HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplementary Material Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Rennick, L. J. Articles by Rima, B. K. Search for Related Content PubMed PubMed Citation Articles by Rennick, L. J. Articles by Rima, B. K. Agricola Articles by Rennick, L. J. Articles by Rima, B. K.

Measles virus minigenomes encoding two autofluorescent proteins reveal cell-to-cell variation in reporter expression dependent on viral sequences between the transcription units

Centre for Cancer Research and Cell Biology, School of Biomedical Sciences, The Queen's University of Belfast, Belfast BT9 7BL, Northern Ireland, UK

Correspondence
Bert K. Rima
b.rima{at}qub.ac.uk

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Transcription from morbillivirus genomes commences at a single promoter in the 3' non-coding terminus, with the six genes being transcribed sequentially. The 3' and 5' untranslated regions (UTRs) of the genes (mRNA sense), together with the intergenic trinucleotide spacer, comprise the non-coding sequences (NCS) of the virus and contain the conserved gene end and gene start signals, respectively. Bicistronic minigenomes containing transcription units (TUs) encoding autofluorescent reporter proteins separated by measles virus (MV) NCS were used to give a direct estimation of gene expression in single, living cells by assessing the relative amounts of each fluorescent protein in each cell. Initially, five minigenomes containing each of the MV NCS were generated. Assays were developed to determine the amount of each fluorescent protein in cells at both cell population and single-cell levels. This revealed significant variations in gene expression between cells expressing the same NCS-containing minigenome. The minigenome containing the M/F NCS produced significantly lower amounts of fluorescent protein from the second TU (TU2), compared with the other minigenomes. A minigenome with a truncated F 5' UTR had increased expression from TU2. This UTR is 524 nt longer than the other MV 5' UTRs. Insertions into the 5' UTR of the enhanced green fluorescent protein gene in the minigenome containing the N/P NCS showed that specific sequences, rather than just the additional length of F 5' UTR, govern this decreased expression from TU2.

Supplementary material is available with the online version of this paper.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Measles virus (MV) belongs to the family Paramyxoviridae, and has a non-segmented negative-stranded (NNS) RNA genome of 15 894 nt containing six transcription units (TUs) encoding eight known proteins (Lamb & Kolakofsky, 2001). There are two envelope glycoproteins, the fusion (F) and haemagglutinin (H) proteins, which interact with the membrane-associated hydrophobic matrix (M) protein via their cytoplasmic tails. Inside the virion is the ribonucleoprotein (RNP) core consisting of the genomic RNA encapsidated by the nucleocapsid (N) protein. The RNP associates with the phospho- (P) and large (L) proteins, which form the RNA-dependent RNA polymerase (RdRp) of the virus. This RdRp functions as both a transcriptase and a replicase. Two further proteins, V and C, are generated from the TU encoding the P protein through editing and use of an overlapping reading frame, respectively (Bellini et al., 1985; Cattaneo et al., 1989).

The six TUs, 3'-N-P/V/C-M-F-H-L-5' (Rima et al., 1986), are separated by trinucleotide intergenic (Ig) sequences. Each TU begins with a 5' UTR (mRNA sense) containing the gene start (GS) signal for transcription initiation. After each ORF there is a 3' UTR with the gene end (GE) and polyadenylation signals for transcription termination (Kuo et al., 1996b; Schnell et al., 1996). Together the 3' UTR, Ig and 5' UTR of the following TU comprise the non-coding sequences (NCS) of the virus. The NCS range in length from 107 nt for the P/M NCS to 1012 nt for the M/F NCS. The 55 nt leader and the N 5' UTR comprise the 3' non-coding terminus (NCT) of the genome, and the L 3' UTR and 40 nt trailer the 5' NCT. Morbilliviruses obey the ‘rule of six’ (Calain & Roux, 1993; Egelman et al., 1989; Radecke et al., 1995; Sidhu et al., 1995). Presumed binding of six nt to one N protein establishes a so-called phase of the nucleotides with respect to the helical RNP core (Kolakofsky et al., 1998). Phase is probably important in transcription of the viral RNAs. Apart from the GS of the F gene, the start site phase of each of the viral genes is conserved throughout the Morbillivirus genus, although different genes have start sites in different phases (Rima et al., 2005).

The RdRp recognizes at least two discontinuous promoter elements during replication. The first is within the terminal 20 nt at the 3' ends of the genome and antigenome (Crowley et al., 1988; Sidhu et al., 1993). The second element, the B box (Crowley et al., 1988), is composed of three hexamers, between nt 79 and 96 (the 14th, 15th and 16th hexamer positions), containing a 5'-GN₅-3' motif (Tapparel et al., 1998). During replication, the RdRp enters the template at the first promoter element and ignores the GS, Ig and GE signals to produce a full-length antigenomic copy of the RNA template. When, during sequential transcription of each gene (Abraham & Banerjee, 1976; Ball & White, 1976), the transcriptase reaches the GE signal, there is a finite probability that it will detach from the (–)RNP. This gives rise to a transcription gradient with many more copies of mRNA generated for promoter-proximal genes (N-P/V/C) than for promoter-distal ones (H-L). The slope of this gradient depends on the virus strain and the cell type infected (Cattaneo et al., 1987; Schneider-Schaulies et al., 1989). Since the transcription gradient is markedly different in the brains of patients with the lethal disease subacute sclerosing panencephalitis (SSPE), it is important to understand aspects of the molecular basis of transcription attenuation. The advent of MV reverse genetics has allowed study of some of these aspects under controlled conditions (Radecke et al., 1995).

