Infection of bovine leukocytes by the apicomplexan parasite Theileria annulata results in alteration of host cell gene expression and stimulation of host cell proliferation. At present, the parasite-derived factors involved in these processes are unknown. Recently, we described the characterisation of a parasite gene (TashAT2), whose polypeptide product bears AT hook DNA-binding motifs and may be transported from the parasite to the host nucleus. We now describe the isolation of a further two genes (TashAT1 and TashAT3) that are very closely related to TashAT2. All three TashAT genes are located together in a tight cluster, interspersed by two further small open reading frames, all facing head to tail. TashAT2 was shown to be expressed in all T. annulata cell lines examined, whereas TashAT1 and TashAT3 were expressed in the sporozoite stage of the parasite, and also in infected cell lines, where their expression was found to vary between different cell lines. Evidence for transport was provided by antisera raised against TashAT1 and TashAT3 that reacted with the host nucleus of T. annulata-infected cells. Reactivity was particularly strong against the host nuclei of the T. annulata-infected cloned cell line D7B12, which is attenuated for differentiation. A polypeptide in the size range predicted for TashAT3 was preferentially detected in host enriched D7B12 nuclear extracts. DNA-binding analysis demonstrated that fusion proteins containing the AT hook region of either TashAT1 or TashAT2 bound preferentially to AT rich DNA.

Theileria annulata and Theileria parva are closely related parasites that belong to the phylum apicomplexa. Both parasite species are transmitted from a feeding tick to their bovine host, and invade white blood cells (Dschunkowsky and Luhs, 1904EF10). The parasite differentiates into the multinucleate macroschizont and stimulates the host cell to undergo unlimited division (Hulliger, 1965EF16). In vivo, infected cells proliferate within the lymph node that drains the site of tick attachment and then spread throughout the bovine host, infiltrating other major organ systems of the body, such as the gastrointestinal tract, kidney and lungs (Irvin and Morrison, 1987EF17). The presence of Theileria-infected cells in these organs is often accompanied by the presence of petecheal haemorrhages that contain infected cells. These resemble multicentric lymphosarcomas, and injection of Theileria-infected cells into SCID (Fell et al., 1990EF13) or athymic mice (Irvin et al., 1975EF18) results in the formation of tumours.

Parasite infection of the host cell is associated with modulation of leukocyte gene expression, including a number of genes encoding transcription factors that are implicated in the control of cell division or apoptosis (Dobbelaere et al., 1988EF9; Baylis et al., 1995EF3; Ole-MoiYoi et al., 1993EF23; Botteron and Dobbelaere, 1998EF5; Heussler et al., 1999EF15). However, although induction of host cell division is known to be Theileria dependent (McHardy et al., 1985EF21), little is known about how the parasite directly modulates leukocyte gene expression or stimulates the host cell to divide.

It has been suggested that as the macroschizont differentiates into the extracellular merozoite, parasite factors involved in host cell division may be downregulated, resulting in the cessation of cell division (Carrington et al., 1995EF6). Thus, the association between parasite and host cellular division would be uncoupled, owing to the removal of the signal that initiates proliferation of the infected lymphocyte. We have previously identified a small gene family whose expression is downregulated during differentiation to the merozoite in T. annulata (Swan et al., 1999EF33). One member of this family, TashAT2 encodes a gene product that bears a predicted AT hook motif DNA-binding domain. Furthermore, experimental data suggest that TashAT2 is transported from the parasite to the host nucleus, implying a role in the modification of host cell gene expression. We present further characterisation of TashAT2 and two other members of the gene family, TashAT1 and TashAT3. All three genes form part of a cluster, encode AT hook DNA-binding motifs (Johnson et al., 1988EF19) and are very closely related in sequence. We present evidence to suggest that TashAT3 and, possibly, TashAT1 could be transported to the host cell nucleus and discuss the possible implications.

Cell culture

The T. annulata (Hidirseyh), T. annulata (Diyarbakir) (Pipano and Shkap, 1979EF24), T. annulata (Ankara) (TaA2) macroschizont-infected cell lines (Shiels et al., 1992EF28) and the cloned cell lines derived from TaA2 (D7, E3, C9 and D7B12) were maintained in vitro at 37°C or induced to differentiate as previously described (Shiels et al., 1992EF28). BL20, an uninfected bovine lymphosarcoma cell line (Morzaria et al., 1982EF22), and TBL20, a T. annulata-infected cell line derived from BL20 (Shiels et al., 1986EF27), were cultured as for D7, except that myoclone super plus foetal bovine serum (Gibco BRL) was substituted for heat-inactivated foetal bovine serum (Sigma).

Cloning and northern blot analysis

A fragment of TashAT1 had previously been isolated from a λgt11 library of genomic DNA derived from merozoites of the D7 infected cloned cell line (Swan et al., 1999EF33). The TashAT1 fragment was used to screen a λDASHII library of D7 genomic DNA using standard protocols. Two λDASHII overlapping clones, based on their restriction maps and hybridisation profiles, were isolated and a 13.4 kbp region sequenced on both strands. DNA sequencing was performed on a Licor 4000 automated DNA sequencer according to the manufacturer's protocol. DNA and protein analyses were performed using the GCG sequence analysis package (Devereux et al., 1984EF8).

RNA from in vitro cultured cells and from purified T. annulata (Ankara) piroplasms was isolated by the Triazol reagent, according to the manufacturer's instructions (Sigma). Sporozoite RNA was isolated from T. annulata (Ankara)-infected Hyaloma anatolicum ticks as described by Williamson et al. (Williamson et al., 1989EF37). RNA was size fractionated by electrophoresis through a formaldehyde-agarose gel and analysed by Northern blotting as previously described (Shiels et al., 1994EF29). Hybridisation was carried out overnight at 65°C according to the method of Church and Gilbert (Church and Gilbert, 1984EF7).

