Journal of General Virology (2003), 84, 1613–1621
DOI 10.1099/vir.0.19128-0
Rabies virus matrix protein regulates the balance of
virus transcription and replication
Stefan Finke, Roland Mueller-Waldeck and Karl-Klaus Conzelmann
Max von Pettenkofer Institute and Gene Center, Ludwig Maximilians University Munich, Feodor
Lynen Str. 25, D-81377 Munich, Germany
Correspondence
Karl-Klaus Conzelmann
[email protected]
Received 29 January 2003
Accepted 6 March 2003
RNA synthesis by negative-strand RNA viruses (NSVs) involves transcription of subgenomic
mRNAs and replication of ribonucleoprotein complexes. In this study, the envelope matrix (M)
protein of rabies virus (RV) was identified as a factor which inhibits transcription and stimulates
replication. Transcription, but not replication, of RV minigenomes or of full-length RV was decreased
by expression of heterologous M. Since RV assembly involving M and the glycoprotein G renders
virus synthetically quiescent, an RV was generated with the M and G genes substituted by
placeholders. Surprisingly, RNA synthesis by this recombinant was characterized not only by an
increased transcription rate but also by a diminished accumulation of replication products. This
phenotype was reversed in a dose-dependent manner by providing M in trans, showing that M is a
replication-stimulatory factor. The role of M in determining the balance of replication and
transcription was further exploited by generation of a recombinant RV with attenuated M expression,
which is highly active in transcription. Regulation of RNA synthesis by matrix proteins may represent
a general mechanism of nonsegmented NSVs, which is probably obscured by the silencing
activity of M during virus assembly.
INTRODUCTION
Genome and antigenome RNAs of rhabdoviruses such
as rabies virus (RV) or vesicular stomatitis virus (VSV)
are enwrapped with nucleoprotein N to form a typical
ribonucleoprotein (RNP) complex. Only RNPs can be
used as a template for the viral RNA polymerase, which is
made up of the enzymatically active L protein and
phosphoprotein P. The polymerase is active in two different
modes of RNA synthesis, transcription and replication.
Transcription involves sequential synthesis of free, subgenomic RNAs, including a short leader RNA, and five
mRNAs in the order 39-le-N-P-M-G-L-59. Replication, in
contrast, gives rise to RNPs, i.e. N-encapsidated full-length
RNAs (Abraham & Banerjee, 1976; Ball & White, 1976). In
contrast to transcription, replication requires ongoing N
synthesis, suggesting that replicative elongation is coupled
to encapsidation of the newly generated RNA (Arnheiter
et al., 1985; Patton et al., 1984).
and other cis-active sequences defining cistrons in the
transcription mode. According to this model, the coupling
of RNP assembly and RNA elongation leads to a selfregulatory system for controlling the balance of transcription and replication (Blumberg et al., 1981, 1983).
RNPs entering a cell are thus not able to replicate until
virus transcription and subsequent translation provide a
critical concentration of N and P protein, allowing a switch
or shift to the processive replication mode. According to
a widely accepted model, newly synthesized N–P complexes
bind to nascent leader RNA and, if present in sufficient
amounts, prevent recognition of the leader/N gene junction
Experimental evidence for intracellular N levels switching
the polymerase to replication is not available for any
nonsegmented negative-strand RNA virus. In minigenome
systems in which all RNP proteins are supplied in trans,
alteration of the amounts and ratios of N, P and L proteins
did not change the ratio of transcription and replication
(unpublished results; Fearns et al., 1997; Wertz et al., 1998),
and VSV transcription initiation may be independent of
leader RNA synthesis at the leader–N junction (Whelan &
Wertz, 2002; Chuang & Perrault, 1997). This strongly
argues in favour of a situation where the availability of
N is just a prerequisite for replication and that other
factors determine whether an RNP is a template for
transcription or replication. These may include modifications of P and L proteins (Das et al., 1998; Pattnaik
et al., 1997; Hwang et al., 1999) or involvement of other
viral and cellular factors. Indeed, in the more complex
negative-strand RNA virus (NSV) respiratory syncytial
virus (RSV) the absence of the viral M2-2 gene product
resulted in enhanced transcription and reduced replication
(Bermingham & Collins, 1999).
Published ahead of print on 25 March 2003 as DOI 10.1099/
vir.0.19128-0.
How transcription and replication of the rhabdoviruses
are regulated is unknown. Apart from the RNP proteins N,
0001-9128 G 2003 SGM
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
Printed in Great Britain
1613
S. Finke, R. Mueller-Waldeck and K.-K. Conzelmann
P and L, RV encodes only two more viral proteins, a matrix
protein, M, and the spike glycoprotein. RV M protein is a
multifunctional protein, playing a crucial role in virus
assembly and budding. M is responsible for recruiting RNPs
to the cell membrane, their condensation into tightly coiled
‘skeleton’-like structures and the budding of enveloped
virus particles (Mebatsion et al., 1999). The long-known
inhibitory activity of M proteins of rhabdoviruses and
other NSVs on viral RNA synthesis in vitro (Carroll &
Wagner, 1979) has been attributed to the function of M in
condensing RNPs prior to virus export. Here, we have used
a reverse genetics approach involving defective and nondefective recombinant RV to investigate the effect of M
protein expression on RV RNA synthesis. We present several
lines of evidence that M is a key factor in the regulation of
RV polymerase functions that exerts opposite effects on
transcription and replication, and thereby tips the balance
toward replication.
METHODS
Cells and viruses. Viruses were grown on BHK-21 clone BSR cell
monolayers maintained in Glasgow MEM supplemented with 10 %
newborn calf serum. The BSR T7/5 cell clone (Buchholz et al.,
1999), which was used for virus rescue, was maintained under
selection conditions [1 mg G418 (ml cell culture supernatant)21].
