By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
The fusion gene of peste des petits ruminants virus (PPRV-F), a paramyxovirus, contains an unusual long 5′ untranslated region (5′ UTR) with a high GC content that is capable of folding into secondary structure proximally to the 5′ cap. Sequence analysis further suggested that the proximal end of this UTR contains a nine-nucleotide sequence which could perfectly complement the 18S rRNA and might affect translation through mRNA–rRNA interaction. Based on these features, we examined the functional role of the proximal PPRV-F 5′ UTR on translational efficiency compared with two other morbilliviruses. From reporter gene assays, PPRV-F 5′ UTR functioned as a strong enhancer of translational efficiency independent of cell and gene specificity. Northern blot analysis of the accumulative RNA levels and mRNA stability suggested that elevated gene expression driven by PPRV-F 5′ UTR was accompanied by an increased mRNA level and enhanced mRNA stability. Deletion analysis identified the complementary sequence and distal nucleotides necessary for the enhancing activity, and results suggest RNA structural conformation is important. Taken together, we conclude that the proximal PPRV-F 5′ UTR functions as a translational enhancer by promoting translation efficiency and mRNA stability.
Under some competitive cellular environments, several viruses developed many strategies to translate their genes [1, 2]. Some RNA viruses contain highly structured motifs that mediate post-transcriptional mechanisms, including the splicing of transcripts, mRNA stability and translation to ensure efficient expression during viral replication [2-4]. Like cellular mRNAs [5, 6], 5′ untranslated sequences with stable RNA structures have been found to participate in viral translation regulation. A stem-loop RNA structure located in the 5′ UTR of retroviruses has a positive effect on viral and reporter gene expression by enhancing mRNA stability and translational efficiency [2, 3, 7]. Additionally, some examples of translational control through the interaction of 18S rRNA and its complementary sequence within the 5′ UTR of mRNA transcripts have also been found in cellular and viral translation [8-10]. Although the presence of these functional domains in viral mRNA has been described, the physiological relevance supporting a proposed mechanism remains to be determined. As some conflicting results of RNA structural prediction do arise in the literature , the critical data from experimental approaches are necessary to determine these functional domains.
Peste des petits ruminants is a highly contagious viral disease of goats and sheep. The causative agent, peste des petits ruminants virus (PPRV), belongs to the genus Morbillivirus of the family Paramyxoviridae . The genome of PPRV is a linear nonsegmented single-stranded, negative-sense RNA. It is well known that a viral RNA-dependent RNA polymerase (vRdRp) produces a decreasing gradient of monocistronic mRNAs for morbillivirus genes closest to the 5′-end of the viral genome . Therefore, the expression of each morbillivirus protein primarily relies on its own mRNA level controlled by vRdRp. Nevertheless, several recent publications have explored the regulatory effect of morbillivirus-UTRs on gene expression and viral replication [14-16]. Morbillivirus mRNAs generally contain open reading frames with short 5′ and 3′ UTRs. However, in contrast to these mRNAs, the matrix gene and fusion gene (F) mRNAs have unusually long 3′ and 5′ UTRs, respectively . It was hypothesized that the compact genome organization of RNA viruses offers an advantage for their rapid replication , and therefore the extra long sequences of morbillivirus-UTRs are likely to be of functional importance.
In the present study, we found that a feature of the long 5′ UTR of PPRV-F gene is its high GC content, which is able to organize into an extensive secondary structure proximal to the 5′ cap structure. Moreover, a nine-nucleotide sequence with perfect complementarity to 18S rRNA was observed at the proximal end of PPRV-F 5′ UTR. Interested by these findings, our aim was to evaluate whether the proximal 5′ UTR sequence of the PPRV-F gene is capable of regulating gene expression particularly the translation mechanism. As indicated by a series of reporter gene assays, proximal PPRV-F 5′ UTR strongly enhanced the translation efficiency. Determination of the accumulative RNA level and mRNA stability by northern blot analyses revealed that the presence of proximal PPRV-F 5′ UTR was able to increase the steady-state mRNA level and the stability of mRNA. The results from this study strongly suggest the presence of translational enhancing activity in PPRV-F 5′ UTR.
