MicroRNAs 221 and 222 target p27Kip1 in Marek's disease virus-transformed tumour cell line MSB-1

MicroRNAs (miRNAs) are a class of 21–25 nt RNAs that can negatively regulate protein expression at the post-transcriptional level by base-pairing with target mRNAs (He & Hannon, 2004). The targets of miRNAs appear to be extremely wide and varied and have been linked to physiological processes including apoptosis, proliferation, differentiation and cell survival (Williams, 2008). In addition, recent evidence implicates the involvement of miRNAs in several tumours (reviewed by Croce, 2008). It has been reported that many miRNAs are aberrantly expressed in cancers and several miRNAs themselves act as oncogenes, while others have predicted tumour suppressor activities (Dalmay, 2008; Gartel & Kandel, 2008). Among the fundamental mechanisms involved in neoplastic transformation, the disruption of the cell cycle regulatory machinery is crucial in several forms of cancer (Viglietto et al., 2002). A number of studies have recently described the capacity of two potentially oncogenic miRNAs, miR-221 and miR-222, to target and suppress proteins that are important in the regulation of cell cycle and cell proliferation in different forms of human cancers. These targets include the cyclin-dependent kinase (cdk) inhibitors p27^Kip1 (Fornari et al., 2008; Galardi et al., 2007; Gillies & Lorimer, 2007; le Sage et al., 2007b; Visone et al., 2007) and p57^Kip2 (Fornari et al., 2008; Medina et al., 2008) and the c-kit receptor (Felli et al., 2005).

Marek's disease (MD) is a highly infectious lymphoproliferative disease of poultry caused by an oncogenic herpesvirus of the family Herpesviridae (ICTVdB Management, 2006). Forms of MD include acute neoplastic disease with lymphomatous lesions in several organs, severe cytolytic forms with extensive lymphoid atrophy, and an acute form of transient paralysis (Gimeno et al., 2001). The immediate nature of the onset of MD lymphomas, sometimes as early as within a few weeks after infection, indicate rapid activation of genes associated with oncogenicity, possibly through virus-induced genes. We and others have previously shown the role of MDV oncoproteins such as Meq in the induction of MD lymphomas (Brown et al., 2006; Lupiani et al., 2004). More recent studies have demonstrated high levels of expression of MDV-encoded as well as host-encoded miRNAs in MD lymphomas and tumour-derived T-cell lines (Burnside et al., 2006; Yao et al., 2007), suggesting a major role for these non-coding RNAs in MD tumours. Although the precise functions of many of these miRNAs are yet to be identified, analysis of the miRNA profile of MD lymphomas would be valuable to help elucidate their roles in oncogenesis. Since primary MD lymphomas are often heterogeneous mixtures of neoplastic T cells and non-transformed cells of other lineages (Payne et al., 1981), the analysis of the whole tumour may not provide the precise miRNA profile of the transformed target cell. However, as MDV-transformed lymphoblastoid cell lines derived from primary MD tumours are homogeneous clonal populations of tumour cells, they can be used more accurately to examine the expression profiles of MD tumours (Buza & Burgess, 2007).

MSB-1, a CD4⁺ T-cell line derived from a spleen lymphoma induced by the BC-1 strain of MDV-1, (Akiyama & Kato, 1974; Hirai et al., 1990) is capable of inducing tumours when inoculated into susceptible chickens (Doi et al., 1976; Lee et al., 1975). We have recently shown that MSB-1 expresses high levels of MDV-1- and MDV-2-encoded miRNAs (Xu et al., 2008; Yao et al., 2007, 2008), many of which are expressed at much higher levels than those in infected chicken embryo fibroblasts (CEF) (Burnside et al., 2006; Burnside & Morgan, 2007; Burnside et al., 2008). These studies suggested that the MSB-1 lymphoblastoid cell line could be used as a model system for analysing the molecular pathways and mechanisms of neoplastic transformation in MD tumours, particularly those involving the miRNA pathways. Analysis of a cDNA library of small RNA from MSB-1 cells showed that, in addition to the MDV-encoded miRNAs, several host-encoded miRNAs are expressed at high levels in MSB-1 cells (Yao et al., 2007, 2008). These included miR-221 and miR-222, both of which have been reported to be upregulated in several types of human cancers where they target the cell cycle regulatory protein p27^Kip1 (reviewed by le Sage et al., 2007a; Chu et al., 2008). Although this dysregulation of p27^Kip1 has been observed in many human cancers, no such observations have yet been reported in any of the virus-induced cancers. On the basis of our observations of increased expression of miR-221 and miR-222 in MSB-1 cells by microarray analysis, we examined whether chicken p27^Kip1 is a target for these miRNAs and affect its expression levels in MSB-1 cells.

