Introduction

Avian coccidiosis is caused by an intestinal protozoa belonging to several species of the genus Eimeria. Protozoan infection results in extensive destruction of the avian intestinal epithelium accompanied by reduced feed efficiency, severe depression in body weight gain, and thus significant economic losses for the poultry industry (Lillehoj 1998; Lillehoj and Lillehoj 2000). Although avian coccidiosis has been extensively controlled by prophylactic medication, alternative control methods are needed due to the emergence of drug-resistant parasite strains (Chapman 1993) and the uncertain future for drug usage in animal husbandry. In this regard, considerable efforts have been made to develop recombinant vaccines (Barriga 1994; Lillehoj et al. 2000; Song et al. 2000; Min et al. 2001). However, control of coccidiosis by field vaccination has been only partially successful due to limited information on host protective immunity, antigenic variation associated with different Eimeria life-cycle stages, and the appearance of mutant strains (Fitz-Coy 1992; Tomley 1994). Recent studies investigating chicken adaptive immune effector mechanisms suggest that field vaccination to control of Eimeria infection may be commercially feasible. This prospect is exemplified by the studies of the intestinal immune response produced subsequent to Eimeria infection (Jakowlew et al. 1997; Choi et al. 1999; Yun et al. 2000a, b; Laurent et al. 2001; Li et al. 2002; Miyamoto et al. 2002). It is now known, for example, that intestinal intraepithelial lymphocytes (IELs) clearly play a central role in protective immunity to avian coccidiosis (Lillehoj and Lillehoj 2000).

Recent advancements in high-throughput gene profiling technology have made it possible to monitor the expression of thousands of gene transcripts in a single hybridization experiment (Freeman et al. 2000) thereby making a significant contribution to an understanding of the molecular processes associated with health and disease (Khan et al. 1999; Young 2000; Lucchini et al. 2001). DNA microarrays have been successfully applied to monitor changes in gene expression following infection with viruses (Morgan et al. 2001), bacteria (Cohen et al. 2000), and parasites (Blader et al. 2001; de Avalos et al. 2002). In light of the recent availability of chicken-tissue-expressed sequence tag (EST) libraries from various tissues, we applied high-throughput gene expression profiling to systematically study host–pathogen interactions in avian coccidiosis, with particular focus on chicken genes associated with protective immunity. In this report, our initial results using cDNA microarray analysis of avian intestinal gene expression changes following primary and secondary infection of chickens with E. acervulina or E. maxima are described.

Materials and methods

Animal, parasites, and experimental infection

Fertilized eggs of White Leghorn SC chickens were purchased from HyVac (Adel, Iowa, USA) and hatched at the Animal and Natural Resources Institute (Beltsville, Md., USA). Hatched chicks were given free access to feed and water and constant light was provided during the entire experimental period. Wild-type strains of E. acervulina and E. maxima, developed and maintained in the Parasite Biology, Epidemiology and Systematics Laboratory (Beltsville, Md., USA), were cleaned by flotation on 5.25% sodium hypochloride and washed three times with Hanks' balanced salt solution (HBSS, Sigma, St. Louis, Mo., USA). Chickens in the experimental group (n=192) were orally infected at 3 weeks of age with 1×104 sporulated oocysts of E. acervulina or E. maxima. Secondary infection was given orally at 9 weeks of age with 2×104 sporulated oocysts of the homologous parasite. Chickens in the experimental group were evenly divided into four groups (n=48) for primary or secondary infection with E. acervulina or E. maxima. Twelve chickens orally inoculated with HBSS alone served as negative controls.

Preparation of RNA from intestinal IELs

Since E. maxima infection results in more severe intestinal pathology than infection with E. acervulina, and oocyst shedding begins at day 5 post-infection (Conway and McKenzie 1991; Choi et al. 1999), IELs were collected from the duodenal region (for E. acervulina) or the section from Meckel's diverticulum to the ileac region (for E. maxima) at 1, 2, 3, and 4 days (12 chickens/day) following primary and secondary infections as described previously (Min et al. 2001). Total RNA from IELs was prepared using TRIzol (Life Technologies, Gaithersburg, Md., USA) and poly(A)+ RNA isolated using the PolyA Tract mRNA Isolation System IV (Promega, Madison, Wis., USA).

