The Echinococcus granulosus antigen B shows a high degree of genetic variability

Echinococcus granulosus larvae secret a polymeric lipoprotein known as antigen B (AgB) into the metacestode hydatid ﬂuid. Three similar AgB subunits have been previously identiﬁed (AgB1, AgB2, and AgB3), and their respective genes isolated, but the actual number of genes encoding AgB subunits remains uncertain. In this study, we characterize the variability of genes encoding the AgB2 subunit, using PCR and RT-PCR followed by cloning and sequencing. We have analyzed 32 cDNA and 34 genomic sequences from a single metacestode, showing a high degree of sequence polymorphism. In addition, we have identiﬁed a possibly new AgB subunit, which we call AgB4. Additionally, we describe an AgB2 genomic clone lacking (i) a segment corresponding to the intron and (ii) a short, 45 bp sequence within exon II. The 45 bp segment encompasses the conserved splicing signals and corresponds to a highly conserved insect promoter motif.

Infection by the metacestode larval stage of the tapeworm Echinococcus granulosus causes an important zoonosis, distributed worldwide, known as cystic hydatid disease.The adult stage develops mainly in the small intestine of dogs and others canids.The metacestode occurs in viscera (mainly liver and lung) of ungulates macropods and humans (as accidental hosts).
Antigen B (AgB) is an extremely abundant component of the hydatid cyst (metacestode) fluid and is a polymeric thermostable lipoprotein of 120-160 KDa (Oriol and Oriol, 1975).The role of AgB in the parasite biology is not completely elucidated, but it seems to be related to the evasion of the hostÕs immune response.In vitro studies have shown that it inhibits neutrophil recruitment, has protease inhibitor activity (Shepherd et al., 1991) and exploits the activation of T-helper cells by eliciting a non-protective Th2 cell response (Rigano `et al., 2001).Recently, it has been demonstrated that AgB induces apoptosis in polymorphonuclear cells from patients with active disease (Rigano èt al., 2002).It is a highly immunogenic antigen used in immunodiag-nostic assays.Recombinant antigens derived from distinct AgB subunits possess variable diagnostic values, with AgB2 showing the highest sensitivity and specificity in immunodiagnostic tests (Virginio et al., 2003).
The first isolated AgB nucleotide sequence (AgB1) was a cDNA cloned by Shepherd et al. (1991), and a second sequence (AgB2) was described by Ferna ´ndez et al. (1996).AgB1 and AgB2 show a hydrophobic stretch from amino acid positions 1-16 and 1-20, respectively, followed by one potential eukaryotic signal for peptidase cleavage (Von Heijini, 1986), indicating that such peptides are secreted.Chemale et al. (2001), using primers specific to the nucleotide sequence for AgB1, have isolated a third AgB gene (AgB3), and raised the possibility that the gene family actually includes even more than three copies.Frosch et al. (1994) identified the first nucleotide polymorphisms in cDNA sequences encoding the AgB1 subunit.Haag et al. (1998) showed that the rates of AgB1 non-synonymous substitutions among isolates from different strains suggest positive selection.The aim of the present study was to unravel the amount, degree and nature of genetic variability for the E. granulosus AgB2 subunit within one single genome.
The analysis of 66 recombinant clones obtained from a single E. granulosus cyst disclosed a high degree of genetic variation.We found eighteen different sequences, represented in Fig. 1.The sequence of clone B2.1geno corresponded to a previously deposited AgB2 cDNA sequence (GenBank Accession No. U15001).Four other genomic clones and six cDNA clones had a high identity to this gene, with minor nucleotide differences (Fig. 1).However, eleven other clones (seven genomic and four cDNA clones), are related to a mRNA sequence previously deposited in GenBank with Accession No. AF252859 (Lu, 2000).
Given its low amino acid identity with AgB2 (70%, Table 1) this second set of sequences is suggested to correspond to a fourth AgB subunit, which we call AgB4.Studies employing polyclonal antisera indicate that non-denatured globular proteins differing by at least 30-40% amino acid sequence usually do not cross-react (Wilson et al., 1977).Comparison of cDNA of the sequences of the cDNA with those of the genomic clones suggested that at least four different gene variants are expressed in protoscoleces.Some cDNA clones with sequences identical to clones B2.1geno, B2.2geno, B4.1geno and B4.3geno were identified (Table 2).
Fig. 2 shows the nucleotide diversities estimated for different gene regions (intron, exon II, and the whole sequence; exon I is excluded from the analyses, because it does not show variation among the clones isolated in this study).Values for all sequences combined are much higher than those calculated separately for the AgB2 and AgB4 groups, reflecting the considerable divergence between the two sets of Table 1 Nucleotide (above diagonal) and amino acid (below diagonal) identities between the four known AgB sub-units a AgB1 (%)    sequences.Note that, within each group, genetic variation sometimes involves only a single non-synonymous substitution (Fig. 1 and Table 2).Also, within each group and between the two groups, the nucleotide diversity was greater in exon II than in the intron.There is evidence that the observed nucleotide variants are not due to PCR errors.Indeed, Haag et al. (2004) have encountered the same substitutions at the same sites in separate experiments, using three independent PCR and different DNA templates.Moreover, high fidelity/proof-reading DNA polymerases were used to prevent nucleotide misincorporations.Moreover, as an external control for misincorporation artifacts, we cloned the amplification products from a nuclear but single-copy gene, encoding a malate dehydrogenase (mdh), using the same reagents.The mdh amplicons also included a small intron and were equivalent in size to the AgB amplicons.We sequenced 21 mdh clones, but no nucleotide variability was found.