Generation of fusion protein and immunoblotting

A 362 bp fragment of TashAT1 starting 63 bp from the first putative translational start site was PCR amplified using ampliTaq polymerase and the primers; 5′-tttaggatccgtaaaatttgcttcttcc-3′ and 5′-gaaggaattctggtggaattttaataaa-3′. The PCR product was then subcloned into the vector pGEX-2TK (Pharmacia), expressed in Escherichia coli strain JM109 as a glutathione-S-transferase (GST) fusion protein and purified on a glutathione-sepharose column using the Pharmacia protocol. Antisera (anti-TashAT1/3) to the fusion protein were raised in New Zealand White rabbits and immunoblotting was carried out as described by Swan et al. (Swan et al., 1999EF33). Signal was detected by ECL using the method provided by the suppliers (Pierce and Warriner).

DNA-binding analysis of TashAT1

A λgt11 library of D7 genomic DNA or purified λgt11 clones were induced to express fusion protein under standard conditions. Nitrocellulose filters were laid on top of the developing plaques, which were then incubated at 37°C for 4 hours. The filters were removed from the plates, washed briefly in TNE 50 (10 mM TrisCl pH 7.5, 50 mM NaCl, 1 mM EDTA, 0.1% NP40), then incubated overnight in blocking buffer (2.5% dried milk, 25 mM Hepes (pH 8.0), 1 mM DTT, 10% glycerol, 50 mM NaCl, 1 mM EDTA, 0.1% NP40) at 4°C. The filters were then rinsed at 4°C in 1× binding buffer (1× BB; 25 mM Hepes pH 7.9, 3 mM MgCl2, 4 mM KCl) and the expressed fusion protein denatured by incubation in 1×BB containing 6 M guanidinium HCl, 1 mM DTT twice for 5 minutes. Renaturation was then carried out by immersing the filters for 5 minutes in denaturation buffer, which was sequentially diluted 1 in 2 with 1×BB, 1 mM DTT, four times. The renatured filters were washed in 1×BB, 1 mM DTT, incubated in 1×BB, 5% dried milk for 30 minutes, rinsed in 1×BB, 0.25% dried milk and washed briefly with HNE (10 mM Hepes pH 7.9, 50 mM NaCl, 1 mM EDTA, 1 mM DTT). Hybridisation was then carried out in HNE buffer containing 500 μg ml-1 poly dIdC filtered through a 0.22 μm filter. Double stranded concatenated probes were generated by ligating oligonucleotides radiolabelled at the 5′ end with 32P using T4 polynucleotide kinase. The oligonucleotides were: CAT1, 5′atgcGCACACAATTTGTAGGGCGAC3′; CAT1m2, 5′atgcGCACACAATTACGAGGGCGAC3′; CAT2, 5′ atgcAGATAAACATGCACACAATTTGTA3′; or CAT3, 5′atgcAGGGCGAC3′ previously annealed to their respective complementary oligonucleotides. Each oligonucleotide had a four base overhang at the 5′ end, either atgc (or gcat in the reverse complement) shown in lower case, for concatenation. Probes were added to the filters and incubated at 4°C for 60 minutes, after which the filters were given three 5 minute washes in HNE buffer, followed by exposure to X-ray film.

DNA-binding analysis of TashAT2

The procedure for determining TashAT2 DNA-binding specificity was based on the method used by Pollock and Treisman (Pollock and Treisman, 1990), but used glutathione-sepharose instead of immunoprecipitation to bind the DNA-protein complex. In brief, a double-stranded 76 mer oligonucleotide with a 26 base pair random core flanked by two specific 25 base sequences (primer F and primer R) was PCR amplified and labelled using 32P dCTP. A part of TashAT2 containing the AT hook domain (amino-acid residues 296-541) linked to GST as a fusion protein (GST-TashAT2) (Swan et al., 1999), was purified and approximately 3 μg of GST-TashAT2 or 3 μg of purified GST alone used per binding cycle. 100 μl of a 50% solution of glutathione-sepharose was washed twice in phosphate buffered saline (PBS) before adding 100 μl of PBS containing the fusion protein. The mixture was then rotated at room temperature for 15 minutes and the beads washed three times in binding buffer (10 mM Hepes pH 7.9, 25 mM KCl, 1 mM EDTA, 50 mM NaCl, 0.1% NP40 plus protease inhibitors) to remove excess protein. 0.4 ng of radiolabelled probe was added to the beads in 25 μl of binding buffer and rotated for 30 minutes at 4°C. The beads were washed in binding buffer three times, followed by phenol/chloroform extraction. The DNA was then ethanol precipitated from the aqueous phase, washed in 70% ethanol and resuspended in 20 μl of TE pH 8.0. 10 μl were taken for PCR amplification using primer F and primer R, and the cycle of protein to DNA binding repeated four times. To test the ability of the amplified sequences to bind GST-TashAT, an electrophoretic mobility shift reaction (EMSA) was set up containing 10 μl of the radiolabelled DNA that was purified during each cycle, 0.5 μg of protein, 1×BB, 5% Ficoll, 1 μg of poly dGdC:dGdC in a 40 μl reaction. The reaction mix was incubated for 30 minutes at 4°C. Electrophoretic separation was then carried out on a 4% polyacrylamide gel cast in 0.5×TBE and run in 0.5×TBE buffer. A mobility shifted band from cycle 4 was excised, eluted into dH2O, PCR amplified and the EMSA repeated. The resulting mobility shifted band was excised, PCR amplified and subcloned into the vector pGEMT-easy for sequencing.