cDNA microarrays

cDNA clones (n=2,600) selected from the chicken EST database (http://www.chickest.udel.edu) were applied to nylon membrane using the procedure of Morgan et al. (2001). A subset of these (n=388) was selected as described below, transferred to 100 μl of Luria-Bertani medium containing 100 μg ampicillin/ml in 96-well plates, and incubated overnight at 37 °C. Approximately 1 μl of each culture was removed for PCR amplification using vector primers, and the products analyzed on 1% agarose gels. PCR products were precipitated in ethanol, washed, dissolved in 50% dimethyl sulfoxide, and spotted onto CMT-GAPS-coated glass slides (Corning, Corning, N.Y., USA) using a Cartesian robot model PlySys 5500 PA workstation (Cartesian Technologies, Irvine, Calif., USA) equipped with a Telechem printhead and stealth quill pins.

cDNA probes, hybridization, and data analysis

To prepare hybridization probes, 10 μg of total RNA was mixed with 2.0 μg of oligo (dT), heated for 10 min at 70 °C, and place on ice for 2 min. A cocktail containing 0.5 mM each of dATP, dGTP, and dTTP, 0.1 mM dCTP, 0.1 mM Cy3-dCTP or Cy5-dCTP (Amersham Pharmacia Biotech), 0.5 mM dithiothreitol, and 400 units of Superscript II (Life Technologies) was added. The mixture (30 μl) was incubated at 42 °C for 2 h; labeled probes were purified on Probequant G-50 microcolumns (Amersham Pharmacia Biotech), precipitated, and resuspended in 10 μl of TE buffer (10 mM Tris-HCl, 1.0 mM ethylenediamine tetraacetic acid, pH 8.0). For hybridization, 10 μl each of the Cy3- and Cy5-labeled probes were mixed with 24 μl of hybridization buffer [25% formamide, 5× saline–sodium citrate (SSC), 0.1% SDS, 20 μg human Cot-1 DNA, and 20 μg poly(dA)10] (DeRisi et al. 1996), the mixture heated for 5 min at 95 °C, and cooled to 42 °C.

Microarray slides were fixed under UV irradiation, incubated in prehybridization buffer (25% formamide, 5× SSC, 0.1% SDS, 1% bovine serum albumin) for 45 min at 42 °C, rinsed with distilled water, and dried. Slides were hybridized with labeled probes for 18 h at 42 °C, then washed twice with 2×SSC containing 0.1% SDS for 4 min at 42 °C, and three times with 0.1×SSC for 2 min at room temperature. After a brief centrifugation at 500 g, fluorescence intensities of the slides were measured using a ScanArray 4000 scanner (Packard BioScience, Billerica, Mass., USA) and the data normalized by within print-tip group normalization (http://www.stat.berkeley.edu/users/terry/zarray/Html/normspie.html). Statistical significance was determined using the Significance Analysis of Microarrays (SAM) software (http://www.stat.stanford.edu/~tibs/SAM/index.html) as described (Tusher et al. 2001).

Results

Microarray construction

Poly(A)+ RNA isolated from chicken intestinal IELs at 4 days following primary and secondary infection with E. maxima was labeled with 32P and hybridized to cDNA nylon microarrays containing 1,100 expressed sequence tag (EST) clones from a concanavalin A-activated splenic lymphocyte cDNA library (Tirunagaru et al. 2000) and 1,500 EST clones from a lipopolysaccharide (LPS)-activated macrophage (HD11) cDNA library (http://www.chickest.udel.edu). From the means of two independent hybridization studies, 388 clones (292 concanavalin A-induced, 96 LPS-induced) were selected that showed at least 1.5-fold increased or decreased transcript levels compared with uninfected controls after global normalization (http://www.geneindex.org). These 388 clones and 12 chicken cytokine genes previously cloned in our laboratory (Lillehoj and Lillehoj 2000; Min et al. 2001) were PCR amplified and applied in quadruplicate to glass microarrays for further analysis.

