INTRODUCTION
Levamisole (LEV) is a broad-spectrum anthelmintic drug widely used to eradicate parasitic nematodes in livestock. However, the high efficacy of LEV against the gastrointestinal nematode Haemonchus contortus in sheep and goats has been compromised by the development of resistance in field populations. Compared to the other two commonly used anthelmintic classes [benzimidazole (BZ) and avermectin (AVM)], the rate of resistance selection for LEV appears to be slower in H. contortus. For this reason LEV is of particular interest to control BZ- and AVM-resistant isolates. In order to sustain the efficacy of LEV treatments, identification of resistance molecular mechanisms, and thus resistance markers, in parasitic nematodes are urgently needed.
LEV acts on nicotinic acetylcholine receptors (nAChR) and resistance is thought to be associated with changes in LEV binding characteristics or in the number of receptors (Coles, East and Jenkins, Reference Coles, East and Jenkins1975; Lewis et al. Reference Lewis, Wu, Levine and Berg1980; Sangster et al. Reference Sangster, Riley and Collins1988, Reference Sangster, Riley and Wiley1998; Robertson, Bjorn and Martin, Reference Robertson, Bjorn and Martin1999). Studies on the mode of action and resistance to LEV in parasitic nematodes were recently reviewed (Martin et al. Reference Martin, Verma, Levandosky, Clark, Qian, Stewart and Robertson2005; Sangster, Song and Demeler, Reference Sangster, Song and Demeler2005).
In C. elegans, many genes can be involved in LEV resistance (Jones and Sattelle, Reference Culetto, Baylis, Richmond, Jones, Fleming, Squire, Lewis and Sattelle2004). Among those genes, some encode nAChR subunits (unc-38, unc-29, unc-63, lev-1 and lev-8) (Lewis et al. Reference Lewis, Wu, Levine and Berg1980; Fleming et al. Reference Fleming, Squire, Barnes, Tornoe, Matsuda, Ahnn, Fire, Sulston, Barnard, Sattelle and Lewis1997; Culetto et al. Reference Culetto, Baylis, Richmond, Jones, Fleming, Squire, Lewis and Sattelle2004; Towers et al. Reference Towers, Edwards, Richmond and Sattelle2005). Nevertheless, studies of the sequence of an H. contortus receptor α subunit orthologous to unc-38 failed to identify differences between resistant and susceptible nematodes (Hoekstra et al. Reference Hoekstra, Borgsteede, Boersema and Roos1997a). Consequently, the molecular mechanisms involved in LEV resistance in H. contortus remain largely unknown and resistance markers are still to be identified.
The goal of the present study was the identification of genes that are differentially expressed in LEV-resistant compared to susceptible nematodes, in order to define potential resistance markers. We used resistant and susceptible isolates of H. contortus from different geographic origins to detect genes that show similar expression in all resistant isolates, but which differ in all susceptible isolates.
A comparative analysis of the transcriptome of susceptible and LEV-resistant H. contortus was performed using the cDNA-AFLP method (Bachem et al. Reference Bachem, van der Hoeven, de Bruijn, Vreugdenhil, Zabeau and Visser1996). This RNA-based technique allows the detection of differences in expression as well as polymorphisms in comparative analysis. This cDNA-AFLP approach proved to be very efficient for the detection of differentially expressed gene in plant parasitic nematodes such as Globodera rostochiensis (Qin et al. Reference Qin, Overmars, Helder, Popeijus, Rouppe Van der Voort, Groenink, Van Koert, Schots, Bakker and Smant2000) and Meloidogyne incognita (Neveu et al. Reference Neveu, Jaubert, Abad and Castagnone-Sereno2003). Here we report the identification of transcript derived fragments (TDFs) that distinguish LEV-resistant from susceptible H. contortus isolates, and a marker of interest was investigated in detail to validate our strategy.
MATERIALS AND METHODS
Nematodes (Table 1)
Three LEV-resistant isolates, Cedara and Kokstad from the Republic of South Africa and RHS6 from Zimbabwe (selected for LEV-resistance under laboratory conditions) were used to find resistance markers. The Moredun strain was used as one susceptible isolate; it has been maintained under experimental conditions for decades without anthelmintic treatments. The Zaire (now Congo) isolate from the Ituri region was also susceptible; it originated from a region where anthelmintic treatments were rare and no LEV treatment had been performed for 20 years at the time it was collected. It was then maintained under experimental conditions for 10 years without any treatment. The Brazil isolate (Fortaleza) and French isolates from the Basque region in the South-West of France were recently introduced into the laboratory (Silvestre et al. Reference Silvestre, Chartier, Sauvé and Cabaret2000). The resistance status was evaluated in pairs of experimentally infected lambs, one being treated with LEV and the other untreated. Faecal egg counts and worm counts at necropsy were used to assess resistance.