The presence of a relatively large number of different sequences (Table 2) may be due to, at least, two possibilities.(1) The AgB gene family has several duplicated genes within a genome, so that some duplications are relatively ancient, such as AgB2 and AgB4, while other duplications are relatively recent, such as B2.1 versus B2.2 or B2.3, or B4.1 versus B4.2 and B4.3.(2) There are ancient duplicated genes (such as AgB2 and AgB4), but additional variation (such as among B2.1, B2.2 and B2.3) is being originated during asexual amplification of the metacestode, so that protoscoleces are not true clones with respect to the AgB genes (see Haag et al., 2004).
Five chimeric clones (three cDNA and two genomic clones) were excluded from analysis because of their possible artifactual origin.PCR-derived chimerism has been reported in previous attempts to isolate AgB genes (Chemale et al., 2001) and other multigene family studies (Bhavsar et al., 1994;Gaedigk et al., 1998).Incomplete extension products often work as primers in the following PCR cycles.Chimeras are produced by heterologous primer annealing, when mixed templates are present in the same reaction (which is the case of a multigene family with closely related gene copies, or of a mixture of DNAs from genetically related individuals).Our interpretation of the excluded clones as chimeric is supported by an additional experiment, in which a primer pair designed to specifically amplify an AgB2/AgB4 chimera did not result in any amplification product from templates derived from six different E. granulosus metacestodes (not shown).
Considering that AgB4 corresponds to another gene copy, the number of nucleotide substitutions within each gene group was analyzed.As expected, transitions are more frequent than transversions (20 in 26 substitutions) and the most frequent transition is C M T (see Table 2).It is interesting that a synonymous C M T transition at position 203 appears in four out of nine AgB4 clones (cDNA and genomic).Since genomic and cDNA clones are obtained in independent experiments, there is a low probability that the transitions at position 203 represent PCR artifacts.We found a non-synonymous A M G transition adjacent to a non-synonymous T M G transversion, at positions 215 and 216, in an AgB2 cDNA clone (B2.8cdna) as well as in all, but one AgB4 sequences.One AgB4 genomic clone (B4.6geno) exhibits the reversed substitutions, suggesting the possibility of gene conversion (see also Fig. 1).
Four out of 10 and 6 out of 16 nucleotide substitutions detected, respectively, in the AgB2 and AgB4 sequences were replacements, indicating a possible adaptive role (Table 2).Haag et al. (1998) suggested positive selection in the evolution of AgB1, which would not be surprising, given that AgB seems to act in immune response evasion.Endo et al. (1996), searching for candidate genes on which positive selection might be operating, obtained 17 gene groups, 9 of which encoded surface antigens of parasites or viruses.
An intriguing finding of the present study was a truncated genomic clone (B2.2geno) with clear characteristics of an mRNA molecule (Fig. 1).It lacked the AgB2 region corresponding to the intron, plus a 45 bp stretch with conserved splicing junctions (GC/AG).The deleted 45 bp sequence in the B2.2geno clone also includes a 3 0 YNYAG conserved splicing signal (Senapathy et al., 1990) and corresponds to an insect promoter motif (http://www.fruitfly.org/seq_tools/promoter.html, October 11, 2003).Although we can not rule out a PCR artifact as a possible cause for this clone, it might indicate that AgB genes duplicate by a mechanism requiring an intermediate mRNA.The 45 bp deletion was found both in genomic and cDNA experiments, suggesting that this AgB2 variant is transcribed.
We have herein described a fourth AgB gene (AgB4), which is transcribed by the E. granulosus metacestode, and seems to encode an additional AgB subunit.We also show that the two genes, AgB2 and AgB4, from a single cyst have considerable variation, which might be adaptive.The surprising discovery of a truncated AgB2 transcribed variant, plus some additional clues, suggest a process involving an intermediate mRNA.

Fig. 1 .
Fig. 1.Alignment of nucleotide sequences from the 18 distinct AgB clones isolated in the present study.The top nine clones, B2.1geno to B2.9cdna, are related to the previously characterized AgB2 gene (GenBank Accession No. U15001); the other nine clones, B4.1geno to B4.9cdna, relate to a cDNA deposited in GenBank with Accession No. AF252859.Genomic and cDNA clones were obtained in independent experiments.Dots (.) indicate identity with the reference sequence.Dashes (---) indicate indels.Primers used to generate the inserts are shown at both ends of the reference sequence.Replacement positions are indicated with asterisks on the reference sequence.When cDNA and genomic clones were identical inside the coding region, only the genomic clone was represented.

Fig. 2 .
Fig. 2. Mean nucleotide diversity (p) and standard deviation (vertical bar) for the combined AgB2 and AgB4 clones and for clones within each group.Calculations are made for the whole fragment, for only the intron, and for only exon II.AgB4 nucleotide diversities are also estimated without clone B4.2geno, because all intron substitutions occur in this clone.

Table 2
Number (N) of cDNA and genomic clones obtained for each AgB variant