Sequence analysis and structural prediction of proximal PPRV-F 5′ UTR
In terms of GC content, the sequences of 5′ UTRs of morbillivirus F gene can be categorized into high and low GC contents: high GC content was found in F 5′ UTRs of PPRV (69%), measles virus (MV) (65%) and canine distemper virus (CDV) (64%), whereas the F 5′ UTRs of phocine distemper virus and dolphin morbillivirus contain lower GC contents of 49% and 46%, respectively. Among these viruses, the 5′ UTR of PPRV-F gene has the most GC-rich sequence with the longest length – 630 nucleotides. To gain perspective on this high GC sequence, we first determined the possible RNA structure of this UTR by using the Zuker algorithm of the mfold program [19, 20]. Prediction of the thermodynamically stable secondary structure of the PPRV-F 5′ UTR suggested the presence of several stem-loops throughout this long UTR with the most stable stem-loops spanning along the 5′ end of mRNA (data not shown). Of particular interest is the proximal mRNA structure near the 5′ cap that would affect translation initiation [21, 22]; the structures of the proximal 85 nucleotides of PPRV-F 5′ UTR with minimal free energy (ΔG) were obtained. For comparative purposes, we included two additional morbillivirus-F 5′ UTR sequences with high GC content, i.e. the proximal 85 nucleotides of MV and the full-length 85 nucleotides of CDV. As shown in Fig. 1A, all these viral UTR sequences were able to fold into secondary structures with different minimal free energies. The secondary structure of proximal PPRV-F 5′ UTR has a free energy of −30.20 kcal·mol−1 which is lower than those of the other two UTRs. As a lower free energy indicates a more stable structure , the proximal PPRV-F 5′ UTR could be able to fold into the most energetically favored RNA structures compared with other UTRs.
Furthermore, we analyzed the potential 18S rRNA binding site in these nucleotide sequences. Interestingly, a nine-nucleotide putative 18S rRNA binding site was found in proximal PPRV-F 5′ UTR, located at nucleotides 29–37, with a perfect Watson–Crick complementarity to highly conserved sequences of mammalian 18S rRNAs (Fig. 1B, upper panel). The proximal MV-F 5′ UTR showed partial complementarity (Fig. 1B, lower panel), whereas none of the CDV-F 5′ UTR sequence is complementary to mammalian 18S rRNA. The existence of a stable RNA structure and 18S rRNA binding site was previously described in some functional UTR to control the translation [8, 23]. Nevertheless, the contrary observation was also demonstrated: certain structures of the mRNA located near the 5′ cap have been found to drastically inhibit translation initiation by preventing the binding of ribosomes to the mRNA [21, 22]. Either an mRNA structure with low free energy or an 18S rRNA binding site near the 5′ end of the UTR is therefore inadequate as an indicator for gene expression as there are many factors involved in mRNA translated into protein. Thus, experimental evidence is required to describe the regulatory effect of the untranslated sequence on gene expression.
The proximal PPRV-F 5′ UTR increases translational efficiency
To precisely characterize the function of these sequences in gene expression, a monocistronic reporter assay involving the firefly luciferase (Fluc) gene was carried out to assess how gene expression is influenced by the upstream insertion sequence. Each of the proximal 85 nucleotides of the morbillivirus-F 5′ UTRs was introduced upstream of the Fluc gene (Fig. 2A); this allowed the in vivo synthesis of capped mRNA, which contains secondary structures near the 5′ cap, terminating at a poly(A) tail. In this study, we used the Fluc activity driven by an empty LUC vector that only differs by UTR inserts as the basal expression level. Embryonic kidney cell 293T (HEK-293T) cells were transfected with LUC vectors with or without insertion of the viral UTR fragments in combination with a control plasmid (pRluc-N1) expressing Renilla luciferase (Rluc). In this context, the regulatory activity of F 5′ UTRs could be monitored using Fluc activity as the readout, while the activity of the Rluc reporter gene served as a control for normalization of transfection variations. As indicated in Fig. 2A, proximal F 5′ UTRs of morbilliviruses showed different expression levels of luciferase reporter gene. Compared with the basal level, the CDV-F 5′ UTR showed an inhibitory effect on Fluc expression whereas the MV-F 5′ UTR had a marginal effect on Fluc expression. Interestingly, the highest luciferase activity was observed from the construct with the PPRV-F 5′ UTR (Fig. 2A). Northern blot analysis was further conducted to detect the accumulative RNA level using an antisense [α-32P] labeled DNA probe corresponding to the coding region of either Fluc gene or Rluc gene. Levels of mRNA were quantified by densitometry using imagej software. The mean band density ratio of Fluc mRNA to Rluc mRNA was expressed as the relative mRNA content (Fig. 2B). Interestingly, results indicated that the level of mRNA content from the construct with PPRV-F 5′ UTR is ~ 1.5-fold higher than those of the other constructs suggesting that the presence of proximal PPRV-F 5′ UTR relatively increases the steady-state of Fluc mRNA and possibly contributes to an increased translational utilization of Fluc mRNA. To better understand the effect of morbillivirus-F 5′ UTRs on translational efficiency and exclude the possible effect from transcriptional regulation, mRNAs bearing an m7GpppN cap structure at the 5′ end and a poly(A) tail (A21) at the 3′ end (Fig. 2C) were in vitro transcribed and directly transfected into HEK-293T cells. At 8 h following mRNA transfection, the results indicate that mRNA carrying PPRV-F 5′ UTR showed an ~ 8-fold higher luciferase activity than the control (LUC), while both MV-F 5′ UTR and CDV-F 5′ UTR had minimal effect on translational efficiency. Taken together, these observations demonstrate that the proximal PPRV-F 5′ UTR plays a role in the steady-state mRNA level and the protein translation process.
