INTRODUCTION
Infection with Leucocytozoon caulleryi, the pathogen in chicken leucocytozoonosis, induces severe illness and results in high mortality in chicken farms in Japan (Morii, Reference Morii1992). L. caulleryi has a life cycle similar to that of Plasmodium spp. causing malaria; it is transmitted by Culicoides biting midges (Akiba, Reference Akiba1960) and has 3 growth stages, schizogony and gametogony in chickens and sporogony in Culicoides (Morii, Reference Morii1992). However, unlike the malaria protozoan, L. caulleryi is not easily transmitted by blood because it does not undergo schizogony in circulating blood cells. Thus, in vitro culture of L. caulleryi in chicken blood has not been successful (Morii, Reference Morii1992). L. caulleryi infection in chickens causes listlessness, green feces, anorexia and often death from haemorrhage by second-generation schizonts. Survivors suffer from severe haemolytic anaemia, owing to erythrocyte destruction by gametocytes, and haemorrhagic anaemia owing to vascular endothelial cell destruction (Morii, Reference Morii1992).
Several vaccination methods for L. caulleryi have been reported (Isobe et al. Reference Isobe, Suzuki and Yoshihara1991; Itoh and Gotanda, Reference Itoh and Gotanda2002; Ito and Gotanda, Reference Ito and Gotanda2004; Ito et al. Reference Ito, Gotanda, Kobayashi, Kume, Sugimoto and Matsumura2005). A recombinant antigen of chicken leucocytozoonosis has been developed (Ito and Gotanda, Reference Ito and Gotanda2004), but the gene(s) has not been identified because of the lack of sufficient genetic data for this protozoan. Recombinant R7 antigen is derived from second-generation schizonts and can be produced in Escherichia coli in large quantities (Itoh and Gotanda, Reference Itoh and Gotanda2002; Ito and Gotanda, Reference Ito and Gotanda2004; Ito et al. Reference Ito, Gotanda, Kobayashi, Kume, Sugimoto and Matsumura2005). However, several problems regarding the safety of and difficulty in production of vaccines have prevented their practical use. For example, the period of protective immunity is only 5 months in chickens (Itoh and Gotanda, Reference Itoh and Gotanda2002); it is therefore desirable to lengthen the period of defensive immunity by methods such as booster vaccination. In addition, there are many difficulties in the development of protozoan vaccines, given their complex life cycle and the host/parasite interaction. Each developmental stage of the parasite is characterized by different sets of expressed antigens, eliciting different types of immune responses (Carvalho et al. Reference Carvalho, Daniel-Riberio and Goto2002). Thus, molecular analysis of L. caulleryi is required to develop more effective vaccines.
Stage-specific gene expression in L. caulleryi has not been characterized because of the difficulty in preventing host genome contamination. Avian apicomplexan protozoa, including Leucocytozoon spp., parasitize bird erythrocytes intracellularly. The result is that extracted DNA from infected avian blood includes 2 sets of genomes, protozoan and host. Therefore, separation of L. caulleryi from infected avian blood cells is needed. Unlike methods for isolation of mammalian blood protozoa, there are no reports of such methods in birds. Attempts to isolate L. caulleryi gametocytes from infected bird blood by gradient centrifugation have resulted in host bird cells still remaining in the target fraction (T. Isobe, personal communication). Flow cytometry is one potential method to detect or separate protozoa from infected blood. Several flow cytometric methods have been applied for the detection of mammalian malarial parasites in infected blood (Saito-Ito et al. Reference Saito-Ito, Akai, He, Kimura and Kawabata2001; Xie et al. Reference Xie, Li, Johnson, Zhang, Milhous and Kyle2007). To the best of our knowledge, flow cytometric analysis of L. caulleryi has not been reported. Only 1 species of avian protozoa, Eimeria tenella, has been analysed by flow cytometry to detect sporozoites and oocysts (Crane et al. Reference Crane, Murray, Gnozzio and Macdonald1988; Fuller and Mcdougald, Reference Fuller and Mcdougald1989). Ficoll density-gradient centrifugation was used in an attempt to separate the blood stage of the avian protozoan for Leucocytozoon smithi (Henry and Dick, Reference Henry and Dick1978). Thus, there are no published data for comparison of dot-plot patterns between infected and uninfected blood for L. caulleryi. Although L. caulleryi-infected chicken blood can be used as a first step to investigate stage-specific gene expression of this protozoan, the presence of host blood cells may contaminate DNA/RNA extraction procedures.
