUBBELS ET AL. J. CLIN. MICROBIOL.

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

controls, the same dilutions of T. annulata DNA in the absence of B. bovis DNAwere analyzed.

Nucleotide sequence accession numbers. The 18S sequences determined inthis study for T. velifera and T. orientalis (northwest China) have been depositedin GenBank under accession no. AF097993 and AF036336, respectively.

RESULTS

18S sequences. The 18S rRNA gene of T. velifera was PCRamplified with general 18S primers, cloned, and sequenced,and its relation with the other Theileria species was determined(data not shown). Possible variation of 18S rRNA gene se-quences within T. mutans was determined by sequencing the18S V4 hypervariable regions of two different isolates (seeTable 1). Both isolates contained sequences similar to thepartial T. mutans 18S sequence previously reported (1). Vari-

ation in 18S between different isolates of the T. orientalis groupwas confirmed by sequencing two isolates from China (seeTable 1). The isolate from northwestern China was identical tothe formerly described type D, and the other isolate fromChina, originally named T. sergenti, was identical to the for-merly described type A (8).

RLB specific oligonucleotides. For each species a specificoligonucleotide was deduced in the amplified V4 region, witha melting temperature between 55.0 and 58.8°C, enabling si-multaneous hybridization (Fig. 1). The B. bovis oligonucleotidehad a slightly higher melting temperature (63.2°C), since anoligonucleotide with a higher melting temperature was re-quired for adequate detection of this parasite. As a control, acatchall Theileria and Babesia species oligonucleotide was alsodesigned in a similar way (Fig. 1). This oligonucleotide was alsoincluded in case a PCR product would not hybridize with anyof the species included, indicating the presence of an unknownor unexpected species. Since high 18S rRNA gene variationoccurs in the T. orientalis group (8), the specific primer forthese parasites is a mixture containing either an A or a T at

FIG. 2. RLB of PCR products obtained from Theileria, Babesia, and other bovine hemoparasite samples. PCR products are applied in vertical lanes. Species-specificoligonucleotides are applied in horizontal rows. Lanes 1 to 7, T. annulata isolates from Turkey (lane 1), India (lane 2), Spain (lane 3), Mauritania (lane 4), Sudan (lane5), Portugal (lane 6), and Bahrain (lane 7); lanes 8 and 9, T. parva isolates from Tanzania and Kenya, respectively; lanes 10 and 11, T. mutans isolates from Tanzaniaand Nigeria, respectively; lane 12, T. taurotragi; lane 13, T. velifera; lanes 14 to 21, T. orientalis isolates from Texas (lane 14), Australia (lane 15), Japan (lane 16), England(lane 17), northwest China (lane 18), China (lane 19), Iran (lane 20), and Korea (lane 21); lane 22, B. bovis; lane 23, B. bigemina; lane 24, B. divergens; lane 25, B. major;lane 26, bovine DNA; lane 27, Trypanosoma congolense; lane 28, Trypanosoma vivax; lane 29, Trypanosoma brucei; lane 30, A. marginale; lane 31, Cowdria ruminantium;lane 32, Ehrlichia canis. Rows: 1, B. divergens; 2, B. bigemina; 3, B. bovis; 4, T. orientalis; 5, T. velifera; 6, T. taurotragi; 7, T. mutans; 8, T. parva; 9, T. annulata; 10, catchallTheileria and Babesia control oligonucleotide.

FIG. 3. Sensitivity of the PCR and the RLB assay. (A) PCR using RLB-F andRLB-R primers on DNA extracted from serial dilutions of T. annulata-infectedblood. Lanes 1 and 13, molecular size markers; lanes 2 to 9, PCR productsderived from DNA extracted from serial dilutions representing 1021, 1022, 1023,1024, 1025, 1026, 1027, and 1028% parasitemia, respectively; lane 10, 0% para-sitemia; lane 11, distilled water control (2); lane 12, PCR positive control of 10ng of genomic DNA of T. annulata (1). (B) Corresponding RLB of PCRproducts derived from DNA extracted from serial dilutions of T. annulata-infected blood. Lanes 2 to 12 are identical to lanes 2 to 12 in panel A; lanes 1 and13 are empty. Rows 1 and 2, T. annulata-specific and catchall Theileria andBabesia control oligonucleotides, respectively.