Profiling local gene-expression changes in Eimeria infected chickens

To analyze changes in gene expression in the intestine of chickens infected with E. acervulina or E. maxima, IELs were collected from the duodenum or ileum at 1, 2, 3, and 4 days following primary or secondary infection. Levels of gene transcripts isolated from the IELs were quantified by microarray using print-tip group normalization and analyzed statistically using SAM software (Tusher et al. 2001). Compared with E. maxima infection, E. acervulina primary and secondary infection resulted in up- or down-regulation of more transcripts, and and either parasite induced changes in a greater number of transcripts after primary infection than following secondary infection (Table 1). Specifically, E. acervulina and E. maxima infection affected the levels of 99 and 51 gene transcripts, respectively, following primary infection, and 46 and 25 transcripts, respectively, following secondary infection (Fig. 1A). Conversely, E. acervulina and E. maxima decreased the levels of 88 and 56 gene transcripts, respectively, following primary infection, and 22 and 37 transcripts following secondary infection (Fig. 1B). Comparing primary with secondary infection, E. acervulina and E. maxima commonly induced 34 and nine transcripts, respectively, and commonly repressed 17 and 24 transcripts, respectively (Fig. 1B).

Table 1. Number of differentially expressed genes in Eimeria-infected chickens
Full size table
Fig. 1.
figure 1

A list of significantly induced (A) and repressed (B) genes following Eimeria infection. Chickens at 3 weeks of age were infected with 1×104 sporulated oocysts of Eimeria acervulina or Eimeria maxima, and a second infection was given with 2×104 sporulated oocysts of the same parasite at 9 weeks of age. At 1, 2, 3, and 4 days following primary and secondary infection, RNA was prepared from intestinal intraepithelial lymphocytes (IELs) obtained from the chicken duodenum (E. acervulina) or ileum (E. maxima) and used in microarray hybridization. Bars Number of identical gene transcripts expressed during all time points of primary (lane 1) and secondary (lane 2) infection. Lane 3 Number of identical transcripts expressed at all time points following primary and secondary infection. Open and closed bars Number of identical transcripts expressed in E. acervulina and E. maxima, respectively. Hatched bars Number of identical transcripts expressed in both E. acervulina and E. maxima. The data were obtained from averages of quadruplicate hybridizations and analyzed using SAM software

Full size image

E. acervulina and E. maxima infection induce or repress common gene transcripts

When considering any of the time points examined following primary infection with E. acervulina or E. maxima, the levels of 37 common transcripts were increased while ten transcripts were induced in common following secondary infection (Fig. 1; Table 2). Conversely, 43 common transcripts were repressed following primary infection and ten were reduced after secondary infection with either parasite (Fig. 1; Table 3). When considering all time points examined following primary or secondary infection with E. acervulina or E. maxima, the quantities of five gene transcripts were commonly induced (CMRF35 leukocyte immunoglobulin-like receptor, zinc-finger gene, PmbA homolog, granulysin precursor, cyclophilin A) and eight were repressed (α-actinin, hypothetical protein F39H12.5, spleen mitotic checkpoint BUB3, interferon-induced granylate-binding protein 2, transcription factor NF-YC subunit, transport associated protein 3, α-adaptin, homobox protein HOX-D8) (Fig. 1; Table 2; Table 3).

Table 2. Identical genes induced in both E. acervulina and E. maxima infections. Clone ID comes from http://www.chickest.udel.edu
Full size table
Table 3. Identical genes repressed in both E. acervulina and E. maxima infections. Clone ID comes from http://www.chickest.udel.edu
Full size table

Changes in cytokine genes following Eimeria infection

Cell-mediated immunity, mediated in part by cytokine production, has been shown to be involved in protection against coccidiosis (Choi et al. 1999; Yun et al. 2000b, c; Lillehoj et al. 2000; Min et al. 2001; Li et al. 2002; Miyamoto et al. 2002). Thus, 12 cytokine genes were included in our microarray analysis in order to monitor changes in their corresponding transcripts subsequent to Eimeria infection. As shown in Fig. 2 , transcript levels for the interleukin (IL)-8, IL-15, and lymphotactin genes were increased at all time points examined following primary or secondary infection with E. acervulina or E. maxima whereas IL-18 and osteopontin gene transcripts were repressed. With the exception of transforming growth factor (TGF)-β4, changes in the levels of all cytokine transcripts examined were similar when comparing primary infection by E. acervulina with that by E. maxima. Similarly, after secondary infection, all transcript levels except those for IL-6 and TGF-β4 showed comparable changes when comparing E. acervulina with E. maxima.