Reporter, miRNA and antagomiR expression vectors.
Reporter plasmids for the p27^Kip1 3' untranslated region (UTR) were produced by ligating the target sequences downstream of the Renilla luciferase gene using the NotI–XhoI site in psiCHECK-2 vector (Promega). The entire p27^Kip1 3'UTR was amplified from chicken spleen cDNA by PCR using forward primer (5'-GGAGGACTCGCGTTTCCTTGCTCAT-3') and reverse primer (5'-CTACATTGCCTGTGTACCTGTATGT-3') and cloned initially into pGEM-T Easy (Promega). To construct miRNA and antagomiR expression plasmids, the chicken U6-3 promoter (GenBank accession no. DQ531569[GenBank] ) was amplified from chicken genomic DNA using forward primer (5'-ATCGATGACAACACAAGCATCGAGC-3') with a ClaI site and reverse primer (5'-CAGTCTCGATGCTCTTATTCGAACTAGTGTTTAAACGGCGCGCCATCGATGAATTCAAG-3') with introduced ClaI, BstBI and AscI sites, and was cloned into pGEM-T Easy vector. Chemically synthesized complementary DNA oligonucleotides corresponding to gga-miR-221, gga-miR-222 and a non-silencing negative control miRNA (miR-NS) that showed no significant sequence identity with any sequences deposited in GenBank with the miR-30 loop, all with BstBI and AscI overhangs, were annealed and inserted into the chicken U6 plasmid. Oligonucleotides for the top and the bottom strands, respectively, were 5'-CGGCATGAACCTGGCATACAATGTAGATTTCTGTGTTTGTTAAGCAACAGCTACATCTGGGTTTCCTTTTTT-3' and 5'-CGCGAAAAAAGGAAACCCAGCAGACAATGTAGCTGTTGCTTAACAAACACAGAAATCTACATTGTATGCCAGGTTCATGC-3' (gga-miR-221), 5'-CGATCGCTCAGTAGTCAGTGTAGATTCTGTCTTTACAATCAGCAGCTACATCTGGCTACTGGGTCTCTTTTTT-3' and 5'-CGCGAAAAAAGAGACCCAGTAGCCAGATGTAGCTGCTGATTGTAAAGACAGAATCTACACTGACTACTGAGCGAT-3' (gga-miR-222), and 5'CGTTCTCCGAACGTGTCACGTCTGTGAAGCCACAGATGGGACGTGACACGTTCGGAGAATTTTTGGAA-3' and 5'-CGCGTTCCAAAAATTCTCCGAACGTGTCACGTCCCATCTGTGGCTTCACAGACGTGACACGTTCGGAGAA-3' (miR-NS). To synthesize the retroviral antagomiR expression vectors, the U6 promoter was cloned into RCASBP-(A)-CN-EGFP (with avian leukosis virus subgroup A envelope) (a generous gift from Dr Jon Gilthorpe, Kings College London) in the reverse orientation using a unique NotI site. Oligonucleotides for the two strands of the antagomiRs were 5'-CGGAAACCC AGCAGACAATGTAGCTTTTTTT-3' and 5'-CGCGAAAAAAAGCTACATTGTCTGCTGGGTTTC-3' (antagomiR-221) and 5'-CGGAGACCCAGTAGCCAGATGTAGCTTTTTTT-3' and 5'-CGCGAAAAAAAGCTACATCTGGCTACTGGGTCTC-3' (antagomiR-222).