FIG. 4. Competition between B. bovis and T. annulata DNAs in the RLB-PCR. Lanes 1 to 7 each contain 500 fg of B. bovis DNA, as well as T. annulataDNA in the following amounts: 0 pg (lane 1), 0.1 pg (lane 2), 1.0 pg (lane 3), 10pg (lane 4), 0.1 ng (lane 5), 1.0 ng (lane 6), and 10 ng (lane 7). Lane 8 is empty(2). Lanes 9 to 15 contain the amounts of T. annulata DNA present in lanes 1to 7, respectively, but no B. bovis DNA. Rows are identical to those in Fig. 2.

VOL. 37, 1999 RLB ASSAY FOR BOVINE THEILERIA AND BABESIA SPECIES 1785

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

position 13, corresponding with position 658 in the alignment(Fig. 1). The location of the oligonucleotide was conserved inall members of this group and could differentiate them from allthe other parasite species (Fig. 2). For each oligonucleotidethe relative reaction was optimized by varying the amount ofthe oligonucleotide applied on the membrane in such a waythat all oligonucleotides resulted in equally intense signalsrelative to the catchall Theileria and Babesia control oligonu-cleotide. These amounts are indicated in Fig. 1.

RLB. RLB-PCR was performed on all Theileria and Babesiaspecies listed in Table 1 except the B. bovis E strain from

Australia. PCR products were hybridized to the membraneand were shown to bind with the specific oligonucleotides only(Fig. 2). Seven T. annulata isolates, two T. parva isolates, twoT. mutans isolates, and eight T. orientalis isolates were detected(Fig. 2). Furthermore, in Fig. 2, row 10, the catchall Theileriaand Babesia control oligonucleotide reacted with all parasiteisolates. Each species isolate was recognized by its correspond-ing oligonucleotide only, and cross-reactions between specieswere not observed.

RLB-PCR performed on DNA from Trypanosoma spp.,DNA from Ehrlichia spp., and bovine DNA did not yield de-tectable products on agarose gels (data not shown). Whenthese PCR products were applied onto the RLB membrane(Fig. 2, lanes 26 to 32), no reaction with any oligonucleotidewas observed. However, the presence of DNA in these sampleswas demonstrated by successful amplification using specificPCR assays for each Trypanosoma species (9) as well as for theehrlichial organisms (28) (data not shown).

RLB sensitivity. After an acute Theileria or Babesia infec-tion, parasitemias usually drop to a very low level and theanimal becomes a carrier. The sensitivity of the RLB assay todetect such carrier animals was determined by using infectedblood from an animal with a known level of T. annulata para-sitemia serially diluted with blood from a noninfected animal.The resulting PCR is shown in Fig. 3A, wherein a signal can bediscerned down to a parasitemia level of 1023%. The corre-sponding RLB is shown in Fig. 3B, wherein the lowest detect-able parasitemia is 1026%, corresponding to 3 parasites per mlof blood.

Blood samples containing two parasite species each werealso tested. It can be imagined that the predominance of oneparasite can influence the results of the RLB-PCR, since both

FIG. 5. RLB of mixed DNA samples from parasites that occur in the sameregion. Lane 1, distilled water control (neg.); lanes 2 through 10, as labeled; lane12, mix I, containing T. annulata, T. orientalis, B. bovis, B. bigemina, and B.divergens; lane 14, mix II, containing T. parva, T. mutans, T. taurotragi, T. velifera,and T. orientalis. Lanes 11 and 13 are empty (2). Rows are identical to those inFig. 2.

