Analysis of Chicken Mucosal Immune Response to Eimeria tenella and Eimeria maxima Infection by Quantitative Reverse Transcription-PCR

Abstract

The recent cloning of chicken genes coding for interleukins, chemokines, and other proteins involved in immune regulation and inflammation allowed us to analyze their expression during infection with Eimeria. The expression levels of different genes in jejunal and cecal RNA extracts isolated from uninfected chickens and chickens infected with Eimeria maxima or E. tenella were measured using a precise quantitative reverse transcription-PCR technique. Seven days after E. tenella infection, expression of the proinflammatory cytokine interleukin-1β (IL-1β) mRNA was increased 80-fold. Among the chemokines analyzed, the CC chemokines K203 (200-fold) and macrophage inflammatory factor 1β (MIP-1β) (80-fold) were strongly upregulated in the infected ceca, but the CXC chemokines IL-8 and K60 were not. However, the CXC chemokines were expressed at very high levels in uninfected cecal extracts. The levels of gamma interferon (IFN-γ) (300-fold), inducible nitric oxide synthase (iNOS) (200-fold), and myelomonocytic growth factor (MGF) (50-fold) were also highly upregulated during infection with E. tenella, whereas cyclooxygenase 2 showed a more modest (13-fold) increase. The genes upregulated during E. tenella infection were generally also upregulated during E. maxima infection but at a lower magnitude except for those encoding MIP-1β and MGF. For these two cytokines, no significant change in expression levels was observed after E. maxima infection. CD3⁺ intraepithelial lymphocytes may participate in the IFN-γ upregulation observed after infection, since both recruitment and upregulation of the IFN-γ mRNA level were observed in the infected jejunal mucosa. Moreover, in the chicken macrophage cell line HD-11, CC chemokines, MGF, IL-1β, and iNOS were inducible by IFN-γ, suggesting that macrophages may be one of the cell populations involved in the upregulation of these cytokines observed in vivo during infection with Eimeria.

Chicken coccidiosis is caused by intracellular protozoan parasites belonging to seven species of Eimeria. These parasites invade and reside in the lining of the intestine or ceca. Parasite development causes diarrhea, morbidity, and mortality, and the impact of coccidiosis on the industry has serious economic consequences. Thus far, chemoprophylaxis has controlled the disease but has been complicated by the emergence of drug resistance. Infection by Eimeria promotes antibody and cell-mediated immune responses. However, cellular immunity mediated by various cell populations, including T lymphocytes, NK cells, and macrophages, plays a major role in disease resistance (27). There is increasing evidence of CD4⁺ and intraepithelial lymphocyte (IEL) involvement during a primary infection, while T-cell receptor α- and β-chain-positive CD8⁺ IEL play a key role in secondary infection (25). The development of a vaccine has been hampered by the lack of understanding of the various components of the host immune system involved in protective immunity.

The low level of homology between chicken genes and their mammalian counterparts has made it difficult to discover immunologically relevant chicken genes. However, there have been increasing numbers of chicken gene sequences appearing in the databases due to the emergence of chicken genome projects. Among the cytokines cloned, one can find genes coding for interleukins (interleukin-1β [IL-1β] [39], IL-2 [36], and IL-8 [20]) and interferons (alpha/beta interferon [IFN-α/β] [34] and IFN-γ [7]) and also for a macrophage growth factor (myelomonocytic growth factor [MGF]) (24) and three isoforms of transforming growth factor β (TGF-β) (16–18). In addition, several members of the chemokine family have recently been cloned: C chemokine (unpublished data), CC chemokines (macrophage inflammatory protein 1β [MIP-1β] [15] and K203 [35]), and CXC chemokines (K60 [35] and IL-8 [20]). A number of receptors have also been identified, including the IL-1 receptor (IL-1R) (12) and a putative chemokine receptor (Chem-R) (13). The development of chicken genome projects in several countries and the use of DNA array technology will undoubtedly expedite the identification of components of the chicken immune response to a variety of pathogens. However, the analysis by reverse transcription-PCR (RT-PCR) of the expression of an available panel of genes will provide initial clues about the development of the immune response to Eimeria infection.

