Microarray analysis of selection lines from outbred populations to identify genes involved with nematode parasite resistance in sheep

Cristina Diez-Tascón1, Orla M. Keane1, Theresa Wilson1, Amonida Zadissa1, Dianne L. Hyndman2, David B. Baird3, John C. McEwan2 and Allan M. Crawford2

1 AgResearch, Molecular Biology Unit, Department of Biochemistry, University of Otago, Dunedin
2 AgResearch, Invermay Agricultural Centre, Mosgiel
3 AgResearch, Lincoln, New Zealand


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Gastrointestinal nematodes infect sheep grazing contaminated pastures. Traditionally, these have been controlled with anthelmintic drenching. The selection of animals resistant to nematodes is an alternative to complete reliance on drugs, but the genetic basis of host resistance is poorly understood. Using a 10,204 bovine cDNA microarray, we have examined differences in gene expression between genetically resistant and susceptible lambs previously field challenged with larval nematodes. Northern blot analysis for a selection of genes validated the data obtained from the microarrays. The results identified over one hundred genes that were differentially expressed based on conservative criteria. The microarray results were further analyzed to identify promoter motifs common to the differentially expressed genes. Motifs identified in upregulated gene promoters were primarily restricted to those promoters; however, motifs identified in downregulated gene promoters were also found in the promoters of upregulated genes but not in the promoters of genes whose expression was unaltered. Protein Annotators’ Assistant was used for lexical analysis of the differentially expressed genes, and Gene Ontology was used to look for metabolic and cell signaling pathways associated with parasite resistance. Two pathways represented by genes differentially expressed in resistant animals were those involved with the development of an acquired immune response and those related to the structure of the intestine smooth muscle. Genes involved in these processes appear from our analysis to be key genetic determinants of parasite resistance.

gastrointestinal nematodes; major histocompatibility complex class II; transgelin; actin


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
ECONOMICALLY, THE MOST IMPORTANT DISEASES of grazing livestock are the infections produced by gastrointestinal nematodes, and, as a consequence, research has concentrated on their control (25). Anthelmintic drenching has been an important therapy for the past 50 years; however, reliance on drugs is currently being questioned; parasites resistant to one or multiple classes of anthelmintics are now common in many livestock species (50), and there is a growing demand from consumers for products free of any kind of chemical treatment (21).

An efficient alternative to anthelmintic use is the selection of animals naturally resistant to intestinal nematodes (2, 36, 37, 65). Nematode resistance has moderate heritability, and sheep selection lines with large differences in fecal egg count between resistant and susceptible lines have been generated. This resistance pertains to a variety of nematode species such as Haemonchus contortus (68), Trichostrongylus colubriformis, and Ostertagia species (42). Selecting for resistance would be substantially simplified if the differences between resistance and susceptibility were known at the gene variant level and animals could be selected based on genotype.

The cascade of immune phenomena developed against nematodes is complex and involves a large number of genes in the model systems where it has been studied (24). It is likely that the ruminant response will be similarly complex. In ruminants, resistance to both infective larvae and adult nematodes has been associated with the development of acquired immunity (10). Thus, while spontaneous expulsion of a primary infection rarely occurs in cattle or sheep, it routinely occurs in previously primed animals although it varies considerably between individuals. This variation might be caused by different magnitudes of the same response mechanism, by the expression of different immune mechanisms, or by a combination of both. Parasite rejection can be antigenic in nature but may also involve processes such as mechanical or enzymatic damage inflicted by the host on larval or adult parasites.

Microarray technology has become an important tool to address complex biological questions by means of parallel measurement and analysis of the expression of thousands of genes. Identifying patterns of gene expression and grouping genes into expression classes can provide insight into their biological function and relevance. We have used microarrays to determine the differential gene expression pattern between lambs genetically resistant to parasitic nematodes and those that are susceptible. A variety of analytic methods were used to ascertain the biological relevance of the results. Lexical analysis and Gene Ontology (GO) were used to find biological processes associated with the differentially expressed genes. Motifs were also found in the promoters of the differentially expressed genes, and these motifs could distinguish differentially expressed genes from those whose expression was unaltered.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Selection Lines and Tissue Isolation
Resistant and susceptible Perendale sheep lines have been selected for high and low fecal egg counts since 1986. Four resistant and four susceptible lambs were exposed to two consecutive natural challenges with gastrointestinal parasites as described previously (9, 41). After weaning, lambs were drenched to ensure the absence of parasites in their digestive tract and exposed to a contaminated field until they averaged 800 eggs/g feces. Then they were drenched and challenged again. At the end of the second challenge, lambs were fecal sampled and slaughtered (Fig. 1). A section of the duodenum ~20 cm from the pyloric sphincter was collected. This tissue was used, as the major proportion of economically important gastrointestinal nematodes in sheep are located in the lumen of the duodenum. The number of adult nematodes in the abomasum and duodenum lumen was counted for each animal (Table 1). Samples of duodenum and mesenteric lymph nodes were collected into liquid nitrogen promptly after euthanasia and stored at –80°C until RNA isolation. Procedures were approved by the AgResearch Invermay Animal Ethics Committee, formally constituted under the New Zealand Animal Welfare Act (1999).



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 1. Field challenge protocol. Fecal egg counts for strongyles (FEC) and nematodirus (NEM) were measured, as were postslaughter adult worm burdens. Tissue samples (duodenum and mesenteric lymph nodes) were collected at slaughter.

 

View this table:
[in this window]
[in a new window]
 
Table 1. Parasite burden at slaughter

 
Array Preparation
A number of bovine cDNA libraries were generated from a variety of tissues from both dairy and beef breeds. These 5'-libraries were single-pass sequenced, generating expressed sequence tags (ESTs). A bovine microarray containing 10,204 unique amplified cDNAs, 31 PCR products, and 137 control spots was used. The cDNAs printed on the array predominantly came from peripheral mononuclear, lymph node, peyers patch, small intestinal mucosa, and spleen libraries. The PCR products consisted of 18 amplified bovine and 13 amplified cervine interleukins and cytokines. The cDNAs were amplified by PCR in a 50-µl reaction volume in 96-well plates containing 55 pmol primers (T7 and T3 for cDNA cloned in pBK-CMV or T7 and Sp6 for cDNA cloned in pCMV-SPORT6), 0.2 mM dNTPs, 1.5 mM MgCl2, and 2.5 units AB Red Hot Taq polymerase. After 36 cycles of 94°C for 45 s, 55°C for 45 s, and 72°C for 60 s, products were extended at 72°C for 5 min. The control spots on the array consisted of the negative control for each PCR used to amplify the cDNAs. A small aliquot of every PCR product was run on a 1% agarose gel to verify the PCR reaction. Products were precipitated with 0.2 M sodium acetate and an equal volume of isopropanol. The 96-well plates were centrifuged at 3,500 rpm for 1 h, and the pellets were washed with 70% ethanol and then dried. When required for printing, the pellets were resuspended in water, transferred to 384-well plates, dried, and then resuspended in 3x SSC printing solution (0.45 M sodium chloride, 0.045 M sodium citrate). The amplified products were printed onto poly-L-lysine-coated glass slides using the ESI array robot, with up to 32 split pinheads depositing 0.6 nl with a 100-µm spot size. After printing, the slides were UV irradiated to cross-link the DNA to the polylysine coating.