BSR cell clones that express M or M and G after induction with
doxycycline (M-on and MG-on, respectively) were grown under
selection conditions by adding G418 (1 mg ml21) and hygromycin
(1 mg ml21).
Recombinant genomes and minigenomes. The minigenome
cDNA plasmid pSDI-CL(NP) (Finke & Conzelmann, 1999) comprises the 39-terminal genome promoter (GP) sequence followed by
the chloramphenicol acetyltransferase (CAT) reporter gene, the noncoding RV N/P gene border sequence, the firefly luciferase reporter
gene and, finally, the 59-terminal antigenome promoter (AGP)
sequence. A 59-copy-back version of pSDI-CL(NP) was generated by
replacement of the 39-terminal GP sequence with a reversed copy
of the 59 terminal 67 nt of the RV genome containing the AGP
sequence. This resulted in a minigenome that exclusively contains
replication promoters.
The pNPgrL cDNA plasmid was constructed by replacement of the
M and G genes in pSAD L16 (Schnell et al., 1994) with the EGFP (g)
and DSred (r) reporter genes (Clontech) resulting in the genome
organization 39-N-P-EGFP-DSred-L-59 (see Fig. 3).
In pSAD LM, the M gene was translocated from its original position
between the P and the G genes to the most distal position on the
viral genome, i.e. downstream of the L gene. This was achieved by
replacement of the L gene transcription stop signal in the previously
described M-deficient pSAD DM (Mebatsion et al., 1999) with a copy
of the N/P-gene border sequence and subsequent insertion of the
M-coding sequence downstream of the new transcription signal,
resulting in a genome organization of 39-N-P-G-L-M-59. A detailed
description of all cloning steps and the final sequences are available
from the authors by email.
cDNA expression plasmids. Expression plasmids pTIT-N, pTIT-P
and pTIT-L have been described previously (Finke & Conzelmann,
1999). pTIT-M and pTIT-G were generated by insertion of the
coding sequences into pTIT (Buchholz et al., 1999) after PCR amplification from pT7T-M and pT7T-G (Conzelmann & Schnell, 1994).
1614
The resulting plasmids encoded the RV genes under the control of
the bacteriophage T7 RNA polymerase promoter and the internal
ribosome entry site (IRES) of encephalomyocarditis virus.
cDNA rescue experiments. cDNA plasmids were transfected into
cells after calcium phosphate precipitation (mammalian transfection
kit; Stratagene) according to the supplier’s instructions. To enable
T7 RNA polymerase-driven expression, either the transfected cells
were infected with T7 polymerase-expressing vaccinia virus vTF7-3
(Fuerst et al., 1986) or the cDNAs were transfected into BSR T7/5
cells that constitutively express T7 polymerase.
Rescue of recombinant viruses was performed as described previously
(Finke & Conzelmann, 1999) by cotransfection of 16106 BSR T7/5
cells grown in 8 cm2 culture dishes with 10 mg of full-length cDNA and
plasmids pTIT-N (5 mg), pTIT-P (2?5 mg) and pTIT-L (2?5 mg). Fresh
culture medium was added 3 days post-transfection and, after another
3 days, cell culture supernatants were harvested and transferred to BSR
cells. Two days after passage, infectious viruses were detected by
immunostaining for RV N protein.
Minigenome virus-like particles (VLPs) were generated from cDNAs
as described previously (Conzelmann & Schnell, 1994) by coexpression
of minigenome RNA and virus proteins N, P, M, G and L in vaccinia
virus vTF7-3-infected cells. In the complementation experiments,
VLPs were transferred to vaccinia virus vTF7-3-infected BSR cells
that were cotransfected with pT7T-N, pT7T-P, pT7T-L (5?0 mg, 2?5 mg
and 2?5 mg, respectively) and with increasing amounts of pT7T-M
(0?0 to 2?0 mg).
RNA analysis. RNA was isolated from cells with the RNeasy Mini
kit (Qiagen). For nuclease treatment, the cells were trypsinized and
pelleted. Pellets of 16105 cells were lysed in 50 ml buffer (10 mM
NaCl, 10 mM Tris pH 7?5, 1?5 mM MgCl2, 1 % Triton). After
30 min incubation at 30 ˚C with 50 ml micrococcal nuclease [20 mg
nuclease S7 (Boehringer Mannheim) in 10 mM Tris pH 7?5, 1 mM
CaCl2], RNA was isolated. Northern blots and hybridizations with
[a-32P]dCTP-labelled cDNAs were performed as described previously
(Conzelmann et al., 1991). Hybridization signals were quantified by
phosphorimaging (Molecular Dynamics Storm).
Immunoblotting. Western blots were performed as described pre-
viously (Finke et al., 2000). Signals were quantified by fluoroimaging
(ECF; Amersham) using a phosphorimager (Molecular Dynamics
Storm).
RESULTS
Inhibition of minigenome transcription
The interdependency of replication and transcription in
non-deficient NSVs impedes investigation of the individual
processes. Therefore, to examine the effect of RV M on
intracellular RV RNA synthesis, we first used a minigenome
system in which all virus proteins are provided in trans
(Conzelmann & Schnell, 1994). To analyse the products
of transcription, a previously described transcriptionally
active bicistronic minigenome, SDI-CL(NP), was used
(Finke & Conzelmann, 1999). SDI-CL(NP) contains two
reporter genes CAT and luciferase flanked by the authentic
RV 39 and 59 ends (Fig. 1A, grey and black boxes, respectively). Accordingly, transcription of SDI-CL(NP) gives rise
to a leader RNA and two mRNAs. To determine the effect of
M on RV replication, a transcriptionally inactive 59-copyback model genome was generated, by replacing the 39-end
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
Journal of General Virology 84
RV matrix protein regulates virus RNA synthesis
replication products from SDI-CL(NP) was not detectable
under these conditions.