Cloning and sequence characterisation of TashAT1 and TashAT3

A fragment of TashAT1 was originally isolated from a λgt11 library representing T. annulata genomic DNA isolated from the cloned infected cell line D7 (Swan et al., 1999EF33). This fragment was used to screen a λDashII library of D7 parasite genomic DNA and several hybridising clones were isolated. Two overlapping clones were selected, restriction mapped and 13.4 kbp of DNA sequence obtained. The restriction pattern of the 13.4 kbp contig coincided with the profile obtained from D7 genomic DNA restriction digested and probed with internal fragments from the TashAT cluster (data not shown). Five ORFs were detected (Fig. 1), three of which shared extensive sequence homology. ORF1 encodes TashAT2 (Swan et al., 1999EF33); ORFs 2 and 4 will be described elsewhere (L.S., D.G.S. and B.R.S., unpublished) and ORFs 3 and 5 are designated TashAT1 and TashAT3, respectively. TashAT1 and TashAT3 encode ORFs of 1401 bp and 2985 bp respectively. Comparison of the DNA sequence of all three ORFs revealed a remarkable degree of conservation to each other. TashAT1 is 98.9% identical to the first 1401 bp of TashAT3. Overall TashAT2 is 83.6% identical to TashAT3 and 67.5% identical to TashAT1. However, TashAT3, from position 1320-2715 bp of the ORF is 99.9% identical to base pairs 1386-2781 of TashAT2. A comparison of the translation product of TashAT1 with that of TashAT3, and of TashAT2 with TashAT3, is shown in Fig. 2A,B, respectively. An analysis of the translation products of both TashAT1 and TashAT3 revealed the presence of a potential signal peptide at the 5′ end of TashAT1/3 and detected possible AT hook DNA-binding motifs (Fig. 2A,C). AT hooks are small (8-10 amino acid residues), semi-conserved basic motifs rich in K, R and P, with a G present in a highly conserved region of the motif (Johnson et al., 1988EF19); they bind preferentially to the minor groove of AT-rich DNA (Reeves and Nissen, 1990EF26) and have been found in a variety of regulatory factors with diverse functions (Aravind and Landsman, 1998EF2). TashAT1/3 have four AT hook motifs (Fig. 2C) contained in a 120 amino acid residue domain, motifs 2, 3 and 4 are identical to each other, and to motifs 2 and 3 of TashAT2, which was previously determined to contain three AT hook motifs (Swan et al., 1999EF33). All four AT hook motifs of TashAT1/3 reflect the semi-conserved basic nature of a typical AT hook (Fig. 2C). Motifs 2, 3 and 4 contain the core sequence RGRP which is present in almost all AT hook motifs (Aravind and Landsman, 1998EF2). Although the AT hook motif containing domains of TashAT1/3 and TashAT2 are very similar, TashAT1/3 contains a 77 amino acid residue insertion between AT hooks 2 and 4 (Fig. 2C). Within this insertion, in addition to AT hook 3, TashAT1/3 also contains two peptides of sequence RPRK that could contribute to DNA binding. Immediately upstream of the AT hook domain of TashAT1/3 there is a region composed of small imperfect repeats, rich in glutamic acid, glutamine, aspartate and threonine residues (Fig. 2A), that is similar in composition to that described for transcriptional transactivation domains (Triezenberg, 1995EF34).

Fig. 1.

Restriction map and schematic of the TashAT cluster.

Fig. 1.

Restriction map and schematic of the TashAT cluster.

Fig. 2.

Sequence analysis of TashAT1 and TashAT3 and their comparison with TashAT2. (A) Sequence of the predicted translation products of TashAT1 and TashAT3. The potential signal sequence is in italics. The AT hook DNA-binding domain is in bold and the putative transcriptional transactivation domain is underlined. Where the sequence of TashAT1 differs from that of TashAT3 is shown above the relevant sequence of TashAT3. The asterisks mark the end of each sequence. (B) A comparison of TashAT2 with TashAT3. Identical residues are shown in bold. The numbers at the end of the comparison denote the region which is 100% identical between the two sequences. A gap generated by the analysis in the sequence of TashAT2 is shown by a series of dashes. (C) A comparison of the AT hook motifs of TashAT1 and TashAT3 with those of the HMGI(Y) protein. Rows 1-4 are the AT hook motifs of TashAT1/3. Rows 5-7 are those of HMGI(Y). Accession Numbers for the TashAT cluster: TashAT1, AJ291829; TashAT2, AJ132045; TashAT3, AJ291830.

Fig. 2.

Sequence analysis of TashAT1 and TashAT3 and their comparison with TashAT2. (A) Sequence of the predicted translation products of TashAT1 and TashAT3. The potential signal sequence is in italics. The AT hook DNA-binding domain is in bold and the putative transcriptional transactivation domain is underlined. Where the sequence of TashAT1 differs from that of TashAT3 is shown above the relevant sequence of TashAT3. The asterisks mark the end of each sequence. (B) A comparison of TashAT2 with TashAT3. Identical residues are shown in bold. The numbers at the end of the comparison denote the region which is 100% identical between the two sequences. A gap generated by the analysis in the sequence of TashAT2 is shown by a series of dashes. (C) A comparison of the AT hook motifs of TashAT1 and TashAT3 with those of the HMGI(Y) protein. Rows 1-4 are the AT hook motifs of TashAT1/3. Rows 5-7 are those of HMGI(Y). Accession Numbers for the TashAT cluster: TashAT1, AJ291829; TashAT2, AJ132045; TashAT3, AJ291830.