The proximal PPRV-F 5′ UTR enhances mRNA stability
To define the mechanism by which the proximal PPRV-F 5′ UTR increases the steady-state mRNA level, we further tested the stability of Fluc mRNA carrying or lacking the proximal PPRV-F 5′ UTR sequence. For comparative purposes, we included the proximal MV-F 5′ UTR and CDV-F 5′ UTR in this experiment. HEK-293T cells were treated with actinomycin D, an inhibitor of transcription, at 24 h post-transfection and total RNA was subsequently isolated at various time points (0, 3, 6, 9 and 12 h) after the treatment. Northern blotting followed by densitometric quantification were used to detect and quantify the luciferase mRNAs. The ratio of Fluc/Rluc mRNA level of each construct is shown relative to the mRNA level at 0 h, which was set at 100% (Fig. 2D). The result showed pronounced differences in stability of these four mRNA molecules that differ only in their 5′ terminal sequences. The level of Fluc mRNA containing PPRV-F 5′ UTR remains relatively stable up to 6 h after addition of actinomycin D, while the Fluc mRNA transcribed from LUC and MV constructs gradually decreased upon actinomycin D treatment with levels reaching 60% of the mRNA level at 6 h. Interestingly, the rapid degradation of Fluc mRNA was observed during the first 3 h when the transcript contains CDV-F 5′ UTR; this fast decreasing mRNA level possibly confers the negative effect observed in the reporter gene expression (Fig. 2A). Together, these observations indicated that the proximal PPRV-F 5′ UTR sequence has a positive effect on the steady-state mRNA level by increasing mRNA stability compared with those of the control transcripts. This positive effect therefore allows us to provide an explanation of the upregulation of the reporter gene translation by PPRV-F 5′ UTR, at least in part, from this mRNA stabilization.
Translational enhancer of proximal PPRV-F 5′ UTR functions in different cell lines and different reporter genes
Previous studies have revealed that some regulatory sequences act in concert with certain specific cellular proteins. The presence/absence of these particular translational factors is able to alter translation activity in different cellular backgrounds [24, 25]. To extend the findings to different cell lines, we further conducted luciferase reporter assays in two additional cell lines: a canine cell line (MDCK) and an African monkey cell line (BSC-I). Compared with the results for the HEK-293T cells (Fig. 2A), transfections of MDCK and BSC-I cells revealed a similar expression pattern in luciferase activity (Fig. 3A). Interestingly, the highest relative luciferase activity driven by the proximal PPRV-F 5′ UTR was observed in all three cell lines derived from different species.
Additionally, we also used a different reporter gene construct to demonstrate the enhancing activity of proximal PPRV-F 5′ UTR (Fig. 3B). The presence of enhanced green fluorescent protein (eGFP) allowed us to visualize translational activity. In this experiment, the amounts of DNA and incubation time after transfection were optimized to eliminate possible artifacts due to over-intensifying fluorescent signals from a high concentration of DNA transfection and prolonged expression time. HEK-293T cells were co-transfected with GFP reporter constructs and pRluc-N1 control plasmid. Results from fluorescence microscopy reveal that all cells were positive for GFP with similar transfection efficiency (Fig. 3B,D). Importantly, transfected cells with the PPRV-GFP construct apparently produced the most intense fluorescence in cells. This enhancing GFP intensity was further confirmed by flow cytometry (Fig. 3C). These data indicate a higher efficiency of gene expression in a cell-type and gene-type independent manner, and strongly support the idea that PPRV-F 5′ UTR is a good candidate for harboring the translational enhancer.