Cells and transfection.
Primary cultures of CEF were prepared from 10-day-old specific-pathogen-free embryos obtained from flocks maintained at the Institute for Animal Health. In addition to MSB-1 (Akiyama & Kato, 1974), lymphoblastoid T-cell lines MDCC-226S (T226S) and MDCC-265L (T265L) derived from lymphomas of birds infected with RB-1B virus were also used in the microarray analysis. MSB-1 cells were grown at 38.5 °C in 5 % CO₂ in RPMI 1640 medium containing 10 % fetal calf serum, 2 % chicken serum, 10 % tryptose phosphate broth, 0.1 % 2-mercaptoethanol and 1 % sodium pyruvate. MSB-1 cells were transfected by resuspending 10⁶ cells in Nucleofector solution T (Amaxa) along with 2 µg plasmid DNA and electroporated using Nucleofector program A-23. DF-1 cells were transfected with 500 ng of the RCAS–antagomiR constructs using Lipofectamine 2000 (Invitrogen) following the manufacturer's instructions. Forty-eight hours post-transfection, supernatants were used for transduction of MSB-1 cells. Transfections in CEF were carried out using 500 ng plasmid DNA and Lipofectamine (Invitrogen) following the manufacturer's instructions.

Microarray analysis of miRNA expression.
Full details of the microarray analysis examining the relative expression of all miRNAs in MDV-transformed cell lines are available from the authors. Briefly, 500 ng purified small RNA from lymphoblastoid cell lines, normal splenocytes or CD4⁺ T cell populations were labelled with either Cy3 or Cy5 dyes using the Array 900microRNA RT kit from Genisphere and hybridized to miRNA microarray described previously (Lawrie et al., 2008).

Dual Luciferase assay.
Reporter vector psiCHECK-2 carrying the 3'UTR sequences of p27^Kip1 were assayed for luciferase expression using the Dual Glo Luciferase Assay System (Promega) following the manufacturer's instructions. The relative expression of target-specific Renilla luciferase was determined by taking the normalized levels compared with background firefly luciferase for each sample transfected a minimum of three replicates±standard error and is representative of at least two independent experiments.

Immunoblotting and Northern blotting.
For Western blotting, cells were lysed in protein gel sample buffer (8 M urea, 2 % SDS, 10 mM Tris/HCl pH 6.8, 0.05 % bromophenol blue) and separated on a NuPAGE 4–12 % Bis Tris gel (Invitrogen) and transferred onto nitrocellulose membranes using an iBlot gel transfer system (Invitrogen). Western blotting (WB) was performed with mouse p27^Kip1 monoclonal antibody (no. 610241; BD transduction laboratories), followed by anti-mouse IgG–peroxidase conjugate (Sigma-Aldrich). The alpha-tubulin monoclonal antibody was used as a loading control. Membranes were developed with an ECL Western blotting analysis system (Amersham). Image quantification was carried out using ImageQuant 300 software (GE Healthcare) to determine the relative levels of p27^Kip1 normalized to the loading control. For Northern blot analysis, total RNA was extracted from cultured cells with TRIzol reagent (Invitrogen) according to standard methods described by the manufacturer. Samples of 20 µg total RNA were resolved using a 15 % polyacrylamide-1xTris-borate-EDTA-8 M urea gel and blotted to a GeneScreen Plus membrane (Perkin-Elmer). DNA oligonucleotides with sequences complementary to candidate miRNAs were end-labelled with [γ-³²P]ATP (Amersham) and T4 polynucleotide kinase (New England Biolabs) to generate high-specific-activity probes. Hybridization, washing and autoradiography were carried out as previously described (Pfeffer et al., 2005; Yao et al., 2007).