Fig. 2.
figure 2

Changes in cytokine transcript levels following Eimeria infection. ○ E. acervulina infection, ● E. maxima infection. DPI Days post-primary infection, DSI days post-secondary infection. Fold changes in transcript levels are expressed as the log2 ratio and were analyzed using SAM software. GenBank accession numbers are: IL-1β (Y15006), IFN-α (U07868), lymphotactin (AF006742), iNOS (D85422), IL-2 (AF017645), osteopontin (X56772), IL-18 (AJ277865), IL-6 (AJ309540), IL-8 (AI980458), IL-15 (AF139097), TGF-β4 (M31160), and IFN- γ (U27465). No data were obtained at 3 DPI with E. maxima for IL-15 and at 4 DPI and 3 DSI with E. acervulina for IL-18

Full size image

Discussion

Eimeria protozoa invade, grow, and reproduce within intestinal epithelial cells of the chicken. In response, the host produces a cellular immune response mediated primarily by local IELs (Bhogal et al. 1989; Lillehoj and Trout 1996; Lillehoj 1998). Intestinal IELs may or may not limit the extent of parasitism depending on a variety of factors including, but not limited to, the species of Eimeria, the infective dose, chicken nutritional status, and environmental stress. As a step toward understanding the complex immune responses resulting from parasite infection, we used cDNA microarrays to profile changes in gene expression in intestinal IELs of Eimeria-infected chickens. In addition, we searched for commonalities in three different parameters: E. acervulina vs. E. maxima, primary vs. secondary infection, and days post-infection. Both species of Eimeria are found commonly in poultry farming and invade different regions of the small intestine (Conway and McKenzie 1991).

In total, we observed up-regulation of 42 chicken gene transcripts and repression of 45 transcripts at days 1–4 following primary or secondary infection with E. acervulina or E. maxima. Generally, primary infection induced more changes in gene expression than did secondary infection, and E. acervulina caused more changes than E. maxima. Of the 42 activated transcripts, five were common among the three parameters analyzed, and of the 45 repressed transcripts, eight were shared. These 13 gene transcripts may encode proteins involved in intestinal anti-parasite immunity since similarities in immune responses among different Eimeria species are well-documented (Lillehoj 1998; Lillehoj and Lillehoj 2000). In this regard, it is interesting to note that some of these transcripts are encoded by immune-associated genes. As many Eimeria-regulated genes do not have defined functions in the infection process, these results provide the foundation for future studies to clarify unknown aspects of the early cellular immune mechanisms following parasite infection.

With regard to expression profile of cytokine genes, previous studies have shown that chicken interferons and interleukins play an important role in the cell-mediated immune response to Eimeria. For example, significantly increased interferon (IFN)-γ mRNA was detected in intestinal cells following infection with E. tenella or E. maxima (Laurent et al. 2001; Yun et al. 2000b). Results of our DNA microarray analysis, in which levels of IFN-γ mRNA increased on days 3 and 4 following primary infection with E. acervulina or E. maxima, supports these findings . Additionally, we observed previously that duodenal IELs obtained from chickens immunized simultaneously with a coccidia DNA vaccine and IL-8 or IL-15 had significantly increased numbers of CD3+ T-cells (Min et al. 2001). Similarly, the number of CD3+ cells was increased in the jejunal mucosa of chicken infected with E. maxima (Laurent et al. 2001). These findings suggest a relationship among Eimeria infection, IL-15, and CD3+ cells. Our microarray data support this conclusion by demonstrating that levels of IL-15 transcripts were up-regulated at all time points examined following primary or secondary infection with E. acervulina or E. maxima.

In conclusion, this study provides the first description of global transcriptional responses elicited in chicken intestinal lymphocytes following Eimeria infection. Our findings clearly indicate that E. acervulina and E. maxima induce similar genetic changes and that the levels of several transcripts were commonly affected in response to both Eimeria species. Furthermore, the expression patterns of cytokine genes followed similar trends after experimental infection. Additional microarray studies with other Eimeria species will provide enhanced insights into the host protective responses against oral infection with Eimeria.