Studies of NNS virus transcription have typically analysed the effect of mutations within the NCS on the expression of reporter genes from bicistronic minigenomes (Edworthy & Easton, 2005; Kuo et al., 1996a, b, 1997; Stillman & Whitt, 1998). In these experiments, the 3' and 5' NCTs of a virus (which contain the replication and transcription promoter signals) have been linked to two reporter gene open reading frames (ORFs), for example luciferase (Luc) or chloramphenicol acetyl transferase (CAT), separated by mutated NCS so that the effect of mutations on the levels (and ratio) of expression of the reporter genes can be analysed. In all studies to date, analysis has been on a population-averaged basis over entire cell monolayers. Such analyses may mask differences in expression within individual cells. Many of these studies analyse transcription at the level of RNA synthesis only, which may mask differences in the translation of protein from the different mRNA transcripts.

In this study, we generated bicistronic minigenomes containing two autofluorescent reporter genes, encoding proteins with distinct excitation and emission spectra, to analyse their expression within individual, live cells. The ORFs were separated by each of the five MV NCS. Expression of these minigenomes has allowed us to describe and quantify hitherto hidden variation in gene expression between cells supporting expression of the same bicistronic minigenome, and to further investigate the previously reported differences in attenuation observed at each of the MV gene boundaries.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Cells and viruses.
Vero cells (European Tissue Culture and Cell Collection) were grown by standard procedures in DMEM medium (Gibco-BRL, Invitrogen Technologies) supplemented with 8 % (v/v) newborn calf serum (Gibco-BRL, Invitrogen Technologies). Fowl-pox virus expressing T7 RNA polymerase (FP-T7) has been described previously (Britton et al., 1996).

Plasmids and constructs.
Plasmids pEMC-Na, pEMC-Pa and pEMC-La, encoding the MV N, P and L proteins under the control of a T7 RNA polymerase promoter, have been described previously (Radecke et al., 1995). The NCS-containing constructs were generated from the plasmid p107MV : CAT (Radecke et al., 1995). Primers used are shown in Supplementary Table S1, available with the online version of this paper. The entire CAT ORF was replaced with an oligonucleotide linker containing restriction sites for MluI, NheI and BsrGI. These sites, along with the naturally occurring EcoRV site, allowed a bicistronic, core vector construct, p(–)MVDI_DsRed2(MN)_EGFP (Fig. 1b), to be produced by sequential insertion of the ORFs for DsRed2 (red fluorescent protein), in the TU1 position, and EGFP (enhanced green fluorescent protein), in the TU2 position (both Clontech, BD Biosciences). Insertion of the MV NCS between the autofluorescent reporter protein ORFs, using the MluI and NheI restriction sites, produced the NCS-containing constructs p(–)MVDI_DsRed2-NCS-_EGFP (Fig. 1c). All MV sequences were amplified from a plasmid containing a full-length cDNA copy of the Edtag virus (Radecke et al., 1995). Constructs with insertions into the P 5' UTR of p(–)MVDI_DsRed2-N/P-_EGFP were generated using p(–)MVDI_DsRed2-N/(M)P-_EGFP containing an engineered MfeI site in the P 5' UTR. Data obtained for the parental and MfeI-containing constructs were not significantly different.

View larger version (28K): [in this window] [in a new window]   Fig. 1. Schematic representation of plasmids containing MV-based minigenome templates. The MV non-coding termini (3' and 5' NCT) surround the ORFs for reporter genes. The minigenome sequences are surrounded by a T7 RNA polymerase promoter (T7), hepatitis delta ribozyme () and T7 RNA polymerase terminator (T) sequences. The plasmid p107MV : CAT (a) was used as template for generation of the bicistronic, core vector construct p(–)MVDIDsRed2(MN)EGFP (b). The restriction sites MluI and NheI were used to insert any of the full-length MV NCS to produce the NCS containing constructs p(–)MVDIDsRed2-NCS-EGFP (c). The NCS sequence comprises the 3' and 5' UTRs separated by an intergenic trinucleotide (Ig), and contains the conserved transcription termination (GE) and initiation (GS) signals.

Transfection and minigenome expression assay.
Vero cells were grown to 80 % confluency in six-well trays, rinsed with Optimem (1 ml; Gibco-BRL, Invitrogen Technologies) and infected with FP-T7 at an m.o.i. of 1 for 45 min. Lipofectamine 2000 (Gibco-BRL, Invitrogen Technologies) was diluted with Optimem according to the manufacturer's instructions and incubated at room temperature for 5 min. A DNA mixture containing the plasmids pEMC-Na (1.2 µg), pEMC-Pa (1.2 µg), pEMC-La (0.4 µg) and the plasmid containing the minigenome sequences (4 µg) was added and liposome–DNA complexes were formed by incubation for 30 min at room temperature. The FP-T7 inoculum was removed and the complexes spotted onto the Vero cell monolayers. Optimem (1 ml) was added to each well. After 24 h incubation at 37 °C, 5 % (v/v) CO₂, the complexes were replaced with Vero growth medium (1.3 ml). Monolayers were observed daily up to 7 days post-transfection (d.p.t.).