Characteristics of Haemonchus contortus isolates
* Artificially selected from the susceptible isolate SHS from Zimbabwe (Hoekstra et al. Reference Hoekstra, Borgsteede, Boersema and Roos1997b).
RNA purification
For each nematode isolate, total RNA was extracted from ten adult H. contortus males. Frozen worms were homogenized in Trizol reagent (Invitrogen) and total RNA isolated as per the manufacturer's recommendations. RNA pellets were dissolved in 50 μl of RNA secure solution (Ambion) and stored at −80°C. RNA samples were checked for contamination with genomic DNA by polymerase chain reaction (PCR) using β-tubulin primers annealing in exons 5 and 6 (Table 2). Amplification of a 286 bp DNA fragment containing intron 5 was never observed and only a PCR product of 219 bp lacking the 67 bp intron 5 was obtained in RNA samples (see Results).
Characteristics of primers used for RT-PCR analysis
* HT1 forward primer was designed in the 5′ flanking region of the TDF (H. contortus EST Hc_d11_13F08).
cDNA-AFLP experiments
For cDNA-AFLP, messenger RNAs were purified from total RNA using the Micropoly(A) Purist kit (Ambion), according to the manufacturer's recommendations.
Procedures for cDNA-AFLP have been described elsewhere (Bachem et al. Reference Bachem, van der Hoeven, de Bruijn, Vreugdenhil, Zabeau and Visser1996). Briefly, cDNA was synthesized with the Superscript cDNA kit (Invitrogen). cDNA samples were digested with HindIII/MseI, restriction fragments were ligated with their corresponding adapters (Semblat et al. Reference Semblat, Bongiovanni, Wajnberg, Dalmasso, Abad and Castagnone-Sereno2000) and pre-amplification was carried out during 25 cycles (94°C, 30s; 57°C, 45s; 72°C, 60s) using primers without selective nucleotide (H+0: 5′ GACTGCGTACCAGCTT 3′, M+0: 5′ GATGAGTCCTGAGTAA 3′). Then, PCR was performed with a trace amount of [33P]-labelled 5′ primer and amplification products were separated on a 5% polyacrylamide gel and analyzed after exposure to X-ray film overnight. Polymorphic fragments were selected when present in the two resistant isolates and absent from the two susceptible isolates, and vice-versa.
TDF analysis
Bands of interest were cut from the dried polyacrylamide gel and soaked in 30 μl sterile water. DNA was recovered after three freeze-thaw cycles using liquid nitrogen. One μl of the diffusate was used as template for re-amplification using the corresponding primer combination that was used for the cDNA-AFLP experiments. Re-amplification products were cloned in the TOPO pCR2.1 vector (Invitrogen) and sequenced by Genome express (Grenoble, France).
Homology searches in databases were performed using the BLAST network service (NCBI, National Center for Biotechnology Information). TDF sequences were compared against all sequences in the non-redundant protein databases using the BLASTX algorithm (Altschul et al. Reference Altschul, Madden, Schaffer, Zhang, Zhang, Miller and Lipman1997). Each TDF sequence was also compared against the EST databases using the BLASTN algorithm.
Semi-quantitative RT-PCR analysis
The primers used for PCR were designed on TDF sequences from the HA1, HA7, HA10, HA17, HT1, HT2, HT15 and HT23 markers (Table 2). RT-PCR experiments were carried out with 500 ng total RNA extracted from adults of all the H. contortus isolates. First-strand cDNAs were synthesized with Superscript III reverse transcriptase (Invitrogen). Total RNAs were denatured by heating at 65°C for 5 min and RT was performed at 50°C for 60 min in a final volume of 20 μl of the following solution: 0·5 mM dNTPs, 40 U RNase inhibitor, 25 ng/μl of T20VN oligonucleotide cocktail (Invitrogen), 5 mM DTT and 200 U Superscript III (Invitrogen).