In this study, we analyzed the local immune response of leghorn chickens to two strains of Eimeria commonly found in poultry farming, Eimeria tenella and E. maxima (33), by qualitative RT-PCR followed by a precise quantitative RT-PCR method.

MATERIALS AND METHODS

Infection of chickens with Eimeria.

Chickens used in this study are specific-pathogen-free White Leghorn (PA12) hatched in our animal facilities and kept in wire cages with water and food ad libitum. Three-week-old chickens were orally infected with 2 × 10⁴ oocysts of E. maxima (strain PAPm11) or E. tenella (strain PAPt38). Animals were killed by cervical dislocation 3, 7, or 13 days after infection.

RNA extraction.

Three-centimeter-long intestinal fragments of duodenum (bottom of the duodenal loop), jejunum (3 cm above Meckel's diverticulum), ileum (3 cm below Meckel's diverticulum), or cecum (median part of the organ) were excised and cut longitudinally. To remove intestinal contents, fragments were washed in ice-cold phosphate-buffered saline (PBS) and immediately immersed in TRIzol solution (Life Technologies, Cergy Pontoise, France) for 3 min under agitation. This technique allows the extraction of RNA from cells of the upper layer of the mucosa as assessed by microscopy. Total RNA extraction was performed according to the manufacturer's recommendations.

RNA standards for quantitative RT-PCR.

For the quantitation of mRNA levels of the genes of interest, plasmids coding for truncated mRNA templates (standards) were constructed. In vitro transcription of these plasmids yields RNAs that carry primer sites identical to those that amplify target RNA. However, the distances between specific 5′ and 3′ primer sites and, therefore, the sizes of the PCR amplification products differ from those of the standard and target RNAs. To generate a truncated template, we used a composite primer made of the upstream primer (5′) followed by a sequence complementary to a region located downstream in the RNA. The corresponding PCR product was obtained with RNA extracted 7 days postinfection from the cecum of an E. tenella-infected chicken as a template, using the composite primer and the downstream primer. The amplified fragment was cloned into plasmid pGEMeasy (Promega, Lyon, France). This procedure was performed for each of the 11 plasmid constructs. Finally, to provide a poly(A) tail and a new unique HindIII restriction site at the 3′ end of the coding sequence, the sequence encoded by two complementary oligonucleotides (5′TCGACA₂₀AAGCTTC and 5′TCGAGAAGCTTT₂₀G) was inserted at the SalI site of the plasmids. To generate standard RNA, plasmids were digested with HindIII and transcribed in vitro using T7 RNA polymerase under conditions recommended by the supplier (Eurogentec, Angers, France).

Oligonucleotide primers for PCR amplification.

Sequences of the oligonucleotide primers used for PCR amplification and the sizes of the predicted PCR products from target and standard RNAs are given in Table 1. Primers were designed based on published sequences and obtained from Eurobio (Les Ulis, France). When genomic sequences were available in the databases, primers were selected to either amplify fragments from cDNA that are distinguishable by size from fragments amplified from genomic DNA or span exon-exon boundaries and therefore do not amplify genomic DNA.

TABLE 1.