Microscopy, image capture and quantification.
Cells were observed using a DM IRBE UV microscope fitted with appropriate filter blocks (Leica Microsystems). For quantification, data were collected by confocal scanning laser microscopy (CSLM) using a Leica scan head and a DM IRBE UV microscope (Leica Microsystems) as previously described (Duprex et al., 1999). Data were collected for living cells at a magnification of x100, and quantification carried out using the Leica TCS NT software. Identical parameters, which did not lead to pixel saturation, were used for imaging and ensured that data collected for all cells and all constructs could be compared. The total intensity of emission in the red and green channels was determined by summing the intensity value for each pixel within an area corresponding to a fluorescent cell. All such values for ‘total green’ and ‘total red’ pixel intensity were corrected for background intensities.

Semi-quantitative assessment of fluorescence.
Cells were observed using UV microscopy to visualize red and green fluorescence simultaneously. Fluorescent cells were assigned to one of three categories, red, yellow or green, depending on the colour of the cell (Fig. 2). All assessments were carried out on the same day in the shortest possible time and at least 300 cells were assessed for each construct. Under these conditions, the assignment of cells was consistent and the data could be used to make a semi-quantitative assessment of the relative amounts of red and green fluorescence in individual cells. Statistical analyses were carried out using GraphPad Prism version 4.00 (GraphPad Software).

View larger version (69K): [in this window] [in a new window]   Fig. 2. Fluorescence observed in Vero cells transfected with various minigenome constructs. Cells were observed at six d.p.t. Images were collected using a digital camera and are shown at x25 magnification for p(–)MVDI-N/P (a–c) and p(–)MVDI-H/L (d–f) using a filter to detect red and green fluorescence simultaneously, and x100 magnification for p(–)MVDI-F/H (g–i) using a filter to detect red and green fluorescence simultaneously (g), green fluorescence only (h) or red fluorescence only (i). The insets (a and d) and numbering (g–i) correspond to the scatter plots from CSLM analysis (Fig. 3).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Gene expression from each minigenome was assessed in Vero cells. Transcription should result in more transcripts from TU1 than TU2, depending on transcription attenuation at the particular NCS present. This, like other minigenome expression assays, is a coupled transcription/translation assay. Significant heterogeneity in reporter gene expression was observed in individual cells expressing each of the minigenomes. The variation in the relative amounts of red and green fluorescence expressed within individual cells also led to variation in the colour of the cells when viewed by UV microscopy (Fig. 2). Cells which expressed high amounts of EGFP without similarly high amounts of DsRed appeared green, cells which expressed high amounts of DsRed without similarly high amounts of EGFP appeared red, and cells which expressed appreciable amounts of both appeared yellow. Data were collected by CSLM to obtain a quantitative measure of the actual amounts of red and green fluorescence expressed within individual Vero cells. The values corresponding to amounts of red and green fluorescence were plotted in scatter diagrams for cells analysed for each of the minigenomes (Fig. 3a–e). Each point corresponds to the actual amounts of DsRed2 and EGFP expressed within an individual cell. Cells expressing the minigenome containing the N/P NCS (Fig. 2a–c) appeared green and yellow, as predicted from the corresponding scatter plot (Fig. 3a), while that for cells expressing the H/L NCS (Fig. 3e) predicts a predominance of yellow and red cells, as was observed (Fig. 2d–f).

View larger version (22K): [in this window] [in a new window]   Fig. 3. Quantitative reporter gene expression in Vero cells transfected with (a) p(–)MVDI-N/P, (b) p(–)MVDI-P/M, (c) p(–)MVDI-M/F, (d) p(–)MVDI-F/H or (e) p(–)MVDI-H/L. (a–e) Data was collected at six d.p.t. by CSLM. Values for the total red and total green pixel intensities in eight optical sections for each cell were determined. Total corrected red (x-axis) and green (y-axis) pixel intensity values were used as coordinate pairs for data points on scatter graphs. All graphs are plotted at the same linear scale. Lines are arbitrarily drawn to give an indication of the observed colours of cells corresponding to the data points. Cells falling between the curved lines are yellow, above the lines green and below the lines red. The points marked 1, 2 and 3 correspond to cells 1, 2 and 3 in Fig. 2(g–i). (f) Histogram representing the average percentages of red (black bars), yellow (white bars) and green (grey bars) cells for each of the five NCS constructs at six d.p.t. (triplicate transfection experiments).

Only small numbers of cells transfected with any of the NCS-containing constructs produced detectable fluorescence. Minigenome expression efficiencies, the percentage of cells available for minigenome expression which actually support expression, were calculated for each of the NCS-containing constructs. For different constructs the efficiencies varied from 0.094 to 0.19 % of cells available for FP-T7 infection and plasmid transfection. These figures are low compared to transfection efficiencies of control plasmids expressing DsRed2 and EGFP under control of a CMV immediate-early promoter, where efficiencies of greater than 60 % were routinely observed (Supplementary Fig. S3, available with the online version of this paper). The minigenome expression efficiencies for the NCS-containing constructs were consistent in replicate experiments. Accordingly, these expression efficiencies are an intrinsic property of the MV minigenome expression system, rather than related to transfection efficiency.