RNA synthesis products of the 59-copy-back genome were
almost exclusively full-length RNAs (Fig. 1B). Resistance
to nuclease digestion further showed that this RNA is
encapsidated in N proteins and represents the product of
replication. In striking contrast to SDI-CL(NP) transcription, increasing amounts of M did not have a detectable
influence on the accumulation of RNPs, even at the
highest M concentration (Fig. 1B).
Overexpression of M suppresses transcription
of full-length RV
Fig. 1. (A) Transcription inhibition by RV M. BSR cells were
infected with SDI-CL(NP). Virus proteins N, P and L were
expressed from transfected cDNAs by vaccinia virus vTF7-3
derived T7 polymerase. Cotransfection with increasing amounts
of pT7T-M resulted in diminished luciferase and CAT mRNA
levels, as determined in Northern hybridizations with cDNA
probes recognizing both CAT and luciferase sequences. (B)
Minigenome replication is not inhibited by increasing doses of
M. Complementation of the copy-back minigenome was done
as detailed in (A). Nuclease resistance confirmed the identity of
signals as RNP-derived minigenome RNAs. Western blots
(bottom) confirmed increasing M protein levels. (C) Signal
intensities were quantified by phosphorimaging and were
normalized with 28S rRNA. The level of transcription products
(open symbols) and replication products (filled symbols) were
plotted as a function of the transfected pT7T-M. n.tr., not transfected but VLP-infected; HV, VLP and helper virus-infected.
of SDI-CL(NP) with a reverse copy of the 59-terminal 67
nt containing the RV antigenome promoter (Fig. 1B, black
boxes). As the antigenome promoter is highly active in
replication but unable to direct transcription, this copy-back
model genome is replicated at high levels. VLPs were
prepared as described (Conzelmann & Schnell, 1994) and
were transferred to BSR cells that were transfected with
fixed amounts of T7 promoter-driven plasmids encoding
RV proteins N, P and L, along with increasing amounts of
pT7T-M, encoding RV M protein.
RNA synthesis from the minigenomes was monitored by
Northern hybridization 2 days post-infection (p.i.). The
probes used in this and in all following hybridizations
recognized both positive-stranded mRNAs and antigenomes
as well as genomes. Substantial amounts of monocistronic
CAT and luciferase mRNAs, as well as bicistronic CAT/
luciferase RNAs transcribed from SDI-CL(NP), were
present in the absence of M (Fig. 1A). With increasing
amounts of pT7T-M, the level of mRNAs decreased in
a dose-dependent manner. At maximal amounts of M
plasmid (2 mg), mRNA synthesis was reduced to ~10 %.
A signal corresponding to genome RNA and representing
http://vir.sgmjournals.org
To determine how RNA synthesis of a non-deficient RV
is affected by M, wt RV SAD L16 was grown in cells,
providing extra M protein. A BSR-derived cell line which
expresses RV M protein after induction with doxycycline
in a dose-dependent manner (M-on; to be published
elsewhere) was infected with SAD L16 and expression of
cell-encoded M (cM) was induced. Total RNA and cell
lysates were analysed after 24 h. cM was detected exclusively
after induction of cells (Fig. 2). Compared to M expression
in virus-infected cells, cM was much less abundantly
expressed. Thus, in induced and virus-infected cells, cM
contributed little to the total M level. Nevertheless, a
significant drop in RV transcription was observed in
induced cells. In spite of clearly reduced mRNA levels, a
decrease in the accumulation of full-length replication
products was not observed (Fig. 2, genome). Since in this
system, in contrast to the minigenome system, replication
depends on the products of RV N mRNAs, the surprisingly
high replication level suggests a positive influence of the
additional cM on RV replication. Comparison of the levels
of mRNAs and genome RNAs by phosphorimaging revealed
that transcription rates (mRNA/genome) were reduced to
53 % for the N mRNA and to 35 % for the L mRNA after
expression of cM. Thus, even low levels of additional M
protein exert a distinct regulatory activity on virus RNA
synthesis in wt RV infection, resulting in a shift towards
replication.
Deregulation of RV transcription and replication
in the absence of M protein
To further dissect the transcription inhibitory effect of
RV M protein, viral RNA production was examined in the
absence of M. We have previously described an M genedeficient RV, SAD-DM, which has been used to study
aspects of M in virus assembly (Mebatsion et al., 1999).
This construct, however, was initially not considered
appropriate for the present study, as it lacks an entire
cistron. We therefore constructed a novel recombinant
genome in which the M gene was replaced with the EGFP
reporter gene. In addition, the G gene was replaced with
the DSred reporter gene, in order to obviate potentially
disturbing interactions of M and G. The cDNA construct
pNPgrL (Fig. 3A) produced functional RNPs in cells
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
1615
S. Finke, R. Mueller-Waldeck and K.-K. Conzelmann
Fig. 2. Overexpression of M in wt virus-infected cells. M-on
cells were infected with wt RV SAD L16 at an m.o.i. of 1. Two
h p.i., M expression was induced by adding doxycycline
(10 mg ml21). After 24 h, RNA was analysed in Northern hybridizations with N- and L- specific cDNA probes, recognizing
both negative and positive strand RNAs. M expression was
monitored as a control in Western blots (bottom). Additional
expression of M resulted in transcription inhibition, whereas
replication stayed high, thus decreasing the transcription rates
(mRNA/genome) to 53 % and 35 % (N and L hybridizations,
respectively) in noninducing conditions.
expressing RV N, P and L proteins from transfected
plasmids as described previously (Finke & Conzelmann,
1999). To complement for the M and G deficiency, cells
were also transfected with M- and G-encoding plasmids
pTIT-M and pTIT-G and were mixed with a BSR cell line
expressing M and G proteins after induction with
doxycycline (MG-on). To select for M- and G-expressing
cells, hygromycin was added to the mixture at 1 mg ml21.