Primers used for RT-PCR analysis of chicken mRNAs

Target mRNA	Accession no.	Primer		Size of PCR product (bp)
		5′	3′	Target	Standard
β-Actin	L08165	5′-CATCACCATTGGCAATGAGAGG-3′	5′-GCAAGCAGGAGTACGATGAATC-3′	353	271
Chem	L34552	5′-GTCTTCTCCTTGGTCATGGTCA-3′	5′-CAAGGCAGAGCTGGCTCCATAA-3′	403	303
Chem-R	AF029369	5′-AGATGAGAGCAACCGCAGCATC-3′	5′-AAGCCAATGGCCTCTGTCACC-3′	350
COX-2	M64990	5′-GGTGAGACTCTGGAGAGGCAAC-3′	5′-GTTGAACAGAAGCTCAGGGTCA-3′	401	300
IL-1β	Y15006	5′-GGCTCAACATTGCGCTGTAC-3′	5′-CCCACTTAGCTTGTAGGTGGC-3′	350	270
IL-1R	M81846	5′-TGATTCTCAAGAATTTACATCATACAT-3′	5′-CTTCTCCTGCTAAATCATTCCTC-3′	353
IL-8	AJ009800	5′-ATGAACGGCAAGCTTGGAGCT-3′	5′-TCACAGTGGTGCATCAGAATTGA-3′	312	238
IFN-γ	Y07922	5′-GCCGCACATCAAACACATATCT-3′	5′-CAGTAGGAGGTATAAATACTTTC-3′	403	302
iNOS	U46504	5′-AGGCCAAACATCCTGGAGGTC-3′	5′-TCATAGAGACGCTGCTGCCAG-3′	371	285
K60	Y14971	5′-GGGCAAGGCTGTAGCTGCTG-3′	5′-TGGTGTCTGCCTTGTCCAGAAT-3′	290	215
K203	Y18692	5′-ATGAAGCTCTCTGCAGTTGTTCT-3′	5′-TCAGTCCCGCTTGACGCTCTG-3′	269	206
Lymphotactin	AF006742	5′-ATGAAACTCCACGCCACAGTTCT-3′	5′-CTTCTTCTGGTAGTACGTCTTCTG-3′	290
MGF	M85034	5′-CTGCAGCTGTGCTGGCGCTG-3′	5′-CCAGCGAGTCGTGGTACGCG-3′	320	189
MIP-1β	L34553	5′-ATTGCCATCTGCTACCAGACCT-3′	5′-TCAGGTAGCTCTCCATGTCACA-3′	322	230
TGF-β2	X59080	5′-TGCAGGATAATTGCTGCCTGCG-3′	5′-AGCTGCATTTGCAAGACTTTACAAT-3′	305
TGF-β3	M31154	5′-CCCTCCTGCAGGAGAAAATCCT-3′	5′-TCAGATCATGAGTGAATGGCTCC-3′	400
TGF-β4	M31160	5′-ACCTCGACACCGACTACTGCT-3′	5′-CTGCACTTGCAGGCACGGAC-3′	340

Quantitation of mRNA levels.

Quantitative RT-PCR was performed as described by Jung et al. (19). Briefly, serial dilutions of known quantities of standard RNA molecules were mixed with 1 μg of total cellular RNA in a total volume of 20 μl and reverse transcribed at 37°C (19). Two microliters of the reaction mixture was used in a 35-cycle PCR except for β-actin, which was amplified for 28 cycles. Annealing temperature was 61°C for all primers except IFN-γ primers (57°C). Sizes of the PCR amplification products differ by 25 to 30% between standard and target RNAs; thus, the products can be easily separated on a 2% agarose gel and visualized by ethidium bromide staining. Band intensities were quantitated by densitometry (GS-670 imaging densitometer; Bio-Rad, Ivry sur Seine, France). Ratios of the band intensities of the PCR products from the standard RNA and target RNA were plotted against the starting number of standard RNA molecules on a double logarithmic scale. When the ratio of the band intensities equals 1, the number of target RNA molecules is equivalent to the number of standard RNA molecules (19). Data are expressed as the number of target mRNA molecules per microgram of total sample RNA. On every RNA sample, a first set of serial 10-fold dilutions of standard was used in the reaction in order to determine the range in which the gene was expressed. Thereafter, six serial threefold dilutions of standard surrounding the estimated value were used. The quantitative RT-PCR was sensitive to 10³ mRNA molecules/μg of total RNA.

Immunohistochemistry of CD3⁺ positive cells.