Immunofluorescence indicates the host nucleus as a possible location for TashAT1/3

Antisera were initially raised against the N terminus of TashAT1; however, it was subsequently determined that this sequence is also present in TashAT3. Therefore, all immunofluorescence studies carried out with this antisera cannot distinguish between TashAT1 and TashAT3. Immunofluorescence analysis of COS7 cells transfected with TashAT2 showed no reactivity using anti-TashAT1/3 indicating specific detection of TashAT1/3 (data not shown).

When anti-TashAT1/3 was used in IFAT analysis of the T. annulata-infected cells, D7B12 and D7 (Fig. 3C,D), immunoreactivity was observed against the macroschizont, displaying a punctate pattern of dots that may represent reactivity with or the region surrounding parasite nuclei, and against the host cell nucleus (Fig. 3C,D). In D7B12 cells, host nuclear reactivity was very bright in some cells, although a large variation in immunoreactivity was observed within each cell population. There was no reactivity with the uninfected control, BL20 (Fig. 3E), whereas with T. annulata-infected BL20 cells (TBL20) (Fig. 3B), faint reactivity with the host nucleus was obtained. During a differentiation time course of D7 cells, anti-TashAT1/3 reactivity against the host nuclei diminished overall, although some cells showed clear fluorescence (Fig. 3E). These cells probably represented undifferentiated macroschizont-infected cells as the differentiation process is known to be stochastic (Shiels et al., 1994EF29).

Fig. 3.

Indirect Immunofluorescence assays using antisera to TashAT1/3. (A-E) Analysed using anti-TashAT1/3 antiserum; (F) Analysed using pre-immune serum. The cells used were BL20 (A), TBL20 (B), D7B12 (C); D7 at 37°C (D), D7 at 41°C for 7 days (E), D7B12 (F). Scale bar: 10 μm.

Fig. 3.

Indirect Immunofluorescence assays using antisera to TashAT1/3. (A-E) Analysed using anti-TashAT1/3 antiserum; (F) Analysed using pre-immune serum. The cells used were BL20 (A), TBL20 (B), D7B12 (C); D7 at 37°C (D), D7 at 41°C for 7 days (E), D7B12 (F). Scale bar: 10 μm.

Differential expression of TashAT mRNA

In order to examine the relevance of the TashAT cluster to T. annulata-infected cells in general, we determined the RNA expression profile of all three TashAT genes in a range of Theileria-infected cell lines, and in the sporoblast/sporozoite and piroplasm stages of the parasite life cycle. The in vitro infected cell lines analysed were the T. annulata (Ankara)-infected cell line TaA2 and cloned cell lines derived from TaA2 (D7, D7B12, C9, and E3); low and high passage cell lines infected with T. annulata (Hidirseyh) and T. annulata (Diyarbakir); and the bovine lymphosarcoma cell line, BL20 and its T. annulata (Ankara)-infected counterpart, TBL20. mRNA species at 2.1 kb, 3.6 kb and 4 kb have been deduced to correspond to TashAT1, 2 and 3 respectively, using probes derived from each of the individual TashAT genes, which gave a more specific signal (Swan et al., 1999EF33; D.G.S., unpublished).

Northern blots were probed with the entire radiolabelled TashAT3 gene, which, out of the three TashATs, has the best overall homology with the other two TashATs. TaA2 and the cloned cell lines derived from it expressed the three mRNA species detected at 4 kb, 3.6 kb and 2.1 kb, corresponding to TashAT2, TashAT3 and TashAT1, respectively (Fig. 4A). Each of the cloned cell lines had essentially the same expression profile as the parent cell line, TaA2. TashAT1 and TashAT2 mRNA levels were more abundant than those of the TashAT3 RNA species; except in the case of D7B12 where TashAT3 mRNA levels were higher, and TashAT2 mRNA levels decreased slightly (Fig. 4A). TashAT1 expression did not alter significantly and was the most highly expressed of the TashATs in TaA2 and the cloned cell lines derived from TaA2. In marked contrast, cell lines T. annulata (Hidirseyh), and T. annulata (Diyarbakir) expressed TashAT1 at low to barely detectable mRNA levels (Fig. 4A). TashAT2 and TashAT3 mRNA species were present in these cell lines, although the levels of TashAT3 mRNA was fainter. TashAT2, however, was the only TashAT message detected in TBL20s. The TashAT probe was also used in hybridisation analysis of T. annulata (Ankara) sporoblast/sporozoite RNA and piroplasm RNA (Fig. 4A). The results indicated that TashAT3 and TashAT1 were expressed by sporoblast/sporozoites, and at this stage of the parasite life cycle, the TashAT3 message was detected at the highest level (Fig. 4A). There was no signal detected in piroplasm RNA by any of the TashAT probes. Hybridisation of the same blots with the gene encoding the parasite large subunit rRNA (Fig. 4B) confirmed that the mRNA levels of all three TashAT genes can vary depending on the cell line and passage number.

Fig. 4.

(A) Northern analysis of T. annulata-infected and uninfected cell lines probed with the entire TashAT3 gene. (B) The same northern blots probed with 2P3 (large subunit rRNA of T. annulata). DIY, Diyarbakir; Hid, Hiderseyh; P, passage number; Pi, piroplasms; Sp, sporozoite.

Fig. 4.

(A) Northern analysis of T. annulata-infected and uninfected cell lines probed with the entire TashAT3 gene. (B) The same northern blots probed with 2P3 (large subunit rRNA of T. annulata). DIY, Diyarbakir; Hid, Hiderseyh; P, passage number; Pi, piroplasms; Sp, sporozoite.