Semi-quantitative PCRs were then performed in a final volume of 20 μl containing 1 μl of 10-fold dilution of reverse transcription reaction products, 250 μM each dNTP, 25 ng/μl each specific primer pair and 1 U GoTaq polymerase (Promega). The reaction mixture was denatured by heating at 94°C for 30 sec, annealed at 55°C for 45 sec, and extended at 72°C for 45 sec for 33 cycles. The mixture was overlaid with mineral oil and amplified in a MJ Research thermal cycler. The PCR products were subjected to electrophoresis through a 2% agarose gel. Sample normalization was assessed using β-tubulin as the reference transcript.
RESULTS
Detection of differentially expressed genes between LEV-resistant and susceptible isolates of Haemonchus contortus
Adult males from two resistant isolates (Kokstad and Cedara) and two susceptible isolates (Moredun and Congo-Zaire) were used in this experiment. Using HindIII/MseI as restriction enzymes and 32 primer combinations (Table 2), approximately 4320 transcript-derived fragments (TDFs), ranging from ∼100 bp to 850 bp, were generated for each isolate (Fig. 1). Among these TDFs, 26 presented a differential pattern between resistant and susceptible nematodes, i.e. 11 were present in both resistant isolates but were absent from or markedly underexpressed in both susceptible isolates. Conversely, 15 TDFs were present in the susceptible isolates and absent from or underexpressed in the resistant isolates.
Example of the transcript derived fragments (TDFs), ranging from ∼100 bp to 850 bp, generated in the Kokstad, Cedara, Zaire and Moredun isolates (generated using HindIII/MseI as restriction enzymes and 32 primer combinations). Among these TDFs, several present a differential pattern between resistant and susceptible nematodes (see on right of the figure).
Examples of such TDFs are presented in Fig. 1. As a first step, the TDFs specifically detected in resistant isolates were characterized as a matter of priority. To date, 8 have been cloned and sequenced. Hypothetical proteins deduced from the six-frame translated fragments were analysed using the BLASTX program; deduced sequence homologies are presented in Table 3. Four TDF sequences presented significant similarities with hypothetical proteins from H. contortus, C. elegans or C. briggsae, while the other TDFs showed no significant homology with known proteins. Half of these 8 TDFs presented significant similarities with H. contortus ESTs. TDFs that did not show any homologies with known genes or ESTs could represent uncharacterized genes, but can also correspond to untranslated regions of mRNA or may be too short to detect significant homologies in databases.
Summary of 8 TDFs that distinguish levamisole-resistant and sensitive isolates of Haemonchus contortus
Validation of candidate markers by RT-PCR experiments
To detect polymorphisms of expression between LEV-resistant and susceptible nematodes, RT-PCR experiments were performed on 1st-strand cDNA from H. contortus adult males. Primers were designed on the sequences of the 8 TDFs identified in the cDNA-AFLP experiments and RT-PCR analyses were carried out on the Kokstad and Cedara LEV-resistant vs. Moredun and Zaire susceptible isolates. Among the 8 TDFs, 7 presented identical expression profiles between resistant and susceptible nematodes (Fig. 2). Conversely, the amplification pattern for HA17 observed during cDNA-AFLP experiments was also obtained through RT-PCR; i.e., we observed an abundant amplicon of the expected size in resistant isolates and a very weak signal in susceptible isolates. This result was confirmed using a second pair of primers designed in the TDF sequence of HA17 (data not shown).
RT-PCR analysis of expression levels of eight TDFs in two LEV-resistant (K, C) and two susceptible (Z, M) H. contortus isolates. (See Table 1 for isolates code).
HA17: a new molecular marker of LEV resistance?
To validate HA17 as a potential LEV resistance marker, a new set of RT-PCR experiments using the primers targeting the TDF HA17 was performed on cDNA from the other resistant and susceptible isolates. The LEV-resistant isolate RHS6 (selected under controlled conditions starting from the susceptible isolate SHS-Zimbabwe, Hoekstra et al. Reference Hoekstra, Borgsteede, Boersema and Roos1997b) and three LEV-susceptible field isolates (Fortaleza (Brazil), Herian (France) and Bordaicoborda (France)) were tested. HA17 expression was readily detected in the three LEV-resistant isolates as shown with an amplification product at the expected size, whereas no or a very low signal was obtained in all susceptible isolates (Fig. 3). As a reference PCR control, we found no change in the expression of the β-tubulin gene. This result clearly suggests that HA17 may be a marker for LEV resistance.