Expression was repeatedly induced until approximately
40 % of the cell culture was infected with NPgrL, as
determined by EGFP fluorescence. Cell culture supernatants
were transferred to fresh MG-on cells for NPgrL stock
production.
RNA synthesis of NPgrL was investigated by Northern
blot experiments after infection of BSR cells at an m.o.i. of
5. In the absence of transcription inhibition, a greater
accumulation of mRNAs, and, as a result of higher virus
gene expression, a higher accumulation of genome RNAs
were expected. However, while the mRNA levels were
comparable to those of wt RV infection (Fig. 3B, see N
mRNA 2 days p.i.), the levels of genome length RNA were
1616
Fig. 3. RNA synthesis in the absence of M. (A) Genome organization of wt RV SAD L16 and NPgrL. In NPgrL, the genes
for the envelope proteins M and G were replaced with the
EGFP and DSred genes. (B) BSR cells were infected with
SAD L16 or with NPgrL at an m.o.i. of 5. One and two days
p.i. RNA was isolated from the cells and was analysed with N
and L cDNA probes. Western blots (bottom) show comparable
N and P protein expression. No M protein was detectable in
NPgrL infected cells. (C) Hybridization signals were quantified
by phosphorimaging and transcription rates are given as mRNA
per genome (SAD L16 at day 1=100 %). Black and white
bars represent N mRNA and L mRNA transcription rates,
respectively.
strongly reduced. Quantification of hybridization signals
revealed an increased transcription rate. Normalized to
SAD L16, the transcription rates of NPgrL were 379 % for
the N mRNA and 501 % for the L mRNA at 1 day p.i. At 2
days p.i., the transcription rate of SAD L16 decreased to
42 %, whereas that of NPgrL was still 181 % and 206 % for
N mRNA and L mRNA, respectively. In spite of this 4?5fold increase in transcription rate, which should result in
rapid accumulation of N, P and L protein available for
replication, total RNA synthesis of NPgrL was always lower
as compared to wt RV SAD L16, suggesting that N, P and
L proteins were not optimally used for replication. The
observation that the complete lack of M is accompanied
by a diminished replicase activity again strongly argues in
favour of a distinct function of M in enhancing replication.
Dose-dependent effect of M on RNA synthesis
To analyse NPgrL RNA synthesis in the presence of
increasing levels of RV M protein, M-on cells were infected
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
Journal of General Virology 84
RV matrix protein regulates virus RNA synthesis
with NPgrL at an m.o.i. of 1. Two days later, doxycycline
was added to aliquots of the infected cells at concentrations
of 0, 0?1 or 1?0 mg (ml cell culture medium)21. Two days
after induction of M expression, viral RNA was isolated
and analysed in Northern blots with RV N- and L-genespecific DNA probes (Fig. 4A). Upon induction with 0?1
and 1?0 mg ml21 doxycycline, the transcription rates of the
N mRNA decreased to 52 % and 22 %, respectively (not
shown). Similarly, a decrease of the L transcription rate was
observed. Since the level of M expression in M-on cells
directly correlates with the doxycycline concentration (not
shown), the decrease in the transcription rates depends on
the level of M protein available.
Normalization with 28S rRNA hybridization signals
revealed that with increasing doxycycline concentrations
the absolute levels of N mRNA decreased to 50 % compared
to noninducing conditions (Fig. 4B). In contrast, absolute
levels of genome-length RNA were 2?3- and 2?2-fold (N
and L hybridization, respectively) increased at a doxycycline concentration of 1?0 mg ml21, demonstrating that M
stimulates RNA replication in NPgrL infected cells. To
exclude effects of doxycycline on virus RNA synthesis,
transcription rates of NPgrL were also determined in
parental Tet-on cells, that only express the reversed
Tet-repressor. In contrast to M-on cells, no reduction of
transcription rates was detectable with increasing doxycycline concentration (Fig. 4C).
RVs with attenuated M expression
According to the above results, RV M has two different
functions in the regulation of RV RNA synthesis, namely a
pronounced transcription inhibition activity, accompanied
by a stimulatory activity on replication. The transient assays
with cM already suggested that small amounts of M were
sufficient to support replication (Fig. 4), stressing the idea
that RV with attenuated M expression might perform better
in total RNA synthesis than RV lacking M. To follow this, a
recombinant virus cDNA was constructed, in which the M
gene was shifted to the very 59-end, i.e. downstream of the
L gene. Due to the steep step in the RV transcription
gradient at the G/L gene border (Finke et al., 2000), M gene
transcription of this virus should be severely attenuated.
Because of the essential function of M during virus assembly
(Mebatsion et al., 1999), it was questionable whether viable
virus could be recovered. However, autonomously propagating virus, dubbed SAD LM and yielding infectious titres
similar to SAD L16, was obtained from cDNA.
BSR cells were infected in parallel with SAD LM and with
SAD L16 at an m.o.i. of 1 and the expression of M and N
proteins was monitored in Western blots. Signal intensities
were determined by fluoroimaging (Fig. 5). Whereas N
protein levels in SAD LM-infected cells were comparable to
wt virus, the levels of M protein were always lower than wt
M levels, indicating that the gene translocation resulted in
attenuated M protein expression. In numbers, the ratio of
M to N was decreased 3?2-, 1?4- and 2?0-fold at days 1, 2
and 3 p.i., respectively.