As well as revealing significant heterogeneity in reporter gene expression in individual cells containing each minigenome, CSLM analysis also revealed significant variation in the data between the different minigenomes (Fig. 3a–e). The average ratios of green fluorescence to red fluorescence for all cells expressing the N/P-, P/M- and F/H-containing minigenomes were similar at 1.5 : 1, 1.2 : 1 and 1.3 : 1, respectively. The value for the H/L-containing minigenome was lower at 0.8 : 1, but this was predicted from the previously reported higher transcription attenuation at this gene boundary (Cattaneo et al., 1987). Surprisingly, the value for the M/F-containing minigenome was 0.2 : 1. This was due to production of little green fluorescence in cells rather than increased amounts of red fluorescence (Fig. 3c). Since only limited numbers of cells could be analysed by CSLM this analysis was extended to develop a robust, reproducible and rapid assay to describe, quantify and allow statistical analysis of the observed variation at the minigenome level.

Semi-quantitative assessment of fluorescence produced in cells reveals significantly lower expression of TU2 from the minigenome containing the M/F NCS
All fluorescent cells within an experiment were screened and assigned as red, yellow or green, based on their colour when viewed simultaneously for red and green fluorescence by UV microscopy. As described above, assignment in this way gives an accurate indication of the actual relative amounts of red and green fluorescence produced in the individual cells (Figs 2 and 3a–e). The data were averaged over three experiments and used to generate histograms of the percentages of red, yellow and green cells generated upon expression of each of the NCS-containing minigenomes (Fig. 3f). The distributions were subjected to statistical analysis (data not shown). These results reflect those of the CSLM-based quantitative analysis. The M/F NCS-containing minigenome produced the highest percentage of red cells, and lowest percentages of yellow and green cells (Fig. 3f). The percentage of red cells observed was highly significantly different to the percentages observed upon expression of any of the other minigenomes. The N/P NCS-containing minigenome produced the highest percentage of green cells (Fig. 3f), as predicted from the CSLM quantitative analysis (Fig. 3a). This percentage was highly significantly different from the percentages obtained upon expression of the minigenomes containing the M/F, F/H and H/L NCS. The percentages of red, yellow and green cells observed upon expression of the minigenomes containing the P/M, F/H and H/L NCS were not significantly different, which reflected the spread of the data in the CSLM analysis (Fig. 3b, d and e).

Truncation of the F 5' UTR is associated with increased gene expression
Regardless of the method of analysis, expression of the minigenome containing the M/F NCS produced significantly different data to all other minigenomes. This NCS is much longer than the others. Hence, we examined the effect of truncating the M 3' UTR (426 nt) and F 5' UTR (583 nt) on minigenome expression. The plasmid p(–)MVDI_DsRed2-M_GE/F-_EGFP [p(–)MVDI-M_GE/F] contains a truncated M 3' UTR in which the 409 nt were deleted, generating a minigenome which contained the GE signal, an authentic Ig and the complete F 5' UTR. The plasmid p(–)MVDI_DsRed2-M/F_GS-_EGFP [p(–)MVDI-M/F_GS] contains a truncated F 5' UTR (556 nt deleted), generating a minigenome which contained the authentic M 3' UTR, Ig and conserved GS signal whilst maintaining the phase for the F gene.

These plasmids, along with p(–)MVDI-M/F, were transfected into Vero cells in minigenome expression assays and assessed by CSLM quantitative and semi-quantitative analyses. The percentages of red, yellow and green cells detected upon expression of the minigenomes containing the M_GE/F and full-length M/F NCS were not significantly different (Fig. 4a). However, upon expression of the minigenome containing the M/F_GS NCS, the percentages of red and yellow cells were highly significantly different from those produced upon expression of the minigenomes containing the full-length M/F or truncated M_GE/F NCS. The percentage of red cells was lower and the percentage of yellow cells higher. The quantitative CSLM analysis (Figs 4b, c and 3c) illustrates that this shift from red to yellow cells is associated with an increase in levels of EGFP (from TU2).

View larger version (27K): [in this window] [in a new window]   Fig. 4. Vero cells were transfected with p(–)MVDI-M/F, p(–)MVDI-MGE/F, p(–)MVDI-M/FGS, p(–)MVDI-N/P, p(–)MVDI-N/(M)P, p(–)MVDI-N/FGSP, p(–)MVDI-N/328P or p(–)MVDI-N/586P in minigenome expression assays. (a and d) Histograms representing the average percentages of red, yellow and green cells for each of the NCS constructs at six d.p.t. as for Fig. 3(f). For p(–)MVDI-MGE/F (b) and p(–)MVDI-M/FGS (c), data was collected at six d.p.t. by CSLM and plotted as for Fig. 3(a–e).