Pieces of jejunum were fixed in PBS containing 4% paraformaldehyde and snap-frozen in liquid nitrogen. Seven-micrometer-thick frozen sections were incubated for 30 min with a mouse anti-CD3 antibody (clone CT-3; Southern Biotechnology, Birmingham, Ala.) diluted 1/100 in PBS containing 0.05% Tween 20. After several washes, sections were incubated with a goat anti-mouse-fluorescein isothiocyanate conjugate (Sigma, Saint Quentin Fallavier, France) for 30 min. Sections were slightly counterstained with Evans blue (1/20,000) before microscopic examination.

Isolation of IEL.

Chicken IEL were obtained as described by Bessay et al. (4). In brief, the small intestine between the duodenal loop and the region immediately prior to Meckel's diverticulum was excised, cut longitudinally, and washed in HBSS (Hank's balanced salt solution; Gibco, Cergy Pontoise, France) medium containing 4 g of glucose per liter and 2% fetal calf serum (FCS). Intestinal fragments of each chicken were treated separately, cut into 1- to 2-cm pieces, and incubated for 10 min in the same medium supplemented with 2 mM dithiothreitol in order to eliminate the intestinal mucus. The supernatant was discarded, and the small pieces of intestine were incubated twice for 20 min at 41°C in medium containing 2 mM dithiothreitol and 3 mM EDTA. Cells in the supernatant were washed and passed through nylon wool to remove most epithelial cells and cellular clusters. Cells were further purified on a Ficoll gradient (Sigma) (density of 1.077 g/ml, 30 min, 1,200 × g) to remove red cells. Cell viability was >95% as determined by trypan blue exclusion.

Purification of CD3⁺ IEL by magnetically activated cell sorting.

IEL resuspended in cold HBSS containing 2% FCS and 4 g of glucose per liter were incubated 20 min with the mouse anti-chicken CD3 with a working dilution of 1 μg/10⁶ cells. After two washes, cells were then incubated for 20 min at 4°C with a rat anti-mouse immunoglobulin G1 (2 μl for 10⁶ cells) conjugated with magnetically activated cell sorting (MACS) superparamagnetic microbeads (Miltenyi, Paris, France). Cells were washed twice in PBS containing 0.5% bovine serum albumin and 2 mM EDTA and applied to the column. The cells were purified as instructed by the manufacturer (Miltenyi). RNA was extracted from CD3⁺ cells as described previously. The mini-MACS separation allowed a purity of CD3⁺ cells of 95% as controlled by flow cytometry.

Expression of recombinant IFN-γ in COS7 cells.

Chicken IFN-γ coding sequence was amplified by PCR from RNA extracted from E. tenella-infected ceca and cloned in the pcDNA3 vector (Invitrogen, Groningen, The Netherlands). The following primers were used: sense, 5′-CTCGAATTCACCATGACTTGCCAGACTTACAACT-3′; and anti-sense, 5′-GTCCTCGAGTTAGCGGCCGCTGCAATTGCATCTCCTCTG-3′.

COS7 cells were maintained in growth medium at 37°C (Dulbecco's modified Eagle medium (DMEM) containing 4.5 g of glucose per liter supplemented with 10% FCS, 2 mM l-glutamine, 50 U of penicillin G per ml, and 50 μg of streptomycin per ml). Plasmids coding for IFN-γ or β-galactosidase (β-Gal) as a control (pCMVβ; Ozyme, Montigny-le-Bretonneux, France) were transfected into COS7 cells using LipofectAMINE (Gibco-BRL, Cergy Pontoise, France) as recommended by the manufacturer. Briefly, serum-free DMEM containing lipid-DNA complexes were added to the COS7 cells for 5 h of incubation. FCS was then added to the incubation medium to reach a 10% concentration. Eighteen hours later, the medium was replaced with fresh growth medium. The supernatants containing β-Gal or IFN-γ activity were recovered 48 h later.

Activation of HD-11 cells with recombinant chicken IFN-γ.

HD-11 cells were maintained at 41°C in growth medium (DMEM containing 1 g of glucose per liter supplemented with 8% FCS, 2% chicken serum, 2 mM l-glutamine, 50 U of penicillin G per ml, and 50 μg of streptomycin per ml).