Fig. 4. Complementation of NPgrL with M. M-on cells were
infected with NPgrL at an m.o.i. of 1. M expression was
induced two days p.i. with the indicated concentrations of doxycycline. (A) Northern hybridization with RNAs isolated two days
after induction with N and L specific cDNA probes. (B) With
increasing doxycycline concentrations the total amount of N
mRNA decreased up to 2-fold, whereas the genome RNA
levels increased up to more than 2-fold. Signal intensities were
normalized with the 28S rRNA signal. (C). To exclude an effect
of doxycycline on RV transcription and replication, M-on cells
and the parental cell line BSR Tet-on were infected with NPgrL
and incubated with doxycycline as described above. Whereas
transcription rates of NPgrL decreased with increasing M
expression to 41 % and 31 %, no reduction of transcription
rates was obtained on Tet-on cells.
http://vir.sgmjournals.org
RNA from the same experiment was visualized in Northern
hybridizations with an N probe (Fig. 6). Compared to SAD
L16, N mRNA in SAD LM-infected cells was much more
abundant, whereas the genome RNA levels were lower. The
transcription rates of SAD LM were 11-, 33- and 23-fold
higher than that of SAD L16 at days 1, 2 and 3, respectively.
In addition, and in contrast to the situation with NPgrL in
the absence of M (see Fig. 3), total RNA production of SAD
LM exceeded that of SAD L16 2-, 5- and 4-fold after 1, 2 and
3 days, respectively. This reflects better accumulation of
genome-length RNA by the available M protein. Thus,
although sufficient M was produced to promote effective
virus particle assembly, the diminished M content tipped the
balance of RNA synthesis to ratios that were previously
observed only for M deletion mutants.
DISCUSSION
Matrix proteins are major components of the envelope of
NSVs and are known to shut down transcription from
RNPs in vitro, as shown for different NSVs such as influenza
virus, measles virus or VSV (Suryanarayana et al., 1994;
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
1617
S. Finke, R. Mueller-Waldeck and K.-K. Conzelmann
Fig. 5. Attenuation of M gene expression in recombinant RV
SAD LM. Cells were infected with SAD L16 or SAD LM at an
m.o.i. of 1 and protein extracts were prepared at indicated time
points. RV N and M proteins were quantified after Western
blotting by fluoroimaging. The signals obtained were normalized
with a signal for cellular calnexin. In the bottom graph, the relative M level is given as the ratio of M to N signal intensities.
White bars represent two independent wt virus infections and
grey bars two independent SAD LM infections.
Watanabe et al., 1996; Carroll & Wagner, 1979). The
function of M in binding and condensing RNPs during
virus assembly and egress provides a plausible explanation
for this kind of transcription inhibition. Tightly coiled,
‘freezed’ RNP structures are not assumed to represent a
proper template for RNA synthesis. Due to the lack of
effective in vitro replication systems for NSVs, data on the
effect of M on replication in vitro are not available. Yet,
condensation of RNPs should similarly halt transcription
and replication in the assembling virus. The present data
on intracellular RNA synthesis of recombinant RV constructs reveal that M has another function prior to ‘freezing’
RNPs, namely as a regulatory protein in tipping the balance
of virus RNA synthesis towards replication. Most importantly, transcription inhibition activity is accompanied by
an opposite stimulatory activity on replication.
The first hint on differential activity of M on transcription
and replication of RV was obtained in experiments using
reporter minigenome RNAs. All virus proteins were provided in trans, ensuring that at least replication occurs
independently of virus transcription. A dose-dependent
decrease in mRNA production from SDI-CL(NP) was
already observed after expression of very little amounts of
M (Fig. 1). The linear decrease of transcription argues
against inhibition by RNP condensation, which would be
expected to need a threshold level of available M protein.
1618
Fig. 6. Deregulation of RNA synthesis of SAD LM. From the
infection experiment in Fig. 5, RNA was isolated and analysed
in Northern hybridizations. The N hybridization revealed
increased mRNA levels in SAD LM infected cells, whereas
genome RNAs were clearly diminished. Transcription rates are
given as mRNA per genome (SAD L16 at day 1=100 %).
Black and grey bars represent SAD L16 and SAD LM transcription rates, respectively.
The low replication activity of SDI-CL(NP), which is also
observable in wt helper virus-infected cells (Finke &
Conzelmann, 1999), prompted us to use the 59-copy-back
minigenome, which allows reliable quantification of replication products. In striking contrast to transcription, even
high levels of M did not abrogate replication of 59 copy-back
RNPs. As condensation of RNPs is expected to shut down
not only transcription but also replication, this result also
favours pre-condensation effects of M.
Effects of additional M were monitored in experiments in
which non-deficient RV was grown in the presence of
additional M expressed from the cell (cM). Relative to RVexpressed M, cM expression was low and increased the
amount of available M only a little. Nevertheless, a marked
inhibition of RV transcription was noted (see Fig. 2), which
is most probably due to higher cM than viral M expression
in the early stages of infection. The more important
finding was that the reduced levels of mRNA observed did
not result in a similar decrease of genome RNA levels. The
latter was rather expected since reducing viral N mRNA
expression below a critical limit might in turn limit
N-dependent replication. An explanation for this not
taking place is that levels of N in wt virus infection are
abundant and far from limiting replication. Since N is not
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
Journal of General Virology 84
RV matrix protein regulates virus RNA synthesis
the limiting factor, other factors should account for keeping
replication in wt RV infections low. The most probable
explanation is a stimulatory effect of M on virus RNA
replication.