These plasmids, along with p(–)MVDI-N/P, were transfected into Vero cells. Cells expressing the minigenomes were analysed semi-quantitatively (Fig. 4d). The percentages of green cells produced from the N/F_GSP- and N/586P-containing minigenomes were both highly significantly different to the percentage produced from the N/P-containing minigenome. However, the percentage of red cells produced by the N/F_GSP-containing minigenome was highly significantly higher than that produced by the N/P-containing minigenome, while the percentage produced by the N/586P-containing minigenome was not significantly different. Also, the percentage of yellow cells produced from the N/F_GSP-containing minigenome was significantly lower than the percentage produced from that containing the N/586P NCS. In CSLM analysis, this cell colour shift correlated with reduced expression of TU2 in cells expressing the N/F_GSP-containing minigenome (data not shown). For those cells the average ratio of green fluorescence to red fluorescence was 1.0 : 1, compared to the value of 1.5 : 1 observed for the original N/P-containing construct. Taken together, these results indicate that the presence of the F 5' UTR-specific sequence significantly shifts the distribution of cell colour from that produced by the N/P-containing minigenome towards that produced by the M/F NCS-containing minigenome, with more red cells and fewer green cells produced upon expression of the N/F_GSP-containing minigenome. This indicates that specific sequences in the F 5' UTR sequence contribute to the significantly different results obtained upon expression of the M/F NCS-containing minigenome.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
The purpose of this study was to investigate variation in viral gene expression at the single cell level. To this end, bicistronic minigenomes containing autofluorescent reporter genes were generated. Since there is a linear relationship between the amount of protein and the resulting amount of fluorescence (Furtado & Henry, 2002; Verkhusha et al., 2003), the amount of fluorescence within individual cells gives a measure of the expression of each TU. To the best of our knowledge, this is the first study within the Mononegavirales to examine protein expression from bicistronic minigenomes in this way. Previous studies with rabies virus, vesicular stomatitis virus (VSV) and respiratory syncytial virus (RSV) used minigenomes to investigate the effect of mutations on the functionality of NCS, primarily the conserved GE, Ig and GS signals (Barr et al., 1997a, b; Finke et al., 2000; Kuo et al., 1996a, b, 1997). However, all these have been evaluated at the cell population level using Northern blot analysis and total cell lysates to determine reporter gene expression. Such analyses may mask significant variation at the single-cell level.

Consistent, reproducible and statistically significant differences dependent on the NCS present between TU1 and TU2 were observed in the minigenome expression assay (Figs 3 and 4). Expression of those containing the P/M, F/H and H/L NCS produced cell colour distributions which were similar, generating many yellow cells that express TU1 and TU2 to high levels. Expression of the minigenome containing the N/P NCS consistently produced cells which expressed high relative levels of TU2. Conversely, expression of the minigenome containing the M/F NCS consistently produced the lowest levels of expression of TU2 within individual cells. The latter was due to F 5' UTR specific sequence. When this sequence was deleted the distribution of red, yellow and green cells shifted to that of the other minigenomes (Fig. 4). Deletion of the M 3' UTR specific sequence had no significant effect on cell colour distribution. Whether the length or the specific sequence of the F 5' UTR was an important determinant of the expression of TU2 was investigated by making a series of insertions into the P 5' UTR of p(–)MVDI-N/P, the construct which gave the highest expression of TU2. Inserting F 5' UTR specific sequence affected the colour distribution produced and shifted it towards that observed upon expression of the M/F NCS-containing minigenome. Increasing the length of the P 5' UTR to the same length as in p(–)MVDI-N/F_GSP with unrelated non-NCS sequence also affected the colour distribution, but this did not reach statistical significance compared with the M/F NCS-containing minigenome.

Removal of the F 5' UTR from mRNA transcripts of MV, canine distemper virus and rinderpest virus has been shown to increase the translatability of F in vitro (Evans et al., 1990; Hasel et al., 1987). This suggests that the F 5' UTR sequence may have a negative effect on translation of the mRNA. The F 5' UTR sequence of MV has a high degree of predicted secondary structure (Richardson et al., 1986) and high sequence conservation between strains (Rota et al., 1994). An in vivo study of the phenotype of a recombinant MV, del5F, containing a 504 nt deletion in the F 5' UTR, (Auwaerter et al., 1996), found that this deletion resulted in decreased peak virus production and a small change in the kinetics of growth (Valsamakis et al., 1998). The conclusion was that the F 5' UTR was not absolutely required for MV replication in that system, but that deletion of the sequence led to a change in the level, but not abrogation of F protein expression, possibly by affecting translation. A more recent study reported that a wild-type virus with a 540 nt deletion in the F 5' UTR had enhanced replication capacity, and caused much stronger cytopathic effect in the host cells (Takeda et al., 2005). This correlated with increased amounts of F protein produced by this virus compared to the parental virus containing the authentic F 5' UTR. This reflects the observations in this study where the presence of the F 5' UTR led to decreased gene expression, and which indicated that it is the specific sequence of that region rather than the length which is responsible. Previous studies on steady-state mRNA levels in lytically infected cells (Cattaneo et al., 1987; Schneider-Schaulies et al., 1989) report that there is no extreme attenuation in the levels of F mRNA relative to the levels of M and H. However, whether the F 5' UTR has its effect on transcription, translation or both remains to be determined.