Before RNA extraction, 10⁶ HD-11 cells were activated for 6 h with 2 ml of diluted (1/30) supernatant from IFN-γ- or β-Gal- transfected COS7 cells. At that dilution, β-Gal-transfected cell supernatant had no effect on nitrate (NO₂) and nitrite (NO₃) release, whereas IFN-γ-transfected cell supernatant induced the maximal level (NO_x = NO₂ + NO₃ = 90 μM) as determined by the Greiss reaction according to the previously described protocol (10).

RESULTS

Inflammatory gene expression in different intestinal regions at homeostasis and during infection with E. tenella or E. maxima.

In pathogen-free animals, genes are differentially expressed in organs reflecting the normal physiologic conditions. Figure 1A shows that while β-actin, MIP-1β, IFN-γ, inducible nitric oxide synthase (iNOS), and MGF were expressed at similar levels in the cecum and the jejunum, IL-1β, cyclooxygenase 2 (COX-2), K60, K203, and especially IL-8 were more highly expressed in the cecum. Only the putative chemokine Chem was expressed at higher levels in the jejunum than in the cecum (Fig. 1A). IL-8 and K60, which belong to the CXC chemokine family, were found to be highly expressed (about 10⁸ copies/μg of total RNA) in uninfected ceca (Fig. 3, day 0). Expression was apparently restricted to the part of the intestine colonized by the parasite, i.e., the cecum for E. tenella (Fig. 1B) and the small intestine for E. maxima. Although E. maxima infects more specifically the midintestinal area, the parasite can spread to the duodenal loop and to the lower ileum if the infection is severe (Fig. 1B). RNA extracted from the chicken jejunum was selected for the gene expression analysis during E. maxima infection. The level of cytokine response was dependent on the dose of inoculation. For example, when chickens were infected with 2,000 or 20,000 E. tenella oocysts, IFN-γ expression in the cecum 7 days after infection increased 50- or 300-fold, respectively, compared to the control value. Values were determined by quantitative RT-PCR on a pool of RNA extracted from five animals. Similar data were obtained with E. maxima infection 7 days after infection with 2,000 or 20,000 oocysts, IFN-γ expression in the jejunum increased 85- or 200-fold, respectively, compared to the control value. The higher dose of inoculation (20,000 oocysts) was used for further experiments with both Eimeria strains.

FIG. 1.

Cytokine expression along the chicken intestine. (A) Ratio between gene expression in the cecum and the jejunum at homeostasis. Pools of RNA extracted from the jejunum or cecum of eight PA12 chickens were prepared. Quantitative RT-PCRs were performed for the 11 genes in both RNA pools, and the number of mRNA copies expressed per microgram of total RNA was determined. For each gene, the ratio between gene expression in the cecum and the jejunum was calculated. Data were obtained from a representative experiment performed twice. (B) IFN-γ mRNA expression in the different parts of chicken intestine infected with 20,000 oocysts of E. maxima (E.m) or E. tenella (E.t). Data presented are from representative chickens; similar data were obtained with four other animals for both Eimeria strains. D, duodenal loop; J, jejunum; I, ileum, C, cecum.

FIG. 3.

Quantitative RT-PCR determination of the time course of gene expression during Eimeria infection. Pools of total RNA extracted from the cecum and jejunum of six to eight PA12 chickens infected with E. tenella (A) and E. maxima (B), respectively, were prepared. Quantitative RT-PCRs were performed for the 11 genes on both RNA pools for the four time points, and the number of mRNA copies expressed per microgram of total RNA was determined as described in Materials and Methods. Ratios between gene expression at days 7 and 0 are indicated in parentheses. Data were obtained from a representative experiment performed twice.

Expression of iNOS, COX-2, and inflammatory cytokines in E. tenella-infected ceca and E. maxima-infected jejunum.