RNA Isolation and Fluorescent Labeling
Total RNA from duodenum was isolated using TRIzol (Invitrogen) with subsequent RNeasy kit (Qiagen) purification. Labeled cDNA was produced from 20 µg of total RNA by anchored oligo(dT)-primed reverse transcription and incorporation of aminoallyl-dUTP, using Superscript II reverse transcriptase (Life Technologies,). cDNA was concentrated to 20 µl on microcon columns (Millipore) and labeled in the presence of either monofunctional N-hydroxysuccinimide (NHS)-ester Cy5 or Cy3 dyes (Amersham). The unincorporated dyes were removed using the QIA-Quick PCR purification kit (Qiagen).

Slide Hybridization and Scanning
Slides were prehybridized for 1 h at 42°C in a solution containing 5x SSC, 0.1% SDS, and 0.25% BSA. Subsequently, they were rinsed twice with deionized water and once with isopropanol and air dried. The Cy3- and Cy5-labeled probes were denatured at 95°C, combined with ULTRAhyb (Ambion) hybridization buffer, and applied to the slides. Incubation was completed for 20 h at 42°C in humidified CMT-hybridization chambers (Corning). After hybridization, slides were washed for 5 min each, first in 2x SSC and 0.1% SDS, then in 1x SSC, and finally in 0.1x SSC. All wash buffers were filtered through 22-µm filters. Before the scanning, they were spin dried at 1,000 rpm for 5 min. The slides were scanned in a ScanArray 5000 (Packard Biosciences), and the dual images were collected in TIFF format. The combination and processing of the images were performed using the GenePix Pro software (Axon Instruments), which included automated and manual flagging of bad spots.

Microarray Normalization and Analysis
The microarray data were then normalized for each individual slide, following the procedure of Baird et al. (8). To determine which cDNAs were differentially expressed, the values for each corresponding EST were analyzed over the 16 slides of the experiment using analysis of variance, removing dye bias effects (Ref. 8; results are given in Supplementary Table S1, available at the Physiological Genomics web site).1 ESTs with a normalized mean log intensity <9 were excluded, as were ESTs with >8 bad spots out of 16. For each remaining EST, the standardized residual of the normalized log ratio of the mean (SR_mean) was then calculated. This is calculated by dividing the normalized log ratio of the mean (C_logRatio_Mean) of that EST by the standard deviation of the normalized log ratio of the mean for all good ESTs on the array (0.178; Supplementary Table S1). On the basis of the Quantile-Quantile (Q-Q) plot for the normalized log ratio of the mean, ESTs were counted as significantly differentially expressed if 3.3 < SR_mean < –2.5, i.e., at or outside the 95% probability thresholds calculated from Michael’s statistic (39). The false discovery rate (FDR) of the procedure was estimated by first fitting an empirical Bayes estimate (55) for the standard errors of the normalized log ratio of the mean. A mixture model was then fitted to estimate the FDR for the modified T probabilities (48). All the microarray information has been submitted into the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus (GEO) website (http://www.ncbi.nlm.nih.gov/geo/). The accession number for the array is GPL1434. Each of the 16 slides has been submitted, and the accession numbers range from GSM30263 to GSM30278 inclusive. The experiment series has been submitted as GSE1728.

Northern Analysis
Total RNA was obtained as described previously from duodenum and mesenteric lymph nodes. Twenty micrograms of RNA per sample were separated by denaturing gel electrophoresis and transferred onto a nitrocellulose membrane using the NorthernMax kit (Ambion). The selected cDNA inserts were obtained from the original plasmid by PCR and resequenced to ensure they represented the expected gene. To use them as probes for the subsequent hybridization, inserts were purified with High Pure columns (Roche Applied Science) and labeled with [{alpha}-32P]dCTP by use of the RTS RadPrime DNA labeling system (Life Technologies). Hybridization was completed using standard procedures (16). The membranes were exposed to autoradiography film and Fugifilm BAS-MP imaging plates. The phosphorimage signal was quantitated using the Fugix MacBAS v.2.0 imaging software.

Microarray Results Interpretation
To annotate the bovine cDNAs, all the EST sequences, including those publicly deposited in NCBI, were assembled into contigs using CAP3 (29) after an initial clustering step using BLAST (4). These contigs were used for all future ortholog identification.

Protein Annotators’ Assistant.
The contig to which each EST belonged was blasted against the SwissProt database (release 44) and the top hit extracted. Any match with an expectation value (E value) >1 x 10–10 was excluded. Of the 50 upregulated cDNA clones, 37 had SwissProt hits with an E value <1 x 10–10. Twenty-seven of these were unique hits. Of the 112 downregulated clones, 67 had SwissProt hits with an E value <1 x 10–10. This corresponded to 65 unique hits. The SwissProt hits for each data set were entered into Protein Annotators’ Assistant (PAA) on the Expasy web site (http://www.ebi.ac.uk/paa/), and keywords with a P value < 0.001 are reported. SwissProt IDs for the same number of clones, chosen at random from the array, were also submitted to PAA and the keywords associated with these genes found. A random sample was analyzed to observe any underlying lexical bias of the genes printed on the array, and significant keywords observed in common between the random sample and the data sets are marked with an asterisk. Simple stemming and phrase construction are undertaken by PAA.

GO analysis.
The corresponding human RefSeq for each contig represented on the array was obtained using BLAST (E <1 x 10–10). Of the 10,204 unique ESTs printed on the array, 8,032 had a corresponding RefSeq with 3,455 unique RefSeqs. The up- and downregulated RefSeqs were submitted to EASEonline (http://david.niaid.nih.gov/david/) along with all the RefSeqs on the array. Expression analysis systematic explorer (EASE) calculates overrepresented functional gene categories in the data sets compared with all the genes on the array. Statistically significant GO terms (P < 0.05) are reported.