Current models of gene expression and transcription in the Mononegavirales suggest a single transcription promoter, either at the 3' end of the genome or directly at the first GS signal, and polymerase dissociation at gene boundaries (Barr et al., 2002; Cordey & Roux, 2006; Vulliemoz & Roux, 2002; Whelan et al., 2004; Whelan & Wertz, 2002). This model predicts fewer mRNA transcripts for EGFP than DsRed2 upon expression of the minigenomes. In particular, the current model of transcription does not predict the detection of cells expressing TU2 without concomitant expression of TU1 (green cells). It is possible that these cells represent an artefact, perhaps dependent on a less than ideal stoichiometry of N, P and L proteins in the cell in the plasmid-driven expression system. However, in an experiment which used tenfold less plasmids, the same colour distribution, including green and red cells, was observed. Experiments using different helper plasmids supplying the N, P and L proteins under CMV promoter control gave the same distribution. Also in experiments using ECFP and EYFP as the two reporter genes, cells were detected which expressed EYFP (from TU2) without concomitant ECFP expression (Supplementary Fig. S3b, available with the online version of this paper). To our knowledge, no earlier studies have been reported in which these single-cell effects could have been observed. In a previous study, Kuo et al. (1996b) used RSV bicistronic minigenomes containing CAT and Luc reporter genes to examine the effects of mutations in the viral GS and GE signals. Interestingly, when the GS signal for TU1 was ablated, expression of both TU1 and TU2 was still detected at levels approximately 10 % that of the wild type containing the authentic GS signal for TU1. This was due to production of an mRNA for TU1 which had the leader sequence attached. Transcription of this mRNA then allowed the polymerase to access the GE and GS signals between TU1 and TU2 and transcribe TU2 as normal. A similar experiment was carried out for the N/P- and P/M-containing minigenomes in this study (Supplementary Fig. S4, available with the online version of this paper). When the GS signal for TU1 was ablated in either construct, many faint green cells (expressing TU2) were detected. Red fluorescence (expression of TU1) could not be detected in these cells. This indicates that this system detects expression of TU2 without concomitant detection of expression of TU1. It is known that MV produces a proportion (approx. 2 %) of polyadenylated N mRNA transcripts that have leader sequence attached in vivo (Castaneda & Wong, 1989, 1990). This leader-N mRNA transcript has been shown to be encapsidated with N protein, not associated with polysomes and unavailable for translation (Castaneda & Wong, 1990). This indicates that the leader-TU1 mRNAs produced from the GS knock-out minigenomes would not be translated and hence no red fluorescence could be produced, whereas the transcripts produced for TU2 (by recognition of the GE and GS signals) could be translated and give rise to green cells. This study, looking at individual cells, can detect expression of a downstream gene without expression of the upstream gene. In the minigenomes investigated in this study (containing the authentic GS signal), differences in the stoichiometry of helper plasmid expression in individual cells may have affected the ratio of TU1 to leader-TU1 transcripts and contributed to the cell colour distributions.

The results presented in this study stand at present alone as significant, reproducible observations of MV minigenome expression in single cells. Even though the conditions for successful replication and transcription of the minigenomes are only established in a very small percentage of available cells, the level in our system is similar to that of other negative-strand viruses such as Ebola virus (Watanabe et al., 2004) and Bunyamwera virus (Bridgen & Elliott, 1996). The suggestion that cells that do not fit the current transcription model result from T7 RNA polymerase-mediated transcription errors is also unlikely, as many minigenomes and helper mRNAs must be generated in the cell and the conditions for minigenome rescue are similar to those in which virus rescue is achieved after faithful transcription of the entire antigenome by T7 polymerase. The variation between cells could also represent differences in the levels of host factors associated with the transcriptase activity, or simply be an intrinsic property of the minigenome system. Whatever the source of the variation, if it is an artefact then the minigenome assays in other studies may also suffer from them and the population averaging that takes place when a total cell monolayer is extracted and assayed will hide this diversity. In this context, it is interesting to note that, if we sum the total red and total green fluorescence intensities for all the cells in the quantitative assessment and hence determine population-averaged fluorescence intensities for each minigenome, the overall ratios of red to green fluorescence are very much in concordance with the earlier reports on transcription attenuation. In the first (Cattaneo et al., 1987), there is a substantial percentage drop-off (64 %) at the N/P boundary and for the remainder of the boundaries attenuation is 20–26 %. In the second (Schneider-Schaulies et al., 1989), these figures are between 21 and 29 % for each gene boundary.

In conclusion, these results indicate that the colour distributions for cells expressing bicistronic minigenomes with two autofluorescent reporter genes are specific to the NCS present, and that differences between the contributions of the various NCS are consistent and reproducible. Hence, we propose that these data reveal a hitherto obscured variation in gene expression between cells in minigenome assays. We are currently extending investigations of the effect of the F 5' UTR on gene expression and are trying to identify the specific sequences involved.

	ACKNOWLEDGEMENTS

 
This work was supported by the Medical Research Council (Grant 9901004), the Department of Education and Learning, Northern Ireland and the Institute for Animal Health, Pirbright Laboratories. We thank Paula Haddock for excellent technical assistance.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Abraham, G. & Banerjee, A. K. (1976). Sequential transcription of the genes of vesicular stomatitis virus. Proc Natl Acad Sci U S A 73, 1504–1508.[Abstract/Free Full Text]

Auwaerter, P. G., Kaneshima, H., McCune, J. M., Wiegand, G. & Griffin, D. E. (1996). Measles virus infection of thymic epithelium in the SCID-hu mouse leads to thymocyte apoptosis. J Virol 70, 3734–3740.[Abstract]

Ball, L. A. & White, C. N. (1976). Order of transcription of genes of vesicular stomatitis virus. Proc Natl Acad Sci U S A 73, 442–446.[Abstract/Free Full Text]

Barr, J. N., Whelan, S. P. & Wertz, G. W. (1997a). Cis-acting signals involved in termination of vesicular stomatitis virus mRNA synthesis include the conserved AUAC and the U7 signal for polyadenylation. J Virol 71, 8718–8725.[Abstract]

Barr, J. N., Whelan, S. P. & Wertz, G. W. (1997b). Role of the intergenic dinucleotide in vesicular stomatitis virus RNA transcription. J Virol 71, 1794–1801.[Abstract]

Barr, J. N., Whelan, S. P. & Wertz, G. W. (2002). Transcriptional control of the RNA-dependent RNA polymerase of vesicular stomatitis virus. Biochim Biophys Acta 1577, 337–353.[Medline]