miR-221 and miR-222 are upregulated in MSB-1 cells
We have previously demonstrated the relative abundance of both miR-221 and miR-222 in MSB-1 cells on the basis of the increased cloning frequency in the small RNA library made from these cells (Yao et al., 2007). As part of a separate study to identify the miRNA expression profiles associated with MDV oncogenesis, we compared the global expression of miRNAs in a number of tumour cell lines (including MSB-1 cells) with the levels in normal splenocytes or purified CD4⁺ T-cells as the reference samples by microarray analysis using miRNA probe sets designed from miRBase v.9.2. These experiments showed changes in the expression profiles of several miRNAs, some of which were common to all MD tumour cell lines, while others were restricted to selected cell lines (the complete sets of results describing the global changes in miRNA expression in MDV-transformed cell lines are available from the authors). These studies showed that both miR-221 and miR-222 were upregulated in MSB-1 in comparison with the T226S and T265L cell lines, where their levels were either unchanged or downregulated (Fig. 1a). The higher expression levels of these two miRNAs in MSB-1 cells were further examined by Northern blot hybridization using individual gga-miR-221 and gga-miR-222 probes. Stronger signals indicating higher expression were evident on the RNA samples from MSB-1 cells, compared with those from uninfected CEF or splenocytes (Fig. 1b).

(29K): Fig. 1. MiR-221 and miR-222 are expressed at high levels in the MDV-transformed cell line MSB-1. (a) Heat map representation of data from two-way cluster analysis of differentially expressed miRNAs (P<0.05) measured by microarray analysis in MDV-transformed cell lines with normal splenocyte RNA (left panel) and purified normal CD4+ T lymphocyte RNA as the reference (right panel). Nomenclature of the miRNAs has been adapted from miRBase (): gga-, Gallus gallus; hsa-, Homo sapiens; mmu-, Mus musculus. Composite P-values of the miRNA dataset from the transformed cell lines against each of the reference samples are also shown. (b) Northern blot analysis of total RNA extracted from CEF, normal splenocytes and MSB-1 cells probed for miR-221 and miR-222.

miR-221 and miR-222 can target and suppress chicken p27^Kip1
Identical in sequence to the human homologues, both gga-miR-221 and gga-miR-222 feature matching seed regions of 7 nt perfect complementarity to the putative binding sites located in the 3'UTR of p27^Kip1 reported previously (Fornari et al., 2008; Galardi et al., 2007; le Sage et al., 2007b). In fact, it appears that this region is evolutionarily conserved across a diverse range of species, suggesting that the interaction between miR-221 and miR-222 with the 3'UTR may be a critical aspect in the regulation of p27^Kip1 expression (Fig. 2a). To determine if the expression of p27^Kip1 in chicken cells could be affected directly by gga-miR-221 or gga-miR-222 alone or in combination, we generated expression plasmids for each miRNA transcribed from the chicken U6 promoter. These constructs, along with a negative control miRNA (miR-NS), were transfected into CEF that have low levels of these endogenous miRNAs and high levels of p27^Kip1. It was found that U6-expressed miR-221/miR-222 could separately or in combination reduce p27^Kip1 protein expression compared with the miR-NS control by approximately 50 % (Fig. 2b). To determine if the p27^Kip1 3'UTR was indeed directly targeted by gga-miR-221 and gga-miR-222, we then constructed reporter vectors that featured the 3'UTR in its native form (p27-3'UTR-WT) and the exact same sequence containing three base-pair mutations in each of the two predicted miRNA binding sites (p27-3'UTR-dbMT) fused to the 3'UTR of the Renilla luciferase in psiCHECK-2 (Fig. 3a). These reporter plasmids were transfected into CEF and MSB-1 cells, and the expression of Renilla luciferase was monitored and normalized against firefly luciferase (Fig. 3b). It was found that the wild-type 3'UTR in CEF was reduced by less than 20 %, whereas nearly 60 % inhibition was seen in MSB-1 cells. Reduced expression of mutant control luciferase reporter in CEF, compared with that observed in MSB-1 cells, is attributable to the low levels of miR-221 and miR-222 in CEF (see also Fig. 2b) as reported previously (Burnside et al., 2008; Yao et al., 2008). To examine the individual activities of the predicted miRNA-binding sites located in the p27^Kip1 3'UTR, we also made two additional reporter constructs (p27-3'UTR-MT1 and p27-3'UTR-MT2) with mutations in one or the other predicted miRNA-binding sites (Fig. 3a). Transfection of these constructs resulted in considerable functional repression of the p27-3'UTR-MT2 reporter construct featuring mutations in the second miRNA-binding site in MSB-1 cells compared with the double mutant control, a pattern that was almost identical to the wild-type 3'UTR vector (Fig. 3c).