A daily time course of gene expression during the infection was performed after inoculation of 3-week-old chickens with 20,000 E. tenella or E. maxima oocysts, which leads to severe infection. Four time points were further selected (days 0, 3, 7, and 13) for the following reasons. On day 3 after E. tenella infection, blood was not detected in the cecum. Maximal upregulation of expression of almost all genes investigated in this study occurred by 7 days postinfection for both Eimeria strains. Finally, by day 13, oocyst excretion had ceased. For each time point, six to eight chickens were used in order to allow for individual variations and to prepare a pool for the quantitative RT-PCR.

Expression of the genes studied during E. tenella infection varied little between chickens in the same treatment group (Fig. 2). During E. maxima infection, there was greater variation in mRNA expression of IL-1β and IFN-γ between the animals at days 3 and 13 postinfection, probably due to a slight shift in the kinetic of the response (Fig. 2). Expression of the proinflammatory cytokine IL-1β was increased 80- and 27-fold 7 days after infection with E. tenella and E. maxima, respectively (Fig. 3). Little or no increase was detected for the IL-1R in infected chickens. Among the chemokines analyzed, lymphotactin, the putative chemokine Chem, and the CXC chemokines K60 and IL-8 exhibited unchanged or modest increases in mRNA expression during infection with either strain of Eimeria. In contrast, the CC chemokines K203 (200-fold) and MIP-1β (80-fold) were strongly upregulated during E. tenella infection, suggesting a role for these molecules in the mucosal immune response. After E. maxima infection, K203 mRNA expression was also clearly upregulated (100-fold) in the jejunum when MIP-1β mRNA expression showed only low upregulation. The putative chemokine receptor (Chem-R) was weakly expressed in the intestinal mucosa compared to the spleen (data not shown). However, a moderate increase in mucosal expression was observed following infection with E. tenella.

FIG. 2.

Qualitative RT-PCR amplification of mRNAs extracted from E. tenella-infected cecum (A) or E. maxima-infected jejunum (B). The data shown are the results of individual RT-PCRs on total RNA extracted from six to eight PA12 chickens at 0, 3, 7, and 13 days postinfection (p.i.). Thirty-five amplification cycles were performed, except for β-actin (28 cycles).

A clear dichotomy between the MGF response to E. tenella and E. maxima infection exists, since a 50-fold upregulation was observed in infected ceca whereas no significant changes were detected in infected jejunum. The strongest upregulation was measured for IFN-γ (300-fold) and iNOS (200-fold) expression 7 days after E. tenella infection. Although IFN-γ was also strongly (200-fold) upregulated during E. maxima infection 7 days after infection, at the same time, iNOS expression was increased only slightly (10-fold) in the jejunum. COX-2 mRNA expression increased 13-fold in the infected ceca, whereas little or no increase was measured in the E. maxima infected jejunum. The mRNA expression of the three TGF-β isoforms did not seem to be regulated during infection with Eimeria, although minor increases can only be seen by quantitative RT-PCR.

Upregulated expression of IFN-γ in CD3⁺ IEL from E. maxima-infected jejunum.

The number of CD3⁺ cells increased in the infected jejunal mucosa, as shown by immunohistochemistry in Fig. 4. Among these cells, an increasing number of CD3⁺ IEL was seen in the infected epithelium (Fig. 4). CD3⁺ IEL from a 7-day E. maxima-infected chicken overexpressed IFN-γ messenger 27-fold compared to CD3⁺ IEL from an uninfected chicken. Quantitative RT-PCR values were 2.0 × 10⁶ ± 1.0 × 10⁶ (n = 4) and 5.4 × 10⁷ ± 2.9 × 10⁷ (n = 4) copies/μg of total RNA from CD3⁺ IEL isolated from control and infected chickens, respectively. β-Actin expression measured in the same samples was stable: 2.1 × 10⁸ ± 1.1 × 10⁸ for all samples analyzed.

FIG. 4.

Localization of CD3⁺ cells in the jejunal mucosa. Frozen sections of control (A) and E. maxima-infected (B) chickens were labeled with a mouse anti-chicken CD3 and then stained with a fluorescein isothiocyanate-conjugated goat anti-mouse antibody. Arrows indicate IEL; arrowheads indicate parasites in the infected mucosa. Slides were slightly counterstained with Evans blue. Magnification, ×400.