Immunoblot analysis using anti-TashAT1/3

To determine whether anti-TashAT1/3 reactivity against the host nucleus in D7B12 cells was due to recognition of TashAT1, TashAT3, or both, immunoblot analysis was carried out on extracts of whole cells and host or parasite enriched nuclear fractions (Fig. 5). A faint band at 180 kDa and a stronger band at 66 kDa were specifically detected in whole cell extracts, relative to pre-immune serum, by anti-TashAT1/3. Partitioning of host and parasite nuclear fractions, showed a clear enrichment of the 180 kDa band in the host nuclear fraction, while the 66 kDa band was detected at increased levels in the parasite-enriched nuclear fraction. Assessment of the origin of these polypeptides, host or parasite, was performed by analysis of an uninfected BL20 extract. Faint recognition of a band at 66 kDa, along with a band of 125 kDa, indicated that the anti-TashAT1/3 antisera crossreacted with bovine-derived polypeptides. It was concluded that the most likely candidate for a parasite-derived molecule specifically detected by the antisera was the 180 kDa polypeptide. From the predicted size of the ORFs, this polypeptide is unlikely to be encoded by TashAT1 but could be derived from TashAT3.

Fig. 5.

Western analysis of cell and nuclear extracts using anti-TashAT1/3 antiserum. (1) D7B12 total cell extract. (2) D7B12 host nuclear fraction. (3) D7B12 parasite nuclear fraction. (4) BL20 nuclear fraction.

Fig. 5.

Western analysis of cell and nuclear extracts using anti-TashAT1/3 antiserum. (1) D7B12 total cell extract. (2) D7B12 host nuclear fraction. (3) D7B12 parasite nuclear fraction. (4) BL20 nuclear fraction.

TashAT polypeptides bind specifically to AT rich DNA

To test TashAT2 for DNA binding, a recombinant protein representing the AT hook region and an upstream basic region of TashAT2 fused to the GST gene (GST-TashAT2) was used in DNA-binding assays with a radiolabelled randomised double-stranded oligonucleotide. Cycles of binding followed by PCR amplification were carried out in an attempt to enrich for DNA with affinity for GST-TashAT2. Protein-DNA complexes obtained with the enriched DNA sequences were visualised by mobility shift gel electrophoresis and after four binding cycles, an increase in the DNA-protein complex was clearly visible (Fig. 6A). The complex from cycle 4 was eluted from the mobility shift gel, PCR amplified and a further mobility shift (Fig. 6A) selected for a more specific DNA-protein complex. DNA eluted from this complex was PCR amplified and subcloned. The sequences obtained from the inserts of 12 subclones are shown in Fig. 6B. Although the sequences are not the same, they are all AT rich and show general similarities. Thus, there are three AT-rich regions discernible in each sequence separated by one to four G or C nucleotides, and a high proportion of the AT-rich regions contain the sequence ATTTA or TAAAT.

Fig. 6.

DNA-binding analysis of TashAT2. (A) Electrophoretic mobility shift assays carried out with the radiolabelled PCR product obtained after each round of binding to the GST-TashAT fusion protein. Lanes 1-4 represent cycles 1-4. Lane 5 represents cycle 4 and lane 6 represents the major band from cycle 4 excised and PCR amplified. (B) DNA sequence obtained after excising and PCR amplifying the band from lane 6 and sub cloning into pGEM7ZF. The TAAAT/ATTTA motif is shown in bold.

Fig. 6.

DNA-binding analysis of TashAT2. (A) Electrophoretic mobility shift assays carried out with the radiolabelled PCR product obtained after each round of binding to the GST-TashAT fusion protein. Lanes 1-4 represent cycles 1-4. Lane 5 represents cycle 4 and lane 6 represents the major band from cycle 4 excised and PCR amplified. (B) DNA sequence obtained after excising and PCR amplifying the band from lane 6 and sub cloning into pGEM7ZF. The TAAAT/ATTTA motif is shown in bold.

As part of a separate investigation into stage differentiation of T. annulata, a λgt11 library of D7 genomic DNA had been screened for proteins that bind a DNA motif located upstream of the Tams1 gene, which encodes the major T. annulata merozoite surface protein (Shiels et al., 2000EF30). One of the concatenated oligonucleotide probes, CAT1 (Fig. 7A), bound to a λgt11 clone that expressed a fragment of TashAT1 containing the AT hook DNA-binding domain. To test whether this binding was sequence specific and whether it related to the specific mobility shifts observed for the Tams1 promoter (Shiels et al., 2000EF30), and to determine which regions of the probe were required for binding to the TashAT1 fusion protein, the purified lambda clone was probed with three more concatenated oligonucleotides (Fig. 7A). CAT1M2 is the same as CAT1 but with a three base pair change, CAT3 is composed of the GC 3′ region of CAT1 and CAT2 overlaps with CAT1 with 10 extra bases 5′ and missing the GC-rich region of CAT3. This demonstrated that concatenated CAT3 bound weakly to the TashAT1 λgt11 clone in comparison with the CAT1, CAT1M2 and CAT2 probes. Binding reactions using the CAT1M2 (Fig. 7B, panel A) and CAT3 (Fig. 7B, panel B) probes are shown. Neither probe displayed affinity for a control λgt11clone purified from the same library (Fig. 7B, panels C,D) that was shown to express recombinant polypeptide by reactivity with bovine antiserum raised against the parasite (data not shown).

Fig. 7.

DNA-binding analysis of TashAT1. (A) Double stranded oligonucleotides used in the binding analysis. (B) Panels A and B show binding of CAT1M2 and CAT3 probes respectively to a λgt11 clone expressing a fragment of TashAT1 bearing the AT hook motif. Panels C and D show binding of CAT1M2 and CAT3, respectively, to a λgt11 clone expressing an unrelated T. annulata gene.