(26K): Fig. 2. MiR-221 and miR-222 can target the chicken p27Kip1 3'UTR. (a) Alignment of nucleotide sequences of gga-miR-221/miR-222 and their putative binding sites in the p27Kip1 3'UTRs of Gallus gallus (NM_204256[GenBank] ), Homo sapiens (NM_004064[GenBank] ), Macaca mulatta (XM_001085433), Rattus norvegicus (NM_031762[GenBank] ), Sus scrofa (NM_214316[GenBank] ) and Felis catus (NM_001009870). (b) Total protein samples from CEF transfected with the U6/miR-221 and miR-222 plasmids were analysed by Western blot using anti-p27Kip1 antibodies (left panel). Quantification of the relative signal intensity for p27Kip1 is shown normalized to the α-tubulin loading control (right panel).

(30K): Fig. 3. The chicken p27Kip1 3'UTR is a target for miR-221 and miR-222. (a) A series of reporter vectors were constructed that feature the chicken p27Kip1 3'UTR sequence inserted downstream of Renilla luciferase in the psiCHECK-2 vector (left panel). Nucleotide sequences of the wild-type and mutant miR-221/miR-222 binding sites (mutated residues shown in bold) located in the p27Kip1 3'UTR (right panel). Luciferase reporter experiments were performed in MSB-1 and CEF, and the luciferase ratio of Renilla to firefly was used to normalize transfection efficiency. (b) Transfection of reporter vectors with either the wild-type or mutated p27Kip1 3'UTR. (c) Transfection of reporter vectors with the p27Kip1 3'UTR with mutated individual miRNA binding sites.

p27^Kip1 is repressed in MSB-1, but can be alleviated using antagomiRs
Since both miR-221 and miR-222 were consistently upregulated in MSB-1, we tested whether these miRNAs modulate the expression of p27^Kip1 in this cell line. Western blotting was used to monitor p27^Kip1 levels in CEF and normal splenocytes, both of which have low miR-221 and miR-222 levels, and in MSB-1 cells (Fig. 4a). These studies showed that there was an inverse correlation between the expression of p27^Kip1 and that of miR-221/miR-222, as both CEF and normal splenocytes produced very dense p27^Kip1 bands compared with the MSB-1 cells. In order to determine that the reduced p27^Kip1 expression in MSB-1 cells is mediated through miR-221 and miR-222, we examined whether specific miR-221/miR-222 inhibitors (antagomiR) could restore the p27^Kip1 levels in these cells. We developed retroviral expression vectors to express RNAs complementary to each mature miRNA sequence similarly to a previously described method that used a lentiviral expression system (Scherr et al., 2007). We used the replication-competent avian retrovirus vector RCASBP-(A)-CN-EGFP to generate antagomiR expression vectors for miR-221 and miR-222. This virus contains an EGFP (enhanced green fluorescent protein) marker and allows for continuous expression of RNAs in infected cells by the chicken U6 promoter, related viruses have been previously used for efficient delivery of short hairpin RNAs in chicken cells (Bron et al., 2004; Harpavat & Cepko, 2006). Transduction of the RCAS-antagomiR viruses into MSB-1 cells resulted in approximately 50–60 % of cells positive for EGFP expression after 14 days culturing (data not shown). Cells were harvested and analysed for the expression of p27^Kip1 by Western blotting (Fig. 4b). The miR-221 antagomiR (anti-221) was able to alleviate p27^Kip1 suppression by about 50 %, while miR-222 antagomiR (anti-222) could only restore about 20 % of the expression. Combination of the two antagomiRs gave similar results as anti-221 antagomiR used alone. Since high levels of miR-221/miR-222 were linked directly to the reduced expression of p27^Kip1 in MSB-1 cells, we then analysed the expression of this protein in the other MDV-transformed cell lines that were shown by microarray analysis to have unchanged or downregulated miR-221/miR-222. It was found by Western blotting that high levels of p27^Kip1 expression was evident in CEF and splenocytes only, whereas MDV-transformed tumour cell lines MSB-1, T226S and T265L had very low levels of p27^Kip1 (Fig. 4c).