A role of M in stimulating replication was further strongly
supported by analysis of RV RNA synthesis in the absence
of M. With respect to the resistance of RV replication to
M in the minigenome system, we expected that replication
of the M- and G-defective virus NPgrL would be at
least comparable to wt RV, or, rather, augmented as a
consequence of enhanced virus gene expression. This was
not the case, as the levels of genome RNA were clearly
decreased in the absence of M. Though sufficient levels of
N and other virus proteins were obviously available, less
full-length replication products accumulated. The lack of
M totally abolishes RV RNP condensation and decreases by
magnitudes the export of RNPs from infected cells through
budding of virus particles (Mebatsion et al., 1999). Thus,
considering the lack of RNP export, the absence of M in
NPgrL must have an even greater effect on replication
than that pretended by the values obtained from phosphorimaging of Northern blots. Therefore, in spite of roughly
comparable N levels (see Fig. 3B), the lack of M is
hampering the use of this available N for replication.
Thus, M in wt virus is not only exerting a selective inhibition on virus transcription but also has a clear stimulatory
effect on replication.
This was further underlined by complementation of the
M- and G-defective virus NPgrL with different amounts of
M. Again, in the presence of M, the opposite effects on
transcription and replication products were obvious.
Importantly, with expression of cM the RNPs gain the
ability to bud off the infected cell as spikeless particles
(Mebatsion et al., 1996). Although in the absence of G
the budding efficiency is approximately 30-fold lower than
in wt RV, a certain efflux of RNPs has to be taken into
account. Nevertheless, increasing amounts of genome RNAs
were observed in cells expressing increasing amounts of
M (see Fig. 4).
As suggested by complementation of defective RV in trans
and as confirmed by the non-deficient virus SAD LM, RV
possesses an M-directed fine-tuning mechanism of RNA
synthesis. In SAD LM-infected cells, M protein expression
was not much less than with wt RV, but the relative ratios
of N and M were changed. In view of this, SAD LM RNA
synthesis showed characteristic and predicted features.
SAD LM grew well, with a short lag at early time points
of infection (not shown), indicating an initial handicap in
replication. This was obviously compensated by a highly
enhanced virus gene expression, resulting in a total RNA
synthesis exceeding that of wt RV. Whereas in wt RV gene
expression is downregulated in favour of genome replication, in SAD LM unrestricted transcription was coupled
with reduced replication activity. The shift of the L gene
to an upstream position probably leads to overexpression
of the polymerase in SAD LM. However, this affects only
http://vir.sgmjournals.org
the overall RNA synthesis but not the ratio of replication
and transcription, as shown previously for other RVs
overexpressing L (Finke et al., 2000). The high transcription phenotype of SAD LM was therefore exclusively due
to the attenuated M gene expression.
For regulatory functions of M in RNA synthesis, two
different targets of M are conceivable: the L polymerase
and the RNP template. Recent models suggest that the
polymerase might exist as a nonprocessive transcriptase
and as processive replicase. In contrast to replication,
Sendai virus transcription requires special reaction conditions to promote processivity on the RNP, indicating that
the processivity of transcriptase and replicase is differentially
regulated (Gubbay et al., 2001). Moreover, recombinant
ambisense RVs indicated that individual replicase and
transcriptase forms of the polymerase complex exist (Finke
& Conzelmann, 1997).
Association of VSV L with different host cell or virus factors
might affect L polymerase activity (Barik & Banerjee, 1992;
Gupta & Banerjee, 1997; Das et al., 1998). Indeed, VSV P
phosphorylation mutants able to support either replication
or transcription of defective interfering particles (Pattnaik
et al., 1997; Hwang et al., 1999) and P antibodies inhibiting
replication but not transcription have been described
(Richardson & Peluso, 1996). The RV P protein is phosphorylated by two kinases (Gupta et al., 2000); however, no
effect of P phosphorylation mutants on RV RNA synthesis
has been described so far. Moreover, no differential activity
on replication and transcription could be demonstrated
with RV N protein phosphorylation mutants (Yang et al.,
1999). As temporal changes in the modification of P, L or
N protein during rhabdovirus infection have not been
shown, their relevance with respect to a physiological shift
between replication and transcription remains unclear.
With the binding of M to the polymerase, either the
formation of replicase could be favoured, or the formation
of transcriptase could be prevented, both provoking effects
on RNA synthesis as described above. In addition, M
might act on cellular factors, which in turn influence the
polymerase complex composition or activity. For instance,
borna disease virus polymerase is differentially influenced
by interferon (Staeheli et al., 2001). For the present regulation of RV RNA synthesis, however, a role for interferon
can be ruled out since all cells used in this study were
interferon incompetent (Schlender et al., 2000).
Another possible mechanism for regulation by M could
involve modification of the RNP such that it is better
suited as a template for replication than for transcription.
Indeed, RV M is able to alter the structure of the RNP by
condensing it into tight coils and keeping it in these
‘skeleton’-like forms (Mebatsion et al., 1999). So, at least
for assembly, M causes a structural transition of RNPs to,
most probably, synthetically inert bodies. To explain the
regulatory function of M exerting opposite effects on
transcription and replication, a model involving an action
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
1619
S. Finke, R. Mueller-Waldeck and K.-K. Conzelmann
of M on intracellular, decoiled RNPs must be favoured.
Indeed, for VSV, M binding to intracellular nucleocapsids
prior to virion assembly has been shown (Flood & Lyles,
1999). Interestingly, the affinity of cytosolic M in binding
intracellular RNPs was approximately 8-fold less than in
binding to virion-derived RNPs. Moreover, membranederived M exhibited little or no binding to intracellular
RNPs. These findings are not only in favour of different
activities of M in virus assembly and in virus RNA synthesis,
but also suggest that the M regulatory activity acts on the RNP.