Motif identification by comparative expression (MICE).
For selected cDNAs, a region of 1,500 bp upstream of the transcription start site for each of the corresponding human RefSeq genes was retrieved from the University of California-Santa Cruz human genome browser (http://genome.ucsc.edu/). This sequence was assumed to contain the human promoter region, as ~75% of human RefSeq core promoters are reported to lie within 1,500 bp of the transcription start site (34). The sequences were masked automatically through the retrieval process for repetitive regions. The corresponding up- and downregulated promoter regions were then analyzed using the MEME motif detection program (6). Motifs had to range in size from 5 to 15 bp, the reverse complement was allowed, and the first 15 motifs were identified. Subsequently, the same up- and downregulated promoter regions plus a randomly selected set of promoter regions (n = 142) were analyzed using MAST (7), with the motifs identified from the two previous analyses, and again allowing the reverse complement of a motif. The results obtained were analyzed with a chi-square test, based on the frequency of RefSeq promoters with a combined up- or downregulated motif significance of P < 0.0001 for each individual promoter in each of the three groups. Finally, the combined motif consensus probability values for matches in each group with values less than P < 0.0001 were analyzed, using the nonparametric Mann-Whitney U-test.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The microarray experiment examined gene expression in duodenum tissue from four resistant and four susceptible Perendale lambs. The lambs were from selection lines established in 1986, which, after 10 years of selection, showed a 7.3-fold difference in parasite burden after natural challenge (41). A full description of the animals is in MATERIALS AND METHODS, and their mean parasite burden and its composition is shown in Table 1.

A factorial design that compared every individual from the resistant group with every one of the individuals from the susceptible group was used. Therefore, every sample was replicated four times, and the number of slides used in the experiment was 16. In every case, two of the four replicates were labeled with Cy3 and two with Cy5, to minimize the bias related to the incorporation of any particular dye (Fig. 2). The labeled cDNA was hybridized to a bovine microarray (see MATERIALS AND METHODS). Cross-species hybridization depends on a number of factors including the length of the probe and the percentage of similarity between the orthologous sequences and the type and distribution of base mismatches (56). Hybridization of sheep cDNA against bovine microarrays was expected to produce specific hybridization under the conditions used in this experiment, because sheep and cattle sequences are highly homologous, especially in the coding region (96% average homology; Ref. 38). It has previously been demonstrated that stringent hybridization conditions, such as those used in this experiment, allow cross-species hybridization between species with sequence homology of 96% or less (1, 56, 63).



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 2. Microarray experimental design. Each susceptible lamb (1S to 4S) was tested against each resistant lamb (5R to 8R) following a factorial design. The design included a dye swap to minimize bias. The 16 resulting slides and the dyes used in every case are shown (Cy3, open cells; Cy5, shaded cells).

 
Of the 10,204 cDNAs analyzed in the present work, 126 showed highly significant differential expression between the resistant and susceptible animals, based on analysis with conservative criteria. Of these, only 14 were more highly expressed in resistant animals compared with susceptible animals (hereafter referred to as upregulated genes), whereas 112 had reduced expression in the resistant line compared with the susceptible line (hereafter referred to as downregulated genes). This distribution is illustrated in the Q-Q plot of the normalized log ratio of the mean (Fig. 3). An FDR of 10% was estimated for this procedure, as described in MATERIALS AND METHODS. To check the quality of the cDNA libraries, a total of 45 clones, representing the extreme values of differential expression, were resequenced. This revealed five sequences that did not correspond to the expected sequence, giving a percentage of misassignment of ESTs of 11.1%. This is consistent with an independent estimate of the error rate of these libraries (Wilson T and Hyndman DL, unpublished results). Any clone that was found to be incorrect when resequenced was not included in any further analysis and was not reported.



View larger version (17K):
[in this window]
[in a new window]
 
Fig. 3. Differenced normal Q-Q plot for the normalized log ratio of the mean plotted against the standardized residuals (95% confidence limits are plotted in red). Data are from the combined results of the 16 resistant vs. susceptible microarray slides. Downregulated genes are at left, and upregulated genes are at right.

 
The human RefSeq (RefSeq release 5.0; http://www.ncbi.nlm.nih.gov/RefSeq/) corresponding to the contig was identified with BLAST. An initial cut-off E value of 1.0 x 10–10 was applied. In cases where the clone matched multiple human transcripts, the best hit was listed; however, further work would be required to unequivocally determine which human-equivalent RefSeq is differentially expressed in sheep duodenum. For the resistant line, 12 of the 14 upregulated cDNAs had an associated human RefSeq. Interestingly, four represent the smooth muscle gene transgelin (TAGLN), while two match the major histocompatibility complex (MHC) class II gene DR-{alpha} (HLA-DRA). The other six genes corresponded to the chemokine (C-X-C motif) receptor-4 (CXCR4), the lysine-deficient protein kinase-1 (PRKWNK1), a hypothetical protein (LOC388078), a Sushi domain-containing protein (SUSD3), a C3H-like zinc finger protein (ZFP36L2), and a solute carrier family member (SLC35D2).

Determination of the number of genes that show significant differential expression was not trivial, as many parameters needed to be considered. Northern analysis confirmed that genes other than the original 14 we considered upregulated showed differential expression (see Fig. 4, B and C). We therefore chose to analyze the 50 most upregulated clones as a representative sample. In total, 40 of these had an associated RefSeq where 6 were likely to be different MHC II genes (DQA1, DQB1, DR-ALPHA) and 5 TAGLN. Twenty-eight unique genes were represented in this data, and these genes are listed in Table 2. Of the 112 downregulated cDNAs, 85 had an associated RefSeq, representing 82 unique genes with a variety of biological functions. Table 3 summarizes some of the putative downregulated genes; the complete list is available as Supplementary Table S2.



View larger version (105K):
[in this window]
[in a new window]
 
Fig. 4. Northern blot of differentially expressed genes. RNA was extracted from duodenum tissue and mesenteric lymph node, separated by denaturing gel electrophoresis, transferred to nylon membrane, and probed for the following genes. A: TAGLN. B: ACTG2. C: HLA-DQA1. D: REG4. (See text for gene descriptions.) Equal loading of the gels is shown in E. Where more than one transcript is visible, the relevant transcript is indicated with an arrow.