(37K): Fig. 4. Chicken p27Kip1 is regulated by miR-221 and miR-222 in MSB-1 cells. (a) Total protein samples from CEF, normal splenocytes and MSB-1 cells were analysed by Western blot using anti-p27Kip1 antibodies. (b) MSB-1 cells were transduced with RCAS viruses expressing antagomiRs against miR-221, miR-222 or both, and total protein samples were analysed by Western blot using anti- p27Kip1 antibodies (top). Relative signal intensity for p27Kip1 is shown normalized to the α-tubulin loading control (bottom). (c) Total protein samples from CEF, normal splenocytes, MSB-1, T226S and T265L cells were analysed by Western blot using anti-p27Kip1 antibodies.

Oncogenic viruses are involved in a significant proportion of neoplasms in animals and are usually characterized by rapid onset and progression, often resulting from the direct effect of virus-encoded proteins. With the discovery of virus-encoded miRNAs, especially in many oncogenic herpesviruses (Gottwein & Cullen, 2008; Sullivan & Grundhoff, 2007), their roles in oncogenesis are widely being investigated. Global analysis of miRNA expression is increasingly used to understand molecular oncogenic pathways and in the diagnosis and prognosis of several human cancers. Apart from a major disease affecting poultry health, MD is a valuable natural disease model for rapid-onset herpesvirus-induced T-cell lymphomas (Osterrieder et al., 2006). The MDV-transformed lymphoblastoid cell line MSB-1 has been used extensively for a range of studies including MDV latency (Morgan et al., 2001), methylation profiles (Kanamori et al., 1987) and p53 dynamics (Takagi et al., 2006a, b). We have also carried out extensive analysis of the miRNA profiles in MSB-1 cells and showed high levels of MDV-encoded miRNAs (Yao et al., 2007, 2008). These studies also revealed altered expression of several host-encoded miRNAs, which were validated recently by global miRNA expression profiling by microarray analysis (results not shown). In all these studies, the host-encoded miRNAs miR-221 and miR-222 were consistently upregulated in MSB-1 cells.