Deletion analysis of an enhancing element within proximal PPRV-F 5′ UTR
To further define the functional region within the PPRV-F 5′ UTR, we employed deletion analysis of the PPRV-F 5′ UTR spanning nucleotides 1–85. In this experiment, the proximal PPRV-F 5′ UTR was divided into three sections: nucleotides 1–27, 28–54 and 55–85. With this in mind, the middle section spanning nucleotides 28–54 contains a nine-nucleotide sequence complementary to 18S rRNA (Fig. 1B). A series of deletion derivatives were generated by deleting one or two of these sections within the UTR and introduced upstream of the Fluc coding region of the LUC vector (Fig. 4A). The ratio of Fluc/Rluc activity of each construct 24 h after co-transfection of HEK-293T cells is shown relative to the activity of the wild-type, which was set at 100% (Fig. 4B). When the deletion of individual sections of the PPRV-F 5′ UTR was tested, it was found that only the truncated PPRV Δ1–27 construct (harboring nucleotides 28–85) was able to maintain the relative luciferase activity as high as that of the wild-type, whereas the truncated PPRV Δ28–54 and PPRV Δ55–85 showed a reduced expression to 20–25% of that of the wild-type construct (Fig. 4B). These results indicate an enhancing element within the proximal PPRV-F 5′ UTR, spanning nucleotides 28–85.
To clarify the specific region required for enhancing activity, LUC vectors with deletions of two sections were examined. A dramatic reduction to 10% of wild-type expression was observed in the construct carrying the proximal section (PPRV 1–27), while the presence of either the middle section (PPRV 28–54) or the distal section (PPRV 55–85) was able to reinstate the expression to 27% and 40%, respectively. These observations suggest that the 18S rRNA complementary sequence in PPRV 28–54 was insufficient to enhance translation efficiency. However, this complementary sequence together with the extension of the distal sequence yielding nucleotides 28–85 (PPRV Δ1–27) was able to maximize the expression of the reporter gene (Fig. 4B).
Northern blot analysis combined with densitometric quantification was subsequently performed to evaluate the mRNA levels (Fig. 4C). With the exception of the truncated PPRV Δ28–54 construct, mRNAs transcribed from constructs carrying either the middle section or the distal section were able to maintain the steady-state mRNA level compared with that of the wild-type construct (Fig. 4C). Importantly, a profound effect on both the steady-state mRNA (Fig. 4C) and the translation efficiency (Fig. 4D) was only found in the sequence spanning nucleotides 28–85. These findings suggest that the enhancing element of PPRV-F 5′ UTR consists of two stem-loop RNA structures with a nine-nucleotide putative 18S rRNA complementary sequence at nucleotide positions 28–85; even so, further experimental evidence is required to support the role of the secondary structure and the mRNA–rRNA interaction.
The structural elements predicted for the 5′ UTR of the PPRV-F gene are mainly multiple stem-loops in the anterior part, and this sequence does not show any similarity to the conserved RNA regulatory elements. This does not imply, however, that less conserved structures will be functionally irrelevant. In the recent work, a translational enhancing activity driven by the proximal PPRV-F 5′ UTR was demonstrated in two different experiments, reporter gene DNA and in vitro transcribed mRNA transfection, and functioned in different cell lines and two reporter genes under the same vector backbone.
A translational enhancer containing highly RNA structures has been previously identified within the 5′ UTR sequence in several RNA viruses [1-3, 26]. Nevertheless, these RNA elements do not share a common secondary structure. As a nine-nucleotide segment that perfectly complements 18S rRNA was also found in the proximal PPRV-F 5′ UTR, we speculate that RNA structural feature and 18S rRNA complimentary are important factors for efficient gene translation; possibly by promoting the attachment of ribosomes as they parallel the function of the bacterial Shine–Dalgarno sequence and 16S rRNA as previously described elsewhere [8, 10]. Besides, we have shown that PPRV-F 5′ UTR can stabilize mRNA transcripts. As described in previous publications [27-29], a role of PPRV-F 5′ UTR in mRNA stability possible due to the relevant specific structure that is more resistant to endonuclease recognition or recruits RNA-binding proteins stabilizing mRNA.