FIG. 6. RLB with PCR products derived from DNA extracted from blood samples from animals infected with two different parasites as indicated. Lane 1,preinfection/day of T. annulata infection (marked by arrow); lane 2, 2 weeks post-T. annulata infection; lane 3, 12 weeks post-T. annulata infection/day of infection withsecond parasite (marked by arrow); lane 4, 1 week post-second infection; lane 5, 2 weeks post-second infection; lane 6, 3 weeks post-second infection; lane 7, 4 weekspost-second infection; lane 8, 5 weeks post-second infection; lane 9, 6 weeks post-second infection; lane 10, 7 weeks post-second infection. Rows are identical to thosein Fig. 2.

1786 GUBBELS ET AL. J. CLIN. MICROBIOL.

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

parasites may compete for the available primers in the PCR.An experiment was designed wherein purified genomic T. annu-lata DNA was added in serial dilution to an amount of purifiedgenomic B. bovis DNA just above the detection limit (Fig. 4).The B. bovis signal was found in all samples (Fig. 4, lanes 1 to7) and varied slightly, although it was added to the master mixcontaining all components except the T. annulata DNA. It is,however, clear that the B. bovis signal was not reduced even inthe presence of a 500,000-fold excess of T. annulata DNA.

Detection of mixed parasites. In order to mimic field situa-tions, two mixtures of DNAs from parasites that are known tooccur in the same region were prepared and subsequentlytested in the RLB. Mixture I contained T. annulata, T. orien-talis, B. bovis, B. bigemina, and B. divergens (parasites thatoccur, for instance, in North Africa, Southern Europe, and theMiddle East). Mixture II contained T. parva, T. mutans, T.taurotragi, T. velifera, and T. orientalis (parasites that occur inEast and Southern Africa). Figure 5 shows the results of bothmixes, together with the signal obtained with each parasiteseparately. All parasites in both mixtures are detected simul-taneously, and no cross-reactions occur (Fig. 5).

Detection of multiple parasites in experimentally infectedcattle. Animals were infected with T. annulata followed byinfection with a second Theileria species or a Babesia species.Blood samples were tested before infection, during the clinicalmanifestation of T. annulata, and during the second infection.Sampling was carried out on a weekly basis until 7 weeks p.i.Reactions in the RLB for these animals are shown in Fig. 6.T. annulata was readily detected at all times. In the coinfectionwith T. mutans, the T. mutans signal was present only at 7weeks p.i. (Fig. 6A, lane 10). Coinfection with one of the twoT. orientalis isolates resulted in strong T. orientalis signals in allanimals from week 3 and week 4 p.i. until week 7 p.i. for theJapanese and the Australian isolate, respectively (Fig. 6B andC). B. divergens coinfection was usually detectable only duringthe clinical manifestation of this parasite (Fig. 6D). However,additional samples were positive between days 25 and 28 p.i.

(data not shown). The secondary B. bovis infection was readilydetected before treatment with Imidocarb but disappearedafter treatment (Fig. 6E). B. bigemina coinfection was detectedfrom 1 week p.i. until week 6 p.i., although only a very faintsignal was found at week 5 p.i. (Fig. 6F).

Detection of parasites in samples collected from cattle inSpain. T. annulata was detected in 26 of 28 animals in Toledo(Fig. 7A) and in 17 of 17 animals in Cadiz (Fig. 7B). T. orien-talis was detected in 2 of 28 animals in Toledo (Fig. 7A, lanes6 and 13) and 15 of 17 animals in Cadiz (Fig. 7B, lanes 1 to 6and 9 to 17), which were all positive for T. annulata as well.One animal in Toledo was positive for both B. bigemina andT. annulata (Fig. 7A; arrow). Two animals in Toledo werecompletely negative (Fig. 7A, lanes 15 and 26).