 

View this table:
[in this window]
[in a new window]
 
Table 2. Upregulated genes

 

View this table:
[in this window]
[in a new window]
 
Table 3. Downregulated genes

 
Northern Blotting Analysis
Northern analysis was used to verify the differential expression of a selection of genes and validate the results obtained by the microarray technique (Fig. 4, A–D). Denaturing gels were loaded with total RNA obtained from duodenum and mesenteric lymph node for each of the animals used in the microarray experiment. Mesenteric lymph RNA was included, as our original microarray experiment had identified the differential expression of a large number of immune function genes. Differential expression was confirmed for the muscle-specific genes TAGLN and ACTG2 in duodenum, with more intense bands observed for the resistant group (Fig. 4, A and B). Expression of TAGLN was increased 2.3-fold (P < 0.001) and expression of ACTG2 was elevated 2.7-fold (P < 0.001) in the resistant animals compared with the susceptible animals. None or very limited expression of these genes was observed in lymph node. The MHC class II-related gene, HLA-DQA1, also showed a higher level of expression in the duodenum of the resistant lambs (1.5-fold, P = 0.014) as expected (Fig. 4C). This gene was expressed in lymph node, although differential expression in this tissue was not evident. Regenerating gene type 4 (REG4) was expressed only in the duodenum and showed a higher level of expression in two samples belonging to the susceptible group (1.8-fold, P = 0.014; Fig. 4D).

PAA
PAA (66, 67) was used to find keywords associated with the differentially expressed genes to aid in the interpretation of the results. This program derives keywords associated with the differentially expressed genes and compares their frequency among these genes to their frequency in the SwissProt database, thereby identifying statistically significant words. Terms pertaining to muscle and immune function were commonly associated with the group of upregulated genes (Table 4). A number of themes were identified in the downregulated genes (Table 5). Most of the themes seem to be specific to either the up- or downregulated genes. In fact, the theme "muscle function" is only associated with the group of genes that are upregulated, whereas the themes "calcification," "pancreas," and "cytoskeleton" are only associated with the downregulated genes. The theme calcification was due to the differential expression of a number of genes, including keratin 14 and REG4, a gene once thought to be an inhibitor of calcium carbonate precipitation (12, 26) but now known not to be involved in that process (22). The identification of the theme pancreas can be explained by the differential expression of two cDNAs corresponding to REG4. This gene, along with other members of the Reg family, has been associated with the gastrointestinal mucosal response to injury in inflammatory bowel disease and colonic adenocarcinoma (23, 27).


View this table:
[in this window]
[in a new window]
 
Table 4. Significant keywords associated with genes that are upregulated in parasite-resistant animals

 

View this table:
[in this window]
[in a new window]
 
Table 5. Significant keywords associated with genes that are downregulated in parasite-resistant animals

 
Interestingly, "immune function," the only common theme for both up- and downregulated genes, contains only one common keyword, "immunoglobulin," which has also been shown to be significant for a random sample of genes on the array (see MATERIALS AND METHODS). It therefore appears that the entire microarray is biased toward this keyword. Unfortunately, PAA cannot currently transparently correct for these biases, because it uses word frequencies derived from all SwissProt entries and not just from genes represented on the microarray. The high representation of keywords related to MHC for the group of genes upregulated in the resistant animals is worthy of note. Similar keywords are not found in the downregulated genes, where the most represented keywords in this category are related to lectins.

Motif Identification by Comparative Expression
The putative promoter regions for 27 of the 28 unique human RefSeq genes corresponding to the upregulated group and 75 of the 82 unique human RefSeq genes in the downregulated group were analyzed to identify common cis-regulatory motifs. This is under the assumption that, because regulatory elements are evolutionarily important, they would be conserved across species. The analysis was performed using MEME (6). A total of five significant motifs were identified, two in the upregulated group and three in the downregulated group (P < 0.05) (Table 6; more in-depth results are given in Supplementary Table S3). The motifs identified in the promoters of upregulated genes (including those motifs that were nonsignificant) were more frequent in the upregulated gene promoters, using MAST (7) under standard stringent conditions [proportion (Pr) = 0.48], than in a randomly selected group (Pr = 0.08, P = 1 x 10–7) or the downregulated group (Pr = 0.05, P = 1 x 10–7), using a chi-squared test (Table 7). Also, the motifs present were more similar to the motif consensus in the upregulated genes than in the random group (P = 0.006) or the downregulated group (P = 0.005), using the nonparametric Mann-Whitney U-test.


View this table:
[in this window]
[in a new window]
 
Table 6. Significant MEME motifs detected in corresponding human RefSeq promoter regions from up- and downregulated cDNAs

 

View this table:
[in this window]
[in a new window]
 
Table 7. MAST analysis

 
The motifs identified in the downregulated set of genes’ promoter regions appear more frequently in the promoter region of these genes under standard stringent conditions (Pr = 0.31) than in the randomly selected group (Pr = 0.13, P = 0.001). However, these motifs are also found in the promoter regions of the upregulated genes (Pr = 0.30, not significant). The motifs, where present, were not significantly different from the consensus motif in any group. The results from the MAST analyses are provided in Supplementary Table S4.

GO Analysis
Statistically significant GO terms associated with genes upregulated and downregulated in parasite-resistant lambs (Tables 8 and 9) were found, using EASE (28). For the network "biological process," GO terms associated with the upregulated group of genes refer mainly to antigen processing and presentation and signal transduction. In contrast, GO terms for the downregulated genes are associated mostly with ion transport. For the "molecular function" network, GO terms relating to MHC class II receptor activity are associated with the upregulated genes, whereas terms relating to cation binding as well as the cytoskeleton describe the downregulated genes. No GO terms were associated with the "cellular component" network in the upregulated genes, whereas extracellular space was associated with the downregulated genes.


View this table:
[in this window]
[in a new window]
 
Table 8. Statistically significant GO terms associated with genes upregulated in parasite-resistant animals

 

View this table:
[in this window]
[in a new window]
 
Table 9. Statistically significant GO terms associated with genes downregulated in parasite-resistant animals

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Genes with elevated expression levels in the duodenum of genetically resistant and susceptible lambs have been identified by microarray analysis. In total, over 100 different genes have been identified from this experiment. Our findings were confirmed by Northern analysis of a subset of the genes. Once we became confident that the microarray was detecting real gene expression differences, deciding which processes are critical to resisting or becoming susceptible to nematode infection became our focus. Of particular interest was the identity of genes in the susceptible lambs that were likely to be associated with the cause of susceptibility rather than those associated with the consequences or pathology of the nematode infection. To do this in a balanced way, trying to avoid personal biases and "favorite genes," we have examined three approaches.