Deletion mutagenesis further showed that a 57-nucleotide section contributes to average luciferase activity and mRNA level comparable to that of the 85-nucleotide proximal PPRV-F 5′ UTR sequence. Interestingly, this 57-nucleotide element contains a nine-nucleotide motif complementary to18S rRNA; however, the nine-nucleotide motif sequence alone did not enhance the translation activity. Therefore, the recent evidence does not support the 18S rRNA complementary theory of PPRV-F 5′ UTR mentioned above. Although the mechanism employed by PPRV-F 5′ UTR remains unknown, these findings allowed us to hypothesize that the observed enhancement of gene expression results primarily from a combined effect on mRNA stability and the possible role of adequate structural features present in PPRV-F 5′ UTR. A site-directed mutational analysis of predicted structure and assay for mRNA–rRNA interaction are necessary to assess this proposed mechanism.
The effect of long morbillivirus-F 5′ UTR has been previously demonstrated. F 5′ UTRs of MV and rinderpest virus have an inhibitory effect on F gene expression [30-33], but controversial results are also reported [15, 32, 34]. Moreover, CDV-F 5′ UTR was reported to contribute to the decrease in F gene expression due to the extensive folding of RNA secondary structures . However, our data indicate that the proximal F 5′ UTR of MV and CDV marginally affect gene translation, and the inhibitory effect of CDV-F 5′ UTR observed from reporter DNA transfection could be explained by a rapid degradation of mRNA. In this study, the translational enhancement of PPRV-F 5′ UTR is reported for the first time. Sequence comparison revealed that the proximal PPRV-F 5′ UTR is conserved among recent circulating PPRVs with 89.4%–100% identity (data not shown) and an analysis of proximal F 5′ UTR derived from different PPRV strains showed a similar effect on translation with no regard to strains (Fig. S1).
The feature of translational enhancer in PPRV-F 5′ UTR would be useful to consider as a translational enhancer for generating expression vectors used in molecular applications. In addition, we speculate that the different translational regulations among morbilliviruses further emphasize their various strategies to control gene expression. Notwithstanding, our data obtained from proximal PPRV-F 5′ UTR were not directly involved in viral replication. It would be interesting to evaluate the effect of the entire PPRV-F 5′ UTR on fusion gene expression as well as on viral replication.
Materials and methods
The human embryonic kidney cell line 293T (HEK-293T), the Madin–Darby canine kidney (MDCK) cell line and the African green monkey kidney (BSC-I) cell line were grown in Dulbecco's modified Eagle's medium (Gibco BRL, Life Technologies Corporation, Carlsbad, CA, USA). The medium contains 10% fetal bovine serum (Hyclone, Logan, UT, USA) and 1% penicillin–streptomycin (Gibco BRL). All cell lines were cultured at 37 °C in a humidified atmosphere containing 5% CO2.
Construction of reporter plasmids
The 5′ UTR DNA fragments under investigation were generated by PCR; furthermore, full-length canine distemper virus (CDV)-F 5′ UTR was amplified from CDV Ondersteeport strain (GenBank ID: AY386316) using appropriate primer sets as shown in Table 1. Additional DNA fragments of 85-nucleotide MV-F 5′ UTR, 85-nucleotide PPRV-F 5′ UTR and deletion derivatives of these were synthesized corresponding to the MV Edmonston strain (GenBank ID: AF266288) and the PPRV (GenBank ID: NC_006383) sequences (Table 1). For generation of the luciferase reporter gene construct LUC, BamHI/NotI firefly luciferase gene fragments from the p2luc plasmid  with or without an upstream sequence of 85-nucleotide PPRV-F 5′ UTR, 85-nucleotide MV-F 5′ UTR and CDV-F 5′ UTR were cloned into pcDNA3.1(+) (Invitrogen, Carlsbad, CA, USA). For the fluorescent reporter gene constructs, pEGFP-N1 plasmid (Clontech, Palo Alto, CA, USA) was modified to allow the cloning of either proximal 85-nucleotide PPRV-F 5′ UTR or full-length 85-nucleotide CDV-F 5′ UTR. The eGFP fragment carrying or lacking viral UTRs was linearized with BamHI/NotI and then ligated into pcDNA3.1(+).