The increased expression of these miRNAs in MSB-1 compared with splenocytes and CD4⁺ T-cells shown by microarray was confirmed also by Northern blot (Fig. 1b). The high level of these miRNAs in MSB-1 cells prompted us to test the functional relevance of these miRNAs in modulating the expression of the putative target molecules. Cell cycle regulatory protein p27^Kip1 has been demonstrated to be a bona fide target of miR-221/miR-222 in several studies (le Sage et al., 2007a; Medina et al., 2008). Since the predicted miR-221/miR-222 binding sites in the 3'UTR of the chicken p27^Kip1 are highly conserved across several species (Fig. 2a), we hypothesized that it is very likely to be targeted by these miRNAs. This assumption was confirmed by overexpression of miR-221 and miR-222 in CEF, which resulted in reduction in the expression levels of p27^Kip1 as shown by Western blotting analysis (Fig. 2b). Increased expression of these miRNAs in MDV-infected CEF has been reported based on the higher frequencies by deep sequence analysis, and the authors have speculated that the resulting downregulation of p27^Kip1 might favour growth and proliferation of infected CEF (Burnside et al., 2008). For a more precise demonstration that these miRNAs are directly involved in the regulation of p27^Kip1 in MSB-1 cells, we carried out reporter assays with p27^Kip1 3'UTR constructs carrying specific mutations in the predicted miRNA binding sites (Fig. 3). These studies clearly demonstrated the functional specificity of the miR-221/miR-222-mediated translational silencing of p27^Kip1 in both CEF and MSB-1 cells. The differences in the luciferase reporter activity between the two cell types are most likely to be the reflection of the levels of expression of the miRNAs (Fig. 1b). These studies also showed potential differences between the two predicted target sites in the 3'UTR of p27^Kip1, as no repression was observed when the first miRNA-binding site was mutated. These data suggested that not only do these miRNAs specifically target and suppress the p27^Kip1 3'UTR, they favour the first miRNA binding site and are able to achieve the same level of suppression as the wild-type sequence when only this site is present (Fig. 3c). Finally, the effect of high levels of miR-221/miR-222 expression in MSB-1 cells on p27^Kip1 expression was also examined directly using retrovirally expressed antagomiRs. The ability of the two antagomiRs to partially restore the p27^Kip1 levels further demonstrated the functional specificity of translational repression by these miRNAs. Although it is likely that a higher level of inhibition may be possible if higher levels of retroviral infection of these cells can be achieved, our data provide further evidence that these miRNAs can target p27^Kip1 in MSB-1 cells.

The role p27^Kip1 plays directly in control of cell cycle and cell proliferation is extremely well documented (Chu et al., 2008; Medina et al., 2008). Promotion of growth and proliferation as a consequence of the downregulation of p27^Kip1 by miR-221/miR-222 has been demonstrated in several types of cancer cells (Fornari et al., 2008; Galardi et al., 2007; le Sage et al., 2007b). It is therefore likely that MSB-1 cells also may use a similar fundamental mechanism for maintaining the transformed phenotype. It was interesting to find, however, that of the cell lines tested here, the increased expression of miR-221/miR-222 was unique to MSB-1 cells and was not a consistent feature of MDV transformation (Fig. 1a). The factors that distinguish MSB-1 from other cell types in relation to the miR-221/miR-222 and p27^Kip1 dynamics are not clear. However, it is possible that certain dissimilarities of MSB-1 cells compared to the T226S and T265L cells might be linked to the observed variation in miR-221/miR-222 expression. These include the co-infection of MSB-1 with BC-1 strain of MDV-1 and HPRS-24 strain of MDV-2 (Hirai et al., 1990; Yao et al., 2007), interference with the p53 pathways due to truncated p53 transcripts (Takagi et al., 2006a, b), differences in the methylation status (Kanamori et al., 1987), and that the RB-1B strain used to transform the other cell lines is more lymphomagenic than BC-1. Also of particular interest was the observed downregulation of p27^Kip1 in all MDV-tumour cell lines despite differential miR-221/miR-222 expression. These data further highlight the universal importance of reduced p27^Kip1 expression, suggesting that other mechanisms are working in these cells to suppress this protein, and that this process may be fundamental to MDV oncogenesis. Nevertheless, our study clearly provides evidence for miR-221/miR-222-mediated translational regulation of cell cycle regulatory protein p27^Kip1 in a herpesvirus-transformed tumour T-cell line. As far as we know, this is first time that the upregulation of miR-221 and miR-222 and their interaction with p27^Kip1 has been demonstrated in a virus-induced lymphoma cell line. Thus, the findings from our studies demonstrate that an oncogenic virus can also exploit at least some of the pathways such as the miRNA-mediated dysregulation of cell cycle, in a manner similar to those seen in some of the common non-infectious forms of cancer.

This work was partly funded by the Biotechnology & Biological Sciences Research Council (BBSRC), UK. Bioinformatics support from Mick Watson, Institute for Animal Health, Compton, UK, is acknowledged.