Initially, keywords associated with the up- and downregulated genes were found using PAA. The advantage of this technique is that it uses an extremely well-annotated set of genes and has strict and constantly updated vocabulary rules (5). Alternatives such as lexical analysis in sources like PubMed and Online Mendelian Inheritance in Man (OMIM), although more comprehensive, suffer the problem of alternative descriptions for the same process. However, this technique is still subject to the insidious problem of assigning gene function to sequences by homology without functional studies being undertaken. This approach was used to identify biological themes associated with the different data sets. PAA identified a large number of muscle- and immune-related keywords for genes upregulated in parasite-resistant lambs. In particular, the immune keywords related to immunoglobulins and MHC class II histocompatibility genes. This indicates that these processes are important in mediating, or are consequential to, resistance to gastrointestinal nematodes. Immune-related keywords were also found to be associated with the downregulated genes; however, in this case, the actual words differed with the emphasis in the downregulated genes being on lectin genes. The word immunoglobulin itself, which was associated with both up- and downregulated genes, was also found to be significantly associated with a random sample of genes analyzed from the microarray. Therefore, it would appear that more immunoglobulin-associated genes are printed on the array than one would expect by chance. Considering that the majority of the cDNAs for the array came from immune tissue libraries, this is not entirely surprising. No other immune keywords were observed in common between the random sample and the data sets. In fact, the only other keywords common to the random sample and the data sets are "lysosome," "repeat," and "signal," indicating that PAA was successful at identifying keywords and themes specific to the dataset.

Analysis of promoter motifs based on expression patterns and DNA sequence within a species has previously been utilized (15, 49). However, the motif identification by comparative expression (MICE) procedure, using homologous human RefSeq promoter regions based on expression of ovine genes, is, to our knowledge, a previously undescribed method for examining expression array data and has wide applicability to a range of experimental situations. This method can be readily validated upon completion of the bovine genome. The major benefit of this comparative technique is that it works across mammalian species and will provide a valuable aid to those who work with a species whose genome sequence is unavailable. Unlike literature-based techniques, de novo motif identification and analysis are not biased by the amount of previous work on a gene or gene family. It also has the benefit that it can identify subsets of up- or downregulated genes that may be involved in a specific pathway and may identify pathways not previously known to be associated with a given biological process. The results suggest that the promoters of the upregulated genes have combinations of motifs that distinguish them from downregulated or unaltered genes. This may indicate that the signaling pathways these unique motifs represent may have upregulated gene expression and directly influence host resistance. Tracing these pathways could identify the causal variants in the signaling pathways influencing genetic resistance. The motifs identified in the promoters of downregulated genes were also found in the upregulated genes’ promoters, but their presence could be used to distinguish these groups from genes whose expression patterns are unaltered by genetic selection and infection. The inference we can draw is that the signaling pathways these putative motifs represent may be responding to the level of infection present, although in some cases the level of gene expression may also have direct effects on host resistance.

GO terms significantly associated with the up- and downregulated genes were found using EASEonline. Once again, GO terms related to antigen processing via the MHC class II system were significantly associated with the genes upregulated in parasite-resistant animals. This further supports the hypothesis that the development of an acquired immune response is critical to developing resistance to parasitic nematodes, although it may be consequential. These data also give credence to the results obtained using PAA. No GO terms relating to muscle function were found associated with this dataset. This is most likely due to the fact that PAA and EASE count the number of genes differently. Whereas PAA will, for example, count all five copies of the TAGLN gene, EASE counts this gene only once. The GO terms associated with the downregulated genes related predominately to ion transport. This is due to the downregulation of two members of the solute carrier family along with ATP synthase and ATPase. The term cytoskeleton is associated with the downregulated genes, using both EASE and PAA.

In summary, the approaches used in this study lead us to believe that our identification of up- and downregulated gene groups from microarray data is describing real events, because all three data interpretation methods identify that these genes have properties that make them significantly different from randomly selected genes on the array. The results also reveal two processes that are important in modulating genetic resistance to parasitic nematodes: the development of an acquired immune response and smooth muscle function, although it is more difficult to identify causal pathways.

The observation that MHC class II genes are more highly expressed by the genetically resistant lambs is consistent with the idea that a key difference between susceptible and resistant animals lies in their ability to mount an early humoral response. In the mouse, a clear interaction between parasites and the MHC has been reported for the helminth Nippostrongylus brasiliensis, which is able to suppress the expression of MHC class II molecules as a mechanism of evasion (20). Evolutionary studies in wild sheep populations infected with gastrointestinal nematodes have led to the hypothesis that the high level of diversity at the MHC locus in sheep is a consequence of antagonistic coevolution between hosts and parasites (47). However, an obvious relationship between MHC alleles and antigen recognition has not been demonstrated in sheep (35), although a correlation has been found in mouse (31). The MHC chromosomal area has also been a common candidate in the search for markers of resistance useful in ruminant selection programs. In this regard, several studies have found an association between parasite resistance and MHC markers (30, 33, 44, 45, 46, 52, 58), although lack of association (13, 18, 57) or genetic linkage (19) has also been reported. Taking into account the contradictory results found in the literature, we suspect that the increased expression of MHC class II in the resistant lambs may only be effective in reducing parasite infection if combined with a set of MHC alleles that are efficient at presenting parasite antigens.

Interestingly, the homeostatic chemokine receptor (CXCR4; upregulated in the genetically resistant animals) is involved in a variety of biological processes (43) and has been the subject of intense study, because it acts as coreceptor for the T-cell line-tropic human immunodeficiency virus (11). Recent studies strongly suggest that CXCR4 channels the antigens into the endocytic pathway for presentation on MHC class II molecules (51). To our knowledge, there is no previous report in the literature that relates this receptor with the immune response to parasites.

Two genes differentially expressed in the intestinal tissue of resistant animals are related to the structure and function of the enteric smooth muscle. In our study, the most represented gene is transgelin (TAGLN), a cytoskeletal protein also known as SM22-{alpha}, WS3–10, and p27, which is expressed abundantly in visceral and vascular smooth muscle cells (3, 32, 53, 59). The second transcript corresponds to actin-{gamma}2 (ACTG2), a specific enteric actin isoform (40). It has been reported that TAGLN binds directly to actin filaments, and its possible role is in the organization of the spatial relationships of actin filaments in smooth muscle cells (54). The simultaneous upregulation of both genes suggests that the resistant lambs may have increased enteric motility. Interestingly, cytoplasmic actin, a ubiquitous actin isoform, is downregulated. Our hypothesis is that this finding might be related to the progression of endocytosis events, whereas the expression of enteric actin and transgelin in the resistant line seems to be more related to the thickening of smooth muscle layers associated with an enhanced contractility of intestinal muscle.