DISCUSSION

We developed an RLB assay based on the 18S rRNA geneof Theileria and Babesia species infecting cattle mainly in thetropics and subtropics. All species included were detected by aspecific oligonucleotide, and no cross-reactions were observed.Neither bovine DNA nor any of the Trypanosoma or Ehrlichiaspecies were detected (although Ehrlichia canis is not a bovineparasite, it was used as a representative, since Ehrlichia boviswas not available). The catchall Theileria and Babesia oligonu-cleotide control is of importance in case a PCR product isamplified but no specific reaction is seen. This was illustratedby including B. major in the RLB (Fig. 2, lane 25). Similarresults could indicate the presence of a novel Babesia or Thei-leria species or strain or the presence of a known species forwhich no probe was included. An example of the latter couldbe B. occultans, which occurs in cattle in Southern Africa (19)but whose 18S rRNA gene sequence is presently unknown.Additional species or strains could be identified by RLB; thePCR product should subsequently be sequenced to identifythese organisms, and if necessary, they can be included in theassay.

Although the T. orientalis group is heterogeneous and dif-ferences in morphology, pathogenicity, and virulence havebeen reported (48), we decided to include a catchall T. orien-talis oligonucleotide for this group. Recently, heterogeneitywas also demonstrated at the 18S rRNA level by Chae et al.(8). Alignments (data not shown) demonstrated that all group

FIG. 7. RLB with PCR products derived from DNA extracted from fieldsamples collected in Spain. (A) Lanes: 1 to 8 and 10 to 29, samples from Toledo;9, Cadiz sample; 30, PCR distilled water control (2); 31, 10 ng of T. annulata asa positive control (1). The sample positive for both T. annulata and B. bigeminais marked with an arrow. (B) Lanes: 1 to 17, samples from Cadiz; 18, empty (x);19, distilled water control (2); 20, 10 ng of T. annulata as a positive control (1).The twice-measured Cadiz sample is marked with an asterisk in both blots. Rowsare identical to those in Fig. 2.

VOL. 37, 1999 RLB ASSAY FOR BOVINE THEILERIA AND BABESIA SPECIES 1787

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

members have in common one part of 18S of the V4 regionthat could differentiate all of them from the other Theileriaspecies. Two representatives of the 13 sequences used for spe-cific oligonucleotide deduction are depicted in Fig. 1. A mix oftwo oligonucleotides containing either an A or a T at position668 (Fig. 1) was used to detect all parasites in this group. It wasdemonstrated that all eight T. orientalis group isolates, of whichat least type A (T. orientalis China and Fukushima, Japan;GenBank no. AB016074) and type D (isolate from northwest-ern China) were tested in our study, hybridized with this oligo-nucleotide mix.

The sensitivity of the RLB assay was established by testingserial dilutions of T. annulata-infected blood samples; it wasdetermined to be sensitive at 1026% parasitemia. The corre-sponding blood sample was 40 ml, which means that the lowerdetection limit of the assay is approximately 120 parasites. Thedetection limit of the T. annulata PCR based on the Tams1gene (13) was 4.8 3 1025%, which is highly comparable withthe sensitivity found for the RLB. For B. bovis, a detection of1027 to 1028% parasitemia was achieved by using 18S PCRand a specific probe followed by very sensitive phosphorimagerdetection (6). These investigators demonstrated that B. bovisparasitemia fluctuates and sometimes even escapes detectionby their assay. Although the RLB assay is less sensitive, detec-tion of B. bovis at time points when the parasitemia rises up to1026% is possible.

The DNAs of those parasites that can occur in the samegeographical region were mixed and appeared to be equallywell amplified (Fig. 5). However, when these parasites arepresent in very different numbers, another situation arises,since competition for available primers may occur in the PCR.The competition experiment between T. annulata and B. bovisdemonstrated that such competition does not occur, at leastnot between these two parasites (Fig. 4). The absence of com-petition can be explained by the excess of PCR primers presentin the reaction mixture. This was also shown in Fig. 5, whereinparasite DNAs tested separately or mixed resulted in signals ofcomparable intensity. The effects of competition between mul-tiple parasites present in one sample may therefore not influ-ence the outcome of the RLB, although not all possible com-binations were tested.