Light and electron microscopy studies have shown that mature gametocytes of L. caulleryi are enclosed by the cytoplasm of the host cell and are surrounded by a distinct 3-layered pellicle in the peripheral blood of infected chickens. However, the nucleus of the host cell is not observed around mature gametocytes (Morii et al. Reference Morii, Matsui, Iijima and Fukuda1981, Reference Morii, Iijima and Fukuda1984). The outer membrane of the gametocyte may originate from the host cell; this is supported by the fact that the outer membrane disappears together with the cytoplasm of the host cell after withdrawal or ingestion by biting midges (Morii et al. Reference Morii, Iijima and Fukuda1984). Mature L. caulleryi gametocytes exit from host erythrocytes 18–24 days after infection (Morii, Reference Morii1992), suggesting that the protozoa and host cells clearly separate.
The aim of the present study was to obtain purified L. caulleryi DNA and/or RNA from infected chickens. We used flow cytometry to separate L. caulleryi gametocytes from infected avian blood cells.
MATERIALS AND METHODS
L. caulleryi-infected chicken blood
L. caulleryi-infected chicken blood was obtained from experimentally infected chickens at day 20 post-infection (Scientific Feed Laboratory Co. Ltd, Tochigi, Japan). Collected blood was stored in an equal volume of Alsever solution (Nippon Biotest Laboratories Inc., Tokyo, Japan) at 4°C until analysis. Uninfected chicken blood (Nippon Biotest Laboratories Inc.) was used as a negative control. Wright-Giemsa-stained blood smears were produced before flow cytometric analysis, and L. caulleryi gametocyte parasitaemia (gametocytes only) was calculated 3 times (n=3 for each sample) by microscopic observation of 1000 blood cells on each slide (×400 magnification).
Flow cytometric analysis
A volume of 5–10 μl of whole infected or uninfected chicken blood was added to 500 μl of sterile distilled water and incubated at 4°C for 1–5 min to induce erythrocyte lysis. A volume of 500 μl of sterile 1·8% (w/w) saline was added to terminate lysis, and SYTO-24 nucleic acid stain was added at a final concentration of 10 ng/ml (Dyer and Day, Reference Dyer and Day2003) and incubated at room temperature for 20 min. The cells were then centrifuged at 1500 g for 5 min at 4°C for collection. The supernatant was removed by aspiration, and the cells were resuspended in 200 μl of Alsever solution. The suspension of stained cells was analysed using an EPICS ALTRA HyPerSort (Beckman Coulter, Inc., Fullerton, CA, USA) equipped with a 488-nm argon ion laser. Green fluorescence was detected in the corresponding FL-1 photomultiplier through a 530/30 bandpass filter. The parasite area, in which gametocytes were plotted, was defined in a 2-dimensional scattergram of green fluorescence (GF) and forward scatter (FSC) primarily by comparing the patterns of events in the 2 scattergrams for uninfected and infected chicken blood.
Cell sorting, observation of sorted cells and DNA extraction
Cell sorting was performed using an EPICS ALTRA HyPerSort (Beckman Coulter, Inc., Fullerton, CA, USA). Sample preparation and cell staining was performed as described above. Stained cells were filtered with a 35μm filter with cell-strainer tubes (Becton Dickinson, Tokyo, Japan) before sorting. We sorted 4000–10 000 cells in the gate specifically identified as infected blood. Sorted cells were collected in 1·5ml tubes containing Alsever solution plus 50% (v/v) fetal bovine serum to observe cell morphology or in 1·5ml tubes containing DNA extraction solution.