IFN-γ upregulates cytokine expression by macrophages.

The presence of a large quantity of IFN-γ in the mucosa is capable of stimulating the synthesis of proinflammatory cytokines and chemokines. We analyzed whether the upregulated gene expression that occurred in vivo following infection with Eimeria could be reproduced by stimulating macrophages with IFN-γ. Lipopolysaccharide (LPS) is a well-known strong inducer of macrophages and was used as positive control. Six hours after stimulation, IFN-γ and LPS activated HD-11 cells upregulated mRNA expression for K203, MIP-1β, IL-1β, MGF, and iNOS (Fig. 5). However, unlike K203, MIP-1β was already well expressed in nonstimulated HD-11 cells (Fig. 5). Although the β-Gal COS7 supernatant dilution used did not induce NO_x release by HD-11 cells as detected by the Greiss reaction, a small nonspecific stimulation of several cytokines and of iNOS gene expression was observed after the incubation. This discrepancy was probably due to the difference in sensitivity of the two methods.

FIG. 5.

Upregulation of K203, MIP-1β, MGF, IL-1β, and iNOS expression in chicken macrophage HD-11 cells stimulated with IFN-γ. HD-11 cells were incubated for 6 h with COS7 supernatant containing β-Gal (−) or IFN-γ (+) activity or incubated for 18 h with or without LPS (5 μg/ml). LPS was used as a positive control of HD-11 cell activation. After RNA extraction and reverse transcription reaction, 35 cycles of PCR were performed and products were resolved on 2% agarose gels.

DISCUSSION

The intestinal mucosa provides both a physiologic and immunologic barrier to pathogens. Coccidia of the genus Eimeria complete their life cycles within the epithelial cells of the chicken intestine. Although E. tenella sporozoites are sometimes found in macrophages or IEL, this is regarded as a route by which the parasite can be translocated within these cells into the lamina propria and gain access to the crypt epithelial cells (37). The first line of defense against Eimeria is provided by the infected epithelial cells and the cells in close contact with them such as IEL and fibroblasts. The RNA extraction method that we used allowed us to detect mainly the immune response in the more apical part of the mucosa, although we cannot exclude some contamination with cells located deeper in the lamina propria.

The inflammation observed in Eimeria-infected intestine is associated with an infiltration of macrophages and T cells (38), accompanied by edema and a thickening of the mucosa (27). IL-1β is a powerful proinflammatory cytokine secreted by many different cell types, with stimulated macrophages being the main producer. IL-1β stimulates the secretion of chemokines by fibroblasts (39), macrophages (32), and epithelial cells (9), which can then attract inflammatory cells including macrophages, neutrophils, and lymphocytes, thus amplifying the immune response. The upregulation of IL-1β mRNA was measured during both E. tenella and E. maxima infection and might contribute to the chemokine upregulation observed. We and others have previously shown that human intestinal epithelial cells upregulate IL-8 mRNA expression after infection with Cryptosporidium parvum and Toxoplasma gondii (6, 22). In the present study, the mRNA levels for the CXC chemokines IL-8 and K60 were unchanged or increased slightly compared to the CC chemokines K203 and MIP-1β. CC chemokines are more specifically involved in the recruitment of macrophages, whereas CXC chemokines participate in the recruitment of neutrophils at inflammatory sites. Our data complement in vivo observations that macrophages are the main inflammatory cells in the Eimeria-infected chicken mucosa (38). Moreover, we have shown that IFN-γ-activated HD-11 cells display upregulated mRNA expression for IL-1β and the CC chemokines. Although macrophages are most probably activated in vivo after Eimeria infection, their relative participation in our RNA extract is not known.