Furthermore, in experimental animals, T. annulata could bedetected simultaneously with T. mutans, T. orientalis, B. diver-gens, B. bovis, and B. bigemina. After Imidocarb treatment theB. bovis signal was completely lost, confirming the earlier find-ings of Calder et al. (6). Our B. bigemina results are similar tothose obtained with the previously published B. bigemina-spe-cific PCR (15). These authors could detect 1027% parasitemia,but signals disappeared at some time points in carrier animals.Fluctuating low parasitemia, as was described for B. bovis (6),was probably the reason for the occasional detection of B.divergens (Fig. 6D). Since B. bigemina was readily detectedafter the acute phase, its level of parasitemia during the carrierstate is probably higher than those of B. bovis and B. divergens.

Screening of field samples collected from cattle in two re-gions in Spain demonstrated the usefulness of the RLB. T.annulata alone or in combination with T. orientalis could bedetected in the majority of blood samples. The presence ofT. annulata was expected, since this parasite is endemic inSpain (50), but T. orientalis was found for the first time inSpain. T. orientalis does occur elsewhere in Southern Europe,as was recently demonstrated in Italy (7). The presence of B.bigemina in one animal shows that the simultaneous detectionof Babesia species is also possible in field samples.

The RLB assay is very practical, since only one PCR isrequired for simultaneous detection of all Theileria and Babe-

sia parasites. When cattle are screened for their infection sta-tus, information can be collected concerning the epidemiologyof pathogenic and nonpathogenic parasites, whereby otherspecies-specific PCR assays are superfluous. The membranewith the covalently linked species-specific oligonucleotides canbe used at least 20 times, reducing costs for screening animals.Finally, we are now in the process of extending the RLB toinclude tick-transmitted ehrlichial species such as Cowdria,Anaplasma, and Ehrlichia, which frequently occur simulta-neously in the same host animal in the field.

ACKNOWLEDGMENTS

We thank E. Pipano, Kimron Institute, Beit Dagan, Israel, for pro-viding a strain of Babesia bovis, J. Dawson, CDC, Atlanta, Ga., for theEhrlichia canis isolate, and Yin Hong and Bai Qi for the T. orientalisisolates from China. A. W. C. A. Cornelissen is thanked for criticalreading of the manuscript and for his support for this work.

This work was supported by a grant from the European Community,Directorate General XII, INCO-DC program, contract no. IC18-CT95-0003, entitled “Application of Recombinant DNA Technologyto Vaccination, Diagnosis and Epidemiology of Tropical Theileriosis.”

REFERENCES1. Allsopp, B. A., H. A. Bayliss, M. T. E. P. Allsopp, T. Cavalier-Smith, R. P.

Bishop, D. M. Carrington, B. Sohanpal, and P. Spooner. 1993. Discrimina-tion between six species of Theileria using oligonucleotide probes whichdetect small-subunit ribosomal RNA sequences. Parasitology 107:157–165.

1a.Allsopp, P. Unpublished data.2. Bishop, R., B. Sohanpal, D. P. Kariuki, A. S. Young, V. Nene, H. Bayliss,

B. A. Allsopp, P. R. Spooner, T. T. Dolan, and S. P. Morzaria. 1992. Detec-tion of a carrier state in Theileria parva-infected cattle by the polymerasechain reaction. Parasitology 104:215–232.

3. Bishop, R., B. Allsopp, P. Spooner, B. Sohanpal, S. Morzaria, and E.Gobright. 1995. Theileria: improved species discrimination using oligonucle-otides derived from large-subunit ribosomal RNA sequences. Exp. Parasitol.80:107–115.

4. Bouattour, A., and M. A. Darghouth. 1996. First report of Babesia divergensin Tunisia. Vet. Parasitol. 63:161–165.

5. Brocklesby, D. W., S. F. Barnett, and G. R. Scott. 1961. Morbidity andmortality rates in East Coast fever (Theileria parva infection) and theirapplication to drug screening procedures. Br. Vet. J. 117:529–531.