Fig. 7.

DNA-binding analysis of TashAT1. (A) Double stranded oligonucleotides used in the binding analysis. (B) Panels A and B show binding of CAT1M2 and CAT3 probes respectively to a λgt11 clone expressing a fragment of TashAT1 bearing the AT hook motif. Panels C and D show binding of CAT1M2 and CAT3, respectively, to a λgt11 clone expressing an unrelated T. annulata gene.

T. annulata infection of the bovine host cell alters bovine gene expression and induces host cell proliferation until the parasite progresses towards differentiation to the merozoite. This results in a marked reduction, and eventual cessation, of host cell division, followed by the destruction of the host cell as merozoites are released. Previous work and this study have characterised a family of DNA-binding proteins encoded by the parasite that display characteristics suggestive of an involvement in the regulation of host cell gene expression and, perhaps, leukocyte proliferation. All three members of the TashAT family possess putative signal sequences as well as nuclear localisation signals, demonstrating that they contain the structural information that would allow them to be secreted from the macroschizont and be transported into the host cell nucleus. TashAT gene expression is downregulated during differentiation at a time point coincident with the initial reduction in host cell division, and immunofluorescence studies suggest that the location of at least two of the TashATs, TashAT2 (Swan et al., 1999EF33) and TashAT3, is likely to be the host nucleus. As well as bearing AT hook motifs, TashAT2 has been shown in this study to bind preferentially to AT-rich DNA. This was demonstrated by the binding of the AT-rich double stranded oligonucleotides by the AT hook domain of TashAT2. Moreover, variations of the sequence TAAAT flanked by GC-rich sequences were often present and repeated in the binding site, perhaps reflecting the presence of three AT hook motifs in TashAT2, which would increase the binding affinity. Interestingly, the binding sites determined for the TashAT2 fusion protein are very similar to the DNA-binding sites determined for the AT hook DNA-binding HMGI(Y) protein in the enhancer/promoter region of the human β-interferon gene (Du et al., 1993EF11). However, the DNA sequence bound in vivo by the complete TashAT2 polypeptide could be more specific and could depend on other host or parasite co-factors binding in a complex. TashAT1 can also bind preferentially to AT-rich DNA and, although the CAT1 and CAT1M2 probes contain a GC-rich region, a concatenated probe representing this region on its own (CAT3) did not bind to the TashAT1 λgt11 expression clone. By contrast, an upstream probe, CAT2, which represented the AT-rich region of CAT1 showed binding activity. These results suggest that the AT-rich region of the CAT1 probe is required for binding. CAT1 forms part of the putative promoter region of Tams1, but there is no further evidence to date, that TashAT proteins bind specifically to the CAT1 region and regulate expression of Tams1. In fact, changing three bases of CAT1 did not alter the binding characteristics of TashAT1 significantly, whereas this alteration did abolish specific binding by polypeptides in T. annulata parasite enriched nuclear extracts (Shiels et al., 2000EF30). One possibility for the detection of TashAT1 by the CAT1 motif is that concatenated DNA-binding probes can isolate proteins that bind to related DNA motifs, as shown by the isolation of HMGI(Y) by a screen with a concatenated octamer motif (Eckner and Birnstiel, 1989EF12).

The AT hook domains of TashAT2 and TashAT1/3 are very similar, but there are some notable differences. TashAT2 has three AT hook motifs, whereas TashAT1/3 has four. Also present in TashAT1/3 are two small basic repeats, similar in sequence to regions found in the HMG1/2 DNA-binding domains (Landsman and Bustin, 1993EF20). It could be proposed, therefore, that TashAT1/3 has a stronger affinity for DNA than TashAT2 and that their sequence specificities will be different.

The individual members of the TashAT family show remarkable sequence conservation. In particular, TashAT1 is virtually identical to the 5′ part of TashAT3, and TashAT2 and TashAT3 are very similar over a region beginning at the AT hook domain. This suggests a very recent duplication event has occurred, possibly owing to adaptation to in vitro cell culture. Southern blotting of DNA derived from a range of in vitro cell lines and in vivo derived piroplasm DNA, however, indicates the existence of all three gene copies (data not shown). Therefore, it would appear that gene duplication has occurred in vivo and is not unique to a few in vitro infected cell lines. Identical copies of genes have been found in other apicomplexan parasites; there are two identical copies of the elongation factor α in Plasmodium knowlesi and Plasmodium berghei (Vinkenoog et al., 1998EF35), of the rhoptry associated protein from Babesia bovis (Suarez et al., 1998EF31) and of the Rop2 gene from Toxoplasma gondii (Beckers et al., 1997EF4).

Northern blot analysis of various cell lines indicated that in general, TashAT2 was found to be the most consistently and highly expressed of the genes. This may be related to the previous detection of anti-TashAT2 reactivity in macroschizont-infected in vivo derived cells (D. G. S. and B. R. S., unpublished). Analysis of RNA from tick derived sporozoites showed differences in the TashAT expression profile. TashAT3 is the major TashAT RNA to be expressed in sporoblasts/sporozoites, while the RNA species representing TashAT1 was also detected. The relationship between these expression patterns and the functional role of TashAT1/3 is not clear but their primary role might be during sporoblastogenesis in the tick salivary gland. Alternatively, they could function to allow the establishment of the parasite following sporozoite invasion of the bovine leukocyte. Either of these roles could involve transport to the host cell nucleus and modulation of tick or bovine gene expression.