Bellini, W. J., Englund, G., Rozenblatt, S., Arnheiter, H. & Richardson, C. D. (1985). Measles virus P gene codes for two proteins. J Virol 53, 908–919.[Abstract/Free Full Text]

Bridgen, A. & Elliott, R. M. (1996). Rescue of a segmented negative-strand RNA virus entirely from cloned complementary DNAs. Proc Natl Acad Sci U S A 93, 15400–15404.[Abstract/Free Full Text]

Britton, P., Green, P., Kottier, S., Mawditt, K. L., Penzes, Z., Cavanagh, D. & Skinner, M. A. (1996). Expression of bacteriophage T7 RNA polymerase in avian and mammalian cells by a recombinant fowlpox virus. J Gen Virol 77, 963–967.[Abstract/Free Full Text]

Calain, P. & Roux, L. (1993). The rule of six, a basic feature for efficient replication of Sendai virus defective interfering RNA. J Virol 67, 4822–4830.[Abstract/Free Full Text]

Castaneda, S. J. & Wong, T. C. (1989). Measles virus synthesizes both leaderless and leader-containing polyadenylated RNAs in vivo. J Virol 63, 2977–2986.[Abstract/Free Full Text]

Castaneda, S. J. & Wong, T. C. (1990). Leader sequence distinguishes between translatable and encapsidated measles virus RNAs. J Virol 64, 222–230.[Abstract/Free Full Text]

Cattaneo, R., Rebmann, G., Schmid, A., Baczko, K., ter Meulen, V. & Billeter, M. A. (1987). Altered transcription of a defective measles virus genome derived from a diseased human brain. EMBO J 6, 681–688.[Medline]

Cattaneo, R., Kaelin, K., Baczko, K. & Billeter, M. A. (1989). Measles virus editing provides an additional cysteine-rich protein. Cell 56, 759–764.[CrossRef][Medline]

Cordey, S. & Roux, L. (2006). Transcribing paramyxovirus RNA polymerase engages the template at its 3' extremity. J Gen Virol 87, 665–672.[Abstract/Free Full Text]

Crowley, J. C., Dowling, P. C., Menonna, J., Silverman, J. I., Schuback, D., Cook, S. D. & Blumberg, B. M. (1988). Sequence variability and function of measles virus 3' and 5' ends and intercistronic regions. Virology 164, 498–506.[CrossRef][Medline]

Duprex, W. P., McQuaid, S., Hangartner, L., Billeter, M. A. & Rima, B. K. (1999). Observation of measles virus cell-to-cell spread in astrocytoma cells by using a green fluorescent protein-expressing recombinant virus. J Virol 73, 9568–9575.[Abstract/Free Full Text]

Edworthy, N. L. & Easton, A. J. (2005). Mutational analysis of the avian pneumovirus conserved transcriptional gene start sequence identifying critical residues. J Gen Virol 86, 3343–3347.[Abstract/Free Full Text]

Egelman, E. H., Wu, S. S., Amrein, M., Portner, A. & Murti, G. (1989). The Sendai virus nucleocapsid exists in at least four different helical states. J Virol 63, 2233–2243.[Abstract/Free Full Text]

Evans, S. A., Belsham, G. J. & Barrett, T. (1990). The role of the 5' nontranslated regions of the fusion protein mRNAs of canine distemper virus and rinderpest virus. Virology 177, 317–323.[CrossRef][Medline]

Finke, S., Cox, J. H. & Conzelmann, K. K. (2000). Differential transcription attenuation of rabies virus genes by intergenic regions: generation of recombinant viruses overexpressing the polymerase gene. J Virol 74, 7261–7269.[Abstract/Free Full Text]

Furtado, A. & Henry, R. (2002). Measurement of green fluorescent protein concentration in single cells by image analysis. Anal Biochem 310, 84–92.[CrossRef][Medline]

Hasel, K. W., Day, S., Millward, S., Richardson, C. D., Bellini, W. J. & Greer, P. A. (1987). Characterization of cloned measles virus mRNAs by in vitro transcription, translation, and immunoprecipitation. Intervirology 28, 26–39.[Medline]

Kolakofsky, D., Pelet, T., Garcin, D., Hausmann, S., Curran, J. & Roux, L. (1998). Paramyxovirus RNA synthesis and the requirement for hexamer genome length: the rule of six revisited. J Virol 72, 891–899.[Free Full Text]

Kuo, L., Fearns, R. & Collins, P. L. (1996a). The structurally diverse intergenic regions of respiratory syncytial virus do not modulate sequential transcription by a dicistronic minigenome. J Virol 70, 6143–6150.[Abstract]

Kuo, L., Grosfeld, H., Cristina, J., Hill, M. G. & Collins, P. L. (1996b). Effects of mutations in the gene-start and gene-end sequence motifs on transcription of monocistronic and dicistronic minigenomes of respiratory syncytial virus. J Virol 70, 6892–6901.[Abstract/Free Full Text]

Kuo, L., Fearns, R. & Collins, P. L. (1997). Analysis of the gene start and gene end signals of human respiratory syncytial virus: quasi-templated initiation at position 1 of the encoded mRNA. J Virol 71, 4944–4953.[Abstract]

Lamb, R. L. & Kolakofsky, D. (2001). Paramyxoviridae: the viruses and their replication. In Fields Virology, 4th edn, pp. 1305–1340. Edited by D. M. Knipe & P. M. Howley. Philadelphia, PA: Lippincott Williams & Wilkins.