6. Calder, J. A. M., G. M. Reddy, L. Chieves, C. H. Courtney, R. Littell, J. R.Livengood, R. A. I. Norval, G. Smith, and J. B. Dame. 1996. MonitoringBabesia bovis infections in cattle by using PCR-based tests. J. Clin. Microbiol.34:2748–2755.

7. Ceci, L., E. Kirvar, G. Carelli, D. Brown, M. Sasanelli, and O. Sparagano.1997. Evidence of Theileria buffeli infection in cattle in southern Italy. Vet.Rec. 140:581–583.

8. Chae, J., J. Lee, O. Kwon, P. J. Holman, S. D. Waghela, and G. G. Wagner.1998. Nucleotide sequence heterogeneity in the small-subunit ribosomalRNA gene variable (V4) region among and within geographic isolates ofTheileria from cattle, elk and white-tailed deer. Vet. Parasitol. 75:41–52.

9. Clausen, P.-H., A. Wiemann, R. Patzelt, D. Kakaire, C. Poetzch, A. Pere-grine, and D. Mehlitz. 1998. Use of a PCR assay for the specific and sensitivedetection of Trypanosoma spp. in naturally infected dairy cattle in peri-urbanKampala, Uganda. Ann. N. Y. Acad. Sci. 849:21–31.

10. Corpet, F. 1988. Multiple sequence alignment with hierarchical clustering.Nucleic Acids Res. 16:10881–10890.

11. De Kok, J. B., C. d’Oliveira, and F. Jongejan. 1993. Detection of the pro-tozoan parasite Theileria annulata in Hyalomma ticks by the polymerasechain reaction. Exp. Appl. Acarol. 17:839–846.

12. De Kroon, J. F., N. M. Perie, F. F. Fransen, and G. Uilenberg. 1990. Theindirect fluorescent antibody test for bovine anaplasmosis. Vet. Q. 12:124–128.

13. d’Oliveira, C., M. van der Weide, M. A. Habela, P. Jacquiet, and F. Jongejan.1995. Detection of Theileria annulata in blood samples of carrier cattle byPCR. J. Clin. Microbiol. 33:2665–2669.

14. Felgner, P., U. Brinkmann, U. Zillmann, D. Mehlitz, and S. Abu-Ishira.1981. Epidemiological studies on the animal reservoir of gambiense sleepingsickness. Part II. Parasitological and immunodiagnostic examination of thehuman population. Tropenmed. Parasitol. 32:134–140.

15. Figueroa, J. V., L. P. Chieves, G. S. Johnson, and G. M. Buening. 1992.Detection of Babesia bigemina-infected carriers by polymerase chain reactionamplification. J. Clin. Microbiol. 30:2576–2582.

16. Figueroa, J. V., L. P. Chieves, G. S. Johnson, and G. M. Buening. 1993.Multiplex polymerase chain reaction-based assay for the detection of Babesiabigemina, Babesia bovis and Anaplasma marginale DNA in bovine blood. Vet.Parasitol. 50:69–81.

1788 GUBBELS ET AL. J. CLIN. MICROBIOL.

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from

17. Figueroa, J. V., and G. M. Buening. 1995. Nucleic acid probes as a diagnosticmethod for tick-borne hemoparasites of veterinary importance. Vet. Parasi-tol. 57:75–92.

18. Gill, B. S., G. C. Bansal, Y. Bhattacharyulu, D. Kaur, and A. Singh. 1980.Immunological relationship between strains of Theileria annulata Dschun-kowsky and Luhs 1904. Res. Vet. Sci. 29:93–97.

19. Gray, J. S., and A. J. de Vos. 1981. Studies on a bovine Babesia transmittedby Hyalomma marginatum rufipes Koch, 1844. Onderstepoort J. Vet. Res.48:215–223.

20. Ishihara, T., and T. Minami. 1978. Bovine theileriasis and babesiasis inJapan, p. 201–209. In T. Ishizaki, S. Inoki, and J. Fujita (ed.), Protozoandiseases. The Japanese-German Association of Protozoan Diseases, Tokyo,Japan.