Experimental infections in ruminants show a variation in the speed at which the parasite is expelled from the host (62). In mouse models infected with Trichinella spiralis, profound thickening of smooth muscle layers, due to hypertrophy and hyperplasia, has been observed (14) and is associated with enhanced contractility of intestinal muscle to facilitate the eviction of worms (17). The magnitude of this contractility has been found to be greater in mouse strains that expel the parasite rapidly than in those that expel the parasite slowly (60). Also in mouse, it has been shown that CD4+ T-cells and MHC class II expression influence worm expulsion and increased intestinal muscle contraction during T. spiralis infection (61). Although this is the first time that TAGLN has been associated with the response to gastrointestinal parasites, the increase of other smooth muscle-specific proteins after nematode infection (in particular, actin and myosin heavy chain) was shown by Weisbrodt et al. (64). In addition, this work showed pretranslational upregulation of gene expression for actin isoforms at the infection area.

The results of this study have outlined a number of interesting observations and results. Further research will be needed, however, to study parasite-resistant animals that are not part of the selection lines (to account for random genetic drift). It is also necessary to distinguish genes responsible for genetic resistance from those elevated simply because of the level of infection. This can be done by rearing animals in a nematode challenge-free environment from birth and ascertaining those genes that are constitutively differentially expressed between resistant and susceptible animals in the absence of a prior challenge.

Our results suggest that highly significant and validated results can be obtained from microarray experiments using genetically divergent outbred sheep lines selected for host resistance to parasites. They also suggest that, at least in some circumstances, cDNA microarrays from one mammal can successfully be used in closely related species effectively. Only minimal numbers of animals are required if appropriate experimental design and analysis techniques are used. We also have provided a robust post hoc interpretation that conclusively demonstrates that the up- and downregulated genes identified have properties distinct from genes randomly selected on the array using three separate tests.


    ACKNOWLEDGMENTS
 
We thank Alan McCulloch for assistance with the EST database and preparation of files for submission into NCBI. We also thank Roger Wheeler, Gordon Greer, Wendy Bain, Mary Wheeler, Chris Morris, and Kevin Knowler for assistance with maintenance of the Perendale selection lines, the animal experiments, and the fecal egg count and adult worm burden data collection.


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: A. M. Crawford, AgResearch, Invermay Agricultural Centre, Mosgiel, Private Bag 50034, New Zealand (E-mail: allan.crawford{at}agresearch.co.nz).

10.1152/physiolgenomics.00257.2004.