Binding of low amounts of M to intracellular RNPs could
affect the function of transcriptase and replicase differently.
The higher processivity of the replicase, which is also able
to read through transcription stop signals, could override
bound M proteins, whereas a more sensitive, non-processive
transcriptase might be blocked by the bound M. Whereas
selective inhibition of transcription could be easily explained
this way, stimulation of replication cannot. Assuming
that transcriptase and replicase exist in an equilibrium,
M-modified RNPs might change this equilibrium by
increasing the concentration of free transcriptase, thereby
increasing intracellular replicase levels. This could also
explain why in our nontranscribing minigenome system
a significant stimulation of replication by M was not
observed, in contrast to all transcribing systems. Another
possibility might be differential action on genome and
antigenome promoters.
The availability of reverse genetics systems for defective and
non-defective RV allowed us to dissociate functions of M
related to virus assembly and regulation of RNA synthesis.
These systems certainly provide tools to further address the
mechanism of M action in regulating RV RNA synthesis. A
promising approach that will be substantially facilitated
by the recently solved crystal structure of the VSV M
protein (Gaudier et al., 2002) is the genetic analysis of M
proteins to identify relevant domains. It also appears
worthwhile to extend these investigations to other NSVs.
In spite of ostensible differences in M structure and the
location of M proteins in VSV and RV virions (Barge et al.,
1993; Mebatsion et al., 1999), it is likely that regulation
mechanisms in rhabdoviruses are similar. Notably, in the
late seventies, a regulatory role of VSV M was suggested
after the analysis of RNA synthesis of temperature sensitive
VSV mutants (Martinet et al., 1979; Clinton et al., 1978). It
will be interesting to also include paramyxoviruses, which
encode nonessential, accessory gene products able to
modulate RNA synthesis, such as the RSV M2-2
(Bermingham & Collins, 1999), or the Sendai virus C
proteins (Curran et al., 1992), in order to determine whether
regulation of RNA synthesis by M is a general feature of
nonsegmented NSVs.
ACKNOWLEDGEMENTS
We thank N. Hagendorf for her perfect technical assistance and B.
Bossert for critical comments on the manuscript. This work was
supported by the Deutsche Forschungsgemeinschaft (SFB 455-A3).
1620
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.
Arnheiter, H., Davis, N. L., Wertz, G., Schubert, M. & Lazzarini, R. A.
(1985). Role of the nucleocapsid protein in regulating vesicular
stomatitis virus RNA synthesis. Cell 41, 259–267.
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.
Barge, A., Gaudin, Y., Coulon, P. & Ruigrok, R. W. (1993). Vesicular
stomatitis virus M protein may be inside the ribonucleocapsid coil.
J Virol 67, 7246–7253.
Barik, S. & Banerjee, A. K. (1992). Phosphorylation by cellular casein
kinase II is essential for transcriptional activity of vesicular stomatitis
virus phosphoprotein P. Proc Natl Acad Sci U S A 89, 6570–6574.
Bermingham, A. & Collins, P. L. (1999). The M2–2 protein of human
respiratory syncytial virus is a regulatory factor involved in the
balance between RNA replication and transcription. Proc Natl Acad
Sci U S A 96, 11259–11264.
Blumberg, B. M., Leppert, M. & Kolakofsky, D. (1981). Interaction of
VSV leader RNA and nucleocapsid protein may control VSV genome
replication. Cell 23, 837–845.
Blumberg, B. M., Giorgi, C. & Kolakofsky, D. (1983). N protein of
vesicular stomatitis virus selectively encapsidates leader RNA in vitro.
Cell 32, 559–567.
Buchholz, U. J., Finke, S. & Conzelmann, K. K. (1999). Generation of
bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is
not essential for virus replication in tissue culture, and the human
RSV leader region acts as a functional BRSV genome promoter.
J Virol 73, 251–259.
Carroll, A. R. & Wagner, R. R. (1979). Role of the membrane (M)
protein in endogenous inhibition of in vitro transcription by
vesicular stomatitis virus. J Virol 29, 134–142.
Chuang, J. L. & Perrault, J. (1997). Initiation of vesicular stomatitis
virus mutant polR1 transcription internally at the N gene in vitro.
J Virol 71, 1466–1475.
Clinton, G. M., Little, S. P., Hagen, F. S. & Huang, A. S. (1978). The
matrix (M) protein of vesicular stomatitis virus regulates transcription. Cell 15, 1455–1462.
Conzelmann, K. K. & Schnell, M. (1994). Rescue of synthetic
genomic RNA analogs of rabies virus by plasmid-encoded proteins.
J Virol 68, 713–719.
Conzelmann, K. K., Cox, J. H. & Thiel, H. J. (1991). An L
(polymerase)-deficient rabies virus defective interfering particle RNA
is replicated and transcribed by heterologous helper virus L proteins.
Virology 184, 655–663.
Curran, J., Marq, J. B. & Kolakofsky, D. (1992). The Sendai virus
nonstructural C proteins specifically inhibit viral mRNA synthesis.
Virology 189, 647–656.
Das, T., Mathur, M., Gupta, A. K., Janssen, G. M. & Banerjee, A. K.
(1998). RNA polymerase of vesicular stomatitis virus specifically
associates with translation elongation factor-1 abc for its activity.
Proc Natl Acad Sci U S A 95, 1449–1454.
Fearns, R., Peeples, M. E. & Collins, P. L. (1997). Increased
expression of the N protein of respiratory syncytial virus stimulates
minigenome replication but does not alter the balance between the
synthesis of mRNA and antigenome. Virology 236, 188–201.