The levels of TashAT1 and TashAT3 mRNA were found to be plastic across different cell lines. Thus, TashAT3 was less abundant than TashAT2 in the majority of cell lines, but more abundant in the D7B12 cell line. A higher level of TashAT3 expression in D7B12 cells was supported by immunoblot and immunofluorescence data. In a similar fashion, TashAT1 mRNA levels were extremely low in the non-cloned cell lines. The reasons for this plasticity in expression are unclear but may be related to the derivation of individual cell lines or clones. Furthermore, it is possible that the host cell background can influence the expression of the TashAT gene cluster, as expression levels were significantly lower in the TBL20 line derived from sporozoite infection of previously immortalised BL20 cells. One possible consequence of the plasticity of TashAT expression is that it may relate to the observed differences in bovine gene expression displayed by individual cell lines and different passages of the same cell line or clone (Adamson et al., 2000EF1; Sutherland et al., 1996EF32).

The presence of three TashAT genes in a small cluster, coupled to the possible presence of TashAT2 and TashAT1/3 in a host nucleus, provokes the idea that at least two (and possibly more) parasite genes are involved in modulation of the host environment by altering the control of leukocyte or tick cell gene expression. Whether these proteins target different subsets of genes is unknown, but modulation of host cell environment could include induction of host cell division, as rearrangements of the AT hook that contains HRX gene fused to a variety of partners are common in several types of leukaemia (Waring and Cleary, 1997EF36). Furthermore, the archetypal AT hook-containing protein, HMGI(Y) and the closely related HMGIC are upregulated in proliferating, non-differentiated cells, and chromosomal translocations involving HMGI genes are found in many types of neoplasia (Hess, 1998EF14). Thus, AT hook DNA-binding proteins play a very important role in the control of eukaryotic gene regulation and proliferation. It is therefore not unreasonable to speculate that the TashAT cluster performs similar tasks in Theileria-infected cells.

We are grateful to Professor Duncan Brown and the C.T.V.M, University of Edinburgh, for providing the T. annulata-infected cell lines; Diyarbakir and Hidirseyh. D. G. S., K. P. and C. O. were supported by grants from the Wellcome Trust, R. S. by a BBSRC studentship, and L. S. by a MRC studentship.