1 The Supplemental Material for this article (Supplemental Tables S1–S4) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00257.2004/DC1. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Adjaye J, Herwig R, Herrmann D, Wruck W, Benkahla A, Brink TC, Nowak M, Carnwath JW, Hultschig C, Niemann H, and Lehrach H. Cross-species hybridization of human and bovine orthologous genes on high density cDNA microarrays. BMC Genomics 5: 83, 2004.[CrossRef][Medline]
  2. Albers GA and Gray GD. Breeding for worm resistance: a perspective. Int J Parasitol 17: 559–566, 1987.[CrossRef][ISI][Medline]
  3. Almendral JM, Santaren JF, Perera J, Zerial M, and Bravo R. Expression, cloning and DNA sequence of a fibroblast serum-regulated gene encoding a putative actin-associated protein (p27). Exp Cell Res 181: 518–530, 1989.[CrossRef][ISI][Medline]
  4. Altschul SF, Gish W, Miller W, Myers EW, and Lipman DJ. Basic local alignment search tool. J Mol Biol 215: 403–410, 1990.[CrossRef][ISI][Medline]
  5. Andrade MA and Valencia A. Automatic extraction of keywords from scientific text: application to the knowledge domain of protein families. Bioinformatics 14: 600–607, 1998.[Abstract]
  6. Bailey TL and Elkan C. Fitting a mixture model by expectation maximization to discover motifs in biopolymers. Proc Int Conf Intell Syst Mol Biol 2: 28–36, 1994.[Medline]
  7. Bailey TL and Gribskov M. Combining evidence using p-values: application to sequence homology searches. Bioinformatics 14: 48–54, 1998.[Abstract]
  8. Baird D, Johnstone P, and Wilson T. Normalization of microarray data using a spatial mixed model analysis which included splines. Bioinformatics 20: 1–10, 2004.[Abstract/Free Full Text]
  9. Baker RL. Genetics of disease resistance in small ruminants in Africa. In: Breeding for Resistance to Infectious Diseases in Small Ruminants, edited by Gray GD, Woolaston RR, and Eaton BT. Canberra, Australia: ACIAR, 1995, p. 119–138.
  10. Balic A, Bowles VM, and Meeusen EN. The immunobiology of gastrointestinal nematode infections in ruminants. Adv Parasitol 45: 181–241, 2000.[CrossRef][ISI][Medline]
  11. Berger EA, Murphy PM, and Farber JM. Chemokine receptors as HIV-1 coreceptors: roles in viral entry, tropism, and disease. Annu Rev Immunol 17: 657–700, 1999.[CrossRef][ISI][Medline]
  12. Bernard JP, Adrich J, Montalto G, De Caro A, De Reggi M, Sarles H, and Daghorn JC. Inhibition of nucleation and crystal growth of calcium carbonate by human lithostathine. Gastroenterology 103: 1277–1284, 1992.[ISI][Medline]
  13. Blattman AN, Hulme DJ, Kinghorn BP, Woolaston RR, Gray GD, and Beh KJ. A search for associations between major histocompatibility complex restriction fragment length polymorphism bands and resistance to Haemonchus contortus infection in sheep. Anim Genet 24: 277–282, 1993.[ISI][Medline]
  14. Blennerhassett MG, Vignjevic P, Vermillion DL, and Collins SM. Inflammation causes hyperplasia and hypertrophy in smooth muscle of rat small intestine. Am J Physiol Gastrointest Liver Physiol 262: G1041–G1046, 1992.[Abstract/Free Full Text]
  15. Bussemaker HJ, Li H, and Siggia ED. Regulatory element detection using correlation with expression. Nat Genet 27: 167–171, 2001.[CrossRef][ISI][Medline]
  16. Church GM and Gilbert W. Genomic sequencing. Proc Natl Acad Sci USA 81: 1991–1995, 1984.[Abstract]
  17. Collins SM. The immunomodulation of enteric neuromuscular function: implications for motility and inflammatory disorders. Gastroenterology 111: 1683–1699, 1996.[ISI][Medline]
  18. Cooper DW, van Oorschot RAH, Piper LR, and Le Jambre LF. No association between the ovine leucocyte antigen (OLA) system in the Australian Merino and susceptibility to Haemonchus contortus infection. Int J Parasitol 19: 695–697, 1989.[CrossRef][ISI][Medline]
  19. Crawford AM, McEwan JC, Dodds KG, Wright CS, Bisset SA, Macdonald PA, Knowler KJ, Greer GJ, Green RS, Shaw RJ, Paterson KA, Cuthbertson RP, Vlassoff A, Squire DR, West CJ, and Phua SH. Resistance to nematode parasites in sheep: how important are the MHC genes? Proc Assoc Adv Anim Breed Genet 12: 58–62, 1997.
  20. Dainichi T, Maekawa Y, Ishii K, Zhang T, Nashed BF, Sakai T, Takashima M, and Himeno K. Nippocystatin, a cysteine protease inhibitor from Nippostrongylus brasiliensis, inhibits antigen processing and modulates antigen-specific immune response. Infect Immun 69: 7380–7386, 2001.[Abstract/Free Full Text]
  21. Dalton JP and Mulcahy G. Parasites vaccines—a reality? Vet Parasitol 98: 149–167, 2001.[CrossRef][ISI][Medline]
  22. De Reggi M, Gharib B, Patard L, and Stoven V. Lithostathine, the presumed pancreatic stone inhibitor, does not interact specifically with calcium carbonate crystals. J Biol Chem 273: 4967–4971, 1998.[Abstract/Free Full Text]
  23. Dieckgraefe BK, Stenson WF, Korzenik JR, Swanson PE, and Harrington CA. Analysis of mucosal gene expression in inflammatory bowel disease by parallel oligonucleotide arrays. Physiol Genomics 4: 1–11, 2000.[Abstract/Free Full Text]
  24. Garside P, Kennedy MW, Wakelin D, and Lawrence CE. Immunopathology of intestinal helminth infection. Parasite Immunol 22: 605–612, 2000.[CrossRef][ISI][Medline]
  25. Gasbarre LC and Miller JE. Genetics of helminth resistance. In: Breeding for Disease Resistance in Farm Animals, edited by Axford RFE, Bishop SC, Nicholas FW, and Owen JB. Wallingford, UK: CAB International, 2000, p. 129–152.
  26. Geider S, Baronnet A, Cerini C, Nitsche S, Astier JP, Michel R, Boistelle R, Berland Y, Daghorn JC, and Verdier JM. Pancreatic lithostathine as a calcite habit modifier. J Biol Chem 271: 26302–26306, 1996.[Abstract/Free Full Text]
  27. Hartupee JC, Zhang H, Bonaldo MF, Soares MB, and Dieckgraefe BK. Isolation and characterization of a cDNA encoding a novel member of the human regenerating protein family: Reg IV. Biochim Biophys Acta 1518: 287–293, 2001.[ISI][Medline]
  28. Hosack DA, Dennis G Jr, Sherman BT, Lane HC, and Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Biol 4: R70, 2003.[CrossRef][Medline]
  29. Huang X and Madan A. CAP3: a DNA sequence assembly program. Genome Res 9: 868–877, 1999.[Abstract/Free Full Text]
  30. Hulme DJ, Nicholas FW, Windon RG, Brown SC, and Beh KJ. The MHC class II region and resistance to an intestinal parasite in sheep. J Anim Breed Genet 110: 459–472, 1993.[ISI]
  31. Kennedy MW. Resistance to parasitic nematodes—how is the MHC involved? Parasitol Today 6: 374–375, 1990.[CrossRef][ISI][Medline]
  32. Lees-Miller JP, Heeley DH, and Smillie LB. An abundant and novel protein of 22 kDa (SM22) is widely distributed in smooth muscles. Purification from bovine aorta. Biochem J 244: 705–709, 1987.[ISI][Medline]
  33. Luffau G, Vu Tien Khang J, Bouix J, Nguyen TG, Cullen P, and Ricordeau G. Resistance to experimental infections with Haemonchus contortus in Romanov sheep. Genet Sel Evol 22: 205–229, 1990.[ISI]
  34. Ma XT, Qian Mp and Tang HX. Predicting polymerase II core promoters by cooperating transcription factor binding sites in eukaryotic genes. Acta Biochim Biophys Sin (Shanghai) 36: 250–258, 2004.[Medline]
  35. McCririe L, Bairden K, Britton C, Buitkamp J, McKeand JB, and Stear MJ. Heterogeneity in the recognition of Ostertagia circumcincta antigens by serum antibody from mature, infected sheep. Parasite Immunol 19: 235–242, 1997.[CrossRef][ISI][Medline]
  36. McEwan JC, Dodds KG, Watson TG, Greer GJ, Hosking BC, and Douch PGC. Selection for resistance to roundworms by the New Zealand sheep breeding industry; the WormFec service. Proc Aust Assoc Anim Breed Genet 11: 70–73, 1995.