Radecke, F., Spielhofer, P., Schneider, H., Kaelin, K., Huber, M., Dotsch, C., Christiansen, G. & Billeter, M. A. (1995). Rescue of measles viruses from cloned DNA. EMBO J 14, 5773–5784.[Medline]

Richardson, C., Hull, D., Greer, P., Hasel, K., Berkovich, A., Englund, G., Bellini, W., Rima, B. & Lazzarini, R. (1986). The nucleotide sequence of the mRNA encoding the fusion protein of measles virus (Edmonston strain): a comparison of fusion proteins from several different paramyxoviruses. Virology 155, 508–523.[CrossRef][Medline]

Rima, B. K., Baczko, K., Clarke, D. K., Curran, M. D., Martin, S. J., Billeter, M. A. & ter Meulen, V. (1986). Characterization of clones for the sixth (L) gene and a transcriptional map for morbilliviruses. J Gen Virol 67, 1971–1978.[Abstract/Free Full Text]

Rima, B. K., Collin, A. M. & Earle, J. A. (2005). Completion of the sequence of a cetacean morbillivirus and comparative analysis of the complete genome sequences of four morbilliviruses. Virus Genes 30, 113–119.[CrossRef][Medline]

Rota, J. S., Wang, Z. D., Rota, P. A. & Bellini, W. J. (1994). Comparison of sequences of the H, F, and N coding genes of measles virus vaccine strains. Virus Res 31, 317–330.[CrossRef][Medline]

Schneider-Schaulies, S., Liebert, U. G., Baczko, K., Cattaneo, R., Billeter, M. & ter Meulen, V. (1989). Restriction of measles virus gene expression in acute and subacute encephalitis of Lewis rats. Virology 171, 525–534.[CrossRef][Medline]

Schnell, M. J., Buonocore, L., Whitt, M. A. & Rose, J. K. (1996). The minimal conserved transcription stop-start signal promotes stable expression of a foreign gene in vesicular stomatitis virus. J Virol 70, 2318–2323.[Abstract]

Sidhu, M. S., Husar, W., Cook, S. D., Dowling, P. C. & Udem, S. A. (1993). Canine distemper terminal and intergenic non-protein coding nucleotide sequences: completion of the entire CDV genome sequence. Virology 193, 66–72.[CrossRef][Medline]

Sidhu, M. S., Chan, J., Kaelin, K., Spielhofer, P., Radecke, F., Schneider, H., Masurekar, M., Dowling, P. C., Billeter, M. A. & Udem, S. A. (1995). Rescue of synthetic measles virus minireplicons: measles genomic termini direct efficient expression and propagation of a reporter gene. Virology 208, 800–807.[CrossRef][Medline]

Stillman, E. A. & Whitt, M. A. (1998). The length and sequence composition of vesicular stomatitis virus intergenic regions affect mRNA levels and the site of transcript initiation. J Virol 72, 5565–5572.[Abstract/Free Full Text]

Takeda, M., Ohno, S., Seki, F., Nakatsu, Y., Tahara, M. & Yanagi, Y. (2005). Long untranslated regions of the measles virus M and F genes control virus replication and cytopathogenicity. J Virol 79, 14346–14354.[Abstract/Free Full Text]

Tapparel, C., Maurice, D. & Roux, L. (1998). The activity of Sendai virus genomic and antigenomic promoters requires a second element past the leader template regions: a motif (GNNNNN)₃ is essential for replication. J Virol 72, 3117–3128.[Abstract/Free Full Text]

Valsamakis, A., Schneider, H., Auwaerter, P. G., Kaneshima, H., Billeter, M. & Griffin, D. E. (1998). Recombinant measles viruses with mutations in the C, V or F gene have altered growth phenotypes in vivo. J Virol 72, 7754–7761.[Abstract/Free Full Text]

Verkhusha, V. V., Kuznetsova, I. M., Stepanenko, O. V., Zaraisky, A. G., Shavlovsky, M. M., Turoverov, K. K. & Uversky, V. N. (2003). High stability of Discosoma DsRed as compared to Aequorea EGFP. Biochemistry 42, 7879–7884.[CrossRef][Medline]

Vulliemoz, D. & Roux, L. (2002). Given the opportunity, the Sendai virus RNA-dependent RNA polymerase could as well enter its template internally. J Virol 76, 7987–7995.[Abstract/Free Full Text]

Watanabe, S., Watanabe, T., Noda, T., Takada, A., Feldmann, H., Jasenosky, L. D. & Kawaoka, Y. (2004). Production of novel Ebola virus-like particles from cDNAs: an alternative to Ebola virus generation by reverse genetics. J Virol 78, 999–1005.[Abstract/Free Full Text]

Whelan, S. P. & Wertz, G. W. (2002). Transcription and replication initiate at separate sites on the vesicular stomatitis virus genome. Proc Natl Acad Sci U S A 99, 9178–9183.[Abstract/Free Full Text]

Whelan, S. P., Barr, J. N. & Wertz, G. W. (2004). Transcription and replication of nonsegmented negative-strand RNA viruses. Curr Top Microbiol Immunol 283, 61–119.[Medline]

Received 20 April 2007; accepted 13 June 2007.

This Article Abstract Full Text (PDF) Supplementary Material Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Rennick, L. J. Articles by Rima, B. K. Search for Related Content PubMed PubMed Citation Articles by Rennick, L. J. Articles by Rima, B. K. Agricola Articles by Rennick, L. J. Articles by Rima, B. K.