Transient transfection, luciferase activity assays and fluorescent microscopy
Transient transfection was performed with Lipofectamine 2000® reagent (Invitrogen) according to the manufacturer's instructions. Briefly, cells were seeded the day before transfection at a density of 1 × 105 cells·well−1 on a 24-well plate. For analysis of translation efficiency and deletion mapping, 800 ng of luciferase reporter constructs or 100 ng of fluorescent reporter constructs were transfected in combination with 50 ng of pRluc-N plasmid (BioSignal Packard, Montreal, QC, Canada). Cells were incubated at 37 °C with 5% CO2. Firefly and Renilla luciferase levels were measured at 24 h post-transfection, using a Dual-Glo luciferase assay system (Promega, Madison, WI, USA) and a FLUOstar OPTIMA microplate reader (BMG Labtech, Offenburg, Germany). For the fluorescence study, live cells were examined at 18 h post-transfection under a fluorescence microscope with the appropriate filter and Renilla luciferase levels were measured as mentioned above.
For RNA transfection, 800 ng of capped RNAs, in vitro transcribed from LUC vector with or without insertion of viral UTRs, was mixed with lipofectamine and luciferase activity was measured at 8 h after transfection.
In vitro transcription
DNA templates for in vitro transcription were made by PCR using Phusion® High-Fidelity DNA Polymerase (Thermo Scientific, Vantaa, Finland). The LUC plasmid with or without the 85-nucleotide morbillivirus-F 5′ UTR was used as templates for PCR. The 5′ primer specified upstream sequence of T7 promoter and the 3′ primer contained a 21-nucleotide poly(A) 3′ overhang. The amplicons span the T7 promoter, the UTR region, the entire firefly luciferase coding region and a stretch of poly(A). The templates were gel purified after PCR. mRNA (G-capped, poly(A)-tailed) was transcribed by T7 RNA polymerase (Promega) in the presence of monomethylated cap analog m7G5′-pppp-G (Epicentre, Madison, WI, USA). The mRNA was checked on a formaldehyde gel and the concentration was calculated.
RNA extraction and northern blot analysis
Total RNAs were extracted from transfected HEK-293T cells using Trizol® reagent (Invitrogen) following the manufacturer's instruction. To remove DNA contamination, RNA samples were further treated with RNase-free DNase (Promega). Three micrograms of extracted total RNA was separated on a 1% denaturing formaldehyde gel and blotted onto a Hybond-N+ membrane (GE Healthcare Bio-Science Corp., Piscataway, NJ, USA). Following UV-crosslinking and pre-hybridization (for 2 h at 68 °C in pre-hybridization buffer 0.5 m sodium phosphate, 7% SDS and 1 mm EDTA), the membrane was hybridized with [α-32P]dATP-labelled gene-specific DNA probes at 68 °C overnight. Probes were generated by PCR with primers designed to specifically target the coding region of the Renilla and firefly luciferase genes (Table 1). After washing steps, the membrane was exposed to a phosphoimage screen (Fuji, Tokyo, Japan) and detected with a Bio-Imaging Analyzer (BAS-2500; Fuji). Densitometric quantifications were carried out using National Institutes of Health imagej software version 1.43 (http://rsb.info.nih.gov/ij).
Analysis of mRNA stability
Stability of mRNA transcripts was assessed after actinomycin D treatment of HEK-293T cells transiently transfected with LUC plasmids either lacking or carrying viral UTRs in combination with pRluc-N1 control plasmid. Actinomycin D (10 μg·mL−1 final concentration; Sigma-Aldrich) was added to the transfected cells at 24 h post-transfection. Total RNA was extracted from the transfected cells at 0, 3, 6, 9 and 12 h after treatment and subjected to northern blot analysis. The Fluc-mRNA was quantified and normalized to Rluc-mRNA level.
For fluorescence study, HEK-293T cells were collected at 18 h post-transfection and resuspended in 1 × NaCl/Pi. The expression of eGFP gene was analyzed with at least 10 000 events (cells) using the FL-1 filter by FACS Calibur flow cytometer (BD Biosciences, Burlington, MA, USA).
Prediction of RNA secondary structure
The predicted secondary structure for full-length CDV-F 5′ UTR, proximal 85-nucleotide PPRV-F 5′ UTR and proximal 85-nucleotide MV-F 5′ UTR was analyzed for minimal free energy and RNA secondary structure using the online mfold program prepared by M. Zuker [19, 20].
We thank Dr Fong-Yuan Lin for technical assistance, and Professor Ralph Kirby, at the National Yang-Ming University, Taiwan, for editorial assistance with the paper.