  37. McEwan JC, Bisset SA, and Morris CA. The selection of sheep for natural resistance to internal parasites. In: Sustainable Control of Internal Parasites in Ruminants, edited by Barrell GK. Canterbury, New Zealand: Lincoln Univ. Press, 1997, p. 161–182.
  38. McEwan JC and Crawford AM. What the bovine genome project means to New Zealand. New Zealand Soc Anim Prod 64: 86–91, 2004.
  39. Michael JR. The stabilized probability plot. Biometrika 70: 11–17, 1983.[ISI]
  40. Miwa T, Manabe Y, Kurokawa K, Kamada S, Kanda N, Bruns G, Ueyama H, and Kakunaga T. Structure, chromosome location, and expression of the human smooth muscle (enteric type) gamma-actin gene: evolution of six human actin genes. Mol Cell Biol 11: 3296–3306, 1991.[ISI][Medline]
  41. Morris CA, Vlassoff A, Bisset SA, Baker RL, and Watson TG. Direct responses to selection for divergence in faecal nematode egg count in young Romney and Perendale sheep. Proc Assoc Adv Anim Breed Genet 12: 413–416, 1997.
  42. Morris CA, Vlassoff A, Bisset SA, Baker RL, Watson TG, West CJ, and Wheeler M. Continued selection of Romney sheep for resistance or susceptibility to nematode infection: estimates of direct and correlated responses. Anim Sci 70: 17–27, 2000.[ISI]
  43. Murdoch C. CXCR4: chemokine receptor extraordinaire. Immunol Rev 177: 175–184, 2000.[CrossRef][ISI][Medline]
  44. Outteridge PM, Windon RG, and Dineen JK. An association between a lymphocyte antigen in sheep and the response to vaccination against the parasite Trichostrongylus colubriformis. Int J Parasitol 15: 121–127, 1985.[CrossRef][ISI][Medline]
  45. Outteridge PM, Windon RG, Dineen JK, and Smith EF. The relationship between ovine lymphocyte antigens and faecal egg count of sheep selected for responsiveness to vaccination against Trichostrongylus colubriformis. Int J Parasitol 16: 369–374, 1986.[CrossRef][ISI][Medline]
  46. Outteridge PM, Windon RG, and Dineen JK. An ovine lymphocyte antigen marker for acquired resistance to Trichostrongylus colubriformis. Int J Parasitol 18: 853–858, 1988.[CrossRef][ISI][Medline]
  47. Paterson S, Wilson K, and Pemberton JM. Major histocompatibility complex variation associated with juvenile survival and parasite resistance in a large unmanaged ungulate population (Ovis aries L). Proc Natl Acad Sci USA 95: 3714–3719, 1998.[Abstract/Free Full Text]
  48. Pounds S and Morris S. Estimating the occurance of false positives and false negatives in microarray studies by approximating and partitioning the empirical distribution of p-values. Bioinformatics 19: 1236–1242, 2003.[Abstract/Free Full Text]
  49. Roven C and Bussemaker HJ. REDUCE: an online tool for inferring cis-regulatory elements and transcriptional module activities from microarray data. Nucleic Acids Res 31: 3487–3490, 2003.[Abstract/Free Full Text]
  50. Sangster NC. Antihelmintic resistance: past, present and future. Int J Parasitol 29: 115–124, 1999.[CrossRef][ISI][Medline]
  51. Schjetne KW, Gundersen HT, Iversen JG, Thompson KM, and Bogen B. Antibody-mediated delivery of antigen to chemokine receptors on antigen-presenting cells results in enhanced CD4+ T cell responses. Eur J Immunol 33: 3101–3108, 2003.[CrossRef][ISI][Medline]
  52. Schwaiger FW, Gostomski D, Stear MJ, Duncan JL, McKellar QA, Epplen JT, and Buitkamp J. An ovine major histocompatibility complex DRB1 allele is associated with low faecal egg counts following natural, predominantly Ostertagia circumcincta infection. Int J Parasitol 25: 815–822, 1995.[CrossRef][ISI][Medline]
  53. Shapland C, Lowings P, and Lawson D. Identification of new actin-associated polypeptides that are modified by viral transformation and changes in cell shape. J Cell Biol 107: 153–161, 1988.[Abstract]
  54. Shapland C, Hsuan JJ, Totty NF, and Lawson D. Purification and properties of transgelin: a transformation and shape change sensitive actin-gelling protein. J Cell Biol 121: 1065–1073, 1993.[Abstract]
  55. Smyth GK. Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Statistical Applications in Genetics and Molecular Biology 3, no. 1, article 3 (2004). http://www.bepress.com/sagmb/vol3/iss1/art3.
  56. Spiewak Rinaudo JA and Gerin JL. Cross-species hybridisation: characterization of gene expression in woodchuck liver using human membrane arrays. J Med Virol 74: 300–313, 2004.[CrossRef][ISI][Medline]
  57. Stear MJ, Tierney TJ, Baldock FC, Brown SC, Nicholas FW, and Rudder TH. Failure to find association between class I antigens of the bovine major histocompatibility system and faecal worm egg counts. Int J Parasitol 18: 859–861, 1988.[CrossRef][ISI][Medline]
  58. Stear MJ, Bairden K, Bishop SC, Buitkamp J, Epplen JT, Gostomski D, McKellar QA, Schwaiger FW, and Wallace DS. An ovine lymphocyte antigen is associated with reduced faecal egg counts in four-month-old lambs following natural, predominantly Ostertagia circumcincta infection. Int J Parasitol 26: 423–428, 1996.[CrossRef][ISI][Medline]
  59. Thweatt R, Lumpkin CK Jr, and Goldstein S. A novel gene encoding a smooth muscle protein is overexpressed in senescent human fibroblasts. Biochem Biophys Res Commun 187: 1–7, 1992.[ISI][Medline]
  60. Vallance BA, Blennerhassett PA, and Collins SM. Increased intestinal muscle contractility and worm expulsion in nematode infected mice. Am J Physiol Gastrointest Liver Physiol 272: G321–G327, 1997.[Abstract/Free Full Text]
  61. Vallance BA, Galeazzi F, Collins SM, and Snider DP. CD4 T cells and major histocompatibility complex class II expression influence worm expulsion and increased intestinal muscle contraction during Trichinella spiralis infection. Infect Immun 67: 6090–6097, 1999.[Abstract/Free Full Text]
  62. van Diemen PM, Ploeger HW, Nieuwland MG, Rietveld FW, Eysker M, Kooyman FN, Kloosterman A, and Parmentier HK. Low molecular weight Cooperia oncophora antigens. Potential to discriminate between susceptible and resistant calves after infection. Int J Parasitol 27: 587–593, 1997.[CrossRef][ISI][Medline]
  63. Wang Z, Dooley TP, Curto EV, Davis RL, and Vandeberg JL. Cross-species application of cDNA microarrays to profile gene expression using UV-induced melanoma in Monodelphis domestica as the model system. Genomics 83: 588–599, 2004.[CrossRef][ISI][Medline]
  64. Weisbrodt NW, Lai M, Bowers RL, Harari Y, and Castro GA. Structural and molecular changes in intestinal smooth muscle induced by Trichinella spiralis infection. Am J Physiol Gastrointest Liver Physiol 266: G856–G862, 1994.[Abstract/Free Full Text]
  65. Windon RG. Selective breeding for the control of nematodiasis in sheep. Rev Sci Tech 9: 555–576, 1990.[Medline]
  66. Wise MJ. Protein Annotators’ assistant. Trends Biochem Sci 25: 252–253, 2000.[CrossRef][ISI][Medline]
  67. Wise MJ. Protein Annotators’ assistant: a novel application of information retrieval techniques. J Am Soc Inf Sci 51: 1131–1136, 2000.[CrossRef][ISI]
  68. Woolaston RR and Piper LR. Selection of Merino sheep for resistance to Haemonchus contortus: genetic variation. Anim Sci 62: 451–460, 1996.[ISI]




This Article
Free upon publication
Abstract
Full Text (PDF)
Supplemental Tables
All Versions of this Article:
21/1/59    most recent
00257.2004v1
Citation Map
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Search for citing articles in:
ISI Web of Science (3)
Google Scholar
Articles by Diez-Tascón, C.
Articles by Crawford, A. M.
Articles citing this Article
PubMed
PubMed Citation
Articles by Diez-Tascón, C.
Articles by Crawford, A. M.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online
Copyright © 2005 by the American Physiological Society.