Adamson, R., Logan, M., Kinnaird, J., Langsley, G. and Hall, R. (
2000
). Loss of matrizx metalloproteinase 9 activity in Theielria annulata-attenuated cells is at the transcriptional level and is associated with differentially expressed AP-1 species.
Mol. Biochem. Parasitol.
106
,
51
-61.
Aravind, L. and Landsman, D. (
1998
). AT-hook motifs identified in a wide variety of DNA-binding proteins.
Nuclei Acids Res.
26
,
4413
-4421.
Baylis, H., Megson, A. and Hall, R. (
1995
). Infection with Theileria annulata induces expression of matrix metalloproteinase 9 and transcription factor AP-1 in bovine leukocytes.
Mol. Biochem. Parasitol.
69
,
211
-222.
Beckers, C. J. M., Wakefield, T. and Joiner, K. A. (
1997
). The expression of Toxoplasma proteins in Neospora caninum and the identification of a gene encoding a novel rhoptry protein.
Mol. Biochem. Parasitol.
89
,
209
-223.
Botteron, C. and Dobbelaere, D. (
1998
). AP-1 and ATF-2 are constitutively activated via the JNK pathway in Theileria parva-transformed cells.
Biochem. Biophys. Res. Commun.
246
,
418
-421.
Carrington, M., Allsop, B., Baylis, H., Malu, N. M., Shochat, Y. and Sohal, S. (
1995
). Lymphoproliferation caused by Theileria parva and Theileria annulata. In
Molecular Approaches to Parasitology
, pp.
43
-56. New York: Wiley Liss.
Church, G. M. and Gilbert, W. (
1984
). Genomic sequencing.
Proc. Natl. Acad. Sci. USA
81
,
1991
-1995.
Devereux, J., Haeberli, P. and Smithies, O. (
1984
). A comprehensive set of sequence analysis programs for the VAX.
Nucleic Acids Res.
12
,
387
-395.
Dobbelaere, D., Coquerelle, T. M., Roditi. I., Eichhorn. M. and Williams, R. O. (
1988
). Theileria parva infection induces autocrine growth of bovine lymphocytes.
Proc. Natl. Acad. Sci. USA
85
,
4730
-4734.
Dschunkowsky, E. and Luhs, J. (
1904
). Die Piroplasmosen der Rinder. Centralblatt fur Bakteriologie, Parasitenkunde.
Infektion. Hyg.
35
,
486
-493.
Du, W., Thanos, D. and Maniatis, T. (
1993
). Mechanism of transcriptional synergism between distinct virus-inducible enhancer elements.
Cell
74
,
887
-898.
Eckner, R. and Birnstiel, M. L. (
1989
). Cloning of cDNAs coding for human HMG I and HMG Y proteins: both are capable of binding to the octamer motif.
Nucleic Acids Res.
17
,
5947
-5959.
Fell, A. H., Preston, P. M. and Ansell, J. D. (
1990
). Establishment of Theileria-infected bovine cell-lines in scid mice.
Parasite Immunol.
12
,
335
-339.
Hess, J. L. (
1998
). Chromosomal translocations in benign tumors: the HMGI proteins.
J. Clin. Pathol.
109
,
251
-261.
Heussler, V. T., Machado, J., Jr, Fernandez, P.C., Botteron, C., Chen, C-G., Pearse, M. J. and Dobbelaere D. A. E. (
1999
). The intracellular parasite Theileria parva protects T-cells from apoptosis.
Proc. Natl. Acad. Sci. USA
96
,
7312
-7317.
Hulliger, L. (
1965
). Cultivation of three species of Theileria in lymphoid cells in vitro.
J. Protozool.
12
,
649
-655.
Irvin, A. D. and Morrison, W. I. (
1987
). Immunopathology, Immunology and Immunoprophylaxis of Theileria infection. In
Immune Responses in Parasitic Infections: Immunopathology, Immunology and Immunoprophylaxis
(ed. E. J. L. Soulsby), pp.
223
-274. Boca Raton FL: CRC Press.
Irvin, A. D., Brown, C. G. D. and Kanhai, G. K. (
1975
). Comparative growth of Bovine lymphosarcoma cells and lymphoid cells infected with Theileria parva in athymic (nude) mice.
Nature
255
,
713
-714.
Johnson, K. R., Lehn, D. A., Elton, T. S., Barr, P. J. and Reeves, R. (
1988
). Complete murine cDNA sequence, genomic structure and tissue expression of the high mobility group protein HMG-I(Y).
J. Biol. Chem.
34
,
18338
-18342.
Landsman, D. and Bustin, M. (
1993
). A signature for the HMG-1 box DNA-binding proteins.
BioEssays
15
,
539
-546.
McHardy, N., Wekesa, L. S., Hudson, A. T. and Randall A. W. (
1985
). Antitheilerial activity of BW720C (buparvaquone): a comparison with parvaquone.
Res. Vet. Sci.
39
,
29
-33.
Morzaria, S. P., Roeder, P. L., Roberts, D. H., Chasey, D. and Drew, T. W. (
1982
). Characteristics of a continuous suspension cell line derived from a calf with sporadic bovine leukosis. In
Fifth International Symposium on Bovine Leukosis
(ed. O. C. Straub), pp.
519
-528. Luxembourg: Commission of the European Communities.
Ole-MoiYoi, O. K., Brown, W. C., Lams, K. P., Nayer, A., Tsukamoto, T. and Macklin M. D. (
1993
). Evidence for the induction of casein kinase II in bovine lymphocytes by the intracellular protozoan parasite Theileria parva.
EMBO J.
12
,
1621
-1631.
Pipano, E. and Shkap, V. (
1979
). Attenuation of two Turkish strains of Theileria annulata.
J. Protozool.
26
,
80
.
Pollock, R. and Treisman, R. (
1990
). A sensitive method for the determination of protein-DNA binding specificities.
Nucleic Acid Res.
18
,
6197
-6204.
Reeves, R. and Nissen, M. S. (
1990
). The AT-DNA-binding domain of mammalian high mobility group I chromosomal proteins.
J. Biol. Chem.
265
,
8573
-8582.
Shiels, B. R., McDougal, C., Tait, A. and Brown, C. G. D. (
1986
). Identification of infection-associated antigens in Theileria annulata transformed cells.
Parasite Immunol.
8
,
69
-77.
Shiels, B. R., Kinnaird, J. H., McKellar, S., Dickson, J., Ben Miled, L., Melrose, R., Brown, D. and Tait, A. (
1992
). Disruption of synchrony between parasite growth and host cell division is a determinant of differentiation to the merozoite in Theileria annulata.
J. Cell Sci.
107
,
99
-107.
Shiels, B. R., Smyth, A., Dickson, J., McKellar, S., Tetley, L., Fujisaki, K., Hutchison, B. and Kinnaird, J. H. (
1994
). A stoichiometric model of stage differentiation in the protozoan parasite Theileria annulata.
Mol. Cell. Diff.
2
,
101
-125.
Shiels, B. R., Fox, M., Mckellar, S., Kinnaird, J. and Swan, D. G. (
2000
). An upstream element of the Tams1 gene is a site of DNA-protein interactions during differentiation to the merozoite in Theileria annulata.
J. Cell Sci.
113
,
2243
-2252.
Suarez, C. E., Palmer, G. H., Hötzel, I., Hines, S. A. and McElwain, T. F. (
1998
). Sequence and functional analysis of the intergenic regions separating Babesial rhoptry-associated protein-1 (rap-1) genes.
Exp. Parasitol.
90
,
189
-194.
Sutherland, I. A., Shiels, B. R., Jackson, L., Brown, D. J., Brown C. G. D. and Preston, P. M. (
1996
). Theileria annulata: altered gene expression and clonal selection during continuous in vitro culture.
Exp. Parasitol.
83
,
125
-133.
Swan, D. G., Phillips, K., Tait, A. and Shiels, B. R. (
1999
). Evidence for localisation of a Theileria parasite AT hook DNA-binding protein to the nucleus of immortalised bovine host cells.
Mol. Biochem. Parasitol.
101
,
117
-129.
Triezenberg, S. J. (
1995
). Structure and function of transcriptional activation domains.
Curr. Opin. Gen. Dev.
5
,
1990
-1996.
Vinkenoog, R., Aparecida-Sperança, M., van Breemen, O., Ramesar, J., Williamson, D. H., Ross-MacDonald, P. B., Thomas, A. W., Janse, C. J., del Portillo, H. A. and Waters, A. P. (
1998
). Malaria parasites contain two identical copies of an elongation factor 1 alpha gene.
Mol. Biochem. Parasitol.
94
,
1
-12.
Waring, P. M. and Cleary, M. L. (
1997
). Disruption of a homolog of trithorax by 11q23 translocations: leukemogenic and transcriptional implications.
Curr. Top. Microbiol. Immunol.
220
,
1
-23.
Williamson, S., Tait, A., Brown, D., Walker, A., Beck, P., Shiels, B., Fletcher, J. and Hall, R. (
1989
). Theileria annulata sporozoite surface antigen expressed in Escherichia coli elicits neutralising antibody.
Proc. Natl. Acad. Sci. USA
86
,
4639
-4643.