To observe cells, they were centrifuged onto glass slides using a Cytospin 3 cytocentrifuge (Shandon, UK); cells were centrifuged at 1000 g for 5 min at room temperature, smear preparations were made and stained with Wright-Giemsa and cells were analysed. Each set of 3 Cytospin slides (n=3) was counted in triplicate, and the percentage of L. caulleryi gametocytes was calculated by microscopic observation of 1000 cells (×400 magnification). DNA of sorted cells was extracted using a QIAamp DNA mini kit (Qiagen Ltd, Crawley, UK) according to the manufacturer's instructions.
Evaluation of separation by cell sorting
To evaluate the efficiency of cell sorting of infected blood, we performed polymerase chain reaction (PCR) of extracted DNA from infected whole blood and sorted cells. A total of 1 ng of DNA was used as a template to amplify the L. caulleryi clpC gene (primer: F, 5′-GAT TTG ATA TGA GTG AAT ATA TAT GGA AAA AC-3′; R, 5′-AAT ATA TTA GTT AAT CTA TTT AAT AAT TC-3′) as described previously (Omori et al. Reference Omori, Sato, Isobe, Yukawa and Murata2007, Reference Omori, Sato, Hirakawa, Isobe, Yukawa and Murata2008) and the avian host 16S rRNA gene (primers and PCR condition were described previously by Imura et al. Reference Imura, Sato, Ejiri, Tamada, Isawa, Sawabe, Omori, Murata and Yukawa2010). We then compared the degree of amplification of each gene between whole blood and sorted cells.
RESULTS AND DISCUSSION
L. caulleryi gametocyte parasitaemia was calculated in 3 independently collected samples at 20 days post-infection and was found to be approximately 0·33% to 0·87%. Results of flow cytometry are shown in Fig. 1. A 2-dimensional scattergram of gametocyte-positive blood showed a unique area, not shown in uninfected blood, as indicated by an asterisk (Fig. 1b). Gametocytes of L. caulleryi were observed in the sorted cell fraction, as shown in Fig. 2; approximately 85% of sorted cells were gametocytes. Host erythrocytes and lymphocytes were found mixed with gametocytes (Fig. 2b and c).
Fig. 1. Flow cytometric analysis of Leucocytozoon caulleryi-uninfected (a) or -infected (b) chicken blood. Lysed chicken blood was stained with SYTO-24. Compared to uninfected blood, a specific area was identified in dot plots of infected chicken blood (indicated by asterisk). Morphological analysis of sorted cells from the gated area (*) of infected blood showed that the cells were L. caulleryi gametocytes (see Fig. 2). The number in each gate indicates the percentage of gametocytes. GF, green fluorescence; FSC, forward scatter.
Fig. 2. Morphology of sorted cells from infected chicken blood at 20 days post-infection. Sorted cells were stained with Wright-Giemsa. Arrows denote Leucocytozoon caulleryi gametocytes; arrowheads indicate chicken erythrocyte; asterisk represents chicken thrombocytes. (a) L. caulleryi gametocytes, (b) chicken erythrocyte and L. caulleryi gametocyte, (c) L. caulleryi gametocyte and chicken thrombocyte. Scale bars=10 μm.
The effect of cell sorting on PCR is shown in Fig. 3. Compared to infected blood without sorting, amplification of the protozoa clpC gene was notably higher in sorted cells, and amplification of the 16S rRNA host gene was lower (Fig. 3; compare band intensity between lanes 3 and 9 (clpC) and lanes 4 and 10 (16S rRNA)). We did not quantify amplification results in the present study; quantitative PCR may be applied for more precise investigation of the effects of sorting.