21. Jacquiet, P., M. L. Dia, N. M. Perie, F. Jongejan, G. Uilenberg, and P. C.Morel. 1990. Presence de Theileria annulata en Mauritanie. Revue Elev.Med. Vet. Pays Trop. 43:489–490.

22. Jianxum, L., and L. Wenshun. 1997. Cattle theileriosis in China. Trop. Anim.Health Prod. 29:4S–7S.

23. Jongejan, F., G. Uilenberg, F. F. J. Franssen, A. Gueye, and J. Nieuwenhuijs.1988. Antigenic differences between stocks of Cowdria ruminantium. Res.Vet. Sci. 44:186–189.

24. Jongejan, F., and G. Uilenberg. 1994. Ticks and control methods. Rev. Sci.Tech. Off. Int. Epizoot. 13:1201–1226.

25. Kamerbeek, J., L. Schouls, A. Kolk, M. van Agterveld, D. Soolingen, S.Kuijper, A. Bunschoten, H. Molhuizen, R. Shaw, M. Goyal, and J. vanEmbden. 1997. Simultaneous detection and strain differentiation of Myco-bacterium tuberculosis for diagnosis and epidemiology. J. Clin. Microbiol.35:907–914.

26. Kaufhold, A., A. Podbieldski, G. Baumgarten, M. Blokpoel, J. Top, and L.Schouls. 1994. Rapid typing of group A streptococci by the use of DNAamplification and non-radioactive allele-specific oligonucleotide probes.FEMS Microbiol. Lett. 119:19–26.

27. Kawazu, S., C. Sugimoto, T. Kambio, and K. Fusjiaki. 1992. Analysis of thegenes encoding immunodominant piroplasm surface proteins of Theileriasergenti and Theileria buffeli by nucleotide sequencing and polymerase chainreaction. Mol. Biochem. Parasitol. 56:169–175.

28. Kock, N. D., A. H. M. van Vliet, K. Charlton, and F. Jongejan. 1995.Detection of Cowdria ruminantium in blood and bone marrow from clinicallynormal, free-ranging Zimbabwean wild ungulates. J. Clin. Microbiol. 33:2501–2504.

29. Kuttler, K. L., and T. M. Craig. 1975. Isolation of a bovine Theileria. Am. J.Vet. Res. 36:323–325.

30. Leeflang, P., J. Buys, and C. Blotkamp. 1976. Studies on Trypanosoma vivax:infectivity and serial maintenance of natural bovine isolates in mice. Int. J.Parasitol. 6:413–417.

31. Leeflang, P., and A. A. Ilemobade. 1977. Tick-borne diseases of domesticanimals in Northern Nigeria. II. Research summary, 1966 to 1976. Trop.Anim. Health Prod. 9:211–218.

32. Medlin, L., H. J. Elwood, S. Stickel, and M. L. Sogin. 1988. The characteri-sation of enzymatically amplified eukaryotic 16S-like rRNA-coding regions.Gene 71:491–499.

33. Melrose, T. R., C. G. D. Brown, S. P. Morzaria, J. G. R. Ocama, and A. D.Irvin. 1984. Glucose phosphate isomerase polymorphism in Theileria annu-lata and T. parva. Trop. Anim. Health Prod. 16:239–245.

34. Morzaria, S. P., S. F. Barnett, and D. W. Brocklesby. 1974. Isolation ofTheileria mutans from cattle in Essex. Vet. Rec. 94:256.

35. Nyindo, M. B., M. Ristic, D. L. Huxsoll, and A. R. Smith. 1971. Tropicalcanine pancytopenia: in vitro cultivation of the causative agent Ehrlichiacanis. Am. J. Vet. Res. 32:1651–1658.

36. Papadopoulos, B., N. M. Perie, and G. Uilenberg. 1996. Piroplasms of do-mestic animals in the Macedonia region of Greece. 1. Serological cross-reactions. Vet. Parasitol. 63:41–56.