Differential expression of the HA17 TDF according to LEV resistance status of H. contortus isolates from various geographical locations. HA17 expression was analysed in three LEV-resistant versus five susceptible isolates by semi-quantitative RT-PCR. (See Table 1 for isolates code).
DISCUSSION
In the present study, we describe a cDNA-AFLP strategy comparing transcripts from LEV-resistant and susceptible isolates of H. contortus in order to identify resistance markers. The comparative analysis of cDNA from two resistant and two susceptible isolates led to the identification of 26 differentially expressed TDFs. Among these candidates, we focused our attention on 8 TDFs that were specifically present in resistant nematodes, because of their potential utility as resistance markers. However, the 3 remaining TDFs specifically present in resistant isolates, and the 15 TDFs specifically present in susceptible isolates, are currently under investigation. These TDFs might constitute additional LEV resistance markers as well as the TDFs already characterized.
Since a polymorphic pattern observed on cDNA-AFLP gels could result from either quantitative (difference in expression level) or qualitative (transcript polymorphism) differences in mRNAs, RT-PCR experiments were performed to investigate the origin of the polymorphisms observed for the 8 TDFs of interest.
Seven of these TDFs presented a similar expression pattern in resistant and susceptible isolates. For those candidates, further studies will be carried out in order to understand the molecular events underlying the polymorphism detected during cDNA-AFLP analysis. Indeed, polymorphic TDFs observed on cDNA-AFLP gels could also result from mutation in the restriction sites or be related to insertion/deletion events.
In contrast, the HA17 TDF was specifically expressed in the LEV-resistant isolates studied during cDNA-AFLP experiments. To validate HA17 as a potential LEV resistance marker, we investigated its expression in additional resistant and susceptible isolates from different geographical origins. A PCR product was obtained only in LEV-resistant isolates, whereas amplification was not detectable or very weak in susceptible isolates. HA17 can thus be considered as a promising LEV resistance marker, but more field isolates must be tested to estimate its robustness. Even though HA17 sequence presents no homology with known proteins or ESTs, its specific expression in all tested resistant isolates could implicate it in a resistance mechanism. Further characterization of HA17 should include gene cloning, promoter studies and functional validation using the RNA interference technique recently developed for H. contortus (Geldhof et al. Reference Geldhof, Murray, Couthier, Gilleard, McLauchlan, Knox and Britton2006; Kotze and Bagnall, Reference Kotze and Bagnall2006) or transformation of C. elegans.
In H. contortus, development of LEV resistance appears to be polygenic (Sangster, Davis and Collins, Reference Sangster, Davis and Collins1991). For that reason, research to discover more resistance markers should be pursued. In this context, the comparative transcriptomic approach described in this paper can be considered as complementary to a candidate gene strategy, based on C. elegans data. Indeed, candidate gene strategy has the advantage of concentrating cloning efforts on a limited number of genes in comparison with the numerous candidates generated using the transcriptomic approach described in this paper. However, extrapolation of C. elegans data to parasitic nematodes remains hypothetical. In contrast, the cDNA-AFLP approach could lead to the identification of novel genes involved in LEV resistance that are specific to parasitic species. Conversely, it must be recognized that many differential TDFs might represent population markers, which are not linked with the resistance phenotype. Optimally, both approaches should be carried out together. In combination with the anticipated completion of the genome sequence of H. contortus, these techniques should accelerate discovery of molecular mechanisms involved in LEV resistance.
Acknowledgements
We are indebted to the following researchers for providing isolates or strains used in this work: J. Van Wyk, Pretoria University, RSA (Cedara and Kokstad strains) under a French-RSA program; Frank Jackson, Moredun Institute, UK (Moredun strain); Fred Borgsteede, Lelystadt Research Institute, NL (RSH6). The Congo Zaire isolate was obtained by one of us (JC) during a World Bank Program/CIRAD with C. Chartier and then maintained in laboratory conditions at INRA Nouzilly. The excellent technical help of Christine Sauvé in preparing isolates from French Basque region and checking for LEV resistance is gratefully acknowledged. C. Neveu is a grateful recipient of a CARS (Consortium for Anthelmintic Resistance SNPs) grant for presenting results in the Glasgow meeting, 4–5 August 2006.