Another set of molecules involved in the mucosal immune response in addition to chemokines are prostaglandins. Prostaglandins are important inflammation mediators and regulators of gastrointestinal fluid secretion (8). Their synthesis from arachidonic acid is dependent on the activities of an enzyme that exists in two isoforms, the constitutive COX-1 and the inducible COX-2. High-level expression of COX-2 can be induced in macrophages and in intestinal epithelial cells by stimulators like IL-1 and TNF (11, 14). In addition, human intestinal epithelial cells produce prostaglandins E₂ and F₂α via the induction of COX-2 following infection with intestinal pathogens such as C. parvum (23). Our present data show that the inducible cyclooxygenase mRNA was moderately upregulated during infection, suggesting that prostaglandin production could occur in response to both strains of Eimeria. However, to confirm that hypothesis, prostaglandins must be measured and their relative contributions to inflammation and diarrhea during coccidiosis must to be investigated.

The TGF-β isoforms are important regulators of inflammation, being proinflammatory at low concentrations and anti-inflammatory at high concentrations (30). These molecules are involved in differentiation and proliferation of T and B cells (21) and have been shown to delay and decrease the barrier disruption caused by C. parvum (31). In a recent study, a slight increase in TGF-β4 mRNA expression was observed in IEL isolated from E. acevulina-infected SC chickens; however, this upregulation was dependent on the chicken strain used (5). Although only qualitative RT-PCR measurements have been performed for the different TGF isoforms, in our hands, no clear upregulation of gene expression seems to occur during E. tenella and E. maxima infection in PA12 chickens.

IFN-γ is a major factor in the development of resistance to Eimeria, as it inhibits E. tenella development in vitro (26) and reduces oocyst production and body weight loss following E. acervulina infection (26, 28). In a recent study, IFN-γ transcript levels were shown to be upregulated in the cecal tonsils, spleen, and intestinal IEL during the course of E. tenella infection (43). In the present study, 7 days after infection, we observed a recruitment of CD3⁺ cells in the lamina propria and the epithelium of the E. maxima-infected jejunum. CD3⁺ IEL isolated from the infected jejunum produced 27-fold more IFN-γ mRNA than CD3⁺ IEL isolated from uninfected jejunum. The strong (200-fold) upregulation of IFN-γ expression in the jejunum of E. maxima-infected chicken was therefore probably due in large part to both the recruitment and stimulation of these cells. This high production of IFN-γ may contribute to clearance of the infection and the development of immunity to reinfection. IFN-γ induces iNOS expression in several cells types, including epithelial cells (40) and macrophages (41). During E. maxima infection, levels of nitrite and nitrate reach peak values at about 6 days postinoculation (3), which concurs with our findings on the levels of iNOS mRNA measured in infected tissues. Although free radical species are produced in response to Eimeria infections, their efficacy against the parasite in vivo is more debatable (1–3). We observed that iNOS mRNA expression was much more important during E. tenella than E. maxima infection. This may contribute to the hemorrhage frequently observed after E. tenella infection, by causing vasodilatation in the cecum (2). The strong upregulation of MGF in infected cecum but not in infected jejunum may contribute to the differences observed in iNOS upregulation in both regions of the intestine. Chickens administered a live recombinant fowlpox virus that expresses MGF exhibited a marked and sustained increase in the number of circulating blood monocytes as well as enhanced phagocytic activity and elevated production of nitric oxide (42). In this study, we showed that recombinant chicken IFN-γ was able to induce both iNOS and MGF mRNA expression in HD-11 cells. These results suggest that IFN-γ and MGF may both contribute to iNOS induction in vivo.

The data presented here give an overview of the immunologically relevant gene-specific response to two strains of Eimeria commonly found in poultry livestock and also provide new insights into a possible use for cytokines as therapeutic agents against this pathogen. The potential uses of cytokine therapy in poultry via delivery with live vectors (viral and bacterial), naked DNA injection, or injection of the recombinant protein is currently being investigated by several groups (26, 29). The populations of cells upregulating cytokine gene expression will have to be identified in order to further characterize the mechanisms by which the natural protective immune response against Eimeria occurs in vivo.