Fig. 3. Comparison of gene amplification (Leucocytozoon caulleryi clpC gene and avian host 16S rRNA gene) between whole blood and gametocytes by polymerase chain reaction (PCR). PCR was performed with 1 ng of each DNA extracted from infected whole blood using a kit (see Materials and Methods section) and sorted gametocytes as templates. Arrows indicate bands showing amplification of the L. caulleryi clpC gene. Lanes 1 to 6: infected whole blood; lanes 7 to 12: DNA extracted from sorted cells of infected chicken blood. M: molecular weight marker. Even lane numbers show 16S rRNA gene, approximately 300 bp (lanes 2 and 8, 25 PCR cycles; lanes 4 and 10, 30 PCR cycles; lanes 6 and 12, 35 PCR cycles). Odd lane numbers, show clpC gene, approximately 400 bp (lanes 1 and 7, 25 PCR cycles; lanes 3 and 9, 30 PCR cycles; lanes 5 and 11, 35 PCR cycles).
In the present study, we used flow cytometry to create 2-dimensional scattergrams of chicken blood infected with the avian protozoa L. caulleryi for the first time. A protozoa-specific area was identified in dot plots of infected blood samples, as shown in previous flow cytometric analysis of mammalian blood protozoa (Saito-Ito et al. Reference Saito-Ito, Akai, He, Kimura and Kawabata2001). Henry and Dick (Reference Henry and Dick1978) separated the blood stage of L. smithi by Ficoll density-gradient centrifugation; however, morphologic findings of microscopic observation of the separated fraction were not shown in that report. Here, we clearly show that the sorted fraction of the specific area in the flow cytometry dot plot contained L. caulleryi gametocytes. The difference in flow cytometry pattern between infected and uninfected chicken blood might be associated with morphological features among gametocytes and other host blood cells. The collected fraction of sorted cells was expected to include only L. caulleryi gametocytes. However, microscopic observation and PCR examination showed remaining contamination by host erythrocytes and thrombocytes. A mix of host-derived cells was reported for the human malaria parasite by flow cytometric detection (Saito-Ito et al. Reference Saito-Ito, Akai, He, Kimura and Kawabata2001). In the present preliminary study, we attempted to compare flow cytometry patterns between L. caulleryi-infected and uninfected chicken blood. Methodological improvements in haemolysis, cell staining, dyes or antibodies and fluorescence-activated cell sorting (FACS) analysis may be necessary. Moreover, the precise quantification of amplified products by qPCR could clearly reveal the efficiency of cell sorting. In addition, flow cytometry analyses using anti-chicken red blood cell antibody staining could clearly confirm gametocytes without stained antibodies and the same antibody could also be used to negatively select against host cells in the FACS assay.
The period for a mature L. caulleryi gametocyte to exit from the cytoplasm of host erythrocytes ranges from 18 to 24 days after infection (Morii et al. Reference Morii1992). Immature or other gametocytes may have remained in our samples because we used blood collected 20 days post-infection. For more effective FACS separation, the use of anti-chicken erythrocyte antibody may be useful to distinguish gametocytes from chicken erythrocytes.
Contamination of host DNA was reduced in flow cytometry-sorted samples compared to samples without flow cytometric sorting. The increased amount of amplified L. caulleryi clpC gene in extracted DNA from the sorted fraction indicates the possibility of effective analysis of protozoa-specific gene expression. Further improvements in the exclusion of host-derived DNA are needed; however, our present results provide a first step for evaluation of stage-specific gene analysis of this highly pathogenic avian protozoan and may aid in the development of genetically attenuated vaccines.
ACKNOWLEDGEMENTS
We thank Drs Tetsuya Mukai, Keiichi Masubuchi, Isao Shibata and Hiroshi Tanabe of Scientific Feed Laboratory Co., Ltd, for supplying L. caulleryi-infected chicken blood and for their helpful suggestions.
FINANCIAL SUPPORT
This study was supported in part by a Grant-in-Aid for Scientific Research (No. 21580406) from the Japan Society for the Promotion of Science and Strategic Research Base Development Program ‘International Research on Epidemiology of Zoonoses and Training for Young Researchers’; a ‘High-Tech Research Center’ Project for Private Universities matching fund subsidy from the Ministry of Education, Culture, Sports, Science and Technology of Japan; Nihon University Research Grants; and a Grant-in-Aid for JSPS Fellows (No. 203886) from the Japan Society for the Promotion of Science.