Finke, S. & Conzelmann, K. K. (1997). Ambisense gene expression
from recombinant rabies virus: random packaging of positive- and
negative-strand ribonucleoprotein complexes into rabies virions.
J Virol 71, 7281–7288.
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
Journal of General Virology 84
RV matrix protein regulates virus RNA synthesis
Finke, S. & Conzelmann, K. K. (1999). Virus promoters determine
interference by defective RNAs: selective amplification of mini-RNA
vectors and rescue from cDNA by a 39 copy-back ambisense rabies
virus. J Virol 73, 3818–3825.
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.
Flood, E. A. & Lyles, D. S. (1999). Assembly of nucleocapsids with
cytosolic and membrane-derived matrix proteins of vesicular
stomatitis virus. Virology 261, 295–308.
Fuerst, T. R., Niles, E. G., Studier, F. W. & Moss, B. (1986).
Eukaryotic transient-expression system based on recombinant
vaccinia virus that synthesizes bacteriophage T7 RNA polymerase.
Proc Natl Acad Sci U S A 83, 8122–8126.
Gaudier, M., Gaudin, Y. & Knossow, M. (2002). Crystal structure of
vesicular stomatitis virus matrix protein. EMBO J 21, 2886–2892.
Gubbay, O., Curran, J. & Kolakofsky, D. (2001). Sendai virus genome
synthesis and assembly are coupled: a possible mechanism to promote viral RNA polymerase processivity. J Gen Virol 82, 2895–2903.
Gupta, A. K. & Banerjee, A. K. (1997). Expression and purification of
vesicular stomatitis virus N-P complex from Escherichia coli: role
in genome RNA transcription and replication in vitro. J Virol 71,
4264–4271.
Gupta, A. K., Blondel, D., Choudhary, S. & Banerjee, A. K. (2000).
The phosphoprotein of rabies virus is phosphorylated by a unique
cellular protein kinase and specific isomers of protein kinase C.
J Virol 74, 91–98.
Hwang, L. N., Englund, N., Das, T., Banerjee, A. K. & Pattnaik, A. K.
(1999). Optimal replication activity of vesicular stomatitis virus RNA
polymerase requires phosphorylation of a residue(s) at carboxyterminal domain II of its accessory subunit, phosphoprotein P. J Virol
73, 5613–5620.
Martinet, C., Combard, A., Printz-Ane, C. & Printz, P. (1979).
Envelope proteins and replication of vesicular stomatitis virus: in
vivo effects of RNA+ temperature-sensitive mutations on viral RNA
synthesis. J Virol 29, 123–133.
Mebatsion, T., Konig, M. & Conzelmann, K. K. (1996). Budding of
rabies virus particles in the absence of the spike glycoprotein. Cell 84,
941–951.
Mebatsion, T., Weiland, F. & Conzelmann, K. K. (1999). Matrix
protein of rabies virus is responsible for the assembly and budding of
http://vir.sgmjournals.org
bullet-shaped particles and interacts with the transmembrane spike
glycoprotein G. J Virol 73, 242–250.
Pattnaik, A. K., Hwang, L., Li, T., Englund, N., Mathur, M., Das, T. &
Banerjee, A. K. (1997). Phosphorylation within the amino-terminal
acidic domain I of the phosphoprotein of vesicular stomatitis virus
is required for transcription but not for replication. J Virol 71,
8167–8175.
Patton, J. T., Davis, N. L. & Wertz, G. W. (1984). N protein alone
satisfies the requirement for protein synthesis during RNA
replication of vesicular stomatitis virus. J Virol 49, 303–309.
Richardson, J. C. & Peluso, R. W. (1996). Inhibition of VSV genome
RNA replication but not transcription by monoclonal antibodies
specific for the viral P protein. Virology 216, 26–34.
Schlender, J., Bossert, B., Buchholz, U. & Conzelmann, K. K.
(2000). Bovine respiratory syncytial virus nonstructural proteins NS1
and NS2 cooperatively antagonize a/b interferon-induced antiviral
response. J Virol 74, 8234–8242.
Schnell, M. J., Mebatsion, T. & Conzelmann, K. K. (1994). Infectious
rabies viruses from cloned cDNA. EMBO J 13, 4195–4203.
Staeheli, P., Sentandreu, M., Pagenstecher, A. & Hausmann, J.
(2001). a/b interferon promotes transcription and inhibits replica-
tion of borna disease virus in persistently infected cells. J Virol 75,
8216–8223.
Suryanarayana, K., Baczko, K., ter, Meulen., V. & Wagner, R. R.
(1994). Transcription inhibition and other properties of matrix
proteins expressed by M genes cloned from measles viruses and
diseased human brain tissue. J Virol 68, 1532–1543.
Watanabe, K., Handa, H., Mizumoto, K. & Nagata, K. (1996).
Mechanism for inhibition of influenza virus RNA polymerase
activity by matrix protein. J Virol 70, 241–247.
Wertz, G. W., Perepelitsa, V. P. & Ball, L. A. (1998). Gene
rearrangement attenuates expression and lethality of a nonsegmented negative-strand RNA virus. Proc Natl Acad Sci U S A 95,
3501–3506.
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.
Yang, J., Koprowski, H., Dietzschold, B. & Fu, Z. F. (1999).
Phosphorylation of rabies virus nucleoprotein regulates viral RNA
transcription and replication by modulating leader RNA encapsidation. J Virol 73, 1661–1664.
Downloaded from www.microbiologyresearch.org by
IP: 130.203.136.75
On: Sun, 18 Sep 2016 09:45:41
1621