37. Perie, N. M., G. Uilenberg, and B. E. Schreuder. 1979. Theileria mutans inNigeria. Res. Vet. Sci. 26:359–362.

38. Qi, B., L. Guangyuan, and H. Genfeng. 1997. An unidentified species ofTheileria infective for cattle discovered in China. Trop. Anim. Health Prod.29:43S–47S.

39. Rijpkema, S. G., M. J. Molkenboer, L. M. Schouls, F. Jongejan, and J. F.Schellekens. 1995. Simultaneous detection and genotyping of three genomicgroups of Borrelia burgdorferi sensu lato in Dutch Ixodes ricinus ticks bycharacterization of the amplified intergenic spacer region between 5S and23S rRNA genes. J. Clin. Microbiol. 33:3091–3095.

40. Saiki, R. K., C.-A. Chang, C. H. Levenson, T. C. Warren, C. D. Boehm, H. H.Kazazian, and H. A. Ehrlich. 1988. Diagnosis of sickle cell anemia andb-thalassemia with enzymatically amplified DNA and nonradioactive allele-specific oligonucleotide probes. N. Engl. J. Med. 319:537–541.

41. Schein, E., G. Buscher, and K. T. Friedhoff. 1975. LichtmikroskopischeUntersuchungen uber die Entwicklung von Theileria annulata (Dschun-kowsky und Luhs 1904) in Hyalomma anatolicum excavatum (Koch 1844). I.Die Entwicklung im Darm vollgesogener Nymphen. Z. Parasitenkd. 48:123–136.

42. Tanaka, M., S. Onoe, T. Matsuba, S. Katayama, M. Yamanka, H. Yonemi-chi, K. Hiramatsu, B.-K. Baek, C. Sugimoto, and M. Onuma. 1993. Detec-tion of Theileria sergenti infection in cattle by polymerase chain reactionamplification of parasite-specific DNA. J. Clin. Microbiol. 31:2565–2569.

43. Uilenberg, G., and B. E. C. Schreuder. 1976. Studies on Theileriidae (Sporo-zoa) in Tanzania. I. Tick transmission of Haematoxenus veliferus. Tro-penmed. Parasit. 27:106–111.

44. Uilenberg, G., B. E. C. Schreuder, and C. Mpangala. 1976. Studies onTheileriidae (Sporozoa) in Tanzania. III. Experiments on the transmission ofTheileria mutans by Rhipicephalus appendiculatus and Amblyomma variega-tum (Acarina, Ixodidae). Tropenmed. Parasit. 27:323–328.

45. Uilenberg, G. 1981. Theilerial species of domestic livestock, p. 4–37. In A. D.Irvin, M. P. Cunningham, and A. S. Young (ed.), Advances in the control oftheileriosis. Martinus Nijhoff, The Hague, The Netherlands.

46. Uilenberg, G., N. M. Perie, J. A. Lawrence, A. J. de Vos, R. W. Paling, andA. A. M. Spanjer. 1982. Causal agents of bovine theileriosis in SouthernAfrica. Trop. Anim. Health Prod. 14:127–140.

47. Uilenberg, G., and R. Hashemi-Fesharki. 1984. Theileria orientalis in Iran.Vet. Q. 6:1–4.

48. Uilenberg, G., N. M. Perie, A. A. M. Spanjer, and F. F. J. Franssen. 1985.Theileria orientalis, a cosmopolitan blood parasite of cattle: demonstration ofthe schizont stage. Res. Vet. Sci. 38:352–357.

49. Uilenberg, G., F. F. Franssen, and N. M. Perie. 1986. Stage-specific anti-genicity in Theileria annulata: a case report. Vet. Q. 8:73–75.

50. Viseras, J., P. Garcıa-Fernandez, and F. J. Adroher. 1997. Isolation andestablishment in in vitro culture of a Theileria annulata-infected cell line fromSpain. Parasitol. Res. 83:394–396.

VOL. 37, 1999 RLB ASSAY FOR BOVINE THEILERIA AND BABESIA SPECIES 1789

on May 23, 2020 by guest

http://jcm.asm

.org/D

ownloaded from